Important Questions About “Most Likely” MH370 Crash Site

Drift model for a crash site at 35S latitude. Black lines are paths for debris and arrow heads are positions on December 31, 2015. (Click on image to enlarge.)

recent report from Australia’s Commonwealth Scientific and Industrial Research Organisation (CSIRO) was greeted with much fanfare. A previous report published in December 2016 predicted that MH370 would be found along the 7th arc at 35S latitude. With new results in, the ATSB proclaimed that they were even more confident in their findings.

Using advanced computer models of how debris from the crash might drift across the Indian Ocean towards Africa, and comparing those results to the location and timing of debris discovered along the shores of Eastern Africa, it was possible to narrow the location to 25,000 sq km of unsearched sea bed. However, the results from this model, presented in December 2016, did not predict that one of MH370’s flaperon would arrive on the French island of La Reunion by the discovery date of July 2015. This perplexed the researchers at CSIRO, who were committed to better understand the discrepancy.

Researchers knew that because the recovered flaperon floated with a portion above the water, the drift path was more heavily influence by wind and waves than debris that floated flat on the surface. The computer model for the flaperon was therefore adjusted for the extra “leeway” by measuring the drift behavior of replica flaperons that were constructed and tested by CSIRO. However, even when the computer model for the flaperon was corrected for this extra leeway, the drift models still did not predict that the flaperon would arrive in La Reunion by July 2015. That was until the American National Transportation Safety Board (NTSB) was able to locate a spare flaperon, which was modified by Australian investigators to resemble the flaperon debris that was recovered. The drift behavior of the modified flaperon was then experimentally measured just as the behavior of the replicas was previously measured. The experimental results said that the computer model should include even more leeway, and the wind would also tend to push the flaperon about 20 degrees to the left. When these effects were included in the computer model for the flaperon, the drift models now predicted an arrival date in La Reunion that was consistent with the discovery date of July 2015, and this was released in a new report. CSIRO felt as though the last missing piece of the puzzle was found.

As reported recently in the media, these results gave CSIRO even higher confidence that its drift models are correct, and MH370 would be found along the 7th arc at 35S latitude. Surprisingly, nowhere in those stories was it reported that at this latitude, the seabed was already searched to a distance of about 20 NM from the arc without finding the plane.

Fellow IG member and co-collaborator Richard Godfrey viewed CSIRO’s drift model with skepticism. He had already performed his own drift study using a computer model he had independently developed, which I have previously discussed and published on this blog, and he came to a much different conclusion. He argued that the timing and location of the debris recovered along the shores of Africa were not consistent with a crash at 35S latitude. Rather, a crash at 30S latitude, well north of the seabed search, was much more likely.

Rather than post Richard’s short comments on this blog, I asked him to prepare a more in-depth critique of CSIRO’s work, as certainly his comments would raise many questions. Within 24 hours, a critique arrived in my inbox, which I publish in full here. In addition to the simulated trajectories of debris, Richard also includes the effect of temperature history on barnacle growth, and comments on the effect of storms in the region. He concludes: Despite the significant contribution in refining the accuracy of the drift model, the new data is interpreted as confirming the findings of the ATSB First Principles Review. The pre-conceived idea, that “other evidence” constrains the MH370 End Point to between 32°S and 36°S is a false assumption. A MH370 End Point at 35°S does not fit the fact that the underwater search has already discounted this location to a 97% level of certainty. An MH370 End Point at around 30°S does fit the available data.

While Richard was preparing his critique, I tried to independently reconcile the differing conclusions of CSIRO and Richard. Basically, CSIRO was predicting a crash site at 35S, and Richard maintained that a search at 30S had a much higher probability of success. Fortunately, the results of CSIRO’s drift studies were made available as KMZ files that could be imported into Google Earth. Using these files, as well as the recent report and the report from December 2016, I was able to piece together some information.

The arrival of debris on the shores of Eastern Africa is highly dependent on the latitude of the crash site. In general, crash sites further to the north along the 7th arc will produce debris that arrives earlier in Africa. After traveling west across the Indian Ocean towards Africa, the debris then tends to travel south. Therefore, debris reaching Eastern Africa would beach last on the shores of South Africa.

The figure at the top of the article shows the position of debris on December 31, 2015, as predicted by the CSIRO model for a crash site along the 7th arc at 35S latitude. The debris is assumed to have “low windage”, which is consistent with the shape of the engine cowling and flap fairing that were discovered in Mozambique and South Africa in December 2015. As can be seen in the figure, CSIRO’s model does not predict that a crash site at 35S latitude would produce debris that would beach as far south along the shores of Eastern Africa as the actual debris that was found. The results of CSIRO’s model are in this respect consistent with the findings of Richard Godfrey.

So why does CSIRO maintain that a crash site at 35S produces debris of the correct location and timing as what was found? The answer lies in a panel from Figure 3.2.1 of the report from December 2016, which is shown below. The vertical axis represents the latitude of potential crash sites along the 7th arc, and the horizontal axis represents the time of arrival along the shores of Eastern Africa, and the color represents the associated probability, with dark blue the lowest probability and red the highest. The red and white bar shown in the figure is aligned along December 2015, which is when the first debris in Eastern Africa was found. And indeed, the colors in the figure do show that for a crash site of 35S latitude, the debris will start to reach Eastern Africa around December 2015.

Probability of reaching Eastern Africa for various crash site latitudes. (From CSIRO, Dec 2016.)

What is not shown in the figure is the timing of when debris will reach various locations along the shores of Eastern Africa. Instead, as can be seen in the title above the figure, all locations along the shores of Eastern Africa from 35S latitude to the Equator are grouped together. But we know that debris will reach locations further south along the shores of Eastern Africa last. In fact, CSIRO’s own model predicts by the end of December 2015, the “non-flaperon” debris, i.e., debris with low windage that floats relatively flat on the water, will reach the shores of Eastern Africa only between 1S and 12S latitudes. On the other hand, the debris was found in Mozambique at 24S latitude and in South Africa at 34S latitude, which is well outside of the range of latitudes predicted by CSIRO’s model.

CSIRO might argue that although a crash site of 35S doesn’t allow debris to reach South Africa by December 2015, a crash site of 30S, as suggested by Richard Godfrey, would have produced debris along the northern shores of Eastern Africa well before Blaine Gibson found the portion of the horizontal stabilizer (nicknamed “No Step”) in March of 2016.  In fact, debris might have arrived well before Mr Gibson’s discovery, and either was not found, was beached and later was again carried out to sea, was caught in offshore eddies, or was found and not reported. In these cases, a distinction should be made between the date of discovery and the date of arrival. Obviously, the arrival must always precede the discovery.

Based on the results of the drift models of both CSIRO and Richard Godfrey, recent claims about the most likely crash site of MH370 should be carefully reviewed by independent investigators.

Update 1 on April 24, 2017.

For those wishing to explore the drift model results in Google Earth, the KMZ files generated by CSIRO are available for the flaperon debris, for non-flaperon low windage debris, and non-flaperon high windage debris. The particular file I used to create the image at the top of the article is the file for low windage, non-flaperon items starting at 35S latitude. Once dragged into Google Earth, simply move the time slider and observe how the particles travel in time.

Update 2 on April 25, 2017.

The Guardian has published an article that discusses our interpretation of CSIRO’s results. “Both CSIRO and the ATSB have been contacted by Guardian Australia for their response.” No response has yet been received by reporter Elle Hunt.

Update 3 on April 26, 2017.

I received the following email from David Griffin of CSIRO:

Dear Victor,

I saw the Guardian article referring to your blog. A few comments:

  1. You are correct that ‘Roy’ was found at an earlier date than the model predicted. But to be fair, the model error is ‘just’ 2 months. I consider Roy’s arrival time – before anything else upstream –  to be something that is simply too hard for any present-day model to convincingly explain. You’ve seen the paths that things take. But those paths should not be interpreted too literally. Our Dec 2016 report mentions that we do not have confidence in the model’s ability to hindcast the arrival times of individual items along the African shore. That’s why we focussed on the more-robust things that the model tells us.
  2. As you correctly pointed out, a 30S crash site would, according to our model, have resulted in debris washing up on Madagascan and Tanzanian shores a full year earlier than was observed. That is a discrepancy that is hard to set aside.
  3. The other factor against 30S that we find very hard to discount is that 30S is right in the middle of the zone targeted most heavily by the surface search in 2014. This is the “other evidence” that Richard overlooked. Please see Section 4 of our Dec report, and Fig 4.2 of the April report. 

Best regards

David

817 Responses to “Important Questions About “Most Likely” MH370 Crash Site”

  1. DennisW says:

    @David/Victor

    On April 21 Higgins reports in (“MH370 drift modelling raises new search hopes for Malaysian airliner”)

    begin cut-paste//

    Transport Minister Darren Chester today played down expectations the hunt might restart soon.

    “I welcome the CSIRO report but it is important to note that it does not provide new evidence leading to a specific location of MH370,” Mr Chester said.

    end cut-paste//

    I think it is very clear at this point that the official search is a dead issue. I am a bit confused by this since the tripartite nations have much more and better information than they did when they initiated the $200M (Aussie Dollar) search. Only possibilities I can see are:

    1> People writing checks have finally realized that the science community over estimated the probability that the ISAT data (or any other data for that matter) can define a terminus. That has been my personal view from the get-go.

    2> Malaysian/Australian politics have taken over as a primary driver of search considerations. It is fairly obvious that Malaysia is not interested in finding the aircraft.

    Nice report Richard.

  2. David says:

    @Victor. The 3.2.1 diagram you use is as you say from 2016 but the new April 2017 has low windage debris arriving in Eastern Africa 6 months earlier. With South Africa a few months behind the equatorial arrivals that would see arrivals there in time. Related, I do not understand why the CSIRO model you use for the Google Earth depiction does not agree with their new 3.2.1.
    What do I miss?

    Along the same lines I notice that Richard Godfrey’s conclusions about African arrivals seem based on diagrams that extend just to August 2015.

  3. David says:

    @Paul Smithson. Your 2&3.The recent simulations (Search and debris examination, top p14) have included examples of loss of AC power on right engine failure. This causes loss of autopilot, via loss of pitot heat, and consequent prime speed data inaccuracy.

    Hence 4. does not arise.

    5,6 Yes.

    7. The APU will power essentials. See Page 5 of:

    https://www.dropbox.com/s/zot9fg5bgwunnk1/Design%20of%20Boeing%20777%20Electrical%20Sysem%2C%20boeing%20and%20Sundsrand00149784.pdf?dl=0

    Depending on what those draw in the circumstances there will be load shedding. IFE is a large power consumer, 22 kVA (Birtles, “Boeng 777: Jetliner for a New Century) but is not in the ELMS load shedding list. That does however include, “electronic seat equipment”. I do not know whether that is connected to IFE but in any case it is down the list, second last to be shed and early back on. So I assume that IFE would be powered.

    The left backup generator and PMGs would be functional also but would not power the SDU.

    8. Would not fly straight. See simulations.

    9. Let’s say the electrical load would have been 65kVA each engine supposing no galleys and otherwise lowered usage. Supposing a thermal efficiency of 30%, fuel to the black box, that equates to 17kg/hr of fuel. Over 7 hrs say, 120kg. Switching off left engine air bleed would save more still. Note the net effect would be to bring the right engine failure and 7th arc forward, all else equal.

  4. David says:

    @DennisW. The Ministers previously did not think that confidence that the crash site had been located was high enough for that.

    What now has changed is that after experimentation with a genuine flaperon CSIRO find it yields a higher probability of the beaching at La Réunion. This is seen by some as lifting confidence in the new search area – though from initial reaction, insufficiently.

    To me it is an invitation to do more work on drift analysis with selected recovered debris, a suggestion they rejected a month ago; but I hope the ATSB will reconsider and be provided funding.

    Thence with satellite data analysis and the help of a couple of unknown unknowns it might not remain a dead issue, though as you say it is now.

  5. Ge Rijn says:

    @VictorI

    I think a good summary about the conflicting conclusions this latest CSIRO-study brings into question.
    Their 35S anomaly within ~1 degree latitude is at least strange and needs further explanation.

    I understand from this study the sequence of graphics is primary based on what the flaperon would do. North and south of 35S there would numerous landfalls on WA till ~32S but only in case of the flaperon (or other high leeway, windage items).

    This did not happen till now we know. So in a way their proof serves like negative proof (which is kind of proof also ofcourse).

    The CSIRO-graphic you show makes this very clear. With 35S there were no landfalls on WA shores but only based on the flaperon and no flaperon could have arrived at Mosselbay.

    I think the problem is the flaperon only represents one end of the debris-spectrum.
    A lot more items should be used with the same method to get a better overall picture.
    The outboard flap section would be perfect for it’s original and in their hands.
    I wonder why they did not choose to take this piece for their experiments anyway.

  6. Andrew says:

    @Paul Smithson:

    “4. Autopilot remains engaged, if I’m not mistaken, using the hot battery?”

    Expanding on David’s comments above, the loss of AC power causes the loss of pitot probe heat. Consequently, the Primary Flight Control System (PFCS) reverts to secondary mode, because the system can no longer be assured of accurate air data due to the possibility of icing. The autopilot is only available when the PFCS is operating in normal mode, so it disengages when the PFCS reverts to secondary mode.

  7. Victor Iannello says:

    @David said, “The 3.2.1 diagram you use is as you say from 2016 but the new April 2017 has low windage debris arriving in Eastern Africa 6 months earlier…What do I miss?”

    The new report addresses the new leeway parameters for the flaperon and also discusses the drift of high (not low) windage debris. In the first study, the “non-flaperon” debris was modeled with a 1.2% windage coefficient. In the new study, the high windage debris was modeled with a 3% windage coefficient, which they say is more appropriate for debris such as seat cushions. The engine cowling debris found in South Africa, for instance, would have low-windage.

  8. Oleksandr says:

    @Andrew,
    @Paul Smithson,

    Re: “Expanding on David’s comments above, the loss of AC power causes the loss of pitot probe heat. ”

    You need to differentiate between the Main AC, and Transfer AC. SDU is lost if the Main L AC is lost. The pitot can be powered by either of the two backup generators, so that the loss of pitot does not occur even if both the IGDs and APU are not functional.

  9. David says:

    @Victor.The diagram I refer to is on their p14. High windage is the next.
    3.2.1 has 1.2% windage + 10 m/s, 10 deg. Another there with like outcome, 20 deg. I miss your point?

  10. David says:

    @Oleksandr. From the ATSB 23rd March in a response I posted, (if),”the left and right AC busses were powered from the right engine via the AC bus tie. In this purely hypothetical scenario, when the right engine failed, this would remove power from the SDU (along with other systems, including autopilot)…

    However they may not be right.

  11. Victor Iannello says:

    @David: Yes, you miss my point. The 1.2% windage + 10 cm/s (not 10 m/s) @ 10 deg is for a flaperon, as it says in the title and in the caption. The 10 cm/s is extra leeway that would not apply to the low windage debris.

    With the new report, there are now three classes of debris:

    1. The flaperon, which has 1.2% windage and extra leeway of 10 cm/s applied at an angle. The effect of extra leeway is now applied at all wind speeds rather than cut off at 10 m/s.
    2. The non-flaperon, low-windage debris (such as the engine cowling debris), which has 1.2% windage and no extra leeway.
    3. The non-flaperon, high-windage debris (such as seat cushions), which has 3% windage and no extra leeway.

    The figure I generated from the CSIRO KMZ file corresponds to non-flaperon, low-windage debris starting at 35S latitude.

  12. Victor Iannello says:

    @DennisW: The Australian political class was twice assured that the plane would be found where the ATSB said it would be. The first time was when the initial 60,000 sq km zone was defined. The second time was when another 60,000 sq km zone was added. If the politicians were very certain that searching the next 25,000 sq km would find the plane, they would proceed. They aren’t proceeding because they don’t want to be embarrassed again. I do believe, however, that the Australians are sincere in their desire to find the plane. In addition to the humanitarian justification, it would help re-build the damaged reputation of the ATSB.

    On the other hand, it is not clear that the Malaysians are sincerely interested in finding the plane.

  13. Andrew says:

    @Oleksandr
    @David
    @Paul Smithson

    RE: “You need to differentiate between the Main AC, and Transfer AC. SDU is lost if the Main L AC is lost. The pitot can be powered by either of the two backup generators, so that the loss of pitot does not occur even if both the IGDs and APU are not functional.”

    My apologies; I misread Paul’s scenario. The pitot probe heat is powered by the L & R transfer buses. If only the L IDG is isolated, then the transfer buses will be powered by the L backup generator when the R engine fails. In that event, the pitot probe heat will remain powered and the PFCS will remain in normal mode. The three AFDCs will also remain powered.

  14. Paul Smithson says:

    @Andrew, David, Oleksandr

    Thank you all for your knowledgeable inputs. Even if pitot heat had to await APU start-up to be restored, this is quite a quick process, is it not (unlike the longer time lag before SDU can become operational)?

  15. Victor Iannello says:

    Update on April 24, 2017.

    For those wishing to explore the drift model results in Google Earth, the KMZ files generated by CSIRO are available for the flaperon debris, for non-flaperon low windage debris, and non-flaperon high windage debris. The particular file I used to create the image at the top of the article is the file for low windage, non-flaperon items starting at 35S latitude. Once dragged into Google Earth, simply move the time slider and observe how the particles travel in time.

  16. TBill says:

    At the risk of over-simplifying, we seem to have 3 independent opinions of MH370 drift models/crash location:

    CSIRO: 35S (relates to ATSB 32.5 to 36S suggested new search area)
    Godfrey/Others: 30S centric
    Chillit Group: 22-27S

  17. Lauren H says:

    Should “Western” be “Eastern?” ie the western Indian Ocean meets the the shores of Eastern Africa or did debris really loop around the Cape of Good Hope to the shores of Western Africa

  18. Victor Iannello says:

    @TBill: I am trying to understand why CSIRO is promoting 35S as the crash site when their own models contraindicate this result. Unless I am misinterpreting their results, the parts found in December 2015 in Mozambique and South Africa are completely unexplained by their drift model for a crash site as far south as 35S.

    When they interpret their drift results, they seem to incorrectly group the arrival of debris to South Africa with the arrival of debris as far north and east as Madagascar as they choose 50E longitude as the eastern limit for debris reaching Africa. Debris hits the eastern shore of Madagascar months before anything reaches as far south and west as South Africa. It is therefore not correct to use the date that debris arrives in Madagascar to explain the timing of debris found in South Africa.

  19. Victor Iannello says:

    @Lauren H: Welcome to the discussion.

    Thank you. You are correct. I meant to say Eastern Africa west of 50E longitude. I have changed the incorrect references to avoid future confusion.

  20. Ge Rijn says:

    @VictorI

    What’s also completly unexplained yet by CSIRO’s conclusions is they predict quite some low-windage debris to arrive on WA shores coming from 35.5S. Which actually is a (IMO sound) prediction other studies made also which led to the conclusion crashs areas south of ~36S could be excluded for no debris is found on Australian shores (still).

    In this way the graphic you show in your topic proves on its own the CSIRO 35S conclusion is contradicting the known facts and therefore must be wrong.

  21. Victor Iannello says:

    @Ge Rijn: I did not point out that the model predicts, for a 35S starting point, that low-windage debris would arrive on the shores of WA because we don’t know for sure if the debris arrived there but was not found as only a small fraction of the debris ends there. But if we use the non-arrival of debris as a criterion, then I agree that it is another strike against 35S.

    I hope you’ve had a chance to study the time sequence using the KMZ files. Unless I am making an error, the engine cowling part found in South Africa in December 2015 could not be from a crash at 35S if the drift model is correct.

  22. Ge Rijn says:

    ..excusse me; ‘must be wrong’ is not the right statement. ‘Must be better explained’ would be better..

  23. Ge Rijn says:

    @VictorI

    I’m working on them. I think the data is great whatever CSIRO’s/Griffin’s conclusions are now which, being scientists, are open to change I assume.
    There’s a lot to it (and said already in the past) but I see the possibility of consensus building combining CSIRO/Griffin, Godfrey, Pattriatchi, MPat (lets not forget him..), Oleksandr and even BrockMcEwen studies defining a 1 or 2 degree latitude area in the near future.

  24. Victor Iannello says:

    @Ge Rijn said, “but I see the possibility of consensus building”

    It’s always good to have an optimist in the room.

  25. Oleksandr says:

    @David,

    Re: “the left and right AC busses were powered from the right engine via the AC bus tie. In this purely hypothetical scenario, when the right engine failed, this would remove power from the SDU (along with other systems, including autopilot)…”

    If the main left and right AC busses are powered by the right IDG, then when the right engine fails, the left engine backup generator feeds both the right and left transfer buses. However, if the left engine failed before that, then neither left nor right backup generator can provide power, and then most of the systems would become depowered.

  26. Oleksandr says:

    @Andrew,

    If the left main bus is manually isolated, and L backup generator is switched to off, does this also remove power from 28V L DC bus? Or L DC bus remains powered by the R DC bus? I am asking this because I saw somewhere semi-accurate information that in a B757 (if I am not mistaken) design such a pilot’s command also implies removing power from the L DC bus by the logic. I have not found anything on this matter with regard to a B777.

  27. Victor Iannello says:

    @Oleksandr: The L and R Main DC Busses provide backup power for each other. See the SmartCockpit reference, including the schematic on page 20/25.

  28. Lauren H. says:

    @DrB – You wondered how much fuel might be used during the descent when using the “Checkpoint” Tables. @Andrew gave you an answer regarding fuel use, and here is something I posted in Jan 2016:

    “Now, one problem with this range from a checkpoint is assumes a normal descent (.84/310/250) of 124 nm over 23 minutes with both engines running. The second problem is the final weight of 174,400 kg is low enough that the LRC speed would have starting to be reduced between the 5th and 6th Arcs. The critical challenge for MH370 is modeling the path after the speed reduction for the lower GW and further reduction after the flameout of the right engine.”

  29. Oleksandr says:

    @Victor,

    The issue is not whether they can provide power to each other. If design of a B777 is similar to older designs (and if this information is correct), then isolating L main generator, L bus tie and switching OFF L backup generator (which should normally be ON), may be interpreted by the power controlling logic as a command to isolate all L buses, including L DC bus. I have not found whether this philosophy is applicable to B777 or not. You may notice that the power control panel does not have DC controlling switches; only AC.

  30. David says:

    @Victor. Re drift rates. Got it thanks

  31. Mick Gilbert says:

    @Oleksandr
    @Victor

    With regards to isolating either DC Bus, I have looked at this matter previously and if I remember correctly (I’m sure Andrew or Don wil correct me if I haven’t) the process of isolating the Left AC Bus does not isolate the Left DC Bus. Per the attached, the DC Buses are powered off their respective AC Transfer Buses via their own Transformer Rectifier Unit and are tied by the DC Bus Tie relay. Isolating an AC Bus does not isolate the respective DC Bus and the DC BTR cannot be opened from the cockpit.

    https://www.dropbox.com/s/egsn993k102xbd2/B777%20Electrical%20Power%20Schematic.png?dl=0

  32. Andrew says:

    @Oleksandr
    @Victor
    @Mick Gilbert

    “If the left main bus is manually isolated, and L backup generator is switched to off, does this also remove power from 28V L DC bus? Or L DC bus remains powered by the R DC bus? I am asking this because I saw somewhere semi-accurate information that in a B757 (if I am not mistaken) design such a pilot’s command also implies removing power from the L DC bus by the logic. I have not found anything on this matter with regard to a B777.”

    In older designs such as the B747/757/767, the operation of the DC Bus Tie breaker(s) was ‘tied’ to the position of the Bus Tie switches. If a Bus Tie switch was selected to ‘ISLN’, the DC Bus Tie breaker would remain open and the associated DC Bus could not be powered from another bus. The B757/767 FCOMs state that “Both BUS TIE switches must be in AUTO for the DC bus tie breaker to close”.

    In the B777, the operation of the DC Bus Tie relay is only dependent upon voltage inputs from the left and right dc buses and output current from the TRUs. If one of the two dc buses has low voltage, the dc tie relay closes, connecting the two buses. The tie relay then remains closed until valid current is detected from both TRUs. Neither the FCOM or engineering manuals mention any connection between the Bus Tie switches and the operation of the DC Bus Tie relay.

  33. David says:

    @Paul Smithson. Correcting the record as to your earlier questions:

    Q4. “Autopilot remains engaged, if I’m not mistaken, using the hot battery?”
    A. It would remain engaged, powered by the left backup generator. In the simulations exploring that ATSB-defined configuration the left backup generator would have to be disconnected already.

    Q8. “Plane flies beyond 7th arc for some minutes under power from left engine.”
    A. Yes, and it would fly straight. Food for thought if the right engine fuel consumption matched the 7th arc, that is if fuel consumption was lower than thought. I notice Dr B in the last post, at 221336 described a circumstance where, “There is too much fuel for exhaustion to occur at 00:17 unless there is a source of extra drag or fuel consumption beyond normal flight. I don’t know what that could be.”

    About your most recent:

    Q. “Even if pitot heat had to await APU start-up to be restored, this is quite a quick process, is it not (unlike the longer time lag before SDU can become operational)?”
    A. Restoration of pitot heat however quick would not re-engage the autopilot automatically, which needs a manual selection.

  34. David says:

    @Oleksandr. Responding to the ATSB note,”If the main left and right AC busses are powered by the right IDG, then when the right engine fails, the left engine backup generator feeds both the right and left transfer buses. However, if the left engine failed before that, then neither left nor right backup generator can provide power, and then most of the systems would become depowered.”

    Yes. I expect the ATSB was thinking of its simulation specification and left unstated that the above outcome entailed disconnecting the left backup generator. I did not spot that.

  35. Andrew says:

    @Oleksandr
    @Victor
    @Mick Gilbert

    In my previous post I should have mentioned that autoland requests also have an effect on the operations of the DC Bus Tie relay, but that’s not relevant to this discussion.

    Circuit diagrams for the B757 and B777 DC Bus Tie relays here (if you’re really keen!):

    B757 DC Tie Relay

    B777 DC Tie Relay

  36. David says:

    @Oleksandr. I disambiguate. Responding to the ATSB note YOU SAID, “….

  37. Andrew says:

    @Paul Smithson
    @David

    RE: “Even if pitot heat had to await APU start-up to be restored, this is quite a quick process, is it not (unlike the longer time lag before SDU can become operational)?

    The APU start sequence takes approximately one minute, but as David noted, the PFCS will not automatically revert to Normal Mode after power is restored. The PRIMARY FLIGHT COMPUTERS Disconnect Switch must be cycled to DISC and back to AUTO to restore Normal Mode operation. The autopilot should be available once Normal Mode is restored, but again it must be engaged manually by pressing one of the Autopilot Engage Switches on the MCP .

  38. David says:

    @Paul Smithson, Andrew. From memory the APU autostarts after loss of all AC. I will confirm that or perhaps Andrew will more quickly.

    If so, without the left backup generator isolated there would be no APU autostart until the left engine failure and so no extension to flight beyond the 7th arc.

    However the SDU is not powered by the backup generator so will have already dropped off line at right engine failure.

  39. Andrew says:

    @David
    @Paul Smithson

    That’s correct; the APU does not autostart unless both AC transfer buses lose power. If the left backup generator is supplying power, then the AC transfer buses will still be powered and the APU will not autostart.

  40. David says:

    @Paul Smithson, Andrew. From the maintenance manual, “The APU will automatically start for these conditions:
    * Airplane is in the air
    * Power removed from the left and right transfer buses.”

  41. Andrew says:

    @David

    Snap!!

  42. David says:

    @Andrew. Yes, stumbled on it quicker than I thought. I scratch my head about that configuration and sequence and what part it could have played in either log-on but no brain wave.

  43. Rao says:

    CSIRO is a scientific research organisation. It used a replica wing to arrive at some numbers. It only played a small part in solving a big and complex problem.

  44. Andrew says:

    @David

    By “that configuration” do you mean “an electrical configuration where the loss of engine power from one engine resulted in the loss of autopilot”, as described in the ATSB report MH370 – Search and debris examination update, or a configuration where the right engine has failed and only the left backup generator is providing power?

  45. David says:

    @Andrew. The second Andrew, which drops the SDU off line yet keeps the autopilot: because I had not encountered it. But it takes someone in the cockpit still and there are plenty of alterrnatives there already.

    To me the main difference it offers for the final log-on is that the SDU is off for longer so can cool more (OCXO drift). Also though the right failure can be even earlier than the fuel loading and PDAs would suggest if the left transfers fuel consumption to the right by offloading electrical load. I would not see that necessarily as being from intent. Switching off left engine bleed air to increase the fuel difference yet more on the other hand would indicate intent; though I can see no reason to do that, or gain.

    Having the autopilot is of course essential to get the aircraft to the 7th arc if unpiloted.

    Flying for longer on one engine would not help get the a/c there on time.
    However as I see it, fuel and timing constraints are being lifted progressively as the most likely crash site moves north east. Just musing while the others sleep.

  46. David says:

    @Andrew. My second sentence does not belong there but to another thought, which was the right IDG being switched off then back on, prompting the 7th arc log-on but without APU start.

  47. Paul Smithson says:

    @Andrew, David, Oleksandr, Mick, Victor

    The “rationale” of the scenario that I am trying to envisage is described briefly below. Perhaps one of you can improve upon it or knock so many holes in it that it dies a natural death.

    1. The loss of multiple systems powered by Left AC is not due to some freak accident that happens to affect those specific systems.

    2. Instead, it is due to some sort of serious electrical problem that cannot be resolved with isolation of individual items and prompts the pilots to remove power entirely from left side to protect the aircraft.

    3. This stabilises the situation and they proceed with checklists/diversion. Just prior to planned landing (at FMT) they take the risk of reapplying power to left side because of items required for safe approach and landing [not sure what these vital “exclusively left’ items are]. In doing so, they deliberately choose to power left side from right IDG – this maybe being a slightly lower risk than re-powering from left IDG if they know they have a problem on that side.

    4. The original problem does recur and incapacitates the pilots quite quickly. So SDU re-boot, FMT and final incapacitation occur more or less at the same time (~182430) and the intended landing is not followed through.

    5. Hereafter, aircraft is flying with electrical power from right engine. I assume that both left IDG and left backup are off/isolated. Because if left back-up was running you wouldn’t have a power cycle when R engine flames out.

    6. Power-cycle of SDU will now arise when right engine flames out and APU auto-starts. Meanwhile, the plane continues flying straight on left engine with fuel remaining.

    In light of your collective comments:-

    Are we sure that autopilot irreversibly [excepting human intervention] drops out with loss of A/C power between R flame-out and APU auto-start? Can pitot heating not be delivered by hot battery? Or perhaps pitot temp needs to drop below a certain threshold to take A/P offline –
    and a few seconds between R power loss and APU start up is not long enough for this to happen?

    BTW, my understanding is that the “60 seconds” power up time refers to period between power-restoration and SDU logon, not APU power availability. With domestic generators on auto-switch, power is available within about 5-8 seconds after mains power is lost

    In any case, what I’m looking for is a scenario whereby SDU power-cycle is associated with R engine flame-out, with plane continuing to fly straight with fuel remaining in L tank.

  48. Andrew says:

    @Paul Smithson

    “Are we sure that autopilot irreversibly [excepting human intervention] drops out with loss of A/C power between R flame-out and APU auto-start? Can pitot heating not be delivered by hot battery? Or perhaps pitot temp needs to drop below a certain threshold to take A/P offline –
    and a few seconds between R power loss and APU start up is not long enough for this to happen?”

    Absolutely sure! The pitot probe heat requires 115V AC power from the left and right transfer buses. As I said previously, loss of the autopilot occurs when the PFCS downgrades to secondary mode on the loss of power to the pitot probe heat system. I’m not entirely sure of the pathway, but I think a pitot heat ‘Failed Off’ signal is sent from the L, C & R pitot ADMs via the via the Flight Controls ARINC 629 Bus to the PFCs, causing the PFCS to downgrade to secondary mode. The ‘Failed Off’ signal is generated by power loss, not a drop in pitot probe temperature.

    “BTW, my understanding is that the “60 seconds” power up time refers to period between power-restoration and SDU logon, not APU power availability. With domestic generators on auto-switch, power is available within about 5-8 seconds after mains power is lost.”

    The APU generator is not on immediate standby. The APU must first start and It takes about 60 seconds for the APU to begin producing electrical power after the start sequence is initiated. The SDU then takes another 60 seconds to begin transmitting a log-on request (see ATSB MH370 – Definition of Underwater Search Areas, dated 3 December 2015, p.10).

  49. Andrew says:

    B777 Pitot Probe Heat circuit diagram shown here:
    B777 Pitot Probe Heat

  50. DennisW says:

    @Paul

    Unfortunately the lack of communication argues strongly against your scenario i.e. it is just plain dumb.

  51. Paul Smithson says:

    @Dennis. My recollection is that at least two of the radios will not function without A/C power on left. In which case no comms may be an undesired side-effect of left side electrical shutdown, done to protect the aircraft.

  52. David says:

    @Paul Smithson. Adding to Andrew, in your hypothetical there would be no SDU recycle at the FMT since there is an IDG running during all that.

    The scenario you seek in your last para for the last log-on is unavailable I think as I said earlier, that is without manual intervention.

  53. Paul Smithson says:

    @Andrew. Allow that pitot heat off prompts the PFCS to drop into Secondary mode. Why can it not continue to fly straight in this condition?

    quote: end-quote

    http://www.davi.ws/avionics/TheAvionicsHandbook_Cap_11.pdf

  54. Paul Smithson says:

    @David. If SDU is powered from Left A/C bus; that bus (and associated left IDG and backup) shut down entirely at IGARI; then power restored to left bus from right IDG via bus tie at FMT, you would get an SDU re-boot, would you not?

  55. Mick Gilbert says:

    @DennisW

    Lack of communication is most assuredly not the “knock out” factor that many might assume it to be. In a quirk of ergonomics over survivability design thinking, on the B777 all the audio, radio and transponder controls are mounted on the Aft Aisle Stand Panel, a 40 centimeter by 50 centimeter “centre console” located between the two pilots. All the splendid redundancy associated with two transponders and five radios (three VHF and two HF) on the B777 comes together at this relatively compact single point of failure. On at least one occasion on a B777 a mishandled beverage spilled onto the Radio Tuning Panel has caused a significant communications failure.

    Ie. It’s not just plain dumb, at all; it’s possible and supported by precedent. (And surely, in the spirit of collegiality, there might have been a better way to express that.)

  56. Don Thompson says:

    Paul,

    Lack of communication is a significant problem: none of the VHF radios require power from the AC busses, all three are DC powered. The ‘left’ HF transceiver & antenna coupler is powered off the Left AC Transfer bus, while the ‘right’ is supplied by (a branch of) the Right Main AC bus.

    WMKK Area Control Centre is contactable via HF and VHF frequencies. Apparently, MAS OCC is also directly contactable via HF.

    There is no suitable landing site in the vicinity of the FMT. Indonesia’s AIPs stated that Banda Aceh is ‘turned off’ at night: no radio, no radio nav aids, the airfield might just be evident as a black hole among any lights from population centres in the area. It was not a moonlit night. No published info on Car Nicobar, so feel free to argue that one (runway strength, radio aids, runway lighting, …). Forget Baaz/Campbell Bay, it’s approx 1500ft long, ditching in the bay might be safer that risking ploughing through the runway.

    If a diversion was an intention why no hint of an attempt to descend for WMKP, the most suitable alternate? The DCA radar track shows that the aircraft just clipped the top of the detection envelope from the Penang & Butterworth terminal area control ATCR-33 radar head on Butterworth airbase.

    BTW: while the three ADIRS pitot probes are powered off the XFR busses, the heaters for the ADIRS AoA and TAT probes are supplied via the main AC busses (& monitored for correct operation, as the pitots are):

    Pitot probe (3 of) heater powered off 115v AC [LEFT|RIGHT] Transfer Bus
    AoA probe (2 of) heater power off 115V AC [LEFT|RIGHT] AC SEC 1 Bus
    TAT probe (1 of) heater powered off 115V AC LEFT MAIN AC Bus

  57. Don Thompson says:

    Adding to previous: the SEA-1 and SEA-2 HF radio stations at Hat Yai, providing Long Distance Operational Control service, would also have been available to the crew.

    (Shame that DXing isn’t such a ‘thing’ in the internet age & ATClive only carries a few HF freqs.)

  58. Don Thompson says:

    Mick,

    Yes, three radio tuning panels and three audio control panels located on the aisle stand. Plus WXR, ATC XPNDR, Rudder trim panel, an MCDU, the flight deck door lock, and sundry other. Precipitously close to the cursor controls and throttle quadrant (position sensors and drive motors)

    Can the chain of necessary coincidences be so long?

    :Don

  59. David says:

    @Paul Smithson. Yes I agree. I had in mind our previous exchanges and your Q2 about a left IDG shutdown.

  60. Andrew says:

    @Paul Smithson

    RE: “Allow that pitot heat off prompts the PFCS to drop into Secondary mode. Why can it not continue to fly straight in this condition?”

    In your scenario, the downgrade to Secondary Mode and disengagement of the autopilot is caused by the loss of AC electrical power to the transfer buses after the right engine fails. The left engine is still producing thrust, but the TAC is not available in Secondary Mode. Consequently, the asymmetric thrust will cause the aircraft to yaw and roll.

  61. Oleksandr says:

    @Andrew,

    “In the B777, the operation of the DC Bus Tie relay is only dependent upon voltage inputs from the left and right dc buses and output current from the TRUs. If one of the two dc buses has low voltage, the dc tie relay closes, connecting the two buses. The tie relay then remains closed until valid current is detected from both TRUs. Neither the FCOM or engineering manuals mention any connection between the Bus Tie switches and the operation of the DC Bus Tie relay.”

    This sounds like you lean to think that the DC tie relay is fully automatic, but not 100% sure. Is there any way to depower DC bus from the cockpit?

    Btw, do you know what power source is used for the check and warning (if required) of the emergency oxygen supply to the cockpit?

  62. Oleksandr says:

    @Dennis,

    “Unfortunately the lack of communication argues strongly against your scenario i.e. it is just plain dumb.”

    Not really. The only difficult part is to explain the silence of the right VHF. The silence of the rest (2 transponders, R & L VHF, and 2 HF) is explainable. Together with downed or malfunctioning ADIRU and followed slow depressurization.

  63. Oleksandr says:

    @Paul Smithson,

    “The loss of multiple systems powered by Left AC is not due to some freak accident that happens to affect those specific systems.”

    My working hypothesis is that the “accident” did occur, and it knocked down part of the equipment. The other part was occasionally switched off by the crew, who thought they were dealing with fire.

  64. Paul Smithson says:

    @Oleksandr. Can you elaborate, please, on how 3 out of 4 radios’ silence is explainable?

  65. Andrew says:

    @Oleksandr

    RE: “This sounds like you lean to think that the DC tie relay is fully automatic, but not 100% sure. Is there any way to depower DC bus from the cockpit?”

    Not at all. I only added that last sentence to highlight the difference between the B757/767 and the B777 manuals. The B757/767 manuals specifically state that the AC Bus Tie switches must be in AUTO for the DC Bus Tie relay to close. The B777 manuals do not. If you have a look at the circuit diagrams in my subsequent post, you can see that the B757 diagram shows the wiring and logic that requires the Bus Tie switches to be in AUTO. The B777 diagram does not.

    For the avoidance of doubt: The operation of the DC Bus Tie relay on the B777 is fully automatic. There is no way for the pilots to remove power from the DC buses other than turning everything off.

    “…do you know what power source is used for the check and warning (if required) of the emergency oxygen supply to the cockpit?”

    28V R DC Bus

  66. Oleksandr says:

    Mick Gilbert,

    Thanks; I am aware of that pretty clear electrical power scheme. Actually I posted it several times. There are, however, many nuances, particularly with regard to the tie logic. Also we do not have complete list of what equipment is powered by what bus.

  67. Oleksandr says:

    @Andrew,

    Many thanks. That clarifies my questions.

  68. Victor Iannello says:

    Update 2 on April 25, 2017.

    The Guardian has published an article that discusses our interpretation of CSIRO’s results. “Both CSIRO and the ATSB have been contacted by Guardian Australia for their response.” No response has yet been received by reporter Elle Hunt.

  69. Oleksandr says:

    @Paul Smithson,

    Re: “Can you elaborate, please, on how 3 out of 4 radios’ silence is explainable?”

    Physical impact of E1 rack.

    If you take a look at the EE-Bay layout, you may find that VHF COMM XCVR L & C units are located close to each other at E1 rack, so that in case of a physical impact, they could be damaged at a time. Alternatively, VHF-L could be damaged, but the impact could also cause a short of the L DC bus, which powers VHF COMM C. Automatic removal of the power from the L DC bus brings down VHF-C, so that both VHF-L and VHF-C would be silent. The same event could cause a short of the L AC transfer bus as the TRU-L is also installed at E1 rack, so that HF-L would be also downed.

    If the crew voluntarily isolated the main L & R AC buses from the IDGs to reduce electrical load due to suspected fire in the EE-bay, HF-R and SDU would become inoperative (HF-R and SDU are inoperative when on backup generator).

    Result: VHF-L, VHF-C, HF-L, HF-R and SDU are inoperative.

    What could cause a physical impact of E1 rack:
    1. Rupture of the LH nose landing gear tire and tire rim (“The missing bolt” hypothesis).
    2. Explosion of the oxygen tank.
    3. Sabotage (explosive in the EE-Bay).
    4. External impact.

  70. Oleksandr says:

    @Victor,

    Re: “The Guardian has published an article…”

    In my post April 21, 2017 at 6:42 am, after taking a quick look at the link posted by ALSM, I made a brief summary: “I do not trust CSIRO’s studies anymore”.

    CSIRO has not provided any substantial argument to take their 35S seriously. Let it rest.

  71. TBill says:

    @David
    “Switching off left engine bleed air to increase the fuel difference yet more on the other hand would indicate intent; though I can see no reason to do that, or gain.”

    Of course I feel bleed air could be off for nefarious reasons, maybe secondarily to increase thrust. The two issues I have with fuel duration models we are developing: (1) we have no data on MH370 jet fuel or energy content; and (2) potential impact if bleed air turn-off during the flight. Perhaps the jet fuel quality is a non-controversial item that ATSB or MY has available and would be willing to supply on request.

  72. Ge Rijn says:

    @VictorI

    Without positivity this blog won’t exist..

    I’ve been through those files.
    In short I see two common demeanors.
    The first is the likelyhood of (non flaperon) debris reaching the shores of Mosselbay and Tanzania is not a criterion between 30S and 39S.
    All those files show green dots arriving there before or around 31 december 2015 (not all beaching but the posibility is obvious to me).
    So I wonder a bit how you’ve got your topic-file from on 35S.

    Starting from ~33S the chances of reaching Mosselbay though get progressively bigger till 29S where it drops flat out.

    A far more obvious criterion is the amount or lack of landings on Australia. For non-flaperon and flaperon-like debris the devide is most obvious at 35S. North of 35S the trend for non-flaperon debris fades and south of this latitude it increases (as perdicted by other studies).

    We know all found flaperon-like items where found on African/Island beaches. The flaperon, outboard flap section and probably those three flap fairing pieces all were high-windage pieces.
    None of them followed the prediction of CSIRO/Griffin by landing on Australian shores from 35S. And none of the non-flaperon pieces did either.

    This facts clearly speak to me against the CSIRO/Griffin conclusion.
    Actually their models are proving clearly their conclusion must be ‘wrong’.

    The best coverage according to the known facts and their data gives ~31S till ~30S.

    I know the WA shore is probably one of the most desolated shores in the world.
    I’ve travelled that coastline from Cape Leeuwin till Exmouth.
    It’s very well possible if debris landed there somewhere (especially north of Geraldton till Exmouth) it never gets found.
    In a way it would be a blessing something obvious got found there.
    It would refine the drift analisis dramatically.
    But after more then 3 years this did not happen still.
    So we can only deal with the facts as they are now.

  73. Victor Iannello says:

    @Ge Rijn said, “The first is the likelyhood of (non flaperon) debris reaching the shores of Mosselbay and Tanzania is not a criterion between 30S and 39S. All those files show green dots arriving there before or around 31 december 2015 (not all beaching but the posibility is obvious to me).”

    For a 35S crash site and with low-windage debris, CSIRO’s model predicts that on December 31, 2015, no debris is within 1800 km of where the engine cowling part was found near Mossel Bay, South Africa. By comparison, for debris starting from 30S and 27S, debris is predicted as far south as Mossel Bay for that date. Although it is only one piece of debris, the timing and location is important for eliminating potential crash sites (assuming CSIRO’s own drift model is accurate).

  74. DennisW says:

    @VictorI

    “No response has yet been received by reporter Elle Hunt.”

    I doubt very much Hunt will get a response. Possibly some form of convoluted reply on the ATSB web site similar to the Higgins treatment and the explanation for why the “consensus” language was changed.

    @Ge Rijn

    It has been about 10 months since any confirmed or highly likely debris has been found. Given the rate of confirmed or highly likely debris finds in the year following the flaperon finding, it seems unlikely that much if any more debris will be found.

  75. TBill says:

    @Victor
    To say 30S, I guess we have to critique ATSB stance that areas north of 32.5S are ruled out due to lack of visual evidence of floating debris after the accident. Either we have to say: (1) Debris was there but was missed or moved by currents, (2) Pilot found a relatively clean way to sink the aircraft; (3) crash was somehow outside boundary of the visual search zone.

  76. Victor Iannello says:

    @TBill: If you look at the cumulative detection probabilities presented in Fig 4.2 of the new CSIRO report, as you proceed north of 31S along the arc, there are squares not at 100% (red), especially north of 27S. But at this point, I don’t have much confidence in ANY probability they assign. For instance, we know the probabilities calculated for the location of the plane, based on DSTG’s Bayesian analysis, were incorrect. I think the kind of study performed by CSIRO can only provide qualitative guidance and relative probabilities.

  77. Mick Gilbert says:

    @Don Thompson

    G’day Don, you asked, “Can the chain of necessary coincidences be so long?”

    Yes, as accidents in general and aviation accidents in particular have illustrated over many decades.

  78. Mick Gilbert says:

    @Ge Rijn

    Much like you I thought that surveillance of the majority of Western Australia’s beaches would have been scant. Back in my air force days I was involved in exercises out of RAAF Bases Learmonth (near Exmouth) and Curtin (near Derby) and in my early days with Rio Tinto I worked at the Argyle Diamond Mine in the Kimberly’s so I’m reasonably familiar with how thinly populated the Western Australian coastline is north of Geraldton apart from a handful of tourist destinations such as Broome. Consequently, I was very surprised to hear about the efforts of the Tangaroa Blue Foundation, “an Australian-wide not-for-profit organisation dedicated to the removal and prevention of marine debris, …”. Not only do they visit and clean up remote beaches, they also maintain an excellent database of their efforts (http://www.tangaroablue.org/database.html). Much to my amazement, they have visited and removed debris from many of Western Australia’s beaches, including some of the more remote ones north of Geraldton, since mid-2014.

  79. buyerninety says:

    @Don Thompson
    Where is the power sourced from for the microphone/ear speaker(s) in the co-pilots oxygen mask?

    @Oleksandr
    A thought for you.
    Perhap the right VHF radio was being setup/used for a different activity than communicating.
    Aviate, Navigate, Communicate.
    Cheers

  80. Brock McEwen says:

    Can folks please post to this forum all they know about US assets (naval or otherwise) which…

    a) are equipped with primary radar, and
    b) were within 350 nmi of a line connecting IGOGU to BITOD.

    I dearly hope – and strongly suspect – any such intel would be stale enough by now to merit getting it out on the table, to help inform path projection – especially if we’re talking about positions during training exercises. MH370 could not possibly have gone where primary radars were in place, correct?

    Getting the US info out on the table may shame other nations into more fulsome disclosure, by setting a world-class standard for transparency and accountability.

    Huge thanks in advance.

  81. Brock McEwen says:

    (“IGOGU to BITOD”…on the night of March 7, 2014, as was hopefully clear from context.)

  82. TBill says:

    @Victor
    I see what you mean, there is an apparent window of opportunity right at 30S unsearched for surface debris, and 26.9S (McMurdo path) is not so red either.

  83. Brock McEwen says:

    @Don: thanks for the additional sourcing.

    Believe me – nobody is keener than I to avoid game-playing with evidence. Hence my polite request for the name of your UK reporter. I like to independently corroborate my sources, too. Can I impose upon you please either to supply the name (by personal correspondence is fine), or explicitly refuse to supply it. I’m not even asking you to justify a refusal – I’ll simply record the refusal, and move on to the next topic.

    Thanks in advance for your time and consideration.

  84. DennisW says:

    @Mick

    Given the flight path recorded by radar and compatible with satellite data from IGARI to 18:25 is there any doubt in your mind the someone was piloting the aircraft?

  85. Brock McEwen says:

    @Mick: can you please provide specific precedents for comms being knocked out as totally and for as long as the official narrative requires. “Significant communication failure” just won’t get the job done. Thanks.

  86. Ge Rijn says:

    @VictorI

    I ran all the non-flap and flap files from the link you posted in your topic: ‘the KMZ files generated by CSIRO’
    Every file between 39S and 29S show green arrows have arrived near or at the tip of SA on 31-12-2015 and also at Tanzania.
    The big variation only happens at/on the WA coast.

    Like your posted 35S graphic (which is different from the 35S KMZ file in missing a lot of green arrows near Moxambique/SA) I only count the green arrows.
    I’m making a mistake?

  87. David says:

    @TBill. An excellent Chevron booklet covers all you ever wanted to know and more about aviation fuel.

    https://web.archive.org/web/20090704014521/http://www.chevron.com:80/products/ourfuels/prodserv/fuels/documents/aviation_fuels.pdf

    MH370 used Jet A-1 and the specs for that are at p17, including minimum specific energy (net heat of combustion). The energy of such fuels is much more mass dependent than volume, which is described briefly at p17. MH370 tank measurements are volume, density and temperature and there are water detectors.

    Combustion characteristics (pages 4 and 40) can also affect fuel consumption.

    Fuel consumption predictors in manuals take no account of specific energy (net heat of combustion) or other variations, indicating that there is little difference amongst suppliers, batches and crude sources (though notice the density specification range of 8.4%, which will affect tanks-full range).

    Post production, what could affect energy value I suppose is dissolved water but I cannot see that being an issue at major airports with their fuel installations and quality control (see storage and handling specs p82).
    Overall then it may well be that differences between fuel loads are unnoticeable.

    @Andrew. Have you noticed differences please in fuel consumption between different fuels on long trips or an airline fuel sourcing based on that?

  88. David says:

    @TBill. On your other question about fuel consumption of bleed air I posted an assessment of that a while back. From recall the gist of it was that the packs consume air at the rate of 280lbs/min.

    Separately, fuel consumption increases 0.7% for each recirculation fan turned off, due to consequent pack air flow increase of 60lbs/min.

    Proportioning from that ratio, supposing linearity or near enough, yields a fuel consumption for total pack flow of about 3¼ %. There might be a better measure of that from elsewhere. The figures I got from the Maintenance Manual. If it is important I will recheck them.

  89. Ge Rijn says:

    @DennisW

    I think your Weibull-prediction is still on schedule..
    The last two confirmed ‘highly likely’ pieces are from december 2016 (the right wing aileron piece) and from february 2017 (the no.7 flap fairing piece).

  90. TBill says:

    @David
    Thank you for the Chevron jet fuel info.
    This Boeing presentation (slide 98) suggests jet engine fuel flow has a correction term for jet fuel energy content.
    https://www.slideshare.net/sanjuvinaykumar/jet-engines-basics

    Of course, we can use the standard assumed energy content, but I would ask what actual data we have? There can be variations in jet fuel quality within the specs. Years ago (TWA800 period) USA typically cut jet fuel heavier because US refineries tended to max out gasoline yield. TWA800 middle tank had EU (lighter) jet fuel from the prior flight, I believe. But not sure trends today.

  91. Mick Gilbert says:

    @Brock McEwen

    Brock, if you’re asking me to provide a precedent for a B777 suffering a total communication failure for over seven hours, I can’t. If there were any direct precedents for MH370 I think we’d have had it solved by now.

    What I can do is provide you with he reference to NASA ASRS ACN: 711779 B777 coffee spill on aft aisle stand panel. That incident demonstrates my point, namely the vulnerability associated with having redundant communication systems come together at a relatively compact single point of failure on the B777. It is a matter that most B777 operators attempt to address by crew instructions (eg beverages to be served to flight crew around the outside ie to the Captain’s left and First Officer’s right) and that Airbus deliberately avoid by having at least one comms control on the overhead panel.

    While it doesn’t apply to Paul’s hypothesised scensrio, it is also worth noting that a comms systems failure doesn’t have to be either complete or persistent if it was accompanied or followed by crew incapacitation.

  92. Mick Gilbert says:

    @DennisW

    G’day Dennis,

    You asked, “Given the flight path recorded by radar and compatible with satellite data from IGARI to 18:25 is there any doubt in your mind the someone was piloting the aircraft?”

    Short answer, it depends on what you mean but, in all likelihood, yes there is.

    Long answer, let’s start in the traditional fashion and define the terms. I take “piloting the aircraft” to mean continuously monitoring the course, altitude and speed of the airplane and making considered decisions about inputting deliberate or correcting unwanted changes to the course, altitude and speed of the airplane. If you had something else in mind, let me know.

    With that out of the way, from IGARI to 1825 UTC we’re talking about two significant changes to the course, altitude and speed of the airplane – the initial turnback at IGARI commencing around 1720:30 UTC and the turn up the Malacca Strait from south of Penang at around 1753 UTC. With regards to the former, there was most assuredly someone piloting the aircraft at that point. The turnback is somewhat consistent with someone commanding an initial diversion to Kota Bharu (WMKC) by selecting it off the FMC Alternate Page (WMKC would have been the No 1 “default” option) and commanding Divert Now, Execute followed a few minutes later by the choice of destination being amended to Penang (WMKP).

    With regards to the the turn up the Malacca Strait from south of Penang, there are at least three possible explanations;
    1. the turn was commanded deliberately at the time (ie around 1753 UTC),
    2. the turn had been deliberately programmed into the FMC sometime before 1753 UTC, or
    3. the turn was the result of an LNAV discontinuity arising from an incomplete approach being programmed into the FMC sometime before 1753 UTC. A truncated BIDMO 1A STARS approach (eg BIDMO-PUKAR-ENDOR-MEKAT-KENDI with PUKAR-ENDOR deleted to straighten and expedite the approach) for ILS04 would have a discontinuity between KENDI and CF04-FF04-RW04. If the discontinuity had not been deleted the airplane would have reverted to Constant Magnetic Heading as it passed through KENDI.

    In sum, there is no compelling evidence to suggest that someone was in continuous control of the airplane subsequent to the turnback at IGARI. So, yes, there is doubt in my mind that someone was piloting the aircraft between IGARI and 1825 UTC.

  93. Paul Onions says:

    ATSB has stated that the flight was consistent with FL300. If we assume that 0017:30 is fuel exhaustion for FL300, then at the more fuel efficient level of FL350, there would be fuel remaining.
    This remaining fuel could equate into extra distance from the seventh arc.

    Example 1. The remaining fuel was used by the APU (which had been turned on by the crew just after Igari). When the left engine finally failed due to fuel exhaustion, the APU connected to the Left Main AC bus. Because the Left GCU is unable to alter the speed of the failing Left IDG to match frequency with the new APU power source, a break power transfer occurred and the Satcom was momentarily depowered. Then the right engine stopped. The APU connected to both Main AC busses, and load shedding of IFE occurred. The autopilot is still engaged trying to maintain LNAV and altitude. At stall speed, flight envelope protection overrides VNAV and pitches the nose down to maintain speed resulting in a descent rate of 4000 ft/min. The 0019:37 BFO is bogus as stated by Inmarsat. With speed regained, the autopilot tries to return to FL350 by pitching the nose up. Speed reduces to stall speed again and the process repeats until the APU stops 13 minutes 45 seconds later. The aircraft is further south from the hotspot.

    Example 2. The Left GCU and backup converter (which both have wiring to the P100 Left power panel near the oxygen bottle) are both inoperative and the right IDG is supplying all the power to the Left AND Right Main AC busses. When the right engine stops due to fuel exhaustion, all AC busses are depowered. The autopilot disengages and the APU autostarts. The APU repowers the Left and Right Main AC busses. Load shedding of IFE occurs. Satcom logs on. The left engine is still operating and alters the trajectory. The aircraft ends further than 40 nautical miles from the arc.

    Why didn’t they just search a 100 nautical mile radius of the Bayesian approach hotspot?

  94. DennisW says:

    @Ge Rijn

    I had not heard about the classification of the pieces you mentioned as “highly likely”. I was basing my comment on the CNN articles linked below.

    http://www.cnn.com/2017/03/08/asia/mh370-debris-found/

  95. Oleksandr says:

    buyerninety,

    I was thinking about such a possibility. Furthermore, if ACARS was switched to VHF-R at some point, this could contribute to the explanation of the absence of ACARS messages after the reboot 18:22. However, in my understanding, normally VHF-C is used for data, but. In addition, Captain’s VHF panel is powered by Captain’s bus, so whoever was in control would be able to switch VHF-R to voice.

    Anyhow, a combination of a single physical impact at the left side of the EE-Bay followed by isolation of the main AC buses by the crew (to prevent potential fire spread) explains almost all the observed features of MH370 flight.

  96. David says:

    @TBill. Thanks. First one needs to know what the LHV of the fuel load is (which will be a mix of previous loads and the new and might vary between tanks), then decide whether precision is needed.

    Determining the LHV before most flights would be impracticable I think though it would be needed in flight testing. In FCOM etc I have seen no place for such an adjustment. Maybe missed it, a current ailment.

    Interesting that they set their projected consumptions on the average though. I suppose encountering fuels below the average is allowed for in contingency fuel.

    The average they quote is 18,580 BTU, 43.22 MJ/kKg. In the below report (1983), p61, they note the maximum as 43.45 (of Jet A and A-1, which have the same minima), the average being 43.12 so close to Boeing. Note though that the difference from average to minimum is just 0.76% and that might be too small to intrude much into your investigations.

    http://www.dtic.mil/dtic/tr/fulltext/u2/a132106.pdf

  97. Oleksandr says:

    Dennis,

    “I had not heard about the classification of the pieces you mentioned as “highly likely”. ”

    The classification was introduced by the Malaysians or ATSB in their debris reports:

    – Unidentifiable (or unidentified);
    – Likely;
    – Highly likely;
    – Almost certain;
    – Confirmed.

    If you had a chance to take a look at my plots, you could notice that I marked all the pieces according to their classification.

    Your wrote to Ge Rijn: “It has been about 10 months since any confirmed or highly likely debris has been found. Given the rate of confirmed or highly likely debris finds in the year following the flaperon finding, it seems unlikely that much if any more debris will be found.”

    I think a lot of debris can be found at the eastern shore of Madagascar. What BG found there is only 5% of what can be found there. Unfortunately, nobody is searching there. Even the famous CSIRO’s study is discarded in this regard. The other ‘hot spot’ is in Tanzania, neat the Mafia Island. I think one of the fragments found by Paul close to that area could be from MH370.

    In other words your WB curve is affected by human actions, or their absence.

  98. Oleksandr says:

    @Dennis:
    @Brock:

    You asked Mick Gilbert:

    “Given the flight path recorded by radar and compatible with satellite data from IGARI to 18:25 is there any doubt in your mind the someone was piloting the aircraft?”

    and

    “can you please provide specific precedents for comms being knocked out as totally and for as long as the official narrative requires. “Significant communication failure” just won’t get the job done”.

    I don’t know why both of you still consider only black and white, and fail to consider anything in between. What if hypoxia affected the passengers and F.O. at around 18:00, while Z. and/or a few other cabin crew members survived to approximately 19:00? Dennis, is there any doubt in your mind that nobody was piloting the aircraft after 19:40?

    Brock, google on SwissAir 111.

  99. Ge Rijn says:

    @Mick Gilbert

    Thanks for the story and the link. They even have a seperate WA-beach clean up department and they collect debris since 2004 on the most remote beaches too like you say. And every find gets cataloged.
    Their database is not excessable by me but I guess they did not find obvious airplane debris yet for I guess they will report it.
    On their way of working:

    http://www.tangaroablue.org/about-us/our-solutions.html

  100. Ge Rijn says:

    @DennisW

    @ALSM posted the latest MOT debris report some days ago:

    ‘Here is an MOT presentation given yesterday to Chinese NOK:
    https://goo.gl/nED3O8

    You can find the pieces there.

  101. Ge Rijn says:

    @DennisW

    It’s on the final pages. The no.7 flap track fairing was actually found 27 januari 2017.

  102. Ge Rijn says:

    @VictorI

    I ran your 35S particular file again and again.
    I’ve come to the suggestion/conclusion your graphic example in your topic much better fits the situation on 31-12-2014.
    Could it be you’ve been mistaken by a year?

  103. Andrew says:

    @David
    @TBill

    RE: “Have you noticed differences please in fuel consumption between different fuels on long trips or an airline fuel sourcing based on that?”

    We have an ongoing issue with westbound ultra-long haul flights (14 hours plus) using more fuel than predicted by the computer flight plans. Variations in fuel calorific value contribute to the problem, but only account for an increase of around 0.3-0.5% of the total fuel burn, whereas the observed increase is around 1.7-2.0%. There’s an ongoing study to determine the causes, but it’s believed that a difference between the logic used by the FMC to determine the ECON cruise speed and that used by the computer flight planning system also contributes to the problem, especially when flying into strong headwinds. Note that this issue has only been observed on ultra-long haul flights, where a small percentage increase in the total fuel burn equates to quite a large increase in the amount of fuel used. On shorter flight of 7 hours or less, the variation is not likely to be significant.

  104. David says:

    @Andrew. Thank you. Interesting.
    @TBill. “Of course, we can use the standard assumed energy content, but I would ask what actual data we have? There can be variations in jet fuel quality within the specs”. Last sentence and 0.3-0.5% should help with that.

  105. Ge Rijn says:

    @Mick Gilbert

    I read your comments on Ben Sandilands site about your 22 april post.
    I started to firts question the CSIRO-study with same arguments on 21 april on @VictorI’s previous topic with follow-ups the next day.

    Is this a case of synchronicity?

  106. Mick Gilbert says:

    @Ge Rijn

    Probably not synchronicity given the causal relationship. I think that it is simply the case that we all read the same paper at around the same time and saw the same flaws.

  107. Victor Iannello says:

    @Ge Rijn: I can’t explain your result regarding the arrival of low windage debris at Mossel Bay, South Africa. If you look at the time slider in the image, the date is 12/31/2015. Others using the same KMZ file have come to same conclusion as me.

  108. Victor Iannello says:

    Update 3 on April 26, 2017.

    I received the following email from David Griffin of CSIRO:

    Dear Victor,

    I saw the Guardian article referring to your blog. A few comments:

    1. You are correct that ‘Roy’ was found at an earlier date than the model predicted. But to be fair, the model error is ‘just’ 2 months. I consider Roy’s arrival time – before anything else upstream –  to be something that is simply too hard for any present-day model to convincingly explain. You’ve seen the paths that things take. But those paths should not be interpreted too literally. Our Dec 2016 report mentions that we do not have confidence in the model’s ability to hindcast the arrival times of individual items along the African shore. That’s why we focussed on the more-robust things that the model tells us.

    2. As you correctly pointed out, a 30S crash site would, according to our model, have resulted in debris washing up on Madagascan and Tanzanian shores a full year earlier than was observed. That is a discrepancy that is hard to set aside.

    3. The other factor against 30S that we find very hard to discount is that 30S is right in the middle of the zone targeted most heavily by the surface search in 2014. This is the “other evidence” that Richard overlooked. Please see Section 4 of our Dec report, and Fig 4.2 of the April report. 

    Best regards
    David

  109. DennisW says:

    @Ge Rijn / Oleksandr

    Do either of you have an accurate (published by ATSB or Malaysians) reference for when debris in the “confirmed” and “highly likely” categories was found? I am not comfortable with any of us assigning dates or the classification of debris. Surely something this important must have an accurate official reference.

  110. Ge Rijn says:

    @Mick Gilbert

    Yes ofcourse.
    Anyway I still agree the CSIRO 31S to 30.25 segment pictured in Godfrey’s paper still makes the best fit for all debris (0,10 and 20 degrees). With only one or no landing on WA.
    The 35S segment only makes a good fit for the 20 degree flaperon debris. With 7 landings on WA for the 0 till 10 degrees debris. Which we know nothing has been found there still.
    So in this way the CSIRO-conclusion contradicts their own data most significantly but IMO (unexpected?) their data helps to build a strong support for the 31S/30.25S segment.

  111. Ge Rijn says:

    @VictorI

    I’ll try to post my result on 35S. Then something must go wrong running it on my computer.
    I already thought it was strange my runs did not show a lot of variation in beachings on the African coast between 39S and 29S.

    @DennisW

    I posted the MOT-link from @ALSM with the debris-items mentioned today at 1.45AM (few posts back).

  112. DennisW says:

    @Ge Rijn

    Yes, I saw that. Unfortunately you had to fabricate a date (you also mis-referenced the item number). That is what actually prompted my question regarding an official accurate document.

  113. Ge Rijn says:

    @DennisW

    I mentioned the no.7 flap track fairing piece and corrected the date 27 januari 2017. And I posted the official MOT-report again.
    I see nothing wrong with that.
    In Holland we say in such a case; ‘don’t look a given horse in the mouth’.

  114. DennisW says:

    @Ge Rijn

    I am not complaining. Just pointing out that the report you reference contains the flap track fairing as item #8 with a date in May 2016. I think your information is correct, but I am not going to base any updates on an unofficial source of info. It is called hearsay. Basically you don’t have an accurate official document you can reference.

  115. TBill says:

    @David @Andrew
    Re: Jet Fuel
    I am under the impression that when an accident happens, checking of jet fuel quality is supposed to be the first step, to make sure the next flights are OK. Admittedly that might not be the exact fuel load on MH370, but should be close. If that fuel quality check step was taken, I would like to see the result. If it was not taken, why not? Did someone know the flight was flying normally?

  116. Ge Rijn says:

    @DennisW

    Just look before you leap and scroll to the end of the MOT-report I (and @ALSM) linked and referre to. It’s on the final pages the last two items.

  117. Ge Rijn says:

    @DennisW

    And to be complete the items are No. 26 and No.27.

  118. DennisW says:

    @Ge Rijn

    Got it. Thanks. I did look before I leapt. Unfortunately your reference to item #7 was incorrect. I am not clairvoyant.

  119. Oleksandr says:

    @Dennis,

    “Do either of you have an accurate (published by ATSB or Malaysians) reference for when debris in the “confirmed” and “highly likely” categories was found?”

    Download summary pdf from:

    http://www.mh370.gov.my/index.php/en/425-summary-of-possible-mh370-debris-recovered-28-feb-2018

    ATSB has also published “expanded” version.

  120. Oleksandr says:

    @Victor,

    Do we have accurate mapping of the complete aerial search? Preferably in shp format, or, at least ‘clear’ plot (not overlaid with other stuff)?

  121. Ge Rijn says:

    @DennisW

    My reference to No.7 flap track fairing (which it is) was confusing I understand now.
    Never mind, we are here to clear up things isn’t it. Not to take things for granted.

  122. Victor Iannello says:

    @Oleksandr: I don’t, but I know somebody that is interrogating CSIRO on this very matter. Mostly, we have maps from press releases.

  123. Ge Rijn says:

    On Griffin’s comment I like to mention the aerial search argument on the ~30S area is a kind of negative argument for there was nothing found in any aerial search anywhere. And the non-aerial searched area around 35S was sonar-scanned without results.

    Further, the Roy-piece was full of large barnacles when first found in december 2015. Not young barnacles but at least 1 year old ones.
    The piece must have arrived there not that long (at most few weeks) before december 2015 spending its time on the ocean before that uninterupted. A lot of barnacles were still attached (although rotting according to the first finder).
    It was picked/eroded clean when it was found in march 2016 again.

    This piece therefore serves as a rather specific time marker IMO, even more than the flaperon.

    Therefore any drift study must have the data that make it possible to include the Roy-piece. The CSIRO-study in fact does clearly but not on 35S.

  124. Ge Rijn says:

    To add a comment on another Griffin’s comment that debris must have landed a full year earlier on Madagskar and Tanzanian shores coming from ~30S seems unlikely.

    I argue if we take the Roy-piece as a time marker at december 2015 being the most far drifted piece full of barnacles than most other pieces arrived well before that time but were only found much later.

    For instance the Tanzania Pemba outboard flap was full of oong time wear, tear, metal corrosion and signs of gone bio-fauling like many other pieces.
    There’s know way to know now for sure but the Pemba piece and many other pieces show signs of having arrived there long before the finding time and long before the Roy-piece.

  125. Ge Rijn says:

    Excusse my Englisch typing is sometimes slower than my Dutch thoughts..

  126. TBill says:

    @buyerninety
    “Aviate, Navigate, Communicate.”

    This policy is part of the problem. MH370 pilot had the freedom to secretly turn off all systems and fly off to the SIO, with lack of communication allowed by the industry. Then someone tries to tell me we cannot assume the pilot did it due to lack of proof. Of course there is a lack of proof.

  127. Perfect Storm says:

    The CSIRO report says:

    « Our earlier field testing of replicas of the flaperon was unable to confirm numerical predictions by the Direction Generale de L’Armement (DGA) that the flaperon drifted left of the wind. Field testing of a genuine Boeing 777 flaperon cut down to match photographs of 9M-MRO’s flaperon has now largely confirmed the DGA predictions, at least with respect to drift angle. The impact of this information on simulated trajectories across the Indian Ocean is that the July 2015 arrival time at La Reunion is now very easy to explain. »

    Why did they use replicas in the first place (instead of a genuine flaperon) ?

    And why wasn’t the MH370 flaperon used, that was found on Réunion ? Yes, it’s a piece of evidence, but they had 1 year to examine it, so I guess they have already done all the work on it. And putting it into the water, let it drift for a while and recovering it later won’t damage it anyway.

  128. lkr says:

    @Victor

    David Griffin’s reasoning about time of arrival in Africa is so flawed by the [seemingly] random pattern of discovery. Not so strangely, the early pieces [like ‘Roy’] are picked up by curious kids whose families have boat trailers. Not so coincidentally at resorts populated by first-world folks with leisure to be curious and some knowledge of first-world concerns like missing aircraft. Some critical debris like the Pemba flaperon and probably the original Blaine Gibson piece seems to have been spotted months earlier by fishermen. And the late Gibson swarm in Madagascar are fouling-free and worn — probably sitting up like cockle-shells for months if not years.

    So ATSB and all the other actors had absolutely no interest over 3 years in ANY organized scheme to collect debris [even the simplest press release/outreach to officials in targeted areas. Or simply someone emailing every resort in the Indian Ocean with pix of debris that could be shown to staff and posted to visitors, with a reporting contact…. THAT might have generated some basis for time-of-arrival. What Griffith refers to as ‘data’ is almost entirely useless for his argument.

  129. Victor Iannello says:

    @Ge Rijn, @Ikr, and others

    My main concern is that researchers are using CSIRO’s results to proclaim with a high level of confidence that the crash site was along the 7th arc at 35S latitude. CSIRO’s results show that crash sites further north would better explain the arrival of “Roy” in Mossel Bay as well as the lack of debris (both high and low windage) found along the coast of Western Australia. (I do commend CSIRO for providing access to the drift model results in a form that others can easily use.)

    As far as debris arriving in Africa too early for a 30S crash site, I agree with others that debris might have arrived but was not discovered until later. For instance, I don’t believe Blaine Gibson found “No Step” just after it arrived to the shores of Mozambique. Likely, it and many other parts arrived much earlier and were either not found or found and not reported. The fact that he found it soon after looking suggests that there were likely many more parts.

    As for the aerial search, I question the 100% efficiency of those efforts based on the comments of others with more SAR knowledge than me. I also see that the stated probability of discovery is not 100% as you proceed north along the arc, including holes in the coverage around 30S and a fall-off of coverage north of 27S.

    The seabed near 35S was also searched to a distance of 20 NM. The final BFO value indicates the plane was in a steep descent near the 7th arc. I believe that if the plane crashed near 35S, it likely would have been found.

    I cannot say with absolute certainty that the plane did not crash at 35S. The purpose of my post was to question the confidence level expressed by CSIRO and the ATSB and to invite others to independently review and evaluate the results.

  130. Ge Rijn says:

    @VictorI

    This CSIRO-study effectually discards its own conclusions about 35S and in fact supports a ~31S/30S segment. I see no way you can interpretate their data otherwise.

    No one can say with absolute certainty where the plane crashed but according the latest CSIRO-data the plane most probably did not crash near 35S.

  131. Ge Rijn says:

    @VictorI

    ‘Likely, it and many other parts arrived much later and were either not found or found and not reported.’

    I suppose you mean; ‘many other parts arrived much earlier and were not found or not found and reported’.

  132. Victor Iannello says:

    @Ge Rijn: Yes. Corrected.

  133. Ge Rijn says:

    @Perfect Storm

    The ATSB/CSIRO did not have the original flaperon available so they did the best the could on the flaperon now.

    But I just wonder why they did not use the Pemba outbourd flap section which is completly in their hands and probably acts like the flaperon too on high windage.

    In fact I think this a question only to answer by CSIRO/Griffin/ATSB.

  134. Richard says:

    @David Griffin

    “The other factor against 30S that we find very hard to discount is that 30S is right in the middle of the zone targeted most heavily by the surface search in 2014. This is the “other evidence” that Richard overlooked. Please see Section 4 of our Dec report, and Fig 4.2 of the April report.”

    My apologies for overlooking Section 4 of your December report and Figure 4.2 of the April report in this context.

    It was not obvious that the comment in brackets in Section 3 referred to a figure in Section 4 about High Windage Items in the April report and Section 4 of the previous December report. In fact you state in Section 3 “assuming the aircraft had crashed near the 36°S-32°S segment of the 7th arc (the zone most consistent with other lines of evidence)” without any reference to either other Sections of the present or previous report.

    The aerial search conducted in the 30°S drift region 20 days after the MH370 incident was less than thorough. My colleague Don Thompson calculates the area covered in six days, 28th March through 3rd April, twenty days after impact with the sea, was approx 555 x 1,111km, 616,605km². Ten aircraft, possibly five hours on task each day, six days: total 300 hours on task, to search 2,055km² per hour. That is a tall task even if the weather and visibility hold for the duration, which it did not at least for 1st April 2104. Please publish the actual areas searched (as opposed to the areas planned) in the vicinity of the 30°S drift region, that compel you to discount this possibility.

    But my main question is: What is your level of confidence in the drift data please?

    Given the number of storms in the Southern Indian Ocean, which may have affected the trajectory of MH370 floating debris items, and the number of days each storm lasted, it could add up to +/- 30 days on any arrival date of a particular debris item from Rodrigues to Mozambique, Tanzania to South Africa.

  135. ALSM says:

    Radio Interview with David Griffin: https://goo.gl/p7juPH

  136. DennisW says:

    @Oleksandr

    “Dennis, is there any doubt in your mind that nobody was piloting the aircraft after 19:40?”

    The information we have does not lead to the conclusion that the plane was not piloted at any time. The flight path suggests the aircraft was piloted between IGARI and past Penang. Someone controlled the FMT. I find the explanation for a communication failure to be very far fetched as would any reasonable person.

    The statistics for hull losses overwhelmingly support human intervention. The simulator data is a slam dunk for the deliberate diversion.

  137. Paul Smithson says:

    @Andrew, David et al

    Thanks to all for your comments on the scenario of SDU re-boot prompted by R engine fuel exhaustion rather than left. On the basis that A/P would disengage due to removal of pitot heat this doesn’t look like a runner for the aircraft to continue flying straight.

    What about a variation as follows: Left IDG shut down but Left backup running. Is it the case that the lower-capacity L back-up will not run the SDU? And if that is the case, could we not have a sequence as follows:

    – Right engine flame-out, with Left backup running; SDU powers down; pitot heat still available and A/P remains engaged
    – APU startup, SDU re-boot/logon
    – Plane continues in powered flight, with A/P, for as long as left tank fuel remains.

    And to take that further: if APU runs even a very short time beyond L flame-out, it provides the power and hydraulics to neutralise the asymmetric trim associated with single engine operation. So that once APU shuts down we have a plane with neutral trim, now in “secondary mode” with limited envelope protection – and therefore the potential for a long and straight-ish glide.

  138. Victor Iannello says:

    CSIRO believes the plane can be located with impressive precision because of a narrow band of westerly flow near 35S along the 7th arc. If you study the path of particles starting at 35S, the flow is indeed west initially, but then reverses itself towards Australia, crosses the 7th arc in an easterly direction, and should result in some debris on the shores of WA. CSIRO ignores this despite the fact that their own model predicts this. We also know that CSIRO model does not predict the discovery of Roy in Mossel Bay.

    The main reason for CSIRO’s belief that 35S is more probable than 30S is that 30S was searched from the air without success. The efficiency of the aerial search might be the key to reconciling the results. I’m glad that Don is hot on that trail.

  139. DennisW says:

    @VictorI

    “The main reason for CSIRO’s belief that 35S is more probable than 30S is that 30S was searched from the air without success.”

    That is a stated reason, but the underlying driver for 35S is to not create a significant incompatibility with their buddies in the DSTG. You don’t get far tossing colleagues under the bus when you are sucking on a government tit.

  140. Victor Iannello says:

    @DennisW: Yes, “We’re confident that it’s just beyond what we’ve already searched” has been used before to extend the search area. My guess is we’ll see non-Australian scientists that are active in this area state their own findings and recommendations. I think the debate is healthy.

  141. Oleksandr says:

    @Dennis,

    Re: “The information we have does not lead to the conclusion that the plane was not piloted at any time. The flight path suggests the aircraft was piloted between IGARI and past Penang. Someone controlled the FMT. I find the explanation for a communication failure to be very far fetched as would any reasonable person.”

    Read my comment again. All the things you listed occurred before 19:40. If you have any specific reason to believe that the plane was actively piloted after 19:40, please name it.

    Re: “the underlying driver for 35S is to not create a significant incompatibility with their buddies in the DSTG.”

    No doubts. I would also add justification of the extensive search at 38S.

  142. DennisW says:

    @Oleksandr

    I am basing my statements on the fact that there is absolutely no reason to believe the plane was not actively piloted. The no pilot scenario requires an elaborate set of circumstances to support. None of which are anything but speculation. Why should I think the plane was not actively piloted? That is what pilots do.

  143. TBill says:

    To some extent I support the ATSB idea to search 32.5-36 S area:
    (1) We do not really know where the plane is so we have to start checking off the boxes.
    (2) It fits DrB’s path, which I consider to be possibly where the concentrated search should have been centered in the first place. Seems to fit BTO/BFO well. I can argue intent to get over Broken Ridge at 31S into deeper water. I can argue intent to go approx. due south heading by a variety of A/P modes (waypoints or just the pilot looking at Longitude like 93S and adjusting to aircraft keep somewhat on the longitude line, etc)

    At some point, if not found, obviously we have some wrong assumptions:
    > Lack of debris sighting is possibly wrong reason to discount 25-35 S
    > Arc7 is possibly not the end point
    > Overall possible intent to hide plane and minimize debris somehow (I have various underwater hiding place theories and one debris minimization theory)

  144. TBill says:

    The only reason to assume the aircraft was unpiloted is because ATSB seemed to make that assumption, and it seems possible ATSB could have secret info to make that assumption, otherwise, why make that assumption? If we had that answer, we might not like it.

    DrB’s path assumes a slow down, which makes sense to me. But I can easily see the speed change later than 1840 and even 1940.

  145. Paul Smithson says:

    @TBill the chief reason IMO to believe “no pilot input” from 18xx is that the BTO and BFO pattern suggest a smooth,~ straight, ~constant speed trajectory. What’s not to like? If the BTO pattern suggests “straight” then it is surely straining credulity that a comparable pattern just happened to have been produced by a “bendy path”. Those path models that correspond with no turn/no speed-change take you to 38S. And I still believe that they were right – but that the aircraft went further.

  146. TBill says:

    “What’s not to like?”

    Because straight path is unlikely answer to the Ghost flight riddle. Jet aircraft do not usually fly straight paths. They fly magnetic curvy paths or waypoints. A straight path automatically assumes an active pilot at the helm dialing in waypoint or a True straight path. Therefore it is illogical to assume a ghost flight on a straight path. Something does not add up.

  147. Paul Smithson says:

    @TBill my current thinking is a constant true heading arising from a route discontinuity. I’m aware that Andrew has tested and found that CMH rather than CTH follows discontinuity. Victor’s PMDG sim, I believe, indicated CTH. Honeywell’s documentation/responses inconclusive.

  148. TBill says:

    @Paul
    Really strange Boeing or ATSB does not at least throw us a bone on this basic question.

  149. ROB says:

    @TBill

    You are talking rubbish, mate! Planes do indeed fly straight paths – ask any airline pilot. My original explanation for the events of March 7th/8th remain completely valid. The recent, illogical and unscientific report from CSIRO is unintentionally serving to back me up. Can’t you work out what’s going on? CSIRO and ATSB have got together to push S35 to the fore. “That’s where it was all along” they cry, but were persuaded to abandon that area early on, and concentrate on the region further south, ie. DSTG Bayesian analysis area. So, once more for luck, this is what happened – under pilot control, the plane turns at IGOGU and flies south in a straight line until fuel exhaustion, the pilot then dives in order to build up momentum, and glides as far south as possible, before entering a stall to make the plane belly-flop into the ocean and sink. CSIRO hade already established that drift analysis allows S38 to S40. But the big big problem with S38 to S40 is that it implies a sentient pilot at the controls, carrying out an evil, pre-planned mission of murder/suicide, which was a definite no-no as far as the Malaysians were concerned (and probably the Chinese were concerned, as well) so to salvage some credibility, and save face, the ATSB and CSIRO concocted S35 as the likely drop zone, knowing full well that the search will never be resumed. Smart, eh?

  150. TBill says:

    @ROB
    We are in violent agreement at least in principle of intentional diversion.
    I was saying ghost flights do not fly straight paths unless pre-programmed to do so. Wasn’t your path close to 35S end point or was it closer to 38S?

    @Paul
    I believe in the 1st Principles review, ATSB indicated curved magnetic might be more expected. That would be consistent with Andrew’s work on the real B777 simulator re: discontinuity magnetic heading.

  151. TBill says:

    PS ROB- I am looking for someone who thinks more like me on intentional “pre-planned mission” and maybe you are the one…

  152. Victor Iannello says:

    @Paul Smithson: The best evidence we have, based on the careful testing done by @Andrew in a Level D simulator, is that a B777 follows a CMH path after a route discontinuity.

  153. Victor Iannello says:

    @Rob said, “the pilot then dives in order to build up momentum, and glides as far south as possible, before entering a stall to make the plane belly-flop into the ocean and sink.”

    If a pilot wanted to glide as far a possible, they would not “dive to build momentum”. Rather, they would gradually descend at about a 3 deg descent angle, where the (L/D) ratio is maximized. Similarly, they would not be “entering a stall to make the plane belly-flop into the ocean and sink”. A stall would put the plane in a configuration that is hard to control, and would require nose down to recover. Rather, they would extend flaps and reduce speed while maintaining some pitch up and lift.

  154. Oleksandr says:

    @Dennis,

    “Why should I think the plane was not actively piloted?”.

    Many reasons, for example:
    (a) The trend in the BTO & BFO after 19:40 suggesting a simple path.
    (b) Not answered SATCOM calls 18:40 and 23:15.
    (c) Never activated cabin ELT.
    (d) No sms/e-mails received from the cabin.
    (e) Absence of any motive to bring the plane to ~30S.
    (f) Possibility to unlock the cockpit door by isolating L DC bus from the EE-Bay.
    (g) No ELT signal after the crash (or ditching, or whatever you prefer).

    You may explain a few of these points, but not all of them at a time. I think we have been chewing this for nearly 3 years.

    But you have not answered my question. What does make you thinking that the plane was actively piloted after 19:40? Any reason?

  155. DennisW says:

    @Oleksandr

    “But you have not answered my question. What does make you thinking that the plane was actively piloted after 19:40? Any reason?”

    I don’t have a reason other than gut feel. I don’t think there is any hard data to support a pilot or a no pilot scenario.

  156. Andrew says:

    @Paul Smithson

    RE: “What about a variation as follows: Left IDG shut down but Left backup running. Is it the case that the lower-capacity L back-up will not run the SDU? And if that is the case, could we not have a sequence as follows:

    – Right engine flame-out, with Left backup running; SDU powers down; pitot heat still available and A/P remains engaged
    – APU startup, SDU re-boot/logon
    – Plane continues in powered flight, with A/P, for as long as left tank fuel remains.”

    It’s not a case of insufficient generator capacity. The SDU is powered by the L Main AC bus, but the backup generators can only power the L & R transfer buses. If the L backup generator is the only power source, then the L Main AC bus (& the SDU) will not be powered.

    In the scenario you outlined above, the L backup generator will continue to power the transfer buses until the L engine flames out. Consequently, there would be no auto-start of the APU until after L engine flame-out.

  157. Paul Smithson says:

    Thanks @Andrew.

    In short, can you not see any scenario whereby the SDU power cycle could be linked to R flame out and the aircraft carries on flying under AP under L engine power?

    Secondly, what are your views on the possibility that the APU could run a few seconds longer than the engine (lower fuel demand, longer line?) – sufficient to neutralise the asymmetric trim that would have prevailed under one-engine inop? And if this could occur, would the aircraft have a decent chance of flying straight(ish) under the limited envelope protection remaining?

  158. Paul Onions says:

    @Paul Smithson

    Read my post on April 25.

    The aircraft can be beyond 40 nautical miles from the Bayesian hotspot.

  159. Paul Smithson says:

    @Paul Onions. I did read it. I share your interest in a location beyond the hotspot. I did not understand the detail of your Example 1, or how it could result in controlled flight beyond the 7th arc. I’d also be glad to hear where you got your xx minutes xx seconds estimate (beyond the arc) from.

  160. TBill says:

    @Andrew
    Your westerly flight excess fuel consumption observation is interesting. Of course as you say, that could be several issues. Presumably it’s a model issue, or maybe differences in temp/conditions between east/west flights. Brainstorming for the moment, lets say possible regional differences in jet fuel quality above and beyond the LHV correction term, not sure what, but let’s say low temp flow differences (if that is even possible). Then one approach would be keep records and grab jet fuel samples at the airports that display the differences, and do some lab testing of various parameters, maybe flow rates at low temps and so on.

    Probably however starting point would be to gather flight parameter data to see what correlates. The first thing is obviously probably more fuel weight and greater flight time on the westbound flights. Then I might look at airspeed, fuel temp, outside temp…not sure if airlines can data log key parameters over a flight.

  161. Brock McEwen says:

    Re: US assets near IGARI/IGOGU, night of Mar. 7, 2014: to prime the pump, let me ask an even easier question: what was the specific training exercise from which the USS Pinckney was called away to assist the search? Even that simple and straightforward snippet of data would help get us started.

    (No, I still have not read Cawthorne’s book. I have vowed never to spend – or take, for that matter – a penny in exchange for MH370 research.)

  162. David says:

    @Victor.”My guess is we’ll see non-Australian scientists that are active in this area state their own findings and recommendations”.

    I doubt that that the science membership of the Search Strategy Working Group is exclusively Australian or those in the Review likewise?

  163. Victor Iannello says:

    @David: What I mean is that oceanographers from other countries that are not associated with the official investigation have already weighed-in with their own drift analyses. With the new CSIRO report published, there is an opportunity for scientific debate.

  164. David says:

    @Victor. Thanks. My surprise was at your response to DennisW’s scurrilous remarks.

  165. DennisW says:

    @David

    “My surprise was at your response to DennisW’s scurrilous remarks.”

    Nothing scurrilous at all. Just the truth. The ATSB and their subcontractors are pathetic by any measure.

  166. David says:

    @DennisW.”….the underlying driver for 35S is to not create a significant incompatibility with their buddies in the DSTG. You don’t get far tossing colleagues under the bus when you are sucking on a government tit”.

    Scurrilous.

    Glock them all.

  167. Paul Onions says:

    @ Paul Smithson

    After the left engine stops due to fuel exhaustion, there is 30 lbs of fuel available for the APU resulting in a maximum duration of 13 minutes 45 seconds. See ATSB report – Definition of Underwater Search Areas 03 December 2016.

    In example 1 (APU on prior to hypoxia), there is “controlled” flight after the seventh arc because the autopilot is still engaged.

    1. APU turned on by crew prior to hypoxia.

    2. One engine fails due to fuel exhaustion (Right engine has greater fuel burn rate, but APU is using fuel from Left tank. Or one engine has less fuel flow due to a failed generator or bleed air). If Right engine fails first, APU powers right Main AC bus. Then as left engine fails, APU powers Left and Right main AC busses. The power transfer from the failing Left IDG (because the left engine is spooling down) to APU results in a break power transfer of the Left Main AC bus. Therefore the Satcom is momentarily depowered and then repowered. The logon sequence commences. The autopilot is still engaged.

    3. With both engines failed the aircraft slows. The autopilot maintains LNAV (heading), and VNAV (altitude) by pitching the nose up. Eventually the aircraft reaches the stall speed. This is where Flight envelope protection overrides the autopilot VNAV and pitches the nose down to maintain flying speed. Note LNAV is not overridden. The aircraft descends at 4,000 feet/minute. This is where the logon request is sent.

    4. Now that the aircraft has regained flying speed, flight envelope protection releases and the autopilot resumes VNAV. The aircraft is now below FL350 so the autopilot pitches up the nose to try and regain altitude. Speed reduces towards stall speed. Flight envelope protection again overrides VNAV and the process repeats. The aircraft continues to head south.

    5. The 00:19:37 BFO should be discounted as stated by Inmarsat in their Journal Of Navigation.

    6. To reduce load on the only available generator (the APU), ELMS sheds the IFE.

    7. Eventually the APU stops due to fuel exhaustion (13 min 45 sec).

    8. All the AC busses are now depowered. The autopilot disengages and the aircraft is out of control beyond 40 nautical miles of the arc to the south.

  168. David says:

    @Rob. I for one remain agnostic about a piloted end but maybe a better reason for one to dive at altitude would be loss of cabin pressurisation at fuel exhaustion. Aside from oxygen, cold.

  169. Brock McEwen says:

    @George: thanks for the reply. Yes, I’d already read those sources thoroughly before asking my question. Which is: what was the NAME of the training mission it was on?

    Here is a listing of naval training exercises (from the source you provided):

    http://www.public.navy.mil/surfor/Pages/TrainingExerciseEvents.aspx

    Was it one of these? If so, which one? And regardless: where, specifically, was the Pinckney on that fateful night?

    We can narrow it down a bit if we learn the exact time it started heading to the coast of Vietnam and the precise moment it arrived there. It is hard to do this when website timestamps don’t include a time zone (all sourcing suggestions warmly welcomed), but if the Pinckney was already swapping out a search crew by 9am on March 9 (per the famous heli photo caption), it seems like the transit time had to have been somewhere between “no time flat” and “a scant few hours”.

    When we’re done with the Pinckney, we will move on to all the rest of the US primary radar assets within range that night.

  170. DennisW says:

    @David

    Just the truth. Plain and simple.

  171. David says:

    @Rob. Of course he might have pointed the aircraft down as others have. In that case though the search area should remain within its twenty miles, that radius 4 times the altitude.

  172. David says:

    @DennisW. We do aim for some integrity down here you know though its not easy, what with having to hang on against gravity by our toes and fly upside down.

    Despite all that you will see from Jack the Insider’s analysis in The Australian today, plus selected comments, that we do worry away at what our values are:

    “I’ve said it many times before. Australia is the best damned country in the world with the exception of some island nations in the Caribbean with some very liberal banking laws.

    For all that Australia can be a confusing place, especially for Australians. There are three tiers of government, all pretty much worthless and in various stages of decrepitude shuffling between inertia, chaos all the way along to abject failure. We have a corporate world blagging its way around a laughably cobbled together regulatory system while trying not to snigger too much. There’s a mutant media that routinely crucifies people, more often that not for no apparent good reason and a taxation system that, frankly, I gave up on a long time ago.

    Last week the Turnbull government announced a range of changes to the visa scheme and rounded it off with tightening requirements for citizenship. At a presser and then again in an interview with Leigh Sales on the ABC, our Prime Minister, resplendent in an electric blue suit, equivocated in response to what were some fairly mild inquiries on his thoughts on Australian-ness and what he considered might be Australian values.

    To be fair to Turnbull the answers are vexed and perhaps, sensing a trap, he felt obliged to stammer out platitudes.

    I appreciate stereotyping saves time and this is important in our busy lives. If we are to lapse into laziness and go down the Family Feud path as a means of understanding the national character, surveying 100 responses, the three top responses are mateship, fairness and egalitarianism.

    See, I would have included alcoholism, too, received the dreaded electronic blurt of failure and missed out on the lounge suite and the hi-fi system.

    The trouble is the three top answers could be seen as roughly appropriated from the French national motto, “Libertè, Egalitè, Fraternitè” (Liberty, Equality, Brotherhood) and the French are bound to take umbrage and tie us up in the courts for years over breach of copyright or some other EU enshrined legal nonsense.

    Looking elsewhere, the motto of the United States is “E pluribus unum”, (In God We Trust), which places Americans’ faith in a supernatural being. And this, when viewed in the context of the current leadership of their country, makes perfect sense.

    But the words that more aptly define American values come from the Declaration of Independence — “Life, Liberty and the Pursuit of Happiness”, which usually means getting off the couch and taking a trip to the local Arby’s on all you can eat night. It could just as well read Life, Obesity and Type 2 Diabetes.

    The difference is Americans have to pursue happiness but in Australia, it’s right here wherever we chose to be.

    The French may have dibs on equality in their motto but they rarely practice what they preach.

    Their language, le Francais, is replete with formality and confusing terms of address.

    No one calls anyone ‘Sir’ in Australia unless they go to expensive private schools or are being tapped on the shoulder by security personnel on licensed premises where the honorific is often accompanied with “You are projectile vomiting on my shoes.”

    Australia has no official motto. Unofficially, it is Advance Australia, a pithy saying lifted from the national anthem often embroidered on cricketers’ shirts next to a beer logo.

    When we contemplate Australian values the best starting point is the ten immutable precepts of Australian-ness:
    1. You want it, when?
    2. Maaaaaaaate
    3. Yeah, nah
    4. Win the toss and bat
    5. No shoes, no shirt, no service
    6. It was like that when I got here
    7. Mate, mate, MATE …
    8. Eh?
    9. She’ll be right
    10. The brassco’s second on the right

    The long answer is if you’re standing on a piece of dirt anywhere between the Indian and the Pacific Ocean somewhere south of Indonesia, you can take it for granted you’re Australian and you can feel, think, do and say whatever you bloody well like.

    It might seem obvious but being Australian is really little more than being able to recite your postcode. If it’s anywhere between 2000 and 9999, you’re Australian and you understand this because Australia Post knows how to track you down, although it could take weeks and sometimes months. You are here and therefore not over there. That’s what makes you Australian.

    Everything we can see, hear, feel, eat and drink is Australian. There are no hordes amassing on the borders to our north, east, west or south because there are no borders. Australia is basically a great, stonking land mass stuck in the middle of nowhere and it is only when you’re standing in it that you appreciate that fact.

    I think we should leave at it that, really, and avoid veering into the perilous practice of stereotyping 25 million people into the space of three or four words.

    In all seriousness, a nation founded on democratic principles, participatory democracy, universal suffrage, compulsory voting, the doctrine of the separation of powers, religious freedom, freedoms of movement, speech and association has got a good thing going.

    These are not vague conceptual elements. They define who we are. They make us who we are.

    It is ironic these institutions — the bricks and mortar of Australian-ness are often unloved and held in open contempt by so many Australians.

    The even greater irony is the man who announced changes to Australian citizenship requirements, the Prime Minister, is mentioned nowhere in the Australian Constitution. Nowhere. He gets a nod only in the broader grouping of ‘ministers for the crown’. This makes me often wonder why another minister doesn’t stride into the cabinet room, give the PM a whack behind the ear and slide into his chair and insist he will be chairing the meeting from now on. Indeed, this might explain why Malcolm won’t swallow his pride and give Tony Abbott a gig.

    In the meanwhile, if you really must ponder what it is to be Australian, you should start by thanking our founding fathers for creating a stern, uncompromising system of government and knocking up an equally inflexible constitution knowing as they did that if we started tinkering with it, we would almost certainly bugger it up.

    And when you’re asked about what Australian values are, the only correct answer is no one knows and it doesn’t matter anyway.
    JACK THE INSIDER
    ColumnistCanberra
    @JacktheInsider

    John 45 MINUTES AGO
    John Excreta taurae sapiente fulgari.

    Jack MODERATORTHE AUSTRALIAN42 MINUTES AGO
    @John Cryptic, John. Wise cooks steal fulgari. That’s according to a Google translator. As you’ve all figured out now, Latin is not my native language.

    Dwight 2 HOURS AGO
    Mum spoke latin, but rarely.

    Kevin 2 HOURS AGO
    E pluribus unum means Out of many, one. But you already knew that maaaate! And No. 5 is only half complete: Men: no shoes, no shirt , no service. Ladies: no shoes, no shirt, free entry.

    ———————————-
    Christopher 2 HOURS AGO
    My postcode starts with a 0 Jack. You’ve just excised the Northern Territory from Australia.

    Bloody unAustralian of you if you ask me. No rock for you.

    Jack MODERATORTHE AUSTRALIAN2 HOURS AGO
    @Christopher Learning all the time. That’s 200,000 people I’ve excised from Australia. My most sincere apologies to all Territorians.

    Martin 2 HOURS AGO
    @Jack Can you excise South Australia too?

  173. Mick Gilbert says:

    @David

    And before the hue and cry, it has been pointed out to the author of that column that “E pluribus unum” is not Latin for “In God We Trust”; he managed to mix up the old Latin motto with the current English one. A few of us down here are old enough to have been taught Latin at school.

  174. David says:

    @Mick, Yes it was pointed out copiousy and a couple gave the history but glad you mentioned it.

  175. Andrew says:

    @Paul Smithson

    RE: “In short, can you not see any scenario whereby the SDU power cycle could be linked to R flame out and the aircraft carries on flying under AP under L engine power?”

    Paul Onions suggested a scenario where a power transfer from the L IDG to the APU caused a momentary power interruption that caused the SDU to log on again. That might be a possibility, but I don’t know if the power break would be long enough to cause another log on.

    “Secondly, what are your views on the possibility that the APU could run a few seconds longer than the engine (lower fuel demand, longer line?) – sufficient to neutralise the asymmetric trim that would have prevailed under one-engine inop? And if this could occur, would the aircraft have a decent chance of flying straight(ish) under the limited envelope protection remaining?”

    As Paul Onions said, the ATSB believes the APU could have continued to run for a maximum of 13 min 45 sec after left engine fuel exhaustion, depending on aircraft attitude and motion, etc. If the APU had been started manually and the transfer buses were continuously powered, then the PFCS would have remained in Normal Mode with the autopilot engaged. On left engine fuel exhaustion, the APU would continue providing power for a short time and any rudder trim inputs previously provided by the TAC would be neutralised.

  176. Andrew says:

    @TBill

    RE: “our westerly flight excess fuel consumption observation is interesting. Of course as you say, that could be several issues. Presumably it’s a model issue, or maybe differences in temp/conditions between east/west flights. Brainstorming for the moment, lets say possible regional differences in jet fuel quality above and beyond the LHV correction term, not sure what, but let’s say low temp flow differences (if that is even possible). Then one approach would be keep records and grab jet fuel samples at the airports that display the differences, and do some lab testing of various parameters, maybe flow rates at low temps and so on.

    Probably however starting point would be to gather flight parameter data to see what correlates. The first thing is obviously probably more fuel weight and greater flight time on the westbound flights. Then I might look at airspeed, fuel temp, outside temp…not sure if airlines can data log key parameters over a flight.”

    I’m not involved in the investigation, but I know that fuel samples from various ports have been analysed along with QAR data from the aircraft, in an effort to determine the cause of the problem. As far as I know, the jury is still out.

  177. Victor Iannello says:

    @DennisW and @David: You’ve both proven you can offend each other (and others at the same time), one with a short insult and the other with a long one. With that established, please move on.

  178. Victor Iannello says:

    @Andrew, @Paul Smithson, @Paul Onions

    The power interruption when the source of power to the AC bus transfers from an IDG to the APU should occur in less than 200 ms. An interruption of power of up to 200 ms should not cause a log-on of the SDU. From the Honeywell manual for the SATCOM:

    Primary power interrupts of less than 5 milliseconds duration have no effect on system operation. For primary power interrupts longer than 5 milliseconds and up to 200 milliseconds duration, the LRU maintains normal operation, except the RF transmit and receive processes do not need to be supported in any LRU during this interval. For primary power interrupts greater than 200 milliseconds duration that causes normal operation to cease, the LRU performs a cold start following restoration of primary power.

  179. Andrew says:

    @Victor

    Thanks – QED!

  180. Oleksandr says:

    @Dennis,

    “I don’t have a reason other than gut feel. I don’t think there is any hard data to support a pilot or a no pilot scenario.”

    “Gut feel” is not an argument for me. There is no hard data, but some indications supporting no pilot scenario (after 19:xx). There is absolutely nothing, which would suggest a piloted scenario (after 19:xx).

  181. Victor Iannello says:

    @Oleksandr: I’d say the jury is still out on whether there was pilot input causing a steep descent at 00:19.

  182. Ge Rijn says:

    @VictorI

    I found the cause of the strange results I got with the KMZ-files in Google Earth. When I ran and saved a file and started the next file, GE ran the previous file together with the next file and so on. Resulting in all files running at the same time in the end.. So, this problem solved..

    Looking at it again I think finding debris on WA shores is crucial to give the CSIRO/Griffin conclusion much better proof they are in the right area.
    At 35S (and more south) till ~32S non-flap-drifters/debris landed on WA shores in respectable numbers.
    And I think the ‘fresh’ large barnacles on the Roy-piece don’t allow a 2 months off periode as with Griffin defends the failure of 35S to predict the Roy-piece.

    I suggest CSIRO conducts an initiative to search for debris on WA shores they have predicted debris would land.
    If only one ‘higly likely’ piece will be found it would make their case a lot stronger.

    Till now only their 31.75S till ~29S covers all debris with none or very few landings on WA.

    So, if they can provide the proof of a WA landing (I think especially around south and north of Geraldton) their case would be a lot stronger.
    Otherwise this CSIRO-study remains to make a much better case for the 31.75S till 29S segments IMO.

  183. TBill says:

    @Andrew
    OK I was going to ask of you needed any help on the fuel issue, but sounds like your engineers are working it. I would be curious to know if data can be logged during a flight, and who can access it. Could lat/long be pulled off by PC and plotted on Google Earth for example, similar to what I do with Flight Sim.

  184. Oleksandr says:

    @Victor,

    One of the strongest arguments against a piloted scenario is non-activation of the cabin ELT. It is hard to believe that F.O. and 12 cabin crew members simply forgot about it. They are trained to activate it in case of an emergency, and they had 6 hours – more than sufficient to realize such a possibility to send a distress signal.

    As you know, I am skeptical about the steep descent. I was not convinced by the oven warm up hypothesis, as there are many unexplained gaps and coincidences, in particular the suggestion to discard a healthy pair of the first BTO and BFO 18:25.

  185. Victor Iannello says:

    @Oleksandr: I don’t think anybody that is proposing a deliberate pilot diversion believes that the FO and crew forgot about activating the ELT for 6 hours.

  186. Victor Iannello says:

    @Ge Rijn: CSIRO states that the lack of debris on WA is consistent with a crash at 35S due to the western current at this portion of the 7th arc. Their drift models say that the debris first travels west, but then later turns and travels east towards WA, with some debris beaching. I can’t make sense of their claims that there is high confidence of a crash site at 35S.

  187. Ge Rijn says:

    @Oleksandr

    Piloted or not-piloted.

    The reasons pro- or contra are at least equaly strong.
    Your arguments and contra:

    (a) The trend in the BTO & BFO after 19:40 suggesting a simple path.
    A pilot won’t normally choose for a complicated path if it’s not necessary. After 19:40 this was obviously necessary anymore.

    (b) Not answered SATCOM calls 18:40 and 23:15.
    A roque pilot/highjacker won’t answer the phone for the risk of giving his identity, motives and location away.

    (c) Never activated cabin ELT.
    A roque pilot/hijacker won’t activate the ELT and would want to prevent others would.

    (d) No sms/e-mails received from the cabin.
    With the SDU and IFE swithed off this was not possible.

    (e) Absence of any motive to bring the plane to ~30S.
    We still don’t know. Motives can be very irrational. See Germanwings and other examples.

    (f) Possibility to unlock the cockpit door by isolating L DC bus from the EE-Bay.
    There would be a need for someone to know how to do this. I don’t know but I assume no pilot or other crew member is trained to perform this kind of technical actions.

    (g) No ELT signal after the crash (or ditching, or whatever you prefer).
    Like mentioned previous. I read ELT’s only have worked in 23% of accident occasions. Very poor reliability I think.

    The strongest argument against a non-piloted flight after FMT ( it was sure piloted before 18:40) is IMO statistically, flights don’t change from piloted to non-piloted flights in 99.9999%(or so) of cases.
    To build assumtions on this very, very, very low chance of possibility I regard as wishfull thinking.

    Why the ATSB took this assumption as their main argument to build their search on remains a mystery to me.

  188. Ge Rijn says:

    @VictorI

    At 35S this only counts for the flaperon-debris not for the 0 degree till 10 degree debris. There they have 7 or 8 landings on WA at 35S.

  189. Ge Rijn says:

    @VictorI

    With non-flap debris the CSIRO prediction is actualy 12 items landing on WA shores from 35S. Also 3 items on the south-west Cape Leeuwin area (been there). This is a much more populated and visited area. Nothing found there either still.

    In retrospect; the lack of debris found in WA weakens their proof considerably. If CSIRO only could come up with one piece of credible WA debris there case would be much stronger and the 31.75S till 29S segments a lot more unlikely.

  190. Oleksandr says:

    @Ge Rijn,

    (a) “A pilot won’t normally choose for a complicated path if it’s not necessary.”

    A pilot won’t normally choose a 6-hr path to the middle of nowhere.

    (c) “A roque pilot/hijacker won’t activate the ELT and would want to prevent others would.”

    How could a locked pilot/hijacker prevent all the 12 cabin crew members + FO from activating the cabin ELT? I think you confused cabin and cockpit ELTs.

    (d) “With the SDU and IFE swithed off this was not possible.”
    SDU was on after 18:25. If not to communicate, then for what?

    Re: “flights don’t change from piloted to non-piloted flights in 99.9999%(or so) of cases.”

    Among accidents or among all the flights? If you take accidents only, Helios flight is a good example when a piloted aircraft became unpiloted: probability 10^-7? In general, there were no disappearances like MH370 in the aviation history.

    Re: “Why the ATSB took this assumption as their main argument to build their search on remains a mystery to me.”

    Can you provide any supportive reason to believe it was a piloted flight till the very end?

  191. Oleksandr says:

    @Victor,

    “I don’t think anybody that is proposing a deliberate pilot diversion believes that the FO and crew forgot about activating the ELT for 6 hours.”

    Then why no distress signal?

  192. Victor Iannello says:

    @Ge Rijn: At 35S, the CSIRO model predicts debris in WA over the range of parameter inputs for windage, extra windage, and extra windage angles. My point is that their conclusion pointing to a 35S crash site because of no debris discoveries in WA is not consistent with their own model. I think you agree.

  193. Victor Iannello says:

    @Oleksandr: Either they were fooled into thinking it was not necessary or they were not able.

  194. Gysbreght says:

    Why was the autopilot disengaged between 1721:13 and 1822:12?

  195. Oleksandr says:

    @Victor,

    “…they were fooled into thinking it was not necessary…”.
    How could they be fooled if the flight was a lot longer than expected and IFE was not functional?

    “…or they were not able”
    Why wouldn’t they be able to activate the ELT?

  196. Oleksandr says:

    “Why was the autopilot disengaged between 1721:13 and 1822:12?”

    Because the ADIRU has failed or it was malfunctioning.

  197. Ge Rijn says:

    @VictorI

    Yes. Their range of 40 till 32S is plenty of ~zero windage/non flap debris landings on WA shores. This only fades clearly after ~31.50S.

    Only their flaperon ~20% windage/leeway avoided the WA shores coming from 35S. The non-flap debris did not (which is the most part of the found debris).
    IMO this only one flaperon-exception proves again the debris did not come from south of 35S. So in this regard it’s usefull again.
    But to base such a solid conclusion just on one (flaperon) piece is not by far enough considering all the other kind of debris in their model.

    IMO CSIRO/Griffin’s dairing conclusion has to be supported by at least one credible WA find.
    I would sure welcome a serious effort to find that piece.

  198. Andrew says:

    @TBill

    RE: “I would be curious to know if data can be logged during a flight, and who can access it. Could lat/long be pulled off by PC and plotted on Google Earth for example, similar to what I do with Flight Sim.”

    Yes, the aircraft record thousands of parameters on every flight. The data is routinely used in Flight Operations Quality Assurance programs to inform airline training/SOP development and to monitor individual aircraft performance. An entire flight can be recreated with the appropriate software. Access to the data is tightly controlled and normally restricted to the airline’s safety and engineering departments.

    Further information here:
    FOQA – Flight Data Analysis of Aircraft for Flight Safety

  199. Brock McEwen says:

    Does anyone know where, specifically, Cawthorne claims live-fire exercises were being conducted that night? (I don’t want to debate the merits of his case – just document its specifics.)

  200. Ge Rijn says:

    @Oleksandr

    a. ‘A pilot won’t normally choose a 6-hr path into the middle of nowhere’

    But if he chose to make this 6 hour flight he would choose to take a simple path and not a complicated one won’t he.

    c. By locking out the co-pilot and depressurise the cabin.

    d.When the SDU was back on line at 18:25 but the IFE switched of before seperately in the cockpit to block this possibility as long passengers and crew were still concious this would be a well thought action from a pilot/hijacker. By 18:25 no one in the cabin would be able to use communications anymore (unconsious or dead).
    It could also explain why the IFE log on was never recieved after 0:19 because it was never switched on again after IGARI.

    Why the SDU went on line again at 18:25 I don’t have a clue. Maybe it was only a side effect of bringing the left IDG on line again for other reasons.

    99.999%(just a guessed number he..) of all flights. The Helios flight is the only commercial flight exception (in aviation hystory) of many millions of flights.
    IMO a very poor argument to build or defend a case like MH370 on.

  201. Victor Iannello says:

    @Oleksandr: Think about it. There are many ways that passengers and crew could have been restrained or incapacitated. The lack of an ELT signal does not mean the plane was not intentionally diverted. As for fooling the crew and passengers, at least initially, there are many reasons for diverting a plane that could have been explained to the passengers and crew that would not have caused the crew to activate the ELT. I never suggested that the crew would have been fooled for many hours.

  202. Oleksandr says:

    @Victor:

    Re: “There are many ways that passengers and crew could have been restrained or incapacitated.”

    How could they be restrained by a single pilot locked in the cockpit? Incapacitated – yes. Then it would be called a mass murder, and the next question would be a motive to fly for several hours after that.

    Re: “The lack of an ELT signal does not mean the plane was not intentionally diverted.”

    But it indicates that whoever was in the cabin was incapacitated. Or not?

    Re: “As for fooling the crew and passengers, at least initially, there are many reasons for diverting a plane that could have been explained to the passengers and crew that would not have caused the crew to activate the ELT”

    So the pilot locked out the co-pilot, and nobody noticed that something was wrong? What about past 23:00, when the plane supposed to land?

    Re: “I never suggested that the crew would have been fooled for many hours.”
    What has happened after the crew and passengers realized that something was wrong?

  203. TBill says:

    I consider turning off ELT one of the actions items of the rouge pilot.

  204. Don Thompson says:

    Concerning ELTs and EPRIBs.

    Transmission of the ELT, mounted in the upper rear fuselage and provided with an external antenna, may be initiated by the flight deck. The ELT doesn’t rely on aircraft power for its operation. One might be expect that if all the voice radios and SATCOM were unusable a resourceful flight crew member might have initiated transmission by the ELT. There is no evidence that the COPAS_SARSAT system detected an ELT broadcast.

    The EPIRB, described as located in a wardrobe in the forward cabin, is designed to be used outside the aircraft. Its part of a grab bag kit to be used after evacuation. Even if this EPIRB was activated inside the aircraft would the fuselage attenuate its signal so that it wouldn’t have been sufficient to be received by a LEOSAR satellite? If it was necessary to hold the EPIRB near a window could that position have been maintained until one of the LEOSAR receivers passed over head?

    It should be straightforward to determine an EPIRB radiated power and what effect an aluminium skin will have on the propagation of its 406MHz signal.

  205. DennisW says:

    @TBill/Oleksandr

    As a PAX or crew member I would not consider the location of the plane to be important. What are the authorities going to do? Shoot it down? My focus would be on what is the pilot going to do or how can his actions be stopped.

    As a PAX or crew member it probably would not even occur to me that the authorities did not know where the plane was at.

  206. lkr says:

    @GR: Concerning debris search in WA, I’d think the volunteer cleaning of remote WA beaches mentioned upthread would have been a more systematic source of debris [if it were there than any other Indian Ocean shore that wasn’t visited by BG! If course it would have helped if ATSB/CSIRO had delegated even one support staff person a few days to contact beach-keepers in Australia [at least! — I think a single clerical support person could have notified at least the resorts around the Indian Ocean in a couple of days.] and set up a notification/chain-of-custody scheme for any leads.

    This should have been done early, certainly after the flaperon recovery in 2015, and it beggars belief that nothing of this sort was done after the 2015 season came up empty. It’s almost as though the quants were so overconfident in finding a crash that they saw no utility in debris and at the same time viewed any diversion into recovering debris as an admission of failure!

    Without slagging the Aussies — they seem to have done a rather solid piece of work with what they had — this seems to be a more general case of groupthink. To get anything at all out of the data at hand in 2014 and early 2015, they had to simplify assumptions — notably taking the satellite data at face value, and assuming a ghost flight [or at least no pilot input]. The problem is that ranks were closed, even when debris started turning up. Ministers had been promised results, and each fruitless month of sea-floor sweeping merely INCREASED their confidence [at least outwardly] that they had tightened the noose to an even more certain find.

    The 35S excitement is part of this process, isn’t it? — “Damn, we were so close”

    There’s nothing to be done about the two [or three?] years of missed opportunity for debris recovery. But I’d like opinions on a couple of points — [1] the ATSB insistence that surface S&R is adequate to eliminate huge swathes of the northern 7thArc, specifically around 30S — that seems to have been treated as very nearly 100% eliminated in their models. So, does anyone want to chime in? We have to assume fuel depletion, so no slick. [2] What do we presume about debris fields in scenarios ranging from ditching to high-speed dive with or without mid-air break-up? What’s the likelihood that a debris field would be missed if the largest fragments were the recovered flaperon, and only a few [say, 10-20] fragments on the surface were as much as 1 meter in length?

  207. Oleksandr says:

    Both the stationary and portable ELTs were capable of transmitting at 406, 121.5, and 243 MHz.

    If a consumer-grade cellphone successfully connected to the network, why would not the cabin ELT? Is it less powerful than a cellphone?

  208. Oleksandr says:

    @TBill,

    “I consider turning off ELT one of the actions items of the rouge pilot.”

    Did he come out from the cockpit for this purpose? Or did he break it down before takeoff?

  209. Victor Iannello says:

    @Oleksandr: An intentional diversion with eventual incapacitated passengers and crew is most likely, in my opinion. Many possibilities exist for the circumstances of the initial diversion.

  210. DennisW says:

    @Oleksandr

    “If a consumer-grade cellphone successfully connected to the network, why would not the cabin ELT? Is it less powerful than a cellphone?”

    Power is not the main consideration. Frequency is the more important. Malay LTE phones operate near 2GHz compared to the 400MHz ELT The aircraft acts as a Faraday cage attenuating signals going out and coming in. At 2GHz an aircraft window would pass the signal with little attenuation. 400MHz would be more difficult, and 121.5MHz would probably not work at all. Wavelength/10 is a rough rule of thumb people use for Faraday cage window size. 400MHz is about 75cm wavelength. A window 7.5cm or smaller would effectively block the signal.

  211. Ge Rijn says:

    @Oleksandr

    The cellphone connection of the co-pilot occured well before 18:25.
    The cockpit ELT could well be switched off by a roque pilot/hijacker and the cabin ELT’s not activated due to decompression.

    And then, what be the use of activating ELT’s at ~30.000ft altitude and ~450knots of speed?

  212. Don Thompson says:

    Oleksandr, and TBill.

    There is no such device as a ‘cabin ELT’. There is an ELT, as I described above, it is fixed equipment with a permanently connected antenna (the aftmost ‘sharkfin’ type antenna on the fuselage top centreline). The EPIRBs are portable, self-contained, devices.

    406MHz is intended to be detected by the LEOSAR satellites. 243 & 121.5MHz are the global VHF comms distress channels, COPAS-SARSAT no longer processes transmissions on these frequencies (as of 2009).

    LEOSAR satellites do not provide continuous lookdown cover the earth’s surface, the detection window is once per hour. GEOSAR & MEOSAR don’t apply.

    Cellphone vs ELT/EPIRB: the cellphone registered to the network within 40km of the base station. The LEOSAR satellite passes overhead at about 850km, it is not continuously overhead.

  213. Ge Rijn says:

    @Oleksandr

    @DennisW provided a more specific answer while I was typing..

  214. Oleksandr says:

    @Don Thompson,

    There is no such device as a ‘cabin ELT’.

    Really? Extract from FI:

    9M-MRO had four ELTs installed:
    • One Fixed ELT. This ELT has the provision to operate on the satellite frequency of 406 MHz when activated. It will also transmit on 121.5 MHz and 243 MHz. This unit is mounted to aircraft structure at the aft passenger cabin at STA 1880. A control switch installed in the cockpit aft overhead panel provides the command signal. If required the flight crew can select the ELT to ON by moving the guarded switch from ARMED to ON.
    • One Portable ELT located in the forward cabin. It is identical to the fixed ELT except that this unit has its own foldable antenna. The portable ELT has a control switch on the front face. It is normally in the OFF position. When needed, the crew can select the switch to the ON position to activate the ELT transmission.
    • Two Slide Raft mounted ELTs. The slide raft mounted ELT will only be available when the slide rafts at doors 1 Left or 4 Right are deployed. The ELT transmission is not satellite enabled. The transmission signal is on 121.5 MHz and 243 MHz which may be monitored with air, sea and ground-based receivers.

    Are you stating that FI provides false information?

  215. Oleksandr says:

    @Dennis,

    Would 243MHz ELT transmitter placed near window work? Compared to 2 GHz cellphone? B777 windows are 380 mm x 250 mm size. Based on you calculation, this is sufficient, isn’t it?

  216. Ge Rijn says:

    @Oleksandr

    And @Don Thompson made it even more specific..;-)

    In the end ELT’s where not activated during the flight, at the crash or after.

    If you considerer those things reliable and think there were people around to activate them, you are right. An ELT signal should have been recieved.
    Only this did not happen, so more logical then is there were no people around and able (or willing) to activate them.

    Or the ELT did not function the way it should, like in almost 80% of earlier cases.

    But more likely the fact that no ELT signals were recieved is ‘negative’ evidence the flight was piloted and the crew and passengers were not able to activate an ELT. And the pilot/hijacker switched/forced this posibility off in an early stage.

  217. Oleksandr says:

    Ge Rijn

    Re: “By locking out the co-pilot and depressurise the cabin.”

    And nobody noticed such an event?

    So your scenario is that a pilot locked out F.O., switched off all the communication devices, killed everyone by depressurizing the cabin, enjoyed performing a few acrobatic maneuvers manually, passed by Penang suddenly affected by nostalgia, and then steered the plane NW to avoid the Indonesian airspace, and then turned toward the SIO accurately avoiding radar detection on his way. Then meditated and contemplated about the sense of life for 5 hours, and after the engines flameout accelerated the plane down at 0.68g?

    Certainly possible. But have you ever heard about similar cases?

  218. Don Thompson says:

    Oleksandr, you’ve answered your own question re ‘cabin ELT’

    The semantics are ELT in aviation, EPIRB in maritime (e.g. door slide/raft & may be triggered by contact with water).

    Whether maritime or aviation, the ELT or EPIRB, in 2014 any transmission will have been relayed by LEOSAR.

    In the meantime I have confirmed that the transmit power for a 406MHz emergency beacon (ELT or EPIRB) is 5W. If the portable device was activated in the cabin, would its signal have been detected by the LEOSAR receiver?

  219. Ge Rijn says:

    @Ikr

    Your comment reflex my thoughts well on the matter. Have to give it some more thought though on your question-signs.
    Indeed, a lot of open questions still.

  220. Oleksandr says:

    @Ge Rijn,

    If you are suggesting that there was nobody alive in the cabin to activate the ELT, then you have a problem to explain the motive. Also try to explain SDU reboot 18:25, and manual piloting from IGOGU to Penang. And the turn NW to avoid Indonesian airspace (what could they do? shut it down?).

  221. TBill says:

    @Oleksandr
    Suffice to say, where my mind is at, I would consider possible tampering with anything that could be tampered with. Shutting data/voice recorders, shutting shutting O2 mask breaker from EE Bay, etc.

    The only MH370 ELT that I think I know how to tamper with myself would be the main ELT that’s supposed to go off when the aircraft hits the water. From what I’ve read, I am thinking that one has a stand-by (off) setting. The other ELT’s are of interest to me.

    I also expect the aircraft could be depressured quite rapidly by manual open of the outflow valves and shutting bleed air. So someone might have to think pretty quick to do anything productive without air.

  222. Oleksandr says:

    @Don Thompson,

    The issue is not in semantics. FI states that the Portable ELT located in the forward cabin was capable to transmit a distress signal at 121.5 MHz and 243 MHz, which are the global VHF distress channels, as you wrote. Are windows of 380 mm x 250 mm size sufficient for this? Forget about LEOSAR.

  223. Ge Rijn says:

    @Oleksandr

    ‘Re: “By locking out the co-pilot and depressurise the cabin.”

    And nobody noticed such an event?’

    Yes, the scenario you discribe is IMO the most plausible.
    And in this case I think everyone noticed the event of the co-pilot being locked out quite soon.
    But by swithing off the SDU, the IFE and decompressurise the cabin no one was able to alarm anyone.

    Only the co-pilot managed to make a contact (at Penang) as far as we know.
    He probably knew the first what was happening when locked out and survived the longest on oxigen-bottles.
    He and the crew preventing at all cost the passengers would panic.

  224. Oleksandr says:

    Ge Rijn,

    “He and the crew preventing at all cost the passengers would panic.”.
    And the crew even have not used portable oxygen bottles, also to avoid panic, I guess…

  225. Ge Rijn says:

    @Oleksandr

    The Helios-steward came into my mind. It would be a similar case..

  226. DennisW says:

    @Oleksandr

    “Would 243MHz ELT transmitter placed near window work? Compared to 2 GHz cellphone? B777 windows are 380 mm x 250 mm size. Based on you calculation, this is sufficient, isn’t it?”

    Actually, I did not calculate anything. Just dragged out a rule of thumb we used to use for cages that were very thin (like a window screen) The link below on the physics stack exchange by a knowledgable poster is pretty complete, IMO. The cutoff frequency is (speed of light) / (twice the window width). A 250mm window width works out to a cutoff frequency of 600MHz. However, this produces exponential decay as the wave passes through the thickness of the material. In our case, this is very thin. Certainly significant energy would escape at 400MHz. The 243 and 121 frequencies would need to be looked at more carefully to see how much energy would escape. My guess is virtually nothing at 121 and some at 243.

    https://physics.stackexchange.com/questions/149607/what-is-the-relationship-between-faraday-cage-mesh-size-and-attenuation-of-cell

  227. Don Thompson says:

    Oleksandr,

    ELT and EPIRB output power at 121.5 & 243 MHz is 150mW, at 406MHz it is 5W.

    (Just for comparison, the aircraft VHF comms transceiver delivers minimum 25W)

    Consider antenna gain, the problems of the window aperture as Dennis describes, RF power…

  228. Oleksandr says:

    @Dennis,

    Using the link you posted, I estimate the decay rates:
    alpha_243MHz = 11.49 m^-1,
    alpha_121MHz = 12.31 m^-1.

    The external diameter of B777 is 6.20 m; the internal 5.84 m. Hence the thickness of walls is 0.18 m. Respectively, the signal decays are:
    decay_243Mhz = 7.9 times;
    decay_121MHz = 9.2 times.

    Assuming that the transmit power for 243/121MHz is also 5W, the escape power would be 0.63W and 0.54W for 243 and 121 MHz channels. These figures are similar to an average cellphone’s emission power.

  229. Oleksandr says:

    @Don Thompson,

    Just saw your post. If 150 mW, then the VHF emission power would be 19 mW and 16.3 mW respectively.

  230. Mick Gilbert says:

    @Ge Rijn

    Re: “It could also explain why the IFE log on was never recieved after 0:19 because it was never switched on again after IGARI.”

    We know that the IFE was powered at the 1825 UTC log-on (first ping) because connections were established by the IFE SMS/e-mail and the Built-In Test Equipment applications so it was most assuredly switched on after IGARI.

    I contend that it is very difficult to reconcile the repowering of the SDU with the IFE functioning at 1825 UTC with the rogue pilot theory. For what is hypothesised as a well thought out and deliberate plan, restoring comms to the passenger cabin (SMS/e-mail) after only an hour of downtime seems reckless if the intent was to ensure that the airplane remained dark.

    Re: “Only the co-pilot managed to make a contact (at Penang) as far as we know.”

    Actually, we don’t know that at all. All we know is that the First Officer’s mobile momentarily registered on a cell tower near Penang. That may simply mean that he hadn’t turned his phone off when he boarded. The cell tower registration appears to have been coincidental with MH370’s turn to starboard from south of Penang; the bank associated with that turn would have improved line on sight from the right-hand (FO’s) side of the airplane to the cell tower.

  231. David says:

    @Victor. Just for the record, my last post to DennisW did not have the intent you attribute to it.

    There was yesterday a Professor Pattiaratchi input which endorses the search area proposed by the review. As I understand it new debris has overtaken their original 28-33degS estimate.

    https://theconversation.com/evidence-mounts-for-a-search-further-north-for-missing-flight-mh370-76584

  232. Mick Gilbert says:

    With regards to the ELTs in general, one of the other considerations is the battery. The operating range for the LiMNO2 6V battery pack is -20°C to +55°C. Accordingly, if there was a depressurisation event then at some point it would be too cold for proper operation. The batteries have a five year activable life and we know that the battery expiry date for both the fixed and portable ELTs on 9M-MRO was November 2014.

    With regards to the fixed ELT, it has two settings;
    ARMED – ready to transmit when automatically activated according to acceleration/time criteria, and
    ON – activated and transmitting.

    It can not be turned off from the flight deck.

  233. lkr says:

    @David: No, Pattiaratchi continues to say that his drift study gives a location between 28S and 33S, and cites a European study bracketing a bit further north. His study didn’t include the 2 most recent finds in South Africa, but I doubt this would drag the source further south.

    He then firmly states that he agrees that the site is between 32.5S-36S. Not, I think, because his data says the southern part is possible, but because it is now understood that a renewed seabed search is politically possible only if all hands take a blood oath ato the same parameters. And, as previously stated here, it’s much better if the new search be contiguous to the old one — the “we were nearly correct and missed by a hair” strategy.

    And to continue my rant, if anyone had cared, they would have in hand a hundred, or hundreds, of fragments, many freshly ashore with epifauna, for a more useful drift study. Not to mention a much better handle on the speed/angle of entry in the water.

  234. DennisW says:

    @David

    I am a bit confused by the Pattiaratchi piece you linked to.

    He identifies three studies:

    1> His own which concluded 28S-33S

    2> A European study which concluded 28-35S

    3> The latest CSIRO work which concluded 32.5S-36S

    He then goes on to claim this constitutes credible new evidence that the plane is in the 32.5S-36S region on the 7th arc. He does not explicitly say why the region from 28S -32.5S, which is compatible with two of the three studies he references can now be excluded.

    I did not see any reference to the “new debris has overtaken” earlier estimates notion.

  235. TBill says:

    @Mick Gilbert
    My understanding (from news reports at the time) is the fixed ELT itself may have an ON/OFF switch. Your info above (from the FI) says no OFF control from COCKPIT, which I was not expecting. I am expecting there is an OFF option at the fixed ELT unit itself.

    Interesting (from FI) that MAS did not order the B777 with the fixed ELT and did the install themselves. I wonder if industry would agree with the installation (location/model etc), and of course I wonder who can access the unit and when.

  236. David says:

    @Ikr, DennisW. The way I read it was that his work and that of the Italians was that he referred to (at the 2nd para under “More Evidence”) which you will see was dated July 2016 and has debris diagrams extending to mid 2015. That identified 28-35degS and 28-33 respectively as the site. Apparently his model could not predict the arrival of 4 items of 22 in places, “…not be well represented in the (his) oceanographic model”. Also, both his ‘Loc 11’ diagram and the 32.5S (or near enough) origin video have plenty of debris reaching the WA shore and obviously would not include the initial NW set westward uncovered more recently.

    The later November 2016 work by the CSIRO for the Review and its April 2017 follow-on incorporated the outcome of that and the new experiments. I assume he now sees his 28-33degS recommendation as outdated, like that of the Italians. The new Review 32.5-36degS is now endorsed by him from conviction I would assume, based on its more up to date data.

    It is as if a pararaph drawing together the earlier work with the new search area is missing, ie immediately before, “Thus based on the results of several independent oceanographic drift modelling studies…all have come to a similar conclusion to that of the ATSB’s first-principles review”. As it reads that is a non-sequitur.

    To me his endorsement is the bottom line. I would expect this to be from conviction, not discipline loyalty, being a form of peer review.

  237. Mick Gilbert says:

    @TBill

    Based on the User’s Handbook, the ELTA FRANCE Emergency Locator Transmitter ADT406 PN 01N65900 as fitted to 9M-MRO has an OFF switch on the unit itself that allows it to be switched off for maintenance. However, once the unit is mounted to the top center of the fuselage at STA 1880 above the aft passenger cabin (roughly above the third last row of passenger seats) it is set to ARMED. You would need to remove the aft passenger cabin ceiling to access the unit to turn it off once it has been installed.

    There is a spring loaded RESET switch for the fixed ELT on the flight deck that ends the transmission of emergency locator signal once the unit has been activated. The RESET switch is not an OFF switch; it simply resets the unit to ARMED after it has been either automatically activated according to acceleration/time criteria or turned ON.

    A fixed ELT is not a standard Boeing item, possibly because of the range of manufacturers (ELTA, Artex, ACK, Kannad, etc) and units available, possibly because of the poor performance record of the units, possibly both. In some respects it is similar to the options available to operators with regards to passenger emergency oxygen systems (gaseous, 12 minute chemical generators, 22 minute chemical generators).

  238. David says:

    Please delete “respectively” third line. The order is the reverse.

  239. David says:

    @Ikr, DennisW. Also, about some detail, he mentions 22 debris items he has investigated of which 4 unpredicted by his model were found in Mauritius and Rodrigues.

    However the debris list of those 22 items includes 4 unidentified, numbers 13, 14, 17 & 21 (see his “Predicted locations…” the second diagram), leaving 18. There are 3 from Mauritius and Rodrigues, not 4. The flaperon at Reunion could not be included in those since it was ‘predicted’.

    Some more work on the article is needed.

  240. Victor Iannello says:

    @David: I believe he has a typo relative to the unpredicted items “4,5,8,18” which I believe should be “4,5,8,10”.

    He does concede that his model does not predict the discovery of item 4, which is the flap fairing found near Mossel Bay, South Africa, in December 2015, although I don’t see an explanation.

  241. Oleksandr says:

    @Mick Gilbert,

    Re: “With regards to the ELTs in general, one of the other considerations is the battery. The operating range for the LiMNO2 6V battery pack is -20°C to +55°C. Accordingly, if there was a depressurisation event then at some point it would be too cold for proper operation.”

    A common mistake is to inevitably associate depressurization with temperature drop. With regard to this, one would need to consider combinations of slow/quick depressurization with temperature control inside the cabin on/off, e.g. 4 combinations. There would be many implications both logic and fuel wise.
    Also, the battery is likely enclosed into a plastic casing of low thermal conductivity, so that its performance would unlikely be an issue for a while after such an event.

  242. Paul Smithson says:

    @Mick. Re phone registration. Note 1: the position of phone registration occurs when plane’s track is tangential to the tower, so minimum velocity with respect to tower. Note 2: there is no sign in the RMP records leaked that phones other than the pilots were investigated – so we don’t know if cellphones of other pax were also registered or not.

  243. Victor Iannello says:

    @Paul Smithson said, “there is no sign in the RMP records leaked that phones other than the pilots were investigated”

    The phone records of both pilots and crew were investigated. However, I do agree that passenger phone records were likely not investigated. In fact, one of the next of kin claims that they were never contacted by the investigative team.

  244. Mick Gilbert says:

    @Oleksandr

    Surely it’s not a mistake to inevitably associate depressurisation with a temperature drop because inevitably there will be a temperature drop associated with depressurisation, it’s a straight up matter of physics relating to adiabatic lapse rate, isn’t it? I agree that the issue is the size and rate of the drop in temperature but I thought my use of the rather non-specific “at some point” would have had that covered.

    Regarding the thermal conductivity of the battery case, funny story, if you were going to get a faster than “usual” cooling of the battery, it would be with the ELTA ADT406 ELT. From what I understand the ELTA 00E64191 LiMNO2 battery pack for the ADT406 ELT is unusual in that the batteries are physically attached to the metal back plate of the ELT; the batteries and back plate come as an integrated unit. When you change the batteries in the ADT406 you also change the back plate. That arrangement would significantly enhance thermal conductivity over a drop in style battery.

  245. Oleksandr says:

    @Mick Gilbert,

    Re: “Surely it’s not a mistake to inevitably associate depressurisation with a temperature drop because inevitably there will be a temperature drop associated with depressurisation, it’s a straight up matter of physics relating to adiabatic lapse rate, isn’t it?”

    No. I specifically mentioned slow or fast. Adiabatic process implies no external source of heating. During slow depressurization cabin temperature can be maintained by heaters, so that the pressure would drop, but the cabin temperature would stay. In addition, the thermal energy would be drawn by the air from surrounding objects, which have much higher heat capacity.

  246. Peter Norton says:

    Concerning the discussion about switching off the ELT:

    – TBill: “I consider turning off ELT one of the actions items of the rouge pilot.”

    – Oleksandr: “Did he come out from the cockpit for this purpose? Or did he break it down before takeoff?”

    – Ge Rijn: “The cockpit ELT could well be switched off by a roque pilot/hijacker and the cabin ELT’s not activated due to decompression.”

    – TBill: “The only MH370 ELT that I think I know how to tamper with myself would be the main ELT that’s supposed to go off when the aircraft hits the water. From what I’ve read, I am thinking that one has a stand-by (off) setting.”

    – Mick Gilbert says: “With regards to the fixed ELT, it has 2 settings;
    ARMED – ready to transmit when automatically activated according to acceleration/time criteria, and
    ON – activated and transmitting.
    It can not be turned off from the flight deck.”

    – TBill: “My understanding (from news reports at the time) is the fixed ELT itself may have an ON/OFF switch. Your info above (from the FI) says no OFF control from COCKPIT, which I was not expecting. I am expecting there is an OFF option at the fixed ELT unit itself.”

    According to this seemingly well-sourced article http://cnn.it/2pg4EOb (edit: and confirmed by Mick Gilbert’s quote from the user handbook), the fixed ELT cannot be switched off:

    “Hijackers or renegade pilots cannot disable some of the emergency beacons, namely, the ones attached to the plane’s airframe. They are powered by batteries and inaccessible to the crew.”

  247. Peter Norton says:

    Another interesting part of the article:

    « What can explain the lack of a signal in the Flight 370 case?
    […]
    The attached beacon on Flight 370 should have activated if the plane crashed. But experts […] say there are numerous reasons why a beacon could fail in an ocean crash. The beacon itself could be damaged by the impact, or its antenna could be sheared from the fuselage, rendering it inoperable. And there’s one other possibility considered even more likely by some: The crash impact may have actually activated the beacon, but the damaged plane sank in less than 50 seconds, the time necessary for it to transmit its first emergency signal.
    […]
    The aircraft could have sunk before the ELT began transmitting. It takes 50 seconds for the ELT to establish the necessary connection. It only takes one half-second data “burst” to indicate there is an emergency. But it can take a half-dozen bursts — at the rate of one every 50 seconds — to provide information that will allow Cospas-Sarsat to triangulate the beacon’s position. “In this case, there wasn’t even one burst, according to the reports that we received,” Lett said. If the plane crashed in the southern Indian Ocean, as Malaysia Airline officials believe, the lack of a distress call could indicate that the plane plunged into the water, or sank quickly, because once underwater, the beacon is ineffective. Likewise, the water-triggered ELTs in the life rafts would be ineffective if they became submerged, according to published Honeywell manuals for the devices. Cospas-Sarsat also notes that beacons must have a relatively unobstructed view of the sky to work properly. “A submerged beacon, or one with its antenna blocked by the body of an aircraft or vessel, is unlikely to be received by the satellites,” the organization said. Said Honeywell Aerospace spokesman Steven Brecken: “Until the recorders are recovered, we don’t want to speculate what could or could not have happened. We ask the same thing you do, why didn’t the ELT operate? We don’t have the answer.”

    In a recent letter to Malaysian authorities, a family group showered them with questions about the ELTs.
    – How many ELTs are on the plane, they asked, adding that they had gotten conflicting numbers.
    – Did Malaysia Airlines conduct maintenance checks ?
    – When was the latest check for MH370’s ELT ?
    – They asked to see the results of those checks.
    – Was the 406 MHz beacon certified ?
    – Was it possible to break the ELT in a crash ?
    – Where exactly was it located ?
    – Would the ELT signal be weakened if it was surrounded by metal ?
    – Was the cable and blade antenna 9G certified ?
    – How much impact is needed to activate the ELT ?
    – Had the crew been trained in the use of ELTs ?
    – Can a beacon unlock “and [float] to the surface of the water ?”
    Many of the questions remain unanswered. »

    ***

    I find the 50 seconds delay noteworthy.
    Also an interesting question whether a sunken ELT can float back to the ocean surface and transmit ?

    Besides, the article claims the fixed ELT to be a Honeywell RESCU 406 AFN
    (as opposed to ELTA FRANCE ADT406 identified in the FI report).
    Is this an error ?
    Strange, because the quotation of the Honeywell spokesman above also implies it was their ELT …

  248. Mick Gilbert says:

    @Peter Norton

    Peter, while I agree with the general thrust of the article – that disabling the fixed ELT would be very difficult (it would require an excursion to rear of the pax cabin with a ladder and tools) – the fact that the author misidentifies the ELT fitted detracts a little from its scholarship.

  249. Mick Gilbert says:

    @Peter Norton

    Peter, unlike a maritime EPIRB, this type of ELT is not designed to float. As I think Ge Rijn has pointed out earlier, the fixed units are notoriously unreliable. One of the significant weaknesses in the set-up is that the ELT, which is mounted on the inside of the upper rear fuselage, is connected to an antenna in the tail; if the connection is severed, as you might expect with a hulk break up, then you’re effectively out of business.

  250. TBill says:

    @Mick Gilbert @Oleksandr
    One of my focus areas is the depressurization temperature math. In a depressurization, the air will get very cold, but it will start warming back up right away because all of the mass of the cabin is still at room temperature. So I am not expecting the equipment to get too cold, unless there was some kind of extended period with outside cold air coming in. It could be uncomfortable but also the low mass of the low pressure air does not have as much cooling capability. So in Helios for example the ACARS was sending equipment overheat messages from some of the EE Bay.

    @Peter Norton
    The FI location of the ELT in the rear of the passenger cabin does not sound inaccessible to me. You’d have to show me a picture of the MH370 B777 installation (or equivalent). My expectation is, if we ever find the aircraft, things like the digital data recorder were switched off by breaker. My guess is the fixed ELT was turned off or otherwise tampered with. I never said the pilot had to be the only person involved.

    I look at this accident not from the perspective of blaming the pilot, but from the perspective of blaming aircraft design and MAS operating policy for allowing a huge window of opportunity for various rouge actions to take place. Almost inviting if there are those wishing to take advantage of the situation (hijackers etc). There should obviously be more safeguards than what MH370 had. Part of my thinking to identify all weaknesses that rouge pilot, with or without an accomplice, might have exploited.

    Someone mentioned tools.
    Which brings up the question, what tools or tool box does a pilot and/or flight crew normally have?

  251. DennisW says:

    @David

    What is clear is that both CSIRO and Pattiaratchi are arguing for the continuation of the search. That is in and of itself refreshing, and more important to me than the details. They are not part of some “let’s bury this thing” decision on the part of the tri-patite group and their subcontractors.

    It is probably a bit aggressive on our part to nit-pick an article written for the general public. My take away is that he is simply saying that a group of independent analysts have come to similar conclusions. Given the stochastic nature of drift physics, I too find this result significant and positive.

  252. Peter Norton says:

    Mick Gilbert: “Peter, while I agree with the general thrust of the article, […] the fact that the author misidentifies the ELT fitted detracts a little from its scholarship.”

    At first, I also thought that the author was simply given false information (i.e. Honeywell RESCU 406 AFN), but the quote of the Honeywell Aerospace spokesman Steven Brecken implies that it was their ELT … ?

    Mick Gilbert: “One of the significant weaknesses is that the ELT, which is mounted on the inside of the upper rear fuselage, is connected to an antenna in the tail; if the connection is severed […] then you’re effectively out of business.”

    What about the attached back-up antenna (PN:02N64070) ?
    The FI report seems to suggest that while the fixed ELT did not have one (despite being recommended), the portable ELT did.

  253. Peter Norton says:

    Perfect Storm: “The CSIRO report says: « Our earlier field testing of replicas of the flaperon was unable to confirm numerical predictions by […] DGA that the flaperon drifted left of the wind. Field testing of a genuine Boeing 777 flaperon cut down to match photographs of 9M-MRO’s flaperon has now largely confirmed the DGA predictions. »
    Why did they use replicas in the first place (instead of a genuine flaperon) ?
    And why wasn’t the MH370 flaperon used, that was found on Réunion ? Yes, it’s a piece of evidence, but they had 1 year to examine it, so I guess they have already done all the work on it. And putting it into the water, let it drift for a while and recovering it later won’t damage it anyway.”

    Ge Rijn: “The ATSB/CSIRO did not have the original flaperon available so they did the best the could on the flaperon now.”

    Would there be any grounds for the BEA not to cooperate with Australia ?
    I would have expected them to lend the Réunion flaperon for such important reasons.

    But even if Australia couldn’t obtain the MH370 Réunion flaperon, I am also wondering why they didn’t at least use a genuine flaperon (instead of replicas). An old spare part would surely be obtainable … ?

  254. DennisW says:

    @Peter

    What I find even more annoying is the lack of commentary from the BEA relative to the nature of the trailing edge damage to the flaperon. Was it flutter or water contact? Larry Vance, crash investigation “expert”, has publicly stated the damage was caused by water contact. No one from the official side has corrected that, so my assumption is that it is true.

  255. Ge Rijn says:

    @Peter Norton

    The latest CSIRO-study did use a genuine flaperon.
    A reason the BEA will not lend the Reunion-flaperon to anyone could be it’s regarded as a piece of evidence in a criminal case (in France).

    This could also be the reason why CSIRO did/could not use the outboard flap section which probably also would drift like a high-windage piece like the flaperon. THe ATSB has this piece in their hands.

    When using the original pieces in this kind of tests, original traces and conditions of the piece could be altered or lost which would render them possibly useless as evidence in a court.

  256. Oleksandr says:

    @TBill,

    Re: “In a depressurization, the air will get very cold, but it will start warming back up right away because all of the mass of the cabin is still at room temperature.”

    Again, you consider quick, nearly instant depressurization. What if depressurization took 20 minutes?

  257. Oleksandr says:

    @Peter Norton,

    I think in your post April 29, 2017 at 7:49 am you mixed up discussions about fixed and portable ELTs, sometimes referred here as the cockpit and cabin ELTs respectively.

  258. Peter Norton says:

    > Ge Rijn: The latest CSIRO-study did use a genuine flaperon.

    Yes, but only now, 3 years after the crash … that was the point made in the posting.
    In the first place, they used replicas, which led to erroneous results.
    That could have been avoided by using the Réunion flaperon (starting in 2015) or at least any real flaperon (starting in 2014), isn’t it ?
    I don’t understand why replicas were used.

    > Ge Rijn: A reason the BEA will not lend the Reunion-flaperon to anyone
    > could be it’s regarded as a piece of evidence in a criminal case (in France).

    That point was also addressed in the posting. That piece of evidence would not be destroyed/damaged, but examined (as to its drift behaviour).

    > Ge Rijn: When using the original pieces in this kind of tests,
    > original traces and conditions of the piece could be altered or lost
    > which would render them possibly useless as evidence in a court.

    Maybe you are right.
    My reasoning was that
    a) … the importance of “not touching” the flaperon would be outweighed by the importance of finding other pieces of evidence (the main wreckage field and especially the flight recorders) and by the importance of solving the case.
    b) … the flaperon would not be damaged by letting it drift for a while and given that it has floated for many months, its condition would hardly be changed by floating for a few days.
    c) … I have seen many aircraft investigation documentaries and whenever there was the need to do testing on recovered parts of the aircraft and/or cargo (=evidence!) it was always done.

    Because of these 3 points, it would have been logical to me to use the Réunion flaperon.

    But even IF that was not possible (in order to preserve evidence), that does not explain, why another genuine flaperon was not used.

  259. Ge Rijn says:

    @Mick Gilbert

    On your comment;

    ‘We know that the IFE was powered at the 1825 UTC log-on (first ping) because connections were established by the IFE SMS/e-mail and the Built-In Test Equipment applications so it was most assuredly switched on after IGARI.’

    You are right the connections with the IFE during the 18:25 log-on were made but I think together with the SDU the IFE was off-line from IGARI till 18:25.
    Anyway also after 18:25 the IFE SMS/E-mail was not used. When coming online again passengers and\or crew must have noticed this soon when screens lit up again.
    And at least the crew must have known by then there was something completely wrong with the flight.

    If they were still alive/consious they must have seen they flew back over Malaysia and passing Penang to open see without an attempt to land.
    IMO at least someone sure would have used the IFE SMS/E-mail to send a distress call. This did not happen (based on known info).

    So IMO this leaves two options. Or no one was consious in the cabin anymore by 18:25 or the IFE was switched off from the cockpit right after the 18:25 log-on. And stayed switched off till end of flight possibly causing the incomplete (non IFE) log-on at 0:19.

  260. Ge Rijn says:

    @Peter Norton

    I agree with you on your explaining post about the (no) use of the flaperon.
    I only tried to give a possible argument why the flaperon was not provided by France. They don’t treat the case as a crash-investigation but as a criminal act.

    I wonder more about why they did not use the Tanzania outboard flap section instead or also.

  261. Victor Iannello says:

    @DennisW said, “Larry Vance, crash investigation ‘expert’, has publicly stated the damage was caused by water contact. No one from the official side has corrected that, so my assumption is that it is true.”

    Larry Vance claims the damage to the flaperon indicated that it was extended (“drooped” is a better description) when it separated from the plane.

    The ATSB, in its Search and Debris Update, dated Nov 2, 2016, refuted this, and concluded:

    The right outboard flap was most likely in the retracted position at the time it separated from the wing.

    The right flaperon was probably at, or close to, the neutral position at the time it separated from the wing.

  262. Peter Norton says:

    @Ge Rijn:
    I agree with you and I appreciated your comment. Sorry in case I came across otherwise.

    > Oleksandr: “I think you mixed up discussions about fixed and portable ELTs”

    I just included all citations regarding ELT switch-off.

    > Oleksandr: “sometimes referred here as the cockpit and cabin ELTs respectively”

    By “cockpit ELT” you mean the fixed ELT, which is not in the cockpit but in the cabin ?

  263. DennisW says:

    @VictorI

    I did see the outboard flap statement. Missed or mentally blocked the comment on the flaperon. To my knowledge there was no damage to the flap similar to the flaperon ragged edge. Still a bit of a mystery to me that the French could clear up I am sure.

  264. Mick Gilbert says:

    @DennisW

    There is indeed trailing edge damage to the section of right outboard flap. Bearing in mind that the recovered piece is just the inboard quarter of the outboard flap, about two-thirds of the trailing edge is missing.

  265. Mick Gilbert says:

    @Ge Rijn

    Re: “… the connections with the IFE during the 18:25 log-on were made but I think together with the SDU the IFE was off-line from IGARI till 18:25.”

    That is supposition, there is no evidence to support the contention that the IFE was ever taken offline. The fact that the IFE SMS/e-mail facility was not used after 1825 UTC is not evidence that the IFE was offline.

    Your two possible options, “… no one was consious in the cabin anymore by 18:25 or the IFE was switched off from the cockpit right after the 18:25 log-on.” make sense.

  266. Mick Gilbert says:

    @TBill @Oleksandr

    In the interest of moving the depressurisation/temperature discussion forward, do either of you have any estimates relating to where the cabin temperature might have settled if the airplane had depressurised at FL350?

  267. TBill says:

    @Mick Gilbert @Oleksander
    The final temperature would be a complicated calculation, with air cool down due to depressurization, followed warm-up at a certain heat transfer rate from the surroundings. I would wonder if Boeing may have complex models for this.

    My guess is the cabin might initially end up 10-20 deg F (5-10 deg C) cooler. I don’t know if the bleed air would be able to warm up the aircraft after that (at the low pressure) and, if bleed air is off, then the cabin will presumably start to slowly lose further heat to the surroundings.

  268. DennisW says:

    @Mick

    Any explanation of the trailing edge damage? Flutter or water impact?

  269. lkr says:

    IMHO, it’s insane to suggest tossing any of the scarce recovered debris back in the ocean for drift studies. These pieces have forensic utility — not as courtroom exhibits or trophy-wall pieces at Le Inspector’s office, but basic, how did it come apart, work.

    And better to do a sea run with a dozen stand-ins than that one original. Best would be to visit a aircraft bone-yard for multiple flaperons, but just as for the debris recovery effort there seems to have been a minimal budget for all things debris.

  270. Mick Gilbert says:

    @DennisW

    G’day Dennis, no, there’s no attempt to explain the missing trailing edge. Mike an Don wrote a preliminary piece on this item (Analysis of the Tanzania Debris, Michael Exner and Don Thompson, June 26, 2016) and it is addressed in some detail in the ATSB’s MH370 – Search and debris examination update of 2 November 2016, but neither address possible causes for the TE damage.

  271. Donald says:

    @Dennis

    Flutter was an arrived at (incorrectly) determination by self-proclaimed ‘experts’ wedded to a non-piloted, no control inputs, accident scenario. The real shame is that these ‘experts’ continue to cling to this fabrication despite all evidence to the contrary. Flutter does not account for the condition of any of the debris as currently understood.

    Maybe Andrew would care to opine on this topic (?), as some very well respected pilots have laughed the notion out of the park.

    @All

    Obviously US intel would have done an exhaustive investigation regarding pax cell phones (if they were not 100% certain as to cause and/or perpetrator). It’s pretty clear to me, at least, that the only connection detected was the FO’s phone. It’s also quite unbelievable that Malaysia would not have investigated the pax phones as well. Again, nothing of note, whatsoever, has been leaked or reported. And the NOK have, to my knowledge, made no noise in this regard, either.

    A locked out FO and rapid depressurization and subsequent incapacitation of pax and crew far and away is the scenario that holds together best. I dare say that everyone should move in that direction. We would be well served imho.

    @Oleksandr.

    To find motive requires you to delve deeply into the mind of the offender. You act as though there is little if anything to be understood about Zaharie’s state of mind and predilections. I would dare say you know next to nothing about the man, and as such may want refrain from the posing of ‘motive’ as some insurmountable obstacle that cannot be adequately accounted for. It can.

  272. David says:

    @Mick. About the ELTs my recolection from having gone into this earlier is that the cockpit switch in test position will isolate the ELT if left there after test. However it is spring loaded so would need to be deliberately held: forcing a soft wedge between frame and switch, super glue etc., requiring someone to plan it.

    In the discussion I might have missed the ELT manuals and information being posted. If so, disregard. If not, the fixed is at:
    http://www.elta.fr/uploads/files/00a2b82a28496c1ff5d6d3513216879f6d4fd557.pdf

    The portable is at:
    http://www.elta.fr/uploads/files/67e22a2ab4f420f72faf85383a6248949dc5d68f.pdf
    and
    http://www.elta.fr/uploads/files/e54bd1565eb0cbcda283ae0e1164711b2fde77cd.pdf

    Activation of the fixed is described here:
    http://www.pprune.org/tech-log/535756-elt-doubts.html#post8376340

    You said, “…there’s no attempt to explain the missing trailing edge”.
    In case I did post a couple of weeks back a study into a closely related issue. My preliminary analysis was that the trailing edge could not come off in flight (flutter, overstress) without the whole flaperon immediately separating. However that would entail the flaperon deploying uncontrollably, at least partly, before failure so would be inconsistent with the ATSB opinion. However that does not entirely rule out that the flaperon and outer flap were not fully housed.

    https://www.dropbox.com/s/i4bygiic796opn6/Right%20flaperon%20separation.docx?dl=0

    If my analysis and the ATSB opinion both stood, the only explanation would be that the flaperon separated when whole and the trailing edge broke off subsequently.

  273. David says:

    In case it helps I did…

  274. Oleksandr says:

    @Mick Gilbert,

    No, I have not estimated, because it depends on many factors and assumptions. One of the main assumptions, in my opinion, is whether heating was on or off. The other critical assumption is how fast air escaped.

  275. Oleksandr says:

    @Peter Norton,

    Yes, fixed ELT. It can be activated from the cockpit.

    The other one is the portable ELT, identical to the fixed one. It could be activated from the cabin regardless what was going in the cockpit. Hence the discussion why no signal was received: due to attenuated signal, or because the portable ELT was simply not activated.

    Both the ELTs are satellite, and VHF-enabled.

  276. David says:

    @Victor. I notice that in the David Griffin interview that ALSM posted, there was no allusion to other opinions though this was after he had commented about your and Richard Godfrey’s assessments.

    http://www.abc.net.au/pm/content/2016/s4660021.htm

    The Pattiaratchi article followed that also and offered fullsome support for the Review search area, though that might well have been written earlier.

    Clearly your opinion and others is that a search in the Review recommended area is less likely to be successful than moving it further north east.

    Since it is unlikely there will be a new search until further confidence is gained, one could argue that the new information needed for that will decide which the new area should be, so there is no harm in remaining passive pending that. Is that how you see it?

    Incidentally I note that David Griffin resisted commenting on whether there should be a new search, restraining his recommendation to searching in the Review’s area were a new search to be conducted. He would not say what the prospects were of finding the aircraft there are other than agreeing they would be more than 50%.

    I think such reticence is professional and admirable, given his personal enthusiasm for the new search to go ahead now but the interview’s effect might well be contra if anything. So, really it should be in my dispassionate view, though like many I suppose, one part of me would like to get it going just on the prospect it might be successful.

    @DennisW. Yes the Pattiaratchi endorement of the Review search area does represent a cross check and might could help lift prospects of a Review-area search. As above, even if there are better prospects elsewhere it does not mean necessarliy that a search in that area will be unsuccessful.

    If on the other hand, as seems likely, there will be no search without more certainty, then it does no harm anyway.

    Harm would come from a new search being unsuccessful since that would be the end of it, more money would have been wasted and NOK hopes raised and dashed again. That is why we need more certainty.

    “It is probably a bit aggressive on our part to nit-pick an article written for the general public”. Well yes but part of my disquiet was unstated, having drilled down further.

  277. Andrew says:

    @Donald

    RE:“Maybe Andrew would care to opine on this topic (?), as some very well respected pilots have laughed the notion out of the park.”

    One so-called ‘very well respected’ pilot laughed the notion out of the park because he claimed that hydraulically driven flight control surfaces couldn’t possibly flutter, but that’s not necessarily true in this case. Each B777 flaperon has two Power Control Units (PCU), each controlled by different electronic units known as Actuator Control Electronics (ACE). The PCUs have two operating modes; normal and bypass. If there is an electrical or hydraulic failure with the PCU or its ACE, the PCU will revert to bypass mode. If both PCUs on a flaperon are in bypass mode, the flaperon can move freely; there is no damping mode. If that were the case for MH370 at the end-of-flight, then I’d say the flaperons could flutter.

    If both engines and the APU eventually flamed out and the aircraft was left powered only by the RAT, then neither of the PCUs on the left flaperon would have had hydraulic power and consequently they would have been in bypass mode. The left PCU on the right flaperon would also have been in bypass mode, but the right PCU should have had hydraulic power from the centre hydraulic system and electrical control from the C ACE. Consequently that PCU should have been in normal mode. However, that assumes the RAT was operating normally. If the aircraft ended up in a high speed descent and exceeded its design envelope, then it’s possible the RAT might have stopped working at some point during the descent. If that happened, then the right flaperon would also have been able to move freely and might have been subject to flutter.

    Of course the foregoing doesn’t mean the observed damage to the flaperon is consistent with flutter. That’s not something I’m qualified to answer.

  278. Oleksandr says:

    @Donald,

    “To find motive requires you to delve deeply into the mind of the offender. You act as though there is little if anything to be understood about Zaharie’s state of mind and predilections. I would dare say you know next to nothing about the man, and as such may want refrain from the posing of ‘motive’ as some insurmountable obstacle that cannot be adequately accounted for. It can.”

    Failure or unwillingness to consider motive or motivation is one of the main factors, which contributed to the fiasco of the search operation. I have predicted this result 2.5 years ago (see JW blog), and yet the IG and ATSB were sure the plane was there. Motive or motivation are required to make assumptions on the path between 18:22 and 19:40. Descents, ascents or turns are not made without a reason. I do not belong to the camp, which postulates “Z. did it” without any further evidence. Don’t make the same mistake again. Motive or motivation does matter.

  279. Victor Iannello says:

    Oleksandr said, “I do not belong to the camp, which postulates “Z. did it” without any further evidence.”

    The simulator data clearly shows that Captain Zaharie Shah created a simulation of a flight that departed KLIA, traveled up the Malacca Strait, turned to the south in the Andamans, and then flew until fuel exhaustion in the SIO. The data related to this simulation was then deleted from his computer. You (and others) might believe that there is insufficient evidence to definitively conclude that captain deliberately diverted MH370. However, to say that there is no evidence for postulating that the captain diverted the plane is false.

    If you have not read it, this paper shows the simulator data is all from the same simulated flight.

  280. Gysbreght says:

    @Andrew: “The PCUs have two operating modes; normal and bypass.”

    The flight control actuators have four operational modes: Active, Bypassed, Damped, and Blocked. From an article by Gregg F. Bartley, Boeing “Boeing B-777: Fly-By-Wire Flight Controls”:

    “An example using the elevator surface illustrates how these modes are used. If the inboard actuator
    on an elevator surface fails, the ACE controlling that actuator will place the actuator in the “Damped”
    mode. This allows the surface to move at a limited rate under the control of the remaining operative
    outboard actuator. Concurrent with this action, the ACE also arms the “Blocking” mode on the outboard
    actuator on the same surface. If a subsequent failure occurs that will cause the outboard actuator to be
    placed in the ‘‘Damped” mode by its ACE, both actuators will then be in the ‘‘Damped” mode and have
    their ‘‘Blocking” modes armed. An elevator actuator in this configuration enters the ‘‘Blocking” mode,
    which hydraulically locks the surface in place for flutter protection.”

  281. Oleksandr says:

    @Victor,

    I think you have already convinced yourself that the flight simulator proves “Z. did it”, so arguing is useless. As discussed, I can accept it a working hypothesis without a proof, but I strongly disagree with your statement that there is some evidence.

  282. Victor Iannello says:

    @David said, “Clearly your opinion and others is that a search in the Review recommended area is less likely to be successful than moving it further north east.
    Since it is unlikely there will be a new search until further confidence is gained, one could argue that the new information needed for that will decide which the new area should be, so there is no harm in remaining passive pending that. Is that how you see it?”

    I haven’t made a public statement as to where I think the plane is. My criticism of CSIRO’s work was related to their drawing conclusions which seem to be inconsistent with their own numerical results, and the precision they claim is possible using their drift models. I also question their assertion about the efficiency of the aerial search, which was an important part of their analysis and conclusions.

    My post was meant to encourage a scientific discussion about their conclusions.

  283. Gysbreght says:

    @Victor Iannello: “… this paper shows the simulator data is all from the same simulated flight.”

    I dont’t see that it does. The paper makes the assumption that the data are all from the the same flight, and builds the McMurdo path on that assumption. The two file fragments at 45S might well be from one simulation, but there is nothing that connects them to the other points.

  284. Andrew says:

    @Gysbreght

    RE: “The flight control actuators have four operational modes: Active, Bypassed, Damped, and Blocked. From an article by Gregg F. Bartley, Boeing “Boeing B-777: Fly-By-Wire Flight Controls”:”

    The PCUs for the various flight controls are not all the same. That article refers to the ELEVATOR PCUs, which are different to the FLAPERON PCUs. The FLAPERON PCUs only have two modes: normal and bypass.

  285. Mick Gilbert says:

    @TBill

    Thank you for that estimate. A cabin temperature of 15°C – 20°C (5°C – 10°C cooler than normal) is a lot warmer than I would have expected for a depressurised airplane at 35,000 feet.

    @Oleksandr

    I know that a multitude of variables might come into play but what do you think might be a ball park figure for cabin temperature say 20 minutes after a rapid depressurisation (depressurisation completed in say 3 minutes) at 35,000 feet with the passenger cabin master reference temperature set at say 25°C?

  286. Gysbreght says:

    @Andrew: The example is for the elevator actuators, the preceding text is more general:

    11.5 Control Surface Actuation
    11.5.1 Fly-by-Wire Actuation

    The control surfaces on the wing and tail of the 777 are controlled by hydraulically powered,
    electrically signaled actuators. The elevators, ailerons, and flaperons are controlled by two actuators
    per surface, the rudder is controlled by three. Each spoiler panel is powered by a single actuator.
    The horizontal stabilizer is positioned by two parallel hydraulic motors driving the stabilizer jackscrew.
    The actuation powering the elevators, ailerons, flaperons, and rudder have several operational modes.
    These modes, and the surfaces that each are applicable to, are defined below.”

    Then follow the definitions of modes Active, Bypassed, Damped, and Blocked, and the elevator surface example.

  287. Mick Gilbert says:

    @Victor

    While we sit on different sides of the table regarding who we think did what when, I have never ruled out the possibility of deliberate malicious action by a flight crew member. With regards to the Captain’s flight sim data, given that the “flight of interest” appears to neither navigate via waypoints nor accurately model fuel consumption what value might have been gained from it for “planning” purposes?

  288. Andrew says:

    @Gysbreght

    Just in case you don’t believe me, try reading the paragraph immediately above the one you quoted, where it explains which modes are used by which PCUs:

    ActiveNormally, all the actuators on the elevators, ailerons, flaperons, and rudder receive commands from their respective ACEs and position the surfaces accordingly. The actuators will remain in the active mode until commanded into another mode by the ACEs.

    Bypassed—In this mode, the actuator does not respond to commands from its ACE. The actuator is allowed to move freely, so that the redundant actuator(s) on a given surface may position the surface without any loss of authority, i.e., the actuator in the active mode does not have to overpower the bypassed actuator. This mode is present on the aileron, flaperon, and rudder actuators.

    Damped—In this mode, the actuator does not respond to the commands from the ACE. The actuator is allowed to move, but at a restricted rate which provides flutter damping for that surface. This mode allows the other actuator(s) on the surface to continue to operate the surface at a rate sufficient for airplane control. This mode is present on elevator and rudder actuators.

    Blocked—In this mode, the actuator does not respond to commands from the ACE, and it is not allowed to move. When both actuators on a surface (which is controlled by two actuators) have failed, they both enter the ‘‘Blocked” mode. This provides a hydraulic lock on the surface. This mode is present on the elevator and aileron actuators.”

    NB. Active mode is the same as Normal mode.

  289. Mick Gilbert says:

    @Gysbreght

    Re: “Then follow the definitions of modes Active, Bypassed, Damped, and Blocked, and the elevator surface example.”

    Read the definitions for Damped and Blocked, you’ll find they end with the sentences “This mode is present on elevator and rudder actuators.” and “This mode is present on the elevator and aileron actuators.” respectively.

  290. Victor Iannello says:

    @Gysbreght: “The two file fragments at 45S might well be from one simulation, but there is nothing that connects them to the other points.”

    This is explained in great detail in the paper.

    The parameters MaxReachedEngineRPM1 and MaxReachedEngineRPM2 are exactly the same value of 32968.9 rpm for data sets 10N in the Andaman Sea and data sets 45S1 and 45S2 in the SIO. Similarly, MaximumGForce is 2.2032 g and MinimumGForce is 0.1453 g for all three data sets. These consistent and exact values link data sets 10N, 45S1, and 45S2. Data set 10N also shows a plane that is banking to the left in a turn towards the south.

  291. Gysbreght says:

    Blocked – In this mode, the actuator does not respond to commands from the ACE, and it is not
    allowed to move. When both actuators on a surface (which is controlled by two actuators) have
    failed, they both enter the ‘‘Blocked” mode. This provides a hydraulic
    lock
    on the surface. This
    mode is present on the elevator and aileron actuators.”

  292. Gysbreght says:

    @Andrew: OK, I believe you.

  293. Andrew says:

    @Gysbreght

    Very clever, but if you bothered to read my earlier post you would see that I was referring to the FLAPERON PCUs. I’ll say it again: THE FLAPERON PCUs HAVE TWO MODES; NORMAL AND BYPASSED.

  294. Gysbreght says:

    @Victor Iannello: “The parameters MaxReachedEngineRPM1 and MaxReachedEngineRPM2 are exactly the same value of 32968.9 rpm for data sets 10N in the Andaman Sea and data sets 45S1 and 45S2 in the SIO. Similarly, MaximumGForce is 2.2032 g and MinimumGForce is 0.1453 g for all three data sets. These consistent and exact values link data sets 10N, 45S1, and 45S2.”

    When are those max/min data generated, and why do they have to be different between simulations?

  295. Victor Iannello says:

    @Mick Gilbert said, “With regards to the Captain’s flight sim data, given that the “flight of interest” appears to neither navigate via waypoints nor accurately model fuel consumption what value might have been gained from it for “planning” purposes?”

    First, the data sets show that waypoints were indeed entered in the simulation. For instance, point 3N is precisely on airway R467 and point 5N is on airway B466 at the Malaysia-Indonesia FIR boundary, and with a track (not heading) that is exactly towards waypoint VAMPI on N571. Similarly, the great circle path that connects points 5N and 45S1, when extended past 45S1, aligns exactly with McMurdo Station, Antarctica.

    The flight might have been a simulation of a flight from KLIA to Jeddah with a diversion, as the fuel loading and SID are consistent with that. Also, Zaharie Shah was scheduled to captain MH150 to Jeddah on Feb 4, 2014. The Shadow Copy Set containing the deleted flight file fragments was dated Feb 3, 2014.

    As for the motivation for creating the simulation, I can only speculate. As he was a pilot, I understand that he would have better tools available for navigating and fuel planning than the PSS 777 model. The fact that we can’t be sure as to why he created the simulation does not alter the fact the he created a flight simulation from KLIA, up the Malacca Strait to the Andamans, and then south into the SIO until fuel exhaustion. That MH370 also is believed to have flown up the Malacca Strait and then turned to the south until fuel exhaustion is too great of a coincidence for me (and many others) to ignore.

    [An earlier version of this comment incorrectly attributed the quote to another commenter.]

  296. Victor Iannello says:

    @Gysbreght: Your questions are answered in the paper, which I co-authored with recognized MSFS expert Yves Guillaume.

  297. Victor Iannello says:

    @Andrew: You were very clear in your words and your citation. Sometimes the best we can do is to present the facts.

  298. Peter Norton says:

    > lkr:
    > “it’s insane to suggest tossing any of the scarce recovered debris
    > back in the ocean for drift studies. These pieces have forensic
    > utility — not as courtroom exhibits or trophy-wall pieces at
    > Le Inspector’s office, but basic, how did it come apart, work.

    a) drift studies ARE forensic work, no ?
    b) They had almost 2 years to do forensic work since july 2015. Not enough time ?

    > And better to do a sea run with a dozen stand-ins than that one original.
    > Best would be to visit a aircraft bone-yard for multiple flaperons

    Agreed. As I said, an old spare part would still be 100x better than any replica.

  299. Andrew says:

    @Victor

    Thank you!

  300. Gysbreght says:

    @Victor Iannello: Your paper doesn’t explain when or how these min/max values are generated or why two flights cannot have the same values.

    The values for simulations 3N and 5N (red box in your table 2) are different from those in 10N and 45S (blue box).

    It would indeed be strange to see identical values for these parameters in a recording for a real airplane, even for the two engines on that plane. But we are not looking at real flight recordings, but at data produced by simulation software. Run that software at the same conditions twice you should see exactly the same values.

    You say in your paper: “As such, all the recorded values of N2 are too high. Nonetheless, the values can still be used as flight markers.” Isn’t that somewhat too convenient?

    As to the Max/Min GForce data, these are also too high/low for normal operation of the airplane by a pilot familiar with its control characteristics. They are probably the result of the simulation recovering from the unstable condition that exists when the simulation is paused, changes are made in flight parameters, and the simulation is then resumed. That process should also be repeatable.

  301. Victor Iannello says:

    @Gysbreght: A pilot does not takeoff with the autopilot engaged. Therefore, even if desired, it is almost impossible to create two flights that experience EXACTLY the same minimum and maximum values of certain parameters, even for an automated flight. I have run many simulations and I have never been able to create identical min/max values, even when I try.

    The maximum rated value for N2 rpm for the GE engines in the PSS 777 model is not correct, as explained in the paper. That doesn’t mean that N2 can’t be used as a marker to identify data sets created by the same simulation.

    We’ve had this discussion before months ago. You are prone to propose theories that are demonstrably false, yet you refuse to actually load FS9 or any other version of MSFS and do the tests yourself, despite the fact that I have offered to pay for the cost of the software.

    My polite recommendation is that you continue this discussion offline with the true expert, Yves Guillaume. I found Yves to be polite, knowledgeable, and generous with his time. Perhaps you can persuade him that you understand MSFS better than he does. Please report back with your findings.

  302. Oleksandr says:

    @Victor,

    You have connected “cherry-picked” samples from the deleted files recovered by FBI from the hard disk. I had an impression that Malay Police and FBI have previously concluded that there is no evidence that these samples belong to a single flight, but not to various flights among hundreds recovered. I recall BG have made a similar statement. In addition, it is not clear why would Malay Police and FBI kept silence for 2 years, if they believed these samples came from the same simulation flight. Also, why was this information released to the public at the very special time? In connection with the latter, I doubt this information is genuine at all.

    I do not understand why you consider this as the evidence of Z.’s involvement.

  303. Gysbreght says:

    @Victor Iannello: Fair enough.

  304. Victor Iannello says:

    @Oleksandr: You accuse me of cherry picking samples from the deleted files recovered by the FBI from the hard disk. This is a blatant lie. I have analyzed EVERY data set that was presented in the report. Since you claim I have cherry picked the data, please produce other data sets in that report that I have ignored. If you cannot, it is proof of your lie.

    The deleted file sets were the ONLY files recovered from the Shadow Volume of the solid state drive (found disconnected from the computer). The other data sets were NOT in the Shadow Volume, and the other data sets were not part of the report. The files of interest were deleted and later recovered by investigators from the Shadow Volume on a single drive that was found disconnected from the computer. This makes the set unique among all the flight files found.

    As for statements made by the Malaysians, if you are willing to believe all statements made by the Malaysians, you should continue to search for MH370 in the South China Sea. There has been a consistent pattern of misinformation, disinformation, and non-disclosure from the Malaysians, including hiding the simulator data and also the cell phone connect at Penang.

    You doubt that the Malaysia police report is genuine. At this point, even the Malaysians don’t refute that the Malaysian police report is authentic. The report was made available to French authorities, who requested it as part of their criminal investigation into MH370. The report was then leaked to French media, and eventually was made public.

  305. Gysbreght says:

    @ictor Iannello: I think the Malaysian police ‘cherry picked’ the data they thought might interest thew inestigation. The shadow volume should have contained more than six fragments of deleted files.

    On the previous subject: A simple way to create different simulations containing identical Min/Max values is to start the simulations from the same .FLT file. That’s what someone would if he was experimenting with the software to diagnose a problem with it.

  306. Gysbreght says:

    Apology: I’m haing problems with the “V” key on my keyboard.

  307. Victor Iannello says:

    @Gysbreght: You are making a claim that there were more flight file fragments in the Shadow Volume that were deliberately omitted from the report by the Malaysians. You have ZERO basis for this claim.

    As for starting the simulation from identical FLT files, I think you are claiming that 10N, 45S1, and 45S2 come from a common parent FLT file, but that 10N is not the parent of 45S1. Again, you have ZERO basis for this claim, as no other file fragments were found. Also, the alignment of 10N, 45S1, and McMurdo, and the banked left turn at 10N towards the south, would all be an incredible coincidence.

  308. Ge Rijn says:

    @VictorI @all

    Does anyone know if Zaharie was scheduled on another flight to Jeddah or Europe after 4 februari till 8 march or some time after?

    I mean, could he have had the opportunity to execute a flight according to the simulator-coördinates in this periode?

    If not, what could be the implications on improvising on a flight to Bejing with a final leg heading to McMurdo after FMT?

    Wouldn’t the FMT had to be made a lot earlier than 10N/90E?

  309. Peter Norton says:

    on the subject of ELTs:

    Mick Gilbert: “unlike a maritime EPIRB, this type of ELT is not designed to float”

    Please correct any errors:

    • The portable ELT is designed to float.
    So a sunken portable ELT could float back to the surface and transmit.

    • Before sinking (or if remaining attached to a floating piece of fuselage but fuselage pieces normally don’t float), the fixed ELT can transmit via external antenna (EA) provided the connection is not severed or via attached backup-antenna (BA) if present. The FI report implies that the fixed ELT lacked a BA. I assume the BA connector on fixed ELTs is normally left unused/empty ?

    Oleksandr: “[…] why no signal was received: due to attenuated signal, or because the portable ELT was simply not activated. Both the ELTs are satellite, and VHF-enabled.”

    @VHF:
    FI report says: “Transmissions on VHF frequency (121.5/243 MHz) are line of sight and effective only in close proximity. […] there may not be such aircraft within [VHF] range at the time of beacon transmission and monitoring 121.5 MHz.”

    @satellite:
    As per the manual, the portable ELT has 2 connectors (EA and BA):
    – BA doesn’t seem to work inside the aircraft: Cospas-Sarsat says an ELT “with its antenna blocked by the body of an aircraft […] is unlikely to be received by the satellites” (see above).
    – Is the portable ELT also connected to the external antenna (while attached to its holding bracket)?

  310. Victor Iannello says:

    @Ge Rijn: ZS did not fly to Jeddah after Feb 4, 2014. He did fly to Beijing on Feb 21, 2014. I don’t know his schedule after March 8, 2014. He also flew to Melbourne on Feb 26, 2014, where there were reports that he met with his daughter for the last time.

    We don’t know for sure whether MH370 was headed towards McMurdo Station, crossing the 7th arc at 27S latitude. This is only a theory based on the path suggested by the simulator data. Richard Godfrey thinks the drift data is a better fit with Wilkins Runway, Antarctica, crossing the 7th arc near 30S latitude.

  311. Ge Rijn says:

    @VictorI

    Then my next question is; where would a Wilkins Runway flight after FMT crossing the 7th arc near 30S latitude place the ~FMT?

  312. TBill says:

    @Oleksandr
    You are correct about slow depressurization being not much temperature loss.

    Basically 3 modes of depressurization:
    (1) Explosive, extremely rapid pressure loss (Adiabatic)
    Final Air Temp T2 = T1 x (P2/P1) exp ((1.4-1)/1.4) = -87 deg F

    (2) Rapid depressurziation
    Final Air Temp T2 = T1 x (P2/P1) exp ((1.2-1)/1.2) = -27 deg F

    (3) Slow Depressurization (Isothermal)
    Final Air Temp T2 = T1 x (P2/P1) exp ((1.0-1)/1.0) = T2 = T1 = 70 deg F
    No Change.

    In Cases (1) and (2) the cabin air will start to warm back up right away because the cabin mass itself (seats/hardware/PAX) will be still at room temperature.

    The unknown issue would be how well the A/C system works at low pressure. Or if the A/C (Bleed air is off) then the cabin temps will trend down due to loss of heat to the surroundings.

  313. Gysbreght says:

    @Victor Iannello: “@Gysbreght: You are making a claim that there were more flight file fragments in the Shadow Volume that were deliberately omitted from the report by the Malaysians. You have ZERO basis for this claim. “

    I’m not making any claims. I’m reminding you that what you are adopting as the basis of your claim is just one of many possibilities.

    A shadow copy contains a snapshot of the entire contents of a drive at the time the shadow copy is created. Wouldn’t it be strange if at the creation of the shadow copy the drive contained only six fragments of deleted files, and that ‘hundreds’ of other simulation files on that drive were created after the shadow volume date and time?

  314. TBill says:

    @Gybreght
    Re: Z Flight Sim Studies
    I think I have have previously suggested the G-Forces and FL400 altitude at 1090E in the simulator studies suggest to me a possible intentional depressurization manuever similar to what might have in reality happened at IGARI. Sharp descent from FL400 could account for the G-forces. I do not know if Victor would agree with me on that.

  315. TBill says:

    PS- I did the FS9 climb/descent cases a while ago, but if I recall it’s the leveling off after sharp descent that could generate the G-Forces. Originally I was thinking steep ascent might be the cause.

  316. Gysbreght says:

    @TBill: The file fragment “10N” at 40003 ft, still climbing at 59.5 fps already had Max/MinGForce 2.2032/0.1453

  317. Victor Iannello says:

    @Gysbreght said, “A shadow copy contains a snapshot of the entire contents of a drive at the time the shadow copy is created.”

    No, that’s wrong. The shadow copy does NOT contain a snapshot of the entire contents of a drive. The shadow volume makes a copy of the contents of a block prior to the first change to that block to give the capability to revert to a prior state of the drive. Unchanged blocks are not backed up. If an unchanged block of data gets corrupted, for instance, there is no way to correct this problem. Therefore, there is no reason for prior flight files to be in the Shadow Volume unless they were written over.

  318. Victor Iannello says:

    I should add, for drive MK25 with the deleted flight file fragments, the date of the Shadow Volume was Feb 3, 2014. I said, “there is no reason for prior flight files to be in the Shadow Volume unless they were written over.” Really, I should have said, “There is no reason for prior flight files to be in the Shadow Volume unless they were written over on or after Feb 3, 2014.”

  319. Gysbreght says:

    If Captain Z. wanted to simulate a diversion from a normal flight to Jedda (or whatever) he wouldn’t pull up to 1.44 g and push down to 0.6 g before engaging the the autopilot, and the autopilot would not produce those values, nor the values in the “10N” fragment.

    If the values for MaxReachedEngineRPM were obtained in the simulations at takeoff, then the takeoff for fragments “3N” and “5N” would have been at a lower ambient temperature than the takeoff for fragment “10N”.

  320. Gysbreght says:

    @Victor Iannello: Please provide references for your posts of 1:20pm and 1:42pm, including evidence that this applies to Captain Z.’s HDs.

  321. Victor Iannello says:

    @Gysbreght said, “If Captain Z. wanted to simulate a diversion from a normal flight to Jedda (or whatever) he wouldn’t pull up to 1.44 g and push down to 0.6 g before engaging the the autopilot, and the autopilot would not produce those values, nor the values in the “10N” fragment.”

    This is a simulation, not a true flight. You have no idea how he flew the plane, and what manual inputs he might have provided, between 5N and 10N.

  322. Victor Iannello says:

    @Gysbreght: Read the RMP report, including the translated Preliminary Examination Report from Folder 1. Then read about how Shadow Volumes work for Windows 7. It’s all available for you if you look.

  323. Oleksandr says:

    @Victor,

    Re: “You accuse me of cherry picking samples from the deleted files recovered by the FBI from the hard disk.”
    I am not accusing anyone in anything. Please show me where I accused you in this.

    Re: “This is a blatant lie.”
    What is blatant lie? It was you, who accused me in stating that there is no evidence that the Captain has hijacked the plane.

    Re: “I have analyzed EVERY data set that was presented in the report. Since you claim I have cherry picked the data, please produce other data sets in that report that I have ignored. If you cannot, it is proof of your lie.”

    Why would I need to do this? What I am asking is only to present a proof that the samples you took belong to the same simulation session. Can you produce such a proof or not? RMP has concluded there is no such evidence. You stated there is such evidence. Please show it.

    Re: “…has been a consistent pattern of misinformation, disinformation, and non-disclosure from the Malaysians, including hiding the simulator data and also the cell phone connect at Penang.”

    I quite disagree with your assessment of Malaysians. There is not more misinformation, disinformation, and non-disclosure than in other countries. Perhaps Malaysia is even more open in this regard than US, Australia or France. Anyway, your personal mistrust does not make any contribution to support your accusations of the Captain.

    In summary, if you are accusing the Captain, you need to present solid evidence. If you have it, please present.

  324. Gysbreght says:

    @Victor Iannello: According to Microsoft there are three ways in Windows 7 that an application can use the Windows Volume Shadow Copy Service (VSS):
    – Complete copy
    – Copy-on-write
    – Redirect-on-write

    Where does the RMP report state which of these methods was used by the MSFS application at the time the Shadow Volume was created?

  325. Joseph Coleman says:

    @Gysbreght

    If you want to produce an exact copy of a hard disk partition, you have to use image writing software such as Norton Ghost or equivalent, that copies and writes to, it’s contents into a single compressed file to another hard drive of an equivalent size. That single compressed file can be uncompressed and copy written using the same software that compressed it to yet another hard drive or overwrite the original partition from which it was copied from.

    Please note that whatever files were on the previous hard drive whether on multiple partitions or just one, can be recovered if they are not written over. The more you rewrite a hard drive the lesser chance of recovering anything from any previous writings.

    For Multiple users for example a call centre, if the multiple computer software such as Windows is faulty or beyond repair, then the same image can be uncompressed to any computers partition either through manual plugin or over its Network.

  326. Victor Iannello says:

    @Oleksander: “In summary, if you are accusing the Captain, you need to present solid evidence. If you have it, please present.”

    I have presented factual evidence that the deleted files found on the MK25, the only deleted files recovered from that drive and found in the Shadow Volume, were related to the same simulation. Contrary to your claims, there was zero cherry-picking. Read the papers I have provided.

    If you believe the Malaysians have been forthcoming providing truthful, basic data related to the MH370 disappearance, including radar data, simulator data, and the cell phone connect, then you and I see the world too differently to have a productive discussion.

  327. Oleksandr says:

    @Victor,

    Here is the statement from BG, who apparently also had access to the full RMP report:

    https://www.dropbox.com/s/v7kd29rchmus5bm/Setting%20the%20record%20straight.doc?dl=0

    Malaysian officials also dismissed this “evidence” after the original publication, which was nothing but a product of the media. For example:

    http://www.smh.com.au/world/mh370-search-no-evidence-a-rogue-pilot-hijacked-and-crashed-plane-malaysia-20160804-gqlhoi.html

    or this:

    http://time.com/4421357/malaysia-airlines-mh370-missing-jet-china-australia-suicide-evidence-captain-zaharie-ahmad-shah/

  328. Oleksandr says:

    @Victor,

    “If you believe the Malaysians have been forthcoming in providing truthful, basic data related to the MH370 disappearance, including radar data, simulator data, and the cell phone connect, then you and I see the world too differently to have a productive discussion.”

    You are right: we see the world too differently. Should this happened in the US, you would not even know about the radar data or cell phone connect.

  329. Victor Iannello says:

    @Gysbreght: Are you are seriously suggesting that Microsoft Flight Simulator 2004, aka FS9, implements an advanced VSS using complete copy? If that is your belief, than it is up to YOU, not me, to prove this. I believe the Shadow Volumes were created using the standard backup and restore features of Windows 7, which use incremental backup technology to save space on the drive, and nothing more extraordinary.

  330. Gysbreght says:

    @Victor Iannello: I’m not “seriously suggesting” anything. I’m asking you to provide the evidence that support your claims.

    You are systematically reading more in the RMP report than I do. Where does it say that the 6 or 7 file fragments containing “interesting” coordinates were the only fragments of deleted files found on the MK25 HDD?

    From the PRELIMINARY EXAMINATION REPORT:
    “The results of all inspections on the hard drive of the computer, it was found that there were 668 file types * .FLT whole. The highest number is in Mk25 hard drive which there were 348 files * .FLT. Followed by the hard drive MK26 112 files, hardsik MK23 K24 of 106 files and 105 files. Many of files * .FLT this is a file that exists during the installation of the game , -., Microsoft Flight Simulator X and it’s not necessarily the file that is created from The game was played after installation
    * .FLT This file containing the configuration and location coordinates flight game in play during the game at the save. The assessment has found 2 coordinate the approaches available MH370 search location. in Altogether there are 7 including 2 coordinate coordinate earlier believed relate to each other. This is based on the name of the aircraft stated in its configuration is the same, “PSS Boeing 777-200LR Malaysia No VC “. All coordinates were found in a file Volume Shadow Information (VSI) named {OOd7ef6c-8bcb-1Ie3-b3f7-ee8a9181afad} {3808876bc176- 4e48-b7ae-04046e6cc752} dated February 3, 2014. This is the VSI File a file that stores information when a state computer the computer is in an unused or ideally more than 15 minutes. Therefore, it can not be ascertained that all the 7 coordinate is from * .FLT the same file. (Sita refer to the appendix “Coordinates”)”

  331. Oleksandr says:

    @Victor,

    Can you pls. comment of the following BG’s comment on RMP report:

    “The official Royal Malaysia Police document concludes there was no premeditated deliberate act.”

    Can you also comment on the comments from the Malaysian officials that there were thousands routes recovered from the simulator, but only a few of them were of interest.

  332. Joseph Coleman says:

    @Gysbreght

    If the file system is corrupt and can’t be accessed or written over i.e. accidental format, and you can’t access the files you require then you can use recovery software on another system to recover these files from the hard drive (external plugin) if they haven’t been already been writer over but the original file names are not present. The recovery software usually recognises what type of files (if they haven’t already been written over then it won’t recognise any) they are but the original file names can’t be identified. So the recovery software names them for you, usually in an order preset by the software.

  333. Victor Iannello says:

    @Oleksandr: Read the data in the RMP report for yourself, as I have done. It’s now publicly available. Analyze the data in detail, as I have done. Then you might have a solid foundation for making statements on this matter. Mindlessly repeating statements made by Malaysian officials and reported by the media is meaningless. You should know better.

  334. Victor Iannello says:

    @Oleksandr: BG made a statement based on the conclusions he found in the RMP report. I base my comments on a detailed review and analysis of the data. If BG has analyzed the data, he hasn’t made his analyses public. He is repeating phrases he found in the report. I don’t fault him. But you have to understand the context of his statements. Also, as a lawyer, he thinks in terms of provable guilt in a court of law. I don’t think there is enough evidence to convict ZS. I do believe, however, there is evidence linking him to a crime.

    As for the “thousands of routes” found on the computer, I will once again repeat what I keep stating, which is summarized in the introductory paragraph of the paper I wrote with Yves, which says:

    Recently, the MH370 Independent Group (IG) released a preliminary assessment of data recovered from Captain Zaharie Shah’s home computer that were related to the Microsoft Flight Simulator(MSFS)game [1]. The data are in the form of fragments of “flight files”, which a user may create during a game to record the state of a game for future reference or to resume play at a future time. The particular flight files of interest were one of hundreds found on the computer on several drives; however, the files of interest were deleted and later recovered by investigators from a “shadow volume” on a single drive that was found disconnected from the computer. This makes the set unique among all the flight files found.

    There were no other deleted files found in the Shadow Volume of drive MK25. There was no cherry-picking. I don’t know how to be more clear.

    Have you bothered to read the RMP report? Or the papers I have referenced? Based on your comments, I suspect you have not.

  335. DennisW says:

    @Oleksandr

    “You are right: we see the world too differently. Should this happened in the US, you would not even know about the radar data or cell phone connect.”

    The Malays have never corrected the radar data (the Lido Hotel slide is all we have), and did not share it with the ATSB. The cell phone connect was never confirmed by the Malays either. It was made public by a leaked report as was the simulator data. No, the Malays truly have behaved poorly and obstructively relative to the MH370 incident.

    As far as the questions you and Gysbreght raise relative to the simulation data points, I put them (the questions) in the “who cares” category. The important thing is the coordinate info and the timing. Your arguments are like a defense attorney claiming his client’s fingerprints on the murder weapon are irrelevant since the client’s fingerprints are on thousands of other things as well.

  336. Victor Iannello says:

    @Gysbreght: It says, “Altogether there are 7 including 2 coordinate coordinate earlier believed relate to each other. This is based on the name of the aircraft stated in its configuration is the same, “PSS Boeing 777-200LR Malaysia No VC “.

    If you want to interpret this as the 7 files were among thousands found in the Shadow Volume with the PSS B777 label, you go right ahead in believing that. I won’t argue any more with you. If you want to believe that FS2004 was written with calls to advanced VSS functions beyond the standard backup for Windows 7, again, you go right ahead in believing that. If you want to believe that points 10N, 45S1, and 45S2 are not related, despite common min/max marker values, an alignment of 10N, 45S1 with McMurdo, and a banked turn at 10N towards the south, again, you go right ahead in believing that. My goal here is not to persuade you.

  337. lkr says:

    @peter,

    Back to my saying it would be “insane” to pitch your best pieces back in the ocean. [I’m not sure if anyone really believes this would have been a smart thing to do — on this forum at least.]
    –It should have been clear that by “forensic” I was referring to lab tests, and materials analysis of the flaperon, flap, and perhaps the tailplane pieces, are probably the best hope to differentiate between high-speed and low-speed entry. That analysis shouldn’t be seen a once and done matter — for instance we now have it that many smaller pieces failed under tension…

    What caught my eye in seeing this is that the plywood and metal “models” give a sense of how low-priority the drift studies were in the first 2 years. This bodging is what you do for a trial run when no one will give you a budget line to do it right with real aircraft components.

  338. Gysbreght says:

    @Victor Iannello: You said today at 8:55 am:

    “Also, Zaharie Shah was scheduled to captain MH150 to Jeddah on Feb 4, 2014. The Shadow Copy Set containing the deleted flight file fragments was dated Feb 3, 2014. “

    If your assumption is correct regarding how MSFS uses Windows VSS, my understanding is that the Shadow Copy Set date means nothing more than that fragments of flight files that were created on unknown date X, and were deleted on unknown date Y, were overwritten with more recent data on Feb 3, 2014. Are you suggesting that there is a connection between unspecified use of the simulator on Feb 3 that resulted in fragments of previously deleted files being overwritten, and Z. being scheduled for the Jeddah flight on Feb 4?

  339. Victor Iannello says:

    @DennisW: The problem with the simulator data is that if we interpret it on its face value, it is so incriminating that it leaves little room for anything but the guilt of ZS.

    When I first learned of the simulator data, I predicted that there would be two camps of people: Those that believed the data was unaltered and ZS was guilty, and those that believed the data was fabricated to frame ZS. I did not foresee that there would be some that would believe the data is unaltered AND the evidence does not suggest ZS is guilty.

  340. Victor Iannello says:

    @Gysbreght: I am suggesting what I wrote.

  341. Gysbreght says:

    @Victor Iannello: I regret that you tend to withdraw when the going gets rough. Not very scientific.

  342. Victor Iannello says:

    Oleksandr said, “You are right: we see the world too differently. Should this happened in the US, you would not even know about the radar data or cell phone connect.”

    I was not trying to compare the honesty of the US government with Malaysia. But since you brought it up, let’s compare the data available from 9/11 to what is available from MH370. All radar data for the four hijacked planes, including primary and secondary radar, is available in the public domain as KMZ files for use in Google Earth. And the cell phone connects from passengers on the planes have also been publicly released. Your statement is therefore patently false.

    Perhaps you should consider whether you wish to continue participating in the discussion here.

  343. Victor Iannello says:

    @Gysbreght: I am tired of explaining to you how your theories are ridiculous. I was patient and polite. I am no longer patient.

  344. TBill says:

    @Oleksandr
    “Should this happened in the US, you would not even know about the radar data or cell phone connect.”

    That is probably unfair to USA/NTSB. First of all, in all likelihood, USA has air marshalls, operations policies, and air defenses that never would have allowed this event to take place in 2014. If it did happen in USA, we would expect open, impartial and thorough investigation/disclosure of likely causes by the NTSB.

  345. Gysbreght says:

    Changing the subject – did anyone notice that the NTSB is the source of the graphic that shows the MH370 trajectory “while the aircraft was being tracked solely by air defense radar systems”, published 26 May 2014 as Figure 1 in ATSB Fact Sheet “Considerations on defining the search area”?

  346. RetiredF4 says:

    It is always amusing how scientists can get into agreement to disagree on details based on facts and attenuated by personal assumptions. But that said as a sidenote I’m interested in an explanation (closer to fact than fiction), why Z would have spent hours in front of an electronic box to simulate a flight to nowhere, when all he had to do is use his expierience from decades flying air transport aircraft. Not that he planned to land on the moon, according to the discussed outcome he planned to crash.

    I posed that question before, and I can’t remember any usefull answer yet.

  347. Gysbreght says:

    @RetiredF4: Be careful!

  348. Mick Gilbert says:

    @TBill

    Thank you for the three depressurisation scenarios; that’s an interesting set of results.

    @Peter Norton

    Correct, once an ELTA ADT406 ELT sinks it’s almost certainly not coming back to the surface. Even the DME SRB-10 ELTs used in the life rafts don’t float per se, they are attached to the life raft by a harness and the inflated raft provides the floatation. The SRB-10 itself is water-proof but doesn’t float.

  349. DennisW says:

    @RF4

    Why would Z have a flight simulator at all? I don’t know of any bus drivers who have bus driving simulators. What percentage of airline pilots have a flight simulator game? For Z it was a hobby that obviously allowed him to create and rehearse flight scenarios. I find it perfectly natural for a man with that mindset to rehearse the flight to the SIO.

  350. Peter Norton says:

    Mick Gilbert: “Correct, once an ELTA ADT406 ELT sinks it’s almost certainly not coming back to the surface. Even the DME SRB-10 ELTs used in the life rafts don’t float per se, they are attached to the life raft by a harness and the inflated raft provides the floatation. The SRB-10 itself is water-proof but doesn’t float.”

    Ok, but we do agree that MH370’s portable ELT (ELTA ADT 406 AP) is designed to float, right ?
    So a sunken portable ELT could float back to the surface and transmit.

    With regard to my previous posting above:
    MH370’s fixed and portable ELT both have 2 connectors (1 for the external antenna, 1 for the attached backup-antenna). Is it standard procedure – and was this the case for MH370 – that
    • on the fixed ELT only the EA connector is used, while the BA connector is unused/empty ?
    • on the portable ELT only the BA connector is used, while the EA connector is unused/empty ?

    Or are they connected to both antennas simultaneously ?

    Something else:
    I have heard numerous “aviation experts” say that the ELTs activate upon both sudden deceleration and contact with water. But judging from the manuals, the latter is untrue. (Only the slide raft mounted ELTs have water sensors.)

  351. sk999 says:

    Gysbreght asks: “… did anyone notice that the NTSB is the source of the graphic that shows the MH370 trajectory ‘while the aircraft was being tracked solely by air defense radar systems’, published 26 May 2014 as Figure 1 in ATSB Fact Sheet ‘Considerations on defining the search area’?”

    Yes, and I posted a comment on it not too long ago (somewhere). Aopparently NTSB and FAA personnel went to KL to assist early on in the investigation (and no, I don’t have a reference handy.) In any case, my conjecture at the time was that the radar data were cast into the form of KML files, which is why the same tracks have kept reappearing on Google Earth maps, even though the projections and annotations have changed. Since then I have discovered that the gx:track element can store all the needed information: lat, long, alt, and timestamp. So my current conjecture (sure to be wrong) is that the files were prepared by the NTSB in this form and those are what were made available to the DSTG.

  352. David says:

    @Andrew. “If both PCUs on a flaperon are in bypass mode, the flaperon can move freely; there is no damping mode. If that were the case for MH370 at the end-of-flight, then I’d say the flaperons could flutter”.

    I would like to go into this a little more. In the general there can be control surface and also wing flutter. As you know there is a flutter envelope, which obviously will be outside the flight envelope. Since the left flaperon with PCUs in bypass will be within the aircraft ‘fallback’ system of flight control I think it reasonable to conclude the left flaperon will not flutter within the flutter envelope. This will have been demonstrated as part of certification.

    Beyond that it may be immaterial as to what can be expected in flutter of all flight controls, damped, blocked or, in the flaperons case if entire, with any of two, one or no PCUs operative.

    In summary, there should be no flutter of any controls or of the entire wing inside the flutter envelope, including with RAT deployed; but there could be outside the envelope. Yes the RAT could become inefficient as its blade tips reach the speed of sound (easy to calculate from diameter and rpm) and it could be that beyond that the drag force on it tears the whole thing off, though I notice the Comoros 767 RAT remained attached in that ditching.

    @Ikr. I think it was you who raised assessment of sequence and cause from fracture surfaces of the recovered flotsam. The only assessment of which I am aware by the investigators was the conclusion about flaperon and outer flap positions when they collided. I would have hoped for far more, though I am unsure whose task this is; whether it is integral with the ATSB search delegation or the business of the accident investigator. I would have hoped that all this (including biological) had been done long since and the reason we have such little outcome is that there was nothing to report. Surely the French have studied the flaperon fracture surfaces? To me the debris should be used for experiments to complement the extant drift data but as you say only if all evidence from the debris has been gleaned.

    Some of us here, three-quarters blind, have attempted reconstructions but without confidence or much in the way of results. One reason has been shortage of information on carbon fibre composite failure appearances. For those who would welcome a better feel for that, the following ICAO info is useful: ICAO Manual of Aircraft Accident Investigation Document 9576 Pt 3, pages 111-9-37 to 62, which I downloaded from the web. For any interested, please let me know if there are difficulties with that.

  353. Mick Gilbert says:

    @Peter Norton

    Re: “Ok, but we do agree that MH370’s portable ELT (ELTA ADT 406 AP) is designed to float, right ?”

    No, Peter, absolutely not. The “portable” ELT is exactly the same as the “fixed” ELT sans fixtures. It is not designed to float. If the portable unit sank it could not resurface.

    I don’t know whether it is “standard procedure” per se but your summary of the antenna configurations for the fixed and portable ELTs is correct; the fixed unit was connected to an external antenna and had no back-up antenna and the portable unit only had a back-up antenna.

    The fixed and portable ELTA ADT406 ELTs are activated either according to acceleration/time criteria or manually via a switch (remotely from the flight deck with regards to the fixed unit or on the unit itself with regards to the portable unit). The DME SRB-10 ELTs attached to the life rafts are activated by the water activation switch when the raft is deployed in the water or manually (there’s no ON switch on the SRB-10 for manual activation, that is achieved by, wait for it … touching the water contacts with a moistened finger.)

    It’s also worth noting that while the ELTA ADT406 transits on 121.5 MHz, 243.0 MHz and 406.028 MHz, the DME SRB-10 only transmits on 121.5 MHz and 243.0 MHz.

  354. Andrew says:

    @David

    RE: “In summary, there should be no flutter of any controls or of the entire wing inside the flutter envelope, including with RAT deployed; but there could be outside the envelope.”

    I agree, but some commentators have suggested that control surface flutter could not have occurred under any circumstances. The point I was trying to make is that flutter might have occurred if the aircraft exceeded its design/certification envelope during its final descent. That is especially true of the flaperons, which are not damped or blocked in the event of loss of hydraulics or electrical power.

  355. sk999 says:

    Was MH150 to Jeddah on Feb 4 the original “SIO” target? But failed for some reason?

  356. DennisW says:

    @VictorI

    “When I first learned of the simulator data, I predicted that there would be two camps of people: Those that believed the data was unaltered and ZS was guilty, and those that believed the data was fabricated to frame ZS. I did not foresee that there would be some that would believe the data is unaltered AND the evidence does not suggest ZS is guilty.”

    Yes, I recall that narrative. I am as baffled as you are. In my view, if the sim data is real and not fabricated to frame Z, then it is a slam dunk against Z. I don’t even have broadcast television to distort (or enhance) my view of forensics.

  357. Donald says:

    @Andrew

    Do the EOF fuel exhaustion simulations sans any control inputs result in the aircraft exceeding it’s design/ certification envelope? I’m an aviation novice so please bear with me. I suppose the BFO suggests this is possible?

    Were Exner’s et al. EOF sim trials able to reproduce such rapid descent rates post phugoid? And does Exner claim flutter post fuel exhaustion, or at some other point during the southern leg?

    Regardless, how it was that flutter became an arrived at determination by non-experts in the field is troubling.

    Thanks for your patience.

  358. TBill says:

    @RF4
    I feel there are many possible reasons for flight simulation to the SIO.

    One visually striking option is to show the fight path on Google Earth with the undersea trenches and structures. Then it is possible to check out flight paths and waypoints that aim at certain features, and deep areas. If you think the pilot was navigating toward a specific undersea target, then you are looking for “proof” that the pilot was aware of the bathymetry. PSS77 with common addons had this Google Earth capability, whether or not Z used it this way, I don’t think we know.

    Many other possibilities including using night sky stars and/or planets for navigation, checking out sunrise at crash time, etc. Just rehearsing a new flight route is something some pilots do at home, according to one reference.
    Another expectation is some elements of the MH370 diverted fight may have been rehearsed in commercial flights prior to 8-March.

  359. Andrew says:

    @Donald

    Unfortunately, the detailed results of the simulations conducted by Boeing on behalf of the ATSB have not been released to the public. Nevertheless, the ATSB noted in its report MH370 – Search and debris examination update, dated 2 November 2016: “Some simulated scenarios also recorded descent rates that were outside the aircraft’s certified flight envelope”. According to the ATSB’s BFO analysis, described in the same report, the final rate of descent could have been anywhere between 3,800 ft/min and 25,000 ft/min.

    There are others here who will be more familiar with Mike Exner’s work than me. However, it’s my understanding, based on the following report, that high rates of descent were observed:
    The Last 15 Minutes of Flight of MH370
    I don’t know if he claims anything about flutter.

  360. Victor Iannello says:

    @sk999: Possibly failed or delayed or something else. We have no way to know. The sim data seems to be closer to a diversion from MH150 than MH370.

  361. Donald says:

    @Andrew

    Thanks.

    @Tbill

    Those are all very plausible (likely, imo) reasons. Another reason would be Zaharie rehearsing the event. Suicides are often preceded by simulation rehearsals. These rehearsal enable the individual to lower or eliminate potential barriers to a successful execution of the planned suicide. In a mass murder/suicide of this complexity, there are myriad potential barriers to be worked through in order to ensure success. Not least of which is the hurdle of the fear of death and the necessary attainment of a comfort level commensurate to the commitment and subsequent execution and follow through.

    Had Zaharie not run any simulations, I would have been baffled.

  362. Donald says:

    @Victor

    Of note is that rehearsals USUALLY precede the actual event by hours or days (weeks or months, though not unheard of, are outliers). As you say, we don’t know what happened or did not happen on MH150, maybe cold feet, needing to see his daughter, crew or co-pilot? We will probably never know.

  363. TBill says:

    @sk999 @VictorI
    Re: Jeddah MH150
    I was personally expecting if Jeddah was the intended flight, then it might have been the later flight departure MH168 added to the schedule below. The MH370 choice then fits in as an alternative late night flight, departing after MH168 so there would be less need (in the case of MH370) to go all the way out to 1090E, assuming Indonesian radar coverage ending around midnight. I don’t think the sim file fragments give us time/date of the simulated flight.

    “Malaysia Airlines from 28JAN14 to 29MAR14 is adding 2nd daily Kuala Lumpur – Jeddah service, on board Boeing 777-200ER aircraft. Reservation for the new MH168/169 service is now open.
    MH150 KUL1145 – 1600JED 772 D
    MH168 KUL2230 – 0245+1JED 772 D
    MH169 JED0500 – 1835KUL 772 D
    MH151 JED1730 – 0705+1KUL 772 D
    MH169 begins from 29JAN14 until 30MAR14.
    From 30MAR14, MH will operate 1 daily flight as MH168/169, which means the oneWorld member shifting operational schedule to night-time departure from KUL.”

  364. Brian Anderson says:

    Re ELTs.

    Just to be clear, it is only the 406 Mhz frequency which is monitored by satellite [COSPAS, Sarsat etc].

    121.5 Mhz is no longer monitored by satellite, and in any case it’s use primarily as an aid for location in SAR. The frequency is monitored by many commercial aircraft, while in the air. 243 Mhz is primarily an SAR frequency used by the military.

    The two 406 ELTs on MH370 are designed to trigger upon a deceleration of about 2.5G or more. However, the data transmission protocol delays the first transmission for approximately 50 seconds after triggering, and then continues to transmit the data block of 520mS at 50 second intervals. If the ELT antenna is underwater before the 50 second delay elapses then no signal will be radiated. Repeating the data block every 50 secs is necessary to allow for the possibility of one of the monitoring satellites to be passing [roughly] overhead. A larger delay may be experienced before this occurs. In the case of a crash into the sea, even if the ELT does trigger, the likelyhood of a successful transmission being received is close to zero.

  365. lkr says:

    @Victor,

    Looking at the ASTB drift report again, I can see why they are so adamant in dismissing most of the area between 28S and 35S. They show 5x to 7x coverage of most of that area by SAR. Taken at face value, they need only claim 50% detection on one pass to get 99+%! My question has all along been just what those boxes really mean in effective coverage, and how the nature of the debris field and ocean conditions would affect the apparent coverage. But one can see how this buttressed the confidence, to the end, that MH370 was in the originally defined box.

    Looking again at this map [Fig 4.2], one detail pops out that I missed previously: There is a tiny blue-green-yellow blip [>10% detection probability to 60%+] that lies in a sea of red [95+%] EXACTLY at 30S. Did anyone else notice this? It could be a processing artefact. If real, it would be ironic indeed if a single missed sweep at that latitude sealed the mystery of MH370…

  366. Gysbreght says:

    @Andrew: you said yesterday at 9:39 pm “There are others here who will be more familiar with Mike Exner’s work than me. However, it’s my understanding, based on the following report, that high rates of descent were observed:”

    Mike Exner posted details of one simulation on the JW blog. In that simulation the rate of descent fluctuated around 2500 fpm from the second engine flameout until 6 minutes later. Six minutes after the second engine flameout the bank angle started to increase rapidly and the rate of descent increased steadily to nearly 20,000 fpm at impact with a bank angle of 85 degrees.

  367. RetiredF4 says:

    @Dennis W
    “Why would Z have a flight simulator at all? I don’t know of any bus drivers who have bus driving simulators. What percentage of airline pilots have a flight simulator game? For Z it was a hobby that obviously allowed him to create and rehearse flight scenarios. I find it perfectly natural for a man with that mindset to rehearse the flight to the SIO.”

    You do find it perfectly natural because you are not pilot. Why would it be perfectly natural for a pilot, but not for a bus driver, a truck driver or a train driver, or the operator of a nuclear power plant? In what kind of professions is it perfectly natural to set up a training gadget at home for the work you have to perform routinly in the job? Your statement is pure assumption because it fits into the biased picture. You like to drive motorcycles. Do you have a simulation software on your computer and do you rehearse your planned tour through the california desert?

    Let me go into some detail. The first question is why he posessed a flight simulation software. We do not know wether and when he aquired the software though, he might have bought it on purpose for himself or his kids or to have fun with friends interested in flying. He might have used it to get fit again after his no-fly period due to his back injury. Those home systems are crap for for a professional pilot concerning flying itself, but can be used as procedural trainer, for example flying an instrument departure and a STAR to an ILS Cat III landing. I guess that more young pilots own and use such a system than older and expierienced pilots, as they lack the vast amount of system and procedural knowledge an expierienced pilot like Z had acquired over his years of flying and they are able to prepare themselves for checkrides or landings on new destinations.

    Flying the simulator at home as a hobby could count, but not with a set up for the 777. If he would have it set up to fly a combat aircraft, an aerobatic aircraft, the Space shuttle or anything other I would buy into your argument.

    That brings me to the second question, why did he set up a complete customized computer system to operate the software? Z was a guy who liked to make things perfect, to repair things, to improve things, to make them work better by using his genius while spending only little money. Anytime a friend showed me his flight sim and his flying abilities on this gadgets like taking of from a carrier, enganging into a air to air fight and recovering to the carrier, I could get no kick at all from it. Operating the switches and commanding the functionns of an aircraft with a mouse and a keyboard while staring at a monitor on your desk is lightyears detached from reality, at least when you are familiar with the real thing. That´s not different for an expierienced high hour commercial 777 pilot like Z was. So he started to improve the input and output devices, contacted other flight simmers and improved the set up. He put the screens at positions representing the approximate locations of the real cockpit. All indications like the postings on his FB page and the drawbacks by software or hardware crashes he had point to the fact, that he was not finished yet. The end game could have been a dark room, lighted instruments and a setup like a cockpit with a real pilot seat, control wheel and rudder pedals. Improving the application of the flight sim software in a more aircraft like setup was his real hobby in connection with the sim.

    As to your argument, that the pure existence of the flightsim in conection with the deleted and recovered navigational points are incriminating enough, would make a great amount of Americans to murder suspects.
    You sure have guns, you acquired them, you used them, you cleaned them after they have been fired. In your amunition box several bullets are missing. In your neighbourhood a homicide happens, the victims death caused by some bullets. Along your case of argument you are already in trouble because you had been in the vicintiy, own guns, have the bullets, some are missing and you even tried to hide the evidence that you shot the gun by cleaning it. Now you are in luck, because you are still alive and able to defend yourself, which Z can´t. Because he might be a victim himself.

  368. Ge Rijn says:

    I’m not sure if it makes any sence but I was thinking, taking Wilkins Runway as an end waypoint heading on a great circle route instead of McMurdo.
    Reciprocal starting at Wilkons Runway at 66S/111E passing over the 45S/104E SIM-point(s) where would this route cross the 7th arc and finaly cross the flight path/heading MH370 took crossing the Malacca straight?

    Roughly it seems to me (on Google Earth) this route passes the 7th arc around 30S/99E and ends up crossing the MH370 flight path/heading not far north/west from Nicobar.

    I have no way of making a precise guess so it could be well off.
    It’s more about the idea.

  369. RetiredF4 says:

    @TBill
    “I feel there are many possible reasons for flight simulation to the SIO.

    One visually striking option is to show the flight path on Google Earth with the undersea trenches and structures. Then it is possible to check out flight paths and waypoints that aim at certain features, and deep areas. If you think the pilot was navigating toward a specific undersea target, then you are looking for “proof” that the pilot was aware of the bathymetry. PSS77 with common addons had this Google Earth capability, whether or not Z used it this way, I don’t think we know.

    Many other possibilities including using night sky stars and/or planets for navigation, checking out sunrise at crash time, etc. Just rehearsing a new flight route is something some pilots do at home, according to one reference.
    Another expectation is some elements of the MH370 diverted fight may have been rehearsed in commercial flights prior to 8-March.”

    Your arguments, viewed from a pilots perspective, hold no water.

    -visualizing routing on google earth
    Is that the way pilots operate their aircraft? Using google earth? Air transport aircraft are equipped and operated to reach the choosen destination with a precision, that they hit the numbers on the tarmac without visual reference at all within a timeframe of p+/- one minute, provided ATC will play along. What would Google earth help in matters, what the onboard systems could not perform?
    To look up deep undersea spots you need no flight sim, you can do it in goolge earth itself, if you wish to visualize such a spot. You can also do a search for deep sea spots. To flightplan to such a spot just note the coordinates and put them into the Nav equipment, and of you go. There is no need to fly along air routes, with transponder off and in a no radar environment after FMT there is no risk to get detected. There is no reason to sit 6 hours in front of an electronic box and get a sore ass in order to expierience, what you think would be important.

    -using night sky stars and/or planets for navigation, checking out sunrise at crash time, etc.
    I have learned ASTRO-NAV in my time, Z sure didn`t. The B777 has no gadgets to perform ASTRO-NAV and it does not need one. Multiple navigation gadgets guide you to whatever place you wish. Sunrise is available in lookup tables for any position and any day in the year, again no reason to waste time in front of a simulator.

    -rehearsing a new flight route is something some pilots do at home
    That might be true for a pilot under training, low hour and low expierience. An expierienced pilot like Z would not do such a thing if you paid him extra money. How much job work do you carry home and create even the electronic gadgets with your own money to perform such work at home?

    – some elements of the MH370 diverted fight may have been rehearsed in commercial flights prior to 8-March
    I have no idea what you are saying with that statement.

  370. RetiredF4 says:

    @Brian Anderson
    “The two 406 ELTs on MH370 are designed to trigger upon a deceleration of about 2.5G or more. However, the data transmission protocol delays the first transmission for approximately 50 seconds after triggering, and then continues to transmit the data block of 520mS at 50 second intervals. If the ELT antenna is underwater before the 50 second delay elapses then no signal will be radiated. Repeating the data block every 50 secs is necessary to allow for the possibility of one of the monitoring satellites to be passing [roughly] overhead. A larger delay may be experienced before this occurs. In the case of a crash into the sea, even if the ELT does trigger, the likelyhood of a successful transmission being received is close to zero.”

    The core question in regards to ELT was not, wether it would or would not work in a crash, but wether it could be used to alert the outside in the flight phase prior FMT. Other traffic and ground stations would have heard transmissions of an activated ELT on 121,5 or 243.0, frequencies which (one of them) have to be monitored by all ATC ground, ship and airborne systems and aircraft. During this phase of the flight until FMT and some mile beyond LOS and range to such systems was no hindering factor.

  371. ulric says:

    I am extremely uncomfortable with 35S.

    In general, I believe that endpoints further South are consistent with scenarios caused by technical failure of systems and components whilst endpoints further North are more consistent with events driven by human agency (of whatever kind).

    My mind returns to the original Inmarsat “hotspot” which is a fair distance North of the positions now being spoken of. The Inmarsat calculations are particularly interesting IMO because they were done very early on in the proceedings and are, as far as I can tell, free of some of the biases and misinformation which have been introduced since then. They don’t for instance, contain any assumptions about motives or the mechanics of the final turn, relying entirely on the ping data.

    Perhaps these original calculations need to be revisited? It seems to me that they should carry weight purely on grounds of their potential neutrality.

  372. Victor Iannello says:

    @Ulric: The requirement that the plane turned to the south before 18:40 along with the assumed speed profile and the assumed probabilities assigned to the number of manoeuvers collectively constrain the path to relatively straight, fast trajectories with endpoints centered around 38S latitude. When those constraints are removed, along with the requirement of close match to the BFO data as well as the BTO data, the endpoints move to the north. I don’t think the ATSB will revisit these assumptions. Of course, many of us are doing that very thing.

  373. Victor Iannello says:

    @Ikr said, “There is a tiny blue-green-yellow blip [>10% detection probability to 60%+] that lies in a sea of red [95+%] EXACTLY at 30S. Did anyone else notice this?”

    Yes, this is not lost on many of us. Also, if you look at Fig 3.2.1 of the CSIRO report from Dec 2016, bottom left panel, you will see that there is a narrow band of latitudes centered on 30S where the non-flaperon debris arrives on Africa sooner than neighboring latitudes.

    One of the arguments against a 30S crash site is that debris would arrive too early in Africa. This could be attributed to the distinction between arrival and discovery. A crash site at 30S would also explain the discovery of the engine cowling part in Mossel Bay, South Africa, in Dec 2015.

  374. Victor Iannello says:

    @Ge Rijn: In a recent paper, Richard Godfrey did consider a great circle path crossing the 7th arc at 30S and leading to Wilkins Runway (YWKS).

    Interestingly, both Barry Carlson and the ATSB had also considered this possibility. From Richard’s paper:

    On 19th August 2014 07:37 Barry Carlson of the Independent Group posted on the ATSB Blog: “Strange as it may seem, the ATSB assessment for the Priority Bathymetric Survey is centered exactly where the GC track to YWKS crosses the 7th arc. If YWKS has never featured in your considerations, then I assume it would now be a further confidence booster in validating the work done to date.”

    Martin Dolan, Chief Commissioner responded on 20th August 2014 09:13: “Thank you, Barry, for your insight. You will be pleased to hear that the search strategy group did consider YWKS as a possible waypoint. The location of the search area however, is based on the analysis of the satellite communications data.”

  375. Ge Rijn says:

    @VictorI

    What hasn’t been considered yet..? 😉

  376. DennisW says:

    @RF4

    So you (and others) think the simulator data is meaningless. I and (others) believe it is quite relevant. We will just have to disagree.

  377. TBill says:

    @RetiredF4
    Come on, everyone here has a sore butt from MH370. Don’t depend on denial tactics. Flight sim cases can go 8x speed, and you can just pick-up the aircraft move it where ever you want on the globe. It does not take long to do experiments. I must speak to Victor…he keeps saying there is little reason for flight sim studies, which undermines his argument to some extent. It is an extremely useful tool, especially for those contemplating taking over an aircraft.

  378. Victor Iannello says:

    @TBill: I didn’t mean to imply that there is little reason for flight sim studies. I meant that if a pilot was planning a flight, the navigation and fuel calculations would best performed using other tools. That’s not to say that ZS didn’t choose FS9 as a tool to plan the diversion. We can only speculate as to why ZS created a simulation of a flight from KLIA, up the Malacca Strait, and into the SIO until fuel exhaustion.

    In the past, I brought up the possibility of his using FS9 as part of suicidal ideation. I think that possibility remains on the table.

  379. Ge Rijn says:

    @RetiredF4

    In a way you give conflicting arguments about Zaharie having a simulator:

    ‘In what kind of professions is it perfectly natural to set up a training gadget at home for the work you have to perform routinly in the job?’

    I think you’re right in general, but he DID set up this gadget at home.

    And: ‘Flying the simulator at home as a hobby could count, but not with a set up for the 777.

    He DID a set up with a B777-200LR specific to this SIM points.

    IMO it’s not in the overall arguments you mention but in the overall context which makes it extraordinary this complex of (a ~month before) deleted SIM-points was found on one removed hard disk he used on his flight-simulator.

    This was not one of maybe many pilots who also had a ‘gun and bullets'(a flight-simulator).
    This was the pilot who flew the ‘gun and bullets’ on the plane that went missing in the SIO.

    Finding deleted coördinates together on one of his hard-disks indicating a similar flight into the SIO with zero fuel end points (what pilot would practise something like that there and for what reason?) must be regarded as highly suspisious and as a possible ‘smoking gun’.

    I think your arguments (although conflicting in some ways IMO) are well counting in general but not in the context of this case.
    Zahari (and his co-pilot) flew the ‘gun and the bullets’.
    That makes a lot of difference.

  380. lkr says:

    Thanks, Victor — I raised the “little blue hole” at 30S specifically because that latitude keeps popping up in independent drift studies [in addition to published studies we can add the series that Olexander posted here a couple of weeks ago, I believe].

    The other argument against northern sites that ASTB/CSIRO keeps raising [as in David Griffin’s letter above] is that debris would have been expected much earlier. Eg, “a 30S crash site would, according to our model, have resulted in debris washing up on Madagascan and Tanzanian shores a full year earlier than was observed. That is a discrepancy that is hard to set aside.” Again, I ask: why does Griffin and colleagues think this is a discrepancy if no one is looking? Strangely the Madagascan [and Moz] debris is found “a year too late” but only when Blaine Gibson is pointed to these coastlines. And none of it fresh — most or all could have lain on the beaches for that year..

    I’m repeating myself and I know this is all old business here, but if time of arrival is seen as an important question, remember that we have only the Reunion flaperon and SA ‘Roy’ to go on as ‘catch of the day’. And I do hope that these questions are posed each time such arguments are made?

  381. Ge Rijn says:

    @Ikr

    I completely agree.
    As I mentioned also before, the RR-piece and the flaperon are time markers that define a more or less specific window of opportunity for all- or most other found debris to have beached.

    Especially the RR-piece I regard as a quite specific time marker.
    For its abondance of large barnacles and its most far distance from the 7th arc most- if not all other found debris must have beached before 22-december 2015.
    The majority of ‘clean’ pieces found arrived earlier and possibly some even almost a year earlier.
    There’s no way to tell exactly how much earlier pieces beached.
    So IMO the CSIRO/Griffin argument on this is rather weak.

  382. ulric says:

    @Victor

    You have consistently considered and investigated scenarios which extend our understanding of what is possible or not possible. Keep asking questions!

  383. RetiredF4 says:

    @Ge Rijin

    Not only my arguments are conflicting, it’s the same the oher way around. That Zahire had a flight sim software and built his own custom made computer setup is a fact, there is no denial on my side. That he used it and he also offered it to friends for flying is a fact too. That the simulator crashed is one or several,times is another fact. That it was forensically examined and some deleted nav points besides other deleted files have been found is a fact too.

    We differ in the interpretation of those facts, and imho it is not as simple as computing 2+2 and coming up with 4, disregarding some unknown variables. I made my case why I think that Z built the sim, to improve the setup of, mainly the input and output devices to a higher standard by cheap means. He liked to do technical things, to repair them and to improve them. I know a lot of classic car enthusiasts who enjoy to restore such a classic car, but when it is finished they nearly never drive it. They get the kick from the resulting excellent state of the car, not from the ability to drive it. When they are finished, they start the next project. I see such a guy each morning in the mirror when I brush my teeth. I know pilots of all sorts, but none of them would plan a mission on his home simulator, most do not even have one. None of them would need to plan a straight flight to a given point like a coordinate in the SIO at all. Flying straight and level until fuel exhaustion is one of the simplest tasks at hand, no rules to follow, no regs to watch, no passengers to worry about, no paper work after the flight. Nothing there what would require prior training. That Z trained this part after FMT is a no brainer. The critical portion of the flight was happening prior FMT, and even an expierienced guy might have trained this timing critical and action loaded part of the flight. You show me this part hidden in the SIm data and I’m with you.

    Why the points in the SIO? Maybe he used them for some simulator calibration profile, checking the resulting flightpath against the expected outcome, concerning variation, fuel used, wind effects, to name a few. He was the guy only trusting his own abilities, I can imagine such a guy checking the software of the sim against his expierience. But those points could have been the result of another person using the simulator, or they could be faked or innocent at all.

    Using the RMP report and especially rhe Sim data as evidence for the Shah did it version and disregarding the findings of the very same report, wich white-washed Z in total, is the real conflicting argument here. The Shah did it fraction believe the parts with the Sim points coming from the one and only source, namely the RMP report uncritical, but do not accept the end finding of this report. The MP is used as the trustworthy witness and observant prosecuter, and at the same time as liar and manipulator of facts. You can’t have it both ways, wether you like it or not.

    Now be carefull to put me in either camp, I really have no idea what really happened, wether it was Z or somebody else. This case stinks wherever you dig. That’s why I stepped back from commenting. But the sim case is grossly overrated as evidence and blocks the broader view on other possibilities.

  384. RetiredF4 says:

    @Dennis W
    “@RF4

    So you (and others) think the simulator data is meaningless. I and (others) believe it is quite relevant. We will just have to disagree.”

    It is not rhe first time, and I think we both can live with it.

  385. Victor Iannello says:

    RetiredF4 said, “The Shah did it fraction believe the parts with the Sim points coming from the one and only source, namely the RMP report uncritical, but do not accept the end finding of this report.”

    First, there were parts from only 7 FLT files found in the Shadow Volume of MK25, so any argument about cherry-picking the files among thousands is wrong. You must also remember that it was the FBI that obtained the deleted simulator data from the MK25 drive. I don’t think the RMP would misrepresent data that was provided to it by the FBI. That would be very dangerous, especially since the Najib administration is under investigation from the FBI for corruption related to the 1MDB fund.

    On the other hand, the RMP had more flexibility in how it might choose to state its conclusions. In Folder 1 of the RMP report, it says: From the forensic examination as of the report date, it is found that there was no activity captured on exhibit JP01 and MK 26 that conclusively indicate any kind of premeditated act pertaining to the incident MH 370.

    There are several interesting things about this statement. First, it only says that the captain’s guilt could not be “conclusively” proven. Second, the suspicious activity was found on drive MK25, not MK26. Third, I find no other reference to exhibit JP01, so it is not clear what that is.

    The deleted data points were from a flight simulation that progressed from KLIA, up the Malacca Strait, and ended with fuel exhaustion in the SIO. They were the only deleted points found in that Shadow Volume. They were deleted on or after Feb 3, 2014. It would be an extraordinary coincidence if ZS did not participate in the diversion of MH370.

  386. lkr says:

    Concerning Z and the simulator, RF4’s points [gun, bullets, dead neighbor] are well-taken. However, I think the case Victor is making is that we found a few pieces of a picture of the dead neighbor, just possibly having been used as a rifle target, in the suspect’s paper shredder..

  387. Gysbreght says:

    “The deleted data points were from a flight simulation that progressed from KLIA, up the Malacca Strait, and ended with fuel exhaustion in the SIO. They were the only deleted points found in that Shadow Volume. They were deleted on or after Feb 3, 2014. ”

    Three opinions that I believe to be false.

  388. lkr says:

    The following may be a great deal of mucking about the obvious, but I don’t think I’ve seen the topic come up in any depth.

    I keep coming back to this for some time: what was the nature of the debris that reached the western shores of the Indian Ocean by mid-2016, and what might that tell us about the initial debris field? Our only real sample again comes from Blaine Gibson’s visit to nw Madagascar in June 2016. According to the Malaysian official site updated to 28Feb17, Blaine collected a total of 5 “likely” to “almost certain” pieces, and 3 possible but “unidentifiable” pieces from two facing beaches [Riake, Antsiraka]. Two more pieces, also “unidentifiable” were turned in from Antsiraka beach later in 2016. I can’t find it now, but I remember Gibson saying he had surveyed roughly 20 km of beach — much of it at speed on an ATV, if I remember correctly. He got a lot of help — most of the pieces were found by local residents when Gibson showed interest.

    So, let’s say that Gibson searched a little over 1 percent of the 1600km Madagascan east coast [surely an overstatement, but he was probably cherry-picking a hotspot for flotsam]. That projects to about 400 identifiable items and another 400 “possibles” being potentially recoverable on Madagascar alone.

    Further, since many models show that 30-50% of recoverable debris items from the 7th arc would come ashore on Madagascar, we have on the order of 1000 to 3000 items afloat for more than a year. That’s surely an understatement of the total in the original debris field. On the other hand, we almost certainly have an overrepresentation of the larger pieces — anything more than 50cm in greatest dimension. These are certainly more likely to attract attention, the more if they have distinctive markings or obviously aerodynamic form. These larger pieces are also much more likely to retain buoyancy for many months, and to remain intact on the beach — both evident in the few, very distinctive pieces turning up in South Africa.

    Reversing my field, if it’s likely that the original debris field [ignoring personal effects, which seemingly are impossible to distinguish from the regular harvest of the Indian Ocean garbage patch] comprised several thousand fragments capable of surviving a year at sea, and perhaps several dozen as eye-catching as a flap or ‘Roy’ or ‘No Step’, the effective population that could reach Western Australia in 2 or 3 months must be quite large. Even sporadic beach-cleaning should have found them if they had arrived.

  389. Brian Anderson says:

    @RetiredF4,

    I agree that there is a reasonable likelyhood that if the portable ELT had been triggered manually [and perhaps held by a window] then a transmission would have been detected.

    So the question is why did that not happen. There is only one portable ELT. It’s location and operation ought to be known by the cabin crew. Did anyone have the presence of mind to access it and trigger it? Or, were the crew incapacitated by the time that a diversion from the expected flight path became apparent.

  390. Oleksandr says:

    @Victor,

    ” It would be an extraordinary coincidence if ZS did not participate in the diversion of MH370.”

    May I suggest you stopping presenting your speculations as facts, and refraining from making false accusations? This style does not fit you.

    So far, you have failed to produce any evidence that the ‘dots’ you connected belong to a single simulation session. I believe RMP with the support from FBI is in better position to conclude that there is no such an evidence.

    Let it rest.

  391. DennisW says:

    @ikr

    While I do not have an opinion on the amount of debris that could be found, I do have an opinion on the amount that will be found. Both the Poisson and Weibull statistics using the frequency of debris found to date suggest that some 50 pieces will be found in the eight year period following the flaperon find. The last several months have actually fallen below that humble pace. My feeling is that it is the result of lack of search activity rather than the amount of debris actually out there. I also believe that additional debris finds will likely not have much if any impact on underwater search decisions.

  392. Oleksandr says:

    @ALSM,
    @Dennis,
    @Bobby,

    Are you confident that BFOs are always affected during warm up?

    If yes, wouldn’t the correct BFO 18:25:27 in the logon sequence be nearly a proof that the SDU was powered prior to this time?

  393. Oleksandr says:

    Dennis,

    Re: “Both the Poisson and Weibull statistics using the frequency of debris found to date suggest that some 50 pieces will be found in the eight year period following the flaperon find. The last several months have actually fallen below that humble pace. My feeling is that it is the result of lack of search activity rather than the amount of debris actually out there.”

    In my estimations there are should be around 50 fragments nearly uniformly distributed along the whole Madagascar eastern shore, 4 fragments in SA (vs. 4 found) and 5 fragments in Mozambique (vs. 6 found) by Dec 31, 2015. A good agreement with your forecast, and with the debris found up to date.

  394. lkr says:

    Dennis: The point of this exercise is to illustrate how observer bias has led to a very unnatural depiction of the drift. And the reasoning from “absence” by ATSB/CSIRO. So the problem of accounting for “Roy” while assuming that no pieces were reaching Tanzania until much later. If thousands of objects remained afloat for over a year, the distributional tail is much wider. And thinking about scale — initial number, size distribution, and of course location on the plane — potentially gives some insight into the type of impact.

    In terms of “how much” — I think you’re considering recovery assuming no depletion of the unrecovered base — not likely if most or all has reached land or sunk.

    And you’re right that this won’t impact any decisions to resume underwater search. That ship may have sailed since only the Australian scientists have much of a stomach for it, and my impression is that the Aussie politicians and bureaucrats are happy to hang the $200 million hole in the ocean on Tony Abbott’s technological optimism.

  395. lkr says:

    Dennis — I should have said that the 50 or so fragments you suggest is realistic for what might eventually be found, if anyone continues to look. But given BG’s little sample that’s probably low by a factor of 10 or 100 for all fragments that made landfall and were sitting up for beachcombers by mid to late 2016.

  396. David says:

    @Brian Anderson. This was a question raised by the Malaysian Air Attendants’ Union, which should have a good idea of what the reaction on board would have been. Speculation has included some of:
    – they might not have been aware what was going on,
    – they were incapacitated as you say,
    – they did try and the signal was not received or was meaningless (timing, brevity or tracking),
    – a hijacker or an armed pilot, aircraft on autopilot, prevented it;
    – it did not occur to them, not being part of their training.

    As to the last,the raising of the question counters that.

  397. Victor Iannello says:

    @Oleksandr: First, I will comment here as I please. If my comments don’t suit you, please find another site.

    You say I have produced no evidence that the simulator points are connected. This is false. Before you make any more comments, you need to read this report carefully from top to bottom so you can comment in an informed manner. And repeating comments made by Malaysian officials is meaningless.

    You say the RMP is better able to make the call about whether the simulator points are connected. In fact, what little analysis they performed is riddled with errors. Basic calculations like fuel level and speed are not correct. I can make this statement because I have studied the simulator data in great detail. I have become proficient in MSFS, installed the same software as ZS, and replicated his results. Yves Guillaume has done the same. You evidently have not.

  398. DennisW says:

    @Oleksandr

    “Are you confident that BFOs are always affected during warm up?

    If yes, wouldn’t the correct BFO 18:25:27 in the logon sequence be nearly a proof that the SDU was powered prior to this time?”

    I think Bobby and ALSM both feel that the 18:25:27 BFO was the result of lucky timing relative to up a power on which occurred very shortly before that time i.e. at the event time the oscillator just happened to be “on frequency” during the turn on transient condition. Another opinion (mine) is that the SDU was powered on long before 18:25:27 or was never powered off at all. I think Holland is also in the Bobby/ALSM camp.

  399. Mick Gilbert says:

    A question for the drift modellers if I may;

    In order to get to achieve the number of recovered items we are seeing do you have any estimates as to how many items were liberated at the point of origin?

  400. David says:

    The ATSB Chief Commissioner loses patience with Higgins. 3 days ago:

    https://www.atsb.gov.au/newsroom/correcting-records/mh370-reporting-by-the-australian/

  401. Brian Anderson says:

    How much debris littered the SIO but was never identified in a surface search? I suggest that it was a great deal more than the few sample pieces that happened to be found during beach searches.

    If I go back to the photographs that were taken by the RNZAF, in particular those from 29 April 2014, and geo-referenced to between 28.4 and 28.8 degS and 97.75 to 96.08 degE, then there are some significant possibilities.

    There are some samples in this Dropbox link . . . https://www.dropbox.com/sh/mmx3959yevxdubx/AAAJ7gWLDkI4ptYmJ9MkcpoHa?dl=0

    In one image the density of flotsam apparent is huge, but unidentified. In another the image shows a substantial piece of debris that is not at all like fishing debris seen in other images.

    It would be interesting to conjecture where this debris might have been on 8 April, bearing in mind the drift patterns in teh area.

  402. Mick Gilbert says:

    @Victor

    Victor, for fear of coming across like the Spanish Inquisition (as some of us know, nobody expects the Spanish Inquisition!) a few questions about the topic du jour (probably more correctly, the topic de la semaine), the Captain’s flight simulator “flight of interest”;

    How do you think these data sets (“fragments of flight files”) were created? The user deliberately saving them? System generated? Restore points after a program/system crash?

    Subsequent to the point at which a data set was created, if the simulation was stopped or aborted and restarted some time later, would the next saved data set appear as part of one continuous simulation?

    In your and Yves’ paper I think that you say that the date/time when these data sets were created is unknown, is that correct?

    Do you know if the hard drive that they were recovered from was the only hard drive physically disconnected from the Captain’s computer?

    Do you know if these data sets were deleted or just overwritten?

    Do you know if these data sets were the only deleted data sets found on the hard drive?

    Do you know if these data sets were all deleted at this time?

  403. DennisW says:

    @David

    I doubt Hood’s comments have anything to do with a loss of patience. He is simply feeling the heat, and wants to deflect it.

    A bold title – “Correcting the Record”. Like this is the last word – we are right, and Higgins is wrong. Something like “Comments on The Australian Reporting” might be more appropriate. The most appropriate course of action would be to say nothing at all. Let other publications/journalists come to your defense. I guess he “ran out of patience” with that happening.

    Hood assures us that a Malay report is coming along with good progress. What he does not say is that the Malays have never published anything resembling a satisfactory report, and this one will be no different. I am quite sure of that.

  404. Andrew says:

    @DennisW

    RE: “A bold title – “Correcting the Record”. Like this is the last word – we are right, and Higgins is wrong. Something like “Comments on The Australian Reporting” might be more appropriate.”

    “Correcting the Record” is the name of the section of the ATSB’s website where it posts rebuttals of the nonsense that is often printed in the media with respect to the ATSB’s activities. The title of the piece that specifically relates to Higgins’ reporting is “MH370 reporting by The Australian”. Is there a problem with that?

  405. David says:

    @DennisW. “Correcting the Record” is the title of the ATSB section where this is logged, not the letter I am more inclined to support Greg Hood than you, having become more than frustrated about Higgins drawing unwarranted conclusions and bullying his way round. He does report useful information but his extravagant dramatisations need to be checked or at least brought to public notice.

    I have found no sign the letter was published by The Australian. If not,I hope this gets taken up with the Press Council here.

  406. David says:

    @Andrew. I missed the snap this time. Slow typer.

  407. DennisW says:

    @David/Andrew

    Yes. I obviously have a problem with the title. Any reasonable person would have a problem with the title, and what it implies. Like this is the answer. No need to question further. Debate over.

    The fact that there is not a single publication in Australia (that I know of) supporting the work of the ATSB says a lot. I am in the same camp. While I think the ATSB is sincere, I don’t think their leadership or decision making is very good. Sorry if I tend to judge people by results. I know that is not politically correct in this day and age.

  408. ALSM says:

    Oleksandr:

    “Are you confident that BFOs are always affected during warm up?”
    Yes, there will always be a warm-up transient as Bobby and I have explained.

    If yes, wouldn’t the correct BFO 18:25:27 in the logon sequence be nearly a proof that the SDU was powered prior to this time?
    Absolutely not. The OCXO was clearly in the process of stabilizing shortly after the return of power, after having been off for some time.

  409. Capt Kremin says:

    As someone who considers that the preponderance of evidence and Occams Razor points to the rogue pilot theory, please consider the following.

    The search area, the fuel consumption data and the default autopilot modes point to the final track of MH370 being a straight line. If this is so, then that track was deliberately flown.

    The aircraft may have been flying in LNAV at some stage after IGARI, however no portion of the aircraft’s track after IGARI was required to be flown in LNAV. Heading Select was adequate to do achieve the described track. The aircraft was at normal cruise speed whilst being tracked, around 480-500 knots groundspeed.

    If the aircraft was flying an LNAV track until the turn south, and then a route discontinuity occurred, the default autopilot modes would revert to HDG HOLD (referenced on magnetic), SPD Mode (current speed hold) and ALT (current altitude hold).

    Heading south in these modes, with heading referenced to magnetic (the default mode) would have, due to changes in magnetic variation, generated a left hand curving track towards Australia. At the defaulted to cruise speed this track would not have fit the handshake rings.

    There are only two ways to generate a straight line south.

    1. Provide a LNAV waypoint for the auto-pilot to track to.

    or,

    2. Lift the guarded HDG Ref switch, switch the heading reference to TRUE, select the HDG/TRK switch to TRK and dial in the required track.

    Neither can be done accidentally or by default. Both require systems knowledge.

    I cant imagine any scenario of an aircraft with some sort of catastrophic malfunction that would dictate this course of action. As a pilot it goes against all common sense.

    The track of the aircraft to and around Penang and north-west along the Malacca Strait confirms the aircraft had, at the very least, visual navigation at this point. The idea then to take the aircraft away from civilisation with seven hours of fuel remaining would be anathema to any pilot.

    As far as the SDU behaviour, it was possibly an attempt by the perpetrator to ensure all communications from the cabin to the ground was impossible until after the aircraft was depressurised. The depressurisation would ideally be a slow event, in order to avoid suffering the Bends.

    This unfortunately, is a much simpler scenario to enact than some of the theories out there.

  410. DennisW says:

    @Oleksandr

    as ALSM says:

    “Absolutely not. The OCXO was clearly in the process of stabilizing shortly after the return of power, after having been off for some time.”

    Of course he is part of a group with a cluster of pins in the map at 38S. He might be right about the OCXO. I claim that it is not something we can know for sure based on the data we have.

  411. David says:

    @DennisW. Our stereo opinion did explain that what you take to be the title, isn’t.

  412. DennisW says:

    @David

    The title is the title. It is pompous and poorly suited to a professional agency. Simple as that.

  413. Andrew says:

    @DennisW

    Ok genius, what would you call it?

  414. ALSM says:

    Summary of significant Debris Drift based predictions:

    Author Pub Date POI Range (Latitude)
    CSIRO Aug 2015 S32-S39
    Henrik Rydberg (IG) Aug 2015 S34
    IPRC Aug 2015 S34-S37
    GEOMAR, et. al. Sept 2015 North of S30
    V. J. García-Garrido, et. al. Nov 2015 None
    IPRC Update Mar 2016 S34-S37
    GEOMAR, et. al. May 2016 “…west of Australia around the 7th arc.”
    Eric Jansen, et. al. July 2016 S28-S35
    UWA July 2016 S28-S33
    CSIRO Dec 2016 S32-S36, S35 most likely
    Richard Godfrey (IG) Feb 2017 S29-S31, S30 most likely
    CSIRO Apr 2017 S30.5-S40 possible, S32-S36 likely, S35 most likely

    It should be noted that all of these predictions begin with a basic debris drift analysis. Based on that alone, they all come to virtually the same broad conclusion (roughly S28-S40 to be inclusive). The final estimates above all depend on assimilating additional data and analysis, such as the early air search coverage, bottom search coverage, lack of debris found on the Western Australian beaches, etc. Thus, the debris drift analysis alone provides no definitive POI, or even a narrow range.

    That said, all of these Debris Drift studies taken together provide compelling evidence that the POI was NE of the area already searched, possibly as far as S28. That means the “pre-debris nav assumptions” that lead to the S37-S38 estimate were incorrect (unless it really is at ~S38 and they missed it). If not at ~S38, but somewhere NE of S35, then we need a different set of path assumptions (heading, alt, TAS, Nav mode) to explain points as far north as S28, consistent with the BTO/BFO observations. Victor and Richard have found some paths and nav assumptions that fit S30, consistent with an Antarctic WP destination assumption. So it is possible to reconcile the BTO/BFO data with points further NE, but only by assuming different nav assumptions and path assumptions between 1825 and 1941.

    All of this leads me to believe it will be impossible to “pin point” the POI combining the best of the path and debris drift models. We are still at the mercy of the remaining assumptions. Seems like nailing the path between 1825 and 1941 is still a worthwhile goal.

  415. ventus45 says:

    Andrew says: May 1, 2017 at 10:14 pm

    @DennisW

    Ok genius, what would you call it ?

    Perhaps if I may chime in ?

    “Bollocks to Bollocks”

  416. DennisW says:

    @Ventus

    I did not comprehend the question, but I will accept your answer.

  417. Andrew says:

    @DennisW

    “The title is the title. It is pompous and poorly suited to a professional agency. Simple as that.”

    It’s a fairly common term used by government agencies and other organisations the world over. As my dearly-departed mother would say, “get over yourself”.

  418. DennisW says:

    @Andrew

    Sorry to hear about your mom. The rest of your comment is BS.

  419. David says:

    @Andrew. I think DennisW has it in mind that the title is, “Correcting the Record” and he has given reasons for disliking that. Dennis, I do not see hoe those reasons apply to the actual title, “MH370 reporting by The Australian”.

    There is more to why this has not been published than meets the eye: I do not think now that it is buried at all.

  420. Mick Gilbert says:

    @DennisW

    “Correcting the record” is a common feature of a number of Australian government websites, it is neither exclusive to nor a product of the ATSB.

    Of the 11 “corrections” issued by the ATSB since June 2015 (the Department of Immigration and Border Protection has issued 28 in the same period), nine relate to MH370. Of the nine, two are general and do not name a media outlet, one relates to the Australian Broadcasting Commission and news.com.au, four relate to Byron Bailey’s contributions in The Australian and three relate to Ean Higgins’ articles in The Australian (one of which also contains material from Captain Bailey).

    Given some of the unadulterated pap dished up by both Ean Higgins and Byron Bailey on MH370 over the years (I’ve previously provided an expurgated summary of some of their nonsense), in my considered opinion the ATSB has been a model of restraint and courtesy.

  421. Andrew says:

    @DennisW

    “The rest of your comment is BS.”

    If you’re referring to the use of the term “Correcting the record” by other agencies, a simple internet search would disabuse you (or anyone else) of that ridiculous notion.

  422. Brock McEwen says:

    @ALSM re: “Seems like nailing…18:25-19:41 is still a worthwhile goal”: it has been a critical goal. Since day one. Hence the criticality of military radar assets of all nationalities and stripes. The path during that time could not have intersected known radar coverage. I could really appreciate some help tracking down the positions of all shipborne, airborne and fixed radar assets that night. Can we please start by firmly establishing the USS Pinckney’s exact position at 17:20 UTC on March 7, 2014? DDG’s radars are, I understand, quite powerful. Thanks in advance for your help.

    Aside: the incompatibility of the debris record with an impact in the zone on which the SSWG blew its entire budget has been known beyond reasonable doubt ever since debris failed to show up on Oz shores by the end of 2014. Some of us have been making this point consistently since 2014. Others started making this point only after search leadership finally admitted this – and spent the intervening years publicly slamming the science, and even the scientists. I find the lack of humility in the latter group utterly fascinating.

  423. RetiredF4 says:

    @Brock McEwen

    Information about the engagement in the search the “USS Pinckney” was reported doing normal training in the south China sea and could relocate to the waters of Vietnam, where the first search started, within 24 hours. That would place it up to 350 – 700 miles away from the action, based on half and full speed.

    USS Kid was reported to be active in the same area and in the same routine training.

  424. Oleksandr says:

    @Victor,

    Re: “First, I will comment here as I please. If my comments don’t suit you, please find another site.”

    I only gave you my suggestion. Whether to take it or not – up to you.

    Re: “You say the RMP is better able to make the call about whether the simulator points are connected.”

    RMP and FBI had physical access to the disk. Did you?

    Re: “I have become proficient in MSFS, installed the same software as ZS, and replicated his results. Yves Guillaume has done the same. You evidently have not.”

    Why is it so difficult for you to grasp that the issue is not only in MSFS?

  425. Oleksandr says:

    ALSM,

    Re:
    ““Are you confident that BFOs are always affected during warm up?”
    Yes, there will always be a warm-up transient as Bobby and I have explained.

    If yes, wouldn’t the correct BFO 18:25:27 in the logon sequence be nearly a proof that the SDU was powered prior to this time?
    Absolutely not. The OCXO was clearly in the process of stabilizing shortly after the return of power, after having been off for some time. ”

    Thanks for confirming. So, you think it is just a pure coincidence that the very first BFO accurately represents speed and heading?

    Why do you think “the OCXO was clearly in the process of stabilizing shortly after the return of power” 18:25 as opposite to what Don Thompson suggested long time ago (that the loss could also be due to improper steering)?

  426. Oleksandr says:

    @Dennis,

    “I think Bobby and ALSM both feel that the 18:25:27 BFO was the result of lucky timing relative to up a power on which occurred very shortly before that time i.e. at the event time the oscillator just happened to be “on frequency” during the turn on transient condition. Another opinion (mine) is that the SDU was powered on long before 18:25:27 or was never powered off at all. I think Holland is also in the Bobby/ALSM camp.”

    Long time ago Don Thompson has proposed an idea that the loss of contact was due to improperly steered antennas as opposite to depowering. The very accurate match of the first BFO to the last known radar speed and heading supports this idea. The confidence that BFOs are always affected during the process of stabilizing shortly after the return of power works against the mainstream assumption that the SDU was previously de-powered. Or not? Was Don’s idea discarded?

  427. Oleksandr says:

    @Mick Gilbert,

    “In order to get to achieve the number of recovered items we are seeing do you have any estimates as to how many items were liberated at the point of origin?”

    I only assumed proportional dependence of the beaching probabilities predicted by my drift model to the number of fragments actually found where they were searched for.

  428. David says:

    @GeRijn. You noticed that at a recent NOK briefing the Malaysians had identified debris items 26 and 27. The investigation results are here:

    http://www.mh370.gov.my/phocadownload/3rd_IS/Debris%20Examination%20300417.pdf

    Note that with item 26 there has been a failure assessment and, “The fitting on the debris appeared to have suffered a tension overload fracture”.

    At first glance that to me suggests shock, assuming the “appeared” is based on a lab examination. The torque reaction on overload would be taken by the locked actuators, ie ‘blocked’ (under RAT or as would be otherwise, assuming speed is above cruise). At this hinge the normal loading would be mostly shear I believe, ie resisting lift, +ve or -ve.

    Also I note their assessment that its finding is consistent with the CSIRO drift model.

    The item 27 findings are again consistent with the CSIRO drift model.

    Again at first glance there is the appearance of it being asymmetrically forced off from the right at the front, twisting off at the rear. This was the inner side. Had it been the outer it might have suggested the engine or a cowling had hit the fairing. I note the fairing was close to the recovered right outer flap piece.

    Their new debris items listing, which now has none with identification outstanding, is at:
    http://www.mh370.gov.my/phocadownload/3rd_IS/Summary%20of%20Debris%20300417.pdf

  429. David says:

    PS I notice the number 7 fairing front part was found in South Africa a year after the rear part of the same fairing was found in Mozambique. Not much travel for that duration yet consistent with the drift model. Local gyres and currents perhaps. (Delayed findings naturally cannot be allowed for in the model).

    This illustrates how accommodating consistent with the model is, in other words how approximate, or am I misreading?

  430. Mick Gilbert says:

    @Oleksandr

    Thank you for that information.

  431. David says:

    PPS Another outcome. Item 26, “The non-metallic honeycomb was soaked with water resulting in the core material to swell”.

    Still soaked. The remaining skin does not have the appearance of being separated from the honeycomb. That would suggest that water can diffuse through the honeycomb, having implications for how low in the water such items got (reduced windage) and how for how long they would float (attrition rate).

  432. Ge Rijn says:

    @Mick Gilbert

    On an estimate about the amount of initial floating debris after impact I think also the following drift model from @MPat in this JW blog-topic gives an indication:

    http://jeffwise.net/2016/07/07/guest-post-where-mh370-search-area-debris-has-historically-gone/

    Out of 177 historical drifters passing through the former search area 39 beached within ~two years. 32 on African shores 7 on Australian shores.
    27 pieces are found so far only on African shores.
    This fairly accurate aligns with the 32 drifters from @MPat’s model.

    So this model predicts an initial start of 177 drifter-like pieces in the former search area is needed to result in the beaching of 32 drifter-like pieces within ~two years on African shores.
    I assume more pieces beached and are waiting to be found but I don’t think more than a 1000 drifter-like pieces were initially present after the impact.

    If so, more than ~5 times the amount of debris (about ~130) should have beached on African shores after ~two years.
    This means more than 100 drifter-like pieces are still lying around somewhere on African shores after more than 3 years without being found.
    I think this could be possible but I think it gets more and more unlikely the bigger you take the amount of initial debris.
    With an initial 10.000 drifter-like pieces there would be ~1000 pieces of yet unfound debris in Africa. 5000 pieces makes ~500 unfound pieces and so on. The more drifter-like pieces after an initial 1000 the more unlikely it becomes IMO.

    So I think an initial amount of max. ~1000 pieces is a fairly correct estimated limit based on this model.

  433. Ge Rijn says:

    @David

    Thanks for the links. This surely needs some thought.
    Come back on it later.

  434. Ge Rijn says:

    @Mick Gilbert

    To be more complete I add the Jeff Wise-topic of 15-july 2016 that followed the previous linked JW-topic.
    It further discribes thoughts on the @MPat model and also includes some of my thoughts back then:

    http://jeffwise.net/2016/07/15/how-we-know-where-mh370-went/

  435. Gysbreght says:

    @Victor Iannello: What were the values for MaximumGForce, MinimumGForce, MaxReachedEngineRPM1, MaxReachedEngineRPM2 that you recorded in the “Recreation of Simulated Flight” in your and Yves Guillaume’s paper dated November 29, 2016?

  436. Gysbreght says:

    @Victor Iannello:
    Source: Copy-on-write method

    In the copy-on-write method, when a change to the original volume occurs (but before the write I/O request is completed), each block to be modified is read and then written to the volume’s shadow copy storage area (also called its “diff area”). The shadow copy storage area can be on the same volume or a different volume. This preserves a copy of the data block on the original volume before the change overwrites it.”

    Can you please explain how your statement “The deleted data points (…) were deleted on or after Feb 3, 2014.” is compatible with Microsoft’s description of How Volume Shadow Copy Service Works?

  437. David says:

    @Ge Rijn. “27 pieces are found so far only on African shores”.
    I wonder how many MH370 pieces washed ashore have not been recognised as possibilities and overlooked or destroyed. Drifters are recognisable, small pieces of honeycomb maybe not be. I would have expected a much higher proportion of unrecognisables than we have seen so far if all actual candidates had been recovered.

    I gather that seat cushions might have been burnt at La Reunion early on.

    How many items from other crashed-at-sea aircraft were recovered because they were near a known crash site, all such items being recovered automatically. Even MH17: how many of its pieces would have floated and of those how many would have been handed in from such a distant crash site?

  438. Don Thompson says:

    @Oleksandr, “Was Don’s idea discarded?”

    Yes, I can confirm that the ‘idea’ was discarded. The outline hypothesis, that I suggested in a note of 1st Oct 2014, was that some non-normal condition involving the HGA may have caused establish of the C-channel (satvoice) calls to fail during the periods evident from the MH370 Data Communications Logs.

    Knowledge gained through later studies disproved that hypothesis:
    That only 60 seconds were allowed for call answer, before the GES/ground network cleared, is consistent with the 3G voice network deployed by Inmarsat (sourced). The ground segment cleared the call, as evident from the C-channel sub-channel exchanges.

    The AES, onboard the aircraft, was using the HGA signal path. The alternative, using the LGA, would not have permitted packet data rates or C-channal use as recorded in the MH370 Data Communications Log.
    The GES received power level measured during the ongoing flight, as described in ‘The Search for MH2370’ paper (Ashton et al, 3rd Oct 2014), was normal. Therefore, there is nothing to indicate that, at any time, the AES was failing to accurately track I3-F1 using the aircraft position and attitude data derived from a fully functioning ADIRS.

    Finally, after much discusssion of the typical warm-up characterisitics of OCXOs, Holland’s paper confirmed that the OCXO used by an MCS-6000 SDU does exhibit an overshoot and subsequent stabilisation of its frequency as the oven settles to working temperature.

  439. Don Thompson says:

    @Gysbreght

    Inappropriate reference. The Technet link that you cite is applicable to environments where NTFS volumes support a database, or similar application, that must remain online during ‘backup’ operations (more properly, ‘data protection’ operations). In such an environment the application is integrated with VSS through requesters and writers to ensure volume IO is quiesced as metadata & any volume blocks are captured with application consistency and ongoing volume changes are saved (usually, to separate devices).

    The use-case for a simple Windows 7 desktop environment, such as the captain’s PC, is much less complex.

  440. Victor Iannello says:

    @Gysbreght: The contents of the Shadow Volume reflect the contents of blocks of data that were overwritten on or after Feb 3, 2014. Before the blocks were overwritten, the original contents were written to the Shadow Volume so that the state of the files on Feb 3, 2014, could be restored. The contents of the Shadow Volume therefore represent a “snapshot” of the content of blocks of data on Feb 3, 2014, that was subsequently modified.

  441. Warren Platts says:

    I was just looking at my google earth, and I see I have a marker at 30.75 S labeled Coriolis end point–from an early analysis of mine that modeled the aircraft traveling “inertially”; that is, where the rotation of the Earth is the only factor affecting its trajectory. I’ve been told repeatedly that such a flight path is impossible for a B777 (but might be possible for a B787). Still, the new drift models make me wonder…

  442. Victor Iannello says:

    @Don Thompson: I’ve already explained to @Gysbreght that it would be extraordinary for Microsoft Flight Simulator 2004 to implement VSS.

  443. Victor Iannello says:

    @Gysbreght: The minimum and maximum values of acceleration are dependent on pilot inputs during manual flight, and could be replicated.

  444. Victor Iannello says:

    @Oleksandr: You propose that we should blindly accept the conclusions of the RMP report because the RMP had access to the physical disk and we do not. You can choose to do this. I will continue to challenge the conclusions of the RMP based on an analysis of the the evidence we have.

  445. Gysbreght says:

    @Don Thompson: Where do I find a description of “The use-case for a simple Windows 7 desktop environment, such as the captain’s PC” ?

    The source that I quoted also states:
    “The Windows operating system includes a VSS provider that uses copy-on-write. ”
    and:
    “The Windows operating system includes a VSS provider that uses copy-on-write. “

  446. Gysbreght says:

    @Victor Iannello: You have not answered my question. Can you please provide that minimum and maximum values of acceleration that occurred in your simulation, and the MaxReachedEngineRPM’s that I requested?

  447. Victor Iannello says:

    @Mick Gilbert:

    >How do you think these data sets (“fragments of flight files”) were created? The user deliberately saving them? System generated? Restore points after a program/system crash?

    Read the paper. It describes how the flight files were manually created by the user almost down to the keystroke. An anomaly that we discovered in MSFS gives us great clues.

    >Subsequent to the point at which a data set was created, if the simulation was stopped or aborted and restarted some time later, would the next saved data set appear as part of one continuous simulation?

    If the simulation was paused, a flight file was properly created, and the simulation was ended, the flight could be resumed at a later time by loading the flight file. A subsequently created flight file would, to the best of my knowledge, be unaffected by the previous stop/restart.

    >In your and Yves’ paper I think that you say that the date/time when these data sets were created is unknown, is that correct?

    Correct.

    >Do you know if the hard drive that they were recovered from was the only hard drive physically disconnected from the Captain’s computer?

    The RMP says of the 5 drives, only MK26 was found connected. It may be that at any one time, ZS was connecting only the drive he was using to boot Windows rather than have a multi-boot system.

    >Do you know if these data sets were deleted or just overwritten?

    The blocks of data in the Shadow Volume reflect overwritten data.

    >Do you know if these data sets were the only deleted data sets found on the hard drive?

    The RMP report, translated from Malay into English, says that in the Shadow Volume on MK25, dated Feb 3, 2014, there were two data sets in the current search location (what I have been referring to as 45S1 and 45S2). It says there was altogether 7 coordinates found, including the 2 previously mentioned, believed to be related because of the same PSS Boeing 777 label. It doesn’t say there were many more coordinates found in the Shadow Volume, and it doesn’t say they used a criteria to filter among those many coordinates. It only mentions the 7 coordinates. In fact, they believe that only 5 of the coordinates of the 7 are related, based on the fuel levels.

    As an aside, I disagree slightly with this assessment. More precisely, and as explained in the paper, 8 data sets were found (labeled in the report as Coordinates 1,2,3,4,5,6,6-2,and 7). Of these, 5 were during flight (Coordinates 1,2,3,4,5) and 3 were of the plane parked at KLIA (Coordinates 6,6-2,and 7). Of the three data sets in which the plane is parked, only 1 (Coordinate 6) has fuel levels in the three tanks consistent with the data sets of the airborne plane.

    The RMP report also says that of the hundreds of flight files that were found on the disks, many were created during the installation of the game. I can verify this. I have hundreds of flight files on my computer that were automatically installed and used for sample flights, tutorials, and missions.

    >Do you know if these data sets were all deleted at this time?

    We know that data blocks were overwritten on or after Feb 3. We don’t know if the data blocks were all overwritten at the same time.

  448. Brock McEwen says:

    @RetiredF4: thanks for the reply. I already have tweets implying 19.8 hours transit time.

    But an announcement of an a priori estimate of a maximum transit time – converted to distances by assumed speeds and paths – is a sub-optimal basis for establishing position with certainty. I think a better strategy is to tie the Pinckney and Kidd to a specific NOTAM. Or, if the “exercise” that night did not involve live fire, we should press the 7th Fleet to demonstrate the kind of transparency and accountability required to get to the bottom of this mess, and prove out their whereabouts that evening.

    @all: who has a comprehensive database of NOTAMs for that part of the world which were in effect that night?

  449. Victor Iannello says:

    @Gysbreght: I answered your question. I created many simulations beyond those with automated flight. In manual flight, I can replicate the min/max values of acceleration and rpm. If you don’t believe me, please do this experiment for yourself.

  450. TBill says:

    @Victor
    Re: simulator
    What about, let’s say MH370 was the pre-planned flight for the diversion, but when MAS added Flight 168 22:30 night flight to Jeddah in Jan_2014, perhaps Z used FS9 to visualize how he would have to change up the route for the earlier take off time. Seems to me the Jeddah route then necessitates going out to 1090E and by using McMurdo for waypoint it goes almost exactly around the Indonesian air space boundary. But anyways that would say there could have been other cases (MH370) lost…who knows how long ago they could have been done, or if FS9 was even used.

  451. Victor Iannello says:

    @TBill: I don’t know. Many scenarios are possible.

  452. Gysbreght says:

    @Victor Iannello: With all due respect, you have stated the obvious but did not answer my question.

    I’m interested in the values of acceleration in the particular simulation you did for Recreation of Simulated Flight in the particular report. I expect that these add some perspective to the values recovered from Z’s disks.

    I’m also particularly interested in the engine rpm’s because they are different in your red and blue boxes. According to your report you programmed the intended flight plan into the FMC and conducted most of the simulation with the autopilot in control. I assume that you conducted a normal takeoff from Kuala Lumpur International airport, cleaned up the airplane, reduced thrust to CLB, and engaged the autopilot. The takeoff rpm would then be set by the engine control system and would not be exceeded in the remainder of the flight under AP control with CLB rating selected.

    I’m not asking for replication of “min/max values of acceleration and rpm”.

  453. TBill says:

    @Ge Rijn
    Over on Reddit there is a nice debris report update from MY. I am confused if we already posted here, so I will not repost.

  454. Victor Iannello says:

    @Gysbreght: The values in automated flight are different than the values when parts of the flight are manually flown. In automated flight, the acceleration stays between 0.8g and 1.2g. With the autopilot disengaged, I can replicate the values from the sim data. In my automated simulation, max g of 1.2 and max rpm of 32,328 rpm was realized in the stepped climb between 5N and 10N, and not the takeoff, which produced a max g of 1.1 and max rpm of 29,195 rpm. The max rpm at takeoff depends on how the thrust is de-rated.

  455. Ge Rijn says:

    @David

    Back on items 26 and 27.

    I think indeed the mentioning of: “The fitting on the debris appeared to have suffered a tension overload fracture” indicate a sudden shock broke off this aileron-piece. With the aileron actuators likely in ‘blocked’-mode or ‘damped’-mode. Implicating there was oil pressure in the system the moment the piece broke away.

    Also note most pieces are discribed as having tension-fractures with pulled out fibres with no sign of kinks.
    This also is an indication of sudden tension overload and not flutter-like fatigue fractures. The latter would show clear kinks in the fibres IMO.

    Item 27 the flap fairing piece is fixed to the wing so it could not have seperated due to flutter anyhow.
    To me it’s telling the remaining seal is intact and still in place.

    Knowing this seal is closing the gap between the adjecting flap fairing track that covers the big hinge from the found outboard flap section when the outboard flap is retracted, tells me either the outboard flap section seperated first followed by this flap fairing or the outboard flap was not retracted leaving the gap open.
    Otherwise this seal would not have survived this way IMO.

    Further the report mentions with many items they are consistend with the CSIRO drift model. And I agree they do. But not all with their 35S ‘hot-spot’.
    With this latest 2 items they now have three items istead of one (RR-piece) their 35S can not predict or explain.
    Agian; their ~31.75S till ~29S can include this latest 2 items too.

  456. Ge Rijn says:

    @TBill

    I think @David already posted the same report?

  457. Gysbreght says:

    @Victor Iannello: Thank you ever so much. That is the helpful information I was looking for.

    As to “The max rpm at takeoff depends on how the thrust is de-rated.”:
    It doesn’t make sense to derate takeoff thrust to less than CLB, it is something that Z wouldn’t do.

  458. Victor Iannello says:

    @Gysbreght: My understanding is that B777 pilots regularly select a high OAT (60C, for instance), to reduce takeoff thrust.

  459. Gysbreght says:

    @Victor Iannello: Perhaps Andrew can comment on what B777 pilots regularly select.

    Is the max rpm of 29,195 rpm you got in takeoff obtained with derated takeoff thrust? If yes, how was it defined?

  460. Gysbreght says:

    @Victor Iannello and Don Thomson

    I have some more questions related to the Shadow Copy. Perhaps I shouldn’t expect an answer to all questions, but I’m hoping you would be able and willing to answer some of them. At the very least these questions serve to explain my problems with what Victor has called “ridiculous theories”.

    1. What is the size of a typical *.FLT file saved by FS9/PSS Boeing 777-LR No VC, and what is the size of a file fragment recovered from the Shadow Volume?
    2. The RMP report gives Feb 3, 2014 as the date of the Shadow Volume. Is that the date of creating the Shadow Volume, or the date that its contents were last changed?
    3. If the Shadow Volume was created on Feb 3, 2014, what events could have caused that creation on that date?
    4. If a file containing one of the fragments recovered by the RMP was not deleted prior to Feb 3, 2014, which application(s) would write to a block allocated to that file, after that file was saved to disk?
    5. If that file was modified and then saved to disk, and it occupied more than one block of disc storage, would all blocks allocated to that file be copied to the Shadow Volume, prior to being overwritten by the modified file?
    6. If that file was deleted in the normal way after Feb 3, 2014, would that deletion affect contents or date of the Shadow Volume?
    7. If that file was deleted prior to Feb 3, 2014, would the “shadow copy provider” that is supplied in the Windows operating system, copy the data contained in a block that had been allocated to a deleted file?
    8. How does the shadow copy provider, that is supplied in the Windows operating system, maintain the “point-in-time” view of a volume that is contained in a shadow copy?

  461. Oleksandr says:

    @Don Thompson,

    Re: “The GES received power level measured during the ongoing flight, as described in ‘The Search for MH2370’ paper (Ashton et al, 3rd Oct 2014), was normal.”

    Holland states in his paper that the signal strength 18:22:27 was weaker than it should be: “The associated received signal level and carrier-to-noise density ratio (C=N0) were also unusually low.”

    Would you be able to comment on this discrepancy?

    Re: “Therefore, there is nothing to indicate that, at any time, the AES was failing to accurately track I3-F1 using the aircraft position and attitude data derived from a fully functioning ADIRS.”

    What about unusually low signal strength? Or the very first BFO mysteriously not affected by OCXO warm up?

    Re: “Finally, after much discusssion of the typical warm-up characterisitics of OCXOs, Holland’s paper confirmed that the OCXO used by an MCS-6000 SDU does exhibit an overshoot and subsequent stabilisation of its frequency as the oven settles to working temperature.”

    Holland suggests ignoring 18:25:27 BFO: “As such, the first BFO was deemed untrustworthy”. But he has not provided any supportive reason. If fact, nobody was able to suggest a reason.

    My conclusion is that your original idea deserves a better place than a dusty shelf. Regardless IG’s consensus.

  462. Gysbreght says:

    Perhaps the end of question 4. would be better expressed by:

    “…, after the application that has created the file has saved it to disk.”

  463. DennisW says:

    @Oleksandr

    “The confidence that BFOs are always affected during the process of stabilizing shortly after the return of power works against the mainstream assumption that the SDU was previously de-powered. Or not? Was Don’s idea discarded?”

    It is possible that the frequency of the OCXO just happened to be correct as the OCXO was over-shooting its nominal frequency as part of a normal recovery settling period. There is really no way to know for sure. It is also possible that there were other issues such as antenna steerage problems, and the OCXO was never powered off. Certainly a log on request during that phase of flight is extremely rare. It is also very unclear how and why the SDU was powered off and on, if that explanation is used.

  464. JohnM says:

    @DennisW and @Oleksandr

    Excuse me for jumping in without having made a contribution before, but I know from experience DennisW has an explanation which is 100% reasonable. This is because I have observed the behaviour of some (non-Inmarsat) ground based satcom kit many times over when I have just switched it on cold. It does exactly what Dennis describes. In fact it overshoots/undershoots several times before stabilising. It transmits ‘successfully’ for short periods until it properly stabilises. This is because the Tx enable circuit on the amplifier output stage only knows if the PLL circuit around the oscillator is within a valid range, which it is briefly several times during stabilisation. It has no more sophisticated idea of stability. Perhaps it could be better designed with a hold off timer – but it isn’t. I guess the extra cost is just not worth it as it never transmits so far off frequency as to interfere with another transmission.

    On the low signal level/signal quality – I can’t see how one could conclude anything particularly useful from this. There are two many unknowns. Assuming other parallel transmission from other aircraft where checked at the relevant time this would eliminate the GES and satellite from impacting the level and quality. This is why the low signal is assumed to be from the aircraft. However, there are no end of reasons for a briefly low signal. With only one sample every hour or so it’s going to be hard to figure anything out from this. Theoretically the aircraft can exceed the antenna tracking speed and hence in general it’s tracking well, but briefly maybe not.

  465. Mick Gilbert says:

    @Victor

    Thank you for those answers. With regards to creating the data sets, I had read your and Yves’ paper a few times and was aware that you had identified a method that produces the same result. I was wondering whether you might have also looked at alternative methods.

  466. Mick Gilbert says:

    @Ge Rijn

    Thank you for those estimates and references.

  467. Don Thompson says:

    @Oleksandr,

    As a reminder, BFO is a measure of the frequency offset from the specified carrier in the burst received from an AES on a given R, T, or C channel.

    Maybe you could explain why you might expect the frequency determination at the GES to be affected by the lower ‘received signal level and carrier-to-noise density ratio’, however, do note that the demodulator locked onto the burst, processed the burst, registered a non-zero (but correctable) Bit Error Rate, while successfully extracting the received ‘signalling unit’.

    9M-MRO’s ‘cold’ start up, at 12:50z 7th Mar, while on the ground at WMKK prior to the MH370 service also resulted in recorded non-zero BER metadata (as did the same event on 23rd Feb). I understand that adjacent buildings (cargo terminals, hangars, for example) can contribute to multipath issues, as can the wing as the aircraft rolls.

    Your question the ‘very first BFO‘: for a burst at what time?

    The widely discussed idea that the aircraft was undergoing a manoeuvre as the 18:25 Log On was processed could contribute to a coincident burst received with low C/No & BER!=0. Subsequent to the Log On Request, the metadata for remaining bursts recorded in the GES log are unremarkable.

    It’s notable that you draw attention to Holland’s comment, copied below:

    Regarding the log-on event at 18:25:27Z for MH370, it was noted that there was a non-zero bit error rate (BER) associated with the log-on request at that time. The associated received signal level and carrier-to-noise density ratio (C=N0) were also unusually low.

    According to ‘The Search for MH370, Table 1 Signalling Message Parameters from Flight MH370 (Malaysian Government, 2014)’ the RxPwr (signal level) values measured for the two in flight, post power interruption, Log On Request bursts were ‘high’ (-52.3 at 18:25:27 and -51.0 at 00:19:29) compared to those recorded during the Log On sessions. That the RxPwr level of the initial Log On Request burst is different should be expected: in its Log On Acknowledge to the AES, the GES advises an EIRP level at which the AES should transmit. It’s possible Holland didn’t understand that point of detail in the AMS(R)S spec.

    :Don

  468. Don Thompson says:

    @Brock

    Have you considered the 15 naval vessels dispersing from India’s MILAN 2014 exercise held in and around the Andaman & Nicobar Islands during the last week of February 2014, any live firing possible from them?

    Or the People’s Liberation Army – Navy (PLAN) Amphibious Task Force undertaking an exercise in the SCS during March 2014?

    Or the PLAN task force that sailed from the SCS out into the Indian Ocean and back again, via the Sunda & Lumbok Straits in Feb-Mar 2014.

    :Don

  469. Don Thompson says:

    @Gysbreght

    Reverse order:

    8) Considering a simple Windows 8 ‘desktop’ using a single NTFS filesystem, the Shadow Copy is maintained as a set of hidden folders within the system volume (an NTFS file system, C:). It is provided, typically, to enable roll back from ‘Patch Tuesday’ updates, driver updates, and similar. 3rd party backup applications may also exploit the service but for complete recoverability a full, offline, volume copy is necessary prior to the Shadow Copy creation. When a Shadow Copy is created, the System simply marks a snapshot of the file system metadata as a point-in-time fromm which it begins to log changes and deletions. The view of a Shadow Copy, at any time, is a composite of the live ‘visible’ file system plus the changes and deletions recorded in the ‘hidden’ Shadow Copy. When a Shadow Copy exists in the file system ‘Previous Versions’ may be reviewed in Windows Explorer.

    7) Again, in this simple case, changed or deleted files are not copied anywhere: the shadow copy service merely manages the file system’s folder/directory and file ‘pointers’. A deleted file isn’t removed, its file system references are simply reserved in the Shadow Copy (there is no segregated, by physical reference, area of the disk device). Once the Shadow Copy is created, any document that is edited and saved is written to free file system blocks while pertinent metadata relating to the previous copy is reserved in the volume Shadow Copy. In this form, the configuration is not intended to be a complete ‘backup’ solution, its simply a mechanism to roll back inadvertent deletions or unwanted changes.

    6) The Shadow Copy tracks changes after its creation. Multiple shadow copies may exist on a volume. For example, it’s feasible to create one at hourly intervals during the working day.

    5) Shadow Copy operates with NTFS constructs, it has no concept of the underlying block storage device. In the single volume case there is no discrete, reserved ‘disk’ area for Shadow Copy, the service simply uses the available NTFS resources in the volume mounted as C:

    4) In the single volume case, no discrimination is made between applications writing data to the file system. However, Windows doesn’t track the swap file changes, as an example.

    3) Creating a Shadow Copy on the volume could be invoked interactively, as part of an application or driver update, or by MSFT’s monthly ‘patch Tuesday’ rollout (which users may accept the system to invoke automatically or only after explicit approval).

    2) & 1) I’ll defer to Victor, should he choose to answer. I haven’t reviewed the RMP documents recently.

    :Don

  470. Andrew says:

    @Victor
    @Gysbreght

    RE:“It doesn’t make sense to derate takeoff thrust to less than CLB, it is something that Z wouldn’t do.”

    “My understanding is that B777 pilots regularly select a high OAT (60C, for instance), to reduce takeoff thrust.”

    “Perhaps Andrew can comment on what B777 pilots regularly select.”

    As Victor said, it is common for pilots to use high ‘assumed’ temperatures (around 60°C) to facilitate reduced thrust take-offs. When the thrust is set to climb thrust after take-off, the FMC automatically selects CLB1 or CLB2, depending on the amount of take-off thrust reduction. In a reduced thrust take-off, the CLB1 or CLB2 RPM is less than the take-off thrust RPM, but the CLB RPM is higher. The FMC selects CLB thrust for an enroute climb, so it would not be unusual to see the associated RPM being higher than the RPM during take-off.

  471. David says:

    @Ge Rijn. Re the item 27 seal, 2/3 or more still being there. I do not follow why it would not remain attached if that section went before or with the fairing rear. Can you amplify?

    It will have been broken when the left side of the front fairing separated from the right. Any ideas? I assume the right stayed at least momentarily but in that case what could cause the left’s separation, and in that condition?

    The rear fairing, item 3, is here:
    https://www.atsb.gov.au/publications/investigation_reports/2014/aair/ae-2014-054/
    It looks like left forward although the identification stencils are depicted further aft. It is possible the front and rear sections separated together. But why would both fairings split lengthways? Asymmetry from whence?

  472. David says:

    One other point about ATSB debris investigations. Some, myself included, have wondered aloud why the ATSB debris assessments mostly have been just about identification. A clue is at the rear fairing reference above, which says, “At the time of writing, ongoing work was being conducted with respect to the marine ecology identification as well as testing of material samples. The results from these tests will be provided to the Malaysian investigation team once complete”. So, hopefully more will emerge in the final report.

  473. TBill says:

    @Capt Kremin
    “As far as the SDU behaviour, it was possibly an attempt by the perpetrator to ensure all communications from the cabin to the ground was impossible until after the aircraft was depressurised. The depressurisation would ideally be a slow event, in order to avoid suffering the Bends.”

    I had not considered The Bends in the depressurization scenarios. I suppose the pilot’s O2 mask would help there.

    @David
    “…hopefully more will emerge in the final report.”

    I guess you are correct David: there is an implication we may learn more in the final report. That’s good and bad. I suppose any serious efforts to 2nd guess or re-interpret the data must wait until the final report.

  474. ventus45 says:

    @TBill

    ” I suppose any serious efforts to 2nd guess or re-interpret the data must wait until the final report.”

    “Wait for the final report” – or variations of it.

    How many times have I heard that over the years, concerning both this, and other accident investigations.

    Those familiar with the ATSB’s “track record” on Mildura, and in particular, the Norfolk Island ditching of the PelAir Westwind, will know full well what I mean. My’s track record on MH370 is many times worse than the ATSB’s.

    My has made numerous “promises” in the past, that have not been honoured. From very early on, and still very near and dear to Victor’s heart, (I presume), was the early promise by MY to him personally, of providing the detailed post Penang radar data. Did you ever get it Victor ? Silly question I know.

    What are we going to do when “the final”, turns out to be as useful as “the interims”. Cry and chuck some tanties like a bunch 4 year olds ?

    No. I think it is high time we pre-empt the “fob off” that the final will almost certainly be, and made a detailed list of what we “demand” is included, in detail, and publish it, very soon.

    To that end Victor, a new thread ? “What we demand to know” or similar ?

  475. Victor Iannello says:

    @John M: Welcome, and thank you for your comment.

  476. Gysbreght says:

    @Don Thompson:

    Thank you for taking the time to answer so many of my questions.

    I would like to understand the process that copied those file fragments to the Shadow Volume where the RMP found them, and how that fits in the timeline of the life of the original files, from the time they were created by FS9 and written to disk until they were deleted. Of particular interest is the possibility that those files were being edited during their life on disk. I believe a *.FLT file can be opened with any text editor such as Notepad and edited at will. If the lifecycle of a *.FLT file had consisted of nothing more than that it was created in FS9 while ‘playing the game’, saved to disk, and then deleted, I don’t understand at which point in that lifecycle a fragment of that file would have been copied to a Shadow Volume.

    My understanding of the process is as follows and please correct me whenever I’m wrong. The RMP recovered incomplete fragments of the original files from a Shadow Volume on the MK25 drive. Why only those fragments and not the complete files? The VSS service is triggered by an application write I/O request. The Volume Shadow Copy Service tells the writers to temporarily freeze application write I/O requests. It copies the contents of the block(s) written to in the Shadow Volume before allowing the write operation to proceed. I don’t understand from your explanation why a write operation occurs to a block that is not empty or free but is allocated to an existing file. Can you elaborate on that point?

  477. Ge Rijn says:

    @David

    I’ll try to amplify..
    First take al ook at this picture I suggest:

    http://www.airliners.net/photo/Japan-Airlines-JAL/Boeing-777-346-ER/576045

    You see the complete no.1 (same no.7 right wing) outboard flap track fairing assembly when the outboard flap is deployed.
    You see the gap between the forward wing-attached piece (item no.27) and the piece that covers the outboard flap hinge.
    When the outboard flap is retracted that gap is closed for the aft piece fits in the forward piece sealed off by the seal.

    Now imagine flap fairing no. 7 (item 27) seperating with a retracted outboard flap before the outboard flap section/fairing seperated.
    The complete item 27 seal would have been pushed forcefully against the adjecting flap fairing piece. Damaging or removing the whole front edge of the piece and seal. I think this is not what happened.
    You see only the bottom part of the seal and edge is damaged.
    This is what I think would happen when the outboard flap was deployed and the gap between both pieces was avoiding a complete break-up of this edge and seal. Or when the outboard flap section and fairing seperated first. For there is only damage to the bottom of the seal and edge.

    The longitudal split of the piece I rather explain by an impact force coming from underneath splitting the piece in halfs. You can also see the structure of item 27 is bend inwards at the rear of the crack.
    If you look carefully to the comparring picture in the report between an intact piece and item 27 you see theere is an overall line in the middle of the intact piece that devides the piece in two halfs.
    I assume the piece is assembled out of two halfs merged together.
    I’ve got another picture which also shows this deviding line on the outside. Look at picture 8 by scrolling down (with thanks to Jeff Wise, again using his tremendous resource of info):

    http://jeffwise.net/2016/03/10/mh370-debris-storm/

    Anyway I conclude for now this piece was damaged and seperated by impact forces coming mostly from underneath causing sudden tension overload damage like most of the other pieces evaluated in the report.

  478. Paul Smithson says:

    “The bends” is caused by dissolved nitrogen in the blood becoming gas bubbles on an abrupt drop in pressure (like opening your can of coke). You need to have been breathing air at pressure for some time to get enough dissolved in the blood in the first place. Given that air was being breathed at a pressure altitude of (say) 7000ft (not pressurised air) and the rather smaller pressure differential in event of decompression at FL350, I rather doubt the bends is an issue. Oxygen mask or not has nothing to do with it.

    As regards BTO thermal transients. I also lean towards the notion of brief “in-spec” nominal frequency, overshoot, then undershoot (with drift) up to final stabilisation. Since all of this is occurring once nominal operating temperature range has been reached (otherwise it wouldn’t start functioning at all, right?), why does the thermal drift magnitude/duration depend on how long it has been off?

  479. Ge Rijn says:

    @TBill

    Although I understand your remark: “I suppose any serious efforts to 2nd guess or re-interpret the data must wait until the final report”, I, like @ventus45, don’t agree.
    I assumme you don’t meant it that literally. When so @VictorI can close this blog till the end of the year.

    My line of thought in this is the final report hasn’t been written yet (I suppose). Therefore every serious effort to 2nd guess or re-interpret the data can have the potential of adding new thoughts and info to the officials who are reading this blog too or are in close(r) contact with independent investigators (like the IG parcitipators).

  480. Andrew says:

    @Paul Smithson

    See the following FAA document:
    ALTITUDE-INDUCED DECOMPRESSION SICKNESS

    Atitude-induced decompression sickness (‘the bends’) doesn’t require an abrupt drop in pressure or the pre-breathing of pressurised air. When I was in air force we had limitations on the maximum altitude that we could fly in unpressurised aircraft for that very reason. We had several incidents of pilots getting decompression sickness after flying at altitudes around FL210.

  481. Ge Rijn says:

    @David

    Small correction; I mean the no.2 left wing flap fairing same as no. 7 (not the no. 1..)

  482. Paul Smithson says:

    @Andrew, thanks for the reference. I stand corrected! Moreover it makes clear that pre-breathing O2 does help insofar as it allows wash-out of dissolved nitrogen from the blood.

  483. TBill says:

    @Ge Rijn @Ventus
    What I meant was, we of course can 2nd guess on the blog, but if I am expecting, for example, US Govt or NTSB someone to step in and give us the truth or an alternate “most likely” scenario, that has to wait until we hear what MY is going to say. What we are missing in this accident is an official list of possible likely causes.

    @Paul Smithson @Andrew
    On decompression sickness, the other question would be does a pilot with a pressurized O2 mask have protection? I don’t know if its the pressure of the surroundings or the pressure in your lungs that tends to cause it.

  484. Paul Smithson says:

    @TBill If I understand correctly,

    its to do with how much nitrogen you have dissolved in your blood (which is more, the higher the pressure air you have been breathing). then how big a pressure drop (bigger drop encourages more nitrogen to come out of solution).

    breathing oxygen in advance works because you have steep partial pressure gradient for nitrogen which encourages it to “wash” out of the blood stream. then if a pressure drop does occur, you don’t have a load of nitrogen in your blood ready to come out of solution.

  485. Ge Rijn says:

    @David @TBill @all

    The very limited information from ATSB and MY about the forensics of the debris with only a few hints: tension overload, pulled fibres with no fibre kinks, show to me they are adressing to a low AoA, relatively slow sudden impact. Not a high speed dive, seperating all those parts by flutter or other aerodynamic overload.
    That’s just impossible regarding the damage on the affected pieces found.

    By these very basic debris reports, it’s obvious to me the plane did not impacted the water with very high speed and very high AoA (~vertical).
    Or it was a AF447 kind of impact or it was an attempted ditch.

    The latter is the most obvious regarding the pieces found to date and the kind of damage they suffered according the reports.

    Denying this obvious facts is IMO denying a very possible reality.

  486. TBill says:

    Re: ELT discussion
    The take away for me is we ought to consider giving the Cabin crew an emergency ELT if a hijacker is locked inside the cockpit, we at least need a distress signal.

    A couple of recent news stories have highlighted the air marshalls on US flights. A couple weeks ago one air marshall accidentally left her firearm in the aircraft restroom. Anyways, what do the marshalls carry beside firearms? Satellite phone? ELT emergency beacon in case someone locks the cockpit door? And by the way, who is going to lock the cockpit door knowing there is possible air marshall telling the ground what is happening? Huge deterrent…

    …something rouge could still happen on a greatly reduced probability, since the safety issue has been recognized and addressed to the extent possible given current aircraft design.

  487. Paul Smithson says:

    @GeRijn. I’m wondering what your professional background is to pass such sweeping and assertive judgements on the basis of some pictures with additional commentary on observations. Do you hold any qualification in accident forensics, materials science, failure analysis or anything similar? I would love to be persuaded by your logic but I cannot for the life of me see how you can assert that x,y,z observation indicates near-certainty with regard to AOA or speed of impact.

  488. Ge Rijn says:

    @Paul Smithson

    Just stay on the content I suggest and make your own logic.

  489. Brock McEwen says:

    @Don: thank you very, very much for those leads. It is greatly appreciated. I will add them to my database.

    Believe me: I am not trying to weave a “shoot-down” theory: I am merely trying to stick pins in a map, and draw range rings around them, so that we can have a comprehensive and reliable picture of military radar coverage of the pertinent regions at the pertinent times.

    And this is strictly in pursuit of path plausibility assessment. When the path was thought to be within a narrow range, the IG engaged in precisely this exercise; now that the search box is searched out, the path proposals are diverging – this begs for a more comprehensive and in-depth study of radar coverage. So all data is most welcome.

    Including the positions of US radar assets. Do you have any information at all regarding those assets?

  490. Don Thompson says:

    @Gysbreght

    You began a sentence with “The VSS service is triggered…”

    That’s the point at which your misunderstanding begins. I’m not prepared to provide a complete tutorial on the subject.

  491. Don Thompson says:

    @Brock,

    The only knowledge I have of Pinckney & Kidd, prior to 7th March, is their port visits in the preceding weeks. You can find that information on each ship’s Facebook page & correlate further with media outlets local to the port locations.

    :Don

  492. Paul Smithson says:

    @Brock. How do you explain non-observation of MH370 by (known location) Indonesian radar between 1800 and 1840? If your starting point is that the plane must have been somewhere that radars couldn’t see, then it wasn’t (anywhere) in the Malacca Straits since the entire track was, in principle, detectable by Thai, Malay, Indo radars all the way through to roughly 92E. That’s not including Andaman based radars, which may have extended the “continually detectable” window still further. I don’t understand how establishment of the position of a US warship is going to change anything.

  493. Brock McEwen says:

    While I have the mic:

    (all values $USD)

    According to IATA, commercial airline revenue for 2014 alone was USD $751B – of which 72% was passenger ticket revenues.

    The figure I’ve heard bandied about for the search to date was $200M – incurred mostly in the first year. While this cost was borne primarily (and inexplicably, if you ask me) by Oz taxpayers, the benefits of intel gained from a successful search accrue to the commercial airline industry – so one would hope amends would be made somehow, somewhen. The airline industry, of course, would in turn recoup this expense from its customers.

    I estimate a search extension up to, say, 26°S would cost roughly $120M more (so, $320M total).

    (This could have been considerably less, had the search been executed more efficiently, by erring on the side of caution when setting the original NE and SW search box bounds – but this is a digression.)

    According to some on this site, continuing up to 26°S would guarantee success, because the ISAT data is guaranteed to be both authentic and accurate. (And military radar is, they would have us believe, arbitrarily porous – but this is another digression.)

    If you spent $400 on plane tickets in 2014, you would have fully funded the search cost to date by paying an 11 cent levy; had you chosen the deluxe, “actually find the plane” option, it would have cost you 17 cents.

    It would be instructive to canvas the general flying public for their sense of the business case for investing the extra 6 cents. We could set up little polling booths on every jetway.

    That the world’s decision-makers would rather charge us 11 cents to not find the plane than 17 cents to find it should tell us something about their true belief in the ISAT data.

  494. Gysbreght says:

    @Don Thompson:

    Thanks for telling me where my misunderstanding begins.

    BTW I found in a paper by the Independent Group dated 2016/08/15 “The data were saved in a Shadow Copy Set, and were last modified on 3rd February 2014”.

  495. Andrew says:

    @Brock McEwen

    RE: “The figure I’ve heard bandied about for the search to date was $200M – incurred mostly in the first year. While this cost was borne primarily (and inexplicably, if you ask me) by Oz taxpayers…”</I"

    There's a common misperception that Australia funded most of the search cost. In actual fact it was Malaysia. According to the ATSB:

    "In 2014, Australia committed $90 million to the search for MH370, including $60 million to support the underwater search activities.

    The People’s Republic of China committed $20 million in the form of funding and equipment.

    In addition to its earlier commitment to match Australia’s contribution of $60 million, Malaysia agreed to fund the balance of the costs associated with searching the entire 120,000 square kilometre search area.

    The cost of the underwater search depended on a number of factors, including challenges with equipment in dangerous waters, weather and the length of the search, and was around $200 million."

  496. Brock McEwen says:

    @Paul: you are preaching to the choir. My personal opinion is that the remaining path options are already strongly counter-indicated by “only” the many primary radar assets you’ve listed. But others would have us believe enough of these military radars were “turned off”, or “asleep at the switch”, or “not operational” to allow MH370 to sneak through undetected. It is for their sake that I am now compiling a comprehensive picture of radar coverage; for those folks, apparently, the presence of US radar would indeed make a difference.

  497. Brock McEwen says:

    @Don: thanks – appreciated. For completeness: do you have any knowledge of Pinckney, Kidd, or any other radar-equipped US asset which could have been within 300 nmi of any IG member-proposed MH370 path on the actual March 7th (or morning of the 8th) itself? Your response only covered dates PRIOR to the 7th. Sorry to be a stickler. Thanks again.

  498. David says:

    Treasure hunt suggestion – posted because it was earlier of interest:

    https://www.uq.edu.au/news/article/2017/05/turn-mh370-search-treasure-hunt

    This does seem outside his normal field.

  499. David says:

    @Ge Rijn. Thanks for your comments. The underwing photo is useful for visualisations. I agree that the damage to the seal at its bottom needs to be accounted for though it may have resulted from being torn away when the frame separated from the missing part forward fairing.

    Also the front section might have rotated downwards while the whole rear fairing, or at least the part not recovered and its fixing, was there still, acting as a fulcrum.

    Besides, with the lip of the rear fairing overlapping the front as in the underwing photo and I suppose the seal at the bottom might well be clear of its abutting face.

    Since there could be these and other explanations, I do not see this adding much to your belief that the flaps were down. That needs to have further grounds stronger that those supporting the current belief they were up; the flaperon/flap witness marks.

    “Or when the outboard flap section and fairing separated first”. Would there have been seal damage in that case?

    “You can also see the structure of item 27 is bent inwards at the rear of the crack”. Not too sure I follow but if you are talking about the frame, to me it looks bent rearwards, indicating the front section levered off from its front and possibly then twisted, its front going left. In the item 27 photo comparing that with the undamaged item the damaged frame tipt is seen as bent outwards also, indicating the other part of the frame (at least ) remained attached to the wing.

    “…there is an overall line in the middle of the intact piece that divides the piece in two halves”. I see no line other than where some dirty water has been. Were this a butt joint there would be fasteners or at least one butt strap. I think that at the JW picture 8 is a water mark too. (Thanks for those photos: I can see item 27 is from the rear fairing bottom not the left side as I had thought).

    None of which means that the fairings were not detached by water forces, which as you say could account for the longitudinal splits; and other breaks for that matter. However to me the forces are magnitudes higher than those of a ditching, even 767 Comoros style (videos post crash show the fairings remained intact there): a much higher speed instead. That was flaps up but the water impact on the forward fairings would be similar.

    The debris origins and asymmetry do to me suggest right wing down but then again the Boeing simulations indicate that to be less likely.

    Thanks for the discussion though I doubt we will get much further. Those with the parts in hand (the rear section and other parts should be returned by the ATSB to Malaysia since they will have responsibility for reconstruction) and with access to Boeing advice should do better.

    To me, the above aside, what is the most striking part of the new evidence is the honeycomb water-logging, apparently not encountered hitherto.

    @ Don Thompson, I see you are reported as believing there was a mid-air break-up, without embellishment. Can you expand on this and the fairing separation in particular?

    https://blogs.crikey.com.au/planetalking/2017/05/03/malaysia-analysis-of-mh370-debris-supports-mid-air-breakup/

  500. DennisW says:

    @David

    Poor suggestion. The reality is that Freescale (now merged with NXP) had a non-GAAP operating income of $2.5B+ in 2016. Freescale also had 20 employees on MH370. If I was the president of Freescale I would certainly be writing a check for $100M. It would be hugely symbolic and appreciated by their employees, by their stockholders, and by their customers. It would amount to a lot less than $100M after tax.

    Suggestions that Boeing fund the search are ridiculous. There is no evidence that anything was amiss with the aircraft.

  501. DennisW says:

    @David

    The reality is that Freescale (now merged with NXP) had a non-GAAP operating income of $2.5B+ in 2016. Freescale also had 20 employees on MH370. If I was the president of Freescale I would certainly be writing a check for $100M. It would be hugely symbolic and appreciated by their employees, by their stockholders, and by their customers. It would amount to a lot less than $100M after tax.

    Suggestions that Boeing fund the search are ridiculous. There is no evidence that anything was amiss with the aircraft.

  502. David says:

    @GeRijn. Correction, 3rd Para, “….with the lip of the FRONT fairing overlapping the REAR…”.

  503. Don Thompson says:

    @David,

    Concerning Ben’s lastest MH370 piece. I don’t understand why Ben Sandilands attributed that comment to me, I have had no direct correspondence with Ben for some weeks. I suspect he simply caught some Twitter traffic originated by Mike Exner that linked to the most recent Malaysian Debris Examination Report.

    That said, I do consider that a high speed descent is the most likely scenario leading to impact and that (destructive) aeroelastic flutter may have been experienced during that descent. As Mike Exner has noted, the debris analysis has described that break lines in ten of the recovered items indicated “pull(ing)” or tension fracture in the composite skin fibres: those parts include trailing edge components, either fragments of moving surfaces or ‘fixed panel’ secondary structures.

    :Don

  504. Don Thompson says:

    @Brock

    To clarify above…

    “The only knowledge I have of Pinckney & Kidd, prior to them joining the search for MH370 , is their port visits…”

    :Don

  505. Ge Rijn says:

    @David

    Thanks. Like you say, it’s about the discussion. And to be more specific I don’t ‘believe’ the plane entered the water in a ~low AoA and relatively low speed. I just argue this is now the more plausible explanation based on the kind of debris found, their locations and the kind of damage in general. Ofcourse strong evidence is needed which only the official investigators can provide (for they have the pieces).

    On your remark; ““Or when the outboard flap section and fairing separated first”. Would there have been seal damage in that case?”.
    I argue that if the outboard flap was retracted or deployed and seperated first together with the aft fairing the sealing would more likely survive the way it did. When deployed I think even more likely for then there is this gap between the front and aft fairing.

    “You can also see the structure of item 27 is bent inwards at the rear of the crack”.
    I don’t mean the frame (which is not bent rearwards if you look at the picture taken sideways in the report) but the honeycomb skin structure at the rear.
    The middle line I’m not sure of what it means, I just see it in both pictures. No way for me to confirm the piece is assembled out of two halfs. Maybe someone else can?

    On water impact and the Comoros-ditch intact fairings and your ‘right-wing first’ remark I ask you to look at the Hudson-ditch pictures I add:

    http://www.9-1-1magazine.com/Pressler-FDNY-Miracle-on-the-Hudson

    Take a look at picture 5 scrolling down in this article (best is to download and then zoom in).
    Also in this crash the first section of the right wing outboard flap broke away with big shunks of its trailing edge missing too (you can see this first outboard flap section hanging up side down under the wing).
    Then you can also see the same flap track fairing is missing a big shunk on its inboard side.

    The next picture shows clearly the Hudson-ditch similar right wing damage compared to MH370 too:

    http://www.dailymail.co.uk/news/article-1120682/Splashdown-The-moment-Captain-Cool-landed-Airbus-Hudson-River–saved-155-lives–VIDEO.html

    You can clearly see the missing first section outboard flap but also the almost completly broken away aileron behind its hinges. The left wing lost its engine but suffered less trailing edge wing damage.
    We know this plane entered the water ~level.

    The next picture shows a more close-up from the Hudson-ditch right wing aileron damage. Scroll down to picture 3:

    http://www.eirinika.gr/article/109669/story-ola-ta-thanatifora-dystyhimata-ton-airbus-a320-apo-1988-eos-simera

    Look at picture 3 scrolling down. You can see the aileron broke of in shunks just behind its hinges. Similar damage you see on item 26 and on the MH370 flaperon and outboard flap section.

    I know it’s no proof I know but IMO this all indicates clearly a ditch-like MH370 crash could have caused the damage we see on most recovered pieces.

  506. Don Thompson says:

    Ge Rijn,

    Might be worth reviewing how N106US was recovered from the Hudson. It languished on the river bank for some time, right wing down deep into the water, before recovery onshore. Just recommending a look, that’s all.

  507. Ge Rijn says:

    @Don Thompson

    Yes I’ve seen the movie before and I know the arguments the wing damage was more likely caused by the river(bed), the ice and the lifting (which was on the wing roots). I think the damage was mainly caused during the ditching cause it’s typically those parts that broke which you would expect during a ditching. Explaining this specific pattern by the river(bed), ice and lifting is IMO looking for a far less likely explanation while the more obvious is at hand.

    The NTSB-report mentions no details about the damage and how it occured.
    If someone has any proof it happened the way you seem to suggest I would gladly see it.

  508. Don Thompson says:

    Ge Rijn,

    Are you sure that it’s not the case that you simply want the damage to the stbd aileron N106US to be explained as directly caused by the landing on water?

    The damage to the stbd aileron is not consistent with damage to any other control surface on N106US. Note that even flap track fairings on both wings were all intact, aerodynamic fixtures protruding below the surface of the wings.

    A large FDNY tug manoeuvred around the starboard side of N106US during the rescue, it could certainly have caused the damage. It’s also likely that the damage was caused by impact with underwater obstacles during airframe’s drift or tow to, or at, the wharf where it was finally recovered from the river.

    While the starboard engine remained attached, the hull’s list to that side was immediate, no images show the stbd aileron’s condition prior to recovery. But then, you probably are aware of that.

  509. Gysbreght says:

    @Ge Rijn: “The NTSB-report mentions no details about the damage and how it occured.”

    The NTSB Docket on the A320 Ditching accident in the Hudson River contains the Structures Group Chairmans Factual Report.

  510. buyerninety says:

    @Ge Rijn
    Like Paul Smithson, the debris information suggests to me a different explanation
    than your conclusion.
    Most of the debris pieces are typically described as exhibiting damage to the
    (composite) fibres “in tension” – this suggests a ‘tearing off or tearing apart’.
    (Very few are noted as having the fibres damaged ‘in compression’, which would have
    been an indication of impact damage.)
    As regards those pieces which came off the wing, this suggests to me the aircraft
    approached or exceeded its airframe ‘airspeed limit’and probably shed parts off its
    wing (and possibly one or both wings may have eventually separated from the passenger
    section).

    @Paul Smithson or all
    I don’t understand the significance of the ‘kinking’ wording (- means ‘twisting’
    of the fibres? or meant to be an indicator of vibration damage?).

  511. Paul Smithson says:

    @buyerninety – i’m sure i’m the wrong person to address that query to!

  512. DennisW says:

    @Don Thompson

    I am sure the debris collected so far has been examined by people skilled in the aircraft crash forensic art. If the damage was not consistent with a high speed termination you have to believe that would have been made known. While I am not in love with the tripartite folks, I do not believe they would intentionally mislead or withhold information that would be so important to the search planning. I don’t think you can second guess them on basis of photographic examination.

  513. Don Thompson says:

    @DennisW

    Apologies, I don’t understand the intent of your comment.

  514. Gysbreght says:

    @buyerninety: “I don’t understand the significance of the ‘kinking’ wording”

    FWIW my guess is that you would see “kinked” fibers in the skin of a honeycomb sandwich panel if you separate the skin laminate from the honeycomb core, and then bend it until it breaks.

  515. lkr says:

    @Dennis: Thanks for the reality check. And a reminder why it is that people here shouldn’t be defensive when they are asked about their expertise!

  516. DennisW says:

    @Don Thompson

    Just saying that I have the same view as you do.

  517. Brock McEwen says:

    @Don: thanks for that. Just for completeness:

    A) what are the first times & positions you have for Pinckney and Kidd UPON joining the search?

    B) trying again: do you have any knowledge of any other US radar detection capability which may have been within 300 nmi of any IG member-proposed path at any time on March 7 or 8, 2014?

    Thanks again.

  518. David says:

    @Ge Rijn. On the damage to the Hudson aircraft, looking at the photos and videos it is hard to sort out what damage occurred during the landings, what was during the rescue and what was in salvage. For example the left outer front flap was missing during rescue but I doubt it came off while landing.

    Also, Sullenberger described the noise of hitting a thick flock of these birds and obviously there could have been impact damage to other than just the engines.

    Further, from the salvage video it seems there was more than one attempt to fit the slings and lift and there might have been damage done then, perhaps including damage to the left wing leading edge flap above the engine.

    Maybe the structures attachments in Gysbreght’s NTSB docket would disclose more.

    As an aside the rescue was lauded but I reckon the salvage also deserved high praise. The ice, current, murky water; getting those cranes slings and barge and divers organised, then the diving and getting the slings in place, controlling the passing the ice problem, dealing with that strong and reversing current, and murky water….

    @Don Thompson. Thanks for clearing that up.

    About flutter causing in flight separations, that would entail the aircraft being at a speed outside the flutter envelope. Following a Gysbreght analysis of an ALSM simulation it was apparent that the load factor in the bottom of its spiral would have exceeded the aircraft’s structural envelope, while speed was still within envelope. Whether a wing came off or not from this, there could have been control surface separation or part thereof in overstress.

    Of the 2016 Boeing results, the ATSB reports, “In some simulations, the aircraft’s motion was outside the simulation database. The manufacturer advised that data beyond this time should be treated with caution”.

    It also said, “Some simulated scenarios also recorded descent rates that were outside the aircraft’s certified flight envelope”. I presume it means the aircraft’s speed as distinct from flutter, though even that is uncertain.

    All in all it seems we need to await a more definitive statement.

  519. Brock McEwen says:

    @All re: derivation of 4600 microsecond incremental offset for R600 BTOs: I’ve been told the 4600 value was derived from “more than 30 records” (per Peter Foley, communicated via Dan O’Malley) from flights over a two-week period prior to March 7, 2014 (per Chris Ashton, communicated directly).

    I vaguely recall a reference to this incremental offset varying from flight to flight. I wish to quantize with precision this instability over time – with particular focus on variability over a two-week period. I wish not to re-invent any wheels: has anyone done any work on this yet? Regardless: can anyone refer me to the best available information on the subject?

    Thanks!

    FYI: if the >30 records indeed generated >30 unbiased and independent estimates of the offset applicable to the flight in question – and 4600 is their mean – then I will happily withdraw my concern over possible incremental error in the placement of Arcs 1 and 7 due to misestimation of this offset. Even if they rounded the 4600 (something they should not have done), I don’t see much incremental impact on the 95% confidence intervals for arc positions. However, if these records were not independent (e.g. they seem always to occur in pairs – to the extent correlation within each pair is strongly positive, the number of truly independent observations reduces by half) or are biased (by, eg, drift over the studied period), we may still have a problem.

  520. buyerninety says:

    @Gysbreght
    That sounds like a reasonable guess.

    @Oleksandr said “Anyhow, a combination of a single physical impact at the left side of the
    EE-Bay followed by isolation of the main AC buses by the crew (to prevent potential fire spread)
    explains almost all the observed features of MH370 flight”.

    I assume your theory includes damage to the ‘Left System Card File’ (‘LSCF’ circuit cards) or the
    cabling to it, located immediately forward of the P100 Power Panel.
    E.g., card file position:

    A6; Left ACIPS (Airfoil & Cowl Ice Protection System), {includes Left Engine Anti Ice}.
    Ice buildup about the engine cowl intake could result in reduced left engine performance.

    A10; Left FODC (Fire/Overheat Detection Card).
    Could reduce or undo aircraft automatic system actions to a fire?

    A12; Left ARINC Signal Gateway

    A14; Left ECS (Environmental Control System)

    A4; HYDIM (Hydraulic Interface Module) C-L
    A17;HYDIM L
    Could affect aircraft control surfaces?

    A18; PSU Linear/Monitor
    A19; PSU Preregulator

    ____________________________________

    I’m a bit surprized you haven’t stated your theory along the lines of, e.g.;
    Circumstance occurs,
    Shah diverts back to Malaysia and reprograms the autopilot (or instructs
    Hamid to reprogram autopilot),
    Hamid dons oxygen mask,
    Shah descends into the EE bay (say to deal with fire),
    Shah is rendered non-concious by smoke inhalation,
    Male flight attendants attempt to retrieve Shah/deal with fire,
    Fire causes loss of outer hull pressure integrity,
    Pressure drops & masks drop (some attendants maybe on portable cylinders),
    Hamid does not descend aircraft due to not noted or inoperative barometric pressure
    indications/pressure warning audible/on whichever displays are still working,
    Fire burns through oxygen line & may have damaged other wires applicable to mask intercom,
    Hamid is rendered unconcious,
    Anyone still concious has no way to open pilots door,
    Eventually all are unconcious.
    Loss of pressure extinguishes fire,
    Aircraft flies on.

    _____________________________________

    Perhaps it would be interesting to consider how Hamid could have lapsed into unconciousness;
    The oxygen mask feed quickly reduced or stopped- say 5 seconds for Hamid to attempt
    to rectify/manipulate the mask controls & realize the feed had indeed failed.
    Remove mask- say 3 seconds.
    Engage the autopilot control- say 1 second.
    Climb out of the copilots seat- (no slide back, so clamber out of seat being careful not
    to bump control column/not to cancel autopilot)- say 3 seconds.
    Retrieve copilot portable oxygen cylinder & place on face & open oxygen valve – say 6 seconds…
    but wait, that’s taken him already to and past 15 seconds, for an average fitness male,
    who is a smoker, so did he get to complete this last action? Apparently not.

    ___________________________
    @David
    I’ll check that reference when I wake up. Cheers

  521. Ge Rijn says:

    @Don Thompson

    On your remark:
    “Are you sure that it’s not the case that you simply want the damage to the stbd aileron N106US to be explained as directly caused by the landing on water?”

    No. I just want to shed more light on this IMO more logical and more obvious possibility concerning all the found debris and their damage.
    While you and some others seem to keep refusing to take a serious look at it to the point of making this possibility ridiculous, this tells me you and some others are more fixed in their believe the damage must have been caused mainly during a high speed dive and high speed impact than I am in a ditching-like crash.

    You know it’s hard to argue with people who have made up their mind already. They become ‘blind’ to consider other possibilities.

    I can go on giving other arguments which support my view but I think I will keep talking to people who are not able anymore to listen because they’ve made up their minds on this already.

    Time (and hopefully the official investigators) will tell in detail what happened that caused the damage.

  522. Don Thompson says:

    Ge Rijn

    In the interest of ensuring a wide scope to view the damage to the aileron on N106US, the BAAA site hosts this image which shows that the right wing tip scooped a sizeable chunk of riverbed. The NTSB published this image which shows impact – compression – damage to the upper, leading edge of the aileron. That image also shows that the lower part of the wingtip fence is missing.

    David,

    Two consequences of possible overspeed must be considered: a) aeroelastic flutter within the fixed structures; and b) motion of undamped PFCS surfaces such as the flaperon and the partially damped ailerons. At what point might these moveable surfaces fail to fair, benignly, in the wing’s slipstream? With no hydraulic power, should the RAT be lost, the aileron PCUs do allow downward deflection of the ailerons as the hydraulic actuator is permitted to retract.

    Brock,

    A) USN and US Embassy KL posted updates for Kidd & Pinckney’s contribution to the search. I understand your desire to be thoroughly independent, other sources may be found.

    B) No.

  523. Gysbreght says:

    @buyerninety: Between the turnback at IGARI and Pulau Perak the aircraft was being tracked solely by air defence radar systems. We have three sources for those primary radar data:
    – The flight path derived by the US NTSB from the unfiltered primary radar data
    – The filtered primary radar data presented in the DSTG book on Bayesian methods
    – The Lido Hotel picture.

    All three sources show that the autopilot was not engaged between IGARI and Pulau Perak. Perhaps it was not available.

  524. Ge Rijn says:

    @Don Thompson

    Thanks for the good images. IMO the damage to the aileron is too typical to being caused by mud or another cause than water forces during the ditch.
    But as I said; who will tell in the end..
    The other picture shows clearly the first section of the outboard flap also broke away like the MH370 outboard flap section.
    And it also shows clearly the first outboard flap track fairing misses a big shunk on it’s inboard side.
    In your posted video I noticed (new to me) also the left wing first section of the outboard flap is missing.

    But I guess this is all much better explained otherwise than just caused by the forces off the water during the ditch.
    Why make it simple if you can make it complicated?

  525. Gysbreght says:

    The NTSB describes analysis of primary radar returns in the Factual Report of Perfomance Group Chairman on the Hudson ditching.

  526. Gysbreght says:

    @David: Thanks. So “kinks” indicate failure under compression load.

  527. David says:

    @Gysbreght. Yes. I find it odd that there are no reports of such. Inertia will generate pushing I would have thought.

  528. Don Thompson says:

    Gysbreght/Buyerninety:

    Between the turnback at IGARI and Pulau Perak the aircraft was being tracked solely by air defence radar systems.

    Not entirely correct: the FI depicts tracks derived using data from the air traffic service, primary radar surveillance heads, at Kota Bharu and Butterworth AB. While both of these heads are Selex ATCR-33 systems, each is attributed with different detection range while the antenna set angle is unknown. The difference in these factors is consistent with a shorter track being detected while 9M-MRO flew within range of Butterworth than when detected within range of Kota Bharu.

    :Don

  529. Gysbreght says:

    @Don Thompson: My wording is copy-pasted from the ATSB ‘Fact Sheet’ that first published the NTSB track. The FI depicts only tracks derived using data from the air traffic service, plus a verbal description of what military radar saw.

  530. Don Thompson says:

    Gysbreght,

    Thank you for providing the links to NTSB’s documents.

    Ge Rijn,

    N106US touched down with wings level. The damage to the right aileron is inconsistent, to an extreme extent, to damage incurred by any other flight control surface on the wing or horizontal stabiliser. The aircraft drifted and was recovered to a location near Battery Park, 6km from the impact point, before it was finally lifted out of the water. I’ll park this digression now.

    :Don

  531. Don Thompson says:

    @Gysbreght,

    So the originator of the excerpt you provided might have been more accurate to state, ‘[the data provided to describe the aircraft track] Between the turnback at IGARI and Pulau Perak was [recorded] solely by air defence radar systems’.

    :Don

  532. Gysbreght says:

    @Don Thomson: Thats what he stated, isn’t it?

  533. Oleksandr says:

    @buyerninety,

    “I’m a bit surprized you haven’t stated your theory along the lines…”

    I have nearly 50-pages note, however there still a few loose ends, particularly with regard to the right DC bus. In addition, a recent Andrew’s comment made me to reconsider flight path model again. Hint: the emergency oxygen supply line to the cockpit is also routed along the left side of the fuselage. Another hint: intercom electronics is also at the left side. I lean to think that the Captain went to the EE-Bay just past 17:30 as the most senior person onboard, while F.O. was instructed to aviate manually – something he was supposed to be trained to do. They had no choice.

  534. Oleksandr says:

    @Don Thompson,

    Thanks for your comments with regard 18:25:27 BFO and Holland’s paper.

  535. buyerninety says:

    @Oleksandr said “Hint: the emergency oxygen supply line to the cockpit is also routed along the
    left side of the fuselage.

    Yes, down between the left (‘east’) walkway & the wheel well wall.
    Another hint: intercom electronics is also at the left side.
    Hadn’t seen this anywhere. Well found.

    I have a number of questions myself I have to formulate & ask Andrew in the coming days.

    @Ge Rijn
    You seem to have some exasperation at David & my answers, so I hope you can accept we are only
    interpreting the evidence as we understand it. Jeff made a comment in one of his recent posts
    about how the ‘IG back in 2014 interpreted the 00:19 BFO value as meaning that MH370 was in a
    steep descent of around 15,000 fpm, and that they believed it impacted the water within
    seconds after the last signaling unit log record’. Therefore it doesn’t seem likely MH370 had
    time to flatten out into a glide.
    Recent evidence that many of the exterior parts recovered had fibres that had failed ‘in tension’
    tends to reinforce the view that parts of the aircraft snapped off as a result of aerodynamic
    forces encountered during a steep dive.
    Cheers

  536. Gysbreght says:

    @buyerninety: “the ‘IG back in 2014 interpreted the 00:19 BFO value as meaning that MH370 was in a steep descent of around 15,000 fpm, and that they believed it impacted the water within seconds after the last signaling unit log record’. Therefore it doesn’t seem likely MH370 had time to flatten out into a glide.”

    Therefore ???

  537. Ge Rijn says:

    @David

    The left wing outer front flap (leading edge) of the Hudson-plane did not seperate during the ditch. This possibly detached while the boats where pushing on the planes wings on the front. It’s gone before the plane sank. No boats pushed the plane on the back side of the wings. It’s all visible in this video:

    https://www.youtube.com/watch?v=5SL1A2d2e7M

    It could be they made several attempts to lift the plane on different locations and damage all those flap gear causing a lot of collateral damage in the wake. I doubt they were so thoughtless to work this way on a plane that still had to be investigated.

    And thanks for the ‘composite damage-report’. ‘Kinks’ occure with compression and bending (combined tension and compression forces) it tells. The majority of tension damage and no kinks suprises me a bit.
    But I think it’s worth to mention two of the most important parts are not included in the MY-report: the flaperon and the outboard flap section.
    And we still have no information on the structural damage in this regard from those items.
    The trailing edge damage of those both pieces though show striking similarity with the Hudson-aileron and item 26 IMO.

    But I agree with @Don Thompson we probably better park this discussion till more conclusive information becomes available.

  538. Ge Rijn says:

    @buyerninety

    Oke one more thing then. I know how you and most others are ‘interpreting the evidence as you understand it’. I respect that but I wanted to make clear this ‘evidence’ has become fixed in some minds it seems to me beyond keeping an open view to other possibilities.

    The 0:19 BFO’s are not ‘holy’ in my view as with it the 15.000fpm descent rate and assumed high speed impact just because it’s an ATSB conclusion based on their data.
    If the debris tells another story (which it does IMO) we have to seriously doubt the 0:19 BFO’s or at least the conclusions drawn from it.

    It’s like with the CSIRO/ATSB-35S conclusion questioned in this @VictorI’s topic.
    We don’t take that for granted also only because CRIRO/ATSB proclame it.

    By the way; tension failures can be caused water the same way as with air only speeds necessary are x-factor lower.

  539. Ge Rijn says:

    @buyerninety

    Something else on the 15.000fpm descent rate estimated with the final BFO’s I have a question.
    If MH370 descended with that rate from ~30.000ft altitude within a minute it would have left another 15.000ft to level out won’t it?

    The ChinaAir flight 006 Boeing 747 managed to do this recovering from a near mach 1 vertical dive and landed safely with no flutter damage on the wing control surfaces (or seperation otherwise) but only serious overload damage on its horizontal stabiliser.

  540. Lauren H. says:

    On 09/26/2016, I posted the following on JW’s blog (Victor also posted something similar around then):

    “It seems many here have forgotten one basic assumption: If the flight after the FMT were piloted until the end, the search area would be too large for a reasonable search. (I believe the piloted search area would have been ten times the current search area and would take 20 years to complete.)”

    Is this statement no longer valid?

  541. Ge Rijn says:

    @Lauren H

    I think it’s probably still valid partly. But we now have more precise drift data and debris reports that limit the crash area better than the Inmarsat data could have done before all this debris was found.
    And I expect more debris to be found and more data on the circumstances of the crash will come foreward some day.

  542. TBill says:

    @Gysbreght
    “therefore” the aircraft must be on Arc7 in the SIO probably north of 38S, but that’s me talking.

    @Lauren H @Gen Rijn
    We have an enormous scientific contribution from Inmarsat defining Arc7, which at least we can say one interpretation is the debris is confirming, and yes, this gives us reasonable future search area. That’s our only hope of finding the plane unless more info is forthcoming. A corollary to that is, if we have searched Arc7 adequately (not yet) then the NoK can at least know the known area was adequatley searched.

  543. Brock McEwen says:

    @Don: thanks – appreciate your patience.

    I of course read those sources carefully before posting here. Unfortunately, the USN has been curiously silent about where Pinckney and Kidd were immediately prior to joining the search. Hence my questions to the IG, some of whom seemed to have contacts which went deeper than Google search engines can reach.

  544. Paul Onions says:

    @Oleksandr said “Anyhow, a combination of a single physical impact at the left side of the EE-Bay followed by isolation of the main AC buses by the crew (to prevent potential fire spread) explains almost all the observed features of MH370 flight”.

    Another scenario is: Oxygen Bottle Rupture similar to Qantas QF30 but with a gradual decompression.

    The oxygen bottle was topped up immediately prior to flight.
    It is mounted on the left side of the nose gear wheel well on the left side of the EE Bay.
    It is located adjacent to the Left AIMS Cabinet (aka the Left electronic brain!).
    Bottle ruptures during cruise like QF30.
    Destroys Left AIMS Cabinet and creates a hole in the nose gear wheel well.
    The Left AIMS Cabinet interacts with many, many LRUs – it supplies:
    (Mode S) Air data for the Left Transponder
    IRS data for the Left High Gain Antenna
    Flight ID for SDU (note Right AIMS does not supply Flight ID)

    Crew are overwhelmed with failures.
    They turn on APU
    The FO turns on his phone since no Comms
    They program a diversion to Mimos then Kendi.
    Then a diversion to Banda Aceh via Nilam-Sanob.
    The aircraft remains at high altitude.
    The gradual decompression goes unnoticed.
    Suffering from hypoxia they pass out.
    The aircraft continues on autopilot past Penang.
    At 1825:27 the aircraft rolls left to start the left turn at Nilam.
    The serviceable right HGA is exposed and logs on (note the aircraft is still heading north west at that moment)
    The BFOs from 1825:34 to 1828:15 should be discounted (according to Inmarsat).
    The aircraft over flies Banda Aceh and ends in SIO
    The electrical power transfer of Left Main AC bus from Left engine failure to APU causes the Satcom to initiate a restart sequence.
    Right engine fails.
    Autopilot remains engaged.
    Aircraft continues south.
    Satcom logs on.
    BFO at 0019:37 should be discounted (according to Inmarsat).
    IFE load shed.
    APU fails 13 min 45 sec later.
    Autopilot disengages.
    Aircraft out of control beyond 40 nautical miles past Bayesian hotspot.
    Never searched.
    RIP MH370.

  545. DennisW says:

    @Paul Onions

    And Shah, being the clairvoyant that he was, anticipated this complex failure mode, and visualized the aircraft path on his simulator.

  546. Donald says:

    @Dennis

    >And Shah, being the clairvoyant that he was, anticipated this complex failure mode, and visualized the aircraft path on his simulator.

    That does indeed seem to be the case. Or just another ‘coincidence’. The sheer number of ‘coincidences’ lending themselves to MH370 being a planned and deliberate act are astounding. The mental gymnastics and absurdities used to explain away these ‘coincidences’ are equally astounding (impressive, not so much).

    I find myself deeply questioning the sincerity, character and ‘impartial’ objectivity of anyone not firmly in the intentional camp. It pains me greatly as 239 human beings have perished and quite a few ‘experts’ seem interested in doing anything and everything possible to deflect the apportion blame on Zaharie.

    These ‘accident’ scenarios are just beyond pathetic and the intent that drives them extremely suspect, imho.

  547. DennisW says:

    @Donald

    Yes. I put any poster in the mechanical failure camp in the “whacko” category. Onions is simply the latest.

  548. TBill says:

    @Paul Onions
    Re: O2 cylinder
    I believe a key difference is Qantas had a metal O2 cylinder.
    The MH370 O2 cylinder was composite (eg; KEVLAR).

  549. Andrew says:

    Why are some people still attracted to the idea of a diversion to Banda Aceh? According to Indoavis, an organisation that provides navigation chart services for Indonesian airports, the airport is closed after 1100 UTC, with no ATC or navigation aids available after that time. Why on earth would the crew of MH370 contemplate going there, when there were far better options at Penang or Kuala Lumpur?

  550. David says:

    @Ge Rijn. First some comment on yours:
    On damage to the right wing rear, the film shows a melee of boats at around 37:38, one behind or over the right wing. I cannot see how the right aileron suffered the damage it did (Don Thompson’s photo) during the ditching, the outer part of the immediately adjacent deployed flap being intact as can be seen.

    “It could be they made several attempts to lift the plane on different locations and damage all those flap gear causing a lot of collateral damage in the wake. I doubt they were so thoughtless to work this way on a plane that still had to be investigated.”

    Think of how they even got the plane alongside (supposing it did not drift there) and how they got those slings under the wings (drill a tunnel through the mud?): amongst the salvage’s challenges that might not have been the foremost. Also, there was underwater camera work, possibly pre-any-damage for that reason.

    Kinks. Surely yes there would have been carbon fibre parts overstressed in direct compression or on the inside of a bend. Of particular interest would be the microscopic appearance of the flaperon upper surface.

    Thank you for your forthright views as to others’ closed minds. I for one do not reject the possibility of a piloted end. However that theory’s effect on search area makes it unworkable I think, for a piloted end brings permutations about the 7th arc log-on source, the dive from altitude (or not, if the final BFOs are discounted), glide and ditching. There is no indication there was a pilot as yet and with no indication there was it remains reasonable I think to suppose there was not, though the representations about the probabilities of search success could be, accordingly, more modest.

    If you do set aside findings about the last BFOs’ descent rates as you are inclined to then you are querying the outcome of Boeing simulations as I see it. If you extend that approach, there could have been even a gradual descent to a flaps down powered ditching, which could have led to an APU autostart and a 7th arc log-on on the water. For this though you would then need to overlook not just BFOs but also the flaperon/flap collision-while-housed evidence. Certainly none of this ‘evidence’ is infallible and that’s how I could agree that any of the above would be possible. The question then is what evidence is it reasonable to challenge? For my part generally I defer to SSWG endorsed opinions at least, given that group’s composition (ie detachment and expertise) and access to source information

    I do not share your confidence in the part drift analysis should play currently even though the weighting afforded its findings has increased somewhat; and I note your answer to Lauren H above.

    Of the two parts of the No.7 fairing which were recovered a year apart, the first found (Mozambique) had no barnacles, so its discovery might even have been delayed, it then being beached even sooner and the gap betweee tne two greater. The second, item 27, (South Africa) had small (maybe juvenile) barnacles so either they had grown recently or were a different species. A third fairing part, item 8 was found five months after the first but at Mauritius; still on the way, no barnacles.

    From just these three examples I see that drift analysis has to process and predict from a small data base of considerable scatter, leaving me (at least) uneasy and looking for a lot more work yet.

    Beyond what you have raised, I could go on about whether the ATSB should answer all the questions we the public would like them to; but no. However, take a look at the extent of the work in looking into the Hudson ditching, which Gysbreght alluded to, and scratch your head as to the extent the public should be included during deliberations: https://dms.ntsb.gov/pubdms/search/hitlist.cfm?docketID=47230&StartRow=121&EndRow=124&CurrentPage=9&order=1&sort=0&TXTSEARCHT=how.

    Yes, to parking it.

  551. buyerninety says:

    @Andrew
    (I believe those people are trying to rationalize the reason for the Final Major Turn,
    and to them WITT is the nearest large airport.)

    Thankyou for sharing your knowledge, we’re all trying to understand flight procedures as
    practised in 777’s, so hopefully we can ask intelligent questions in the following weeks.

    For the moment though, here are several ‘rough and ready’ questions.

    In the 777 FMC ARPT page, how many airports are usually present? I’ve read that if an airport
    is in the airports page, the navigation aids for that airport are also present.

    When making changes on the LEGS page(s), I’ve read that when using the ERASE function, that
    “Note: If you make multiple changes all of them will get erased. This function reverts
    everything back to the last executed state.”
    Do you see any way that part of an active route, say (WMKK-VABB) AGOSA to IGOGU, could remain
    whilst the remainder of that route, or the remainers of the LEGS of that route, were erased or
    deleted (all after IGOGU)?
    If this were possible, and all the remaining waypoints were deleted, wouldn’t the destination
    airport still remain however (VABB)? Would the FMC then proceed to use the remaining information,
    which may be the heading (occuring after the last deleted waypoint) to VABB?

    If you have LOC autotune set, but the VOR/DME that is received has no VOR, but a DME is being
    received, does it attempt to fly the DME arc? Does it just ‘touch’ the (DME) arc and fly
    onwards?

    If you have LOC autotune set, and the radio tunes to a TACAN, What happens?
    (Does it attempt to fly (touch) the (TACAN equivalent) DME arc, despite the absence of VOR?)

    Do you know if, in the Russian miliary equivalent of TACAN, the DME portion of the signal is
    captured, or is it non-compatible? (unlike the ‘DME portion’ of Western TACAN)

  552. buyerninety says:

    EDIT;
    In the 777 FMC ARPT page, how many airports are usually present? I’ve read that if an airport
    is in the airports page, the navigation aids for that airport are also present.
    “The ALTNpage displays a list of up to four alternate airports. The source
    of alternate airports can be:…
    …3. Automatic selection from the nav database ”

    So how many potential airports could the ‘nav database’ hold?

  553. paul smithson says:

    @Andrew. That is a very good question and one that continues to puzzle me. Hijack? Unwilling to risk penang without transponder, TCAS, comms? Still running checklists, problems not yet stabilised and want to keep clear of populated areas and active airways?

  554. Andrew says:

    @buyerninety

    “I believe those people are trying to rationalize the reason for the Final Major Turn,
    and to them WITT is the nearest large airport.”

    Thanks – I understand that part, but I think it’s already been pointed out that WITT would not have been a suitable alternate late at night, even if the situation was dire. If the aircraft turned back after suffering some kind of problem after IGARI, then WMKP or WMKK would have been far better options. I can’t think of any sensible reason for going to WITT!

    “In the 777 FMC ARPT page, how many airports are usually present? I’ve read that if an airport
    is in the airports page, the navigation aids for that airport are also present.”

    I think you mean the ALTN page, in which case it shows four alternate airports. The airports must be in the navigation database. Any associated navaids should also be in the database.

    “Do you see any way that part of an active route, say (WMKK-VABB) AGOSA to IGOGU, could remain
    whilst the remainder of that route, or the remainers of the LEGS of that route, were erased or
    deleted (all after IGOGU)?”

    Yes, it’s certainly possible to insert a route and then delete everything after a certain waypoint.

    “If this were possible, and all the remaining waypoints were deleted, wouldn’t the destination
    airport still remain however (VABB)? Would the FMC then proceed to use the remaining information,
    which may be the heading (occuring after the last deleted waypoint) to VABB?”

    Yes, the destination would remain. There would be a discontinuity between IGOGU and VABB. After passing IGOGU, the aircraft would maintain the heading it was flying as it passed overhead IGOGU.

    “If you have LOC autotune set, but the VOR/DME that is received has no VOR, but a DME is being
    received, does it attempt to fly the DME arc? Does it just ‘touch’ the (DME) arc and fly
    onwards?”

    I’m not sure I understand what you mean, but the FMC doesn’t actually use the navaids to steer a course. It uses a combination of GPS, IRS and radio positions (depending on what’s available) to derive an FMS position. The required course is programmed into the FMC via the LEGS page. If a DME arc is programmed as part of an approach or departure, then the aircraft will use the FMS position to navigate its way around the arc, regardless of any navaids that might or might not be available.

    “If you have LOC autotune set, and the radio tunes to a TACAN, What happens?
    (Does it attempt to fly (touch) the (TACAN equivalent) DME arc, despite the absence of VOR?)”

    As above.

    “Do you know if, in the Russian miliary equivalent of TACAN, the DME portion of the signal is
    captured, or is it non-compatible? (unlike the ‘DME portion’ of Western TACAN).”

    As far as I know there is no Russian military equivalent of the TACAN systems used in the West. Russia does have VOR/DMEs, which are the same as those used in Western countries.

    “So how many potential airports could the ‘nav database’ hold?”

    I don’t think there’s a fixed number, but suffice to say it’s a lot! The databases in our aircraft have a huge number of waypoints, navaids, approach & departure procedures for airports all over the world.

  555. Victor Iannello says:

    Andrew asked, “Why are some people still attracted to the idea of a diversion to Banda Aceh?”

    I’d expand that and say it is unlikely that there was an emergency diversion to ANY airport (e.g., Kota Bharu or Penang) as there is no evidence of a descent while MH370 was in range of radar. The groundspeeds are simply too fast.

  556. Andrew says:

    @Paul Smithson

    “That is a very good question and one that continues to puzzle me. Hijack? Unwilling to risk penang without transponder, TCAS, comms? Still running checklists, problems not yet stabilised and want to keep clear of populated areas and active airways?”

    Yes, but with no navaids and no lighting, WITT would have been unusable at night. Period.

    If I had some kind of dire problem, I’d be looking to stay as close to home as possible rather than diverting to some far flung airfield with unknown conditions, especially one that was unlikely to be usable. This all happened very late at night, at a time that Penang and KL would have been relatively quiet, so I very much doubt that traffic would have been an issue. If I had checklists to run, etc, I’d hold somewhere at a non-standard level to avoid any potential conflicts. In a serious emergency, I’d be looking to keep things as simple as possible to get the aircraft on the ground ASAP. In this case, the most likely choice for a diversion airfield would have been KL or Penang, either of which would have been very familiar to both pilots.

  557. Andrew says:

    @Victor

    Don’t get me wrong; I’m not saying there was a diversion. However, IF the crew was looking for somewhere to divert, then I think that WITT would have been extremely unlikely.

  558. buyerninety says:

    @Andrew said; -“If a DME arc is programmed as part of an approach or departure, then the aircraft will
    use the FMS position to navigate its way around the arc, regardless of any navaids that might or might
    not be available.”

    I understand that the 777 FMS can’t fly a VOR, a human pilot is required to ‘fly the needles’. But is
    this also true for the DME? Can’t the autopilot fly to a DME arc? and when it reaches the arc,
    do I understand you to mean it will actually fly along the arc? (surely not, or it would do
    continual 360° circles around the navaid!) If it follows the DME arc, what determines when it ceases
    to follow the arc (-is it when the FMS determines {IRS, GPS-wise} that it is passing over the GS)?
    (Note- does it need to identify the VOR to know where the GS is located, or could it use only the
    IRS, GPS derived position to know this? {Please consider ‘knowing’ as meaning ‘being aware of when it
    is crossing the GS’, not actually trying to fly down the GS…}

    “…many potential airports could the ‘nav database’ hold?”
    -I don’t think there’s a fixed number, but…approach & departure procedures for airports all over the world.
    Interesting, so potentially, the airports with their navaids, for say, WITT and YPCC, would be held
    as data in the FMC. A further question, though – 777-300ER might have AIMS2. MH370, a 777-200ER, was
    built in 2002, and seems to have had an AIMS(1) (unknown if it had a memory expansion added later.)
    Do you have any comments as to airports capacity of AIMS(1)?

  559. Victor Iannello says:

    @Andrew: No, I didn’t think you were proposing that there was a diversion to an airport.

    If you are willing to propose a scenario, I think many of us would be interested to hear your thoughts.

  560. TBill says:

    fyi Richard has a nice paper on possible diversion airports

  561. Victor Iannello says:

    @Andrew: I’d add that an emergency descent can be performed with the minimum of instruments and without using the A/P or A/T. Simply put the throttle levers in the idle position and trim to a reasonable airspeed, like 310 kn. If a faster descent rate is required, do the same but use speed brakes. I assume any B777 pilot would know exactly how to perform an emergency descent.

  562. Oleksandr says:

    @Victor,

    “The groundspeeds are simply too fast.”

    This is a common mistake. Radar data suggest that the plane was not at a constant level. The cellphone contact was unlikely if the plane was at high altitude. The high average speed does not justify impossibility to descent to around 7 km altitude. Even DSTG analysis states: “Collectively these data points suggest that the aircraft may have slowed down at some point between 18:02 and 18:22.” (p.20).

  563. Oleksandr says:

    @Andrew,

    You need to keep in mind that if L AC bus was isolated (say as a preventive measure taken by the crew), jettison pumps would not work. Immediate landing at Penang with 30+ tonnes of kerosene would be too risky, especially if ground emergency services were not alerted.

  564. Oleksandr says:

    @Paul Onions,

    “Another scenario is: Oxygen Bottle Rupture similar to Qantas QF30 but with a gradual decompression.”

    If you read my previous comments, I have included this into my list of possible causes. However, I think it is impossible to distinguish between all the 4 causes I listed until the main debris are found and analyzed. My primary speculation of the cause: rupture of the LH nose landing gear wheel rim. “The missing bolt” theory.

  565. Oleksandr says:

    @Dennis,

    “I put any poster in the mechanical failure camp in the “whacko” category.”

    I did exactly the same with the simulator “data”.

  566. Victor Iannello says:

    @Oleksandr:

    “Radar data suggest that the plane was not a constant level”.

    I don’t believe any level data that was measured by the military radar, which also recorded a flight level of FL447 as it flew past Penang. I also don’t believe the speed data that was presented in the DSTG Bayesian analysis, as there was disagreement even with the known speed from the recorded ADS-B data.

    “Even DSTG analysis states: ‘Collectively these data points suggest that the aircraft may have slowed down at some point between 18:02 and 18:22.'”

    The DSTG ignored the 18:22 radar point because the radar data was deemed inaccurate at that range. For this reason, all their calculations started at the 18:02 radar point. The 18:25 BTO values do indicate that the plane turned away from the subsatellite radius, which most of us explain by a lateral offset manoeuver. You probably missed most of this discussion between @DrB, @DennisW, maybe others, and me, and I won’t repeat it here.

    “The cellphone contact was unlikely if the plane was at high altitude.”

    If you read the RMP report, you will see that cell phone connects were measured at the maximum tested flight level of FL240. My guess is that sporadic connects are possible at even higher altitudes.

  567. Oleksandr says:

    @Victor,

    Re: “I don’t believe any level data that was measured by the military radar, which also recorded a flight level of FL447 as it flew past Penang.”

    Then why do you believe the plane passed via VAMPI and MEKAR?

    Re: “You probably missed most of this discussion between @DrB, @DennisW, maybe others, and me, and I won’t repeat it here.”

    No, I did not miss. I just disagree with you conclusions.

    Re: “The 18:25 BTO values do indicate that the plane turned away from the subsatellite radius”

    This could be a result of combined turn and descent (descent also causes increase of the distance). There are many emergency maneuvers that fit BTOs and BFOs, not only the lateral offset. Earlier I have shown that a spiral descent fits BTO and BFO perfectly, but this is not the emergency maneuver I consider in the present.

    Re: “If you read the RMP report, you will see that cell phone connects were measured at the maximum tested flight level of FL240”.

    Yes, but the connection at FL350 is very unlikely. Also you probably missed Dennis’ comments in this regard. In general, I would first consider what is more likely rather than “handing” on what is less likely.

  568. Don Thompson says:

    Victor wrote: “My guess is that sporadic connects are possible at even higher altitudes.”

    A typical cell base station antenna is an electronically ‘focussed’ array. While the antenna constrains the radiated power within a vertical dispersion of +/-20º, the antenna designs naturally generate additional lobes radiating up to high elevations. The range from the antenna site to the aircraft was approximately half the expected maximum range of the antenna site. We don’t know the specific characteristics of the antenna in Bandar Bharu/Air Itam so cannot discount that this connection, and registration with the network’s Location Based Service, occurred. It is entirely feasible.

    The RMP used a Beech King Air B350 to support the tests carried out by MCMC. It is another one of those anomalies, why the tests were not executed at the representative altitude.

  569. Victor Iannello says:

    @Oleksandr:

    “Then why do you believe the plane passed via VAMPI and MEKAR?”

    You obviously don’t understand what I believe. I believe the 18:02 radar capture and the BTO/BFO data. Any radar data after 18:02 is suspicious. The DSTG says they were given only one capture after 18:02, and the accuracy is questionable. As per the DSTG, my paths do not use any radar data after 18:02.

    “Also you probably missed Dennis’ comments in this regard. In general, I would first consider what is more likely rather than “handing” on what is less likely.”

    I missed nothing. Dennis and I discussed this privately, too. By the way, you are quoting Dennis, but didn’t you just call Dennis “whacko”? (I don’t think Dennis is whacko, although we might disagree on a few points.)

    I don’t think the plane flew by Penang at 520 kn and low altitude. It’s that simple.

  570. Victor Iannello says:

    @Oleksandr said, “Immediate landing at Penang with 30+ tonnes of kerosene would be too risky, especially if ground emergency services were not alerted.”

    In an emergency, a pilot would land heavy rather than fly longer to burn off fuel. See this Boeing publication.

  571. Oleksandr says:

    @Victor,

    Re: “You obviously don’t understand what I believe. I believe the 18:02 radar capture and the BTO/BFO data. Any radar data after 18:02 is suspicious. ”

    No, I don’t. If you state that any radar data after 18:02 is suspicious, they why do you insist on the path via VAMPI and MEKAR in LNAV mode? I don’t understand why would someone claim that the radar data is unreliable with regard to the altitudes, but the same time accept the horizontal accuracy of 0.5 km to state that the route was via VAMPI and MEKAR?

    Re: “In an emergency, a pilot would land heavy rather than fly longer to burn off fuel.”

    Your reference is inappropriate. We have discussed this many times. The issue is not overweight. I am tired to repeat this again and again. Let me try one more time. Imagine heavily damaged nose landing gear and 30 tonnes of kerosene; no ground emergency services ready or even aware. What can happen? I don’t understand why it is so difficult to grasp this. Hint: if you can’t imagine, check what has happened with Emirates Boeing-777 in Dubai last summer.

  572. Gysbreght says:

    With 573 responses reloading the page is becoming exceedingly slow again.

  573. Andrew says:

    @buyerninety

    “I understand that the 777 FMS can’t fly a VOR, a human pilot is required to ‘fly the needles’. But is
    this also true for the DME? Can’t the autopilot fly to a DME arc? and when it reaches the arc,
    do I understand you to mean it will actually fly along the arc? (surely not, or it would do
    continual 360° circles around the navaid!) If it follows the DME arc, what determines when it ceases
    to follow the arc (-is it when the FMS determines {IRS, GPS-wise} that it is passing over the GS)?
    (Note- does it need to identify the VOR to know where the GS is located, or could it use only the
    IRS, GPS derived position to know this? {Please consider ‘knowing’ as meaning ‘being aware of when it
    is crossing the GS’, not actually trying to fly down the GS…}”

    I’m not sure we’re talking about the same thing. When pilots talk about “flying a DME arc” they mean just that; the aircraft flies an arc at a constant distance from a DME station. Normally it would only be part of the circle between two radials, not the full 360°! For example, have a look at the chart for the
    ILS04 at Penang. The aircraft might fly inbound on the 155 radial from VPG to intercept the 15DME arc. It would then follow the 15DME arc to intercept the localiser for RWY04.

    If the pilots want to fly that approach, they would program it into the FMC. The autopilot would initially fly the approach in LNAV mode and follow the ground track of the 15DME arc, but the FMC is providing commands based on the FMC position. That position is essentially the GPS position because it is the most accurate position source available. Strictly speaking, the autopilot is not using the DME; the only radio navaids the autopilot can track are the ILS localiser and glideslope. To fly the final part of the approach, the pilots would arm the autopilot localiser & glideslope modes as the aircraft approaches the localiser, and the autopilot would then intercept the localiser & glideslope beams.

    “Interesting, so potentially, the airports with their navaids, for say, WITT and YPCC, would be held
    as data in the FMC. A further question, though – 777-300ER might have AIMS2. MH370, a 777-200ER, was
    built in 2002, and seems to have had an AIMS(1) (unknown if it had a memory expansion added later.)
    Do you have any comments as to airports capacity of AIMS(1)?”

    Potentially yes, although I’m not sure about YPCC. The databases on older versions of the aircraft are smaller, but they still contain the data for a significant number of airports all around the world.

  574. Andrew says:

    @Victor

    “If you are willing to propose a scenario, I think many of us would be interested to hear your thoughts.”

    It’s early morning here and I’m about to head off to work for the day. I’ll try to write something later today.

    Cheers,

  575. Victor Iannello says:

    Oleksandr said, “If you state that any radar data after 18:02 is suspicious, they why do you insist on the path via VAMPI and MEKAR in LNAV mode? I don’t understand why would someone claim that the radar data is unreliable with regard to the altitudes, but the same time accept the horizontal accuracy of 0.5 km to state that the route was via VAMPI and MEKAR?”

    First, the horizontal accuracy of the radar can be much better than the vertical accuracy. The horizontal accuracy is essentially determined by the azimuth and the timing. The altitude determination requires an elevation beam pattern and possible scanning.

    Secondly, after the DSTG report, I have questioned all radar data after 18:02, so much that I posted an article that discussed possible paths in the Malacca Strait. Not once in that post did I use VAMPI and MEKAR to generate the paths. Rather, starting from the 18:02 radar point, I found constant speed, great circle paths that satisfied the BTO and BFO data at the log-on without manoeuvers or descents.

    “Imagine heavily damaged nose landing gear and 30 tonnes of kerosene; no ground emergency services ready or even aware. What can happen?”

    Even if we accept that the pilot could not jettison fuel and chose to remain flying to burn fuel (which I don’t accept), it makes zero sense in an emergency to fly away from a suitable airport like Penang. The pilot would enter a holding pattern to burn fuel and keep the airport close, not fly up the Malacca Strait.

  576. Victor Iannello says:

    @Gysbreght: The page loads in about 3 seconds on my computer, and about the same on my smart phone. As added comments have no graphics, they should load fast. What kind of computer and internet connection do you have?

  577. Oleksandr says:

    @Victor,

    “First, the horizontal accuracy of the radar can be much better than the vertical accuracy. The horizontal accuracy is essentially determined by the azimuth and the timing. The altitude determination requires an elevation beam pattern and possible scanning.”

    No doubt. Suppose the Butterworth radar provided some data. Do you believe it could not differentiate between, say, FL200 and FL350 at 60 nm distance, but provided horizontal accuracy of 0.5 km at the distance of 200 nm? Suppose it was some other remote radar, perhaps Thai. Same question.

    In your post you linked you wrote: “The radar data presented to the NOK at the Lido Hotel in Beijing on March 21, 2014, has been used by independent investigators to justify a path along airway N571”. This is exactly what was happening – justification. And this is wrong, regardless whether the Lido image is correct or not (I believe it is correct). The Lido image is insufficient to state anything with regard to N571 or waypoint. At least because of the accuracy of the radar data at long distances.

    Re: “The pilot would enter a holding pattern to burn fuel and keep the airport close, not fly up the Malacca Strait.”

    Exactly! That is what the Captain may have attempted to do in my opinion.

  578. Victor Iannello says:

    Oleksandr asked, “Do you believe it could not differentiate between, say, FL200 and FL350 at 60 nm distance, but provided horizontal accuracy of 0.5 km at the distance of 200 nm?”

    Please re-read what I wrote. I think the radar point at 18:22 is useless, and NOT accurate to 0.5 km. I don’t know how to say this any clearer. I also think the radar-derived altitude data is almost useless, which puts the plane at FL447 near Penang, not the low altitude that you assert is dictated by the cell phone connection.

    “Exactly! That [enter a holding pattern to burn fuel and keep the airport close] is what the Captain may have attempted to do in my opinion.”

    If the captain attempted to enter into a holding pattern near Penang, it makes no sense that the plane flew up the Malacca Strait. Penang is a large, inhabited island with visual landmarks at night that would be very familiar to the captain. Even with no navigational aids, he could certainly stay close to Penang if he chose.

  579. Mick Gilbert says:

    @Victor

    Regarding your contention that there was no diversion because there was no descent, the airplane’s track was broadly consistent with a diversion to Penang for ILS04. If they were diverting to Penang and were not faced with an obvious depressurisation problem then they would not have commanded a descent until about 10 minutes after initiating the diversion. A lot can happen in 10 minutes. Surely it is not beyond the realm of possibilities that the crew were diverting to Penang but were either incapacitated before they could command a descent or they commanded a descent which was not actioned because of a flight control problem (eg autothrottle disengaged) and they were subsequently incapacitated before they either noticed the failure to descend or could take corrective action.

    If you weigh the evidence on the balance of probabilities I don’t think that you can rule out a diversion simply because of the lack of an observed descent, particularly so when you consider that other evidence supports the possibility of a diversion.

  580. Mick Gilbert says:

    @Andrew

    Mmmm … it seems that I’ve seen that 10DME arc somewhere before.

  581. Victor Iannello says:

    @Mick Gilbert: If the pilot was not able to descend to Penang because he was incapacitated, how did the plane turn northwest up the Malacca Strait and then later south to the SIO?

  582. Brock McEwen says:

    @Victor: did you – or anyone in the IG, as far as you know – archive all NOTAMS in effect either on or near the night of March 7, 2014, for all airspace within, say, 300 nmi of a line connecting IGOGU to IGARI?

    If not: did you (or IG, AFAYK) archive ANY such NOTAMS?

    I’m particularly interested in live-fire exercises, given the greater likelihood of accompanying long range radar assets – but I’ll gratefully take whatever you’ve got.

    Thanks in advance.

  583. Mick Gilbert says:

    @Victor

    Re: “If the pilot was not able to descend to Penang because he was incapacitated, how did the plane turn northwest up the Malacca Strait and then later south to the SIO?”

    The turn from south of Penang up the Malacca Strait can be explained by the crew programming a modified STARS approach for ILS04 where the discontinuity between KENDI and CF04 had not been removed. With speed and altitude controlled by the MCP, the FMC would have commanded an early turn towards KENDI and then reverted to a CMH as it passed through KENDI and on up the Strait flies the airplane.

    After Andrew’s recent post on DME/VOR navigation, I was struck by the similarity between the 10DME arc and the 01:52:35 MYT PSR trace. So a possible alternative worth exploring is that the crew programmed a 10DME approach rather than a STARS. I’m wondering what the outcome might have been as the airplane passed through the 212 radial given the altitude of the airplane and absent a crew intervention to arm the autopilot localiser and glideslope modes; FMC reversion to CMH?

    And if we’re setting aside the radar data from the Malacca Strait, how well does a turn through KENDI onto a CMH of 300-ish° end up looking against the 1825-1828 UTC BTO/BFO? Back of the fag packet, it’s got to produce a track of around 297° (ie parallel to N571) to the north of N571 (ie something that looks very similar to an offset).

    That still leaves the FMT but I’m not Robinson Crusoe on that one.

  584. David says:

    @Ge Rijn. I said, “If you do set aside findings about the last BFOs’ descent rates as you are inclined to then you are querying the outcome of Boeing simulations as I see it.”

    Rethink. While the Boeing descents do qualify the final BFOs as possible they do not demonstrate they are valid. That would require a descent sample large enough to be representative and which showed that those of suitable extent and timing were not unusual.

    Expanding on that, separate the results from the more recent simulations into two groups, the first assuming loss of autopilot at fuel exhaustion, the second when autopilot loss was at first engine failure. In the first group 2 of 7 had such descent rates (ATSB). Of the second it was 3 of 3. However the ATSB also has said the chances of the second are low. Thus the overall probability of encountering final BFO type descents from these 10 is say 35-40%. Based on that and supposing the simulations to be realistic, there is less than a 50/50 chance they support descent rates which explain the final BFOs.

    The Definition of Underwater Search Areas of December, 2015 does not say how many of the earlier Boeing simulations satisfied final BFO descent rate criteria.
    Furthermore, aside from the small sample it is also unclear what the criteria are for adjudging descents, both rates and timing. What accuracy is required and do excessive descent rates disqualify?

    In summary, for these reasons alone the simulation information released is insufficient to draw the conclusion I did. That could only be done if earlier simulations did have a high proportion satisfying statistically suitable descent criteria. I know of no statement to that effect.

    I hope others will correct my interpretation if flawed. Otherwise there can be room for doubt whether the final BFOs have been substantiated by simulations at all.

  585. Andrew says:

    @Oleksandr

    “You need to keep in mind that if L AC bus was isolated (say as a preventive measure taken by the crew), jettison pumps would not work. Immediate landing at Penang with 30+ tonnes of kerosene would be too risky, especially if ground emergency services were not alerted.”

    “The issue is not overweight. I am tired to repeat this again and again. Let me try one more time. Imagine heavily damaged nose landing gear and 30 tonnes of kerosene; no ground emergency services ready or even aware.”

    Fair enough, but if the crew did decide to reduce the aircraft’s weight before landing and found they couldn’t jettison fuel, they would most likely hold as long as necessary to reduce the weight, as Victor said in another post. The two most logical places to do that would be Penang or KL. There is not a snowball’s chance in hell that I would go to Banda Aceh, which was the point of my post!

  586. Gysbreght says:

    @David: “In the first group 2 of 7 had such descent rates (ATSB). Of the second it was 3 of 3. However the ATSB also has said the chances of the second are low. “

    I believe the ATSB erroneously omitted track #9 when they said that the second group consisted of 7,8,10.

    Track #9 turns right and is almost identical to #10. The ATSB said earlier that the airplane always turned left in the simulations. Only in the unlikely electrical configuration of the second group they found that “the aircraft descended in both clockwise and anti-clockwise directions”. It is also highly unlikely that #9 would be identical to #10 if it belonged to the first group. Another stonewalled question I put to the ATSB.

  587. Gysbreght says:

    @David: That leaves #3 as the only track that achieved the descent rate of the final SATCOM transmission. But when asked for details of the scenario simulated: again stonewall.

  588. David says:

    @Gysbreght. Thanks. I follow.

  589. buyerninety says:

    @Mick where is this ‘CF04’ that’s clouded your thinking?

  590. Mick Gilbert says:

    @buyerninety

    CF04 is the first waypoint of the basic ILS04 approach – CF04-FF04-RW04.

    Let me explain it to you as it was explained to me;

    ‘CF’ stands for ‘course fix’ and ’04’ is obviously the runway designator. The approach charts show several different ways of getting onto finals – from KENDI, from overhead VPG VOR, from a 10 or 15 DME arc. When programming the approach the pilot selects ILS 04 and then has the option of selecting a ‘VIA’, where one of the above options can be selected. If the pilot does NOT select a ‘VIA’, then it is possible (but not definite – depends on the database coding) the FMC would simply show the STAR as far as KENDI, followed by a discontinuity and then the first waypoint of the ‘basic’ ILS approach, which would be the course fix. The course fix isn’t marked as such on the charts, but in this case it would probably be the fix at 7.8DME VPG, which is between KENDI and the runway threshold.

    There you go, now your thinking is as “clouded” as mine.

  591. Victor Iannello says:

    @Brock: I am not aware of any relevant NOTAMs. I don’t know what research others have done.

    @Mick Gilbert said, “So a possible alternative worth exploring is that the crew programmed a 10DME approach rather than a STARS.”

    There is no way to program an approach along a DME arc rather than choosing individual waypoints that are at various bearings and constant distance from a fix. It makes no sense for a pilot to do this, and even less sense in an emergency. It would be much easier to choose the ILS approach, and possibly choose a STAR and possibly choose a transition.

    “And if we’re setting aside the radar data from the Malacca Strait, how well does a turn through KENDI onto a CMH of 300-ish° end up looking against the 1825-1828 UTC BTO/BFO?”

    If the pilot had chosen the BIDM1A (BIDMO) STAR for the ILS04 runway, the waypoints would be BIDMO, PUKAR, ENDOR, MEKAT, KENDI, RTE DISC, CF04, FF04, RW04. I think the best representation of what was actually flown near Penang is the radar data presented in the RMP report, which has been digitized by @sk999. Using @sk999’s data, the plane passed near ENDOR, FF04, and OPOVI, but not MEKAT or KENDI. The radar data near Penang is not consistent with a route discontinuity at KENDI.

    “That still leaves the FMT but I’m not Robinson Crusoe on that one.”

    As for the unexplained turn towards the SIO, I don’t think those that are in the camp of an incapacitated pilot can simply ignore that it occurred.

    @buyerninety asked, “@Mick where is this ‘CF04’ that’s clouded your thinking?”

    The coordinates of CF04 are 5.153656,100.148056. Similarly, the coordinates of FF04 are 5.191003,100.181583. These waypoints align with RW04 for an ILS approach.

  592. Ge Rijn says:

    @David

    First to set something straight. I’m not inclined to set aside the final BFO’s or the descent rates. I’m not at all in a position to do that by any means.
    I just stated (and state again) that IMO we seriously have to question the final BFO’s or at least the conclusions the officials have drawn from them (high speed dive and impact), if the (kind of) debris, the amount of debris found so far and the kind of damage on them, is pointing to another scenario.
    Which it clearly does IMO, just my opinion based on my (and others) observations and arguments, nothing more.

    On the simulations you mention I think the first group is counting, for the left IDG had to be on line again after the 18:25 log-on to power the SDU. Assuming the SDU stayed powered till left engine fuel-exhaustion it would be only then when the AP disconnected I think. Not after the right engine stopped.
    Anyway a ~60% chance (6 out of 10) simulations that don’t meet the descent rates the BFO’s dictate doesn’t speak for the reliability of those BFO’s if I understand you well.

    Then again I think a descent rate of 15.000fpm calculated between those final BFO’s 8 seconds is not necessarally proof MH370 dived high speed in the water. It’s just 8 seconds. There could have been enough altitude left to level out in some way.

    Then finally, all simulations are based on no consious pilot at the controls after the AP disengaged.
    And this is IMO just an irrational assumption based on no evidence at all (as far as we know of).

    I appreciate you pick up on this BFO/descent rate/simulations subject.

  593. TBill says:

    Working a ISBIX path that looks good to me at the moment:

    A/P is set for waypoint ISBIX at M.84 arrive approx 20:00 and then pilot sets heading for 180S CTH (True/Heading). The winds aloft at first take MH370 west to about 93E but then the winds change direction and send MH370 east ending up at about 32.5S and 94.5E on Arc6.

    Obviously it seems that a “final minor turn” course correction could be made after 19:41.

  594. Victor Iannello says:

    @Mick Gilbert: I looked more carefully at the transitions available for WMKP/ILS04. Indeed, there is D155L, which provides waypoints that follow the 10 DME arc for VPG, and D130Q, which provides waypoints that follow the 15 DME arc for VPG. However, neither of these arcs agrees with the recorded radar data, which seems to be at 7 DME as the recorded path crosses the runway course. (I still believe the only way a DME arc can be programmed is to use a series of waypoints.)

    To avoid confusion, here is the radar data near Penang that @sk999 has digitized from the radar data presented graphically in the RMP report. I have no idea how accurate these data are, but they are arguably the best we have. If you or anybody else has better data, or has reason to believe these data are in error, it would be helpful to know.

    5.352,100.624
    5.301,100.506
    5.287,100.478
    5.263,100.436
    5.249,100.413
    5.218,100.360
    5.220,100.289
    5.195,100.216
    5.195,100.075
    5.236,99.942
    5.296,99.852

  595. Andrew says:

    @ Victor et al,

    ”If you are willing to propose a scenario, I think many of us would be interested to hear your thoughts.”

    Unfortunately, I don’t have anything radically new to offer, just a few thoughts on some of the theories that have been presented to date.

    The various failure scenarios (eg cockpit fire, oxygen bottle rupture, nosewheel blowout) can explain the aircraft’s loss of comms and the initial diversion from the flight planned track, but in my opinion they fail to adequately explain the NW track up the Malacca Straight and the FMT. The NW track might be explained by an FMC route discontinuity followed by flight on a constant magnetic heading (although that now seems doubtful), but I am yet to see a plausible explanation for the FMT. I don’t believe the theories about a diversion to Banda Aceh gone wrong are credible, as I have already explained.

    That leaves the ‘rogue pilot’ theory. If a rogue pilot was responsible for the disappearance, then it’s possible the other pilot was either locked out of the cockpit or otherwise incapacitated some time before the aircraft ‘went dark’ at around 1720 UTC. A standard ACARS report was sent by SATCOM at 1707 UTC, but the next report, expected at 1737 UTC, was not sent, suggesting the SDU was depowered at some time between 1707 and 1737 UTC. I suggest the L Main AC Bus was depowered about the time the aircraft passed IGARI (1720 UTC), when the aircraft’s Mode S symbol also dropped off the ATC radar screens. The aircraft would then have been diverted back towards Penang.

    The pilot might then have depressurised the aircraft with a view to rendering the cabin crew and passengers unconscious or dead. The passenger emergency oxygen system would have activated when the cabin altitude exceeded 13,500 ft, but the chemical generators would only have had a duration of around 22 minutes. The occupants are likely to have begun losing consciousness before that time due to insufficient oxygen partial pressure. In a depressurisation, the cabin crew are trained to sit down and don the nearest drop-down mask, so I think it’s unlikely any of them would have used the portable oxygen. They would have been unaware of what was happening and I suspect they would have become hypoxic and lost consciousness at much the same time as the passengers. Death would have followed soon afterwards.

    With the rest of the aircraft occupants unconscious or dead, the pilot would have been free to restore power to the SATCOM, resulting in the SDU initiated log-on request at 1825 UTC. In the meantime, he would have flown the aircraft around the southern tip of Penang and then NW up the Malacca Strait. Once clear of the northern tip of Indonesia, he would have turned the aircraft south towards the SIO. At some point I assume he also re-pressurised the aircraft and removed his oxygen mask.

    The psychologists seem to think that a pilot intent on suicide would need to see it all or most of the way to the end. It’s possible the pilot remained conscious after the engines flamed out, but the evidence seems to suggest the aircraft was not piloted during its final stages of flight. The aircraft would have depressurised during the final descent, but I guess it’s also possible he hastened that process and rendered himself unconscious some time before the aircraft crashed.

    In summary, as much as I would like to think that MH370 was the result of some kind of technical failure, I don’t believe that such a scenario is supported by the evidence. That leaves us with the rogue pilot theory, which does fit the evidence. Ultimately, the file fragments found on the Capt’s home simulator are too big a red flag to ignore.

  596. Victor Iannello says:

    @Andrew: I think that is all quite sensible, as I suspected it would be. Thank you for being bold enough to express your thoughts.

  597. Mick Gilbert says:

    @Victor

    Re “There is no way to program an approach along a DME arc rather than choosing individual waypoints that are at various bearings and constant distance from a fix.”

    I don’t think that’s correct, Victor. Can you be sure that the 10DME and 15DME approaches were not in the MAS database.

    Re “If the pilot had chosen the BIDM1A (BIDMO) STAR for the ILS04 runway, the waypoints would be BIDMO, PUKAR, ENDOR, MEKAT, KENDI, RTE DISC, CF04, FF04, RW04.”

    Agreed but having chosen and programmed the BIDMO 1A STAR it would not be unreasonable to subsequently clean it up and straighten it by removing PUKAR-ENDOR (and possibly even BIDMO) in order to expedite the approach.

    Re “I think the best representation of what was actually flown near Penang is the radar data presented in the RMP report …”

    I disagree, I think that the 0152:35 MYT PSR trace is the best and only primary evidence, everything else has been modelled and smoothed to join up the track of the airplane back across the Malay Peninsula with the 1802 – 1822 UTC Malacca Strait radar track that is now being questioned. That flattened turn in the RMP and the DSTG 10 second radar data track makes no sense in any expected mode of navigation.

    Re “The radar data near Penang is not consistent with a route discontinuity at KENDI.”

    To the contrary, allowing for some margin of error, the 0152:35 MYT PSR trace is not inconsistent with an early turn towards KENDI off a track of 240-ish towards MEKAT.

    Getting back to how well a turn through KENDI onto a CMH of 300-ish° ends up looking against the 1825-1828 UTC BTO/BFO, I’m guessing that it’s a pretty good fit and doesn’t require an offset manoeuvre.

    As for the FMT, I am certainly not ignoring it. I think that whether we’re in the deliberate malicious action camp or the inflight emergency camp we’re in general agreement that there was some sort of manual intervention from the flight deck just prior to 1825 UTC. I’ve addressed what that intervention might have been and how it might fit with an FMT at length previously.

  598. Victor Iannello says:

    @Mick Gilbert: “I don’t think that’s correct, Victor. Can you be sure that the 10DME and 15DME approaches were not in the MAS database.”

    They are, as I said in my last comment. But the RMP radar data doesn’t support either 10DME or 15DME.

    “I think that the 0152:35 MYT PSR trace is the best and only primary evidence, everything else has been modelled and smoothed to join up the track of the airplane back across the Malay Peninsula with the 1802 – 1822 UTC Malacca Strait radar track that is now being questioned. That flattened turn in the RMP and the DSTG 10 second radar data track makes no sense in any expected mode of navigation.”

    Perhaps our basic disagreement is the radar data. You wish to use the PSR data from the FI as the best available. I think the data in the RMP report is better because we can resolve the coordinates better. Ideally, the data sets should agree within the margin of error of the measurement. And I don’t believe there was any smoothing of the data to join with the 18:02 to 18:22 radar points because those points weren’t presented nor even acknowledged to exist (which I suspect they don’t).

  599. Mick Gilbert says:

    @Victor

    Thank you for that additional information, Victor. Your 8.21am and my 8.40am crossed paths.

  600. Victor Iannello says:

    @Mick Gilbert: You’re welcome. I am still thinking about ways that STAR/transition/approach options might be reconciled with the radar data. I think KENDI is too far from the path, but there might be other options. On the other hand, if the radar data is very imprecise, it’s hard to rule out anything.

  601. Andrew says:

    @Victor

    “Thank you for being bold enough to express your thoughts.”

    You’re welcome – I have my hard hat on in anticipation!

  602. buyerninety says:

    @Mick
    Clar. Glide Slope intercept 7.8DME ILS is marked on ‘ILS (LOC/DME) RWY 04’, ;
    http://aip.dca.gov.my/aip/AIP2016/graphics/17525.pdf

    (I suppose someone else will point out, if I don’t, that KENDI {at 11 DME} is situated
    on the 222° radial, & a reciprocal magnetic heading to that 7.8DME fix would be 042°!)

    Anyway, let me float a hypothesis as to the actions that led to MH370’s flightpath.
    (Not every ‘i’ is dotted or ‘t’ crossed, so ‘broad strokes’ only follow , OK?)

    Shah reprograms the FMC (maybe Hamid programed some of the details).
    He didn’t do a simple DIVERT TO, from present position, (for reasons not fully clear).
    He selected a route from the pre-stored route list, probably WMKK-VABB, and modified it,
    e.g.; -GUNIP-VAMPI- to say, instead -GUNIP-WMKC-WMKP-VAMPI- (maybe including a few other
    wapoints to/near those airports), and perhaps waypoints after IGOGU were deleted in ,or
    as a result of
    , that reprogramming process.
    This, then, was (at some point) made the Active Route.
    He or Hamid may have set the radio to autotune airport navaids.
    The autopilot captured the navaid as the aircraft came up to WMKC.
    The aircraft (at 30000+ feet) approached to the north of WMKC and turned & eventually
    overflew the GlidePath for WMKC, probably on a heading of approximately 235°,
    perhaps as seen here;
    http://aip.dca.gov.my/aip/AIP2016/graphics/17065.pdf
    It flew onwards across Malaysia (on that approximate heading), until (somewhat before
    Mount Bintang) it captured the navaid for WMKP, flew initially on a course to the navaid,
    then approached to the south of WMKP and overflew the GlidePath for WMKP, probably on a
    heading of approximately 267°, perhaps as seen in the first reference of this post.
    (It may have aimed to cross at the 7.8DME ILS fix point.)
    (Mick has suggested it may have additionaly flown a DME arc during this transit).
    The aircraft, having autotuned and sequenced two navaids that were along its Active Route,
    then continued to fly that Active Route, VAMPI-MEKAR-NILAM etc..

    This hypothesis may explain certain facets of MH370’s flightpath, ergo;
    1.) Why it flew towards WMKC and then moved north of WMKC.
    2.) Why it did not take a heading of approximately 239 to 246° (range of expected
    headings if flying from circum-WMKC space to WMKP), but rather flew a heading of 237° or
    slightly less.
    3.) Why it changed heading towards WMKP, as it came within approximately 50Nm of WMKP.
    4.) Why it flew towards WMKP and then moved south of WMKP.
    5.) Why it may have flown on a heading of 267° over the WMKP GlidePath (and possibly for
    a very short time afterwards, rather than turning towards VAMPI somewhat sooner).

    Cheers

  603. TBill says:

    @Andrew
    I am thinking your analysis is quite similar to pilot Ewan Wilson in his book with Geof Thomas “Goodnight Malaysian 370” which I have adopted as the best preliminary explanation of MH370, based on what we currently know.

    Re: Left Bus Turn off
    Thinking back to the Helios accidental depressurization scenario, I believe there were over-heat warning messages on ACARS from the EE Bay, because the thin air does not provide adequate cooling for certain air-cooled electronic equipment.

    >>Could the Left Bus Turn Off action be intended to prevent damage to some air-cooled devices possibly such as back-up or primary flight computers. or other key equipment (SDU)?

  604. Brock McEwen says:

    @all: what is standard operating procedure (SOP) in response to an incursion into restricted airspace? If it varies by country, what is SOP for each nation into whose FIR MH370 is hypothesized to have flown?

  605. Brock McEwen says:

    @all: I do appreciate the huge difference between an FIR and a restricted airspace. I’m using the FIR above merely to select a large yet relevant sample of SOPs; nothing more.

    @Victor: thanks. Any idea where I could lay hands on a comprehensive database of NOTAMS in effect March 7, 2014? I’d like to pore through and confirm the relevance of each.

    As you’ll know if you followed my convo with Don, I think we need to broaden our definition of relevance to include radar range. A military asset could have nothing whatsoever to do with CAUSING MH370’s fate, yet still greatly help us DETERMINE its fate, by virtue of the radar coverage rings around its position at the time.

    Thanks in advance for any direction you can offer.

  606. Victor Iannello says:

    @Brock: To find the NOTAMs, I would have to do an internet search just like you. I am very supportive of knowing all the radar assets that could have seen MH370.

  607. Victor Iannello says:

    @buyerninety said,

    “He or Hamid may have set the radio to autotune airport navaids.”
    “The autopilot captured the navaid as the aircraft came up to WMKC.”
    “it captured the navaid for WMKP, flew initially on a course to the navaid”

    I’m not sure what you mean.

    In LNAV mode for a B777, the navigation system is flying a route, roughly defined as a series waypoints connected by legs, and uses GPS, inertial navigation, and navigational aids to determine the current position. The plane would not “capture a navaid” and fly towards it unless the navaid was entered as a waypoint in the route. For a precision landing, the ILS or LOC frequency and course are automatically programmed based on the airport and runway using the ARRIVALS page. To use the LOC or ILS mode requires arming by pressing the LOC or APP buttons when in range and capturing the radial and the slope.

  608. buyerninety says:

    @Victor Iannello
    As an initial suggestion, (referring to the diagrams used in the post), it seems for WMKC & WMKP,
    the waypoint name to input may be e.g. -GUNIP-IKBIPG-VAMPI- etc if intending to use ILS,
    or e.g. -GUNIP-VKBVPG-VAMPI- etc if intending to use VOR/DME.

    I acknowledge that the FMS may request a more specific designation – in which case this information
    may be applicable (Referring to this document):
    http://docshare01.docshare.tips/files/27270/272700051.pdf
    page 6-50;
    “SELECT DESIRED WPT
    The SELECT DESIREDWPT page is automatically displayed when an identifier is entered into the CDU
    and the nav database contains more than one component (waypoint, NDB, VOR) with that identifier.
    The SELECT DESIRED WPT page is displayed to let the pilot select the desired nav database fix.”

    (Please note that if a more specific designation is not requested by the FMS, that the pilot may
    not have made the designation more specific. A reason for not making the designation more specific,
    could be that the pilot wished to leave the decision to decide which airport to land at until a
    later time – if the circumstance developed to be extremely serious, then WMKC would be chosen. If
    less serious, then the (further in time & distance) choice of WMKP could be chosen instead of WMKC.)

    I would draw to your attention certain other parts of that document which may be relevant;
    page 8-5;
    “The radio tuning function can be handled automatically by the FMS”…
    page 8-6;
    “VOR L/VOR R…
    — R (route autotuning) – The FMS-selected navaid is the next VOR, previous VOR, or a downpath VOR
    on the active route
    and within 250 NM of the current aircraft position.”
    page 8-7;
    “ILS–MLS…
    When an ILS, LOC, back course approach, or an ILS/LOC runway is entered in the active
    flight plan, the appropriate frequency/course is displayed
    in small font with a caret,
    followed by PARK. This display becomes active when the aircraft is within 200NM of the
    top-of-descent or the aircraft is more than halfway along the active route, whichever
    is closer to the destination…
    When a valid frequency is manually or automatically tuned, PARK is replaced with the
    frequency/course display (large font). The FMS autotunes the ILS/MLS associated with the
    active flight plan destination runway
    when the aircraft is within 50 NM of top-of-descent,
    or within 150 NM direct distance of the runway threshold (whichever is greater), or when
    active in descent.
    PARK indicates that the ILS tuning is not active, it is in a standby mode waiting for the
    proper conditions before the tuning becomes active.
    The tuning status following the frequency/course is A for automatic tuning, M for manual
    tuning, and PARK for selected but not active.”

    Of course, we should seek comment from Andrew on these matters.
    __________________________________

    I note that an alternate scenario may be applicable, with WMKK-VABB still made the Active Route
    but no waypoints for WMKC or WMKP added – if autotuning of navaids were possible/occurined, such
    that the FMS were autotuning the closest VOR/DME or ILS navaids, the FMS captured the navaid &
    then departed & flew onwards from the navaids of WMKC and WMKP (having sequenced those navaids).
    (In this case, having WMKK-VABB as the Active Route could prevent a ‘Not On Intercept Course’
    warning by the FMS.)
    ________________________________

    EDIT;
    In a previous post, I omitted to list as a precondition, that the MCP had a (lower) altitude
    limit (remaining) set on the MCP – thereby the aircraft never descended below that altitude
    limit (if, at all, any procedure would normally action a descent, in the absence of an MCP
    altitude limit).
    Obviously, people have differing viewpoints at to why no descent occurred .
    Cheers

  609. Andrew says:

    @TBill

    “I am thinking your analysis is quite similar to pilot Ewan Wilson in his book with Geof Thomas “Goodnight Malaysian 370” which I have adopted as the best preliminary explanation of MH370, based on what we currently know.”

    It’s interesting that the scenario I proposed is similar to Ewan Wilson’s. I have not read Wilson’s book, nor any other relating to MH370.

    “Could the Left Bus Turn Off action be intended to prevent damage to some air-cooled devices possibly such as back-up or primary flight computers. or other key equipment (SDU)?”

    That’s an interesting question. In the event of a depressurisation, the low airflow caused by the reduction in air density would probably cause the equipment cooling system to automatically reconfigure to the override mode, triggering the EQUIP COOLING OVRD EICAS message. In override mode, the system normally uses cabin differential pressure to provide cooling airflow to the equipment. However, if the cabin is depressurised, there will be little or no airflow. The associated checklist does not provide any guidance about turning off equipment; it only says: “After 30 minutes of operation at low altitude and low cabin differential pressure, electronic equipment and displays may fail.”

    Would a pilot shut down an entire bus to reduce the heat load in the MEC? Never say never, but I think its doubtful. Pilots are not taught about the power sources of individual components and that information is not listed in the flight crew manuals, so there’d be no knowing what equipment would be affected by shutting down the bus. Further, the major heat producers in the MEC have multiple, redundant power supplies, so shutting down the bus would probably have little effect. For what it’s worth, the SDU and other SATCOM components are located above the mid-cabin ceiling, close to the SATCOM antennae.

  610. Victor Iannello says:

    @buyerninety: You are confusing “autotune” with automatic capture of a VOR. Automatic capture of a VOR is not the way the FMS of a B777 operates. The only way to automatically fly towards a VOR (or any other waypoint) is to enter it as a waypoint with the A/P in LNAV mode. Or, if you want to follow a radial towards a VOR (or any other waypoint), you would set the course to the waypoint and then use HDG SEL to fly an intercept course. For an approach to an ILS runway, you would capture the lateral and vertical path by first arming and flying an intercept, as I said before.

    You can choose to put the Navigational Display (ND) in VOR mode, in which case you would see the “needle” and you could either manually fly or use HDG SEL to stay on course. But the VOR radial will not be flown automatically unless it is part of route.

    I don’t think you are going to believe me, so perhaps @Andrew can chime in.

  611. Andrew says:

    @buyerninety

    There are two conditions that must be met for FMC ILS autotuning to occur:
    • an ILS, LOC, back course, LDA (localizer-type directional aid), or SDF (simplified directional facility) approach has been selected to the active route, and
    • the aircraft is within 50 nm from the top of descent, 150 nm of the landing runway threshold, or the FMC is in descent mode.

    The first of those conditions requires that an ILS approach procedure be selected into the FMC active flight plan. That can only be done by selecting an approach via the DEP/ARR pages at the destination airport. The FMC will not autotune an ILS simply because an ILS station (eg IPG) has been inserted as a waypoint in the flight plan.

  612. Victor Iannello says:

    @Andrew: Am I correct in saying that even when the ILS frequency is autotuned (meaning the frequency is automatically entered and an “A” appears next to the frequency on the Nav Radio page of the CDU), the plane won’t automatically follow the ILS path until it is within range and armed by pressing APP for the vertical and horizontal paths or LOC for the horizontal path? Even after it is armed, the path has to be intercepted before it is captured.

  613. Andrew says:

    @buyerninety
    @Victor

    I think we crossed paths there Victor, but yes, I agree. The FMC will autotune various VORs as the aircraft proceeds along its route. The VOR stations the FMC autotunes might be associated with waypoints (eg VPG) that have been inserted in the flight plan, or they might be off track. However, the only time the FMC will autotune an ILS is when an ILS procedure has been inserted at the destination.

    The FMC never uses the raw data from a navigation aid to provide steering commands to the autopilot. It uses the FMC position to work out where it is and then provides steering commands to fly between the waypoints that are entered in the flight plan. For example, let’s say you want to fly direct to VPG VOR. You’d execute a ‘Direct To’ in the FMC LEGS page, with VPG as the active waypoint. The FMC would autotune the VOR and provide steering commands to the autopilot in LNAV mode. However, those steering commands aren’t based on the VOR ‘needle’; they are based on the FMC position, which is derived from a combination of the GPS, IRS and radio navaid positions, depending on the most accurate information available.

  614. Andrew says:

    @Victor

    “Am I correct in saying that even when the ILS frequency is autotuned (meaning the frequency is automatically entered and an “A” appears next to the frequency on the Nav Radio page of the CDU), the plane won’t automatically follow the ILS path until it is within range and armed by pressing APP for the vertical and horizontal paths or LOC for the horizontal path? Even after it is armed, the path has to be intercepted before it is captured.”

    Yes, that’s correct. If you left the autopilot in LNAV mode and didn’t arm the APP modes, then it would still fly the horizontal path defined in the flight plan, but it would not actually ‘capture’ the localiser beam.

  615. David says:

    The ATSB letter about Higgins has been published in the Australian today.
    “Fate of Flight MH370

    I am writing to express my concern regarding the manner in which The Australian’s journalist Ean Higgins continues to report on the search for Malaysia Airlines Flight MH370 — in particular, I am concerned at the negative impact this reporting is having on the knowledge of the search by the families of those on board the aircraft.

    Under international conventions, the government of Malaysia is responsible for investigating the causal factors behind the loss of MH370. I understand their investigation team is well progressed in developing a draft report. The role of the Australian Transport Safety Bureau, at the request of the government of Malaysia, is to lead the underwater search.

    Many of The Australian’s articles by Higgins report criticism of the ATSB’s search strategy. Byron Bailey and others, for example, contend that the aircraft was being actively controlled at the end of its flight and was glided to a location well away from the area which has been searched.

    The analysis of leading international experts used to determine the search area reveals the aircraft probably hit the surface of the ocean reasonably close to where the aircraft’s last transmissions were made with a satellite (what is known as the seventh arc).

    This analysis is based on solid evidence and has been extensively peer reviewed and published in a scientific journal. The analysis of the transmission data, when complemented by the recently published CSIRO drift-modelling analysis, provides the best possible definition of an area in which the aircraft is likely to be located. This is contrary to the views of Bailey and others.

    Higgins has also criticised the ATSB for not releasing a series of emails that relate to this analysis. I encourage your readers to view the ATSB’s response to Higgins’s Freedom of Information request on the “correcting the record” page on its website. The search has been extraordinarily difficult and challenging work. The Australian’s reporting undermines the good work of many dedicated professionals, committed to finding the aircraft.

    Greg Hood, chief commissioner, ATSB, Canberra, ACT”

    This is a shell of its former self. It no longer includes that Higgins, “consistently attempted to create the appearance of contention between the search strategy experts and members of the ATSB’s team where none exists”, his use of a “particularly objectionable” journalistic tactic, and “harassing and intimidating approach”.

    Instead it objects to Higgins putting the view that the aircraft could have been glided, on the implied grounds that a non-glide is what the experts concluded (“This analysis is based on solid evidence and has been peer reviewed…”). Except that to my knowledge it isn’t. Glide is to do with search width. For that the experts have worked from what is an assumption, no active pilot. However reasonable, it remains an assumption and surely that leaves room for Higgins to put another opinion?

    What happened?

    The original is here:
    https://www.atsb.gov.au/newsroom/correcting-records/mh370-reporting-by-the-australian/

  616. Andrew says:

    @David

    RE:“What happened?

    I suspect The Australian edited out the bits they didn’t want the public to see.

  617. ALSM says:

    David
    The fact that there was rapid descent has nothing to do with the question of whether there was a live pilot. The rapid descent in no way depends on pilot vs. no pilot. The data tells us there was a rapid descent. It does not tell us if there was anyone at the controls.

  618. Mick Gilbert says:

    @David

    Re “What happened?”

    Frankly that sort of nip tuck cosmetic editing is rife at The Australian. One notable example that I haven’t mentioned previously was from an Ean Higgins story “Darren Chester seeks briefing on hunt for MH370” dated 20 February 2016. In a late February 2016 interview with The Times here’s what Martin Dolan, then Chief Commissioner of the ATSB, said:

    “We’re not at the point yet, but sooner or later we will be — and we will have to explain to governments what the alternative is, and the alternative is, frankly, that despite all the evidence as we currently have the possibility that someone was at the controls of that aircraft on the flight and gliding it becomes a more significant possibility, if we eliminate all of the current search area. In a few months time, if we haven’t found it, then we’ll have to be contemplating that one of the much less likely scenarios ends up being more prominent. Which is that there were control inputs into that aircraft at the end of its flight.”

    Here’s how that was reported in The Australian by Ean;

    “The alternative is, frankly, that despite all the evidence, the possibility that someone was at the controls of that aircraft and gliding it,”

    There was wholesale chainsawesque editing of the original statement without even the slightest hint to the reader as to where words and phrases had been removed. Just one of the phrases removed, “… becomes a more significant possibility, if we eliminate all of the current search area”, sets an entirely different context for what Dolan had said; the phrase sets both a condition (the elimination of the current search area) and a probabilistic qualification (that it becomes a more significant possibility).

    I regularly spend three to four times as much time fact-checking articles in The Australian as it takes me to read them. The irony is made all the more delicious by The Australian’s slogan, “For the informed Australian”!

  619. David says:

    @Andrew. Maybe not as straightforward. Earlier I had asked ‘The Australian’ why the letter had not been published. Their (immediate) response was, “we are dealing with this letter in conjunction with the ATSB”.

  620. David says:

    @Mick. Thanks.I had not seen yours.

    More delicious to me than The Australian’s slogan are your untiring efforts to yard them to it.

  621. Mick Gilbert says:

    @David

    To be fair to The Australian, when I wrote to them about about the matter I’ve just referenced their Deputy Editor, John Lehman, contacted me for a chat. After getting the “So, what’s your connection to the ATSB?” bit out of the way, we had a very amicable discussion about what constituted “doctoring” with regards to quotes. While we disagreed about the veracity and accuracy of some of The Australian’s coverage I was thoroughly impressed by John’s professionalism and by his taking the time to deal with the matter personally over the phone.

  622. David says:

    @ALSM. Yes the current interpretation of the final BFOs is in the First Principles Report so has endorsement, hence there was a high descent rate. The conclusion that that continued depends in part on what caused it and also for how long it continued. The simulations do not prove anything much. Those that met BFO criteria were few, some may have been outside the simulation data base and anyway their nature assumed no pilot.

    The new assessment that the aircraft ended closer to the arc than earlier thought rests on few simulations and some were outside the certified flight envelope.

    I agree it is likely that the high descent rate continued but that remains an assumption.

    Irrespective, the scrubbing of the other criticisms of Higgins and the turning of the focus to imply that Higgins should not put an alternate view is what caught my eye.

  623. David says:

    @Ge Rijn. From what Gysbreght says, the number of ‘hits’ in the simulations could reduce from 2 to 1 of 7, leaving aside those of configurations the ATSB describes as unlikely. Also I am unsure whether these/it, exhibiting descent rates between the maxima and minima at the BFO timings, were the same that met the 8 secs 10,800 fpm difference both maxima and mimima required; and if so with reasonable accuracy.

    All this does not sit well with what was implied in the Review conclusion that, “Results from recent simulations showed high rates of descent broadly consistent with the BFO analysis”.

    However we are on the same page with your, “First to set something straight. I’m not inclined to set aside the final BFO’s or the descent rates. I’m not at all in a position to do that by any means.”

  624. David says:

    @Andrew. Points about your scenario.

    “..the cabin crew are trained to sit down and don the nearest drop-down mask, so I think it’s unlikely any of them would have used the portable oxygen” . New thanks, at least to me.

    “..the pilot would have been free to restore power to the SATCOM, resulting in the SDU initiated log-on request at 1825 UTC”. Why would he want to do that? Even up fuel flows? Also, does a left AC bus depowering of itself explain how the aircraft ID was not retrieved?

    “That leaves us with the rogue pilot theory, which does fit the evidence”. You did not mention hijackers, that is whether the pilot might have been compelled. Not as tight a fit with the evidence I would think but still in the mix?

    More generally, why would a rogue pilot pick a flight to Peking that entailed a turn back? To attempt to answer myself I suppose he might have run simulations in a musing, fantasising sort of way rather than it being for planning (though I still think there was a chance he was testing the system). The turn back would then have been a spontaneous decision, or to mark the political import of that day, or was for display. But a long trip west or NW would have got him away from all radars without risk of suspicion.

  625. Mick Gilbert says:

    With regards to cabin crew actions in the event of a depressurisation, MAS train that if a descent is not evident after the pax cabin masks have dropped the nearest cabin crew member to the flight deck is to immediately don a portable oxygen, enter flight deck using the emergency access code and ensure that the flight crew are receiving oxygen. Further, the first post depressurisation action item for cabin crew is to obtain the nearest portable oxygen bottle.

    Given the very highly experienced cabin crew on MH370 (the average length of service was over twenty years, the Inflight Supervisor had 35 years experience and even the most junior member had 13 years experience) I don’t think that it would be unreasonable to assume that subsequent to a depressurisation event one or more cabin crew would have obtained and donned portable oxygen bottles.

  626. Donald says:

    @David

    >You did not mention hijackers, that is whether the pilot might have been compelled. Not as tight a fit with the evidence I would think but still in the mix?

    More silliness? Hijackers whose motive was to disappear a 777 in the SIO? Or maybe Z was running dress rehearsals on his home simulator for just such a scenario? Ready to thwart their best laid plans for the greater good? Prepared for every possible eventuality?

    I think we can and should leave ‘hijackers’ out of the mix. Not only is in not a tight fit, it’s patently absurd.

  627. David says:

    @Donald. “Hijackers whose motive was to disappear a 777 in the SIO?” I do not think I even implied that was their motive or their intention was realised.

  628. Donald says:

    @David

    Eh, sure. It’s just ridiculous given what is known. But do carry on.

  629. Mick Gilbert says:

    @Donald

    “I think we can and should leave ‘hijackers’ out of the mix. Not only is in not a tight fit, it’s patently absurd.”

    Is that because there’s never been a hijacking of a commercial flight that has seen it end in the ocean a long way from its destination? Or that no hijacking has ever gone wrong and concluded with an unintended outcome?

  630. Gysbreght says:

    One week after MH370 went missing, the police seized computer and HDD’s from Z’s home. Was anyone at home in that week?

  631. Gysbreght says:

    MK25 was not connected to Z’s computer. Perhaps someone ‘borrowed’ it?

  632. Ge Rijn says:

    @Gysbreght

    Interesting questions. You suggest someone could have put it there to provide a ‘smoking gun’ in the week after the police seized the computer?

    It could be maybe. To me it’s a bit puzzling why Z. (if being the culprit) would disconnect a harddisk from his computer with possible ‘smoking gun’ evidence and not completely get rid of it by throwing it in a river or so.

    Another question might be: was MK25 disconnected but still in the computer or completely removed out of the computer lying around somewhere?

  633. Ge Rijn says:

    ..I put the question for I cann’t remember reading somewhere MK25 was an external harddisk. Was this already confirmed somewhere?

  634. Mick Gilbert says:

    @Gysbreght
    @Ge Rijn

    According to the RMP report the data sets for the “flight of interest” were found on exhibit Mk 25, a Scandisk Extream 240GB SSD (back in 2014 this internal HDD was a bit of a favourite amongst gamers because of its high speed and relatively low price). The report states that the inspecting officer “found that there are five (5) hard disks inside of this computer. However, only one hard disk that is connected (MK26) to this computer at the time it is confiscated.”

  635. Oleksandr says:

    @Donald:

    Re: “Hijackers whose motive was to disappear a 777 in the SIO?”

    Why not? Get something or someone they needed, descend to 3 km altitude over the Malacca, exit via EE-bay belly access door, and steer the plane to the SIO to get rid of any evidence. If the hypothetical hijackers selected slow ascent to a specified FL via MCP, they would have sufficient time to do it.

    Re: “Or maybe Z was running dress rehearsals on his home simulator for just such a scenario?”

    There is no evidence that Z. was running any rehearsals. Sadly, I see only desperate desire to hang Z. to justify the failed search operation, and make him responsible for the disappearance without any sounding proof.

  636. Oleksandr says:

    @Andrew,

    “…but I am yet to see a plausible explanation for the FMT.”.
    Believe it or not, there is a trivial explanation for it… The right DC bus is more troublesome.

  637. Oleksandr says:

    @Victor,

    Re: “I think the radar point at 18:22 is useless”

    We obviously disagree on this. It is very useful in my opinion. It might inaccurate, but it gives a good idea of the plane whereabouts, its heading and speed. The current working hypothesis of the IG and ATSB does not explain the remarkable matching of 18:25:27 BFO to the radar-derived velocity 18:22, and in this context your statement looks interesting.

  638. Donald says:

    @Mick

    >Is that because there’s never been a hijacking of a commercial flight that has seen it end in the ocean a long way from its destination? Or that no hijacking has ever gone wrong and concluded with an unintended outcome?

    No. It’s because the plane went incommunicado and completely dark at the perfect moment. As if it was intentionally done by whoever had command of the cockpit. Imagine that?

    And it’s because the PIC was flying simulations at his home to a place far far away (the SIO, in case you’ve forgotten) where the airplane and it’s passengers just so happened to ‘arrive’ at.

    But let’s keep the inane ‘hijackers’ theory alive despite no evidence, no motive, no hidden gremlins in wheel wells etc…must have been those pesky Russians.

    And Mick, you’d find good company over on JW’s forum what with the ‘hijackers’ and all.

    @Oleksandr

    >Why not? Get something or someone they needed, descend to 3 km altitude over the Malacca, exit via EE-bay belly access door, and steer the plane to the SIO to get rid of any evidence. If the hypothetical hijackers selected slow ascent to a specified FL via MCP, they would have sufficient time to do it.

    About what I expected from you.

    >There is no evidence that Z. was running any rehearsals. Sadly, I see only desperate desire to hang Z. to justify the failed search operation, and make him responsible for the disappearance without any sounding proof.

    Desperate desire? I don’t even know how to reply other than to say that I honestly believe you to be on the Malaysians payroll, or dumber and more naive than a sack of bricks.

  639. Mick Gilbert says:

    @Donald

    Slow your roll there, Digger. I’m not a proponent of the hijack theory, I’m trying to maintain an open mind and assess the evidence as dispassionately as possible. You might eliminate hypothesises based on your preconceptions, pardon those of us who don’t.

  640. Victor Iannello says:

    @Donald: No need to name call. Please.

  641. Victor Iannello says:

    @Oleksandr said, “We obviously disagree on this. It is very useful in my opinion. It might inaccurate, but it gives a good idea of the plane whereabouts, its heading and speed. The current working hypothesis of the IG and ATSB does not explain the remarkable matching of 18:25:27 BFO to the radar-derived velocity 18:22, and in this context your statement looks interesting.”

    OK, let’s just say the 18:22 radar point is of questionable accuracy and not argue about its utility.

    As for the working hypothesis of the IG, there is none. Individual members have proposed theories. There is a healthy range of opinions within the group. Richard, Mike, Don, and I have been more vocal about our opinions than others within the group, but we don’t speak for the others, or for each other.

  642. Ge Rijn says:

    @Mick Gilbert

    Thanks for the answer on the Mk 25 hard disk.
    It then seems quite unlikely to me someone else put it in that computer in the week before the police seized it.

  643. Ge Rijn says:

    @Oleksandr

    “descend to 3 km altitude over the Malacca, exit via EE-bay belly access door”

    This small door opens inwards and closes like a plug in the fuselage.
    I think it can not be opened during flight in a pressurerized cabin.

    And how do you think someone can leave the plane through that small door during flight not breaking his legs or back instantly the moment he was half out?
    It has been done but that was a B727 with its back stairs door lowered just above stall speed. Something quite different.

  644. Victor Iannello says:

    When considering the possibility of separation of parts from MH370 due to overspeed in a steep descent, it is useful to consider the recovered parts from SilkAir 185 in Dec 1997, which was flying from Jakarta to Singapore, and descended at high speed into the Musi River in Southern Sumatra. The US NTSB concluded that the captain deliberately lowered the nose and crashed the plane. Indonesian investigators disagreed, saying the evidence was inconclusive.

    Prior to the crash into the river, parts separated due to the high speed, and were found on land some distance from the crash the site. Therefore, we can compare the type and condition of parts expected from a high speed impact versus separation from high speed prior to impact.

    These pages, taken from the investigative report, show pictures of the recovered parts. Fragments of composite honeycomb panels recovered from the river resemble similar panels from MH370, and are shown in Fig 8b. By contrast, parts of the horizontal and vertical stabilizer recovered on land are shown graphically in Fig 10a and photographically in Fig 11. The size and condition of these parts resemble the flaperon and flap track fairing recovered from MH370.

  645. Victor Iannello says:

    @Ge Rijn, @Oleksandr

    A parachute jump using the bulk cargo door has been proposed before. The door opens from the inside, has dimensions of about 3’x4′, is accessible by using a hatch in the passenger compartment, and would put the jumper clear of wings and engines. The door open inwards, so the cabin would have to be depressurized.

  646. Ge Rijn says:

    @VictorI

    I can see sizes of several parts from SilkAir 185 are comparable to several MH370 pieces but I see no pieces as clearly identifiable. No flaperon or clear other flap pieces or pieces with undamaged edges or a hardly damaged monitor-mounting.
    Then the sheer amount of which looks like fuselage (picture B) and cabin debris (picture C) must be many thousands of pieces probably even tens of thousands.
    With thousands of honeycomb/floatable pieces.
    IMO if this happened to MH370 a very lot more debris should have been found by now.

    The final drawing shows only pieces from the tail section seperated that were found on land. If this is really all, than no flaps or other trailing edge parts of the wing seperated during the dive that were found on land or the water. Which is opposite to the MH370 debris. Most pieces are from the wing (trailing edge and engine) and only two from the tail section.

    IMO the damage and debris left after the Aisiane 124 crash landing is still a far better comparison.
    Flaperon seperated, outboard flap seperated, aileron damaged/seperated and so on. Even the vertical stabilizer shows comparable damage at the same spot. And the pictures of the cabin show some monitor mountings seperated and a same kind of closet was partly destroyed which reminds of the Rodrigues piece. Including many interior panels being smashed.
    Also a door (left aft) seperated completely on impact. When this also happened to MH370 pieces could have floated out.

    I suggest you and others take a close look at all the pictures in the following series:

    https://www.flickr.com/photos/ntsb/sets/72157634528378705/

  647. Ge Rijn says:

    I want to add this Flickr series of Asiana 214 damage is not complete (particularly cabin pictures). Thet all can be found easily on the net.

  648. Victor Iannello says:

    @Ge Rijn: You are comparing the damaged parts from SilkAir 185 and MH370 too literally. One was a B777, the other a B737. The fact that no wing control surfaces separated from the B737 doesn’t mean it couldn’t have occurred for the B777. Rather, I was comparing the size and condition of the honeycomb panels and the flight control surfaces for the two aircraft, which seem quite similar, in my opinion. But in any event, I see your point.

  649. DrB says:

    @Oleksandr:

    On May 6 you said: “Re: “The 18:25 BTO values do indicate that the plane turned away from the subsatellite radius” This could be a result of combined turn and descent (descent also causes increase of the distance). There are many emergency maneuvers that fit BTOs and BFOs, not only the lateral offset. Earlier I have shown that a spiral descent fits BTO and BFO perfectly, but this is not the emergency maneuver I consider in the present.”

    Please provide a combination of speed, track and descent rate (including a “spiral descent”) that produces the 273 Hz BFO measured at 18:25:34.

    Without this match, you do not have a “perfect” fit. You have a terrible fit.

    As an aside, I also believe that you can’t match the 18:25:27 and the 18:25:34 BFOs with an accelerating ASCENT with a B777.

    I maintain that OXCO overshoot is the ONLY viable explanation we have at this time for the 273 Hz BFO, and this can only occur after an extended (~1 hour) unpowered period at high altitude.

    So here is my simple question: Can anyone provide a VIABLE alternative explanation for the 273 Hz BFO?

    If not, then I would suggest that you and Dennis accept this point and move on. If your personal biases prevent you from accepting it, that’s OK, too, but you would help your case if you could demonstrate at least one alternative explanation that has not already been disproved.

  650. Victor Iannello says:

    @DrB: Welcome back! Do you have any preliminary results that you can share regarding paths that are compatible with fuel load?

  651. Oleksandr says:

    @DrB,

    Re: “Please provide a combination of speed, track and descent rate (including a “spiral descent”) that produces the 273 Hz BFO measured at 18:25:34.”

    Apparently you did not read the whole discussion. You know that nothing except a rapid descent or, alternatively, corrupted in some way BFO can explain this.

    An example of the spiral descent is here:
    https://www.dropbox.com/s/3eais38phjo9h0j/descent_1822_1841.jpg?dl=0

    This is only the top of iceberg – I have many more… Virtually one only needs to find a proper starting time and a proper descent rate to fit a holding pattern. Left, or right – does not matter – all may fit.

    Re: “I maintain that OXCO overshoot is the ONLY viable explanation we have at this time for the 273 Hz BFO, and this can only occur after an extended (~1 hour) unpowered period at high altitude.”

    You sound like a broken record. See diagrams in my technical note, where I explored possibility of zero compensation term to explain the abnormal value of 273 Hz:

    https://www.dropbox.com/s/r551bp495n2juoc/TN-ABFO-Rev1.0.pdf?dl=0

    Heading from NILAM to approximately IDKUT with the descent of approximately 5 m/s matches the BFO of 273 Hz under the assumption of zero compensation term. Note that BFO is very sensitive to the heading. Supporting reason: If the ADIRU was in the process of alignment, the AES would likely use zero velocity to compute the compensation term. I trust 8 seconds is sufficient for the ADIRU to cancel the alignment process, as it would sense motion, so all the subsequent navigation data would be sourced from SAARU and GPS, and hence the compensation term would be correct again. The very first value might still be correct because the data were still sourced from the SAARU and GPS.

    Re: “Can anyone provide a VIABLE alternative explanation for the 273 Hz BFO? ”

    I have provided it long time ago. Do you read anything except your own papers?

    Re: “I would suggest that you and Dennis accept this point and move on. If your personal biases prevent you from accepting it, that’s OK, too, but you would help your case if you could demonstrate at least one alternative explanation that has not already been disproved.”

    I will not accept your and Mike’s theory until you or someone else is able to explain 18:25:27 BFO. Can you explain it or not? I can’t recall my explanation of zero compensation term was ever disproved. If you can – please do so.

  652. Oleksandr says:

    @Victor,
    @Ge Rijn,

    “A parachute jump using the bulk cargo door has been proposed before. ”
    Yes, it is not a novel idea. But I recall that time a jump was considered from FL350, but not a descent to some altitude for this purpose.

    Re: “This small door opens inwards and closes like a plug in the fuselage. I think it can not be opened during flight in a pressurerized cabin.”

    Why would the cabin be pressurized in such a scenario?

    Re: “And how do you think someone can leave the plane through that small door during flight not breaking his legs or back instantly the moment he was half out?”

    Do you agree that this depends on the TAS at the moment of jumping? Enter a new TAS, ascent rate, and the final altitude on the MCP, and jump out when the plane is still slow at low altitude. Air speed of 180 kph at zero altitude results in approximately g acceleration of a human’s body. At 500 kph and 3 km altitude it would be 5.7g.

  653. Oleksandr says:

    @Donald,

    Re: “I honestly believe you to be on the Malaysians payroll”.

    Jeff Wise honestly believes that I am a Russian troll. A Russian troll on the Malaysian payroll? I simply travelled too much to judge people by their passports. Malaysians, btw, the most friendly nation in my opinion.

  654. Victor Iannello says:

    Oleksandr: “I will not accept your and Mike’s theory until you or someone else is able to explain 18:25:27 BFO.”

    Here’s what’s already been proposed as an explanation. On power-up, the transmit is inhibited until the temperature of the crystal is within a small error band. That occurs at 18:25:27. The temperature subsequently overshoots and the inhibit is not again set. We therefore see a good value, a peak, and settling of the BFO back to the good value.

    “But I recall that time a jump was considered from FL350, but not a descent to some altitude for this purpose.”

    I don’t recall anybody so bold as to propose a jump from FL350. Any reasonable jump scenario includes a descent.

    By the way, Zaharie Shah was an experienced paraglider. And there is no need to set a slow climb rate. The FMS can be programmed with altitude and speed constraints at waypoints such that the plane could automatically follow a slow, low path followed by an ECON climb and cruise until fuel exhaustion.

    “Malaysians, btw, the most friendly nation in my opinion.”

    Yes, nice people saddled with a corrupt government.

    “If the ADIRU was in the process of alignment, the AES would likely use zero velocity to compute the compensation term.”

    The ADIRU will not align while airborne.

  655. Oleksandr says:

    @Victor,

    Re: “On power-up, the transmit is inhibited until the temperature of the crystal is within a small error band. That occurs at 18:25:27. The temperature subsequently overshoots and the inhibit is not again set.”

    If the transmit is allowed when the temperature is within a small error band, why the next transmit is allowed when the temperature is already not within this small error band?

    Re: “The ADIRU will not align while airborne”.

    It will not complete the alignment process. The alignment will be cancelled once the ADIRU senses motion, but I do not see any reason why the alignment process can’t be initiated in the air. The alignment takes time – up to 15 minutes in polar regions. It is physically possible to switch off/on the ADIRU from the EE-Bay.

  656. Don Thompson says:

    Oleksandr,

    You’re now simply making ‘stuff’ up.

    To go from a suggestion that the AES was powered down and later restored by unknown hand using the circuit breakers in the MEC, to a suggestion that the ADIRU was intentionally power cycled, is really preposterous and demonstrates a flawed comprehension of the ADIRS.

    Take a break.

  657. Victor Iannello says:

    @Oleksandr asked, “If the transmit is allowed when the temperature is within a small error band, why the next transmit is allowed when the temperature is already not within this small error band?”

    Transmissions are inhibited during warm-up for obvious reasons. Once the oscillator is deemed warmed-up and the inhibit is removed, the temperature controller should ensure that the oscillator remains close to the setpoint temperature. The designers chose a simple start-up algorithm that produces acceptable performance. With this explanation, the BFO sequence at the 18:25 logon is explained without the need for invoking manoeuvers or intermittently incorrect navigational data.

  658. DennisW says:

    @DrB

    “If not, then I would suggest that you and Dennis accept this point and move on. If your personal biases prevent you from accepting it, that’s OK, too, but you would help your case if you could demonstrate at least one alternative explanation that has not already been disproved.”

    I don’t have a personal bias. Why would I? When ALSM says something is “very clear” when it is not very clear just annoys me greatly because it is bullshit.

  659. Oleksandr says:

    @Don Thompson,

    “You’re now simply making ‘stuff’ up. To go from a suggestion that the AES was powered down and later restored by unknown hand using the circuit breakers in the MEC, to a suggestion that the ADIRU was intentionally power cycled, is really preposterous and demonstrates a flawed comprehension of the ADIRS.”

    No, it is the IG made stuff up, resulting in the waste of time and public funds, when the simple logic worked against your 38S. Now your are trying to find an excuse for your failure. You have already demonstrated gaps in your knowledge of the ADIRU and navigation system, so better leave this topic.

    Btw, I have never suggested that “the AES was powered down and later restored by unknown hand” – you confused me with someone else. Perhaps with yourself?

  660. Oleksandr says:

    @Victor,

    Sounds reasonable, thanks.

    Re: “The designers chose a simple start-up algorithm that produces acceptable performance.”.
    Why no overshot on the ground?

  661. lkr says:

    @GR: “I can see sizes of several parts from SilkAir 185 are comparable to several MH370 pieces… Then the sheer amount of which looks like fuselage .. and cabin debris … must be many thousands of pieces probably even tens of thousands.
    With thousands of honeycomb/floatable pieces.
    IMO if this happened to MH370 a very lot more debris should have been found by now.”

    I’ll take you at your word that you’re Devil’s advocate for a ditching scenario, not wed to it. But your lawyering is sometimes a bit intense. We already had a round when you seemed to agree that a thousand honeycomb/floatable pieces could have reached the western coastline. My suggestion of perhaps 10x wasn’t reasonable, simply because the most of the 777 exterior — the fuselage and wings — are entirely Al alloys with a bit of Ti.

    The argument that “a lot more” should have surfaced isn’t persuasive — if no one is looking. What we know is that when someone [always BG] is looking, it was possible to find a piece or putative piece every couple of kilometers on a suitable beach in Madagascar. How would that be possible if no more than few hundred pieces entered the . A larger cloud of debris also makes the pieces recovered in SA more likely — as the statistical tail [head?] of a much larger population.

    The interior pieces are another matter — here, I’d suggest that high-speed impact needs to be reconciled with failure to be observed in SAR. But I’d expect most of the interior debris to break up and sink long before Africa, and the surviving pieces much more likely to be ignored as usual junk-wrack..

  662. Don Thompson says:

    @Oleksandr,

    Do play out the scenario that might suggest the ADIRU experienced a power cycle, otherwise was reset & underwent an attempt for airborne re-initialisation.

    Why no overshot on the ground?

    Review Holland’s paper, ‘The Use of Burst Frequency Offsets in the Search for MH370’, section ‘IV. EFFECT OF SDU STARTUP ON THE BFO’.

    Holland notes that the BTO decay trend was evident during five Log-On events recorded in the GES Logs (Internal study regarding SATCOM ground-station logs,” MH370 Flight Path Reconstruction Group – SATCOM Subgroup).

    The time & date of the five events referred by Holland, Table III in his paper, correlate with 9M-MRO positioned on the ground at KLIA.

  663. paul smithson says:

    I asked this earlier without response. If tx inhibited until within temperature tolerance, why does Holland expect that drif magnitude and duration is affected by powered down duration? Would we not expect to see similar drift behaviour on reaching temperature range? And could this not occur after any power down duration sufficient to take it out of temp tolerance range?

  664. Oleksandr says:

    @Don Thompson,

    Re: “Holland notes that the BTO decay trend was evident during five Log-On events”

    Take a look at Fig 8 and 9. The 6 of 7 presented sequences have exhibited the decay trend, which included the first value. The 7th sequence, which is the logon in question, is the only one when the very first BFO value does not fit the decay trend. Basically Holland shows in these figures that the BFOs during the logon 18:25 do not fit the overall decay trend.

  665. Oleksandr says:

    One more implication is with regard to 00:19 logon. The explanation suggested by Victor (May 8, 2017 at 5:34 pm) implies that the second BFO, which is -2 Hz, could also be affected by the overshot, which has nothing to do with the ambient temperature.

  666. Ge Rijn says:

    @VictorI

    Some questions I hope you have conclusive answers to.
    I have guesses but cann’t be sure about them.
    Do you know (exactly) how many drifters (green arrows) are present in the latest CSIRO KMZ-file animations starting in each segment?
    Is this number the same in each segment?

    Is this latest CSIRO-model also based on the historical drifters from the Global Drifter Program like the previous (2016) CSIRO-model?

    (n.b. In your topic the ‘previous report published in December 2016’ also shows the ‘recent report’..).

  667. Don Thompson says:

    @Oleksandr

    Basically Holland shows in these figures that the BFOs during the logon 18:25 do not fit the overall decay trend.

    The above is simply your interpretation.

    To understand why the BFO overshoot characteristic isn’t evident during each Log On procedure, it’s also necessary to understand that procedure implies a variable delay after the AES reaching a ‘ready’ state, factors for which include the OCXO attaining working temperature, and the AES transmitting the Log On Request SU. The longer this delay, the closer the oscillator is to stability.

    Holland’s Fig 8 and Fig 9 show the effects of 9M-MRO’s AES OCXO overshoot as positive, relative to the later stabilised frequency.

  668. Oleksandr says:

    @Don Thompson,

    Re: “The above is simply your interpretation.”

    How come? I only commented what is shown in these plots. All the 6 “calibration” curves exhibit the same monotonic trend within the first 3 minutes. The 7th curve, which is a subject interest, is different. If it was in other way around, I would not have objection (i.e. if one of the calibration curves is non-monotonic).

    Re: “The longer this delay, the closer the oscillator is to stability.”
    This has nothing to do with non-monotonic behavior of the 7th curve.

  669. Andrew says:

    @Mick Gilbert et al

    ”With regards to cabin crew actions in the event of a depressurisation, MAS train that if a descent is not evident after the pax cabin masks have dropped the nearest cabin crew member to the flight deck is to immediately don a portable oxygen, enter flight deck using the emergency access code and ensure that the flight crew are receiving oxygen. Further, the first post depressurisation action item for cabin crew is to obtain the nearest portable oxygen bottle.”

    ”I don’t think that it would be unreasonable to assume that subsequent to a depressurisation event one or more cabin crew would have obtained and donned portable oxygen bottles.”

    In the event of oxygen mask deployment, cabin crew are trained to immediately don the nearest oxygen mask and secure themselves at the nearest seat, on a passenger’s lap if necessary (I kid you not – there are spare masks at each seat row). The reason for the urgency is because the time of useful consciousness (TUC) is perilously short at high altitude. The TUC for healthy individuals is usually quoted as 30-60 seconds at FL350. However, it will be much less if there is a rapid decompression and/or the individual is engaged in any kind of physical activity.

    The cabin crew’s ‘post decompression’ duties are performed after the aircraft has descended to a lower altitude and the Commander has announced the aircraft is safe. The portable oxygen bottles are primarily intended for first aid use by passengers that might need supplemental oxygen after the decompression and subsequent descent. If the aircraft did not descend and a crew member attempted to enter the flight deck using the emergency access code, then:

    • In the rogue pilot scenario, the pilot could deadlock the flight deck door to deny access.

    • It is unlikely the portable oxygen bottles would be of much benefit to the crew or passengers if the aircraft remained at FL350. The continuous flow masks attached to the bottles are the same as the drop-drown masks and are not designed for continuous use above FL250.

  670. Don Thompson says:

    @Oleksandr

    How come?

    Refer above.

    Monotonic, or non-monotonic, “curve”

    Holland didn’t depict a curve, he charted the BFO data points and straight lines.

  671. Brock McEwen says:

    @all: I’ve today renewed my request to the ATSB for all data underpinning the “4600 microsecond” incremental offset, and asked for three more data items:

    1) detailed raw data from the flight simulator runs referenced in the ATSB’s Nov. 2016 report,

    2) detailed raw data from the manufacturer establishing 3:40 and 2:00 as, respectively, the prior and current point estimates for system reboot time (time from flameout to logon request), and

    3) verification of the width of the already scanned portion of the search box for whose widening the ATSB currently advocates (near 35s).

    I require all of the above in order to properly calibrate an integrated stochastic model of end-of-flight dynamics, from which to extract insight I hope might inform search area assessment.

    I will post the ATSB’s response, if any, to this forum.

  672. David says:

    @Brock McEwen. Gutsy play

  673. Mick Gilbert says:

    @Andrew

    I agree wholeheartedly; I’m sure that if it had have been a case of deliberate malicious action the perpetrator would have deadbolted the fligt deck door. All I’m saying is that I don’t think that it would be unreasonable to assume that subsequent to a depressurisation event one or more cabin crew would have obtained and utilised portable oxygen bottles. How effective they were in maintaining consciousness is problematic. By corollary, it would not be reasonable to assume that all of the cabin crew were unconscious shortly after the depressurisation event; Helios 552 serves as an example of a cabin crew member maintaining consciousness for more than two hours after the airplane failed to pressurise.

    Moreover, if it was a case of deliberate malicious action, then given MAS procedures, it is not unreasonable to assume that at least one cabin crew member would be promptly alerted to the fact that something was amiss.

  674. Brock McEwen says:

    @David: in what way?

  675. Andrew says:

    @Mick Gilbert

    Fair enough.

  676. David says:

    @Brock McEwen. Take your 1). They have said they will not release that information, being Boeing’s.

    However I meant it as complimentary. No harm trying and you never know…

    I particularly liked your ambitious aim, “to properly calibrate an integrated stochastic model of end-of-flight dynamics, from which to extract insight I hope might inform search area assessment.”

  677. Victor Iannello says:

    @Andrew, @Mick Gilbert: That was a good exchange of information and thoughts. Thank you.

  678. Brock McEwen says:

    @David: thanks for your kind words and encouragement.

    @ALSM: re: MH370 end of flight sims you ran in early 2015:

    1) exactly how many trials did you run?

    2) of these, how many happened to exhibit descent rates within what the ATSB set in its Nov. 2016 report as hard bounds at each of 00:19:29 and 00:19:37 (which for purposes of your sims we should describe as “2nd engine flameout plus, respectively, 120 and 128 seconds”)?

    3) for only those trials which DID satisfy those bounds, how far from Arc7 (FL0) was impact?

    Alternatively, I would be thrilled to accept a raw data dump of your sim runs, and rummage through them myself. Whichever you prefer.

    Profuse thanks in advance for your time and attention.

  679. TBill says:

    @Victor
    Did you consider 180S True Heading from BEDAX?
    Seems to me 180S from ISBIX works at CTH constant speed for Arcs3-7.
    Thus 180S from BEDAX almost works for all Arcs2-7.
    I am thinking if MH370 goes faster BEDAX to past ISBIX and one speed step change down it might work, but not able to check that for few days. I am thinking Brian Andersons calc shows a fast speed, but I think maybe a slow down between Arcs2 and 3 complicates the calc.

  680. Mick Gilbert says:

    There is a Boeing Flight Operations Technical Bulletin that might provide some useful context for two matters that have been previously discussed;

    – the possibility that the airplane was flown by hand (ie with the autopilot disengaged) at around the time of the initial turnback from near IGARI, and

    – whether the MCP may have been set to TRK rather than the default HDG.

    Boeing Flight Operations Technical Bulletin 777-20 dated November 15, 2005 (https://www.dropbox.com/s/pcqthetdsqx4bxw/Boeing%20Flight%20Operations%20Technical%20Bulletin%20777-20%20dated%20November%2015%2C%202005.pdf?dl=0) was issued to address AFDS Mode Control Panel (MCP) Faults. The background statement is as follows:

    “Several operators have reported in-flight events where various AFDS pitch and roll modes such as VNAV, LNAV and HDG SEL become unselectable or cease to function normally. Typically these faults do not generate a failure annunciation to the flight crew and may be caused by an MCP hardware (switch) problems. … ”

    The guidance provided for handling an AFDS anomaly is to “… attempt to correct the problem by disconnecting the autopilot and selecting both flight director switches to OFF. This will clear all engaged AFDS modes. When an autopilot is re-engaged or a flight director is selected ON, the AFDS default pitch and roll modes should engage. …”.

    A period of manual flight while the crew attempted to troubleshoot an AFDS anomaly might explain the temporary speed and altitude excursions suggested by the DSTG analysis of the 10 second radar data.

    The guidance goes on to state that “If the above action does not correct the fault, desired flight path can be maintained by selecting an alternate pitch or role mode.” There is a table that lists suggested alternate autopilot modes or crew techniques to be used for inoperable or faulty autopilot modes; for an LNAV mode failure it recommends “Use HDG SEL (or TRK SEL on 777 airplanes) …”.

    Accordingly, if the crew were attempting to troubleshoot a persistent or recurring ADFS anomaly it would not be unreasonable to assume that the HDG/TRK Reference Switch was selected to TRK. Once the switch to TRK was made it should be retained even if the autopilot was subsequently disengaged and the flight director switches were cycled. I imagine that it would also then be the default mode if LNAV was re-engaged and subsequently encountered an END OF ROUTE or DISCONTINUITY error.

  681. Gysbreght says:

    @Mick Gilbert: Thanks for giving thought to that aspect. In a few hours the IG will wake up to write their predictable response.

  682. Don Thompson says:

    @Mick

    How would such a Boeing Flight Ops Support Technical Bulletin be incorporated into an airline’s documentation provided for flight crew?

    Is there a subsequent Service Bulletin from Boeing or Rockwell-Collins to deliver a ‘fix’ for the root cause of the problem for which the Technical Bulletin describes a workaround?

    I imagine that it would also then be the default mode if LNAV was re-engaged and subsequently encountered an END OF ROUTE or DISCONTINUITY error.

    The FMC computation of path is independent of the AFDS settings. The HDG/TRK Reference Switch pertains to path guidance computed by the AFDS as set by the Heading/Track Selector. If the AFDS is set for FMC sourced Lateral Navigation, the FMC defines the path including post-EOR/DISCONTINUITY.

  683. David says:

    @Brock McEwen. About your earlier 2) and the 2 mins (more literally 1:59) elapsed time from fuel exhaustion to the log-on request, the addition of the ATSB’s “approximately” a minute for APU start and “approximately” another for SDU start. That is a vague word.

    I timed an APU start on the ground at a minute, give or take a couple of seconds, from its winding-up sound on a Youtube video. It takes between 30 and 40 secs for its air inlet to open fully and, according to the maintenance manual the APU start will not commence until that completes. I have been unable to reconcile that 1:30 plus with the ATSB 1 min. Conceivably the APU start time at altitude is shorter though I have my doubts that would be the answer.

    Since the BFO times are known, should the “approximate” times in fact add to longer than 1:59 that brings fuel exhaustion forward by that amount (also see the note below). More important though is the timing of those descents in the simulations (2 groups thereof) which are consistent with the BFOs; ie those which fall between the maxima and minima derived from the BFOs. If in fact the simulation timings when these descents were measured were at 1:59 after AC loss (and 8 secs after that) you will see that there could be a problem if in fact the APU or SDU fire-up aggregate was longer.

    I do think it likely that with simulation results coming under scrutiny during the Review, these having influenced the new search area’s width as they did, this can be explained. Unfortunately though, compared with descent rate derivation in the Search and debris examination update, the simulation coverage is cursory, apparently from information ownership reasons. It would be nice though to confirm the outcome should you succeed in obtaining source data.

    2 notes about the above:
    • Further on fuel exhaustion timing, note that APU start commences not at left engine flame out but when AC power drops off line. The IDG drops off line as the engine drops below idle but I do not think it is clear when the backup generator does.
    • About your 2 mins in place of 3:40, that results from changed advice from the SDU manufacturer (Definition of Underwater Search Areas Dec 2015, p9 footnote 6).

  684. Don Thompson says:

    @Brock,

    Have you set out a definition for ‘raw data’ when applied to your questions?

  685. Gysbreght says:

    I’m wondering what could have caused the autopilot to be disconnected or not available:

    1. A pilot input on the control wheel, column, or pedals would disconnect the autopilot, requiring the pilot to re-engage it.
    2. An electrical configuration change that made the autopilot unavailable.
    3. The high-acceleration manoeuvre that the DSTG observed in the radar data.

    Regarding the latter, could the attitudes and accelerations in that manoeuvre have been so extreme that the PFCS changed to secondary mode?

  686. Mick Gilbert says:

    @Don Thompson

    Re: “How would such a Boeing Flight Ops Support Technical Bulletin be incorporated into an airline’s documentation provided for flight crew?”

    A very good question, Don, and one that Andrew would be best placed to answer (I had only just asked Andrew pretty much the same question via another channel prior to checking on developments here).

    Re: FMC and AFDS settings.

    We know that when the FMC encounters an END OF ROUTE or DISCONTINUITY error while in LNAV it reverts to a constant magnetic heading, ostensibly because that’s the default MCP setting. I’m wondering if the MCP HDG/TRK Reference Switch was selected to TRK whether that would cause the FMC to revert to a constant magnetic track.

  687. Mick Gilbert says:

    @Gysbreght

    Re:

    Have a look at the ATSB’s Aviation Occurrence Report – 200503722, In-flight upset event 240 km north-west of Perth, WA, Boeing Company 777-200, 9M-MRG, 1 August 2005 (https://www.atsb.gov.au/publications/investigation_reports/2005/aair/aair200503722/).

    The incident involves a sister ship to 9M-MRO that experienced a quite extraordinary “inflight upset” (the term hardly covers the event). From a normal climb out from Perth passing through FL365 the airplane suddenly pitched nose-up to +17.6º and climbed through 38,590 ft at a vertical speed increasing to 10,560 fpm. A/P overspeed and stall protection activate together! Despite the best efforts of the crew the airplane reaches 41,480 ft and CAS drops to 158 kts. A series of oscillations follow – pitch down to -7.4º with rate of descent reaching 7,824 fpm then pitch up to 12.7º. Eventually the crew managed to disengage both the autopilot and autothrottle and get back to Perth.

    If you turn to page 36 of the report and look at Figure 9: Relevant parameters displayed over a five-minute period, you’ll see altitude and speed excursions that are not markedly dissimilar to your interpretation of the DSTG’s Figure 4.2.

    The problem was traced back to an accelerometer that had failed 4 years earlier; at the time of the incident a second accelerometer failed and a latent software anomaly allowed the ADIRU to use data from previously failed accelerometer. The initial upset caused the autopilot to disengage, the crew ran into problems because the autothrottle kept re-engaging because it remained armed.

  688. Don Thompson says:

    @Mick

    We know that when the FMC encounters an END OF ROUTE or DISCONTINUITY error while in LNAV it reverts to a constant magnetic heading, ostensibly because that’s the default MCP setting.

    No, it’s a mistake to conflate the operation of the FMC and the AFDS-MCP as you suggest above.

  689. Don Thompson says:

    @Mick,

    It’s a coincidence that you should highlight the 9M-MRG event caused by the ADIRU faults, I reviewed it only yesterday.

    During that incident there was no un-commanded autopilot disengage. The crew’s first action was to manually disengage the autopilot.

    The erroneous data, generated by the ADIRU, was not detected by the ADIRS ‘self-monitoring’ between the ADIRU and SAARU. The erroneous parameters output by the ADIRU are not generated by the SAARU and, therefore, no comparison was possible.

  690. Ge Rijn says:

    @VictorI @others

    I’ve been counting the drifters in the latest CRIRO-model animation on KMZ-file 35S non flap (2dp).
    Assuming these drifters and model are also based on the Global Drifter Program data base of historical drifters.
    I count around 290 drifters starting on the 7th arc spreading out.
    I assume this number is the same in every other segment.

    Then I counted the landings on the African coast and Islands and on WA on 31-12-2015 for the 35S non-flap (2p)segment, the other 35S non-flap segment and the 35S flap segment.
    Then the 30S non-flap segment and the 30S flap segment.

    I think it shows something interesting:

    35S non-flap (2dp)segment : 41 African landings 12 WA landings
    35S non-flap other segment: 188 African landings 27 WA landings
    35S flap segment : 133 African landings 9 WA landings

    30S non-flap segment : 95 African landings 5 WA landings
    30S flap segment : 123 African landings 0 WA landings

    First the model seems to predict an average of 120 pieces of drifter-like debris arrived on African shores till 31-12-2015 and an average of ~13 on WA shores coming from 35S out of a drifter-like debris field of only ~290 pieces.

    Coming from 30S the average is 109 on African shores and ~2 on WA shores.

    Compared to the 27 pieces found so far on African shores and zero on WA shores the 30S numbers are reflecting reality a lot better.
    But this numbers also indicate IMO only a relatively small debris field of ~300 drifter-like pieces is needed and likely to give these results.

    It also suggests at least ~70 drifter-like pieces must be still lying around somewhere since december 2015 on African coasts and islands and maybe also a few on the North-coast of WA.

  691. Victor Iannello says:

    @GE Rijn: I count around 366 drift paths for 35S low drift, non-flaperon parts.

    I suspect that many more parts landed on shores that were not recovered. I base this on the repeated successes that Blaine Gibson had in finding the parts. For instance, No Step in Mozambique was found soon after he started searching. Another part in Madagascar was found just after it beached as Blaine found it on his second scan of that part of the beach hours after his first scan. I suspect many parts beached, were not recovered, were carried back out to sea, and later beached.

  692. Gysbreght says:

    @Mick Gilbert: ATSB Aviation Occurrence Report – 200503722, Findings: “Other outputs sourced from the ADIRU such as altitude, airspeed, mach
    number, wind speed, wind direction, vertical speed, pitch attitude, roll attitude,
    heading, temperature, and drift angle remained valid for the flight.”

    In the case of MH370, could the airspeed or attitude have reached a value where it was considered invalid?

  693. Victor Iannello says:

    @Mick Gilbert: I missed something. In the case of the faulty ADIRU in 9M-MRG, the pilots disengaged the A/P and A/T, communicated with ATC, and successfully landed in Perth. How is this like MH370?

  694. Oleksandr says:

    @Victor,

    You asked: “I missed something. In the case of the faulty ADIRU in 9M-MRG, the pilots disengaged the A/P and A/T, communicated with ATC, and successfully landed in Perth. How is this like MH370?”

    Extract from Don Thompson’s comment above: “The crew’s first action was to manually disengage the autopilot.”

  695. Oleksandr says:

    @Don Thompson,

    “It’s a coincidence that you should highlight the 9M-MRG event caused by the ADIRU faults, I reviewed it only yesterday. ”

    You should have reviewed it at least 2 years ago… However, in contrast to 9M-MRG, failure of the ADIRU of 9M-MRO could be due to an external impact, not ADIRU’s internal issue.

  696. Andrew says:

    @Don Thompson
    @Mick Gilbert

    “How would such a Boeing Flight Ops Support Technical Bulletin be incorporated into an airline’s documentation provided for flight crew?”

    Boeing issues several different types of bulletins, including Flight Operations Technical Bulletins and Flight Crew Operations Manual Bulletins.

    The Flight Operations Technical Bulletins provide information that is not critical for flight safety, but is nevertheless useful for airline operations or training. They are often used to provide background information on changes to procedures or new procedures. It’s up to the individual airlines how they choose to disseminate the information to their flight crews. It seems that MAS keeps of a copy of the Technical Bulletins on their aircraft.

    Flight Crew Operations Manual Bulletins are issued to provide urgent information or procedures that have a direct bearing on the safe operation of the aircraft, pending formal revision of the FCOM. They are included in a separate section of the FCOM.

  697. Don Thompson says:

    @Oleksandr

    Haha! I omitted to include ‘again’ above, that is, ‘I reviewed it again…’

    The report includes an interesting note concerning calculation of accelerations & the ADIRU’s location in the airframe.

    I am satisfied that, after 18:25z, 9M-MRO’s ADIRS operation was sound and the AES received the necessary data to operate satisfactorily.

  698. Oleksandr says:

    @Don Thompson,

    “The report includes an interesting note concerning calculation of accelerations & the ADIRU’s location in the airframe.”

    There are many interesting notes, though this report is far from being comprehensive… The ADIRU’s location is on the left front rack next to the main battery (E3 rack, the lowest shelf). The ADIRU, VHF-L, and LH nose landing gear wheel are approximately on the same line, which was a starting point for my “tire rim rupture” hypothesis (aka “the missing bolt”).

    Re: “I am satisfied that, after 18:25z, 9M-MRO’s ADIRS operation was sound and the AES received the necessary data to operate satisfactorily.”

    Are you sure about that? The ADIRU transmits on the 2 of 3 ARINC channels (left and right). The 3rd channel is allocated for SAARU. In principle AES may receive necessary data from SAARU and GPS to operate satisfactory. Though I have not found anywhere that AES can use data from SAARU and GPS, I think it would be a logical redundancy measure without any cost implications, because if the ADIRU fails when the plane is in far from ground VHF coverage, SATCOM would still be available.

  699. Brock McEwen says:

    @Don: I did not actually use the word “raw” in my email to the ATSB. I’ll flip you the email itself, so you’ll have the exact wording.

    What I meant – which I hope I made clear in the email – is that I need more than just summary descriptions and statistics – I need seriatim records (of flight sim runs, prior flight BTOs, reboot time trials, etc), so that I can properly shape he distribution of each variable.

    I don’t care if it is “processed” for clarity of meaning – in fact, I’d prefer it. What it cannot be is “processed” to remove detail.

  700. Victor Iannello says:

    TBill: That’s one of many possibilities.

  701. Mick Gilbert says:

    @Victor

    Re: “I missed something. In the case of the faulty ADIRU in 9M-MRG, the pilots disengaged the A/P and A/T, communicated with ATC, and successfully landed in Perth. How is this like MH370?”

    G’day Victor,

    No, I don’t think you missed anything, that’s an appropriate summary of the crew response and outcome of the 9M-MRG incident. I raised the incident in response to Gysbreght querying reasons for auto-pilot disengagement and his list of possible reasons did not include deliberate manual disengagement because of an obvious fault. The 9M-MRG incident is an excellent example of exactly those circumstances involving an airplane that just happened to be from the same stable as 9M-MRO.

    For fear of stating the obvious, MH370 remains a mystery, at least in part, because there is no previous aviation event that is directly comparable from the perspective of either deliberate malicious action or response to an inflight emergency. No pilot suicide or response to an inflight emergency has entailed the time to end-of-flight and manoeuvring that we see with MH370. For that reason I think that it is important to look at aviation events that might be in some way partly analogous; that approach applies equally to fleshing out both the deliberate malicious action and response to an inflight emergency hypothesises.

    If nothing else the 9M-MRG incident report makes for an interesting read, moreso if you’re a fan of roller-coasters.

  702. Don Thompson says:

    @Oleksandr,

    ADIRS interfaces to the three (L,C,R) A629 Flight Control busses. In normal operation the SAARU provides a pass-thru, while performing a comparison check, of ADIRU data to the Centre bus.

    GPS is not an alternate source for position data used within the avionics, the ADIRU is the source. GPS is used during alignment of the ADIRU and, during flight, as an input to the ADIRU’s process for reducing drift errors.

    The AIMS DCGF relays the required set of position and attitude parameters over two redundant A429 busses to the AES.

    So, yes, I am satisfied that, after 18:25z, 9M-MRO’s ADIRS operation was sound and the AES received the necessary data to operate satisfactorily. Holland’s work confirmed the OCXO’s characteristic for overshoot and the effect of aircraft vertical speed on the received frequency of bursts is well understood.

  703. Mick Gilbert says:

    @Andrew

    Thank you for the information regarding the handling of Flight Operations Technical Bulletins and Flight Crew Operations Manual Bulletins.

  704. Mick Gilbert says:

    @Don Thompson

    Re: “During that incident there was no un-commanded autopilot disengage. The crew’s first action was to manually disengage the autopilot.”

    Actually, Don, I think that there’s some ambiguity around the initial disengagement of the autopilot, and I say initial disengagement because the autopilot was disengaged six times in 22 minutes, five of those following crew actions to re-engage it.

    According to the Sequence of Events (pp.27-28), the initial autopilot disengagement is described as:

    “17:03:12 – 17:03:29. A/P overspeed and stall protection activate together and the AFDS pitch mode goes to FAIL resulting in A/T changing to speed mode. The A/P disengages and the thrust levers retard slightly before returning to original 65º position. A/P disconnect is again pressed …”

    There’s no clear indication in that statement of an initial crew intervention disengaging the autopilot; I imagine that may have happened as the autopilot tried to manage the conflicting overspeed amd stall protection activations. “A/P disconnect is again pressed” indicates that there was most certainly a crew intervention subsequent to the initial disengagement.

    We then have this sequence of re-engagement/disengagement;

    17:04:17 – 17:04:58. … During descent, stick shaker activates a
    number of times prior to A/P re-engaged at 17:04:39 in VNAV/LNAV mode resulting in immediate activation of A/P overspeed
    and stall protection and pitch mode failure. A/P disconnected five seconds later. …

    17:05:37 – 17:05:41. … First officer (F/O) engages A/P in vertical speed and track hold mode. Pitch mode fails and stall protection active. Pitch mode is changed to flight level change mode.

    17:05:47. Vertical speed reaches 4.400 fpm and A/P disconnected at FL386.

    17:23:04 – 17:23:06. 9M-MRG at FL200. right A/P engaged in vertical speed and heading attitude hold mode. Aircraft pitches nose up and the pitch mode fails and stall protection activated. A/P disconnected.

    17:24:20 – 17:24:31. 9M-MRG at FL197. left A/P engaged in flight level change and heading hold mode. Aircraft pitches nose down and banks to the right. A/P disconnected.

    17:24:51 – 17:24:58. A/P re-engaged in vertical speed heading hold modes with immediate pitch mode failure.

    17:24:59 – 17:25:03. 9M-MRG at FL167. A/P pitch mode changed to flight level change mode and pitch mode failure ceases. A/P disconnected at 17:25:04.

  705. Mick Gilbert says:

    @Gysbreght

    Re: “In the case of MH370, could the airspeed or attitude have reached a value where it was considered invalid?”

    Sorry, I’m not qualified to comment on that. Don?

  706. Don Thompson says:

    Mick,

    9M-MRG inflight upset: History of Flight & Summary sections of report both state that it was a crew member who first disengaged the autopilot. That is, it did not disengage as a consequence of invalid data.

    and Gysbreght

    “In the case of MH370, could the airspeed or attitude have reached a value where it was considered invalid?”

    Does my most recent response to Oleksandr help?

  707. lkr says:

    @Victor: “I suspect that many more parts landed on shores that were not recovered. I base this on the repeated successes that Blaine Gibson had ….I suspect many parts beached, were not recovered, were carried back out to sea, and later beached.”

    Thanks for chiming in here. GR is a bulldog on this, but I think he’s just wrong in imagining a very small debris field. Unless you accept conspiracy theories about Blaine Gibson [I’m referring to those that start out “Blaine speaks Russian”], his is the only semi-quant sample. If BG can gather a piece of MH370 from every 2-4 km he covers, and even he [or Pattiaratchi, as it seems, advising Gibson] has cherry-picked very promising areas, the population of buoyant debris capable of reaching the western shores was much larger than the 70 or so GR imagines. Anyone working with the drift at least has to take BGs ‘census’ seriously in scaling the original population and to a lesser extent, the size distribution of buoyant debris.

  708. Victor Iannello says:

    @Ikr: Yes, I agree with your logic.

  709. Victor Iannello says:

    @Mick Gilbert: Two years ago, I looked at the radar data closely and concluded that the civil PSR data presented in the March 2015 FI was consistent with MH370 flying straight segments with several small turns until the large turn south of Penang, as shown in this figure (yellow line). I suspect MH370 was flying with autopilot engaged and either in HDG SEL or TRK SEL roll mode. More recently I have concluded that the radar data from the time that the transponder became inoperable up until the first civil PSR capture at 17:30:37 is either inaccurate or fabricated. The unrealistic dip in speed, the unrealistically sharp turn at the turnaround, and the intermittent, shaky lines are poor attempts to fill in the blanks. However, I do believe the civil PSR data is valid.

  710. TBill says:

    @Mick Gilbert
    “We know that when the FMC encounters an END OF ROUTE or DISCONTINUITY error while in LNAV it reverts to a constant magnetic heading, ostensibly because that’s the default MCP setting. I’m wondering if the MCP HDG/TRK Reference Switch was selected to TRK whether that would cause the FMC to revert to a constant magnetic track.”

    That an interesting question that has not come up to date, because it has been hard to find any magnetic setting (heading vs. track) that meets the Inmarsat ping rings. Right now I like the path ISBIX 180S True Heading, the wind gives me shape I need to hit the rings.

    If you look at hypothetical straight, due South tracks on Google Earth, down say 93E, 93.5E, 94E etc. I think it shows, to meet the ping rings, you need a mild “S”-shape curve where the wind at first takes MH370 Westerly but later switches Easterly below 20S. The problem with Magnetic Heading is after Arc5 the swing to the East is to severe. The problem with Mag or True Track is between the first Arcs 2-4, MH370 really needs to shift more westerly with the winds (towards say 93E) to meet the rings timing. My current thinking.

  711. sk999 says:

    I have no opinion on the significance of the 7 save-points found in the Volume Shadow Information file on disk MK25.

    However, Microsoft’s Shadow Copy service seems to be pernicious in retaining data that a user unsuspectingly thinks has been wiped from his or her disk. The following (pdf p. 25) is an account of a person who thought they had committed the perfect computer crime, but was done in by a Shadow Copy.

    http://www.verizonenterprise.com/resources/reports/rp_data-breach-digest-2017-perspective-is-reality_xg_en.pdf

  712. DrB says:

    @VictorI,

    An unexpected and tragic death in the family has occupied me for the past month. However, I am still following your blog and am continuing to engage in modeling fuel consumption at arbitrary air speeds as a precursor to more accurately predicting combinations of altitude and airspeed that match the observed endurance.

    I noted that you and Oleksandr had some success in using aerodynamic equations to allow estimation of fuel flows at speeds in between holding and LRC. I believe Gysbreght showed a plot of those results matching the Aero Fuel Mileage curve roughly but not exactly. Therefore I have continued my fuel flow modeling to see if I might improve upon those results.

    My goal is to achieve ~1% FF accuracy compared to the Boeing tables for LRC, Holding, and also the Checkpoint tables, plus matching the Aero Figure 1 plot (if only we had other similar plots!). The Checkpoint tables are actually quite useful, despite the fact that the fuel is rounded to 100 kg and the times to 1 minute. I have found a way to fit all the tabulated points simultaneously and to remove the effect of most of the quantization errors. These checkpoint tables allow the existing Boeing LRC table to be extended to lower weight, speeds, and altitudes.

    I am still wanting to investigate further a fixed KIAS at a lower altitude (below FL250) to see if 1.5% PDA with a single speed can produce a MEFE at 00:17. We already know a combination of Holding and ECON can do this, but neither ECON nor Holding alone above FL250 can. Maybe a fixed KIAS can at a lower altitude, but then it will also have to match BTO/BFO.

    I am using some basic physics equations for density, lift, drag, etc. in developing a compact empirical model for B777/Trent fuel flow as a function of weight, altitude, and air speed. I’m not an aerodynamicist, and I am also trying to avoid arbitrary polynomials. Maybe there is an in-between model that can work. If I can get a close match to the Boeing data we have, I’ll post it and let Gysbreght and you (and anyone else who wants to) evaluate it.

    Mick,

    At an End of Route or at a Route Discontinuity, the FMC reverts to a constant heading. According to Honeywell, whether it is magnetic or true depends on the position of the NORM/TRUE switch on the console at the time of the discontinuity. As Don said, it does not depend on any MCP switch settings.

    @Oleksandr,

    I have also considered either zero frequency compensation or mis-timed frequency compensation in the AES as a possible cause of the 273 Hz BFO. Following my question on this subject to the ASTSB, they replied with information from Honeywell indicating that the frequency compensation term is required for the AES to log on and furthermore, there is no difference in the SDU software for doing this at the log-on request, at the log-on acknowledge, and at all the other normal transmissions. In other words, they explicitly say that the frequency compensation is always done the same way. In my view, this rules out both of our theories related to improper frequency compensation causing the 273 Hz BFO; those theories are not viable if you believe the manufacturer’s response to my question. I also think it is highly unlikely that intermittently bad ADIRU data caused the 273 Hz for the simple reason that BFOs just prior and shortly after appear unaffected.

    Victor had it exactly right when he said: “With this explanation, the BFO sequence at the 18:25 logon is explained without the need for invoking manoeuvers or intermittently incorrect navigational data.”

    Mike’s theory of OXCO thermal warmup is consistent with the observed frequency transient errors, both on the ground, and in flight. This effect must occur; about that there is no doubt. The only question to be answered was whether a repeatable thermal transient effect, depending only on starting OXCO temperature error and on the time since power was applied, could explain all the warm-up transient data. My paper on this subject demonstrated that this is indeed the case. The in-flight log on at 18:25 is unique in that it is the only one done in flight, and it demonstrates that the power was off for a considerable time beforehand. It is no mystery why the log on request BFO at 18:25:27 had a small frequency error, even though the one 8 seconds later had a much larger frequency error. When the power is off in flight for an extended period of time, the SDU becomes cold soaked. When power is applied, it takes a longer than usual (on the ground) amount of time to heat the OXCO oven to the operating temperature. During this extended oven warm-up period, the other necessary tasks are completed by the SDU so it is simply waiting for the oven to reach the operating temperature band before transmitting the log-on request. That happened at 18:25:27 when the temperature error was fairly small. The temperature control servo then overshoots the operating point, and this causes the next log-on acknowledge transmission to be made at a higher temperature and at a higher frequency (thus the 273 Hz). By the time the next transmissions at occur at 18:27, the temperature is settling down to the operating point, and the frequency errors are small and decreasing.

    You (Oleksandr) said: “Take a look at Fig 8 and 9. The 6 of 7 presented sequences have exhibited the decay trend, which included the first value. The 7th sequence, which is the logon in question, is the only one when the very first BFO value does not fit the decay trend. Basically Holland shows in these figures that the BFOs during the logon 18:25 do not fit the overall decay trend.”

    As explained above, the 18:25 log on is unique because the OXCO was cold. Yet it and all the ground-based log-ons match the same transient frequency curve once allowance is made for the starting OXCO temperature. See my Figure 3 in my paper here:

    https://drive.google.com/file/d/0BzOIIFNlx2aUUUVycGJLM21SX2c/view?usp=sharing

    You said: “Re: “The designers chose a simple start-up algorithm that produces acceptable performance.”. Why no overshot (sic) on the ground?”

    All the ground log-ons referenced by Holland always show overshoot (high temp = high BFO).

    You said: “Apparently you did not read the whole discussion. You know that nothing except a rapid descent or, alternatively, corrupted in some way BFO can explain this.”

    No “rapid descent” can possibly explain a +273 Hz BFO. Neither can “an ADIRU . . . in the process of (in flight) alignment”.

    You said: “I will not accept your and Mike’s theory until you or someone else is able to explain 18:25:27 BFO. Can you explain it or not?”

    I have already done so in February and March, if you bothered to read it. See my Figure 3 referenced above. It demonstrates that the 18:25:27 BFO is ~25 Hz low because of the thermal transient effect. Similarly, the 18:25:34 BFO is ~110 Hz high, the 18:27 BFOs are ~30 Hz high, and the 18:28 BFOs have negligible warm-up errors.

    @Dennis,
    You said: “I don’t have a personal bias. Why would I?”

    Only you know why you do. Calling other people who disagree with you “whacko”, as you do, appears to me to be based on personal bias. If not based on that, then based on what?

    You also said: “When ALSM says something is “very clear” when it is not very clear just annoys me greatly because it is bullshit.”

    Not everything you disagree with or do not understand is “bullshit.” You might try understanding it first, asking questions to clarify, and then offering constructive criticism or perhaps a different theory. Ideas which have no merit will fade away without invective. I did not initially accept Mike’s theory because I had only a vague notion of how the OXCO and SDU worked, and his explanation was only qualitative. My questions were responded to by Mike and by Don, who has been extremely helpful in this regard. Then I went off and studied the problem on my own, producing a frequency model that could be applied to the BFO data. Somewhat to my surprise, a straightforward servo model can explain quantitatively all the log-on BFO data. Furthermore, we know (not guess or assume) that frequency transient errors occur, and they appear related to the start-up OXCO temperature. That gives me confidence in the theory, and this theory then can be used to determine the impact of the warm-up effect on the 18:25-18:28 and the 00:19 BFO data (as I have done).

  713. Mick Gilbert says:

    @TBill

    As Don has pointed out I’m barking up the wrong tree on the FMS defaulting to anything other than CMH at an EOR/DISCONTINUITY error; it would appear that regardless of MCP HDG/TRK Reference Switch position the default mode is always CMH.

    Regarding “it has been hard to find any magnetic setting (heading vs. track) that meets the Inmarsat ping rings.”, am I correct in assuming that that would be where you quite logically invoke a constant air speed? I’ve wondered (but do not have the tools to explore) what a longer period of one engine operations than suggested by the ATSB (ie right engine flame-out much earlier than 0002 UTC) would do to some of the CMH scenarios; what does an FMT onto a CMH of 180° from near 7°3.00′ N 94°10.00′ E at around 19:23:00 UTC at M0.83 with the right engine flaming out at say 2330 UTC look like?

    Further, DrB has previously stated, “There is an alternative scenario that ends up in the same neighborhood. It uses Magnetic Track. I have noticed for a long time now that CMT and CTH routes have nearly equal curvature. … One should ask, why would a pilot set a CMT? The only possibility that appeals somewhat to me is if it were exactly 180.0 degrees CMT. …”.

    On the basis that the crew may well have selected the MCP HDG/TRK Reference Switch to TRK in response to an AFDS anomaly per the Boeing guidance, an FMT using CMT is not beyond the realm of possibilities. Using an albeit early version of Barry Martin’s flight path model, an FMT onto a CMT of 180° from near 7°3.00′ N 94°20.00′ E at around 19:23:00 UTC at M0.83 appears to produce a good fit to the BTO/BFO data (I think).

  714. DennisW says:

    @DrB

    Anyone who uses the description “very clear” is being careless with words, IMO. There is almost nothing that is very clear about this incident. Not long ago it was “very clear” how pilots like to fly airplanes which resulted in a bunch of pins in the map at 38S. Not long ago, the IG rejected the consideration of anything related to motive or causality. Well, obviously that approach did not work out very well.

    As far as the “whacko” description is concerned it has nothing whatever to do with disagreeing with me. The poster made an elaborate failure mode description of the accident and never even mentioned the simulator data. If he had even claimed that the simulator data was irrelevant for some reason I would be fine with that (even though I think an aircraft failure scenario is very unlikely). It is not opinions I am objecting to. It is the arrogance with which the opinions are presented, and the blatant exclusion of information that does not support those opinions.

  715. DrB says:

    @VictorI,

    On my brand-new-just-out-of-the-box HP Z book it takes about a minute to load the current thread over my slow 10 Mbps internet link. I suspect others who, like me, don’t have high download speed available, would benefit from a new thread.

  716. Mick Gilbert says:

    @Victor

    I couldn’t agree more, the radar data has been “problematic” from the get go. It has been akin to being given 220 pieces of 1,000 piece jigsaw puzzle to work on and sometime after you think you’re seeing something you learn that 110 pieces don’t even belong to that puzzle. I also agree with the back to basics approach of focusing on the FI PSR traces, as rudimentary as they are, as the best data available.

    With regards to @sk999’s digitized radar data from the RMP report, that trace is about 50 nm long, about three times the length of the 01:52:35 MYT PSR trace from the FI.

  717. MH says:

    @DrB- maybe try a difffent browser or actively use the down arrow or side scroll to force the page update.

  718. Mick Gilbert says:

    @Don Thompson

    Regarding the 9M-MRG incident, yes, you’re correct, it was a manual disconnect. I had it in the back of my mind that two envelope protections from opposite ends of the envelope (the A/P overspeed and stall protection) activating together would have forced an automatic autopilot disengagement but that wasn’t the case.

  719. Oleksandr says:

    @Don Thompson,

    You wrote: “In normal operation the SAARU provides a pass-thru, while performing a comparison check, of ADIRU data to the Centre bus.” (May 10, 2017 at 4:37 pm).

    Also previously you stated: “SAARU monitors the ADIRU, it tracks the ADIRU’s data as transmitted on the L & R A629 Flt Ctrl busses. When the ADIRU data is good, the SAARU simply passes that data on, along the C A629 Flt Ctrl bus. Should the SAARU detect ‘bad’ data from the ADIRU it transmits its own processed data on the C A629 bus.” (April 22, 2017 at 5:23 pm).

    I think you misunderstand how navigation system of a B777 works. The SAARU provides only reduced navigation data. It does not compute position and many other parameters (see, for example, FCOM, 10.20.15, or AMET Refresher Course). Therefore, the SAARU cannot in-principle monitor all ADIRU’s output, and replace with its own data if it detects ‘bad’ data. In fact, ADIRU itself is comprised of fault containment areas, and it is capable to assess its own ‘health’ status, and it does not require SAARU for this. SAARU’s function is not to monitor ADIRU, but to serve as the backup source of navigation data.

    Extract from FCOM (11.20.4):
    “The SAARU is a secondary source of critical flight data for displays, flight control systems, and other systems. If the ADIRU fails the SAARU automatically supplies attitude, heading, and air data.”

    The only thing, which is not explicitly stated here, is whether AES is among these other systems or not.

    You wrote: “GPS is not an alternate source for position data used within the avionics, the ADIRU is the source. GPS is used during alignment of the ADIRU and, during flight, as an input to the ADIRU’s process for reducing drift errors.”

    Compare this with, for example, FCOM 11.20.1:
    “If the ADIRU becomes inoperative during flight, the EICAS displays the message NAV ADIRU INERTIAL and the FMC uses only GPS data to navigate”.

    “If flight, the FMC position is continually updated from the GPS, navigation radios, and ADIRU”, FCOM 11.31.6.

    Am I missing something? Is FMC not “within the avionics”?

  720. Gysbreght says:

    @Don Thompson: “Does my most recent response to Oleksandr help?”
    No, it doesn’t. Im not thinking of an ADIRU fault. I’m thinking of the system operating as designed rejecting data that are outside the envelope for which it is designed. Along the lines of the A330 in AF447 rejecting AoA data when indicated airspeed dropped below 60 kt, and rejecting airspeed itself when less than 30 kt. (Label NCD is No Computed Data). In the 9M-MRG incident the minimum airspeed was 158.5 kts CAS and the attitude did not exceed 17.6º nose-up and 7.4º nose-down. What if those values were grossly exceeded in the MH370 high-acceleration manoeuvre?

    The ATSB Occurrence Report on the 9M-MRG upset is not a good investigation of the incident. The “Technical Analysis” describes the observed symptoms but fails to explain how a faulty accelerometer causes the airplane to behave as observed. That disinterest and lack of understanding is symptomatic of the ATSB’s and DSTG’s handling of the MH370 radar data, end-of-flight simulations and fuel available to the APU after main engine fuel exhaustion.

  721. Ge Rijn says:

    @Ikr @VictorI

    I may have my views and be persistent in finding indications to support them (who hasn’t?), but I still try look objectively to the information and data at hand within my limitations.

    I understand the ‘BG’-argument and other arguments but when only taking the drift-data from this latest CSIRO-model it shows predicting facts that are at least not conflicting the facts of reality concerning the debris finds.
    On the contrary I would say.
    It shows a relatively small debris field is sufficient to give the results it shows and are in line with the current reality.
    A big debris field of thousands of drifter-like pieces is not in line with current reality for then a multiple amount of pieces should have arrived and more than 27 found by now.

    I took a different approach and took a look at the situation on the end of the time line at 26-6-2016.
    I took the ~290 drifters I counted to start with (it’s more about relation). The following ~numbers show up:

    30S flap: 240 landings on African coasts 0 on WA.
    30S non-flap-dp: 265 landings on African coasts 0 on WA
    30S non-flap-2dp: 215 landings on African coasts 5 on WA

    35S flap: 240 landings on African coasts 9 on WA
    35S non-flap-dp: 235 landings on African coasts 28 on WA
    35S non-flap-2dp: 110 landings on African coasts 14 on WA

    Notable is from 35S quite some landings have by then also taken place on the Maldives and Indonesian shores.
    Taken into account we are almost a year past 6-2016 the numbers above would most probably a bit higher by now.

    IMO this CSRIRO-model predicts based on ~290 drifter-like pieces to start with an average of at least:

    An average of ~240 pieces should have landed on African shores and ~2 on WA shores by 26-6-2016 coming from 30S.

    An average of ~195 pieces should have landed on African shores and ~17 on WA shores by 26-6-2016 coming from 35S.
    Noted that around 10 pieces landed by then on the Maldives and Indonesian shores.

    Concluding based on this model with a starting drifter-like debris field of around 300 pieces there will be a minimum of ~200 pieces still lying somewhere on African shores since 6-2016.
    And ~17 when coming from 35S and ~2 coming from 30S on WA shores.

    27 found so far ~200 still not found after more than 3 years.

  722. Victor Iannello says:

    DrB: I am sorry about the loss of a family member. Nothing here is time-critical. Family first.

  723. Ge Rijn says:

    I like to add that I also read every comment looking for and thinking over arguments and data that conclusively exclude a relatively low speed, low AoA impact resulting in a relatively small debris field.

    I still haven’t seen them convincingly.
    The high speed dive and impact is only based on the final BFO’s and a dubious small amount of simulations as far as I can see.

    Still the BFO’s are under fierce debat although there seems to form concensus on the ‘off’ 18:25 BFO’s being caused not by descent or something else but by a depowered OXCO and SDU for at least 35 minutes.

    Still nothing tells me that the similar ‘off’ 0:19 BFO’s could not have a similar cause or another than a high speed dive.
    I hope someone can give a clear explanation why the ‘off’ 0:19 BFO’s can only be caused by a high speed descent and not by something else (f.i. like the 18:25 BFO’s)

  724. Victor Iannello says:

    @Ge Rijn: The data and analysis presented in Ian Holland’s paper suggests that the warm-up transient produces abnormally high BFO values. On the other hand, the values at 00:19 are less than what is expected for level flight.

    If you propose that the BFO values at 00:19 are not consistent with a high speed descent, then you need to explain why you believe that Ian Holland’s paper is incorrect.

  725. David says:

    @Ge Rijn. “I hope someone can give a clear explanation why the ‘off’ 0:19 BFO’s can only be caused by a high speed descent and not by something else (f.i. like the 18:25 BFO’s)”

    Pages 8-12:
    https://www.atsb.gov.au/media/5771939/ae-2014-054_mh370-search-and-debris-update_2nov-2016_v2.pdf

  726. Don Thompson says:

    Oleksandr wrote: “Am I missing something?

    Yes, I suggest you are. You are missing how redundancy is ensured by the ADIRS, and that the FMC has ten, discrete, navigation update modes exploiting the ADIRU, GPS, and radios in various combinations. The FMC computes, and tracks, the position uncertainty of each source and selects a source on the basis of that uncertainty.

    Your position is that the operation of the ADIRS may have been comprised and as a consequence, somehow, GPS derived data may have been substituted in the data set transmitted to the AES. I disagree, there is no basis for conjecture that multiple data sources may be exploited by the AES. GPS position data from the MMRs is interfaced directly via A429 connects to the FMCF partition within AIMS (and also to the GPWS).

    @Gysbreght asked, “In the case of MH370, could the airspeed or attitude have reached a value where it was considered invalid?

    I will defer to an eventual DFDR read-out on that question.

  727. DennisW says:

    @David

    A cut-paste from you link above.

    begin//
    A statistical analysis of the BFO error from all the 20 previous flights of 9M-MRO identified that the distribution was approximately Gaussian (see DST Group book – link above) with a standard deviation of 4.3 Hz. ±3 standard deviations (12.9 Hz) is a conservative choice for the error.
    end//

    Figure 5.4 of the DSTG Book can also be referenced which shows why a statistical analysis of BFO data is essentially meaningless, and is the reason why Allan Variance is used to describe oscillator behavior not mean and variance relative to a normal distribution. While that does not really impact the conclusion the ATSB is drawing relative to the BFO values at 00:19, the inclusion of a statement like that undermines their credibility.

    Figure 5.4 also casts doubt on the error bars in the ATSB BFO graphics. Again not particularly relevant to what the ATSB is saying for the observables art 00:19, just an inconsistency in published data However, it is relevant to the early work done by many of the people here who correctly determined that straight paths to 38S were the only solutions with low single digit BFO error that also satisfied the BTO constraints.

    When I reflect on the events that got us to where we are today, the biggest disconnect that I see is the formation of the SSWG itself. Why not simply contract with an organization such as Metron who has a pedigree in analytics of the type needed for this investigation? Not saying the SSWG is not comprised of a lot of very smart people. They simply don’t have a track record either individually or as an organization in forensics of the type needed for this incident. My guess is that Metron would have told the ASTB that the location of the aircraft cannot be defined with sufficient precision to undertake an underwater search given the data available. That would have delayed the start of the search until the debris we now have could be added to the forensic evidence. A search further North might then be justified and undertaken, and the aircraft might have been found.

    Now the use of satcom and drift analytics are too late with too little. The people writing checks are not going to be swayed by the DSTG, or DrB, or Dr Iannello, or anyone else using these methods.

  728. Gysbreght says:

    @DrB: I’m sorry to hear about the loss in your family.

    You wrote: “I believe Gysbreght showed a plot of those results matching the Aero Fuel Mileage curve roughly but not exactly.”

    I did more than that. I provided the TSFC model that is a complement to Professor Obert’s CD curves.

    “I am using some basic physics equations for density, lift, drag, etc. in developing a compact empirical model for B777/Trent fuel flow as a function of weight, altitude, and air speed.”
    Without a model of engine fuel efficiency as a function of thrust and airspeed it’s hard to see how your basic physics equations can get you any further.

  729. Ge Rijn says:

    @David

    Thanks for the link. I had a closer look. It all seems sound and well but I keep wondering about certain aspects I probably don’t understand.

    First of them is the predicted BFO at 0:19:29 of 260Hz and the measured BFO of 182Hz.
    While the final ‘ping’-BFO at 0:11 is just above the 250Hz line I think there is no way to know for certain when the plane started descending after 0:11.
    I mean the earlier the descent started after 0:11 the lower the descent rate would be reaching 182Hz at 0:19:29 I’m thinking?

    And then at 0:19:37 they take again the predicted BFO of 260Hz.
    Why not only the measured BFO of 0:19:29 of 182Hz compared with the -2Hz BFO instead and derive the descent rate from there (182Hz)?
    What would be the descent rate in those 8 seconds?
    Just thinking loud now..

  730. buyerninety says:

    @Victor, Andrew
    Thankyou for your clarifications/comments regarding radio navigation.

    @Mick , Oleksandr
    Victor said;
    “here is the radar data near Penang that @sk999 has digitized from the radar data presented
    graphically in the RMP report….I have no idea how accurate these data are…If you or
    anybody else has better data, or has reason to believe these data are in error, it would
    be helpful to know.”

    Here is what I see in the RMP report radar points graphic;
    Modern Police forces are able to determine from mobile phone cell tower network data, a distance
    to a mobile phone (to an accuracy of something like approximately 5 to 10 metres, if intervening
    structures are not interfering with the mobile signal – which should not be a factor for MH370.)

    The map graphic at appendix K-2, of MH370’s radar detections (@about 1 minute intervals) as it
    transited south of Penang, also displays an additional flagged point named ‘Fariq phone detected’.
    The position could be reasoned as having been sited on the ‘radar positions’ track, by the RMP
    having taken into consideration the time of the phone registration and by the RMP having derived
    the distance the phone was from the tower. (The actual known distance, however many XXXXX feet it
    is from the tower, is not necessary for us to know, for the purposes of this discussion).

    Unfortunately, at the time the graphic was produced, the RMP apparently believed that MH370 was
    at altitudes of 44700 feet and therefore represented the known distance to the phone, as being
    (an invisible line) laid out from the tower antenna up to a point at 44700 feet altitude.
    This represented phone location on the graphic is noticeably inside the arc or line of the radar
    points, and not in line with the radar points.
    I would suggest the phone location should alternatively be considered as still being the same
    known distance, but represented as (an invisible line) from the tower antenna up to a point of,
    say, 35000 feet altitude. This results in the new phone point appearing (from overhead) to be
    located further outwards (away) from the tower – in a rough approximation, the new phone point
    is now about +10% further away from the tower, which locates the new phone point on, or slightly
    outside, the arc of the radar points, but more generally conforms to the line the radar points
    make.

    We could consider that the line of radar points may be somewhat inaccurate, and further we
    could move the arc of radar points slightly outwards so the new phone point smoothly conforms to
    the line of the radar points.
    Interestingly Mick, if we note that Victor has previously stated he believes the radar points
    seem “to be at 7 DME {where} as the recorded path crosses the runway course” {glideslope}, & then
    if we note where our new ‘slightly moved outwards’ line of radar points crosses the glideslope,
    that brings us about onto the 7.8DME (Nm) position on that glideslope, (which Mick has remarked)
    is the location of a waypoint fix (as marked on certain Penang approach charts).

    This is all somewhat rough figuring, of course.
    If you attempt to measure distances on the appendix K-2 graphic, be aware that the scale marked
    on that graphic seems to be slightly incorrect, and you may find it helpful to determine your
    own scale, by measuring between two ‘prominent features’ on the graphic and relating the distance
    between the features to some other map of your choice of Penang. (As an example, Google Earth, at
    a scale of 5km, may match the distances on the graphic if you set the graphic pdf to a zoom size
    of ~89%.)(As an example, two ‘prominent features’ to use to measure between, could be the tip of
    the SouthWest, and the tip of the SouthEast, promentaries of Penang Island.)

    In the discussion above, I substituted an altitude of 35000 feet instead of 44700 feet. I believe
    that presently, most readers would regard MH370 as having been within the 28000 to 37000 feet
    range of altitudes ( – and a possibly later handwritten note addition to Appendix J-1, ‘CREW’S
    ACTIVITIES IN A TIMELINE’, page 6 of 8, translated as ‘radar Flag Hill (army) 35100 feet‘,
    suggests that the RMP, at a time later than when the K-2 graphic was produced, also came to
    realize that MH370 flew lower than those figures of 44700 feet that they were earlier informed of). of).
    Cheers

  731. TBill says:

    @Ge Rijn
    I have not yet closely focus on very end of flight BFO, but for most of my paths I like a optional turn to due East. That could also imply rapid descent if there is a planned landing zone. I hold open options or attempt to minimize debris but maybe not by slow landing.

    @VictorI
    I am also using slow internet connection lately and now see the problem. But I get faster connection in some locations.

  732. buyerninety says:

    @Ge Rijn
    I didn’t forget you, but Davids posts seemed to cover most of your questions,
    except that although damage ‘in tension’ could be caused by contact with aircraft
    parts in glide, there does not seem to be much ‘in compression’ damage to the
    composite fibres (except the front wheel door? wasn’t it, that did exhibit ‘in
    compression’ damage.)
    Cheers

  733. DennisW says:

    @buyer90

    “Modern Police forces are able to determine from mobile phone cell tower network data, a distance
    to a mobile phone (to an accuracy of something like approximately 5 to 10 metres, if intervening
    structures are not interfering with the mobile signal – which should not be a factor for MH370.)”

    Really??

    I would challenge you to provide a reference for that claim. My wording here takes into account the recent criticism I have received for my use of language on this site by DrB and others.

  734. Oleksandr says:

    @Bobby,

    I am sorry to hear that. Perhaps it is better to postpone our discussion. Family is more important. Take care.

  735. Oleksandr says:

    @Don Thompson,

    You wrote (May 11, 2017 at 9:02 am): “Yes, I suggest you are. You are missing how redundancy is ensured by the ADIRS, and that the FMC has ten, discrete, navigation update modes exploiting the ADIRU, GPS, and radios in various combinations. The FMC computes, and tracks, the position uncertainty of each source and selects a source on the basis of that uncertainty. ”

    Citation from your earlier post May 10, 2017 at 4:37 pm:
    “GPS is not an alternate source for position data used within the avionics, the ADIRU is the source. GPS is used during alignment of the ADIRU and, during flight, as an input to the ADIRU’s process for reducing drift errors.”

    Citation from FCOM 11.20.1 I posted earlier:
    “If the ADIRU becomes inoperative during flight, the EICAS displays the message NAV ADIRU INERTIAL and the FMC uses only GPS data to navigate”

    I am not sure what you tried to say. If you said something wrong, please admit it. If you insist that FMC is not avionics, please state this explicitly. The AES has nothing to do with ADIRS redundancy. As long as position and attitude data are available, I don’t see a reason why would the AES not be able to use this data. I admit it may ‘listen’ to the ADIRU only, and discard any data available from GPS and SAARU in any circumstances. There is such a possibility. However, I don’t see any reason, because it would mean, at least, that in the polar regions SATCOM may be inoperative for more than 15 minutes after power up. Also, it would mean that in case of the ADIRU failure over the ocean, the crew would not have means of communication.

    Please state this explicitly and provide any reference that the AES does not use position data from GPS even if the ADIRU is inoperative.

  736. paul smithson says:

    Inmarsat geo doesn’t have coverage in polar regions anyway.

  737. Don Thompson says:

    Oleksandr,

    I generalised in order to focus on the importance of ADIRS and its inherent redundancy.

    The ADIRS provides a highly redundant and autonomous inertial reference for the B777, once initialised on the ground the aircraft can operate satisfactorily without external augmentation. However, GPS does augment the ADIRS in the FMCs navigation function.

    Considering your concerns, a) that ADIRU initialisation takes longer near the poles: that is of little consequence. If ADIRU alignment is necessary, time has to be allowed, the aircraft will not move until the ADIRU aligns. (Remember, at high latitudes geostationary SATCOM operation isn’t possible.) And b), for redundancy in communication over oceanic regions: 9M-MRO was equipped with two HF radios, in addition to the AES.

    My sources include the B777 AMM, B777 Training Manual, ARINC 429P1 (which includes data word formats, label assignments, & data sources).

    I have no issue with your citation from the FCOM, 11.20.1: that the quoted EICAS message is displayed and that the FMC has GPS data available. As I described the MMR-GPS data outputs are interconnected over A429 connection directly to the FMCF partition in AIMS (and the GPWS). Note: directly, point to point not via the A629 busses.

    The AES is directly connected to the DCGF hosted on AIMS, the DCGF translates ADIRU data from the A629 Flight Control busses and forwards the IRS data via A429 to the AES.

    The NAVIGATION volume of the AMM, in chapter 34-61-00 includes a section titled ‘FMCS – FMCF INTERFACES – 1’, and provides an Output Overview and Output Details (schematic summary): the AES is not listed.

    That is all for now.

  738. Don Thompson says:

    @Paul,

    High latitudes: indeed, no footprint of coverage from Inmarsat’s geostationary orbits.

  739. Oleksandr says:

    @Don Thompson,

    Re: “As I described the MMR-GPS data outputs are interconnected over A429 connection directly to the FMCF partition in AIMS (and the GPWS). Note: directly, point to point not via the A629 busses.”

    AMET B777 Refresher presents a diagram, according to which both the GPS communicate with the ADIRU via AIMS and ARINC 629 Bus 3, schematically:

    L/R GPS Antenna -> L/R MMR AIMS ARINC 629 Bus 3 ADIRU.

    In other words ARINC 629 still receives data from GPS.

    Re: “The NAVIGATION volume of the AMM, in chapter 34-61-00 includes a section titled ‘FMCS – FMCF INTERFACES – 1’, and provides an Output Overview and Output Details (schematic summary): the AES is not listed.”

    The fact that the AES is not listed does not mean anything. In the case of 9M-MRG a “detailed” ATSB’s report did not include the whole list of ADIRU’s output parameters. You will not find the magnetic heading in that report, for example.

  740. Andrew says:

    @Oleksandr

    And if all that’s not enough to convince you, in my experience as a pilot operating the B777, the SATCOM is not available for use until ADIRU alignment is complete.

  741. Andrew says:

    @Oleksandr

    “In other words ARINC 629 still receives data from GPS.”

    That’s somewhat irrelevant to the AES, because it uses ARINC 429 data provided by the AIMS data conversion gateway function (DCGF). The DCGF converts inertial reference data from the ADIRU into ARINC 429 data for high gain antenna beam steering. The AES does not receive ARINC 629 data.

  742. Mick Gilbert says:

    @Victor
    @Don Thompson

    A general question regarding the PSR data please; what level of accuracy should we reasonably expect for that data. I’m as guilty as the next bloke for drawing razor-sharp tracks for the PSR traces associated with the transit across the Malay Peninsula, realistically what sort of margin of error should I incorporating?

  743. David says:

    @DennisW. “Why not simply contract with an organization such as Metron who has a pedigree in analytics of the type needed for this investigation? Not saying the SSWG is not comprised of a lot of very smart people. They simply don’t have a track record either individually or as an organization in forensics of the type needed for this incident”.

    About the people involved, according to the ATSB June 2014 report (updated July 2015), in the period March-April 2014, “…the Australian Maritime Safety Authority (AMSA) and the ATSB jointly determined a search area strategy correlating information from a Joint Investigation Team (JIT2) located in Malaysia and other government and academic sources” and, “The Joint Investigation Team comprised specialists from Malaysia, China, US, UK and France”. The search moved to its underwater phase on 28th April 2014 and a tender for a search was released on 4th June 2014.

    From the same report, “Since May 2014, a search strategy group, coordinated by the ATSB, has been working towards defining the most probable position of the aircraft at the time of the last satellite communications at 0019. The group brought together satellite and aircraft specialists from the following organisations:
    • Air Accidents Investigation Branch (UK)
    • Boeing (US)
    • Defence Science and Technology Organisation (Australia)
    • Department of Civil Aviation (Malaysia)
    • Inmarsat (UK)
    • National Transportation Safety Board (US)
    • Thales (UK)

    These worthies surely would be familiar with the expertise available worldwide and with previous experience, including Metron’s. In the June 2014 ATSB report p34 there is a footnote about the AF447 search, “Metron Scientific Solutions Report: Search Analysis for the Location of the AF447 Underwater Wreckage 20 January 2011″.

    When you say, “They simply don’t have a track record either individually or as an organization in forensics of the type needed for this incident”, what is the qualitative difference between that group and those settling what to do about AF447, noting the SSWG had the added AF447 experience before it?

    Possibly Metron was consulted? And what if they had indeed said search success probabilities were low? Do you stick around in the hope something turns up, like, maybe, wreckage retrieval, a couple of years later? How do you demonstrate the probability there will be gain from doing so that which is worthy of NOK patience and potential delay in cause discovery; the wreckage possibly sinking into sludge and its deterioration otherwise it prejudicing the evidence it might yield?

  744. TBill says:

    @Mick Gilbert
    As far as your early engine flame-out scenario, I agree in principle that the Arc6 is less certain due to the end of flight situation. So I am not looking for exact Arc6 match. If I see a good match, I’ll take it though.

    I feel fairly strongly there could have been speed/minor heading changes at least up Arc3. So I am only holding Arc3-4 leg and Arc4-5 leg as no changes right now.

  745. Victor Iannello says:

    @ventus45: I don’t have permission to view the figures on that site, so I am only guessing about your hypothesis. If you are proposing that the radar data after 18:02 is non-existent, I think that is a possibility. If you are proposing that the plane continued flying southwest after Penang, later turned to the south, and ultimately crashed near 41S latitude, I think neither the BFO data at 18:25 nor the drift models support this.

  746. Victor Iannello says:

    @Mick Gilbert: Don has studied the radar hardware much more than me. I’ll let him respond, if he chooses.

  747. DennisW says:

    @David

    There is no substitute for pedigree. The organizations you site failed (miserably), plain and simple. I have this childlike mentality of ranking successful endeavors higher than endeavors that failed. My bonus always worked that way.

    Having said the above. It is useful to go back and ask about the problem statement itself.

    1> If the ATSB was asking “we are going to start an underwater search, where should we begin?” Then the IG end points were the best choice (given that they did not have the data from previous flights).

    2> If the ATSB was asking “should we start at search based on the data?”. Then the answer is no, IMO, and I truly believe this is the answer they would have received from Metron.

  748. DennisW says:

    @David

    …more

    You can go back to posts on the Wise and Duncan sites and revive arguments I had with Victor and others. My position was that initiating an underwater search “based on the spreadsheets of a bunch of geeks” was an incredibly stupid thing to do. I stand by those statements. The IG, DrB, and the majority of the SSWG are scientists. You don’t let these people make spending decisions. I am a geek myself by training, but running a several hundred million P&L taught me to take a realistic path. Again, my bonus depended on it.

  749. Andrew says:

    @DennisW

    That might be true if it was a business decision, but it wasn’t and I somehow doubt it was the ‘geeks’ that made the decision to launch the search. Would the politics have allowed them to sit back and do nothing?

  750. DennisW says:

    @Andrew

    Every decision is complicated – science, politics, …

    At the end of the day intelligent people do what makes sense. The decision to start an underwater search at the time was started was just plain stupid.

  751. ventus45 says:

    @Victor

    There does seem to be some issue with AuntyPru that precludes some people from viewing it that I do not understand.

    In any case, I have just finished transferring all those posts above to my blog. Try this: http://ventus45.blogspot.com.au/2017/

  752. DennisW says:

    @Andrew

    “That might be true if it was a business decision, but it wasn’t and I somehow doubt it was the ‘geeks’ that made the decision to launch the search. Would the politics have allowed them to sit back and do nothing?”

    I am not an Aussie taxpayer. I am just pointing out stupidity.

  753. TBill says:

    @ventus45
    I think you have to have Login ID on with AuntyPru to see the graphics, which is why I joined a few months ago…

  754. ventus45 says:

    @TBill
    That would explain it.

    @Andrew &@DennisW
    AF447 set the benchmark in media/public expectations. The politicians could not have said “too hard – no way” and got away with it.

  755. Andrew says:

    @ventus45

    “The politicians could not have said “too hard – no way” and got away with it.”

    I agree.

  756. Ge Rijn says:

    @buyerninety

    Thanks for your comment: “except that although damage ‘in tension’ could be caused by contact with aircraft parts in glide, there does not seem to be much ‘in compression’ damage to the composite fibres”.

    I meant with my earlier comment on the subject of damage that pieces could have been damaged and ripped off with mostly tension-failure by water also in a ditch-like scenario with low vertical speed and relatively low horizontal speed.
    Only the speed necessary horizontally through the water surface would be very much lower than the speed necessary though air.

    A low vertical impact speed, low AoA and relatively low horizontal speed on the water surface (ditch-like scenario) could IMO explain the small amount of compression damage and all other observed factors (I won’t repeat them..) better then caused by high speed air damage and seperation during a dive.

    Comparable damage and seperation is only shown by the relatively low speed crashes (landings) of Asiana 214 and the Hudson-ditch.
    Not by the the high speed dives of SilkAir and ChinaAir 747.
    In both latter cases no trailing edge wing parts seperated during the dive.

    And I would like to mention we still have no detailed information on the two most important pieces regarding compression and tension failure: the flaperon and the outboard flap section.
    I would sure like to see what kind of tension and compression failures the broken trailing edges of those pieces show.

    I think if those pieces show mostly tension failure on the bottom side of their broken trailing edges and more compression damage on their top sides we could have a conclusive answer.

  757. David says:

    @Ge Rijn. “Just thinking loud now”.

    A descent post 0:11 and prior to the two final BFOs would be undetected and would reduce the net descent rate increment of both, surely, though not the 8-secs-apart difference between them of a nominal 10,800fpm. In the extreme, the aircraft before 0:19:29 might have been descending at a rate equal to or even greater than one consistent with that BFO, though I think that assumes an active pilot.

    As to this ‘nominal’ 10,800fpm, that is a sum of the effects of;
    • BFO tolerance, which for this aircraft was taken to vary by ±12.9Hz but which over these 8 secs was assumed constant. I presume there are grounds for assuming that constancy, a subject beyond me. However were that to change transmission to transmission from one extreme to the other the descent rate difference would be 1530fpm,
    • and then there is direction of flight. Since the aircraft could well be in a spiral with high turn rate, the assumption of that being constant over those 8 secs is likely to be errant. That would be joined by;
    • ground speed. I see from fig3 that that can have an uncompensated effect, its extent depending on direction. My (unshared) opinion that the descent rate increase over those 8 seconds had a large nosing-down component would alter groundspeed if true, airspeed constant. Airspeed in fact would increase some but that is unlikely to be in proportion. Outcome: groundspeed would slow with a nose down, IMO. Even so, from their fig3 my guess is that the combined effect of direction and ground speed change would not amount to much; a few Hz and maybe 200fpm.
    • Finally there is OCXO warm up. I have not followed that closely but I gather there would be no transmission when not within frequency limits. I assume that the designed-in Doppler compensation for satellite separation/closing is to keep transmission bandwidth narrow. But with this OCXO drift that raises speculation in my head that were drift frequency change not limited, the SDU could waste its efforts transmitting outside receiver frequency range. The extreme of their assessed drift is 130Hz, or 7650fpm. Because I have confidence they know what they are about I raise this just as part of my, ‘just thinking aloud now’, like yours.
    Incidentally to employ use of Mick’s word ‘delicious’, fig3 in its heading says it is not to scale – yet there are no scales.
    However I assume that were there, we would see that there is little drift change over 8secs.

    What all this means though is that the 10,800 fpm might itself vary some.

    Returning to those simulations and any which displayed consistency with the final BFOs, aside from quite a variable descent rate range they could fall within, there was this 10,800fpm, which was a fixture within that Report’s context. Thence if there had been a 10,800fpm difference observed, nothing much more nor less, the question then was whether this was consistent with the timing of earlier AC power loss: not much good if outside that. As I posted recently, even the ATSB assessment as to when this occurred depends on “approximate” times for APU start and SDU reboot, leaving the whole consistency question somewhat woolly.

    While leaving prior descent aside, putting the rest of the above together suggests that the ATSB section indicating that simulations meet, simultaneously, BFO descent rate criteria, descent rate increase criteria – and timings – needs expansion.

  758. David says:

    @Ge Rijn. The figure without scales as you will see is 4.

  759. Don Thompson says:

    @Oleksandr,

    You asked for sources, I listed relevant information, you are contradicting Boeing published sources. I’m done.

    @Ventus(?)

    Nope, I’ll pass on engaging with your Str of Malacca surveillance musings.

    @Mick

    Radar errors: short answer, it depends. I have seen no published ‘tolerances’ for the systems utilised in the region. Fundamental factors…

    The Kota Bharu & Butterworth terminal area PSRs give no altitude discrimination, a corrected ground position may be considered from slant range if an altitude can be assumed. Transmitter power and antenna tilt would guide an estimate of the detection envelope, we don’t have any idea for the latter. A number of contributors made slant range corrections that resulted in a ground track straightened from the ‘bump’ that is apparent in the FI’s playback images of 9M-MRO’s track in the vicinity of DCA’s Kota Bharu PSR.

    The military ‘3D’ PSRs, the Penang Hill RAT31-DL and the Bukit Puteri Martello, exploit beam forming by a phased array antenna to discriminate altitude. The ‘physics’ of the beam forming technique implies, that at short range, interference lobes will exist and these result in a source of error for altitude discrimination.

    :Don

  760. Ge Rijn says:

    @David

    Thanks for your explanation and view which gives me further thoughts about those BFO’s etc..

    I conclude from your comment the predicted 260Hz at 0:19 is a virtual number and an unknown for it’s not known where the plane started descending after ( or maybe even sometime before?) 0:11.
    Then the 0:11 BFO of ~250Hz is the last actualy measured BFO before 0:19:29 isn’t it? Making some basic calculations which probably have a basic failure I don’t understand yet..

    When taking the 0:11 BFO as a starting point and using only nominal numbers there would be 250-182=68Hz of difference till 0:19:29.
    When taking 1.7Hz=100ft/min descent rate this would make 68:1.7=40Hz till 0:19:29 which would make 100ft/min.x40=4000ft descent from 0:11 till 0:19:29 (?).

    Now assuming the plane was on 34.000ft altitude at 0:11 then by 0:19:29 it would have been on 30.000ft altitude after which the steep descent started leading to the steep 10.800ft/min descent rate during those 8 sec. till 0:19:37.

    Would you agree there is no way of knowing how long this 10.800ft/min descent rate lastet after 0:19:37?
    An could you then agree there could have been more then enough altitude left to level out in some way (phugoid,or controlled)?

  761. Ge Rijn says:

    @TBill

    Would an optional turn and descent to the East after 0:11 affect the 0:19 BFO’s much compared to the measured North and South examples used in the report? I guess not but I wonder.

    The final BFO’s indicate a steep descent took place between 0:11 and 0:19:37 that’s clear. To me how long that steep descent lasted and till what altitude is not known. I also think it could be part of a controlled descent for I still see no reason why the plane must have been not pilot cotrolled till the end.

    Than the report also mentions several simulations in which the plane stayed airborne for 20 minutes.
    As far as I know there is only one example with a similar plane (Airbus 330) that almost did the same.
    Air Transat flight 236 stayed airborne for 19 minutes after loss of both engines. But this was a controlled glide.

    I wonder how the B777 in those simulations stayed airborne for 20 minutes uncontrolled by a pilot.

  762. Mick Gilbert says:

    @Don Thompson

    Thank you for your post on the radars, Don. I wasn’t aware that the Kota Bharu and Butterworth PSRs give no altitude discrimination; that perhaps explains both the unrealistically high altitude and very high ground speeds recorded for the trace to the south of Penang.

  763. Oleksandr says:

    @Don Thompson,

    “…you are contradicting Boeing published sources”

    I am not, but you are. I do not claim that my understanding is correct, but you have already demonstrated that your understanding of the B777 ADIRS is wrong: you statements often conflict with FCOM and other literature.

    The source of the diagram you claimed to be wrong this time is here, 7-6.

  764. Oleksandr says:

    For some mysterious reason the link above appended to Victor’s blog hyperlink. The correct link is:

    http://www.aviationlearning.net/files/B777%20REFRESHER%20GE%20RR.pdf

  765. Oleksandr says:

    @Andrew,

    “That’s somewhat irrelevant to the AES, because it uses ARINC 429 data provided by the AIMS data conversion gateway function (DCGF). The DCGF converts inertial reference data from the ADIRU into ARINC 429 data for high gain antenna beam steering. The AES does not receive ARINC 629 data.”

    My point was that the GPS data can reach ARINC 429 through the AIMS. Don stated this contradicts to Boeing sources. The DCGF could in theory ‘capture’ this data, or the ADIRU could relay it back to the ARINC 429, when it identifies itself as inoperative, either due to the alignment in progress, or when it detects internal failure. This is my speculation, which might be incorrect.

    You wrote: “in my experience as a pilot operating the B777, the SATCOM is not available for use until ADIRU alignment is complete.”

    That is very important aspect. Do you observe this always or sometimes?According to FCOM: “the bulletin issued by Boeing (D632W001-TBC, No. TBC-81R3), states that there are known cases of ADIRU heading error anomaly, when, in particular, SATCOM, TCAS and ACARS may become inoperative in-flight, or their functionality can be degraded.”

    What is a reason why SATCOM is not available during the alignment, in your opinion? I may suggest several reasons:
    (a). The AES receives a message that the ADIRU is being aligned, and does not proceed.
    (b). The AES does not receive nav data, and it does not proceed.
    (c). The AES proceeds, but the antennas may not be properly steered.

  766. David says:

    @Ge Rijn. What I was describing in answer to your speculation was how with a pilot the aircraft could have replicated the BFOs: they do not define flight since 0:11.

    The 182 Hz was real and the ATSB described how the descents could have resulted from 260hz estimated for 0:19:19. Had it been descending much at the 6th arc that would have been apparent from its BFOs, again allowing the ATSB underpinnings of flight direction etc.

    In talking about the 0:11 frequency remaining static to the 7th arc you overlook the about-linear increase with time they have plotted and you lose me there. That increase is 32 hz/hr. (As to why that would be linear I cannot say. Others might. Going backwards though, if linear it would cross the zero Hz time line at about 1620).

    The 68 Hz you mention would not take account of that linear change to 7th. I suspect you might need to look into the Holland paper for a deeper take on this. Mind you I am unsure I understand you fully here.

    “Would you agree there is no way of knowing how long this 10,800ft/min descent rate lasted after 0:19:37? And could you then agree there could have been more than enough altitude left to level out in some way (phugoid,or controlled)?”

    If a pilot had been in a 4000fpm descent before 0:19:29 he would then be at about the minimum needed for consistency with that BFO and would need to increase that by the nominal 10,800 fpm over the next 8 secs for consistency with the 0:19:27. For my part I would expect that from the high altitude you mention he could pull out from this and even the maximum descent rate and yes there is no way of knowing how long the high descent rate lasted.

    However the altitude of the 7th arc log-on is unknown, except as might be gained from BFO-consistent simulations.

    I think it possible that at high speed the aircraft would at least start to pull out of a steep dive without a pilot though I doubt it would if in a steep spiral.

  767. Victor Iannello says:

    @Oleksandr said, “For some mysterious reason the link above appended to Victor’s blog hyperlink.”

    You didn’t include “http://” in the URL. Mystery solved.

  768. Victor Iannello says:

    @ventus45: After reading your blog post, I stand by my previous comments.

  769. Oleksandr says:

    @Victor,

    Thanks for correcting the link. Probably I omitted “http” when I copied/pasted it.

  770. TBill says:

    @Ge Rijn
    Have to get back later on that…

  771. Don Thompson says:

    @Oleksandr

    Thank you for referring me to the Airline Maintenance and Engineering Training company’s B777 Refresher Course Technical Training Manual.

    I think I understand where the ‘disconnect’ arises.

    The AMET TTM comprises 234 pages to cover the entire aircraft. The Boeing 777 Aircraft Maintenance Manual, ATA Chapter 34 Navigation, extends to 877 pages and 111 of those pages are concerned solely with the ADIRS.

    I will make a precis of the specific AMM sections to expand on the information you have referenced in the AMET TTM, it’ll likely be next week before I can prepare that.

    :Don

  772. Don Thompson says:

    @Mick

    Radar plot around south of Penang: I assume you refer that accompanying the MCMC’s analysis of the cellphone LBS registration? I very much doubt that plot derived from the Butterworth air traffic PSR.

  773. Mick Gilbert says:

    @Don Thompson

    G’day Don, yes, I was refering to the altitude and speed data shown in Diagram 3: Area Covered by Sector 2 of BBFARLIM2 Base Station and MH370 Flight Path. Where do you think they’ve got that data from if not from the Butterworth PSR?

  774. David says:

    @GeRijn. Your last BFOs’ posting revisited.

    The actual and linear plot of BFOs at Fig5 for 0:11 are both near enough 250Hz. Fig 1 displays the BFO error bars of ±13Hz applied to all. Since there is no OCXO drift at 0:11 and a southerly course, so no correction from those, the ascent or descent range within which the 250 applies is ±13*100/1.7 or ±760fpm. This might be biased a little by the 2-3 Hz residual for a southerly course of Fig 3, say 150fpm. So at 0:11 the aircraft could have been descending at 900fpm as I understand it and still been measured at 250Hz, as with an ascent of 600fpm.

    Translating that across to 0:19, the effects there of the possible northerly course and of OCXO drift need to be added for Case B, resulting in the two Table 2 extremes.

    I think you might be mixing descents with descent rates with the 4000ft/fpm.

    About the lack of IFE connection which should have occurred at 21:06 a couple of explanations are that antenna steering was lost, or could not keep up in a spiral, or the antennae were blanked, or that the aircraft hit the sea before those 1½ mins from 19:37 were up. A Case B 0:19:37 descent rate would do that.

    I add a speculation (only), not being much into electronics, that the descent rate after the 7th arc log-on increased to the point that it modulated the carrier to a frequency beyond transmitting or receiving specs; even ruling out Case B. It is something I think you have given thought to earlier, come a glide. I suppose the pilot could have switched it off but why?

  775. David says:

    @Ge Rijn. Above about Boeing simulations I noted, “the ATSB section indicating that simulations meet, simultaneously, BFO descent rate criteria, descent rate increase criteria – and timings – needs expansion”.

    I should now add that the aircraft, when meeting all those criteria, should be in an an attitude and spiral rate which allow communication with the satellite.

  776. Andrew says:

    @Oleksandr

    “My point was that the GPS data can reach ARINC 429 through the AIMS. Don stated this contradicts to Boeing sources. The DCGF could in theory ‘capture’ this data, or the ADIRU could relay it back to the ARINC 429, when it identifies itself as inoperative, either due to the alignment in progress, or when it detects internal failure. This is my speculation, which might be incorrect.”

    There is nothing in the manuals to suggest that GPS data is available to the AES the way you propose. The multi-mode receivers (MMR) send GPS data to the following via ARINC 429 buses:

    1. L & R AIMS cabinets.
    2. Electronics unit in the on-side electronic flight bag system.
    3. Ground proximity warning computer.

    The GPS data is used by the Flight Management Computing Function (FMCF) in each AIMS cabinet for position updating. GPS data is also sent from the AIMS cabinets to the ADIRU via the Flight Controls ARINC 629 bus to calibrate the ADIRU’s internal sensors. There is no mention of GPS data being ‘available’ to the AES via the Data Conversion Gateway Function (DCGF) or any other means. IRS data, on the other hand, is specifically mentioned and is sent to the AES by the DCGF via ARINC 429 for high gain antenna beam steering.

    The SATCOM System Interface Diagram shows all the interfaces to the various components of the AES. GPS/MMR data is not listed. Note: “GP 1 data” is time/date data provided by the #1 general purpose bus data from the L AIMS cabinet.

    “Do you observe this always or sometimes? According to FCOM: “the bulletin issued by Boeing (D632W001-TBC, No. TBC-81R3), states that there are known cases of ADIRU heading error anomaly, when, in particular, SATCOM, TCAS and ACARS may become inoperative in-flight, or their functionality can be degraded.”

    Always. I can’t find any reference to the ADIRU heading error anomaly to which you referred. I suspect the bulletin is no longer current. Are you able to provide a link or some other reference?

    “What is a reason why SATCOM is not available during the alignment, in your opinion?”

    I can only speculate that it’s because the ADIRU is not aligned and therefore not providing IRS data to the AES. Our aircraft do not have a low gain antenna, so it is not possible to send or receive SATCOM data unless the ADIRU is aligned. The same thing happens during our post-flight duties, which require us to send a Commander’s report. The FCOM states that the ADIRU must remain ON for about one minute after the report is sent.

  777. Ge Rijn says:

    @David

    I think I see what you mean. 250Hz or 260Hz to start with doesn’t make much difference the way you put it.
    My wondering was that I didn’t quite understood why they took a predicted 260Hz number instead of the known 0:11 ~250Hz while they can not know for sure where the plane was between 0:11 and 0:19:29.
    And you’re right I mixed up the descent rate with a 4000ft descent I see now..

    So resuming from your explanation I conclude now there is no way of knowing ~ what the descent rate, direction or altitude of the plane was between 0:11 and/at 0:19:29.

    I wonder is it possible to derive/calculate a minimum and maximum altitude of the plane at 0:19:29/182Hz?

    On the lack of IFE-connection at 0:21:06 I suggested earlier the possibility the IFE was switched off in the cockpit after the 18:25 SDU log-on.
    Is this possibility debunked in between?

  778. Ge Rijn says:

    @David

    To add on your IFE speculation. I don’t think the speed of any descent rate would have effect on the modulation of the carrier-wave which travels with the speed of light.

  779. David says:

    @Ge Rijn.”I wonder is it possible to derive/calculate a minimum and maximum altitude of the plane at 0:19:29/182Hz?”

    To my knowledge, no. The 7th arc position depends on a height assumption, being a slant range from the satellite.

    “On the lack of IFE-connection at 0:21:06 I suggested earlier the possibility the IFE was switched off in the cockpit after the 18:25 SDU log-on. Is this possibility debunked in between?”

    No. From the Definition of Underwater Search Areas Dec 2015 p10:
    “The fact that the expected IFE system transmission was not received could be due to:
    • the IFE system being selected off from the cockpit overhead panel at some point after the 18:25 logon…….”

  780. Ge Rijn says:

    @David

    A reason for the pilot/hijacker to switch off the IFE right after the repowering of the SDU (left IDG) at 18:25 could be obvious if he depressurized the cabin after IGARI with the goal of killing all in the cabin.
    He could not have been quite sure yet everyone was unconsious or dead at 18:25.
    Switching off the IFE would be a logical action to prevent any communication from the cabin via the IFE also after 18:25.

  781. David says:

    @Ge Rijn. Your 1.45. Thanks. My shortest lived speculation ever.

  782. Ge Rijn says:

    @David

    Thanks. Clear. I did not know this IFE possibility was mentioned in the Dec 2015 report.

    Another reason the pilot/hijacker could not be sure everyone in the cabin was unconsious or dead at 18:25 is the possibility at least the co-pilot was still alive regarding his phone connection only ~25 minutes earlier.

    He would have been the first one to know what was going and the only one of the crew not trained to sit down and grab an oxigen mask the moment they came down (like @Andrew pointed out).
    He might have grabbed a portable (like the steward in the Helios flight) and survived the longest. Possibly banging on the cockpit door in the meantime.

  783. Ge Rijn says:

    @Andrew

    Then a question comes up in my mind regarding my previous comment you probably can answer if you like.

    When those masks come down are all passengers ordered by the crew (after they put their masks on probably..) to strap themselfs in?
    And does the crew the same?

  784. Andrew says:

    @Ge Rijn

    There is an automatic pre-recorded announcement over the PA in the event of a depressurisation. If that fails for some reason MAS procedures state that the cabin crew are to shout “Mask on, fasten seat belts”. Cabin crew are trained to don the nearest oxygen mask and sit at the nearest seat or between seats if necessary.

  785. Ge Rijn says:

    @Andrew

    Thanks. So in such an event all passengers would be constraint in their movements while fixed in their seats with a mask covering their mouths.
    Can you confirm if the crew is also required to put on a seat belt if possible?

  786. Mick Gilbert says:

    @Ge Rijn

    Regarding the IFE, surely any even halfway planned deliberate malicious act would have seen the IFE switched to OFF as soon as the plan was put into effect and never subsequently turned back ON. There is no logical reason for having the IFE ON at 1825 UTC and then turned OFF. In fact there is no logical reason for a malicious perpetrator taking the somewhat drastic action of depowering the Left AC Bus for any reason in general, leave alone doing so specifically just to disable the SATCOM. Preventing ACARS using the SATCOM to send downlink messages is easily achieved by deselecting SATCOM from the ACARS Manager page 2/2; preventing the IFE SMS/email function from using the SATCOM is even more easily achieved by selecting the IFE to OFF.
    However, if for some arcane reason a perpetrator decided to take the extra precaution of shutting down the SATCOM by depowering the SDU why would they then somewhat cavalierly repower it within 30 minutes flying time of Penang?

  787. Andrew says:

    @Ge Rijn

    “Can you confirm if the crew is also required to put on a seat belt if possible?”

    Yes, they are required to sit down and secure themselves by any means possible.

  788. Ge Rijn says:

    @Mick Gilbert

    By depowering the left AC bus several systems would shut down at the same time in one action. The satelite link would be lost with the SDU, so the IFE also. The CVR would be depowered and the main cabin lights would go out.
    All important if a pilot/hijacker wanted to go dark in one go and block all communications at the same time.

    I think if a pilot/hijacker could accomplish this in one action he would rather choose to do so. There would have been no need to switch of the IFE seperately at that time. And it would be a very easy move to switch off the IFE in the cockpit just after the repowering of the left AC bus at 18:25.

    For the repowering of the left AC bus (if that’s what he did) at 18:25 I see no obvious reason also.
    Maybe he wanted the TCAS on line again being on a busy regular flight path above the Malacca Straight? Or the possibility of jettisoning fuel? Both are lost also after depowering the left AC bus according the following Jeff Wise-topic:

    http://jeffwise.net/2016/05/16/the-sdu-re-logon-a-small-detail-that-tells-us-so-much-about-the-fate-of-mh370/comment-page-1/

  789. Ge Rijn says:

    @Mick Gilbert

    I forgot another possible reason for repowering the left AC bus at 18:25.
    Also the left Window heat (captains seat) is lost after depowered AC bus.
    Maybe this caused icing/visibility problems on that window and risk of cracking if sustained too long.

  790. Niels says:

    @All
    In previous weeks I’ve been focusing on optimizing my explicit path generation tool and using it to study (near) straight paths, i.e. having a (near) constant true track. I’ve added the option to add a linear offset to the Doppler residual function and intend to study all offset functions combining (+8, +4, 0, -4, -8 Hz) at both 19:41 and 00:11 UTC (so 25 paths). The procedure that I follow is to minimize the variance in the track by choosing the optimal 19:41 latitude. This gives a unique path for every linear offset function added to the D(t) input function. Some first results can be viewed through the link.

    https://www.dropbox.com/s/blm5ym5is76dsr1/StraightPaths_linoffset_D_130517.pdf?dl=0

    Note that the two extremes (+8, +8) [Hz] and (-8, -8) [Hz] for the 19:41, 00:11 offset in D give a range of 00:11 latitudes between S28 and S39 degrees. Note also that taking the D(t) function without any added offset gives a 00:11 latitude near S34 degrees. For those interested in path ending near S30 degrees I’ve added a path based on an offset (-4, -6) [Hz]. Feedback is welcomed.

  791. Joseph Coleman says:

    @Niels

    Do you have the FMT info leading to the -7.4 Latitude at 19:41 as seen in your link?

  792. Niels says:

    @Joseph Coleman
    In these calculations I only consider the BFO/BTO data between 19:41 and 00:11 UTC. There will be several possible reasons to eliminate some of the paths found, and clearly a S7.4 latitude at 19:41 could be one of them.
    Personally I feel 19:41 latitudes north of equator as more likely, but perhaps others have a different view on this.
    So for the moment I’ll try focus and report on what are possible paths given the data between 19:41 and 00:11 UTC.

  793. Ge Rijn says:

    @Niels

    Not exactly my cup of tea but while I’m in to it the last days so deeply I had a go also on your path calculations just to learn from it and give it a try.

    It seemes to me your ~30S/97E and 5.9N latitude at 19:41 would put the FMT around ~8N and ~92E if the flight path from Penang over the Malacca Straight is not turning around 18:40 but going straight on with a descent past Nicobar as @VictorI has proposed.

    Roughly estimating on Google Earth the 19:41 lat/long would than be at 5.9N/~92.30E. But I cann’t find clear enough if this is on or near the 19:41 ‘ping-ring’. Is it?

    It seems also this makes a nice great circle route overall to Wilkins Runway as a final waypoint.

    Does this make any sence to you?

  794. Joseph Coleman says:

    @Niels

    Thanks for your graphs. They all look near straight paths.

    Just by looking at the similarities below, you could see a nearly a constant speed in a nearly constant direction. The changes in range (BTO) are Equally proportional.

    Percentage breakdown from a whole BTO difference of 6900ms from 19:41 to 00:1929 Assuming BTO 18400 at 00:1029 is used.

    BTO from 19:41 to 00:1929

    19:41 to 22:41 Difference 3040 ms
    44.05797101449275%

    22:41 to 00:1059 difference 3500ms
    50.72463768115942%

    00:1059 to 00:1929 difference 360ms
    5.217391304347826%

    Sat Path 19:41 to 00:1929

    19:41 to 22:41 difference 56.651km
    43.80581953712797%

    22:41 to 00:1059 difference 66.32km
    51.28244782443958%

    00:1059 to 00:1929 difference 6.352km
    4.911732638432452%

    Total sat movement 129.323km (great circle)

    Percentage breakdown of whole Sat movement (great circle) from 19:41 to 00:1929.

  795. DrB says:

    @DennisW,

    You said: “It is not opinions I am objecting to. It is the arrogance with which the opinions are presented, and the blatant exclusion of information that does not support those opinions.”

    Fair enough, but sometimes statements made in ignorance might appear arrogant to someone much better informed. I would you suggest you simply point out what information has been excluded, and why that information changes the conclusions.

    You also said: “However, it is relevant to the early work done by many of the people here who correctly determined that straight paths to 38S were the only solutions with low single digit BFO error that also satisfied the BTO constraints.”

    There are also non-straight paths which have small BFO/BTO errors. I have previously described a CTH path with BFO errors from 18:28 – 00:11 of 7 Hz Peak and 3 Hz RMS and BTO errors of 29 microseconds RMS.

  796. DrB says:

    @MIck Gilbert,

    You said: “As Don has pointed out I’m barking up the wrong tree on the FMS defaulting to anything other than CMH at an EOR/DISCONTINUITY error; it would appear that regardless of MCP HDG/TRK Reference Switch position the default mode is always CMH.”

    As has been posted here several times, Honeywell says the default EOR/DISCONTINUITY navigation mode is CMH when the NORM/TRUE switch is in NORM and CTH when the switch is in TRUE.

  797. Mick Gilbert says:

    @DrB

    Regarding “There are also non-straight paths which have small BFO/BTO errors. I have previously described a CTH path with BFO errors from 18:28 – 00:11 of 7 Hz Peak and 3 Hz RMS and BTO errors of 29 microseconds RMS.”

    You’ve previously noted that “There is an alternative scenario that ends up in the same neighborhood. It uses Magnetic Track. I have noticed for a long time now that CMT and CTH routes have nearly equal curvature.”

    Would I be correct in assuming that a CMT of 180° produces a similar outcome error-wise to your CTH scenario?

  798. DrB says:

    @Mick Gilbert,

    No, the CMT has somewhat larger errors than CTH when using ECON or best holding speeds. I have not yet tried an arbitrary CAS. My results were near 180 degrees, but not exactly. I want to work on this some more once I have a usable fuel model for CAS.

  799. Mick Gilbert says:

    @DrB

    Thank you for that clarification.

  800. DennisW says:

    @DrB

    “There are also non-straight paths which have small BFO/BTO errors. I have previously described a CTH path with BFO errors from 18:28 – 00:11 of 7 Hz Peak and 3 Hz RMS and BTO errors of 29 microseconds RMS.”

    Yes. However “back in the day” it was fashionable to construct paths with low single digit BFO errors. Our filtering criteria have changed. No doubt about that.

  801. DrB says:

    @MH,

    You said: “@DrB- maybe try a difffent browser or actively use the down arrow or side scroll to force the page update.”

    Thanks for the suggestions. I tried both Chrome and Internet Explorer. Same result – about 77 seconds. The down arrow is easier to use than the side scroll, but the speed is still the same.

  802. DrB says:

    @Gysbreght,

    You said: [START] You wrote: “I believe Gysbreght showed a plot of those results matching the Aero Fuel Mileage curve roughly but not exactly.”
    I did more than that. I provided the TSFC model that is a complement to Professor Obert’s CD curves.

    “I am using some basic physics equations for density, lift, drag, etc. in developing a compact empirical model for B777/Trent fuel flow as a function of weight, altitude, and air speed.”

    Without a model of engine fuel efficiency as a function of thrust and airspeed it’s hard to see how your basic physics equations can get you any further. [END]

    Of course I agree that the engine efficiency model is necessary. That is what I am working on now.

    What I need is a set of equations (or a 3-D table to be interpolated) that provide an accurate estimate of fuel consumption as a function of three variables: air speed, flight level, and aircraft weight under the assumption of ISA conditions. Your two curves may allow that information to be derived but you have not provided any generalized equations for that purpose, so I can’t use what you have done so far. I am hopeful that all the information needed is actually contained in the Boeing LRC/Holding/Checkpoint tables and the Aero Figure 1, assuming that the appropriate functional form for interpolation is known from first principles and reported measurements using similar engines.

  803. Mick Gilbert says:

    @Ge Rijn

    As noted in the Jeff Wise post that you’ve referenced;

    “… some o