MH370 Safety Report Raises Many Questions


As expected, the Safety Investigation Report on MH370 offered no explanation on the cause of the disappearance. “The answer can only be conclusive if the wreckage is found,” Kok Soo Chon, head of the MH370 safety investigation team, told reporters. However, Malaysian investigators did surmise the plane was intentionally diverted, likely due to unlawful interference by a third party. The Malaysia investigators also believe the disappearance could not have been a deliberate act by the pilots based on their background, training, and mental health.

As all the passengers and crew were cleared,  who was this third party that diverted the plane? How can Malaysian investigators ignore that the captain had the best opportunity and capability to divert the plane? How does the compressed timeline of the diversion fit any other possibility if the diversion was intentional? It is understandable that the Safety Report did not apportion blame to the captain. However, it is not understandable that the report deflected blame to an unnamed third party.

The report, including the appendices, is 1,423 pages, and it will take some number of days for independent investigators to thoroughly pore over the entire contents and provide thoughtful comments and analysis. However, based on an initial review, there are some technical questions and inconsistencies that are apparent:

Radar data

The report provides more details about the radar data, but Malaysia fails to provide the raw military data that would allow an independent review. (The civilian radar data was previously made available through an unsanctioned release, and published in a previous blog post.) The military radar includes speed and altitude data, in addition to latitude and longitude at each capture. However,  some of the variations in speed and altitude are beyond the capabilities of the plane. For instance, the measured speed and altitude at 18:01:59 UTC are 589 knots and 58,200 ft. One minute later, the speed and altitude are  recorded as 492 knots and 4,800 ft. The investigative team was warned that the altitude and speed extracted from the data are subjected to inherent error. The only useful information obtained from the Military radar was the latitude and longitude position of the aircraft as this data is reasonably accurate. How the military radar data can be so far out of calibration is unexplained.

After all this time, we still can’t be sure what radar data is available as MH370 passed over the Malacca Strait. According to the main body of this report, the radar target disappeared at 18:01:59 near Pulau Perak, and re-appeared at 18:15:25 along airway N571 between waypoints VAMPI and MEKAR, disappearing at 18:22:12 about 10 NM past MEKAR. However, in an Appendix 1.6E, Boeing reports that after 18:01:59, there was only one capture at 18:22:12. It appears that the radar data provided to Boeing is consistent with the data provided to the ATSB, but Malaysia claims there were additional captures along airway N571 between 18:15:25 and 18:22:12.

We also can’t be sure when the unidentified radar targets captured over the Malay peninsula were first recognized as MH370. According the Safety Report, On the day of the disappearance of MH370, the Military radar system recognised the ‘blip’ that appeared west after the left turn over IGARI was that of MH370. Even with the loss of SSR data, the Military long range air defence radar with Primary Surveillance Radar (PSR) capabilities affirmed that it was MH370 based on its track behaviour, characteristics and constant/continuous track pattern/trend. Therefore, the Military did not pursue to intercept the aircraft since it was ‘friendly’ and did not pose any threat to national airspace security, integrity and sovereignty. In light of the claim that the military was fully aware of the path of MH370, it is not explained why the initial Search and Rescue operations were coordinated in the South China Sea to the east of Malaysia, and proceeded for some number of days before they were moved west of Malaysia to the Indian Ocean.

Pilot Simulator Data

Citing the Royal Malaysian Police (RMP) report from May 2014, the Safety Report says It was also discovered that there were seven ‘manually programmed’ waypoint4 coordinates … that when connected together, will create a flight path from KLIA to an area south of the Indian Ocean through the Andaman Sea. These coordinates were stored in the Volume Shadow Information (VSI) file dated 03 February 2014. The function of this file was to save information when a computer is left idle for more than 15 minutes. Hence, the RMP Forensic Report could not determine if the waypoints came from one or more files.

What is not mentioned is that certain values that were common to the data sets indicate that these files were likely from the same simulation. This was first documented in a technical paper I co-authored with Yves Guillaume, and summarized in a previous blog post. Also included in the previous blog post were statements from the ATSB indicating that the chronological order of the data sets matches a flight departing KLIA, flying over the Malacca Strait, continuing past the Andaman Islands, turning to the south, and exhausting fuel in the SIO, in the same order that the progressively depleting fuel levels suggest. The time values also indicate that the flight session lasted for about one hour, suggesting the position of the aircraft in the simulation was manually advanced and the fuel levels manually depleted.

There are other strange facts surrounding the simulator data. For instance, Malaysia does not explain why only fragments of the data files were recovered, as the reconstruction of the entire contents of the data files should have been possible using the data in the Shadow Volume. The missing portions of the data files include important information that would be stored in the flight management computers (FMCs), including flight plans. Also, the time stamp data that was included in the data files that were made available to the ATSB by the Australian Federal Police was not included in the RMP report.

The RMP conclusion that there were no unusual activities other than game-related flight simulations is odd considering the extraordinary coincidence that a simulated flight including a departure from KLIA and ending in fuel exhaustion in the SIO was recovered after the disappearance.

Investigation of the Flaperon

After the flaperon was recovered from Reunion Island, which is sovereign territory of France, French scientists performed tests to determine the flaperon’s provenance, to analyze the attached marine biology, to determine its buoyancy and other hydrodynamic characteristics, and to investigate the damage to determine how the flaperon separated from the aircraft. To the French team, it appeared that the flaperon impacted the water while still attached to the aeroplane and that at the time of the impact it was deflected. The implication is that water forces from an attempted ditching caused the flaperon to separate rather than in-flight during a high speed descent. The French were careful to advise that this is only an hypothesis because of the limited data made available to them by Boeing, and because of the complicated dynamics of the impact mechanics.

Nonetheless, if it can be proven that the aircraft was in a controlled descent at the time it impacted the ocean, the distance from the 7th arc could be greater than 120 NM, as there could have been a long glide after fuel exhaustion. A controlled descent after fuel exhaustion would also leave open the possibility of pilot navigational inputs after 19:41, and possible paths would include crossing the 7th arc over a large range of latitudes. These unknowns would make it very difficult to define a new search area of a manageable size. For this reason, it is critical that mechanism that led to separation of the flaperon be determined with a higher level of certainty.

First Officer’s Cell Phone Connect

The First Officer’s cell phone registered on a cell tower as MH370 passed to the south of Penang Island. Although it would be unlikely that a cell phone connection would persist long enough to complete a call, a cell phone registration of short duration and at cruise altitude is not that uncommon.

Considering the large number of Malaysian passengers and crew that were likely carrying cell phones compatible with the Malaysian cell network, and with some fraction of those phones likely in an operational configuration during the flight, it is odd that other cell phone registrations did not occur. It is unexplained whether or not a systematic review of the cell phone numbers of all passengers and crew was ever performed.

Final Comments

It is unlikely that another search for MH370 will occur unless a case can be made that there is a reasonable probability of success. Unfortunately, the Safety Report raises more questions than it answers, and it will be difficult to use the information in the report to define a search area of a manageable size. The answers to some questions will not be known until the flight data recorder is recovered. However, the answers to other questions are known to Malaysian authorities today. Any remaining chance to find MH370 squarely rests on the willingness of the new Malaysian administration to cooperatively work with official and independent investigators.

629 Responses to “MH370 Safety Report Raises Many Questions”

  1. flatpack says:

    I may be a pernickety old ham (in fact I am fairly sure that I am) so this may be inconsequential…

    Page #303 notes KL ACC {Lumpur Radar (Sector 3+5)} is on radio frequency 132.5 MHz

    Page #316 effectively confirms this as the frequency used “This was especially so after he (the Lumpur Sector 3+5 Radar Controller) had tried to establish radio communication with MH370 by making a ‘blind transmission’ on the VHF radio frequency 132.5 MHz”

    Page #309 – Note 4
    “When Lumpur Approach Control transferred MH370 to Lumpur Radar (Sector 3+5), the Sector 3+5 radio frequency was transmitted by Lumpur Approach Control and MH370 read back the radio frequency, “Night one three two six Malaysian err… Three Seven Zero”.”

    Emphasis mine.

    So, the summary in the second paragraph of Note 5 on Page #310 is not accurate in that it does not note that the last frequency readback made by the PIC was apparently wrong.

  2. Victor Iannello says:

    @Hank said: I don’t know the lowest speed in ground effect that a zero fuel 777 can achieve. What is the best way to submerge a 777 to minimize surface debris?

    Ground effect should be significant all the way down to stall speed. As for how to minimize surface debris, I would open the outlet valves, try to ditch perfectly (low speed, nose up) and sink the plane in one piece. I think trying to dig the nose into a wave would possibly break up the plane.

  3. flatpack says:

    @Nederland

    ‘This could also explain why a) radio transcript section 4&5 appear blurred (Malaysian errrr 3-7-0, repeated cofirmation of altitude)’

    Not to mention that the radio frequency readback was apparently wrong (see my post above).

    I shall take the liberty of repeating my last post from the previous thread:-

    “The Team noted that the medical records or reports of the flight crew
    obtained from the MAS Medical Centre facility did not include medical records or reports from other medical facilities. In fact, the Team has found a medical record of the PIC from another private medical centre which was not recorded in the MAS Medical Centre. The records from the MAS Medical Centre as well as the records from private clinics regularly visited by both the flight crew also seemed to be mainly records related to minor ailments such as coughs and colds and may not be reflective of the complete medical record of the individuals in question.”

    Main Report Page #359

  4. Nederland says:

    RE ELTs and why a ‘soft ditching’ may not have seemed suitable to a potentially-suicidal pilot flying:

    “In general, Emergency Locator Transmitters (ELT) are intended for use on land or on the surface of water, and neither portable nor fixed ELT signals are detectable when the ELT is submerged in deep water. Portable ELT is equipped with a floatation device and can be activated by immersion in water. For effective signal transmission, the antenna of the ELT must remain above water.”

    In an earlier ICAO report, it was also noted that the ELTs could be switched off (although with some effort).

  5. TBill says:

    @Victor
    I am holding some comments on the report until I review, but I was moved by the article you posted yesterday on Twitter:

    https://abcnews.go.com/US/final-report-missing-flight-370-yield-answers/story?id=56887426

    “The aviation community continues to be desperate for answers to the vanishing of 239 people on their way to Beijing, but Monday’s report is unlikely to satisfy their questions.

    In fact, some are afraid a conflict of interest exists with the Malaysian government writing such a report on their own aircraft piloted by government-paid pilots.”

    “Since the airline was owned by the government, and given that the most likely cause was a deliberate act by the pilot, the Malaysians would prefer that the issue fade from memory,” said (ABC News contributor Col. Steve) Ganyard.

    “It’s a shame because the government likely has evidence that would point to cause or motive but won’t release anything that would be politically embarrassing.”

  6. Victor Iannello says:

    @TBill: Yes, I tagged that article on Twitter because I thought it was an uncommonly good one from ABC. Geoffrey Thomas also had a good one entitled “Malaysia exonerates MH370 pilots and walks away”.

  7. TBill says:

    @Victor
    Quite interesting about the radar and IGARI turnback.

    Malaysia now seems to be saying their military radar analysis “system” (computer) knew it was MH370 turning back at IGARI and therefore did not signal an intruder alarm. So that is why they did not pursue it.

    This is quite contradictory to PM Razak on 15-March-2014 said they did not know the unknown blip was MH370 until Inmarsat was able to confirm with the sateliite ID and radar path matching up at 18:22 to 18:25.

  8. Victor Iannello says:

    @TBill: On any given day, Malaysian officials say whatever is most convenient. We’ve seen this many times. The same Inspector General of Police (IGP) that ran the RMP and exonerated Najib Razak (despite $700M diverted from 1MDB accounts and deposited in Najib’s personal account) was in charge of the RMP when the crew of MH370 was investigated. The conclusions from the RMP report are not worth the paper they are written on. Yet, the Safety Investigators relied upon that report to exonerate the pilots.

  9. David says:

    @TBill. When a friendly aircraft alters course unexpectedly and heads towards a major defence base, transponder off, it is odd that the investigation offers no comment on the Malaysian military assessment that it remained friendly.

    The logic overlooks 9/11.

  10. Hank says:

    @victor

    Ditching in SIO would not be like the “Miracle on the Hudson.” Ditching in rough sea seldom go well. Issue is what PIC believed to be best way to end the flight. Seems that lots of thought and simulation was done. He must have thought about the end phase. How would experienced large aircraft pilots approach the problem? Maybe a transport pilot on blog has thought about it.

    Clearly Malaysia never wanted to consider a pilot suicide – a very bad outcome. Best for them if the aircraft is never found. It was odd that ATSB never addressed possible piloted flights and only dealt with ghost ships to define the probable search. Maybe pushback to not consider anything else? I didn’t know about the degree of cover for what was on the PC simulation. From your post regarding the simulation it is hard to believe that this is anything but a PIC suicide mission and very well executed.

  11. Sabine Lechtenfeld says:

    @David, I fully agree with your comment above. The scenario you describe is totally unbelievable. How could a plane, which had been sharply turned around and was heading westwards towards the peninsula where Kuala Lumpur is, and then towards an important airbase, have been deemed friendly and not worthy of an adequate response??
    There is a lie hidden somewhere. The plane was either not detected in real time, and the Malaysians are too embarrassed to admit this, or there have been early reactions which the public is not supposed to know about.
    We will probably never know for sure, though. The totally unwarranted half baked attempt to exhonerate the pilots although they are the most likely candidates for having turned around the plane, considering the time table of crucial events, speaks volumes 🙁

  12. Niu Yunu says:

    Hank: “I didn’t know about the degree of cover for what was on the PC simulation.”

    If they wanted to cover it up, wouldn’t they have covered it up entirely (i.e. never revealed the PC simulation in the first place) ?

  13. Richard Godfrey says:

    @flatpack

    Although the main body of the report states that the radio frequency of Lumpur Radar is 132.5 Mhz both on page 303 and page 316, the Appendix 1.18E states the frequency is 132.6 Mhz consistently on all 29 pages. Appendix 1.18D states the handover at 16:46:39 was to “Malaysian Three Seven Zero contact Lumpur Radar One Three Two Six”. Appendix 1.18F states the Blind Transmission at 17:41:23 UTC was also on 132.6 MHz.

    I think there are 2 typos in the main report, which your eagle eye has spotted, and 132.5 MHz should read 132.6 MHz in both cases.

    However the PIC read back the correct frequency according to the Appendix 1.18D at the handover to Lumpur Radar.

  14. Richard Godfrey says:

    @Victor

    I find it incredible that the Malaysian Report states the private address of certain relatives of MH370 passengers. It shows how little care was taken in the production of the report.

    Or was this a deliberate action against the next of kin?

  15. flatpack says:

    A while ago I wondered whether CLI was displayed by the cockpit SATCOM and could be the reason that the 18:40 call was not answered. Don Thompson replied that calling line identity was not signalled from the GES to the AES during the C-channel setup.

    I noticed that in the report on Page #382 a ‘priority level’ is mentioned, see below:-

    ‘Two Ground-to-Air Telephony Calls were placed to the cockpit from the MAS Airline Operations Centre at Airline Operational Communications (AOC) Q10 priority level at 1839 and at 2313 on 07 March 2014. Neither of the calls was answered.’

    Is this AOC Q10 priority level a ‘real thing’ and is it transmitted and displayed in the cockpit? Can calls be filtered using it?

  16. Pavel says:

    I suppose there is a mistake in “Radar data” section:

    “For instance, the measured speed and altitude at 18:01:59 UTC are 589 knots and 58,200 ft. One minute later, the speed and altitude are recorded as 492 knots and 4,800 ft.”

    4,800 ft. is too low (they never mentioned so low altitude for NH370 in official reports). Could it be 48,000 instead?

  17. Victor Iannello says:

    @Richard Godfrey: It is often hard to distinguish between Malaysian incompetence and malice. In this case, I think it was incompetence.

  18. Victor Iannello says:

    @Pavel: The low altitude of 4,800 ft is not a typo in the text. Look at Figure 1.1B. I agree that the progression of reported altitudes (and speeds) did not actually occur, and should be ignored, as was recommended by investigators.

  19. Don Thompson says:

    @Flatpack

    Priority is definitely a thing: packet data (ACARS and ISO-8208 SSN) correspondence and voice connections are subject to prioritisation. The Aeronautical Mobile Satellite (Route) System – AMS(R)S is a ‘safety’ system that may also be used for operational and amenity services. The priority levels range from ‘distress/urgency’, thru ‘flight safety’, ‘regularity/meteorological’, to ‘amenity’.

    IIRC, Q10 = AOC = ‘regularity/meteorological’. Priority is not a selectable ‘thing’ on the part of a calling or called party, it’s a mechanism for pre-emption in circuit/datalink access. Given that only one SATVOICE call was attempted at any time during the flight, there was no contention for circuit/datalink resources, no pre-emption was necessary, and the signalled priority was superfluous.

  20. Don Thompson says:

    On Twitter Sumisha Naidu of Channel News Asia, @SumishaCNA quoted Minister Loke as saying ““Every word recorded by the investigation team will be tabled in this report. We are committed to the transparency of this report. It will be tabled fully, without any editing, additions, or redactions”

    Bearing that statement in mind, do review the ACARS Traffic Log (Appexdix 1.9A) presented with the Factual Information of March 2015 and the ACARS Traffic Log (Appendix 1.9A) presented with the report of 2018-07-30.

    The Log, publised yesterday, has definitely been edited from its original form. This most recent copy prepends a page to the 2015 Log.

    That additional page illustrates the instruction for VHF data to be disabled prior to 9M-MRO’s departure to operate MH370. I have learned, from other contemporary reports discussing aero comms that VHF data coverage in SE Asia is provided by AEROThai (note address BKKXCXA) as a partnership with ARINC. So customers of SITA effectively pay a premium on VHF data traffic as they’re ‘roaming’ beyond SITA’s coverage. Hence, SATCOM was actually the preferred AOC data medium for MAS B777 operations.

    The filter used to create the message report at Page 1 of 14 is different to that used to create the message report comprising Pages 2 thru 14.

    The filter for Page 1 is the aircraft registration, 9M-MRO.

    The filter for Pages 2 thru 14 is the a/c registration AND SITA’s message processor, QXSXMXS. QXSXMXS provides ACARS correspondence via SATCOM.

    An ACARS Log, filtered by only the a/c reg 9M-MRO, for the period between 2014-03-07 12:50Z thru to, say, 2014-03-08 01:20Z might illustrate any further attempts to message the aircraft, just as voice call attempts were made.

  21. Andrew says:

    @flatpack

    RE: “Is this AOC Q10 priority level a ‘real thing’ and is it transmitted and displayed in the cockpit? Can calls be filtered using it?”

    Expanding on Don’s comments, priority level Q10 is also termed ‘Operational Low’ or simply ‘Low’ and is the lowest priority given to SATCOM calls for operational purposes. It is typically used for operational control purposes such as redispatch, maintenance and other calls of an administrative nature. There are two higher priorities, ‘Emergency’ (Q15) and ‘Operational High’ (Q12) that are used for distress/urgency communications and ATC clearances respectively. There is also a lower non-operational priority, ‘Public’ (Q9), assigned to public phone calls.

    The priority of an incoming SATCOM call can be displayed on one of the cockpit CDUs and most calls are Q10, or Low priority. Higher priority calls are very rare in my experience.

  22. Victor Iannello says:

    The apparent errors in the altitudes measured by the military data needs further scrutiny and explanation. Some here have ardently defended the altitudes that were reported in the Factual Information (FI) released in March 2015. Even before the release of the Safety Report, many of us have questioned their accuracy for many reasons, including the fact that they don’t seem to match up with the deduced altitude across the Cone of Silence at Kota Bharu, as Mike Exner has pointed out many times.

    It appears that the altitude data that was included in the FI were values that were selected because they were in a reasonable range, and the extreme low (e.g., 4,800 ft) and high (e.g., 47,500 ft and 58,200 ft) values were conveniently omitted. I think there is a story here that needs to be better understood.

    It may be that Malaysian military radar performs poorly, and for political and strategic reasons, there was an attempt to hide this. (If military jets wish to intercept a target, discrimination of altitude would be important if radar operators are providing assistance.) Or, perhaps the military radar is accurate, and multiple targets were captured. I suspect the former, but how can we be sure?

  23. Victor Iannello says:

    @Andrew: Can you confirm that in your experience there is no way to identify an incoming SATCOM call?

  24. Andrew says:

    @Victor

    Not in my experience. On the rare occasions I’ve had an incoming call, the caller could only be identified by answering the call.

  25. Richard Godfrey says:

    @Don

    Well spotted! The new page added to the ACARS Log with 3 new items and 2 repeat items does make you wonder what else is missing.

    1. Complete ACARS Log.
    2. Raw Radar data.
    3. Complete passenger mobile connections.
    4. Complete PIC home PC data.
    5. Content of calls made by PIC before take off and previously to MAS Engineering.
    6. Information withheld by Indonesia.
    7. …

  26. Andrew says:

    @Victor

    However, there are several references to caller identification in the following documents:

    ICAO Satellite Voice Guidance Material

    FAA AC 20-150B Airworthiness Approval of Satellite Voice (SATVOICE) Equipment Supporting Air Traffic Service (ATS) Communication

    Perhaps Don can explain.

  27. flatpack says:

    @Don Thompson
    @Andrew

    Thank you both for your responses re AOC Q10 priority level etc.

    I don’t really subscribe to a ‘political hostage taking gone wrong’ scenario but I wanted to fully bottom out the possibility that the SATCOM was re-booted for a reason, vis to enable a ‘side-channel’ capable of little more than sending a go/no-go message without any call being answered (and thus recorded).

    Both CLI and priority level, were they actually displayed in the cockpit, could have been manipulated to achieve this. I am guessing here but my inferrence is that the higher priority levels are hard coded to particular landlines (such as ATC) whereas a burner phone dialing in through a public gateway would be assigned Q9.

    There is still just a slim chance that no Q9 public call in a particular timeslot meant something.

    My other option, the IFE SMS gateway, which was apparently enabled (if I have read it correctly) and which apparently carried no traffic at all is pretty depressing to contemplate.

  28. Gysbreght says:

    Here is Fig. 1.1B (Military Radar Data) plotted to scale:

    https://www.dropbox.com/s/7h9tm1i2sb0yag1/MY_Mil_Radar.pdf?dl=0

  29. TBill says:

    @Victor
    Regarding 4800-ft at Paula Perak I am wondering if they get around to reconciling that with the eyewitness report…standing on the top of the island the viewer could have touched it (slight exaggeration I know)

  30. TBill says:

    @David
    @Sabine
    I mean discounting the radar turnback also looks suspicious, that they maybe knew MH370 was up to mischief. There are the rumors from TimR that a plan for alternate but rogue safe landing was known in some circles, and also some rumor that Hish told the 4 Corners reporters of the record that some were told to “keep their eye on MH370” before the take-off.

  31. TBill says:

    Correction: “off” the record

  32. Victor Iannello says:

    @TBill: I am of course very familiar with @TimR’s claims. However, I don’t recall who made the claim that Hishamuddin said they were aware before the flight that they should watch MH370. Can you please provide details?

  33. Sabine Lechtenfeld says:

    @TBill, yes, please elaborate! I have never heard that rumor…

  34. Gysbreght says:

    Asking for the raw Radar data may be asking too much. According to the Safety Investigation report:

    “The only useful information obtained from the Military radar was the latitude and longitude position of the aircraft as this data is reasonably accurate.”

    Perhaps requesting latitude/longitude data at the times indicated in Figure 1.1B is more likely to get a favourable response.

  35. Victor Iannello says:

    @All: The wording in the Safety Report has a bit different tone than what Kok Soo Chon said at the press conference. The Safety Report implies a diversion by one or both of the pilots is most likely. On the other hand, at the press conference, third party intervention is deemed most likely in light of the background investigations of the pilots.

    From the Press conference:

    “We have examined the pilot, the flight officer. We are quite satisfied with their background, with their training, with their mental health, mental state. We are not of the opinion that it could have been an event committed by the pilot.

    “But at the same time we cannot deny the fact that there was an air turnback. We cannot deny the fact that, as we have analysed, the systems were manually turned off with intent or otherwise.

    “So we feel that there’s also one possibility that could account for all these … No matter what we do, we cannot exclude the possibility of a third person or third party or unlawful interference.”

    From the Safety Report:

    There is also no evidence to suggest that the aircraft was flown by anyone other than the designated MAS pilots. However, the Team does not exclude the possibility of intervention by a third party.

    and this:

    It should be recognised that there is a significant lack of evidence available to the Team to determine with any certainty the reasons that the aircraft diverted from its filed flight plan route. However, the change in flight path likely resulted from manual inputs. The lack of evidence includes the exact location and disposition of the main aircraft wreckage and the evidence that it could provide, the information recorded on the Flight Data Recorder, Cockpit Voice Recorder and other recording devices on the aircraft and the absence of any aircraft voice or data transmissions that could indicate why the aircraft flew to the Southern Indian Ocean.

    Without the benefit of the examination of the aircraft wreckage and recorded flight data information, the investigation was unable to identify any plausible aircraft or systems failure mode that would lead to the observed systems deactivation, diversion from the filed flight plan route and the subsequent flight path taken by the aircraft. However, the same lack of evidence precluded the investigation from definitely eliminating that possibility. The possibility of intervention by a third party cannot be excluded either.

    In retrospect, neither the Safety Report nor the words of Kok at the press conference eliminate the possibility that one or both pilots diverted the plane.

  36. Don Thompson says:

    @Andrew, @flatpack, @Victor

    Concerning Caller ID/calling line identity and SATVOICE.

    Refer to SVGM, sect 4.1.2: “Note Caller ID is not currently available in aircraft avionics; however, manufacturers will consider this human factors element for ease of use in future designs.

    The Inmarsat C-channel sub-channel signalling doesn’t define a means to signal a terrestrial network CLI to the AES. The Inmarsat terrestrial network does exploit CLI as a means to assign priorities to ground-to-air calls (eg mapping known airline ops control centre numbers/ranges to appropriate Q/priority number) and also to permit a terrestrial call to be completed through the AMS(R)S network.

  37. Nederland says:

    The Lido image contains a time stamp “0207:16” at the circumference of the circle, right hand side. Right next to this, there is an outlier “0207:06”

    https://www.ibtimes.com/sites/www.ibtimes.com/files/2014/03/21/mh370-search-briefing-3.jpg

    According to the FR, there were no blips recorded or detected between 0203:09 and 0215:25.

    This means the Lido image is not “real”.

  38. Don Thompson says:

    @flatpack, and perhaps @sabine.

    @flatpack – what if the prior arrangement for a go/no-go step involved an ACARS message to be sent to the aircraft after that hour of ‘radio silence’? That may have initiated a next step…

    I note that much speculation involves the PIC having deep knowledge of satcom technology: that the Log On Interrogations were intentional breadcrumbs, and that the final Log On is somehow synchonised with an intent to dive and pull-out into a glide so as to bluff interpretation of the metadata. I’m more inclined to believe that, if there was a prior arrangement with an accomplice on the ground, that the airborne side of the compact screwed up. I have doubts that the consequence of disabling ACARS while taking the SDU offline for an hour was fully understood. The consequence being that no ACARS correspondence was possible after the failed ground-to-air message initiated at 18:03.

    Of course, the intent may have simply been to proceed directly to the sIO, but in that case why reconfigure the aircraft power distribution back to normal?

  39. TBill says:

    @Victor
    @Sabine
    As far as the “keep an eye MH370 on” reference, is something my memory globbed onto, presumably from a post by someone here in the last 12-months. I think I have repeated several times as it stuck with me.

  40. TBill says:

    P.S.- …and you know what they say, memory is the first thing to go…hmm, or is it the second thing? I forget…

  41. Victor Iannello says:

    @TBill: I did a search in the comments for “eye on MH370”. There were three hits, each one your submission:

    http://mh370.radiantphysics.com/2018/01/21/the-search-for-mh370-begins-again/#comment-12358

    http://mh370.radiantphysics.com/2018/05/16/sixty-minutes-australia-story-on-mh370-is-a-sensation/#comment-16259

    http://mh370.radiantphysics.com/2018/07/30/mh370-safety-report-raises-many-questions/#comment-17273

    Other than your comments, I don’t think this was ever discussed on this site.

  42. Victor Iannello says:

    @TBill: Perhaps this is causing the confusion:

    ABC’s Four Corners program quoted Malaysian Defence Minister Hishammuddin Hussein as saying that Malaysia’s civil aviation authorities called the military asking them to keep an eye on the plane but that the military allowed the plane to glide out to sea.

  43. Victor Iannello says:

    @Don Thompson: You’ve proposed in the past that disabling the SATCOM link in the ACARS Manager screen combined with power down of the SATCOM prevented the receipt of ACARS messages in a way that was possibly not understood by the captain. It’s an interesting idea. The functionality of the ACARS under those conditions was not easy to test before hand. Nonetheless, not being able to test this configuration introduced a tremendous amount of risk. Would the captain have done this? I don’t know.

  44. TBill says:

    @Victor
    Yes that sounds like it could be the basic source information, but I feel like I was remembering someone’s paraphrasing of that.

  45. Dan Richter says:

    What surprise me is that radar data suggesting left turn are from Western Hill, Penang on the other side of peninsula. RMAF Air Defence Surveillance, Sqdrn 321 Martello S-743D Radar at Bukit Puteri, Jertih N5º47’9” E102º30’16”, which is much closer to IGARI on the east side of peninsula, didn’t see MH370.

    https://scontent-vie1-1.xx.fbcdn.net/v/t1.0-9/32253599_1811877282202284_5747545769745842176_n.jpg?_nc_cat=0&oh=83b4220f3ac5e7765762902e40426f33&oe=5C11BB49

  46. flatpack says:

    @Don Thompson

    Yes I agree that your ACARS ‘prior arrangement’ makes sense with regards to ‘the airborne side of the compact screwed up’ because of an incomplete understanding of the effects of the actions taken.

    It may be falsifiable given that there may be a logfile (somewhere other than Inmarsat) that records failed ground-to-air message attempts initiated after 18:03.

    ACARS messages should normally originate from within MAS, though presumably there are service provider access gateways that could be misused by an external unauthorised actor.

    The downside of utilising such a channel would be that, if it had actually worked, it would have been logged by Inmarsat. Also the pool of capable actors would be much, much, smaller than one capable of simply ringing a SATPHONE.

    If the intent had simply been to proceed directly to the SIO, the benefit of reconfiguring the aircraft power distribution back to normal could have been simply to allow the phone to ring but not answer.

    This could lull MAS ops into thinking that the plane was still in the air but with comms problems, thus delaying a search. Another benefit would be to indicate when MAS ops had worked down their SOP to the ‘use the SATPHONE’ action.

  47. Don Thompson says:

    @Dan Richter

    The detail for the RMAF’s operational use of the air defence radars isn’t known. I have speculated that, normally, these radars are surveilling airspace at the outer extent of their range, say 180nm to 250nm, rather than 0nm to 250nm. That’s readily achieved by defining a window for the receipt of the returned ‘echoes’.

    If that was the case Bukit Puteri would have been the source of the targets in the vicinity of Pulau Perak, while Western Hill tracked the turn back from IGARI and the track from VAMPI to MEKAR.

  48. Nederland says:

    The DGA result that the flaperon came off in a ditching rather than a crash was first reported on 4 September 2015 by Florence de Changy in Le Monde. She was reporting about a meeting of the French next of kin with the anti-terrorist judge Alain Gaudino and the technical expert François Grangier.

    “Il a en revanche estimé que la manière dont la pièce était « tordue » laissait penser à un amerrissage plutôt qu’à un crash, ce qui indiquerait que l’avion est resté sous contrôle jusqu’au bout.

    Dans le cas du crash de Germanwings, les plus gros débris faisaient environ 30 centimètres. Le flaperon, lui, mesure environ 2 mètres.”

    “On the contrary, he estimated that the manner in which that piece was ‘torn off’ lets one think of a ditching rather than a crash, something which indicates that the aircraft remained under control until the end.

    In the case of the Germanwings crash, the largest piece of debris was around 30 centimetres. This flaperon measured around 2 metres.”

    https://www.lemonde.fr/asie-pacifique/article/2015/09/04/le-flaperon-retrouve-a-la-reunion-appartient-bien-au-boeing-777-du-vol-mh370_4746144_3216.html

  49. flatpack says:

    @Andrew

    You posted on July 31, 2018 at 9:00 am

    ‘The priority of an incoming SATCOM call can be displayed on one of the cockpit CDUs and most calls are Q10, or Low priority. Higher priority calls are very rare in my experience.’

    That being the case, a prior arrangement to use a public phone (indicating priority Q9) by a specified time would suffice.

    The pilot(s) would be in a position to check and confirm that MAS ops would display Q10 and indeed to confirm that a public phone would display Q9. This could even be done on the ground.

    Presumably the log files for all planes used by both pilots could be checked for ‘test calls’.

  50. Don Thompson says:

    @flatpack,

    Pls excuse me butting in on the question to @Andrew: the SITA CLI log for both SATVOICE calls was included in a previously released document. The CLI associated with both calls, 18:39 and 23:13, were MAS OCC numbers.

  51. flatpack says:

    @Don Thompson

    What I was (unclearly) trying to suggest was that there may have been precursor events such as SATVOICE calls to test the priority level displayed on the cockpit CDU using a public phone.

    The post above at 4:29pm by Victor notes the risks of not testing a particular configuration. Any such tests would probably have been carried out in the preceeding weeks and thus I suggested that the log files for all planes used by both pilots could be checked for ‘test calls’.

  52. David says:

    @Victor. To me, there has been a substantial if nuanced change in the likelihood accorded to the final BFO’s being valid and the final DSTG-assessed descent rates actually being realised. If so that increases the prospects of a piloted glide.

    While the Malaysian report at page 381 says of the 00:19 log on request and acknowledgement BFOs that they, “could have been due to uncompensated vertical velocity” it adds they also, “could have been due to OCXO drift”. The impression conveyed by this is that possibilities have a like standing. Page 391 conveys the same, both adding that there could have been a mix of the two, that further reducing the confidence that the descent rates widely supposed hitherto are indeed valid. Of course OCXO drift has been discussed extensively here but the descent rates have been treated as highly likely all the same.

    Likewise for the formal search. The First Principles Review of November 2016 (report p17) placed some reliance on these descents, noting; “The analysis of the BFO from the final two SATCOM transmissions which indicated that the aircraft was likely to be on an unstable flight path”, that being one basis for its subsequent conclusions about search area.

    Another was, “… recent simulations showed high rates of descent broadly consistent with the BFO analysis. These simulations indicated that the aircraft was likely to be within 15 NM of the 7th arc.”

    It is peculiar that the Malaysian report makes no mention of these most recent simulations (and indeed we were hoping that there would be substantiation of how consistency with the descent rates and their timing had been deduced). It is not just a Malaysian report though since all accredited reps have signed off on it. Do those most recent simulations now have no standing?

    My conclusion is that the underpinning of further searching has been weakened.

  53. David says:

    @Victor. Another weakening has been page 384 where it concludes the most likely reason for the 00:19 log-on was a power cut to the SATCOM and notes that this could have been because, “some form of generator transfer may have occurred.”

  54. Richard Jones says:

    The only hope left is either that someone who has concealed evidence decides to release it or detailed mapping of the ocean floor trips over the wreckage. Thanks to all those who have put an enormous effort into studying and analysing the currently available evidence. Where money and politics are in play truth often becomes a casualty. A tragic and perhaps shameful episode in aviation history.

  55. TBill says:

    @David
    @Richard Jones
    The flip side of fading hope for finding aircraft is that it seems unlikely a 3rd party hijacker would have a flight plan designed soley to hide a MAS aircraft by gliding from Arc7…instead the plane should be findable.

    Therefore this possibly argues for checking 20-25 South if someone really thinks it is 3rd party, we should be able to find it.

  56. TBill says:

    @Victor
    “@All: The wording in the Safety Report has a bit different tone than what Kok Soo Chon said at the press conference.”

    What you say makes sense. The report had to be revewied by the parties, and the radio interview that @ventus45 posted a few months ago indictaed the draft report would say pilot hijacking among other possibilities. I am thinking the report review parties would not agree with the draft if Malaysia failed to consider internal causes.

  57. DennisW says:

    @Richard

    With the regime change it is unlikely that someone who previously concealed evidence is going to come forward and risk prosecution. At this point my feeling is everyone in Malaysia would simply like this incident to go away. The pilots have been “cleared” there is no one to hang for this disaster. I doubt OI will proceed, but you never know.

    Certainly a lot of people put in a lot of work, but my feeling is that it was not wasted. I think we all learned something from the experience.

  58. Barry Carlson says:

    @flatpack,

    Any call by SATVOICE to or from the aircraft will have an Inmarsat log record.

    The simplest method to convey a message from the ground to the aircraft would be via HF on a predetermined frequency. The aircraft’s transceiver needs to be on, set to the preselected frequency and muted in SELCAL mode. On the ground, a suitable HF transceiver capable of sending the appropriate SELCAL allocated to 9M-MRO is all that is required.

    Whoever was to be communicating with 9M-MRO would need to know the aircraft’s SELCAL code. A list of MAS aircraft SELCAL codes wouldn’t be hard to come by, and a mobile phone call to that person confirming the flight was about depart Kuala Lumpur need only mention MRO, and any reference to HF could be avoided.

    Coincidentally, MAS Operations Dispatch has the appropriate equipment located at KLIA, installed in 2013 complete with a log periodic HF antenna normally pointing towards Chennai.

  59. David says:

    @TBill. Yes, there are ramifications to the 3rd party possibility. In fairness it could not be ruled out entirely. Whether that is reasonable grounds not to point the finger at the more likely person or whether it was an excuse not to point the finger will remain unknown unless one of the accredited reps leaks about that.

    Were a third party alive and all alone, as you say it is unlikely he could glide far without autopilot and even if he could, whether he would be so motivated. I speculate that his continuing on to an isolated spot would suggest suicide as a motive, not terrorism or extortion.

  60. Andrew says:

    @flatpack

    RE: “That being the case, a prior arrangement to use a public phone (indicating priority Q9) by a specified time would suffice.

    The pilot(s) would be in a position to check and confirm that MAS ops would display Q10 and indeed to confirm that a public phone would display Q9. This could even be done on the ground.”

    My understanding is that most operators configure their SATCOM systems such that incoming ‘public’ calls (Q9) are not routed to the flight deck. ‘Public’ priority calls are normally restricted to outgoing calls from the cabin. Some configurations also allow the flight crew to select ‘public’ priority for an outgoing call when the number is manually dialed.

  61. Andrew says:

    @Don Thompson

    RE: ‘Refer to SVGM, sect 4.1.2: “Note Caller ID is not currently available in aircraft avionics; however, manufacturers will consider this human factors element for ease of use in future designs.”

    The Inmarsat C-channel sub-channel signalling doesn’t define a means to signal a terrestrial network CLI to the AES…’

    Thanks. I guess the regulators are trying to drive the manufacturers towards providing a means for caller ID. Out of interest, do you know if the Iridium system has that capability? Some websites suggest that it does have that capability for ATC voice communications, provided the airline and service provider have the necessary hardware.

  62. Victor Iannello says:

    @David: I don’t think you should put much weight on the statements that Malaysia has made about the final BFO values. The statements appear to reflect earlier thinking, and seem to have been made without consideration of the work of Ian Holland. I would offer that same caution in regards to the end-of-flight simulations.

  63. DennisW says:

    @VictorI

    I have been a major critic of BFO data from the onset. Mike Exner and I have a long history of knit picking in that regard. At this moment I agree with Mike’s interpretation. I think the last BFO values indicate a rapid descent and likely impact near the 7th arc. I am much less inclined to put a lot of weight into the French interpretation. I still believe that best chance of finding the aircraft is to search North near the 7th arc.

  64. David says:

    @Victor. The simulations are years out of date if that is the reason and the Holland work likewise. Haven’t they been watching? Not just the Malaysians but all those accredited reps.

    I agree that that may be the reason but note that this is the formal report to the ICAO, whatever the reason. It is this that many will refer to.

    Some diatribe pending about damage assessments.

  65. DennisW says:

    @VictorI

    BTW, I took Methods of Theoretical Physics from Feshbach himself, and audited Morse’s thermodynamics classes (awesome man).

    You are a youngster, but a good guy. I am proud of you.

  66. TBill says:

    @David
    “Another weakening has been page 384 where it concludes the most likely reason for the 00:19 log-on was a power cut to the SATCOM and notes that this could have been because, “some form of generator transfer may have occurred.”

    Soemthing like that is what I am thinking could be possible. A psuedo fuel exhaustion to extend/glide far beyond the last ping

  67. Paul Smithson says:

    @TBill. I continue to ponder what scenario might possibly allow flight terminus beyond 7th arc search width without pilot input. Scenarios along the following lines:

    A) power interruption was enough to get the APU to fire up but not enough for A/P to drop and revert to secondary law. For example, as fuel runs out the engine does some stuttering over a period of more than a minute (like a car running out of fuel – cuts in an out for a while before giving up altogether). This is enough to power-cycle the SDU and prompt APU to power up. But engine never spools down enough to interrupt pitot heat. Result is that APU is up and running with A/P engaged and in normal law, enabling a straight glide without power. This could only work if APU auto-start is triggered by power instability or similar rather than total cessation of power from the engine generators. It would also require the BFO to be anomalous and NOT indicative of rapid descent.

    B) the power cycle was caused by flame out of first engine (R), the L engine generators having been isolated. Is it possible (L) power could be re-established automatically AND APU auto-started? It wouldn’t be a normal seamless “hot switch” because L IDG and backup was previously offline/isolated. Bringing L power back on is sufficient to power cycle the SDU but not interrupt the pitot heat so A/P never drops. Plane flies on under L engine for ?5 minutes then glides a couple of minutes with APU running and AP engaged. Again, you would also need 0019 BFOs to be anomalous (no steep descent).

    In either case, my preferred route is over the DSTG hotspot and a further 80-150NM beyond 7th arc, but you could equally posit the same end of flight at any position along the arc. In both instances, the scenario doesn’t quite tie together – you need to power-cycle the SDU but somehow A/P remains engaged [how?] and the 0019 BFO has to be “wrong” [how?]. Perhaps others can come up with a variation on the theme that ticks more boxes.

  68. Gysbreght says:

    The ATSB observed in one simulation conducted in 2016 high rates of descent and downward vertical acceleration in an ‘unstable’ phugoid. The cyclic nature of phugoid motion implies that half a period later in that phugoid the aircraft would be climbing and the acceleration would be upwards.

  69. Richard Godfrey says:

    The age of the barnacle colony found on the Flaperon is put at 476 days.

    “The size of the biggest specimen indicated that the initial settlement could have occurred 15-16 months prior to the discovery of the flaperon at Reunion Island.” Main Report – Page 396

    “By extrapolation, the colonisation of the larger specimens (36 mm) on the flaperon dates back 476 days (Figure 6), i.e. 15-16 months.” Appendix 2.6A – Page 8

    “The growth of the specimens stopped on the day of the discovery of the flaperon on Réunion Island, on 29/7/2015. The initial colonisation therefore dates back to 10/04/2014, or 33-34 days after the disappearance of the Boeing 777-200 on flight MH370.” Appendix 2.6A – Page 9

    The growth of the barnacle colony indicates that the sea water temperatures at the start of the growth were 28.5°C ± 1°C and at the end of the growth were 25.4°C ± 1°C.

    “At the beginning of their growth, the barnacles were immersed in waters with a temperature close to 28.5 +/- 1°C. Temperature distribution maps in the months preceding the discovery of the flaperon suggest that it has drifted in waters located East-North East of Reunion Island.”

    “Temperatures during the growth of the youngest valves and the terminal fringe of the biggest adult valves (25.4 +/- 1°C) were consistent with temperatures observed off the Reunion Island. These results suggest that the barnacles ended up their developments in waters whose thermal characteristics were similar to waters close to Reunion Island, before the discovery of the flaperon.”

    The Flaperon was found on Reunion at 20.9162°S 55.6492°E. The average sea water temperature taken from 4 GDP drifter buoys passing Reunion in the month of July between 19.9°S and 21.9°S and 54.6°E and 56.6°E was 24.685°C. The average sea water temperature taken from 4 GDP drifter buoys passing Reunion from the north east in the month of June between 18.9°S and 20.9°S and 55.6°E and 57.6°E was 25.426°C. The average sea water temperature taken from 3 GDP drifter buoys passing Reunion from the south east in the month of June between 20.9°S and 22.9°S and 55.6°E and 57.6°E was 24.646°C. A trajectory of the Flaperon approaching Reunion from the north east fits with the sea water temperatures of the recent barnacle growth found on the Flaperon.

    The Flaperon originated somewhere near the 7th Arc on 8th March 2014. The barnacle growth started around 1 month later. The average sea water temperature taken from 6 GDP drifter buoys passing within 1 months journey time of 40°S near the 7th Arc in the months of March/April was 14.759°C and a maximum of 18.023°C. The average sea water temperature taken from 6 GDP drifter buoys passing within 1 months journey time of 30°S near the 7th Arc in the months of March/April was 21.700°C and a maximum of 24.522°C. The average sea water temperature taken from 4 GDP drifter buoys passing within 1 months journey time of 20°S near the 7th Arc in the months of March/April was 25.842°C and a maximum of 27.364°C. By 15°S the average sea water temperature is 27.715°C and at 10°S is 29.400°C.

    This would suggest that the Flaperon originated from waters north of 20°S.

  70. David says:

    @Nederland. The Malaysian report effectively gazumped the French report (Appendix 1.12A-2 Item1) and its flaperon-down assessment, citing its position when there was a collision with the flap.

    For my part the ATSB went too far in its statement that the flap and flaperon separated when not deployed. At the most it should have said that the two collided when not deployed. Separation of one might well have preceded that and in fact neither may have separated as a result of the collision. Also they were not in the neutral/retracted positions: had they been the bottom of the flap would not have impacted the top of the flap.

    Besides, there is more to that as is made clear in this French photo:
    https://www.dropbox.com/s/8og6ss1a8muuazw/French%20flaperon%20report%20page%2058.jpg?dl=0

    The ‘spar’ projection forward of the flaperon rear spar (that the ATSB indicates holed the flap) is bent in compression and has hit something sideways and in the opposite direction to the rear spar damage. However there is no sign of that impact on the flap forward (which would be at or above the height of the rear spar penetration (the photo below shows the flap inverted)) or of what tore off the skin between those two projections:
    https://www.atsb.gov.au/media/5772936/mh370_inboard-section-of-flap-9.jpg

    Again for my part the ATSB conclusion re the collision it depicts is in doubt. However that does not mean the flap could have been extended during a ditching. Its retraction is demonstrated by its internal damage, from the fully inserted (or close to) auxiliary support track, given the likelihood that that occurred at separation. However it does mean that the ATSB’s flap/flaperon collision might have nothing to do with it: either flap or flaperon could have been well separated by the time of any collision between them.

    As to the Vance theory that the flap was earlier forced from deployed to retracted in a ditching, aside from the mechanical difficulty of forcible reversal of the flap’s actuating ball screw there is a brake to inhibit that. Had the design intended that retraction could alleviate overload, there have been no need for the automatic retraction with airspeed arrangement it has to prevent that.

    More generally what the French report offers is thoroughness. That has evaded the Malaysians – and also the ATSB, to the extent that any detailed damage assessment was its responsibility for the parts it identified.

    The damage assessments in this Malaysian report are superficial. ICAO should expect better under its Annex 13 requirements. For example, what of the relationship between flaperon separation and that of the panel above it which separated? What of the implications of the flap and flaperon trailing edges where they are proximate, ie the widest sections of both, not actually being broken off as widely as where they both are narrower? As to timing and cause, what to make of the flap internal damage, of the flap chord-wise break at the outer end of this flap section and the twist and bend in the pivot link there?

    Where is there mention of the vortex generator, reportedly made available for collection by Madagascar some time ago? Where is the damage report on that?

    Is it likely that panels came off the engine on a ditching, as with the Comoros 767? What of the sizes of the biggest of these and that of the flap fairing pieces, the flaperon and the flap. Mid-air break up?

    Another posting addressing specifics of the French report will follow.

  71. Sabine Lechtenfeld says:

    @Richard Godfrey, thanks for going thoroughly through the barnacle assessment.
    Doesn’t the conclusion that the flaperon originated from waters north of 20°S pose a bit of a problem? It’s nevertheless a very interesting conclusion. I have been in contact (through my sister who is a marine biologist and scientific journalist) with a German barnacle specialist back in 2015 after the flaperon has been discovered. After having assessed the barnacle situation as thoroughly as it is possible by looking at photographs, he had gone public in German media with the claim that he could help to determine from where the flaperon originated if he could have access to the barnacles. Nothing came of it, but his preliminary analysis is very compatible with the conclusions of the French report. Since the search for the plane has been much further south at the time, he claimed the barnacle situation would tell him that the search was taking place at the wrong locations. He said that the plane must’ve crashed a good deal further north. Apparently his diagnosis solely through photographs has been spot on!
    The Geomar drift studies also seem to have been on the right track.

  72. Paul Smithson says:

    @Richard G. Great insights. That is very interesting indeed. If the findings are reliable in terms of sea surface temp inferred (28.5+/-1) and timing I think it is the best-discriminator yet of probable latitude of origin.

    I have looked through historic sea surface temperature using earth.nullschool.net (link below). Unfortunately it doesn’t contain SST data for April 2014. However, for April in subsequent years, the latitude at which 27.5 and 28.5 deg centigrade sea temperatures were found (in vicinity of 7th arc) are as follows:-

    27.5 degrees: 2014 n/a, 2015 17S, 2016 18S, 2017 16S, 2018 18S
    28.5 degrees: 2014 n/a, 2015 14S, 2016 16S, 2017 13S, 2018 15S
    for 29.5 degree sea surface temperatures, you need to be getting up to 12 or 13 degrees south.

    So at first appearance, this suggests the flaperon had already reached north of 20S, probably nearer 15S, and possibly even further north – by April.

    Do we think that “charted” datum SST matches temperature in the top foot of the sea where the barnacles were growing or is it (like ocean surface current) a thicker layer? If there is a big temperature gradient in the top few feet of sea then maybe this data is not quite as exciting as it appears.

    https://earth.nullschool.net/#2015/04/07/2100Z/ocean/primary/waves/overlay=sea_surface_temp/equirectangular=100.47,-25.93,497/loc=103.261,-14.026

  73. David says:

    @Nederland. I hope you do not mind me directing this all at you for no apparent reason.
    In more detail about the French report on the flaperon I add:
    First, they seem unaware that its inboard hydraulic actuator would have been in by-pass unless the APU was running still. That means that it would take no loading from the flaperon and the outer actuator would take about twice normal. There would be torsion in the flaperon in consequence. It also means that the inboard actuator would extend freely on flaperon hinge breakage: the outer actuator would not. This would lead to lesser loading of the inboard hinge, at the expense of the outer, which also carries the load from a bigger area. So, this is likely to break first, though hinge bending also played a part.

    Second, unless there had been a control input since the autopilot disengagement on left engine run down, there would be no movement of the flaperon from that last commanded, any deflection from neutral being for bank control at that earlier instant.

    Third, they assess the added mass needed for the broken trailing edge to be underwater, their expert Dr Poupin indicating that being necessary for barnacles to attach and grow there as they have. They find from flotation tests that this would be 5 kg. From x-rays disclosing residual salt in the honeycomb they determined there would have been 3.1 kg of salt water which then had evaporated, leaving that salt. Water ingress is what I predicted would cause the trailing edge submersion, nullifying flaperon windage. I believe the difference needed (the trailing edge must have been submerged by something), well could have come from salt water which subsequently leaked out very slowly, so leaving no salt residue as a tell-tale.

    Fourth, to me the centre of gravity they portray of the recovered flaperon is too far aft to suppose the flaperon would descend nose down as they conclude (subject to experimentation), if it came from a height. To me most likely it would spiral.

    Fifth, no notice has been taken hitherto of these scrape marks of increasing amplitude down the flaperon’s top inboard rear surface:
    https://www.dropbox.com/s/vd5tj2gl6wdyq7d/Flaperon%20scrape%20strike%20damage%20inboard%20end%2C%20French%20report%20page%2041.png?dl=0

    If these were caused by it being struck, or striking, part of the inner flap or its mechanism, judging by the looks of the their witness lines and impact the last strike might have taken off the flaperon trailing edge, or a fair chunk of it. A failure like this would be consistent with a thrust upwards resulting in a strike from above at the inboard end.

  74. Victor Iannello says:

    @DennisW: Thank you for your comment. An individual that is 57 isn’t often referred to as a “youngster”.

    Herman Feshbach was a force at MIT, and was Chairman of the Physics Department when I was an undergraduate. A candid statement he made (not pertaining to me in particular) in a small gathering was also the reason that I did not major in Physics, which is what I intended to do when I entered. But that’s another story best told over a beer.

  75. Victor Iannello says:

    @David said: The simulations are years out of date if that is the reason and the Holland work likewise. Haven’t they been watching? Not just the Malaysians but all those accredited reps.

    My guess is that major portions of this Safety Report were written years ago, and there was no attempt to update those sections as more information and new insights came to light. For instance, the Appendix containing the Boeing analysis looks as though it was not updated since 2014.

    I have done more work analyzing the end-of-flight simulations. After the excitement regarding the Safety Report passes, I’ll share some of that.

  76. Richard Godfrey says:

    @Paul Smithson

    The GDP Drifters measure the sea water temperature between 20 cm and 30 cm below the surface.

  77. Don Thompson says:

    @Andrew,

    Iridium & Calling Line Identity Presentation: it’s my understanding that CLIP was incorporated on Iridium network prior to Oct 2013 per an update to the Iridium ISU AT Command Reference, dated 2014. Iridium SDUs employed a common, reference design L-Band Transceiver (LBT) for access to the network.

    An Iridium SDU fitment was to be compatible with an ARINC 741 install which does not define a requirement for CLI.

  78. TBill says:

    @Paul Smithson
    “So at first appearance, this suggests the flaperon had already reached north of 20S, probably nearer 15S, and possibly even further north – by April.”

    That starts to get very close to Xmas Island and Jean-Luc’s MH370-Captio.net explanation for the flaperon. There was the tropical storm bearing south so northward motion seems a little hard to imagine on the surface . Also 38S end point is weakened.

    Also @Richard’s numbers of barnacle growth time tend to say SIO crash location makes perfect sense instead of a conspiracy to plant fake aircraft parts in the SIO.

  79. Nederland says:

    @David

    Thank you for your detailed observations on the flaperon report. What would be your main conclusion? Was the system still powered or not?

  80. Richard Godfrey says:

    @TBill

    You stated “@Richard’s numbers of barnacle growth time tend to say SIO crash location makes perfect sense instead of a conspiracy to plant fake aircraft parts in the SIO.”

    I agree with your conclusion that fake aircraft parts were not planted.

    I believe the discoveries by Johnny, Blaine, Schalk, Niels, Liam and many others were genuine finds by observant and diligent citizens.

  81. Victor Iannello says:

    @David said: Azharuddin Abdul Rahman has resigned.

    That’s unfortunate. From my perspective, he was one of the good guys that sincerely was trying to solve the mystery, and he also supported the follow-on search by Ocean Infinity. There had to be a fall-guy, and he was it. On the other hand, the military comes away from this scot-free, despite many deceptive and conflicting claims. For instance, the report claims that the military tracked MH370 in real time, and chose not to intercept it because it was deemed to be friendly. Yet, the SAR operations proceeded in the South China Sea for days. The timeline of events simply does not match the military’s claims.

  82. Dan Richter says:

    @Paul Smithson
    So at first appearance, this suggests the flaperon had already reached north of 20S, probably nearer 15S, and possibly even further north – by April.
    August 1, 2018 at 6:45 am

    Based on my search of 20 051 NOAA buoys with 50 mil. GPS positions I found common debris origin at position 11S 84E.

    Map shows how many NOAA buoys originated in SIO from the LAT LON 1°x1° for: Flaperon; Flap fairing, Tailstabilizer, Engine cowling, Wing flap and Mossel Bay debris to their place of finding (rectangle area +- 1.5°). The final map is a result of multiplication of number find buoys in given SIO rectangle 1°x1° for each debris.

    NOAA buoys
    https://www.2i.cz/images/2018/07/24/MH370_6_1184.jpg

    KMZ file here (use slow download)
    https://ufile.io/1xx0a

  83. Paul Smithson says:

    @Richard G. Thanks once again for highlighting these findings and for your response above. If there aren’t unforeseen problems with this inference of sea temperature to which flaperon was exposed in April 2014 then I think you have put the spotlight on uniquely powerful evidence of a (very) northerly flight terminus.

    A couple of further questions. Does the report shed any light on the estimated duration of exposure to 28.5 degrees +/- 1, or confidence limits on the start of this period? The more tightly that period can be pinned down the more useful the information will be.

    Do you know of any public-access charts of SST temperatures for the relevant area and time period? It’s frustrating that earth.nullschool SST data doesn’t seem to go back to 2014.

  84. Paul Smithson says:

    @TBill. It’s ironic that Richard’s post came so soon after my musings on 38S! If the “flaperon sea temperature in April” is correct, then a southern terminus can be safely and finally ruled out and we could have a high level of confidence that the terminus lies to the north. It might take a little while to “bracket” with greater confidence the latitudes that are consistent with the barnacle evidence. But I suspect we will be looking at 20S and northwards, potentially all the way up to Indonesian waters. Perhaps subsequent debris drift trajectories could help to narrow this down further.

  85. Don Thompson says:

    @Paul S,

    SST: see GHRSST.org. NOAA hosts an archive of their products.

    Discussing ACARS, Victor replied “Nonetheless, not being able to test this configuration introduced a tremendous amount of risk. Would the captain have done this? I don’t know.

    I don’t know either. I’m simply suggesting that it should not be assumed that the perp had deep knowledge of the operation of ACARS over SATCOM. If there was any prior planning, the datacom system operation may have been misunderstood and not even deemed a risk to ‘the plan’.

  86. John McGrath says:

    Just a note referring back to radar data and the lack of data from the Australian Jindalee (JORN) system. It seems to me to be unlikely that there was no data al all from this system. Perhaps more likely that the Australian government is trying to hide the full capabilities of JORN? This is an excerpt from the Wikipedia page (I Know, not a truly reliable source!)

    URL: https://en.wikipedia.org/wiki/Jindalee_Operational_Radar_Network
    Theories regarding Malaysia Airlines Flight 370[edit]
    In May 2016, the JORN FAQ file / Fact Sheet was updated by the RAAF to address questions about Malaysia Airlines Flight 370. According to the update, “JORN was not operational at the time of the aircraft’s disappearance.” The update also stated that MH370 would have been unlikely to be detected by the system due to radar range, ionospheric conditions and a “lack of information on MH370’s possible flight path towards Australia.”[33][34]
    However, in the immediate aftermath of the March 8, 2014 disappearance, information regarding JORN’s status was not released. This led to months of speculation. On March 18, 2014, sources cited by The Australian said that JORN was not tasked to look toward the Indian Ocean on the night of the disappearance of MH370 as there was no reason for it to be searching there at that time.[35] On March 20, 2014, it was reported that Malaysian Defence Minister (also Acting Transport Minister) Hishammuddin Hussein requested the US to give any information from the Pine Gap base near Alice Springs, possibly alluding to JORN as well.[36] On March 19, 2014, it was reported that an Australian Defence Department spokesman said it “won’t be providing comment” regarding specific information on tracking MH370 by JORN.[37] However, several days prior to this, Australian Foreign Minister Julie Bishop told the Australian Parliament, “All our defence intelligence relating to Flight 370 has been, and will continue to be, passed on to Malaysian authorities…”[35][36]
    In March 2015, before the July 2015 discovery of MH370 debris on Réunion Island, aviation technology expert Andre Milne requested information from JORN to prove or disprove that the aircraft ended up in the Indian Ocean,[38] but he received no response from the Australian government in 2015. It was made public in May 2016 that JORN was not operational at the time of the disappearance.[33]

  87. Don Thompson says:

    @John McGrath

    Wikipedia needs an edit. I suggest that Milne received no response as he has long been recognised as a delusional crank.

  88. Paul Smithson says:

    @Don. Thanks for the link, though I haven’t been able to use it to find SST charts for the Indian Ocean in April 2014 and I don’t have the time right now to figure out how. The long and short of my earlier review of earth.nullschool SST for 2015,16,17,18 is that a temperature of 28.5 +/-1 in April typically corresponds with ~15S, +/- 2 degrees. And if rate of drift gives you (say) an average of 0.2kts northing over ~40 days, that would point to terminus at 18S +/-2 degrees. All very thumb-in-the-air for now…

  89. Don Thompson says:

    @Paul,

    I overlooked that you were looking for charts, it appears that the archives hold data in ‘netCDF’ format. Sometime through this odyssey, I vaguely recall extracting data from such files.

  90. TBill says:

    @John McGrath
    In addition to what Don said, Australia has said JORN was not believed to be operating that night and even if it was they would have had to know in advance about where and what to look for. So are not expecting any secret data is being hidden by Australia from JORN.

    We only wish we had as much clarity about Indonesia/Singpaore AWACS radar.

  91. Victor Iannello says:

    @Andrew: It may be a minor point, but after looking at the simulations that were performed for the turn at IGARI, I do think the turn could have been accomplished with the autopilot engaged. Just before initiating the turn, the initiation of a V/S climb to FL400 at 6,000 fpm would reduce the speed and allow a smaller radius turn at the 25° bank limit, and match the timing of passing the custom waypoints they considered.

  92. DennisW says:

    @all

    Well the last few days have certainly been unsettling to me. It is hard to say the information we have now is new information (maybe it is new to us). You have to ask why the search went the way it did.

    It could simply be that the underwater search was started too soon, and the effects of “sunk cost” dominated the thinking of the decision makers. It is very hard to abandon a course of action you have spent considerable funds pursuing. When the flaperon was found the conclusion of the drift modelers was that it “could have come” from the area being searched. That conclusion morphed into the flaperon “supporting” the area being searched. Studies which showed that the priority area was low probability were basically ignored.

    There was never, IMO, a moment when the initial priority search area was seriously reconsidered.

  93. TBill says:

    @Victor
    Re: IGARI turn
    I am very interested in that as I did many turn cases myself.
    I’ve been doing it with 25 deg bank in A/P, but I did not have the military radar data all I had was the civil radar data. So I have to study.

  94. Niels says:

    @VictorI
    Regarding your Oct. 2017 post: Would it be possible to calculate the mean BFO error and the variance of the BFO error for each of the 16 LRC based paths in addition to the information provided in fig. 2?
    I would be interested to see if there is a minimum in the variance and for which 00:19 latitude (and mean BFO error) this would be.

  95. Sabine Lechtenfeld says:

    @Dennis, I agree with your sentiments.

  96. Victor Iannello says:

    @Niels: In the post, I supplied a CSV file the included the RMS error for each path. If you are looking for the contributions to the RMS error, i.e., mean error and variance about the mean, I will have to do some more digging.

  97. Niels says:

    @VictorI
    Victor, yes, I’m interested in the mean error and the variance about the mean. It could possibly give some information about the bias frequency offset (after 18:25 logon), although oscillator drift will “contaminate” the info.

  98. DennisW says:

    @Niels

    Regarding your Oct. 2017 post: Would it be possible to calculate the mean BFO error and the variance of the BFO error for each of the 16 LRC based paths in addition to the information provided in fig. 2?

    Mean and variance are possible to compute relative to BFO. However, those metrics are meaningless for a non-stationary and non-ergodic process.

  99. Niels says:

    @DennisW
    I understand your remark, however with the wide range of latitudes we are now considering (and the large values of mean BFO error implied for “straight path” solutions relative to possible oscillator drift) I don’t think these numbers are necessarily completely meaningless.

  100. Niels says:

    @Richard Godfrey
    While I agree that the data on surface water temperature derived from the isotopics analysis on the barnacle capitulum is useful, I’m not convinced by the growth time estimate (fig. 6 in appendix 2.6a) and the logarithmic growth curve fit. The reference is based on a different genus (L. anserifa) and perhaps more important seems to be based on three data points for just one specimen. A good reference should be based on many specimen and include error bars to indicate inevitable variations in size. So the 476 days estimate to reach 36 mm should be considered, at best, as a very crude estimate by Dr. Poupin. Personally, I’m not convinced that this 476 days estimate could be used to indicate a crash location.
    Perhaps with better quality growth curves for the right genus and for a range of temperatures, a better estimate of the growth period could be made.

  101. paul smithson says:

    Thanks @niels. These comments are exactly the sort of close critique to which the barnacle temp/time evidence needs to be subjected in order to understand how tightly it might constrain terminus latitude. I haven’t had time yet to read in detail.

  102. TBill says:

    @all
    @Paul Smithson
    Jeff Wise has now chimed in with his different interpretation of the barnacle growth, which is I think probably conlficts. So we need a referee.

  103. Richard Godfrey says:

    @Paul Smithson, @Niels,

    I stand by my previous statement, that with the benefit of hindsight, I conclude that the MH370 end point must be near the 7th Arc between 20.0°S and 25.0°S, with a hot spot around 21.5°S.

    I agree that barnacle growth is nothing more than a hint that the MH370 end point is further north than 25.0°S.

    I also agree that the drift analysis and barnacle analysis are both subject to many unknowns and assumptions.

  104. Niels says:

    @Richard Godfrey

    With reference to a previous comment:
    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17198

    I’m a bit surprised you already came to a conclusion; can you please explain how you arrived at that position?
    I thought the drift analysis of the flaperon with windage modelling in accordance with Griffin et al. was not yet completed.

  105. Richard Godfrey says:

    @Dan Richter

    You stated “Based on my search of 20 051 NOAA buoys with 50 mil. GPS positions I found common debris origin at position 11S 84E.”

    11S 84E is 2,559 km west of the 7th Arc at 11.0°S 107.4448°E. Are you saying the satellite data is wrong?

  106. Victor Iannello says:

    @TBill: We have two French biology studies saying different things. Poupin says the largest barnacles were around 36 mm, indicating an age of 476 days. Blamart and Bassinot say that the largest barnacles were 20-25 mm, indicating an age of several months. The data set from an earlier study (Evans, 1958) used by Poupin to determine the age consisted of 3 points from a closely related species. The points, expressed as (mm,days) were (8,8), (21,40), and (25,107). Poupin extrapolated these points using a log relationship to (36,476).

    I don’t see how any of this can be used with any level of certainty to estimate the age of the barnacles. The growth rates flatten with time, and small differences in size lead to large differences in age. Of course, this kind of ambiguity allows people “flexibility” in their interpretations.

  107. flatpack says:

    Data 3

    I wrote upthread that it was depressing to contemplate and I think that my tentative conclusions are not going to be welcomed by the ‘alternate but rogue safe landing’ crowd.

    Excerpt from the Safety Investigation Report (emphasis mine):-

    “The IFE equipment set up two ground connections over SATCOM [for the SMS e-mail application and Built-In Test Equipment (BITE) application] after the SATCOM re-established the link at 1825 UTC, 07 March 2014 (normal), but not after the SATCOM re-established the link at 0019 UTC, 08 March (abnormal). At no time during the flight was any user data sent over the link by means of the SMS/e-Mail application.

    I conclude that the PAX and cabin crew were incapable of using the SMS/e-Mail application by 1825 UTC.

    If there were a go/no-go message, then it would have had to be around 1800 UTC at the latest (and thus probably not via SATCOM).

  108. Richard Godfrey says:

    @Niels

    You stated “I’m a bit surprised you already came to a conclusion; can you please explain how you arrived at that position? I thought the drift analysis of the flaperon with windage modelling in accordance with Griffin et al. was not yet completed.”

    In my view, there are automated flight paths that end north of 25.0°S near the 7th Arc, that have not been previously considered in the underwater search for MH370, such as posted by Victor Iannello. The drift analysis, that Victor kindly published, supports a MH370 End Point between 20.0°S and 25.0°S. Further analysis regarding the difference between drogued and undrogued buoys, did not substantially change the proposed search area. I also believe that the barnacle analysis does not help to define a MH370 End Point.

    I believe that the interpretation of the BTO data leads to a clear definition of the 7th Arc as published by Mike Exner. I do not believe the aircraft was actively piloted after 19:41 UTC. I believe the interpretation of the BFO data shows a steep descent at 00:19:37 UTC and that the search width of 22 NM is adequate. In my view, it is highly unlikely, after fuel exhaustion, that the aircraft first entered a steep descent and then recovered to glide further away from the 7th arc, than was previously searched. I do not believe that Fugro or Ocean Infinity missed the underwater debris in their searches to date.

    I have explained my position in a 83 page paper and 48 comments in the last 3 posts on Victor’s blog.

    I am not stating like Griffin et al. that the MH370 End Point can be precisely defined, such as “This location is 35.6°S, 92.8°E.” I am stating that the MH370 End Point is between 20.0°S and 25.0°S within 22 NM of the 7th Arc, with the benefit of hindsight and discounting other options.

    I am not saying I am right and respect that you obviously have a different view.

  109. flatpack says:

    Freefall

    Previously on this thread (July 31, 2018 at 12:42 pm) Gysbreght posted a link to the Military Radar Data plotted to scale.

    To me sawtooth = subjugation.

    To anyone who questions that MH370 really got down in the weeds sometime around 1800 UTC, I suggest that they review the Safety Investigation Report section 1.1.5 titled Detection of Hand Phone Signal.

    Note that those tests were carried out using a King Air 350 which I suspect would have lower effective RF shielding for a typical seated passenger holding a cellphone normally.

  110. Victor Iannello says:

    @flatpack: Do we know for sure that the IFE on 9M-MRO allowed passengers to send email and text messages? My recollection is that there was some capability to use credit cards for purchases, but perhaps somebody else knows the precise information.

  111. Victor Iannello says:

    @flatpack: The tests in the King Air 350 were to look for connected phone calls, which requires a much longer connection than a registration on the tower. As I have said before, I have experienced momentary connections in-flight while cruising at 35,000 ft. I know the connections occurred because I got text messages welcoming me to the cell network of a particular country.

    As for the altitude progression measured by the military data, the most likely conclusion is the military radar system cannot discriminate altitude, as the change in altitudes combined with the speeds are impossible for a B777, “subjugation” or not. Either that or other aircraft than 9M-MRO were captured.

  112. Dan Richter says:

    @Richard Godfrey says:
    August 2, 2018 at 2:02 pm

    @Dan Richter

    You stated “Based on my search of 20 051 NOAA buoys with 50 mil. GPS positions I found common debris origin at position 11S 84E.”

    11S 84E is 2,559 km west of the 7th Arc at 11.0°S 107.4448°E. Are you saying the satellite data is wrong?

    11S 84E is just a common debris origin i.e. most of the debris had to travel through this position but it doesn’t mean it is the crash site. I don’t trust in the satellite data. What surprise me that Safety Report doesn’t mentioned reliable eyewitness Mike McKay who saw “burning object” and wrote a letter to Vietnamese authorities about his sighting. On September 22, 2014 he wrote to me:

    Dan,
    I wrote what I thought might help the search but within two days of my email being leaked the search in the South China Sea was called-off. On the basis of the (belated) primary military radar readings coming to light, the search was then moved to the Andaman Sea. I think the South China Sea search was called-off prematurely and as such I would agree with your assessment. Acting on my sighting, the Vietnamese sent out one flight only (six days after MH370 disappeared) and then were told to stop looking.
    I have been hoping for some evidence to prove that is was not MH370 I saw, but unfortunately there has been no proof yet.
    Mike.

    Possible debris route through Sunda Strait {confirmed by NASA too, http://oceanmotion.org/html/resources/drifter.htm select March and click on map near Vietnam coast)
    https://www.2i.cz/images/2018/04/05/MH370_Mike_Sunda_Reunion.jpg

  113. Victor Iannello says:

    @Dan Richter: This is not Facebook, where all theories, no matter how unlikely, are treated equally.

    Please explain how MH370 could have crashed in the South China Sea without the discovery of debris along the shores of Malaysia, Indonesia, Thailand, Vietnam, or Cambodia. Please explain the capture of the aircraft by civil and military radar. Please explain why the Inmarsat BTO and BFO values are incorrect.

  114. TBill says:

    @flatpack
    DrB covered some of the IFE in his known facts posts, and I threw some comments about the IFE. MH370 had the most basic IFE, which if I recall, allowed SMS/email to first class, and only if that capability was activated. So the lack of messages is perhaps not too surprising. Might be interesting to compare IFE SMS records to MH371 (9M-MRO prior flight from Beijing same day) if we have that info from Inmarsat.

    I think you have made some good points above…you may be onto some good thinking. As far as contact, I wonder if Iridium phone could be to work as a clandestine approach, with some kind of window antenna.

  115. flatpack says:

    @Victor Iannello

    From page 121/122 :-

    The SATCOM provides the Satellite link for the following functions:
    …snip…
    • Cabin Packet Data (Data-3) – Interface via the Panasonic System 3000i IFE equipment:
    – SMS/e-Mail
    – BITE-offload

    And from page 123 :-

    “At no time during the flight was any user data sent over the link by means of the SMS/e-Mail application.”

    I do not know for sure that the IFE on 9M-MRO allowed at least some passengers to send email and text messages however my view is that the specific references in the Safety Investigation Report would not have been included were it not so.

    It is possible that only a subset of passengers had the full IFE functions (such as SMS/e-Mail) available to them.

    I do recall reading somewhere about the BITE because I was curious about ‘offload’ but cannot presently find the details.

    Any BITE messages should been reviewed because it would seem possible that pulling circuit breakers would leave the IFE OS with a powerfail flag on re-start, perhaps triggering a file system check. This may not be evident as the dump may be cryptic or even deliberately obfuscated.

  116. Victor Iannello says:

    @flatpack: Yes, I know what’s in the Safety Report. That doesn’t tell us definitively that there was SMS/email capability that was available but not used by the passengers.

  117. flatpack says:

    @Victor Iannello

    You wrote;-

    ‘The tests in the King Air 350 were to look for connected phone calls, which requires a much longer connection than a registration on the tower.’

    Indeed. One wonders why the tests were so conducted when, according to the RMP, there was only a registration on the tower.

  118. Victor Iannello says:

    @flatpack: I wonder about a lot of things. For instance, look at this exchange between Grace Nathan, who is the daughter of MH370 passenger Anne Daisy, and Malaysian officials during the recent briefing with the NOK.

    When asked why the plane flew so high or so low at certain points the answer received was that the speed or height of the aircraft did not concern the safety team as they were just concern[ed] with the location of the aircraft o establish whether it turned back. Which demonstrated a lack of concern with how the height and speed affects fuel consumption which in turn may lend clues as to when and where the plane ran out of fuel.

    When asked why no other phone numbers were investigated other than the Co-pilot’s, the response was ‘that’s a good suggestion.’

    Read Grace’s entire comment.

  119. flatpack says:

    @TBill
    @Victor Iannello

    TBill wrote:-

    ‘MH370 had the most basic IFE, which if I recall, allowed SMS/email to first class, and only if that capability was activated.’

    I appreciate that features could be selectively ‘activated’ however the report states that the IFE equipment set up a ground connection over SATCOM for the SMS e-mail application…

    Hey ho, I suppose it could be a bug.

  120. flatpack says:

    @Victor Iannello

    Thanks for the link. I was going to comment; ‘That’s curious’. Perhaps ‘That’s incurious’ would be more apposite.

    One thing I wonder about is the report I read quoting the security guard at the pilots housing complex as noting that he gave a military salute on passing through the security barrier on the way to the airport that day, rather than the customary muslim salutation.

    The report on the other hand seems to find no change in behaviour presaging the events.

  121. Niels says:

    @Richard Godfrey
    Thank you for explaining. In fact there are some important points where we agree. Like you I think the 7th arc is clearly defined and that it is likely that the crash location is within 22 NM around the arc.
    However, I’m still considering different explanations why the wreckage has not yet been found. This includes the possibility that Fugro or OI were not able to detect the wreckage in difficult terrain.

    Your drift analysis is very important both the compare with CSIRO results as well as to extend CSIRO results to a larger range of latitudes. You state that “The drift analysis, that Victor kindly published, supports a MH370 End Point between 20.0°S and 25.0°S. Further analysis regarding the difference between drogued and undrogued buoys, did not substantially change the proposed search area”.
    This is where I have concerns. First of all, your additional calculations show that there is a significant difference in expected displacements based on drogued and undrogued drifter data, see for example:

    https://www.dropbox.com/s/3trlgjev99qcat8/Drogued%20vs%20Undrogued%20Days%20to%20Reunion.png?dl=0

    https://www.dropbox.com/s/4xars6l8xn5izah/Drogued%20vs%20Undrogued%20Days%20to%20Africa.png?dl=0

    Secondly, from our discussion it became clear that the effects of wind and waves had not yet been properly incorporated in your drift analysis of the flaperon. It is especially the estimated transit time of the flaperon which in CSIRO’s analysis is used to restrict the origin of the debris to south of S30 degrees. Therefore, I really encourage you to also complete the drift analysis for the flaperon. The results may indicate that CSIRO overlooked a crucial range of latitudes in their analysis, however this can only be concluded if your methodology is sound.

  122. Don Thompson says:

    @Victor, flatpack

    Onboard video clips from two MAS 777-2H6ER passengers showing the IFE ‘Connect’ menu. Onboard an MH141 service and MH51.

    Chargeable at $1.50 using a credit card. One of those clips may show the c/card swipe slot in the side of the IFE handset.

    It’s my understanding that the ‘IFE/PASS Seats’ switch on the flight deck P5 panel does not isolate core IFE system components such as the EPESC involved with initiating the Data-3/ISO-8208 SSN connections. If that switch was released, the core systems would continue to function but not the seat components.

  123. David says:

    @Victor. Re Grace’s comment, Larry Vance has suggested there be an ICAO initiated inquiry.

    I have been touting for a Malaysian royal commission but that is dreaming.
    Possibly beyond the scope here I think but on this Vance has a point. However I do not know that the ICAO has the power, much less motivation.

    From yesterday’s ‘The Australian’ (Ean Higgins again)

    “One of the world’s most respected air crash investigators has called for the international aviation watchdog to commission a fresh and independent investi­gation into the disappearance of Malaysia Airlines Flight MH370, saying the Malaysian government’s inquiry failed to properly assess the evidence.

    Canadian Larry Vance, who led some of the biggest inter­national air accident investi­gations over three decades, also said a “comprehensive criminal investigation” was required into the loss of the aircraft, in which 239 people perished in an event a Malaysian government report this week acknowledged involved human intervention.

    “An organisation such as the International Civil Aviation ­Organisation should do a thorough inquiry into the circumstances of this occurrence and the investigation that followed,” Mr Vance told The Australian.

    “If there is no such inquiry, then aviation safety has taken a step backwards.”

    The Malaysian investigation found the aircraft, bound for Beijing from Kuala Lumpur, was ­deliberately flown off course just as it disappeared from secondary radar with its transponder turned off on March 8, 2014.

    Primary radar and automatic satellite “handshakes” determined it flew back over the Thai-Malaysian border and up the Strait of Malacca, before another turn on a long track south to end in the southern Indian Ocean.

    The Malaysian report said investigators could not conclude why the aircraft disappeared, but effectively excluded mechanical failure and accidents such as suggestions a cargo of tropical fruit and lithium batteries combined to cause a fire. Mr Vance described these ­aspects as “positive”, saying the report “helps put to rest many of the speculative and far-out-there theories that have circulated about what might have happened”.

    “The report also makes it clear that this was the result of human intervention, and not some sort of mechanical event or an intervention from outside the aeroplane.”

    Mr Vance disagreed with the statement of chief Malaysian investigator Kok Soo Chon when he in effect excluded the two ­pilots, particularly the captain, Zaharie Ahmad Shah, from the widely held suspicion that one of them hijacked the plane.

    Instead, Mr Kok said, it was possible an unknown “third party” had intervened.

    Mr Vance and other aviation experts have rejected the “third party” possibility as fanciful, and claim Mr Kok emphasised it to cast doubt on the dominant view that a pilot from the Malaysian government-owned airline had taken 238 other people to their deaths.

    Mr Vance said the Malaysian report failed to properly assess the implications of parts of the aircraft found washed up and recovered on the other side of the Indian Ocean, which he said conclusively proved a pilot flew MH370 to the end and ditched it.”

  124. Victor Iannello says:

    @flatpack: Regarding the PIC’s psychological profile, the Safety Report states:

    The PIC’s ability to handle stress at work and home was reported to be good. There was no known history of apathy, anxiety, or irritability. There were no significant changes in his lifestyle, interpersonal conflict or family stresses.

    Yet, the report mentions nothing about his relationship with Fatima (Tim) Pardi and her family despite Malaysian officials interviewing her four times. From a story in the The Australian by Amanda Hodge:

    Fatima also alluded to how Zaharie played an almost fatherly role to her children, but that at his request, they decided to take a step back from their close friendship and the relationship he had with her kids, in the weeks leading to the disappearance of the plane.

    She said they had exchanged messages related to a “personal matter” just two days before the ill-fated flight on March 8, 2014.

    However, Fatima, 35, did not reveal the subject of their last WhatsApp discussion before the flight, according to The Australian.

    “That last conversation was just between me and him. I don’t want to talk about it,” she was quoted as saying, adding however that “Captain Zaharie had not seemed stressed”.

    “I’m afraid what I say will be misunderstood. It was a personal matter, a private issue.”

    This interview of course proves nothing. But it does run counter to the assessment in the Safety Report (which was extracted from the RMP Report) that says there were no changes in the captain’s life in the weeks prior to the disappearance. Somebody doesn’t seem to be telling the truth.

    Article in The Australian (sometimes behind a paywall):
    https://www.theaustralian.com.au/business/aviation/mh370-pilots-friendship-with-mystery-woman-revealed/news-story/1e5d5b18a3a87e4765830c311f1e87ac

    Recap article in Free Malaysia Today (no paywall):
    http://www.freemalaysiatoday.com/category/nation/2016/09/07/woman-talks-of-close-friendship-with-mh370-captain/

  125. Don Thompson says:

    To be a little more emphatic, the MAS B777-2H6ER (and 747-400) fleet underwent a cabin upgrade during 2005 and 2006 that included the Matsushita/Panasonic Avionics 3000i IFE system.

  126. Victor Iannello says:

    @Don Thompson: So let me put it a different way. Considering there were no passenger emails or texts after 18:25, we are left with one of the following possibilities:

    1) Passengers were incapacitated at that time.
    2) Passengers were not incapacitated, but nobody chose to pay for an email or text message.
    3) Passengers wished to send messages, but the system was not operational because certain functionality was de-selected by the IFE/Pass Seats switch on the overhead panel in the cockpit.

    Is this correct? If option (3), could that explain the missing IFE log-on at 00:19? Or does the log-on to the IFE occur independent of the switch position?

  127. Don Thompson says:

    Mr Vance said the Malaysian report failed to properly assess the implications of parts of the aircraft found washed up and recovered on the other side of the Indian Ocean, which he said conclusively proved a pilot flew MH370 to the end and ditched it.

    That certainly is a case of the pot calling the kettle black.

    Vance’s assessment, made only with access to publicly available images, has been strongly refuted here.

    Perhaps Vance’s motive has been to position himself as lead for an ‘independent’ investigation.

  128. David says:

    @Don Thompson. Yes I think an investigation into the investigation is warranted, with its disconnections, inconsistencies and inactions but not of course for the Vance specifics of the last paragraph.

  129. Don Thompson says:

    @Victor,

    I agree with your three scenarios.

    Concerning 3): the IFE system components are distributed throughout the aircraft, and the system provides interfaces to the Cabin Service System, the Passenger Address System, among other aircraft systems.

    The AMM information shows that only power for the IFE cabin peripheral and seat position components are isolated by the IFE/Pass Seat control: the Seat Electronics Boxes (SEB, which includes the handset or ‘PCU’), the Area Distribution Boxes, RF distribution amplifiers (for video), video modulator units, and the video system control unit. If the SEBs are powered down, the passenger has no means of input.

    The IFE’s core processing, interface and management LRUs, such as the EPESC and the Pax In-flight Information Contoller (PIIC), are not isolated by releasing the IFE/Pass Seats switch.

    However, ELMS could shutdown the EPESC. But do note that the aircraft started at 2014-03-07 at 12:50, while on the ground at WMKK, on APU power and the IFE system connections were made at that time.

    My assumption has been that the core systems do remain functional even if the peripheral and seat electronics are powered off. Therefore, I do expect that the ISO-8208 connections would be initiated.

    @flatpack questioned what the full 7th March Stratos Log showed of IFE correspondence. During the previous MH371 service each successful GES Log On was followed by the two ISO-8208 connections. The departure message was recorded after take off. A (long) BITE message was recorded prior to landing at WMKK. No pax amenity messages were recorded in the Log.

    ISO-8208 correspondence is readily identifiable in the Log: a T-ch ISU contains a datalink Logical Channel Number that is specific to either ISO-8208 connection.

  130. Victor Iannello says:

    @Don Thompson: Since no email or text messages occurred during MH371, which was normal in all respects, the absence of emails and texts after 18:25 means nothing. However, the missing IFE log-on subsequent to the 00:19 log-on is likely due to either the APU running out of fuel or the plane impacted the ocean.

  131. Don Thompson says:

    @David,

    I’d settle for a competent investigation. There is little point in trying to take the Malaysians to task, and I’m not sure there’s is any supra-national recourse for that.

    The ICAO sets out recommendations for member countries to implement. In 2014, Malaysia had not then acted upon the ICAO’s recommendation that states should establish of an independent air accident investigation body.

    Hence the clumsily titled ‘The Malaysian ICAO Annex 13 Safety Investigation Team for MH370’ still under the auspices of the Ministry of Transport.

    Malaysia quietly instituted an AAIB in 2015.

    The only accident that appears to have been investigated by AAIB-MY is the crash of a Eurocopter AS-365 Dauphin. The investigation’s interim report described that the turbine engined Dauphin was fueled with avgas. A common theme is apparent.

  132. David says:

    @Don Thompson.”I’d settle for a competent investigation”. So would I were one within sight.

  133. TBill says:

    @Victor
    “However, the missing IFE log-on subsequent to the 00:19 log-on is likely due to either the APU running out of fuel or the plane impacted the ocean.”

    Can we say that? I thought the other possibility was IFE was switched off. I admit I am starting to get paranoid the PIC knew we’d think that lack of IFE meant crash near the final ping (fooled us). I still think it could be near Arc7, so we should search.

  134. Don Thompson says:

    @TBill wrote “I thought the other possibility was IFE was switched off.

    Perhaps read the function of the ‘IFE/Pass Seats’ switch as ‘IFE at the passenger seats’, not the entire IFE system.

  135. Andrew says:

    @Victor

    RE: “It may be a minor point, but after looking at the simulations that were performed for the turn at IGARI, I do think the turn could have been accomplished with the autopilot engaged. Just before initiating the turn, the initiation of a V/S climb to FL400 at 6,000 fpm would reduce the speed and allow a smaller radius turn at the 25° bank limit, and match the timing of passing the custom waypoints they considered.”

    Is that something you’ve tried with your simulator? I’d be surprised if it produced a result much different from Sessions 3 & 4 described in the Safety Investigation report, where the speed was allowed to reduce during the turn. The aircraft made the exit waypoint in both those simulations, but took too long to do so. There would also be a significant risk of stalling the aircraft during such a manoeuvre at high altitude. The AFDS is relatively slow to react to changes in the selected VS when operating in VS mode.

  136. Andrew says:

    @Don Thompson

    For what it’s worth, the Boeing Service Bulletin for the installation of the IFE/PASS SEATS and CABIN/UTILITY switches states the following (my emphasis):

    ‘When selected OFF, the IFE/PASS SEATS switch will remove power from all the components of the IFE (in-flight entertainment) system and the passenger seats (seat motor power, personal computer power outlets and telephones).’

    The FCOM also states that selecting the IFE/PASS SEATS switch OFF ‘removes power from… all IFE components.’

    However, neither document defines what ‘all IFE components’ actually means. Are you able to post or email a copy of the AMM info on the switch? My copy pre-dates the SB.

  137. haxi says:

    @All,

    I want to share a link to a new National Geographic documentary on the search of MH370 which I think worth watching. It’s part of Nat-Geo’s “Drain the Oceans” series.

    In it, we see nicely-made CGIs of the ocean floors in and around the search zone, as well as a vivid animation of end-flight scenario based on Boeing’s simulation results.

    It also features Peter Foley, Blaine Gibson and David Griffin.

    https://www.nationalgeographic.com/tv/watch/fe70898178fdcafb39ccb7b324c81e17/

  138. Richard Godfrey says:

    @Niels

    You stated “Secondly, from our discussion it became clear that the effects of wind and waves had not yet been properly incorporated in your drift analysis of the flaperon.”

    I am not aware of any error in my drift analysis of the Flaperon.

    Since it is clear to you that I have made an error, perhaps you could be more specific what the error is?

  139. Richard Godfrey says:

    @Niels

    For the sake of clarification:

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17170

    This post on 29th July 2018 4:52 am is correct together with all its attachments.

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17176

    This post on 29th July 2018 11:24 am was incorrect as I stated subsequently on the 30th July 2018 at 8:05 am and should be ignored.

  140. Dan Richter says:

    @Victor Iannello says:
    August 2, 2018 at 3:55 pm

    @Dan Richter: This is not Facebook, where all theories, no matter how unlikely, are treated equally.

    Please explain how MH370 could have crashed in the South China Sea without the discovery of debris along the shores of Malaysia, Indonesia, Thailand, Vietnam, or Cambodia. Please explain the capture of the aircraft by civil and military radar. Please explain why the Inmarsat BTO and BFO values are incorrect.

    This is easy to explain because of the Bernoulli’s principle. In narrow straits flows water much faster so debris can’t wash up on the beaches there. Although there are many radar stations along the alleged route, most of the radar stations didn’t saw MH370 at all. For example RMAF Air Defence Surveillance, Sqdrn 321 Martello S-743D Radar at Bukit Puteri, Jertih 5.785833N, 102.504444E, which is close to IGARI on the east side of peninsula. I am a specialist in data transmission, who used to work for Motorola and with experience with Iridium satellite system, would never count on Inmarsat BTO and BFO. By the way when I was sailing in 2016 from Cape Verde to St. Maarten in Caribbean in Atlantic Ocean I was able to get the data transmission over Iridium and download GRIB weather forecast (50 kb) without any problems. My colleagues they have two Inmarsat satellite phones were barely able to get the voice connection. They could forget the data transmission. Based on Inmarsat BTO and BFO they found MH370 in 4 years, isn’t it.

    This is the typical movement of NOAA buyos going south from the Vietnam coast. Picture shows 79 NOAA buoys passing +-1° position of Mike McKay burning “object” at 8.3511 N 108.1614 E
    https://www.2i.cz/images/2018/07/22/SCS_Mike_Sunda_Strait.jpg

  141. David says:

    @Nederland. Conclusion re flaperon. Elusive so far.

  142. Don Thompson says:

    @Dan, Dan…, Dan!

    While your image does depict drifter paths extending down the Karimata Strait, there is no track that tracks out into the Indian Ocean via either the Lombok or Java Straits.

    It is not known which TUDM long-range RADAR sites tracked 9M-MRO after its turn-back, that detail hasn’t been released.

    To add to your catalog of satcom performance incidents, seven oceanic control centres suspended the use of CPDLC (five of them suspended use of ADS-C) over Iridium AMS(R)S in Oct 2017 due to performance issues through the Iridium space and ground segment.

    Do note that appreciation of the doppler effect is fundamental to the Iridium system also, for the intra-constellation Ka-band links and the subscriber devices. The early generation GDP buoys, prior to those embedded with GPS receivers, also relied on the doppler effect for the ground stations to determine their position. Similarly, COSPAS LEOSAR exploits the doppler effect to determine position for non-GPS beacons.

    McKay’s position would have put an aircraft, at or near IGARI, very close to the horizon, not high in sky. The McKay myth exemplifies the problem with so-called ‘eye-witness’ reports: they’re unreliable.

  143. Don Thompson says:

    @Andrew

    Yes, I’ll collate via email. I’d certainly appreciate your help to get further clarity.

  144. Niels says:

    @Richard Godfrey
    Both from your report and from your comment at July 29th 1.27 pm (see link below) I initially concluded that you apply the 16 degree bearing to the left to both the 1.2% of the 10m wind as well as to the constant 10 cm/s term. You suggested that the total windage vector should be added to undrogued drifter velocities.

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17178

    As we agreed later, Following Griffin et al. the angle to the left of the wind should only be applied to the constant 10 cm/s windage vector. The 1.2 % x 10m wind vector should be aligned to the wind, see also Victor’s comment:

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17183

    After checking your table 10 in the report it gets a bit confusing. It looks to me that you only used the 1.2% of 10m wind (at 16 degree left of the wind) and omitted the constant 10 cm/s contribution. Maybe this should be clarified first.

  145. Victor Iannello says:

    @haxi: The Drain the Oceans episode on MH370 featured Peter Foley, Blaine Gibson and David Griffin, as you say. It also featured Andy Sherrell, who helped the BEA find AF447, and also helped the ATSB and Ocean Infinity plan their search missions and analyze the data. He is an incredible resource.

  146. Victor Iannello says:

    @Andrew: There were several differences between the simulations I ran and what was reported in the Safety Report. I started the climb and the turn before the “entry waypoint”, and I allowed the speed to drop lower than 400 knots. With a V/S climb at 6,000 fpm and a target altitude of FL410, the speed decreases to near stall during the climb, and the speed during the turn is between 380 and 400 KTAS.

    I also don’t understand why in their Session 6, the speed is set to 250 KIAS with the autothrottle engaged, yet the speed decreased to stick shaker speed during the manual turn with a 35° bank. To me, that suggests there was a climb during the turn, as there should have been more than enough thrust at 250 KIAS and FL350 to maintain speed.

    But in the end, I don’t think it really matters whether or not the autopilot was engaged during the turn. Even with the A/P engaged, the pilot inputs were extreme, as a high V/S climb at FL350 would suggest.

    Another possibility is the military radar data that captured the turn is not correct, either due to measurement errors, or the data was an inaccurate extrapolation from other radar data. (I’m being kind by not using the word “fabrication”.) We’d be better able to judge that if we had more than two points during the turn.

  147. Victor Iannello says:

    @Dan Richter: As I said, the farcical comments found on Facebook won’t be allowed here.

    First, what you describe is not Bernoulli’s principle, which deals with the relationship of changes in speed and changes in pressure (or level in an open channel). Rather, the speed increases in the strait because mass is conserved, which means velocity increases as the flow area decreases for incompressible flow. But that does not explain why debris was not found on the shores of Malaysia, Indonesia, Thailand, Vietnam, or Cambodia where there is not a narrow strait.

    Second, the drifter paths in the South China Sea that you reference do not explain the discovery of debris in East Africa.

    Third, even if you dismiss all the military radar, we still have the civil radar, which shows an aircraft with no transponder traveling back over the Malay peninsula.

    Fourth, your recommendation to ignore the BTO and BFO data because you have anecdotal evidence that the Iridium satellite network performs better than the Inmarsat satellite network is irrelevant. You provide no reason to dismiss the BTO and BFO data, despite the work that Inmarsat and the DSTG did in validating the BTO and BFO models using historical flight data.

    If we are to accept your premise that MH370 crashed in the South China Sea, we would have to unjustifiably dismiss all the evidence we have after 17:21 UTC.

    Future comments that you submit will be subject to moderation.

  148. Paul Smithson says:

    @Victor. Can you clarify where these two points are? I’m presuming that these represent turn entry and turn exit? Do we know what the implied turn diameter is [I guess we can derive this from what they have simulated]? What relation, if any, do they bear to the points depicted in the “right angle turn”?

    My own suspicion is that there was some extrapolation/coasting on intermittent military radar fixes over this period. This gives us an inaccurate time/position at start of turn – overshooting IGARI. And if you take that start position/time as “actual” then you find that you cannot get the aircraft to do it at normal speeds/bank angles.

    I have previously posted a simulated turn track based on M0.78, FL340 and bank angle 25 degrees, turn starts 17:22:07, ends 17:24:50 that was able to match the apparent diameter of the turn, the maximum extent NE, and the arrival time back at ’17:30:33′. IAS is 271, TAS 460 (comparable to speed when first picked up by Kota Bharu primary radar). As noted previously, this corresponds with ECON descent speed, while the direction of travel was initially directly at Kota Bharu – ie consistent with initial intent to divert directly to KB before changing course and speeding up (around 1734).

    https://www.dropbox.com/s/gnen985lz247txr/igari%20turnback%20simulation.jpg?dl=0

    I wish that they had provided the reference lat/long/timestamps for the turn start/end that this whole series of simulated turns in the report is trying to fit!

  149. Gysbreght says:

    @Victor Iannelo: “I also don’t understand why in their Session 6, the speed is set to 250 KIAS with the autothrottle engaged, yet the speed decreased to stick shaker speed during the manual turn with a 35° bank.”

    The stickshaker is riggered by angle of attack, not airspeed. In a turn with a 35° bank the load factor is about 1.22 and stall speed and stickshaker speed increase by about 10% relative to 1g.

  150. Gysbreght says:

    Correction: “triggered”

  151. Richard Godfrey says:

    @Niels

    I am obviously not making myself clear.

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17170

    This post linked above on 29th July 2018 4:52 am is correct together with all attachments.

    This post supersedes the results presented in the drift analysis, but does not represent a change in my conclusion.

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17176

    This post linked above on 29th July 2018 11:24 am was incorrect as I stated subsequently on the 30th July 2018 at 8:05 am and should be ignored.

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17178

    This post linked above on 29th July 2018 1:27 pm was similarly incorrect as I stated subsequently on the 30th July 2018 at 8:05 am and should be ignored.

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17183

    Victor’s interpretation linked above posted on 29th July 2018 at 4:40 pm is correct and that is why I stated subsequently on the 30th July 2018 at 8:05 I stand corrected.

    I worked for 2 months on the research, analysis and production of my paper. Unfortunately I made one mistake, which I have spent a further 3 days working on and corrected.

    Instead of your continued witch hunt, maybe you would like to publish the results of your research with the correct results. You obviously believe my work is incomplete and the methodology is questionable, as you state “however this can only be concluded if your methodology is sound”. Why don’t you produce a complete work based on a sound methodology?

    However, I disagree with your unfounded allegation and insinuation. I claim that my work is complete and my methodology is sound. You have no right to tell me what to do, for example “A minimum (as I understand it all is a lot of work) would be to have c. in addition to a.” It is my choice alone, if I wish to do any further drift analysis. I see no point in playing with numbers, as you suggest in your comment linked below:

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17198

    I stand by my conclusion as stated, that the MH370 End Point is between 20.0°S and 25.0°S within 22 NM of the 7th Arc, with the benefit of hindsight and discounting other options.

  152. Victor Iannello says:

    @Paul Smithson said: I wish that they had provided the reference lat/long/timestamps for the turn start/end that this whole series of simulated turns in the report is trying to fit!

    I agree. Depending on what the radius of the turn needs to be to match the radar data, there are many possibilities.

    Can you clarify where these two points are? I’m presuming that these represent turn entry and turn exit? Do we know what the implied turn diameter is [I guess we can derive this from what they have simulated]? What relation, if any, do they bear to the points depicted in the “right angle turn”?

    I know only as much as is contained in the Safety Report, as presented starting on page 276. The entry and exit waypoints are: N07.05.7 E103.47.1 and N07.12.7 E103.38.7, which I interpret as N07°05.7′ E103°47.1′ and N07°12.7′ E103°38.7′, which is similar to the format required to enter custom waypoints in the FMC of the B777.

  153. Amble says:

    The Safety Report, page 368 states “ For the passengers, oxygen could be supplied by chemical oxygen generators located in passenger service units (PSUs). A door with an electrically operated latch…”

    The circuit breaker was not flight deck accessible. Would turning the left electrical circuit off prevent the automatic deployment of the cabin oxygen masks in the event of a depressurisation?

  154. TBill says:

    @Amble
    “Would turning the left electrical circuit off prevent the automatic deployment of the cabin oxygen masks…”

    Good question, but to date, the answer to that question has been “no”. The only way to disable O2 masks would be to cut the breakers in the MEC bay.

  155. Victor Iannello says:

    @Gysbreght: For no vertical acceleration, the angle of attack and speed are related. One determines the other. Even if the stick shakes based on the measurement of the angle of attack, it is the stick shaker speed that is indicated in the PFD. If you want to argue that the use of the term “stick shaker speed” is incorrect, please take that elsewhere. I won’t argue with you for the sake of arguing, as I’ve told you many times.

    The simulation was performed with autothrottle set to 250 KIAS at FL350, 250 MT, and a 35° turn. Under those conditions, the engines should have enough thrust to maintain the speed, and the stall speed should be much less than 250 KIAS, even with 1.22g wing loading. The fact that the stick shaker speed was reached suggests the speed was less than 250 KIAS, which to me suggests a climb.

  156. TBill says:

    @Victor
    In the recent past I was quite adament that the turn at IGARI was consistent with ALSM’s postulation of a climb to FL400+. There is an apparent huge slow down, and then speed up. This can be seen vividly in @sk999’s digitized path spreadsheet (which does not strictly curve through IGARI but is quite good generally).

    I’d be curious to hear ALSM’s ideas if the new report changes anything about his recent work with the civil radar data.

  157. Victor Iannello says:

    @TBill: @sk999’s path was generated by integrating the speed and track curves presented in the DSTG report. Those values were the output of a Kalman filter that includes performance limits of the plane. I would not use those results to infer anything about the details of the turn.

  158. Paul Smithson says:

    @Victor. Ah, ok and thanks – I’d missed those coordinates in my skim read. Here they are superimposed on my earlier figure. I don’t know what to make of it, to be honest. These positions appear to be much closer together than the diameter of the turn as best as we can tell from the “right angle turn” depiction. My simulated turn begins quite a bit earlier than theirs though the end position looks close.
    https://www.dropbox.com/s/mb96f0yfzigu0yt/igari%20turnback%20waypoints.jpg?dl=0

  159. Victor Iannello says:

    @Paul Smithson: An alternate interpretation of the coordinates is N07°05’7″ E103°47’1″ and N07°12’7″ E103°38’7″, which displaces the waypoints a bit.

    Adding to the confusion is this statement:

    However, there were issues with the entry waypoint being off the direct track IGARI to BITOD (to the south-east) and this resulted in the ‘aircraft’ being in a slight right bank when overflying the waypoint and then starting the left turn. This would have increased the time to make the left turn as the ‘aircraft’ had to roll through level, before rolling west.

    The waypoints appear to be to the north of the IGARI-BITOD path, not south. But perhaps I misunderstood the meaning of the sentence.

  160. Gysbreght says:

    @Victor Iannello: “I won’t argue with you”

    Then you’ll just have to wait for Andrew.

  161. Richard Godfrey says:

    @Victor, @Paul,

    Assuming the coordinates are N07°05.7′ E103°47.1′ and N07°12.7′ E103°38.7′, then they are 20.17 km apart.

    Using the csgnetworks.com Aircraft Turn Information Calculator for a speed of 472 knots, a 180° turn requires a bank angle of 31° and takes 129.8 secs to complete to fit within the 20.17 km (10.9 NM) diameter. I used a stall speed of 130 knots.

  162. TBill says:

    @Paul S
    Your IGARI turn graphic is very nice.

  163. Richard Godfrey says:

    @Victor, @Paul,

    The above calculation puts the time at the end of the turn back at 17:24:37 UTC. This is based on a time at IGARI of 17:20:31 UTC (per ADS-B data) and a time at the start of the turn back of 17:22:27 UTC calculated from a Ground Speed continuing at 472 knots (per ADS-B data) for the 68.8 km from IGARI to the start of the turn back.

    Assuming MH370 followed a track of 237°T (not 273° as stated in the report page 4 para 2), following parallel to flight route M765 (per report), coasted at 17:37:12 UTC (per report), then the average Ground Speed over the 200 km from the end of the turn back to the Malaysian coast (just at the border to Thailand) was 514.9 knots.

    Somebody wanted to fly as fast as possible back toward Malaysia.

  164. Paul Smithson says:

    @Richard G. Thanks for solving that error in the report on course at end of turn, parallel to M765 of “273”. Of course 237 makes sense as a simple transposition of digits – nearly reciprocal of 058.

    I can’t agree with your mean groundspeed calculation, though. I make it 94NM from end of turn waypoint (17:24:37 according to yours, above) to last point before Kota Bharu cone of silence (at 17:36:43) per ME’s primary radar file, position 6° 18.696’N 102° 21.159′. 94NM / 0.2017hrs => 466kts.

  165. Victor Iannello says:

    From Free Malaysia Today

    GEORGE TOWN: A former air force pilot has called for an inquiry to find out why the armed forces did not reveal that Flight MH370 turned westward to the Indian Ocean, noting that this cost a week of wasted efforts to search for the plane in the South China Sea…

    The radars also spotted the plane at Pulau Perak, a small island southwest of Langkawi. The reports show that it was flying at the dangerously low level of 1,463 metres…

    “Only an RCI can determine whether the government purposely hid the military radar information at the instruction of foreign powers.”

    It’s good that people are highlighting the obvious inconsistency of detecting MH370 real time but not alerting civil aviation and search authorities until days passed.

    The implication that a foreign power coerced Malaysia to hide the radar data is an interesting twist.

  166. Paul Smithson says:

    Calculation for groundspeed from position of turn exit waypoint (decimal minutes version) and your time of 17:24:37 to first position on the primary radar file at 17:30:33. I make it 44.86NM in 5 mins 56, yielding groundspeed of 454kts.

  167. TBill says:

    @Victor
    “GEORGE TOWN: A former air force pilot…”

    Yes indeed very interesting. If you go back and listen to Razak on 15-March and then Hish on 4 Corners, Hish says (paraphrasing) I will give you the same explanation as PM Razak, go ahead ask him, that we saw an unknown radar blip but until the JIT used the satellite data to prove it was MH370, we did not know that.

  168. flatpack says:

    @Victor Iannello

    You noted to Don Thompson the following:-

    ‘Since no email or text messages occurred during MH371, which was normal in all respects, the absence of emails and texts after 18:25 means nothing.’.

    It does not surprise me that no use was made of the SMS or email facility during an essentially normal flight. This service seems to me to be overpriced and of limited utility if everything is going well.

    It would seem to me to be much more useful were there a change in flight plan mid-way.

    Whilst near real time arrival/diversion information may be available from the internet to those waiting to meet passengers at the destination, it would be extremely useful to alert them that this was likely and to re-schedule business meetings etc.

    In summary, significantly abnormal flight = reason to use SMS/email

  169. Victor Iannello says:

    @flatpack: Sorry, but I’m not convinced the capability was available on MH370 after 18:25. Perhaps it was. I don’t know. It’s hard for me to believe that the price would be so prohibitive that NOBODY chose to use it on MH371, including passengers in first class.

  170. Paul Smithson says:

    @Victor. In response to your earlier. Whether dd.mm.ss or dd.mm.mmm the first waypoint lies comfortably north east (left) of the IGARI-BITOD line. So that statement in the report about it lying SE therefore plane must still be banking right doesn’t make any sense at all to me.

    By my measurements, the less conventional dd.mm.ss for both waypoints gives you a diameter of 20.86kms versus 20.14kms for the dd.mm.mmm version (rounded to one decimal). The larger turn diameter would still require AOB a shade over 30 degrees at 472kt TAS.

  171. Paul Smithson says:

    * to two decimals!

  172. Richard Godfrey says:

    @Paul

    On page 9 of the report it states under point 2 that the target P3362 coasted at 17:37:12 UTC. I took this to mean that MH370 crossed the coast at this time at 6.2321°N 102.1261°E, which is 108 NM from the end of the turn back at 7.21167°N 103.6450°E.

  173. Victor Iannello says:

    @Paul Smithson: I think they mean that if the plane was following the legs IGARI-WP1-BITOD, as it flew near WP1, it would be cutting the corner to reach BITOD in a slight bank to the right, so flying direct to WP2 after WP1 means the plane would take additional time to reach the left bank limit. That wasn’t obvious to me the first time I read it.

  174. paul smithson says:

    @Richard. My understanding is that the target was tracked… then coasted.. then dropped. So the “coasting” is an extrapolation of target position until it drops altogether. This terminology has been used throughout in the FI. Nothing to do with crossing the coastline IMO.

  175. Richard Godfrey says:

    @Paul

    Coasting has two meanings:

    1. Taking your foot of the accelerator but keeping going.

    2. Reaching or following a coastline.

  176. Victor Iannello says:

    @Paul Smithson, @Richard Godfrey: By coasting on the radar screen, they mean continued to appear on the radar screen with an extrapolated position despite no captures. Targets continue to appear for 10 seconds after the last capture, which is why P3362 first appeared at 1730:37, coasted (was last detected for that segment) at 1737:12, and was dropped (disappeared from the screen for that segment) 10 seconds later at 1737:22.

  177. Richard Godfrey says:

    @Victor, @Paul,

    According to the radar expert definition of coasting, the whole report seems to be full of coasting!

    Somewhere there is a point of truth and the rest is made up to make it seem like an extrapolation of the truth.

    I thought the report was written for public circulation and not radar experts.

    I just proved I am not a radar expert!

  178. Paul Smithson says:

    Just to confirm RG’s estimates of time at exit waypoint above. We should be able to estimate reasonably accurately the time recorded by miltary radar for positions referred to in the report as Entry Waypoints and Exit Waypoint.

    Distance from IGARI to Entry Waypoint is 15.27NM which takes 117 seconds at groundspeed of 470kts. If time “over IGARI” is accepted as 17:20:31, then time at Entry Waypoint is 17:22:28.

    The safety report tells us that there was a 2 minute 10 second time elapsed between Entry Waypoint and Exit Waypoint. This is independent of any assumption about path, speed, bank angle and it gives us time at Exit Waypoint of 17:24:38 (1 second earlier by Richard’s calculations).

    From that time and position we can determine speed over a decent period (nearly 10 mins) to the start of the KB primary radar. As stated earlier, not matter which way you cut it this is going to give us a slow speed – consistent with our earlier best estimates of speed at the start of the KB primary radar (450-460kts).

    The claim of 515 knots groundspeed from turnback to coastline/KB is simply incompatible with the evidence as we know it.

  179. Gysbreght says:

    17:36:43 to 17:38:43 was the cone of silence in thew Kota Bharu Civil radar data obtained by ALSM.

  180. Victor Iannello says:

    @Paul Smithson said: The claim of 515 knots groundspeed from turnback to coastline/KB is simply incompatible with the evidence as we know it.

    Who said that? From the first PSR capture at 17:30:33 to the start of the Cone of Silence at 17:36:43 above Kota Bharu, I estimate the average speed as 487 knots, but accelerating from a lower value to a higher value over that interval.

  181. Andrew says:

    @Victor

    RE: ”I also don’t understand why in their Session 6, the speed is set to 250 KIAS with the autothrottle engaged, yet the speed decreased to stick shaker speed during the manual turn with a 35° bank.”

    It’s not stated in the report, but my guess is the stick shaker was intermittent, caused by the pilot momentarily increasing back pressure on the control column. From the FCOM:

    For turns up to 30 degree of bank, the pilot does not need to add additional column back pressure to maintain altitude. For turns of more than 30 degree of bank, the pilot does need to add column back pressure.

    If the pilot relaxed the back pressure during the turn with the AOB greater than 30°, the aircraft would descend. If he then increased the back pressure to regain the initial altitude, he might trigger the stick shaker for a short period.

  182. Niels says:

    @Richard Godfrey
    These type of calculations for modelling the drift of debris across the ocean are difficult, even for specialists. I hoped by constructive criticism I could help you a bit in the right direction. As becomes clear from your 9.21 am post in the end it did not work.

    With loss of constructive communication I will have to reconstruct the way you modelled windage for the flaperon by myself. It becomes clear by comparing figures 19 and 20. These are identical curves, only shifted 53 days with respect to each other.
    This is confirmed in the following comment, where you state: “In the original study, with mainly drogued data, I assumed the windage for the Flaperon would get it to the Reunion area 53 days earlier”

    http://mh370.radiantphysics.com/2018/07/20/godfrey-drift-model-says-mh370-might-have-crashed-further-north-on-arc/#comment-17176

    The 53 days are based on a 35.5% increase in the drift speed, as becomes clear from the text between figures 19 en 20: “Windage accounts for an average increase in speed of 35.5% in the 150 days travelling with the South Equatorial Current towards Reunion and floating debris will arrive on average 53 days earlier”

    The calculation of the 35.5% is illustrated in table 10. From this table it becomes clear that indeed you only took the 1.2% windage into account (at 16 degrees left of the wind; which should have been aligned with the wind) and omitted the 10 cm/s. This 10 cm/s may look like a small number, however it is equivalent to 0.19 knots and should be compared with your average drift speed of 0.48 knots (without wind) and 0.65 knots (with 1.2% windage).

    In addition it should be noted that wind effects are not only important while in the south equatorial current, they will also influence the transit from the more southerly latitudes towards this current. In your paper you ignore this aspect of the contribution of wind.

    Altogether I conclude that your drift modeling for the flaperon is wrong, which is a serious issue as the estimated transit time for the flaperon is one of the key indicators for the origin of the debris.

    Earlier we already agreed that the drift of “low-windage” items should best be modelled based on undrogued drifter data and not on the “mixed” dataset which is dominated by drogued drifter data. So, also for this type of items there are issues with the results presented in your paper.

    The proper incorporation of the effects of wind and waves is difficult and critical for the reliability of the results obtained. Imo you did not (yet) manage to correctly address this aspect.

  183. Victor Iannello says:

    @Andrew: Thank you. I was considering the case of level flight followed by a climb. It could have been a descent followed by a climb if he pulled back hard enough to regain altitude.

  184. TBill says:

    @Paul Smithson
    I am OK with 17:20:31 for IGARI. I believe the FI says that the left turn started almost immediately after 17:21:13 so in the past I normally waited just a little bit no later than 17:21:30 to start the turn.

    FI:
    “At 1721:13 UTC [0121:13 MYT] the Military radar showed the radar return of MH370 turning right but almost immediately making a constant left turn to a South Westerly
    direction.”

    The new waypoints suggest maybe a little later turn which may not be totally consistent with the “almost immediately” verbage. So when would you say the the started?

  185. Andrew says:

    @Victor

    RE: ”I think they mean that if the plane was following the legs IGARI-WP1-BITOD, as it flew near WP1, it would be cutting the corner to reach BITOD in a slight bank to the right, so flying direct to WP2 after WP1 means the plane would take additional time to reach the left bank limit.”

    I haven’t plotted the waypoints to see how they relate to the flight-planned track, but I interpreted their statement to mean the aircraft was in a slight banked turn to the right as it crossed the ‘entry’ waypoint. From the FI: At 1721:13 UTC [0121:13 MYT] the Military radar showed the radar return of MH370 turning right but almost immediately making a constant left turn to a South Westerly direction. That would have increased the time it took to roll into the left turn.

  186. Victor Iannello says:

    @Andrew: I don’t know. I was referring to the simulation and the position of waypoints WP1 and WP2 relative to IGARI and BITOD. They programmed the path IGARI-WP1-BITOD, and then changed the route to fly direct to WP2 after crossing WP1. As a right hand turn is required to change from the IGARI-WP1 track to the WP1-BITOD track, this produced the slight bank in their simulation. I don’t know if they did this to replicate the behavior of MH370. It’s not clear.

  187. Richard Godfrey says:

    @Niels

    I have tried several times to answer your criticisms, but you are not listening.

    You make more and more requests, which result in a lot of work for me.

    I am simply not willing to follow your whims any longer.

    It has become evident, that you have your own agenda.

    You are surprised that I reached a conclusion, but you have concluded that I am wrong.

    I am fine with the fact, that we disagree.

  188. Gysbreght says:

    @Andrew: “I haven’t plotted the waypoints to see how they relate to the flight-planned track, …”

    https://www.dropbox.com/s/o7w68y9alcw7iq0/SIR_Sims.pdf?dl=0

  189. TBill says:

    @Gysbreght
    Nice

  190. Gysbreght says:

    @TBill: Thx. Just plotted the points in an old chart (2015?).

  191. DrB says:

    @Richard Godfrey,
    @Victor Iannello,

    Have you computed the BFO errors and average leg speeds for a path to 20S that matches the BTOS?

    I believe the BFO errors continue to grow in magnitude as one moves north along the Arc from a minimum near 32S. That is one area of concern for the 20S region. It would also be good to check to see if the necessary ground speeds can be reconciled with the known aircraft endurance.

    At present we seem to require at least one “unusual” happening to get a fully consistent solution and end point.

    At the northern portion of the Arc, say from 20S-26S, we require that the BFO offset must be shifted (from 19:41 to 00:19) by something like 10 Hz compared to the pre-18:25 BFO offset. If such a BFO offset was also present at 18:25-18:28, this impacts the interpretation of the BFOs then and the conclusion that a lateral offset occurred then. Have you tried to fit a different maneuver then with a shifted BFO offset?

    At 31.6S there is an excellent CMT solution to match the BTOs/BFOs, but we now know the aircraft is unlikely to be there within 22 NM of the Arc. In this second case the “unusual” event must be a glide beyond 22 NM. Presumably this could occur either by a pilot recovering from an initially steep descent or by phugoid oscillations ( such as previously illustrated by Gysbreght) with an unusually well-trimmed rudder such that a large bank angle did not develop in secondary flight control mode.

    The choice of search latitude thus depends on the choice of which “unusual” event is thought to be more likely (BFO offset step or extended glide), as well as on the weight given to the ocean temperatures inferred from the barnacle growth study.

  192. Paul Smithson says:

    @TBill. In response to your earlier question. My original simulation produced end of turn at 17:24:50. The position was nearly “abeam” the Safety Report’s position but situated 1.4NM further “out” (to the NE).

    If we take as “fact” that the turn end time was 17:24:37 or 38 then the plane needs to arrive there 12s earlier than in my simulation. That is achieved by commencing the turn ~6s earlier, at 17:22:01, position 7.0371 103.7500 (on the IGARI-BITOD line), if we keep speed FL340/M0.78 and AOB 25 then my turn diameter will still create a turn terminus position a bit further “out” than Exit Waypoint.

    If you want to end precisely at Exit Waypoint starting the turn from the IGARI-BITOD line and max AOB of 25 then turn calculator says you need TAS not more than 440kts, or 26 AOB and 450kts TAS. Or if you accept that Exit Waypoint might be 0.5NM out ie turn diameter up to 12.7NM, you can achieve it with 450TAS/25 AOB or 460TAS/26AOB. In fact turn diameter is marginally tighter than the calculator tells you because of the headwind during the turn from ~070.

    In any event, the speeds required are close to the inferred groundspeed of 454kt between Exit Waypoint and start of the KB primary radar trace. No crazy bank angles or odd turn shapes required. In fact it all seems to me to hang together rather well.

    The question foremost in my mind is why would a pilot SLOW the plane to ~M0.75/FL340, make a 180 degree turn, maintaining this slow speed for 11 minutes (from start of turn to 1733)?

  193. Paul Smithson says:

    *correction – slow speed for 9 minutes, 1722 to 1731 [not 11].

  194. Paul Smithson says:

    Another oddity. Safety Report (hereafter SR) page 298 says final SSR drop was at 17:21:13, 3.2 nm [sic] after passing IGARI. If the plane was doing 470-ish, it should have been 5.5NM beyond IGARI 42 secs after “passing over” it at 17:20:31. Something wrong here. For 3.2nm it should be 24-25 seconds after IGARI.

  195. Andrew says:

    @Gysbreght

    Thanks for posting your graphic with the waypoints.

    Thanks also for your earlier paper with your thoughts about phugoids. Do you see any possibility of a phugoid that would take the aircraft outside the area that has already been searched?

    @Richard Godfrey
    @Paul Smithson

    RE: “Using the csgnetworks.com Aircraft Turn Information Calculator…”

    The problem with the classic turn formula is that it assumes an instantaneous angle of bank. In reality, the aircraft is quite slow to roll into and out of turns, especially in LNAV. Consequently, the turn rate and radius predicted by the formula are somewhat less than would be seen in the aircraft.

  196. Paul Smithson says:

    @Andrew. Thanks for pointing that out. What sort of roll rate would you expect for LNAV? How long does it take to go from wings level to 25 bank angle?

  197. TBill says:

    @Paul
    If I follow you, you are talking about U-turns without a climb. Right now I am with @ALSM and @Victor that I think MH370 was climbing in the turn and could be going under 400-ish at the slowest point. That could conribute to shortening the turn diameter at the end. Admittedly I have not played yet with hitting the new waypoints.

    Seems like you are showing however that 25 deg turn A/P setting possibly works within the accuracy of what we know about the data. If MH370 was slowing due to climb maybe 25 deg A/P works.

  198. Andrew says:

    @Amble

    RE: “Would turning the left electrical circuit off prevent the automatic deployment of the cabin oxygen masks in the event of a depressurisation?”

    As @TBill has already mentioned, disabling the left side of the electrical system does not prevent automatic passenger oxygen mask deployment. The system receives power from the AC STBY BUS and the DC CAPT FLT INST BUS, both of which are normally powered by the left side of the electrical system. However, those two buses power many items of essential equipment. Consequently, considerable redundancy is built in to the system to ensure they remain powered in the event of a problem with the left side of the electrical system. Ultimately, they will be powered by the aircraft battery if all other power sources are lost.

  199. Victor Iannello says:

    @DrB: I haven’t looked at the 20S path in depth, other than the work on great circle paths. Eyeballing the chart from the previous post, it looks like the RMS BFO error would be around 12 Hz for such a path.

    On the other hand, the path based on the planned landing on Cocos Island has the RMS BFO error at 6.0 Hz, with most of that due to the mean error of -4.6 Hz. The BFO error at 00:11 is -11.9 Hz, or 7.3 Hz from the mean. Does that disqualify the path? I don’t think so. Also, as best I can tell, the fuel endurance is acceptable.

    If you accept that the final BFO values are valid, I don’t see how the impact could have been far from the 7th arc for uncontrolled flight. If magically the plane remained near perfect rudder trim, a level-wing phugoid could not develop an average 0.7g downward acceleration over an 8-second period ending in a 15,000 fpm descent (regardless whether of acceleration was upwards at that instant). The bank would have to be appreciable (>30 deg), the dynamics would be unstable, and the flight path would be curved.

    Even if the final BFO values are wrong and an extended phugoid developed with no pilot inputs, likely the path would be curved. Regardless of whether the rudder trim was precise, the Safety Report explains that the TAC doesn’t go to zero after the second (left) engine flames out because the instant at which the A/P disengages and the flight control mode goes to secondary, the TAC is non-zero, and that value is frozen. (Through an intermediary contact with Boeing, I learned that although the TAC does not operate when the flight controls are in secondary mode, the rudder deflection commanded by the TAC at the transition to secondary mode persists. I don’t think this is widely known, and not all Level D simulators accurately reproduce this.) This almost guarantees that an extended phugoid with level wings will NOT develop. I’ll have more to say about this in a future post.

    If there was a controlled glide after fuel exhaustion, it really doesn’t matter whether or not the final BFO values are valid. The descent could have started above FL400, and only several thousand net feet need be lost in the high speed descent and recovery climb. A glide from FL350 could still last longer than 105 NM. And it would be difficult to choose a latitude to conduct the search, as there could have been other pilot inputs between 19:41 and 00:19. That makes it almost impossible to define a search area.

    Again, I am trying to keep an open mind, but at this point it seems that our best chance is to search north along the arc. We do have some months remaining before the start of the search season, so perhaps we will have new insights before that then.

  200. Andrew says:

    @Paul Smithson

    RE: “What sort of roll rate would you expect for LNAV? How long does it take to go from wings level to 25 bank angle?”

    Off the top of my head, the roll rate is around 2-3 deg/sec, so around 10 seconds.

  201. ventus45 says:

    “We do have some months remaining before the start of the search season, so perhaps we will have new insights before that then.”

    The “search season” is latitude dependent due to weather. It is most constrained in the south, less so in the north.

    Assuming that OI do actually intend to continue the search, I don’t think we really do have “some months remaining” to make up our minds. Since the jury is still out on where to look next, vacillating is not really an option. It might be more useful to look at the strongest theories we have, for both wider searches at southern latitudes, and extending the +/- 22nm arc search north, and prioritizing both, with seasonality in mind. In other words, develop justifiable “seasonal” search options.

  202. Gysbreght says:

    @Andrew: “Do you see any possibility of a phugoid that would take the aircraft outside the area that has already been searched?”

    Sure, anything that keeps the wings level, either DrB’s “unusually well-trimmed rudder”, or a non-pilot person at the controls managing to keep the wings level.

  203. TBill says:

    @DrB
    I believe there probably are full-out active pilot cases that can reduce the BFO error Victor is reporting to get to 22 South. Victor has kindly provided a well-thought hybrid case that is passive pilot, but has pre-programmed COCOS maneuver. Victor’s BFO error does not bother me because I believe you can get to that same spot with less BFO error.

  204. DrB says:

    @Victor Iannello,

    Thank you for addressing my questions.

    You said: “If magically the plane remained near perfect rudder trim, a level-wing phugoid could not develop an average 0.7g downward acceleration over an 8-second period ending in a 15,000 fpm descent . . . .” What is the basis for this statement?

    I tend to trust the final BFOs. It’s too bad no in-flight SDU shutdowns were done as tests to determine typical BFO behavior.

    I don’ t have much confidence in arbitrary great circle paths to the northern part of the Arc. The BTO errors for a path to 20S require a fairly large offset jump at the 18:25 SDU reboot. So far, we have a total of two paths which are unique: your Cocos overshoot to 22S and my CMT to 31.6S. I think those are more likely candidate toutes. We are still left with a choice between a change in the BFO offset and an extended glide.

    There are three reasons why I think the 31.6S CMT has merit:

    1. The best-fit 181.2 deg CMT corresponds exactly to the actual magnetic track while on a 180.0 deg true bearing in the FMT region, and there are waypoints in the right place with identical longitudes allowing this. I don’t think this course matchup is a coincidence.

    2. The TRACK HOLD could have been done by either pilot or a member of the air crew or by a relatively untrained person circa 19:00. Perhaps this is the simplest way to send the aircraft deep into the SIO using the A/P.

    3. The CMT solution best fitting the BTOs/BFOs is unlike all other solutions for other navigation methods. That is, the best-fit residuals behave in a unique fashion. They are partially correlated over extremely fine course variations of a few hundredths of a degree. This correlation produces high-angular-frequency grating lobes in the error statistics. I don’t think this is some improbable effect of random noise in the data. To me it is an indication of being very close to the true route.

    Yet, the aircraft was not found by OI within 22 NM of the Arc at 31.6S. If the aircraft passed there, then either the BFOs are wrongly interpreted or a glide occurred, piloted or unpiloted (by some means not yet understood). There is still at least one piece of the puzzle missing.

  205. DrB says:

    @TBill,

    You can invent flight paths that have zero BFO errors if you allow turns and speed changes just prior to the ping times. However, they are not credible because the pilot would not know the times when those pings would occur. Even allowing only one or two arbitrary maneuvers after 19:41 seems to prevent a unique solution. If that occurred, you could only limit the solution space to the portion of the 7th Arc that is reachable with the fuel on board. That’s what ATSB did in one of their figures.

  206. Gysbreght says:

    For the record, just a clarification of my post on August 3 at 9:11 am:

    When I wrote “stall speed and stickshaker speed increase by about 10% relative to 1g” I meant the speed at which the stickshaker is triggered. That speed may differ grom the stickshaker speed indicated on the PFD speed scale, if the latter is calculated for 1g, not accounting for the actual “g” level.

  207. TBill says:

    @Paul
    Report p.3
    “The Bangkok radar target drop occurred at 1721:13 UTC [0121:13 MYT] and Viet
    Nam’s at 1720:59 UTC [0120:59 MYT].”

    perhaps the 3.2-nm is based on the Vietnam timing?

  208. Gysbreght says:

    @Victor Iannello, you wrote: “Through an intermediary contact with Boeing, I learned that although the TAC does not operate when the flight controls are in secondary mode, the rudder deflection commanded by the TAC at the transition to secondary mode persists.”

    Many thanks for sharing that information. That issue was extensively discussed here some time ago and is really important for understanding the uncontrolled simulations, in particular those where the first engine flame-out caused the loss of the autopilot.

  209. Dan Richter says:

    @
    Victor Iannello says:
    August 3, 2018 at 8:31 am

    @Dan Richter: As I said, the farcical comments found on Facebook won’t be allowed here.

    First, what you describe is not Bernoulli’s principle, which deals with the relationship of changes in speed and changes in pressure (or level in an open channel). Rather, the speed increases in the strait because mass is conserved, which means velocity increases as the flow area decreases for incompressible flow. But that does not explain why debris was not found on the shores of Malaysia, Indonesia, Thailand, Vietnam, or Cambodia where there is not a narrow strait.

    Second, the drifter paths in the South China Sea that you reference do not explain the discovery of debris in East Africa.

    Third, even if you dismiss all the military radar, we still have the civil radar, which shows an aircraft with no transponder traveling back over the Malay peninsula.

    Fourth, your recommendation to ignore the BTO and BFO data because you have anecdotal evidence that the Iridium satellite network performs better than the Inmarsat satellite network is irrelevant. You provide no reason to dismiss the BTO and BFO data, despite the work that Inmarsat and the DSTG did in validating the BTO and BFO models using historical flight data.

    If we are to accept your premise that MH370 crashed in the South China Sea, we would have to unjustifiably dismiss all the evidence we have after 17:21 UTC.

    Future comments that you submit will be subject to moderation.

    Victor, I appreciate your great work very much. I do not understand that after four years of fruitless search based on pings, the Safety Report does not include drift analyzes and eyewitness testimonials. It looks like the investigators have a black belt accross their eyes and they are not willing to accept and work with other evidence. I have the evidence that debris could easily traveled from South China Sea through Sunda Strait to equatorial streams and washed up in the beached in Reunion, Mozambique and South Africa. See the book “Prostar Sailing Directions 2005 Borneo, Jawa, Sulawesi and Nusa Tenggara Enroute”, ISBN-10: 1577856546 describing Sunda Strait: Tides-Currents. During the Northwest Monsoon, the current direction varies through Selat Sunda. From April to September, the currents are SW to S at a rate of about 0.75 knots. From October to March, the currents are NE to N at the same rate. When the winds are from NW through N to NNE, the current sets S to SW at up to 1.3 knots.

    The NOAA buoy 56435 traveled from the burning object position in SCS to Sunda Strait area in 61 days, so we can assume that the flaperon could pass through strait around May 8, 2014 (8.3.14 +61 days). In May, the current flows to S to SW at a rate of about 0.75 knots.

    I do not mind if you will ban me to access this forum. My goal is to help find MH370 as soon as possible. Together with other people, we are trying to rent a Boeing airplane in Kuala Lumpur, seats paid by volunteers, to let if fly along the supposed route around Indonesia and back to Kuala Lumpur to verify tracking by all civilian and military radar stations in the area. One of my colleagues had a meeting in Kuala Lumpur a few days ago to arrange renting a plane.

  210. Victor Iannello says:

    @Gysbreght said: That speed may differ grom the stickshaker speed indicated on the PFD speed scale, if the latter is calculated for 1g, not accounting for the actual “g” level.

    The indication of the MMS does not vary. However, in a level turn, the stick shaker speed indicated on the PFD moves up as the wing loading increases.

    Andrew suggested something different. He mentioned possible additional loading for initiating a climb to recover altitude lost in the turn. The indicated stick shaker speed does not move due to vertical acceleration, and the stick may shake due to high angle of attack at a speed above the stick shaker speed.

  211. Victor Iannello says:

    @Dan Richter: Do you have permission to turn off the transponder and the cooperation of all the military in the area to share the raw radar data? What path, speed, and altitude do you intend to fly?

  212. Victor Iannello says:

    @Victor: If magically the plane remained near perfect rudder trim, a level-wing phugoid could not develop an average 0.7g downward acceleration over an 8-second period ending in a 15,000 fpm descent . . . .

    @DrB: What is the basis for this statement?

    Simulations and analyses. If you consider a level phugoid with near constant angle-of-attack (which is not entirely accurate), the amplitude and period of the variation in vertical speed would make that unlikely. I haven’t heard anybody here debate that. You can’t get to g-levels higher than around 0.3 without introducing a bank. If you accept the final BFOs, either the plane impacted close to that point or the glide was controlled.

    In any event, the Boeing simulations indicate that the wings do not stay level after fuel exhaustion. If their modeling of the TAC at the time of the loss of AC power is accurate, then a long-lasting wings-level phugoid cannot occur.

  213. Victor Iannello says:

    @DrB: Let me also say that for no pilot inputs, I think the scenario in which the left IDG is isolated at the time of the flame out of the first (right) engine is most likely to produce the observed downward acceleration and descent rate. TAC is lost after the right engine flames out, and the residual TAC input will not completely compensate for the thrust asymmetry, and the plane rolls gently to the right. After the left engine flames out, the residual TAC causes a stronger roll to the left, leading to high, variable descent rates.

    More later. It’s the subject of a new article that I will be posted in coming weeks.

  214. Victor Iannello says:

    @David asked: Where is there mention of the vortex generator, reportedly made available for collection by Madagascar some time ago? Where is the damage report on that?

    I apologize for not addressing this question sooner, as it got lost in the flurry of other questions.

    Last we heard from Malaysian officials, the vortex generator is being held in Madagascar as evidence in the investigation into the death of the Malaysian honorary diplomat Zahid Raza. That indicates that the authorities in Madagascar don’t dismiss the possibility that his death was somehow related to MH370 debris.

  215. Victor Iannello says:

    DrB: Previously, I said After the left engine flames out, the residual TAC causes a stronger roll to the left, leading to high, variable descent rates.

    A more accurate statement is after the left engine goes to idle thrust, there is a roll to the left. We have to remember that the autothrottle will remain engaged after the autopilot is disengaged. I believe that part of what we see in the Boeing simulations for the cases in which the left IDG isolated is there is a descent that causes an increase in speed past the setpoint speed of the autothrottle, which leads to commanded idle thrust to the left engine, which in turn leads to a roll to the left due to the residual TAC input of left rudder. The interaction of the autothrottle, the vertical speed, the airspeed, the rudder position, and the thrust asymmetry leads to very unstable flight. I believe the timing and values of the final descent rates can all be explained for the configuration in which the left IDG is isolated. In fact, I think it’s possible the left engine was still running (with idle thrust) when MH370 impacted the ocean.

  216. vodkaferret says:

    on the question of searching further North vs wider::

    1) if you believe the BFO/BTO data cannot be trusted, then the possible search area becomes unfeasibly large.

    2) if you believe it to be accurate but that the pic (whoever that was) recovered the plane from a dive and glided on then the search area becomes unfeasibly large. ( there is a subset of people who subscribe to the glide theory and make rather precise predictions as to landing place – no-one can say they are wrong unless searched, but I would argue most of those predictions are overly confident given the information available)

    3) there is still a portion of the 7th arc with feasible flight paths that match BFO/BTO and arguably other data that is as yet unsearched.

    4) given the areas in question if the already searched zone were to be extended out to (say)120nm from the arc, even OI’s impressive capability could only search a comparatively small portion in a single season. whereas to cover from the currently searched area up to and even including Indonesian waters in the previous 25nm from the arc could be accomplished in a single season.

    therefore given the currently available information, if the search went north, close to the arc and failed then then it would have ruled out everything except areas too large to be feasibly searched, and in the absence of any substantial new data / analysis the search could reasonably be concluded.

    whereas if the search went wider along currently searched portions of the arc and failed, it would mean more than one season and would still leave the northerly part close to the arc unsearched, hence always leaving a serious doubt.

    Therefore to me the logical option would be to search North, on the grounds a failure to find the plane would at least close out the most feasible search zones and allow for some measure of closure. Searching wider and failing however would not accomplish this. Given all the uncertainties and lack of data it’s probably better to prioritise a new area based on what it can rule out in the future, at least as much as what one believes in terms of its chances of actually finding the plane.

    Hope that makes sense.

  217. Victor Iannello says:

    @vodkaferret said: Hope that makes sense.

    Yup.

  218. DennisW says:

    @vodkaferret

    +2

  219. Richard Godfrey says:

    @vodkaferret

    *3

  220. Dan Richter says:

    @Victor Iannello says:
    August 4, 2018 at 7:32 am

    We would like to use official flight path based on Safety and Factual Information Reports (KUL, IGARI, left turn, Penang and around Sumatra back to KUL). FlyGlobal are the most likely source of charter B777s. Cost would be $15,000+ per “block hour”. Of course the demo flight needs to be coordinated with official civil & military authorities. We are working on it but anyone who can be of any help would be welcomed. I was asked: Do we have any members in London who could go to Inmarsat’s offices in London?

  221. TBill says:

    @DrB
    “You can invent flight paths…”

    But we are all inventing flight paths.
    A passive flight without turns is as much speculation as anything else, if not more speculative at this juncture.

  222. DennisW says:

    @TBill

    I think what DrB meant was “inventing” a turn to improve the BFO fit as opposed to a turn (i.e. Victor’s path) that has some logic behind it.

  223. Tim says:

    Victor,
    Don’t forget re your electrical configuration at the end. Even if the Left IDG is isolated, the left backup gen should still be working providing power to a transfer bus. So the flight controls should not drop to secondary mode when the R eng flames out.

  224. flatpack says:

    @TimR

    Upthread someone posted:-

    ‘There are the rumors from TimR that a plan for alternate but rogue safe landing was known in some circles’

    Please can you (or anyone else) point me to a post that sumarises this plan and/or any post that assess it’s likelihood. Thanks.

  225. Sabine Lechtenfeld says:

    @vodkaferret, I like your moniker and also your sentiments re: future searches. You nailed it. Searching the rest of the 7th arc – even all the way up to the Indonesian coast – is manageable for OI. If the search doesn’t produce the plane, we know at least that the plane cannot have crashed anywhere near the 7th arc for reasons which need to be discussed.Even a negative search result will yield more knowledge.
    I need a vodka with lemon now! Cheers 🙂

  226. vodkaferret says:

    @Sabine et al.

    Thanks. I have very little to contribute here as most of the scientific / technical stuff is way above my pay grade, but occasionally I see somewhere I may be able to help. In this case that “where should we search next?” and “where do we think the plane most likely is?” are two separate questions and shouldn’t be conflated.

    Enjoy your vodka 🙂

  227. vodkaferret says:

    by “way above my pay grade” I hope it’s clear I mean “baffles the f*** out of me ‘ 🙂

  228. David says:

    @Victor. I have trouble with the phugoids at 30 deg bank. First, the amplitude of the phugoid will need to be doubled to realise the same descents and vertical acceleration.

    Second, the lift vector will be less than 1 g on the nose ‘down’. When that 30 deg off vertical vector is combined with gravity’s at vertical, that will result in a sideslipping acceleration will it not and quite different dynamics to a phugoid?

    I think the ATSB notion of a less stable phugoid was in place of steep bank, not a part of it: a phase that occurred when wings were reasonably level.

  229. Richard Godfrey says:

    @TBill

    We are not inventing flight paths.

    We are looking for flight paths, that fit the satellite data, aircraft performance, weather information, fuel range and endurance and other evidence such as the drift analysis, taking into account the area already searched.

    DrB has optimised the flight paths and concludes a 181.2 CMT track to 31.6S and considers that the search width was too narrow.

    Victor has shown a flight path that fits a MH370 end point further north at 22S.

    Both considerations are right, larger search width and/or further north.

    My money is on further north because in addition the drift analysis shows between 25S and 20S fits the floating debris discovered.

  230. David says:

    @Victor. Correction to my above. In place of ‘doubled’, ‘substantially increased’.

  231. David says:

    @Victor. Vortex generator. You said, “Last we heard from Malaysian officials, the vortex generator is being held in Madagascar..”
    Is that an update of the below please?

    https://www.airlineratings.com/news/malaysia-finally-collect-crtical-mh370-debris/

  232. Victor Iannello says:

    @David: My understanding is the parts have not yet been sent to Malaysia as of days ago.

  233. Victor Iannello says:

    @Tim: Thanks, Tim. Yes, you are right. The loss of the autopilot and the autostart of the APU is triggered by loss of power to the transfer buses. After the right engine flames out, that would require both the left IDG and the left backup generator to be isolated.

  234. Victor Iannello says:

    @Dan Richter: What altitude profile do you intend to fly?

  235. Andrew says:

    @Victor

    RE: “The indication of the MMS does not vary. However, in a level turn, the stick shaker speed indicated on the PFD moves up as the wing loading increases.

    Andrew suggested something different. He mentioned possible additional loading for initiating a climb to recover altitude lost in the turn. The indicated stick shaker speed does not move due to vertical acceleration, and the stick may shake due to high angle of attack at a speed above the stick shaker speed.”

    In the scenario I suggested, the pilot increases back pressure on the control column to recover the lost altitude. In doing so, he momentarily increases the AoA to increase lift (and wing loading). The manoeuvre margin (ie the difference between MMS and the top of the red band, or stick shaker activation speed) is a function of the AoA margin to stall warning. In other words, if the pilot increases back pressure on the control column to increase the AoA, the stick shaker activation speed displayed on the PFD will increase.

    Some aircraft also have a dedicated AoA indication at the top right hand corner of the PFD. It’s not mentioned in the MAS FCOM, so I assume the MAS aircraft did not have that option installed.

  236. Andrew says:

    @Victor

    RE: “We have to remember that the autothrottle will remain engaged after the autopilot is disengaged.”

    The L & R autothrottle servo-motors (ASM) are powered by the L & R 28V DC buses respectively. If the L IDG and backup generator are selected OFF, both ASMs will lose power when the right engine flames out.

  237. Neville says:

    @haxi says

    I want to share a link to a new National Geographic documentary on the search of MH370 which I think worth watching. It’s part of Nat-Geo’s “Drain the Oceans” series.
    https://www.nationalgeographic.com/tv/watch/fe70898178fdcafb39ccb7b324c81e17/

    The documentary was worthwhile, but I was told I could only watch it if “I interacted with advertising”. Later I was told I could only continue to watch it if I had a TV contract, which I do not have. It seems I hadn’t “interacted” with the advertising sufficiently.

  238. David says:

    @Victor. “My understanding is the parts have not yet been sent to Malaysia as of days ago.” Thank you.

    I hope there is some explanation as to why not and why in lieu it could not have been picked up (rhetorical).

  239. David says:

    @Nederland. What happened to the flaperon? There are many possibilities but I have found none consistent with all the evidence.

    I lay out some thinking in the below in case anyone can help with a solution.
    Warning: heavy going:

    https://www.dropbox.com/s/lgws5w8br6bzol9/Flaperon%20Separation%2C%20MH370.docx?dl=0

  240. Victor Iannello says:

    @Andrew: Really? I haven’t check the schematics, but I’m surprised that the batteries don’t power the ASMs. Does the autothrottle disengage, or does the thrust setting simply not change? What happens when the APU comes online? Is autothrottle not available under RAT power?

  241. Andrew says:

    @Victor

    I emailed you a diagram of the autothrottle system showing the power supply. The autothrottle system is not a critical item and there is no requirement for power supply redundancy for the ASMs. The aircraft can be dispatched with one or both ASMs inoperative.

    The L & R DC buses are normally powered by their respective XFR buses via the L & R TRUs. If the XFR bus on one side is unpowered, the DC bus on that side is powered by the opposite side DC bus via the DC BUS TIE relay. If both XFR buses are unpowered, the L & R DC buses are not powered. The STBY power system, including the RAT and the batteries, does not power the L & R DC buses.

    Power to the ASMs is restored if the APU autostarts and provides power to the R XFR bus.* However, IIRC the TMCF servo loop monitor will have set an inhibit when the power previously failed. The inhibit is only removed if the A/T ARM switch on the MCP is cycled.

    *In this scenario only the R XFR bus is re-powered because the L BUG is selected OFF. The L & R DC buses will both be powered by the R XFR bus via the R TRU.

  242. Andrew says:

    @Victor

    ERRATUM: The last paragraph in my previous post should read:

    Power to the ASMs is restored if the APU autostarts and provides power to the XFR buses. However, IIRC the TMCF servo loop monitor will have set an inhibit when the power previously failed. The inhibit is only removed if the A/T ARM switch on the MCP is cycled.

    Ignore the note with the asterisk.

  243. Gysbreght says:

    @David: You wrote “I have trouble with the phugoids at 30 deg bank. First, the amplitude of the phugoid will need to be doubled to realise the same descents and vertical acceleration. “

    Why 30 deg bank?
    Amplitude doubled relative to what?

    In the paper I posted yesterday I show that in the simulation that achieved the final BFOs rates of descent the bankangle was 5 degrees on average.

  244. Paul Smithson says:

    Folks. Is the conspicuous silence about low speed from the turnback to ~1731 because a) folks don’t believe it b) don’t find it surprising c) don’t know what to make of it? I suggest that it is both interesting and important. Why would he do that??

  245. Nederland says:

    @David

    Thank you.

    Question: If the flaperon (and presumably the outboard flap, too) came off in flight as a result of flutter, then how would you explain the damage to either trailing edge? What could have caused this other than impact with the water?

  246. Tim says:

    Re the autothrottle servo power.
    All the different configs considered over the years……autopilot on/off, autothrottle on/off I don’t think we’ve ever considered the case of just one autothrottle servo working. This brings in a new level of complexity. Something else that must be put to the test in a simulator.

    Fo example, how would an aircraft preform in a phugoid when only one autothrottle is working? Does it explain the 2 turns we know about and would there be many more turns we don’t know about?

    And surely, now as the IGARI turn is an autopilot off manoeuvre, the rest of the flight is likely to be a passive autopilot off flight.

  247. David says:

    @Nederland. Were the flaperon to flutter within the flutter envelope because the inner actuator was inoperative, the trailing edge could separate. Should it, as per an earlier post the flaperon might well deploy immediately and then separate from overstress. However in so doing its separation would not of course be from the ATSB’s neutral.

    Outer flap separation could result from the flaperon hitting it. All conjectural though. Other than that while the French reckon a separated flaperon would descend nose down from a separation on high they indicate also that that needs trialling. The trailing edge might come off on the flaperon striking the sea after descending flatter.

    Or it could come off in shock such as at wing breakage, mid air or collision with the sea.

  248. Gysbreght says:

    More on end-of-flight simulation bank angles:

    The chart shows the bank angle calculated from the track radius of turn at 460 kt TAS. The autopilot was lost at the first engine flame-out, and the initial groudspeed derived from the track was 426 kt. The airplane first banked 3 degrees to the right, then banked left and continued in a left turn until it ended in a tight spiral dive, probably due to the second engine flaming out with the airplane remaining trimmed for one-engine-inoperative.

    CAUTION:The assumed airspeed of 460 kt TAS is probably approximately correct for the first 2 minutes. After that the airplane will increasingly accelerate downwards due to the increasing bank angle, the airspeed will increase and consequently the bank angle that corresponds to the track radius will be greater than that shown on the chart.

    https://www.dropbox.com/s/rkqxcefjlnzqm7s/Trk8_bankangles.pdf?dl=0

  249. Gysbreght says:

    @Victor Iannello: You wrote on August 4, 2018 at 7:47 am –
    If you consider a level phugoid with near constant angle-of-attack (which is not entirely accurate), …

    I trust your upcoming paper will elaborate on that statement and am looking forward to that.

  250. TBill says:

    @Paul Smithson
    Possiblities-
    IGARI slow down was:
    > climb to high altitude
    > Loss of engine
    > Never happened (IGARI turn did not start until closer to BITOD)
    > Other reasons?

  251. Victor Iannello says:

    @Gysbreght: Constant AoA won’t occur unless (Cl/Cd) is constant, which won’t occur over the range of Mach numbers and bank angles we expect to see. It’s a simplification that allows analytical solutions but it is only an approximation.

    I agree qualitatively with your description of what occurred in the simulation. The key is that when the thrust is loss from the left engine, the residual left rudder from the TAC causes a roll to the left, resulting in an increasingly steep descent. As we don’t know exactly when the left engine flamed out relative to the timing of the flame out of the right engine, timing of the increasingly steep descent as measured by the BFO can be explained.

  252. Gysbreght says:

    @Victor Iannello: Thanks for your reply. I don’t agree but will keep my arguments to myself.

  253. Nederland says:

    @David

    But would the flaperon and outboard flap fall down fast enough to cause the entire trailing edges to break off on impact with water?

  254. Gysbreght says:

    @Nederland: I think an airfoil with the shape and weight distribution of the flaperon, in free fall, will pitch up continuously and descend in a succession of loopings.

  255. flatpack says:

    @Paul Smithson

    You suggest that the apparent low speed from the turnback to ~1731 is both interesting and important.

    I agree but take it slightly further and consider that the flight profile from the turnback through to the apparent extremely low pass over Pulau Perak is the critical phase of the flight and that fully comprehending it is key to everything.

    Like TBill I think there was a ‘climb to high altitude’ and I have previously advocated a simulated ‘engine loss’ so I cannot argue with that other than to suggest it happened later on than he implies.

    Personally, I think there were also ‘extreme manoeuvres’ intended to subjugate the PAX and particularly any crew still sustained by portable oxygen bottles. These may be thought of as nasty versions of the vomit comet profile.

    Gysbreght posted a link to a plot of the Military Radar Data earlier on this thread at July 31, 2018 at 12:42 pm. Unfortunately for my ‘extreme manoeuvres’ theory, Victor posted that ‘the change in altitudes combined with the speeds are impossible for a B777’ and suggests that the military radar system cannot discriminate altitude or other aircraft than 9M-MRO were captured.

    The laws of physics are just that but whether flight envelope protection or engine rev limits can be sufficiently ‘gamed’ by a committed pilot is unclear.

  256. Gysbreght says:

    It seems to me that the discussion of “coasting” lends credibility to an idea that I floated a long time ago: that the sharp corner in the ATSB/NTSB chart of the military radar post-IGARI track is the result of forward projection from the point where the radar echo is lost, and backward projection from the point where it reappears.

  257. Gysbreght says:

    @flatpack: If you discard the impossible, what is left is possible.

  258. TBill says:

    @Gysbreght
    Interesting I do not know, but the Safety Report seems to adamantly say the radar itself did keep track of MH370 as the same aircraft going around IGARI; in other words, the radar kept the same code number for the aircarft…there was no flight interruption, as far as the radar equipment was concerned. Maybe a slight gap but the radar was not fooled.

  259. Gysbreght says:

    @TBill: The Safety Report uses the terms “coasting” and “coded as” only in relation to civil primary radar.

  260. Gysbreght says:

    Safety Report: “It became very apparent, however, that the recorded altitude and speed change “blip” to “blip” were well beyond the capability of the aircraft. It was highlighted to the Team that the altitude and speed extracted from the data are subjected to inherent error.”

    What a revelation! Are the points in Figure 1.1B “blip” to “blip”?

  261. flatpack says:

    @Gysbreght

    On your plot of the Military Radar Data, is the Civil Radar GS (ie the green line) based on the assumption of a constant height of 43kft?

    If this were recalculated using the military height data, would it reduce the difference between the green and red lines, as I suspect?

    I do realise that the military height data is sparse and suspect but maybe there is some mileage in it.

  262. Paul Smithson says:

    @Gysbreght. Are they blip to blip? I think that’s what’s implied, isn’t it? As we saw with the primary radar, range resolution of 0.1NM will immediately give you quantisation of ~ 90knots in speed if the blip periodicity is 4 seconds. Better speed estimates would surely have been obtainable across multiple “blips”.

  263. Gysbreght says:

    @flatpack: Q1: Yes; Q2: Don’t know.

  264. Paul Smithson says:

    As regards coasting/extapolation/interpolation. Yes, I agree that this explanation provides a good fit for some of the anomalies in the “right angle” turnback at IGARI.

    – From IGARI the radar trace continues for 3.7NM (~28s / 7 radar sweeps). I suggest that this was coasting and it had not detected that the target had changed course from the incoming 025.
    – Next we have a discontinuity (abrupt angle) where the radar seems to have determined that a change of course to the right has taken place. This again is approximated to a straight line segment, before another discontinuity.
    – The NE extent of the turn is represented a straight line that lies nearly perpendicular to the bearing of Western Hill. I don’t think it is a coincidence that the target has been re-acquired by radar on this segment when distance to radar is nearly constant. The backward extrapolation of that line and forward extrapolation of the previous one gives you the distinctive “corner”.
    – The radar seems to do a better job of following the target through the final part of the turn and this trace is represented as a curve. Yet even here we have an anomaly. Once the turn is completed and back to a nearly straight path, the path “jumps” outwards to the NW. Again, this smacks to me of an extrapolated course at turn exit that gets “corrected” (jumps to the NW) once there are enough steady-course captures to permit the signal processing to discern more accurately the range and track of the target.

    If the above is true, then the most parsimonious course-fit for the turnback would run as follows. Turnback to the left occurred after the IGARI turn was completed, from the IGARI-BITOD line. The NE extent of the turn (tangent to the outer part of the turn) coincides with the military radar depiction. The return course is perfect reciprocal of IGARI-BITOD (057/237). Diameter of the turn can be determined by backward extrapolation of the course after the “jump”. All of which gives you a straightforward 180 degree turn on to reciprocal heading, that starts at ~17:22:00, performed at ~460kts and AOB25, ending 17:24:37. Maintaining the same speed, the plane arrives on time at the beginning of the KB trace 1730-1731. Only after that do we see the continuous acceleration to ~525-530 groundspeed towards Penang.

  265. David says:

    @Nederland. “But would the flaperon and outboard flap fall down fast enough to cause the entire trailing edges to break off on impact with water?”

    While as Gysbreght says it might pitch up continuously it is possible too that being asymmetric in plan it will rotate, ie spiral.

    Either way I agree with what you imply that the trailing edge being detached after separating mid air looks much less likely than before flaperon separation. While a deployed ditching looks to be commonsense to many, contra that is the apparent collision between the flaperon and flap when both were deployed and the high energy impacts on the flaperon upper surface.

    There is no resolution of that on offer, so far.

    I have taken another look at what the French report at its p19 means when it says, “Plate n°40 shows a slight distortion of the outboard hinge fitting towards the outside of the aeroplane. The inboard hinge fitting also shows minor distortion towards the outboard hinge fitting (outside of the aeroplane, see plate n°41).”

    I think I have misinterpreted that description so aim to draw up a new version excluding flaperon inboard movement as initiating the sequence. Unfortunately that will not help with what could have done the initiation other than water strike in a ditching. Also, it will not alter what followed, but it will reduce inconsistency with the outer flap damage.

    The new version also will withdraw the need for a steep-sided wave trough to account for the main impact being at the flaperon’s inboard end in a flaperon-deployed ditching, noting the alleviating effect of hinge angle in a photo I have used here before:
    https://www.dropbox.com/s/2zxxvjbhzp219k3/dihedral%20and%20flap%20hinge%20angles.docx?dl=0

    “Deployed” here does not mean the flaperon would be drooped, as would accompany flap extension.

  266. David says:

    Line 7 “both were” should read “neither was”

  267. Niu Yunu says:

    Victor Iannello: “Please explain the capture of the aircraft by civil and military radar.”

    Victor, what is uenxplicable about MH370 having been racked by civil radar? As I understand it, civil radar tracked the flight until the transponder was disabled and could not track it anymore beyond that moment.

    Victor Iannello: “even if you dismiss all the military radar, we still have the civil radar, which shows an aircraft with no transponder traveling back over the Malay peninsula.”

    Civil radar is capable of tracking aircraft without transponder ??

    Doesn’t civil radar only have secondary radar requiring a working transponder?
    I thought only military radar has primary radar ?

    I’m somewhat baffled as the quoted statements seem to run counter to long-established knowledge on your blog.

  268. Andrew says:

    @Niu Yunu

    RE: “Civil radar is capable of tracking aircraft without transponder ??

    Doesn’t civil radar only have secondary radar requiring a working transponder?
    I thought only military radar has primary radar ?”

    Civil ATC radar includes both primary and secondary surveillance radar, however, the controllers normally use the secondary returns. As I understand it, the primary returns are usually suppressed on the controller’s display, but are nevertheless recorded and available if necessary. In the MH370 case, the Malaysian FI report (pp.3-4) states the following:

    “From 1730:37 UTC [0130:37 MYT] to 1744:52 UTC [0144:52 MYT] a primary aircraft target was captured by the Terminal Primary Approach Radar located to the south of Kota Bharu Airport runway…”

    “The primary aircraft targets above have been confirmed by DCA and its maintenance contractor, Advanced Air Traffic Systems (M) Sdn. Bhd. (AAT), that it was Kota Bharu Primary Radar Station which captured them.”

  269. Nederland says:

    @David

    Presuming that the ATSB is correct in their assessment that the outboard flap was retracted and that also there were no hydraulics available to extend the flaps, could it be that the airplane glided and came down in a tail-down attitude? The tail section dissolved first, therefore the vertical stabiliser panel came off too, with parts of the tail section hitting the flaperon upper surface.

    I’d also say it’s still best to look further north, and close to the arc, as the next step. The end of flight scenario is not really conclusive, there is even the possibility the aircraft was controlled and ditched, flaps down, close to the arc. Like the French report on marine life, the Australian report also says that the debris was drifting in tropical waters, twelve months or so before getting stranded. (ATSB, Operational Search, Appendix G)

    Thereafter, it could make sense to look into specific areas c. 90 nm beyond the 7th arc that have not been covered during the surface search, are reasonably far north, consistent with drift patterns, and provide a credible (piloted) flight path (rather than autopiloted straight path).

  270. Gysbreght says:

    @PaulS: The Safety Report has this footnote on page 2:

    “Waypoint – A specified geographical location used to define an area navigation route or the flight path of an aircraft employing area navigation. Waypoints are identified as either:
    • Fly-by waypoint – A waypoint which requires turn anticipation to allow tangential interception of the next segment of a route or procedure, or
    • Flyover waypoint – A waypoint at which a turn is initiated in order to join the next segment of a route or procedure.”

    The radar data show, and the report narrative confirms, that the airplane passed over waypoint IGARI before startung to turn right to join the IGARI-BITOD route segment. That seems to have been ignored in the specification of the entry and exit waypoints for the turn simulations.

    Another picture of the radar trace from the ATSB report of June 2014:
    https://www.dropbox.com/s/83on19lhfl7dhxa/Fig2_Radar_detail.jpg?dl=0

  271. haxi says:

    @Victor,

    Just read Naren’s post. The released report is not a “Full Report”, as the Investigation Team told Chinese NOK members last week (after strong demands from the NOK at a meetup in Beijing) that they would share a more detailed analysis on fuel consumption, although they didn’t say when. The NOK source estimated the fuel analysis might come in October.

  272. Gysbreght says:

    @PaulS:

    BTW, I don’t really care whether it can be shown that the turnback could have been accomplished with autopilot engaged. A single person wanting to turnback immediately grabs the control wheel. He would not start fiddling with course selector or CDU keyboard. The radar data in its entirety, considering track, speed and altitude, clearly shows that the autopilot was off.

  273. Victor Iannello says:

    Niu Yunu: In addition to what @Andrew said, I wrote an article discussing the civil PSR data that Mike Exner was able to obtain.

  274. David says:

    @Victor. Cocos fly-by. You have been curious as to whether there was infrasound evidence of that, or a lack of when there should have been. As you know, Lawrence Livermore Laboratories looked at the infrasound record from the Cocos Islands’ sensor but were unable to find trace of aircraft which approached and departed Cocos at known times. Hence they deduced it was pointless looking for any signature generated by MH370 in the vicinity.
    https://casis.llnl.gov/content/pages/casis-2014/docs/poster/Kane1-CASIS-2014.pdf

    Taking the 7th arc crash site as something over 500NM from Cocos and deducing a time bracket when MH370 would have passed by to reach that, I find there is a blip in the LL Cocos record that falls within that, shown in the second trace down, under “Cocos Waveforms”, in their above poster. Given the LL experience, this being an aircraft signature is a long shot but bearing in mind that detections are sensitive to conditions and that an aircraft passing might have led to a different infrasound carry (height, wind, engine noise) than an approach or departure it does not seem absolutely impossible that this could be of interest.

    The blip, at 10000 secs, is expanded here:
    https://www.dropbox.com/s/jrtsa320i1tgyoh/Cocos%20infrasound%20record%20approaching%202400%20UTC%2C%207th%20March..png?dl=0

    You can see it in the H1 – H4 records,
    https://www.dropbox.com/s/ei0rjpnoprvos8d/Cocos%20Hydrophones%201-4.jpg?dl=0

    though it is indiscernible in the others:
    https://www.dropbox.com/s/qk8m57tp7ulefn5/Cocos%20Hydrophones%205-%208.jpg?dl=0

    10000 secs after 20:00:00 makes it 22:46:40. This is 1 hr 33 min before the earliest time of the crash, when, had it gone 550 NM beyond Cocos for example, it would have averaged 355 KTAS during that time.
    In other words should there be other reason to think that the aircraft might have passed Cocos at that time, this could be looked at more closely, though not otherwise since most probably that would not be worth it.

  275. Victor Iannello says:

    @haxi: In that piece, Naren did an excellent job describing the Safety Report and the clumsy way the NOK were briefed in Malaysia. All should read it.

  276. Victor Iannello says:

    @David: Thank you. It’s hard to use this as a path discriminator.

  277. Paul Smithson says:

    @Gysbreght. Regarding programmed waypoints. I sought clarification previously on this point from pilots. The aircraft flies along the legs from waypoint to waypoint in its programmed flight path. It does not fly OVER every waypoint because this would result in every course change being over-shot. Instead, the turn is begun slightly before the programmed waypoint so that mid-way through the turn the waypoint is abeam. Assuming 470kts and AOB25, the aircraft would have begun the programmed turn about 15 seconds before IGARI. At mid-turn with IGARI abeam, the aircraft should have been ~0.3NM to the right (SE) of IGARI waypoint.

  278. Paul Smithson says:

    BTW, I don’t really care whether it can be shown that the turnback could have been accomplished with autopilot engaged. A single person wanting to turnback immediately grabs the control wheel. He would not start fiddling with course selector or CDU keyboard.

    Really? Would you not consider selecting DIVERT NOW (if that was your intention) or, alternatively, selecting a 180 degree change in heading?

    The radar data in its entirety, considering track, speed and altitude, clearly shows that the autopilot was off.

    I don’t think it “clearly shows” anything of the sort.

  279. Gysbreght says:

    @Paul Smithson: RE Your August 6, 2018 at 8:05 am:

    Yes, I know the normal practice is waypoint Fly-By as you describe.

    The point is that’s not what MH370 did at IGARI.

  280. Gysbreght says:

    I suggest you plot waypoint IGARI on the chart I posted, with lat/lon grid.

  281. Gysbreght says:

    Another aspect where the description of the turnback simulations is deficient concerns the bank angle. The default max is 15 degrees, and the pilot can select max 25 degrees. The right turn starting just after IGARI was flown with 15 degrees or less of bank. So when did the pilot select max 25 degrees?

  282. Gysbreght says:

    @Victor Iannello: The link in your post at 7:57 am is the same as 3 minutes earlier.

  283. Victor Iannello says:

    @Paul Smithson: Other than in a STAR or SID, it would be rare to see a fly-over waypoint in a flight plan. (I don’t know how to manually enter one in a route using the CDU, if that’s even possible.) If MH370 flew over IGARI and turned towards BITOD only after passing IGARI, that would indicate that the aircraft was NOT flying in LNAV on a DIRECT to IGARI-BITOD path. If on autopilot, more likely it was on a TRK/HDG Hold, or TRK/HDG SEL. The turn towards BITOD would have been TRK/HDG SEL. However, there would be no reason for the crew to be in any autopilot mode than LNAV/VNAV at that point.

    This is significant because the first actions taken for the diversion could have occurred before reaching IGARI. This compresses the timeline between the last radio communication and the diversion, and would leave little time for a takeover of the flight deck from a third party.

  284. Paul Smithson says:

    @Gysbreght. You appear to think that the aircraft followed the path indicated in the military radar trace. It cannot have done so literally in every respect because it is impossible and even a Kalman-filtered version of it proved to imply infeasible changes in speed. What is “perfectly clear” is that this radar trace includes several discontinuities. That implies that parts of the trace are “coasting” whereas others are likely a line joining up a gap where target had been dropped. So I think we can very safely say that the aircraft did NOT fly the turn shape depicted.

    Since the target clearly dropped/became intermittent what we CAN expect is a short coasting segment before the drop. I suggest that this is exactly what is seen at IGARI on the military trace.

  285. Victor Iannello says:

    @Gysbreght: Thank you. The link should be correct now.

  286. Gysbreght says:

    @Paul Smithson: I believe the narrative of the report is correct on that point. I also believe that the well defined radar track correctly depicts the airplane path at least until “Last secondary radar data 1722”.

    Did you plot IGARI on the chart and do you believe that it just a coincidence that it is smack-on the radar trace?

  287. Andrew says:

    @Gysbreght

    RE: ”Another aspect where the description of the turnback simulations is deficient concerns the bank angle. The default max is 15 degrees, and the pilot can select max 25 degrees.”

    Only in HDG SEL/TRK SEL modes (at high TAS). In LNAV mode, the AFDS will limit the bank angle to 25° (30° in a holding pattern), subject to buffet/thrust limits. The Safety Investigation Report states that LNAV mode was used for the turn back simulations.

  288. Victor Iannello says:

    @Andrew said: The manoeuvre margin (ie the difference between MMS and the top of the red band, or stick shaker activation speed) is a function of the AoA margin to stall warning. In other words, if the pilot increases back pressure on the control column to increase the AoA, the stick shaker activation speed displayed on the PFD will increase.

    Sorry I missed this comment. I think we are in agreement that the MMS-stall margin as displayed on the PFD is a function of wing loading.

    In a turn, it is easy to see the barber pole shrinking as the stall speed creeps up. Whether or not the AoA option was displayed in the PFD of 9M-MRO (along with the red tick that indicates the stall AoA), my guess is that pilots are looking more closely at speed. The maneuver margin goes to zero just as the AoA reaaches the stall value.

    However, in the PMDG 777 simulation, if I enter into a steady descent and quickly pull back to load the wings, I can get the stick to shake even if there is still maneuver margin shown in the PFD. While the AoA indicator shows the instantaneous value, the maneuver margin responds more slowly. For a level, steady turn, the two are consistent.

  289. Gysbreght says:

    @Andrew: Thank you for putting that straight. I apologize for writing off memory.

  290. Paul Smithson says:

    @Gysbreght. Um, no, it’s not a coincidence. If the radar fails to follow the target when it is turning and instead projects a straight “coasting” path then of course it will fly directly over the waypoint of the leg it was on.

  291. Gysbreght says:

    @Paul Smithson: How about the secondary radar data?

  292. TBill says:

    @Andrew
    @Niu Yunu
    Andrew that explanation makes sense to me…I have a family member AT controller on duty 9/11 I think he recently told me they were able to watch the second plane go down the Hudson river to NYC on primary radar. We were of course discussing MH370 at the time.

  293. TBill says:

    @David
    One thing about that solitary blip at COCOS, it could be MH370 in the sense I do not expect much other air traffic at that time of night in March. If you look at FR24 tolday you might see other traffic, but look again after daylight savings time goes into effect around November, and I believe that probably clears the air traffic out of the area an hour earlier.

  294. Gysbreght says:

    @Andrew:

    Safety Report Figure 1.1A “MH370 over waypoint IGARI at 1720:31 UTC”.

    Is that compatible with the airplane being in LNAV?

  295. Gysbreght says:

    @Paul Smithson: If I understand you correctly, you explain the radar trace by assuming that the radar lost the target before IGARI, projected its path to a point beyond IGARI, then picked up the target again on the projected path north of IGARI while it was turning steadily towards a heading south of BITOD to intercept the IGARI-BITOD route segment, when the target was actually passing south of IGARI in the normal Fly-By mode according to flight plan.

    Frankly, I’m beginning to doubt your sincerity.

  296. Paul Smithson says:

    @Gysbreght, would you mind sharing your kmz file of the “SSR” and the source doc that it is derived from? I recall seeing it depicted in reports but haven’t seen anything with LAT/LONG or timestamps beyond last ACARS.

  297. Gysbreght says:

    @Paul Smithson: I don’t have a kmz file. What I posted is a cropped enlargement of a detail from Figure 2 of ATSB report “MH370 – Definition of Underwater Search Areas”, dated 26 June 2014, available on the ATSB website.

  298. David says:

    @Victor. Yes the infrasound blip is by no means stand-alone evidence.
    However I meant to point out where it is in the H1-H4 traces in case that was not obvious. Here that is, for the record:

    https://www.dropbox.com/s/5npt135wkbj8uv7/Cocos%20infrasound%20H1-H4%20blip%20pointed%20out.jpg?dl=0

  299. Andrew says:

    @Gysbreght

    RE: ”Is that compatible with the airplane being in LNAV?”

    The track change at IGARI is small. In LNAV mode the aircraft should have passed very close to, but not directly overhead IGARI. @Paul calculated ~0.3nm inside IGARI. I very much doubt the radar had sufficient resolution to tell the difference. As far as anyone watching the display was concerned, the aircraft passed ‘over’ IGARI.

  300. Gysbreght says:

    @Paul Smithson: That chart was first published in an ATSB ‘factsheet’ dated 26 May 2014, Figure 1: “MH370 Flight path derived from Primary and Secondary radar data”; Source: NTSB/Google; without the latitude/longitude grid and in poorer resolution.

  301. Niu Yunu says:

    @Andrew
    @Victor
    Thanks all for the clarification!
    Victor, I read your article, but somehow I thought the whole time that the data you presented came from military radar, not civilian radar (as I was absolutely sure that civilian radar only has secondary radar at their disposal). Sorry for my confusion.

  302. Paul Smithson says:

    1) Page 4. “The Mode S symbol of MH370 dropped off from radar display at 1720:36 UTC [0120:36 MYT], and the last secondary radar position symbol of MH370 was recorded at 1721:13 UTC [0121:13 MYT].” I’m no radar expert, but that last secondary radar position is 59 seconds after “Mode S” drops = 10 radar cycles @4s. Coasting?

    2) Mode S is a Secondary Surveillance Radar process that allows selective interrogation of aircraft according to the unique 24-bit address assigned to each aircraft. So from 17:20:36 [possibly up to 4s earlier] the aircraft unique address has been lost – so aircraft can no longer be interrogated by the SSR, right? Why do we think that positions between then and 39s later are valid?

    3) The FI and SR do not tell us when the right turn at IGARI began or precisely how close to IGARI the “over IGARI” nomenclature means.

    I’m not nit-picking for the sake of it. I believe that the simulated turn that I provided earlier is the best path explanation of how the aircraft got from IGARI to 1731 and I think it provides an excellent fit to the data – including correctly predicting the exit position/time and the speed at which it must have been flown.

  303. Victor Iannello says:

    @Andrew: Using the ADS-B data published by FR24, it’s clear that the turn was anticipated as it would be for a fly-by waypoint. The plane was in LNAV mode as it passed near IGARI.

  304. Paul Smithson says:

    correction: 17:20:36 to 17:21:13 = 37s, so ~9 sweeps @4s

  305. Andrew says:

    @Victor

    Thanks.

    RE: ”it would be rare to see a fly-over waypoint in a flight plan. (I don’t know how to manually enter one in a route using the CDU, if that’s even possible.)”

    Airbus FMCs provide that capability. Boeing FMCs do not! A fly-over waypoint would only be seen if it was coded as part of a SID or STAR, as you mentioned.

  306. Paul Smithson says:

    @Gysbreght. If that is the source to which you are referring then I don’t think it tells us anything very precise at all about what SSR shows or doesn’t show around IGARI.

  307. Gysbreght says:

    @Andrew: According to the radar track (SSR at that point) the airplane starts the turn at IGARI. The airplane needs to start turning 4.3 NM before IGARI to join the IGARI-BITOD route in fly-by fashion.

  308. Paul Smithson says:

    @Victor. Thank you for reminding us of the ADSB dataset which Sladen kindly assembled and posted at https://github.com/sladen/inmarsat-9m-mro/blob/master/ads-b/all-combined.csv

    This confirms the approach from last ACARS towards waypoint IGARI was precisely on-track towards IGARI, perhaps 3s late to have IGARI abeam at 17:20:31. The change of course to 028 seems to shows the start of the turn where and when you would expect. And the last heading of 040 is roughly halfway through the turn – which is also what you would expect for that time. So the programmed turn at IGARI began exactly as expected, at least as far as 17:20:36. 10 seconds later (15s after IGARI) would expect the aircraft to have completed its turn and be tracking 059 towards BITOD.

  309. Victor Iannello says:

    @Paul Smithson: If MH370 was not in LNAV at IGARI, that would be a really big deal. The review has value.

  310. Greg says:

    118/5000
    MH370 was in the LNAV by using the shortcut for IGARI according to the ATC command but without the continuation setting. And that’s all.

  311. Greg says:

    What was the relationship between the pilots.
    Important information is on top ..
    Read the quotes from the FR carefully.
    You are an intelligent guy, you will have your own opinion.
    FR Page 47
    “The Pilot-in-Command (PIC) signed in for duty at 1450 UTC [2250 MYT], 07 March 2014 followed by the First Officer (FO) who signed in 25 minutes later. The MAS Operations Despatch Centre (ODC) released the flight at around 1515 UTC [2315 MYT].
    The PIC, an authorised examiner for the Department of Civil Aviation (DCA), Malaysia, was conducting the last phase of line training for the FO, who was transitioning to the Boeing 777 (B777) aircraft type from the Airbus A330. As the FO was certified functional during his last line training flight, no additional pilot was required as safety pilot on MH370. It has been established that the PIC had assigned the FO to be the Pilot Flying for this flight.”
    FR Page 332
    “2. 1450 UTC [2250 MYT] PIC of MH370 signed in for duty.
    3. 1515 UTC [2315 MYT] FO of MH370 signed in for duty. MAS Operations Despatch Centre (ODC) released flight.
    As per operational requirements.”
    FR Page 265
    “(1) Operation Control Centre
    The Operation Control Centre (OCC) was where the briefing of flight crew and cabin crew took place. A team of Licensed Aircraft Despatchers were stationed in this Department.
    Besides the crew formalities required prior to departure, the flight crew would be working in tandem with the assigned despatcher to review all documentations related to the assigned flight which influenced the decision on the finalised routing and fuel ordered by the Captain of the flight.”

  312. Gysbreght says:

    The FR24 data don’t show any turn. Have the course data been added manually (See Sladen’s README file)?
    https://www.dropbox.com/s/esdtqhb8lp073iu/FR24_ADS-B.pdf?dl=0

  313. Andrew says:

    @Gysbreght
    @Victor
    @Paul Smithson

    RE: Gysbreght’s comment “According to the radar track (SSR at that point) the airplane starts the turn at IGARI. The airplane needs to start turning 4.3 NM before IGARI to join the IGARI-BITOD route in fly-by fashion.”

    I plotted the flight planned tracks and the position of IGARI on the chart that @Gysbreght mentioned:
    https://www.dropbox.com/s/wf670bt2t011oti/Fig2_Radar_detail.JPG?dl=0

    I agree, the plot does show the aircraft overflew IGARI instead of turning beforehand. I can’t say if that’s due to inaccurate plotting or the way the aircraft was flown.

  314. Andrew says:

    The resolution of the image in the previous file wasn’t great. This one’s a bit better:

    https://www.dropbox.com/s/kaoy6z9ixm1fkjb/Fig2_Radar_detail2.png?dl=0

  315. David says:

    @TBill. Cocos infrasound spike. Thanks, though it could be from something quite different to an aircraft. It would take some research to get at that and even then there might be no confident result.

    However in examining any postulated 7th arc position, it might be interesting to use this as a guide to the implications in case it should be from MH370.

    Another example. 380 miles (700 km) beyond Cocos to that last aircraft transmission: 1.55 hrs for that, average speed thence 245 KTAS.

    Even including any slow down after right engine failure, low level after Cocos is indicated.

  316. Victor Iannello says:

    @Andrew, @Paul Smithson, @Gysbreght: We need the right data to get the right answer. The final data points from FR24 were likely MLAT rather than ADS-B, and Paul Sladen chose to ignore those points due to the lower resolution. I’ve solicited Mike Exner’s help in getting the SSR data around IGARI that is graphically shown in official publications.

    As I said before, if MH370 was not in LNAV at IGARI, that would be a really big deal. This review has value, even if this has been discussed before.

  317. sk999 says:

    All,

    I have stepped away a bit from MH370, so the following may not be keeping up with the current discussion.

    It should be remembered that there are two sets of SSR data being presented in all of the reports – one set from KL ATCC (civilian) with data points every 4 sec (give or take), and one set from military radar (presumably the RAT-31DL on Western Hill) with data points every 10 sec. The latter is what is shown, e.g., in the ATSB ‘factsheet’ dated 26 May 2014, Figure 1, and even earlier in a figure on Hishammudin Hussein’s Facebook page: https://www.facebook.com/media/set/
    ?set=a.10152221414824355.1073741996.72613804354&type=3 – this conclusion being based on textual descriptions given in multiple reports. From what I can tell (not an in-depth study), the two sets are entirely consistent.

    Regarding the flyby at IGARI (it’s a flyby waypoint), quite a while ago I compared the heading from the last reliable data point in Sladen’s compilation (at 17:20:18) with predictions based on ADS-B data for other flights, including a random Boeing 777 flight (turn rate 0.48 deg/sec) and MH360, an Airbus 330, on Oct 16, 2016, which flew the exact same flight profile as MH370 (direct IGARI, turn rate 0.32 deg/sec to BITOD). I estimated that the heading should have changed from 25 to at least 36 degrees at the 17:20:18 point, whereas the actual heading at that time was only 28 degrees. My conclusion was that the diversion was already in progress, before IGARI. Not stong evidence, but anyone here is free to check or refute that conclusion – there’s a wealth of information available on FR24 and FAW.

  318. Andrew says:

    @Victor

    RE: “However, in the PMDG 777 simulation, if I enter into a steady descent and quickly pull back to load the wings, I can get the stick to shake even if there is still maneuver margin shown in the PFD. While the AoA indicator shows the instantaneous value, the maneuver margin responds more slowly.”

    Interesting – In Level D simulators the minimum speed (ie top of the red band) displayed on the PFD pretty well matches the point at which the stick shaker activates. If the pilot pulls too hard while manoeuvring at high altitude, the minimum speed increases straight away and the stick shaker activates when that speed equals the current speed.

    Here’s what Boeing has to say:

    Of particular interest are the minimum speed amber and red bands, or barber pole. At low speeds on Boeing-designed airplanes currently in production, these indications are based on sensed AOA and the AOA margin to stick shaker. At higher Mach numbers, most airplanes with fixed AOA stall warning schedules show margins to stick shaker or margin to initial buffet, whichever corresponds to the highest speed. On these airplanes, the margin to buffet at higher Mach numbers is calculated by the FMC.

    On newer models, such as the 777 and 767-400, the amber and red bands show margin to stall warning at all times because the stall warning schedule generally follows the initial buffet boundary at higher speeds up to cruise. The position of the amber and red bands is always a function of AOA margin to stall warning.

    The speed tape is designed to provide the flight crew with situational awareness of the flight envelope.
    It shows the crew where the airplane speed is relative to the limits (i.e., maximum placard speeds or minimum stall warning speed, as well as the maneuvering capability available).

    http://www.boeing.com/commercial/aeromagazine/aero_12/attack_story.html

  319. ventus45 says:

    sk999
    you said: “I estimated that the heading should have changed from 25 to at least 36 degrees at the 17:20:18 point, whereas the actual heading at that time was only 28 degrees. My conclusion was that the diversion was already in progress, before IGARI.”

    So, are you suggesting, that the A/P had, most probably, already been switch off, a little bit before IGARI (say 5 miles or so) to prevent the normal Lnav “pass Igari close abeam (to the south)” onto the Bitod track, such that the aircraft was then hand flown directly over Igari, then slightly turned to the right (slightly to the north of Igari – not south), but that right turn was not far enough to actually head for Bitod, and then the left turn was executed, manually, ie, hand flown ?

  320. sk999 says:

    ventus45,

    All we know is what the plane’s instrumentation said it did. How it got there, your guess is as good as mine. I opine nothing.

  321. Paul Smithson says:

    @sk999, @Victor, @Gysbreght, @Tbill, @Andrew

    1) Gysbreght – my reference to start of a turn refers to a change of heading from 025 (all previous points) to 028 and not to a discernible lateral shift off track.
    2) sk999. My expectation of turn dynamics is as follows. For GS 472 and AOB25 you complete 360 turn in 334s yielding rate of turn 1.08 degrees per second. Course pre IGARI 025, course IGARI-BITOD 059 so 34 degrees of turn to be made, with halfway point (040.8) expected abeam IGARI. At 17:20:18 expected heading is 026.8 – ie turn has just begun. The reason that our expectations are different (and that you conclude A/P already disengaged) is that you have a different rate of turn to me (0.32/s vs 1.08/s). This could easily be resolved by @Andrew’s experience, or by @Victor or @TBill on their sims.
    3) According to my assumptions above, if 17:20:31 IGARI abeam is accurate, we would expect track 026.8 (027) at 17:20:18 as compared to previous steady state track of 025. At IGARI we would expect track 040.8 (041) – and still turning right at 1.1 degrees per second.

  322. Andrew says:

    @Paul Smithson

    RE: “My expectation of turn dynamics is as follows. For GS 472 and AOB25 you complete 360 turn in 334s yielding rate of turn 1.08 degrees per second.”

    The AOB the AFDS uses in LNAV mode isn’t fixed; it depends on the size of the track change and the groundspeed. The maximum AOB is 25°, but for smaller track changes, such as occurred here, the AFDS would use a lower AOB. I’m guessing it would be somewhere around 15°, which would obviously generate a lower rate of turn. Perhaps someone with a simulator can try it out and report back.

  323. Paul Smithson says:

    Thanks @Andrew. It will be interesting to see if Victor or TBill can give it a go in the sim and see what the standard LNAV turn looks like.

  324. Andrew says:

    @Paul Smithson
    @sk999

    Further to my previous comments, be careful using A330 figures as a guide to what might be expected in the B777; the turn rates are very different, as you can see in sk999’s examples. The A330 tends to use a painfully shallow AOB, so much so that it’s not uncommon to be pinged by ATC for deviating too far off track!

  325. Gysbreght says:

    @Paul Smithson:
    ” – my reference to start of a turn refers to a change of heading from 025 (all previous points) to 028 and not to a discernible lateral shift off track.”
    The last two course values in the table, 28 and 40, are obviously fake, invented by a creative mind and added after MH340 became an accident. The same has occurred on similar forums, just read the README file. Credibility zero.

    @Andrew: The radar trace has a constant radius of 14 NM. That requires 13,7 degrees of bank at FL350, ISA+10.7C, M.82, 484kTAS.

  326. David says:

    @Nederland. A glide to a tail down attitude. In a slow speed ditching it would be tail down as would replicating AF447 as I remember. Were the flaperon and tail to collide I would have thought would be from the flaperon ‘flying’ aft on detachment.

    For my part I am hoping that something emerges to give a greater incentive for a new search than is apparent as yet. In so doing, automatically that would solve the length vs width conundrum.

  327. flatpack says:

    @Greg

    The 25 minute sign-in gap intrigued me when I first came across it a long time ago. I thought it just possible that the PIC had arrived early and used the time to make any necessary physical preparatory steps for the jacking.

    The ODC release time had completely escaped me. Taken at face value the FO did not take any part in the doco review. Procedural irregularities aside, this would consolidate the ‘position of authority’ that the PIC held over the FO and make the FO less likely to question any unusual activity.

    So was the PIC early or was the FO late?

  328. Richard Godfrey says:

    @Victor

    I have compared the Kota Bharu raw radar plot with the ADS-B data from 17:17:01 UTC to 17:21:13 UTC.

    It would appear that the MH370 started the turn (toward BITOD) around 27 seconds or 6.556 km before IGARI, passing 0.970 km south of IGARI at a 18° bank angle.

    Details in the linked Excel: https://www.dropbox.com/s/aq3eitl768d03sn/Turnback%20Analysis.xlsx?dl=0

  329. Gysbreght says:

    Strange that I noted this fly-by/flyover issue many times over the years and that never made any waves in the pond here.

  330. Paul Smithson says:

    @Richard G. Thanks for sharing those calculations. I’m not clear where these very precise “Kota Bharu raw radar plot” is sourced from. Could you please clarify the source and how you arrived at such fine resolution lat/long used to infer changes in track?

  331. Richard Godfrey says:

    @Paul

    The track is calculated directly from the raw x-y radar data.

    I am not at liberty to reveal the source.

  332. Paul Smithson says:

    @Richard G. Thanks. It’s interesting to see how consistent this is with other sources. Am I right in thinking that this is secondary radar,though? If it really is primary, I can’t understand how position resolution can be good enough to calculate actual track to a fraction of a degree over 4 second steps. And what is it that you have interpolated if the lat/long comes from the raw data? If it is primary, do we also have positions beyond IGARI?

  333. Richard Godfrey says:

    @Paul

    This is Primary Radar data.

    I have the habit to calculate to a precision an order of magnitude beyond the accuracy.

    Take the results of a comparison to ADS-B with a pinch of salt and two teaspoon’s of sugar.

  334. Victor Iannello says:

    @Richard Godfrey: Excellent. The SSR data from Kota Bharu was exactly what was needed to resolve this. I trust that data more than the graphical representations of the path. The military data suffers from inaccuracy, interpolation, extrapolation, and fabrication.

  335. Paul Smithson says:

    This data also perfectly matches the 3s discrepancy that I noted earlier between FR24 positions approaching IGARI and the backward projection of IGARI abeam (at 472kts). If these radar positions are right, the precise time at which IGARI was abeam ought to have been 17:20:34 and not 17:20:31.

    Why do you interpret from this data start of turn at 17:20:04 though? The track over segment has changed only +0.1 degree by 17:20:18 and it is only after then that we see the number creeping systematically up.

    I’m also interested in how you have worked out bank angle. The numbers that you have presented indicate a tightening turn. Rate of change of track (degrees per second) goes from 0.5 or 0.6 to 1.7 per second (3s to 17:20:30). Even if radar periodicity is actually constant at ~4 you still have an increasing rate of turn.

  336. Paul Smithson says:

    Thanks, Victor. I think it’s pretty clear that this must be SSR not PSR 😉 If it was PSR, @Richard would not doubt be able to tell us exactly where it went from there!

  337. Victor Iannello says:

    @Andrew: Thank you the facts from Boeing. In a B777, the margin to stick shaker speed is accurately displayed on the speed tape, and corresponds to the the margin to the AoA at stall.

  338. Victor Iannello says:

    @Paul Smithson: To be clear, I don’t think the KB SSR data is either Mode S extended squitter or ADS-B. Rather, it looks like Mode C, i.e., the measured range and azimuth are combined with the (pressure) altitude to estimate X,Y.

  339. Paul Smithson says:

    I admire your optimism @Victor. If Mode C transmits altitude only and position comes from “reflection” then these positions are much too consistent for that. Would you not need to have accuracy <1/100 NM to be able to discern such subtle changes in track across 4s?

  340. Richard Godfrey says:

    @Paul

    If you look at the spreadsheet I linked, at 17:20:00 the track is 25.51 degT and 4 seconds later at 17:20:04 the track is 26.57 degT, which is more than a degree of turn.

    How can you not conclude that a turn had commenced?

  341. Paul Smithson says:

    @Richard. I am looking at your spreadsheet. From 17:20:00 your track angles show (by step): 25.51, 26.57, 26.05, 26.57, 26.57, 25.67 (calculated, 17:20:19). Hence no significant or systematic change in track angle by that stage. From 17:20:18 we go: 25.7, 27.1, 27.6, 29.9 etc. What am I missing?

  342. Paul Smithson says:

    On the basis of your single pair track delta, you could equally have claimed that a turn started at 17:19:07, or 17:19:14, or 17:19:26… I’m more convinced by a monotonic trend than a single point change.

  343. Richard Godfrey says:

    @Paul

    Interesting that I look at the same data and see a sustained change.

  344. Victor Iannello says:

    @Paul Smithson: Between 17:17:01 and 17:20:19, the track angle has a mean of 25.5° and SD=1.4°. The next values are 27.1, 28.1, 31.8, and 38.0. The final value at 17:20:31, representing the average track over the preceding 4 seconds, shows an increase in track angle equal to 9-sigma from the mean. Clearly, a turn was underway.

  345. Paul Smithson says:

    @Victor/Richard. I’m in no doubt at all that the data shows that a turn was underway by the end. I only quibble on RG’s interpretation to pinpoint when it commenced. In any event, it’s of no great consequence. The data demonstrate that the aircraft started its turn at IGARI exactly as you would expect.

    The important points are:
    a) This new data is at odds with the military trace – which shows the aircraft initially continuing straight on over IGARI ~4NM and this lends credence to the interpretation that the military radar was coasting at that stage;
    b) At the time of Mode S drop there is no sign that the aircraft was doing anything other than flying its programmed turn at IGARI.

    I would, in addition, like to say special thanks to RG for sharing this important data, even if the source is “unattributable”.

  346. Gysbreght says:

    At about 17:20 the track changes about 5.5 degrees to the right, and remains on that heading until the end at 17:20:15. Are quantization errors in the data similar to those given to Mike Exner?

  347. TBill says:

    @Paul
    My IGARI FS9 sim case shows turn to IGARI starting 2.7 secs before IGARI, which is probably too late compared to real B777. Recall that @Andrew previously compared real B777 for @DrB’s turn at IGOGU and it was smooth, whereas Victor and I had and over-shoot on both FS9 and FSX. The reason for the over-shoot is apparently a slightly too slow turn anticipation in FS9/FSX. We might be able to tune it by either (1) moving to waypoint DOTEN by hand when we think the turn should start (my usual approach) or fiddling with the Bank Angle setting. But I don’t think we are expecting fidelity to the specific aircraft (B777 vs. Airbus) like Abdrew is describing. We just get a third opinion (Microsoft “game” which I despise that term as it applies to MH370).

    One thing I would say is, due to turn anticipation, I never really found a way to automate the IGARI turn. I always have to manually move to the next waypoint where I think the turn started. So if the PIC put a KB area waypoint after DOTEN in the A/P (which is the way I do it) the A/P could possibly start anticipating this ahead of IGARI, since DOTEN is so close.

  348. Paul Smithson says:

    @Victor & @Tbill. It would be beneficial – to confirm that this part of the record is “as expected” – to run a simulation of such a programmed turn in LNAV. How long before the waypoint does turn commence? How long does it take to roll from wings level to maximum AOB? What is that max AOB? And what does the resultant rate of change in track angle look like through the turn? Any chance that one of you could run a simulation and tell us the result?

  349. Paul Smithson says:

    @Tbill. Our posts crossed. Thanks anyway for trying.

  350. Richard Godfrey says:

    @Paul

    Many thanks for your special thanks, but …

    If it is of no great consequence, why quibble about the timing?

    Why quibble about the source being “unattributable”, when you had no problem previously debating other radar data from the same source.

    It appeared to me that you were, and maybe still are, in denial of any turn starting before IGARI, quoting the discrepancies with the military radar data.

    Below is a link to a graph of the data. The Malaysian report stated that MH370 reached IGARI at 17:20:31 UTC and the graph only runs to 17:20:30 UTC. If you think that MH370 reached IGARI at 17:20:34, I won’t quibble about that.

    The key take away is that a turn clearly started well before reaching waypoint IGARI.

    https://www.dropbox.com/s/w8hqrt4dpalwq6x/Turnback%20Analysis%20Graph.png?dl=0

  351. lkr says:

    Haven’t seen Dennis on this site for about a week. I hope he’s OK, with the huge wildfires sweeping through his area. Air quality is terrible even here, 3-400 miles north of the biggest fires.

  352. TBill says:

    @Richard
    @Paul
    My sense is Paul debates the climb scenario at IGARI, preferring to think of it as a slow down for yet unknown reasons. So Paul tends to opt for a later turn at IGARI, which leaves slightly less time for the climb and subsequent speed up.

    A couple months ago I personally studied the IGARI turn, and I concluded it looked to me like there was a climb/slowdown/speed-up, consistent with @ALSM’s interpretation of the KB civl radar. Nobody at that time really supported what I was saying, and @ALSM only got luke warm (at best) acceptance of his calcs, on the blog here at least. I was the only one totally supportive.

    With the Safety Investigation Report (abbrev: SIR?) I am confused and waiting to hear if @ALSM and others still support @ALSM’s altitude calcs a few months back.

  353. Paul Smithson says:

    @Richard. I think you misunderstand the points I was making.

    You asserted that the data shows a turn starting at 17:20:00. Reviewing the same data, I believe it shows turn starting at 17:20:18. It doesn’t matter a great deal.

    “It appeared to me that you were, and maybe still are, in denial of any turn starting before IGARI, quoting the discrepancies with the military radar data.” Not at all,you will see if you re-read my posts. I am pointing out that a) the data are consistent with the turn that was expected at IGARI b) they don’t match the military data.

    The reason that this is important to me is that I believe the aircraft completed its turn on to the IGARI-BITOD leg, made a 180 turn at 460kts, and flew at the speed all the way to 17:31. I reached this conclusion by painstaking search for a path that could get you from IGARI, around the turn and back to 17:30:33 on time and on trajectory, without overshooting the maximum NE or NW extent of the military trace. My simulation estimated M0.78. It now transpires that the optimum solution is a few knots slower.

    The predictions of that solution have now been proven correct by several new bits of data, viz:

    1) Time and position near IGARI (was in little doubt but now re-confirmed)
    2) Programmed IGARI turn was completed (data confirms this to half-way point)
    2) Time and position of the Exit Waypoint (my position was abeam 1.5NM, correct to less than 1s and on identical track)
    3) Speed required to get from Exit Waypoint to 17:30:33 (matches what I had predicted for the whole leg from turnback to 1731)
    4) Speed observed for the first 60-90 seconds of the Kota Bharu primary radar (as above).

    So far it’s a pretty damn good hypothesis.

  354. DennisW says:

    @ikr

    Thanks for thinking of me. Fire is very close now (a couple of miles). Air quality is horrible. I am near the point of baling. Ami baled a couple of days ago with all our important docs and some memorabilia. I am hanging around only to support local EMS. Nothing I can do to alter what will happen.

  355. Gysbreght says:

    TBill said: “With the Safety Investigation Report (abbrev: SIR?) I am confused and waiting to hear if @ALSM and others still support @ALSM’s altitude calcs a few months back.”

    After a long gestation, I have finally come round (about mid-june 2018) to accepting ALSM’s ideas of altitude as a workablwe hypothesis. This is a plot of the resulting groundspeeds that I posted at the time:

    https://www.dropbox.com/s/dbvvje8xg55t85f/PSR_43kft_speeds.pdf?dl=0

  356. Victor Iannello says:

    @DennisW; Thanks for updating the group. Hopefully the winds keep the fire away from your ranch. Please be safe.

  357. Victor Iannello says:

    @Paul Smithson: One of the points of this exercise was to estimate whether or not the turn towards BITOD was initiated before reaching IGARI, as it would in LNAV mode. Using the KB SSR data, I think you are agreeing that the turn began before IGARI, and the waypoint was not overflown.

  358. Richard Godfrey says:

    @Paul

    It does not look like the turn to BITOD was completed.

    From the ADS-B data I shared earlier, MH370 turned to a track of 40 degT just after IGARI.

    Half a minute later, the last ADS-B data still shows a track of 40 degT.

  359. Paul Smithson says:

    @Richard, IIRC Sladen reckoned that the last reliable ADS-B was 17:20:35 showing speed 471 and track 040. That data and the anomalies that followed have been discussed to death elsewhere.

  360. Victor Iannello says:

    @Jean-Luc Marchand: Is your group involved with the re-opened French investigation?

  361. flatpack says:

    So, it seems to me that a vanilla chandelle is effectively excluded. Does anyone still advocate one?

  362. Victor Iannello says:

    @flatpack: We have been discussing the right turn at IGARI towards BITOD. The details of the subsquent left hand turn remain unknown.

  363. Richard Godfrey says:

    @Paul Smithson

    The ADS-B data I used was from Paul Sladen.

  364. Paul Smithson says:

    “@flatpack: We have been discussing the right turn at IGARI towards BITOD. The details of the subsquent left hand turn remain unknown.”

    Except that you need to get to Exit Waypoint by 17:24:38 on track ~237 and fly from there at a mean groundspeed of 454kts (with a slight tailwind) to 1731. Turns out that there is a perfectly simple way of doing that, starting from the IGARI-BITOD track, as I am getting tired of pointing out. Hint: no chandelle involved.

  365. Victor Iannello says:

    @Paul Smithson: Yes, what you are proposing is a possibility. You are also estimating that there was a deceleration. A rapid climb would be consistent with that. I freely admit that I don’t know what happened at that turn. Don’t take that as being ignored.

  366. Richard Godfrey says:

    @DennisW

    Please look after yourself! If the fire is that close, are you not in a mandatory evacuation area?

  367. flatpack says:

    Correct me if I am wrong; my interpretation is that there was one continuous manoeuvre, that there was not enough time to a right turn, return to level flight for a short while, then do a longer left hand turn (with or without climbing). Thus the pilot rolled straight through.

  368. sk999 says:

    Super Secret radar data converted to long, lat, using the Vincenty Direct formula (and a position for the KB radar of 102.2940E, 6.1640N)

    17:17:01 103.385 6.515
    17:17:05 103.389 6.522
    17:17:08 103.392 6.530
    17:17:12 103.396 6.537
    17:17:16 103.400 6.545
    17:17:20 103.403 6.552
    17:17:24 103.407 6.560
    17:17:27 103.410 6.568
    17:17:31 103.414 6.575
    17:17:35 103.417 6.583
    17:17:39 103.421 6.590
    17:17:43 103.425 6.597
    17:17:47 103.428 6.605
    17:17:50 103.432 6.613
    17:17:54 103.436 6.620
    17:17:58 103.439 6.628
    17:18:02 103.443 6.635
    17:18:06 103.446 6.643
    17:18:09 103.450 6.650
    17:18:13 103.454 6.659
    17:18:17 103.457 6.666
    17:18:21 103.461 6.673
    17:18:25 103.465 6.680
    17:18:28 103.468 6.689
    17:18:32 103.472 6.696
    17:18:36 103.475 6.703
    17:18:40 103.479 6.711
    17:18:44 103.482 6.718
    17:18:48 103.486 6.726
    17:18:51 103.490 6.733
    17:18:55 103.493 6.741
    17:18:59 103.497 6.748
    17:19:03 103.501 6.756
    17:19:07 103.504 6.764
    17:19:10 103.508 6.771
    17:19:14 103.511 6.779
    17:19:18 103.515 6.787
    17:19:22 103.519 6.794
    17:19:26 103.522 6.802
    17:19:29 103.526 6.810
    17:19:33 103.529 6.817
    17:19:37 103.533 6.825
    17:19:41 103.537 6.832
    17:19:45 103.540 6.840
    17:19:48 103.544 6.847
    17:19:52 103.547 6.855
    17:19:56 103.551 6.863
    17:20:00 103.555 6.870
    17:20:04 103.558 6.877
    17:20:08 103.562 6.885
    17:20:11 103.566 6.892
    17:20:15 103.569 6.900
    17:20:19 103.573 6.907
    17:20:23 103.577 6.915
    17:20:27 103.582 6.922
    17:20:30 103.587 6.929

    Mean time offset from Sladen’s FR24 compilation is 0.4 sec, with an oscillatory dispersion of +/- 1.5 sec (approx.). Mean track offset is 1.5 times the wingspan of a Boeing 777-200ER, with an rms dispersion of 1 wingspan.

  369. DennisW says:

    @RG

    Yes relative to evacuation status, but you cannot be forced to leave. Plus that I know just about everyone is the sheriff’s department. Very few people have left the area except those with severe breathing difficulties.

  370. Nederland says:

    @DennisW

    I hope it works out well for you.

  371. Paul Smithson says:

    @sk999. Bless you.

  372. Paul Smithson says:

    @sk999. for what it’s worth, I make your lat/long 1.5 seconds “ahead” in position relative to RG’s positions. Projecting forwards to IGARI, the positions that you have just posted DO get you abeam IGARI at 17:20:31 after all. kmz is below. hover over the dots to see respective timestamps.
    https://www.dropbox.com/s/dmuyg90pi4gkla8/sk999%20kbssr%20track%20from%20RG%20data.kmz?dl=0

  373. Andrew says:

    @sk999’s data in Google Earth, with the flight planned tracks in yellow:

    https://www.dropbox.com/s/hr4vyujeet6vsvv/IGARI%20TURN.jpg?dl=0

  374. TBill says:

    @Andrew
    Thank you I also did it on SkyVector. I am not seeing any evidence of starting left turn before IGARI, are you? Looks like a turn to BITOD. Now then, what I do is set the turn diameter to 25 deg in preparation for the tight left turn, so I don’t know if the turn looks a little sharper.

  375. Andrew says:

    @TBill

    @sk999’s ‘super secret radar data’ shows the aircraft was initially tracking directly towards IGARI, but started turning right before it reached the waypoint, presumably to intercept the outbound track to BITOD.

  376. Dazza says:

    Victor, what is your opinion on the choice of a nominal position of 0,64.5E used in models (eg Dr B’s, Richard Godfrey’s) in the calculation of the compensation Doppler in the AES?
    I only ask this because:
    The satellite(3-F1) is/was drifting east (and developing increasing inclination) as it reached its end of life but was situated at 63.9-64E when introduced.
    If an astronomer was to request the position of the satellite on 7/3/2014 he would be told, “Nominally at 0,64.5E but with an inclination of (approximately) 2 degrees”.

    If the SDU was “asked” the position of the satellite it would refer to a table embedded in the AES (used also by the beam steering unit for the HGA) and could possibly reply “0,64E” if the table had not been updated.

    One wonders that even if software revisions were promulgated for the AES there would have been a necessity for the “table” to be altered, for in the description of compensation to the transmit frequency in The Search for MH370 by Ashton et al:

    “The aircraft terminal adjusts its transmit frequency to compensate for the Doppler induced on the uplink signals by the aircraft velocity. Aircraft latitude, longitude,track and ground speed are used to calculate the Doppler shift the signal would
    experience if the satellite was at its nominal location over the equator. This only partially compensates for the Doppler associated with the aircraft’s velocity as it does not allow for vertical movement (which introduces discrepancies when the aircraft is climbing/descending) and the satellite is rarely at its nominal location: these small errors are immaterial to the communications performance but do affect the BFO.
    This is δfcomp in Equation” …

    The penultimate sentence seems to suggest that communications performance is not too worried about (small) variations in the satellite position. (…and doesn’t need any alteration to the “table”-my interpretation)

    There appears to be only one set of figures from this flight that provides a solid test at level flight: that at 17:06:43 where ACARS provides latitude, longitude, track /course, and groundspeed(468Kts) can be calculated from a 5 minute earlier earlier ACARS. Inmarsat, Dr B and Richard Godfrey come up with a BFO error of -2, -1.3, -2.7 Hz respectively when compared to the measured BTO of 132Hz. These are very good results and well within generally accepted tolerances. However, using 0,64 (and not 0,64.5) we get the following BFO errors: N/A,1.8,0.6Hz. These latter results are just as tolerable.

    I have addressed this question to you Victor as convenor, and hope that others will not be offended. If this matter was dealt with and resolved in earlier posts I am probably wasting your time.

    I would expect that if 0,64E was more appropriate it would move many calculated final flight paths further up the seventh arc and/or change the residual RMS BTO’s for LNAV navigation amongst other things.

  377. Don Thompson says:

    @Dazza,

    Interesting thought concerning I3-F1’s longitude.

    The GES maintains a System Table in which each Satellite’s longitude is listed. As an AES prepares to execute its a Log On, it syncs to the system’s station management channel and, if necessary, updates its local copy of the System Table over the air.

    Independently of data releases by parties involved in the investigation or providing analysis in support of the search, I received a log of data broadcast on the station management channel, including the System Table. This log showed that the Longitude for I3-F1 was set as 64.5°. At the time this log was captured, Dec 2015, and since the loss of 9M-MRO, Inmarsat had updated the System Table with non-zero data for the inclination and right ascension parameters but longitude had remained unchanged at 64.5°. The chronology of the changes were later confirmed.

    As an aside, the system table entry for JCAB’s MT-SAT region is notable: the MT-SAT region was designed to be served by two satellites (and two GES) in a redundant configuration. The two satellites, Himawari-6 and -7, were spaced approx 5° of longitude apart but the System Table defined a ‘virtual’ position between the two. It certainly worked.

  378. Victor Iannello says:

    @Dazza: Thanks for your comment. As Don explains, the nominal longitude as broadcast in the System Table was run down years ago. As an aside, the statement “Aircraft latitude, longitude,track and ground speed are used to calculate the Doppler shift the signal would experience if the satellite was at its nominal location over the equator” is in general false. It’s true for the AES in MH370 only because the inclination for I3F1 as broadcast in the System Table was zero at the time of the disappearance. If a more accurate value of inclination and right ascension (which is subject to the resolution, i.e., number of bits, of the stored value) was broadcast, the BFO would be an even poorer discriminator of the groundspeed and track. We’re fortunate that the inclination in the System Table was set to zero.

  379. Gysbreght says:

    Do I understand correctly that the last airplane position recorded by the Kota Bharu radar was at 17:20:15 UTC ?

  380. Victor Iannello says:

    @Andrew: What is disturbing is that near IGARI, the SSR data from Kota Bharu doesn’t agree with the path depicted in the official reports. After four months, we are still debating the facts surrounding the military radar data. Malaysia needs to release the damn data. Nothing in that data set can be more embarrassing than the measured altitudes they already released.

  381. Victor Iannello says:

    @Gysbreght: We have the (X,Y) position (in NM) from Kota Bharu at 17:20:31 as (77.16,45.75).

  382. Gysbreght says:

    @Victor Iannello:

    Thank you for your reply. I apologize for my question, I suppose I have been misled by Richard Godfrey’s file, which says:
    17:20:30 X=77.16 Y=45.75 Track=37.97 Interpolated at GS=471 kt, Track=34.87, for 3 seconds.

  383. Gysbreght says:

    @Victor Iannello: Also the lack of response to my posts of August 7 at 10:48 am and 12:29 seemed to confirm that my understanding was correct. Apologies again for my misunderstanding.

  384. Richard Godfrey says:

    @Victor

    You stated: “We have the (X,Y) position (in NM) from Kota Bharu at 17:20:31 as (77.16,45.75).”

    This should read “17:20:30 as (77.16,45.75)” according to the source data.

    For some reason @Gysbrecht feels grievously misled by me???!!!!

    I do not know where he gets the fact that the last data was at 17:20:15 from.

  385. Victor Iannello says:

    @Richard Godfrey: I don’t know, but I assure you that there is no value in arguing with him.

  386. Gysbreght says:

    @Richard Godfrey: The comments at 17:20:19; 17:20:23; 17:20:27; 17:20:30 all read: “Interpolated at GS = 471 knots, Track = xxx°, for yy seconds
    “. So how could I possibly know which aircraft positions are interpolated/extrapolated and which are as recorded in Kota Bharu? Why did you not correct my apparently erroneous understanding in my posts of August 7 at 10:48 am and 12:29?

  387. Richard Godfrey says:

    @Gysbreght

    Each comment refers to the Latitude and Longitude calculation in the same line as the comment.

    The comment does not refer to any other line, only the line where it is found.

    This is obvious to a school child.

  388. Gysbreght says:

    @Richard Godfrey: I’m referring specifically to the comment on line 62 (17:20:30 UTC).

    Please answer my second question.

  389. Victor Iannello says:

    @Gysbreght: Knock it off. People here are not responsible for correcting all your mistakes.

  390. Gysbreght says:

    @Victor Iannelo: Thank you for allowing me to express my views.

  391. Richard Godfrey says:

    @Gysbreght

    You stated “Why did you not correct my apparently erroneous understanding in my posts of August 7 at 10:48 am and 12:29?”.

    1. I took a cursory look at your linked .pdf in the comment at 7th August 2018 10:48 am, but your point was not clearly stated, in fact not stated at all. I decided not to waste my time, trying to work out what you wanted to say.

    2. Your question on quantisation errors at 7th August 2018 at 12:29 am was not addressed to me. It seemed that it was addressed to Mike Exner or perhaps Victor, but again you did not make that clear.

    3. I am not responsible for checking the accuracy of meanwhile over 393 comments on this post alone, this includes your 50 comments. It may not have occurred to you, but I also have other things to do in life.

    4. The radar source file was not in a standard Excel format. It appears that when you import an Excel in this particular format into Numbers it truncates the msecs from a data time value, the actual time at 17:20:30 UTC is 17:20:30.850, which Victor rounded up to 17:20:31 understandably and my spreadsheet in Numbers truncated to 17:20:30. This was not an attempt to mislead you or anyone else intentionally. This was purely a software error between Apple and Microsoft for the particular format used in this case.

  392. Gysbreght says:

    @Richard Geoffrey: Thank you for taking some of your precious time to address this mere mortal.

    You have still not explained why line 62 that gives the X,Y values of a position (Line 62: 17:20:30 X=77.16 Y=45.75 Track=37.97) that you attribute to the Kota Bharu radar has that comment:

    Interpolated at GS=471 kt, Track=34.87, for 3 seconds.

  393. Gysbreght says:

    @Richard Geoffrey: To clarify my last comment:

    Either that position was derived from the preceding line as defined in the comment, or it comes straight from the Kota Bharu log, and the comment is not applicable.

  394. Gysbreght says:

    @Richard Godfrey: Please accept my sincere apologies for mis-spelling your name twice.

  395. Gysbreght says:

    @Victor Iannello: You have repeatedly said “if MH370 was not in LNAV at IGARI, that would be a really big deal”.

    You should not simply embrace Richard G’s turnback analysis as “problem solved” and brush off any critical questions that may arise.

  396. Victor Iannello says:

    @Gysbreght: The track angles calculated from the X,Y data from Kota Bharu show that the turn began before IGARI with high confidence (9-sigma). (I have access to the same source file as Richard.) The evidence points to a path defined by Direct to IGARI to BITOD in LNAV mode, with the turn anticipated before IGARI. (If there was a fly-over at IGARI, that would be a really big deal.) If somebody produces evidence to the contrary, we’ll all consider it.

  397. Gysbreght says:

    @Victor Iannello: If one ignores dubious FR24 data, the X,Y data from Kota Bharu presented in Richard G’s file as I legitimately interpret them, show a course change of about 5 degrees to the right before IGARI, but nothing like that shown in Richard’s file and sk999’s and Andrew’s depiction of it.

    Time for bed for me now, see you tomorrow.

  398. flatpack says:

    @Victor Iannello

    What do you make of the following segment of paragraph six on page four of the report?

    “At 1801:59 UTC [0201:59 MYT] the data showed the “blip” on a heading of 022°, speed of 492 kt and altitude at 4,800 ft. This is supported by the “blip” detected by Military radar in the area of Pulau Perak at altitude 4,800 ft at 1801:59 UTC [0201:59 MYT].”

    So is the ‘blip’ supporting itself or is there another Military radar out there? The second instance of the ‘blip’ does not seem to have a heading or speed associated with it.

  399. sk999 says:

    R.E. the nominal longitude of the 3-F1, the duration of flight MH371 is long enough that one can actually treat the nominal longitude as a free parameter and solve for it. Doing so, I get 64.54 deg with a 1-sigma error of 0.03 (using all BFO data from 1:00 to 6:54). So yes, it is likely that the SDU is using 64.5 and we can rule out that it is using 64.0 with very high confidence. Secret info from Don Thompson not required.

    R.E. the ADS-B FR24 data, there are 2 data points after MH370 passed IGARI, at 17:20:35 and 17:21:03. The latitude and longitude are not consistent with a plane flying at 472 knots, instead imply a lower speed. We do not know in detail what FR24 does when the ADS-B link from an aircraft is lost, but the position of the 17:20:35 point is consistent with a timestamp of 17:20:32 – just a second after passing IGARI. So the conjecture is that FR24 propagates the last received ADS-B info to the next schedule timestamp. Thus, the the track angle of 40 deg was actually recorded 1 second after passing IGARI. This track angle is entirely consistent with that from the KB SSR. The track angle at the final time stamp (17:21:03) simply repeats the last recorded value and is otherwise bogus. Make sense?

  400. Victor Iannello says:

    @flatpack: I don’t know how to interpret that blurb. It sounds as though there are two radar sources that measured an altitude of 4800 ft near Pulau Perak. The “heading” (probably meant to say “track”) of 22° is also odd. Perhaps they meant to say 292° and dropped a digit?

    Perhaps others can offer their ideas.

  401. Victor Iannello says:

    @sk999 said: Secret info from Don Thompson not required.

    The System Table is broadcast in the open without encryption (as is all the traffic on the P-channel). In this case, it was received and recorded by an amateur with a known identity and shared. The data is not “secret info”.

  402. Victor Iannello says:

    @Gysbreght: Using the KB SSR data, the average track between the last two points at 17:20:27 and 17:20:31 is 38.0° +/- 1.4°, representing a change in track of about 12.5°. We have no reason to doubt the X,Y values for these last two points, as the previous values agree closely with the ADS-B data.

  403. Gysbreght says:

    @Victor Iannello: Maybe you have KB SSR data after 17:20:15, I don’t.

    I have to go by Richard G’s calcsheet which says as school child can see that the X,Y positions at 17:20:19; 17:20:23; 17:20:27; 17:20:30 have been calculated from the 17:20:15 position at GS = 471 knots in steps of progressively increasing track angles which seem to have been interpolated from two values taken from Paul Sladen’s ADS-B file.

  404. Richard Godfrey says:

    @Victor

    I have compared and aligned the radar data from each radar source at Genting RSR, KLIA TAR and Kota Bharu TAR together with the ACARS and ADS-B data from 16:41:43.720 UTC to 17:20:30.850 UTC.

    Here is a link to an overview of the 991 data points:

    https://www.dropbox.com/s/rn4xxq2qwh9pnek/Combined%20Position%20Data%20to%20IGARI.pdf?dl=0

    Here is a link to a plot of the last 10 radar points giving location and time (now extracted to the nearest msec):

    https://www.dropbox.com/s/6cmhbznhndhmogq/Combined%20Position%20Data%20to%20IGARI%20Detail.pdf?dl=0

    It would appear that a turn started around 10 seconds before reaching waypoint IGARI.

  405. Victor Iannello says:

    @Gysbreght: How many times do we need to tell you that the data points up to and including 17:20:31 were SSR targets from Kota Bharu? You are once again arguing for the sake of arguing.

  406. Gysbreght says:

    @Victor Iannello; @Richard Godfrey:

    Please explain to a school child for the data points after 17:20:15 what the comments mean in Column J of RG’s spreadsheet.

  407. TBill says:

    @flatpack
    @Victor
    I am trying to find the Palua Perak eyewitness account discussion if any. If the plane were flying so low, we should have been told that a long time ago.

  408. Victor Iannello says:

    @Gysbreght: Those comments were added by Richard to help understand the data. They were NOT included in the source file. If those comments are causing you such confusion, just ignore them. I will state this one last time for you: The data points up to and including 17:20:31 were SSR targets from Kota Bharu. (Note that the final, untruncated time is 17:20:30.850 to the nearest millisecond.)

  409. Victor Iannello says:

    @TBill: There was a claim in the RMP report of an eyewitness on Pulau Perak seeing MH370.

  410. Gysbreght says:

    @Victor Iannello:

    You have not explained what those comments are intended to convey, but I’ll let it rest.

    Thank you for clearing up the confusion caused by those comments.

    It is sowhat strange that those points describe a perfect circle while earlier points tend to scatter a bit around a straight line.

  411. Richard Godfrey says:

    @Victor

    Please find below a link to the data that I used in my previous comment on 9th August 2018 at 06:18 am:

    https://www.dropbox.com/s/j1ci9nyp57oelxj/Combined%20Position%20Data%20to%20IGARI%20Detail%20Data.png?dl=0

    Liability Disclaimer

    All data is supplied without guarantee. The user must satisfy himself or herself that the data is suitable for his or her enquiry. The data has been produced and processed from sources believed to be reliable, no warranty expressed or implied is made regarding accuracy, adequacy, completeness, legality, reliability or usefulness of any information. The provider shall not be held for the accuracy of the data provided. The provider is not liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to: procurement of substitute goods or services; loss of use, data, or profits; or business interruption) however caused and on any theory of liability, whether in contract, strict liability, tort (including negligence or otherwise), or any other theory arising in any way out of the use of this data, even if advised of the possibility of such damage. This disclaimer of liability applies to any damages or injury, whether based on alleged breach of contract, tortious behaviour, negligence or any other cause of action, including but not limited to damages or injuries caused by any failure of performance, error, omission, interruption, deletion, defect, delay in operation or transmission, computer virus, communication line failure, and/or theft, destruction or unauthorised access to, alteration of, or use of any record.

  412. Gysbreght says:

    Thank you, Richard

  413. DennisW says:

    @Richard

    There is no getting around the fact that you could be considered “actively negligent” for even posting the information.

  414. @Victor: Is your group involved with the re-opened French investigation?

    No, it is the result of an initiative of the French Department of Justice. You might know that Ghislain Wattrelos (husband and father of victims) sue the State of France. This is the reason also why the flaperon is kept within the same Department of Justice as an evidence for this sue case.

    (Sorry for the late reply, I was not connected … on holidays 🙂

  415. flatpack says:

    @Andrew
    @Don Thompson

    Re the IFE/PASS SEATS switch. Have you gleaned any further insight as to what ‘all IFE components’ actually means? Thanks.

  416. Victor Iannello says:

    @All: I’ve heard that some are having problems accessing the site. I’ll look into it.

  417. Jay says:

    Knowing what we know about, with the possibility of the crash substantially more north, what are the chances that this eyewitness sighting was actually relevant?

    According to the testimony of 6 Swiss Citizens making a cruise between Perth and Singapore via Jakarta, the following evidences were spotted on March 12 while approaching the Sunda Strait:

    1430LT – latitude 6°, longitude 105°, speed 17,7 knots:
    life jacket, food trays, papers, pieces of polystyrene, wallets,
    1500LT: a huge white piece of 6 meters long to 2,5 meters wide with other debris,
    1530LT: two masts one meter long with small flags on top, red and blue,
    2030LT – latitude 5°, longitude 107°, speed 20,2 knots.

    This testimony has been submitted to the Chinese Authorities by these 6 Swiss Citizens without any feedback until today. However, they have been in touch with the Australian Authorities. – B3A (Bureau of Aircraft Accidents Archives)

  418. mash says:

    re: The Art of War

    What if the initial erratic flight path patterns during/after turnback (and before final south turn) were all deceptional moves? They really served no other practical purposes – too excessive for depressurisation, and too in-effective for radar evasion.

    Then it is more likely that a ‘third party’ is trying to pretend that it is the ‘first party’; rather than the other way round – naturally.

    More than that, it is more likely that that ‘third party’ is ‘something’ rather than ‘someone’ (i.e. the meaning of pretending – to be something [some category] else).

    In conclusion, the deceptional patterns indicate an ‘extraordinary’ event unfolding; the secondary mission purpose is to hide the primary mission purpose …

    [erratic patterns: left/right, high/low, fast/slow (at least as ‘recorded’).]

  419. TBill says:

    @mash
    Maybe the pilot knew he had to stay within MY military radar range. By doing that, the aircraft radar blip was continuously coded as MH370-Friendly-no-alarms.

  420. Don Thompson says:

    Concerning IFE and passenger seat power.

    Prior to the Service Bulletin that added the IFE/Pass Seats & Cabin/Util switches to the P5 panel, an IFE Power switch existed at the Purser Station in the cabin. This IFE Power switch removed power from the Seat Electronics Boxes & Area Distribution Boxes.

    After implementing the SB for the additional switches on the flight deck P5 panel, the IFE/Pass Seats switch replicates the action of the Purser Station switch AND removes power from all video displays (eg the cabin bulkhead mounted units), seats mechs, seat AC power outlets, and the IFE units in the Purser Station itself.

    Power is not removed from IFE units in the MEC when either the Purser Station or P5 switches are released.

  421. flatpack says:

    @mash

    Deception? All of it!

    My take; the ‘first party’ tried to make us think he was struggling to control a ‘wounded bird’ or that ‘third party’ intervention had occurred.

    For example; the “heading changes” on the military radar track detailed on page 4 of the report (also noted by Victor as “Possible small turns” on the civilian radar track).

    These ‘oscillations’ could be interpreted as a pilot trying to control the plane primarily using engine thrust variation as, I think, a DHL pilot managed to do after being hit by an anti-aircraft missile.

    They could also be interpreted as an inexperienced ‘bad actor’ going up a steep learning curve.

  422. flatpack says:

    @Don Thompson

    Thanks for the information.

    As an aside, does the Purser Station include a SATPHONE?

  423. Victor Iannello says:

    @All: David Griffin alerted me that he has updated his thoughts about MH370 in consideration of the failed OI search. He concludes:

    We do not think that our drift modelling supports extending the search either northwards or southwards. We can not completely rule out the possibility that the aircraft is located either north or south, but the available information suggests to us that the intermediate latitudes are much more likely. We leave it to relevant domain experts to assess whether the search should extend either farther from the 7th arc, possibly at just a selection of latitudes, or focus on the so-called ‘holidays’ (gaps in the sonar coverage where terrain is difficult). We hope to revisit our conclusions at some future point but in the meantime express our sincere regrets that the committed efforts of many dedicated people have not yet solved this enduring mystery.

  424. Gysbreght says:

    Based on Richard Godfrey’s August 9, 2018 at 11:04 am, and Andrew’s estimate of roll rate:

    https://www.dropbox.com/s/00oqqwvkf7sjtem/Roll_Bank_Track.pdf?dl=0

  425. Gysbreght says:

    @Richard Godfrey:

    How many studies and analyses did you post on this blog? Did you ever add a Liability Disclaimer? So why this time?

  426. Victor Iannello says:

    @Gysbreght: Have you no shame? Move on.

  427. Gysbreght says:

    [You might be obsessed by the comments that Richard G added to the spreadsheet, but I’ve asked you to move on.]

  428. Richard Godfrey says:

    @Victor

    I have analysed the MH370 turn back from the radar data we have received, tracking both the outbound and inbound flight. The raw data for the outbound flight is from the Kota Bharu Terminal Approach SSR Radar with a nominal range of 90 NM and the raw data for the inbound flight is from the Kota Bharu Terminal Approach Primary Radar with a nominal range of 60 NM. These radar systems are co-mounted at 6.819756°N 103.009384°E.

    The results are shown in the linked graphic:

    https://www.dropbox.com/s/c8ve5q9x64vaj8s/Combined%20Position%20Data%20Turnback.pdf?dl=0

    The track from the Malaysian Report via their entry point and exit point for the turn back is, supposedly based on their military radar, with a 172° turn at a bank angle of 31°, which appears to arrive 29.661 secs too early for the inbound data capture at Kota Bharu at 17:30:33:310 UTC, based on maintaining the air speed of 478 knots.

    On the other hand, the track with a 180° turn at a bank angle of 20° appears to arrive 52.306 secs too late based on maintaining the air speed.

    After the turn, the wind will have the effect of increasing the ground speed from 471 knots outbound to between 485 knots and 486 knots inbound, depending on the bank angle, if the air speed of 478 knots is maintained. In this case, a turn at a bank angle of around 27° fits the timing of the inbound data capture at Kota Bharu at around 17:30:33:310 UTC.

    We are fortunate to have the raw data from the civilian radar systems. However, the turn back was executed beyond the range of the civilian radar systems. Without the Malaysian Authorities disclosing the raw data from the military radar, we are not able to conclude the precise nature of the turn back.

    It has not escaped my notice, once again, that the turn back was executed in the HCM FIR without announcing a handover via VHF to HCM FIR as instructed, without announcing a turn back or distress via VHF to KUL FIR, whilst ensuring there was no incursion into BKK FIR, whilst ensuring at no time MH370 was beyond the range of the Malaysian military radar and therefore being registered and maintained as a friend and not a foe. This is a well planned turn back by someone with familiar knowledge of all regional civilian and military radar systems.

    Why the Malaysian Authorities refuse to disclose the military radar data is beyond comprehension. Why the Indonesian Authorities refuse to even comment on the Malaysian Report is also beyond comprehension. What are both hiding?

  429. DennisW says:

    @Richard

    What are both hiding?

    As ICAO noted in their initial comments on the flight, the early response seems to have been improperly directed by people at high levels in the Malaysian government, and not the usual people normally associated with that responsibility. That was a very atypical response for ICAO. My sense is that the nature and reason for the diversion was known at high levels. Yes, I believe things have been buried, and will probably never see the light of day.

  430. Richard Godfrey says:

    @DennisW

    Not even with the new government?

  431. DennisW says:

    @Richard

    Dissappointing response so far. Overall impression is the desire to let the incident fade.

  432. Richard Godfrey says:

    @DennisW

    More importantly, is the wild fire threatening you and your home contained?

  433. Gysbreght says:

    The chart below shows a turn left at 26 degrees of bank, roll to the left at 3 deg/s starting immediately when the airplane after turning right is on track to BITOD. Turn diameter in still air is 14 NM. Wind has not been taken into account. After rolling back to wings level the airplane is on the reciprocal track at 17:23:47.

    https://www.dropbox.com/s/op12woahxf2ry8g/Roll_Bank_Track-2.pdf?dl=0

  434. DennisW says:

    @RG

    Not yet, but the situation is improving. High winds forecast yesterday did not occur. I am monitoring fire progress closely using IR satellite data (MODIS and VIIRS). The satellite data is more timely than Cal Fire updates.

    Thanks for asking.

  435. paul smithson says:

    I am relieved to see that others have woken up to the FACT that there was a significant slowdown from turnback to 1731 and, in all probability, a simple 180 turn on to reciprocal heading from the IGARI BITOD leg. There are multiple methods with independent constraint assumptions that lead us to this solution and these can be triangulated to estimate time, position, mean GS and AOB. Now that we have precise positions for both 172031 and 173033 I conclude that the plane:
    1. Completed it’s turn on to the IGARI BITOD leg as scheduled flying at groundspeed 471kts.
    2. At 172158+/-2 seconds the plane made a 180 turn at 25 Degree bank angle, mean GS of 444, ending it’s turn nearly abeam but a little outside Exit Waypoint.
    3. From there the plane continued at the same TAS, now achieving steady GS of 456kts inclusive tailwind. This speed was maintained until at least 173130.
    4. From 173150 there is clear evidence of the start of the steady speed up to the very high speeds thereafter.
    5. There are no viable solutions for 180 turnback with AOB closer to 30 except with multiple speed changes up, down, up, down.
    6. The above ignores for simplicity the time taken to roll to 25 AOB and to reduce speed from 471GS to 444GS. This does not materially affect the solution.

    I’ll be glad to hear from others if they have alternative turnback models.

    Now, I ask again, particularly of the PIjacker enthusiasts, why slow down to TAS444 for 9+ minutes from 1722 to 1731. I think it’s beyond reasonable doubt that this occurred. I think the leading explanation may be INITIAL intent to land at KB, resulting in slowing to econ descent speed IAS 271-273 but without commencing the descent.

    Finally, I note that the altitude at which TAS444 matches econ descent speed is 31000feet (equivalent FL320 that day near KB), with ISA+10.8. At this altitude M0.87 easily achieves the very high ground speeds seen later.

  436. Gysbreght says:

    @Paul Smithson: If there is one thing certain in this mystery, ECONomy was of no concern to whoever did it.

  437. Richard Godfrey says:

    @Paul Smithson

    So the entry and exit points defined in the Malaysian Report, supposedly from military radar, are incorrect in your view?

  438. flatpack says:

    @paul smithson

    Or, any such slowdown was made to lead later analysis work to the conclusion that there was initial intent to land at KB.

  439. paul smithson says:

    @th I think it’s demonstrable that exit is pretty good but entry cannot be. By my calculation you exit turn with exit wpt abeam +/-2s but the turn diameter takes you a shade wider than exit wpt.

  440. Andrew says:

    Geoffrey Thomas in this morning’s The West Australian (also on airline ratings.com):

    Malaysia’s MH370 report a ‘cover-up’

  441. Andrew says:

    @flatpack

    RE: ”As an aside, does the Purser Station include a SATPHONE?”

    It seems to be an airline option; some airlines do, some don’t. I don’t know how the MAS aircraft were configured.

  442. Victor Iannello says:

    @Paul Smithson: I think the decrease in airspeed at the turn and the PSR data from Kota Bharu are best explained by a steep climb, not an intention to descend. For the required rapid descent to land at Kota Bharu, the MCP altitude would be set to the lower altitude, FLCH would be selected, the MCP speed would be increased to Vmo, and the speedbrakes deployed. I don’t see evidence for that.

  443. Victor Iannello says:

    @Andrew: Thanks for the link to the Geoffrey Thomas article. I am glad that some are questioning the conclusions in the Safety Report instead of just regurgitating the information.

  444. Andrew says:

    @Victor
    @paul smithson

    A TAS reduction of 27 kts only needs a CAS reduction of 17 kts at FL350/ISA+10. That could be achieved relatively quickly in level flight with the thrust at idle.

  445. Victor Iannello says:

    @Andrew: Why would a pilot with an intention to land reduce speed without descending? With A/P engaged, you’d increase the MCP speed and perform a FLCH descent at Vmo with speedbrakes. If A/P is not engaged, you’d go to idle thrust and trim for speed. I don’t see how flying level and reducing speed indicates an intention to land. What am I missing?

  446. Andrew says:

    @Victor

    I didn’t say the pilot had an intention to land! The speed reduction might simply have been an attempt to minimise the turn radius (however small that reduction), followed by a speed increase to get out of the area as quickly as possible.

  447. Victor Iannello says:

    @Andrew: Got it! I agree that the reduction in speed could have been to reduce the turn radius. But also, our analysis of the PSR data at Kota Bharu seems to indicate that MH370 was higher than FL400. Although idle thrust would have reduced speed before or during the turn, the reduction in speed at the turn and the high altitude near Kota Bharu are both consistent with a climb during the turn around.

  448. DennisW says:

    @Victor

    Not all of us agree that the flight level at KB was higher than 40,000′. I think everyone’s analytics are flawed (but mine, of course).

  449. paul smithson says:

    @victor.
    A) If slow down from climbing you are suggesting
    he was climbing from 1722 to 1732?
    B) just happened to produce constant ground speed
    C) acceleration to m0.87 would have got you away a lot faster than slowing 30 knots, whether by climb or otherwise.

  450. paul smithson says:

    Start of the turn, where I estimate speed reduction began, was 121NM out of KB along the track.

  451. Victor Iannello says:

    @Paul Smithson:
    A) I doubt it. But at high altitude, the acceleration will be slow.
    B) How do we know the ground speed was constant?
    C) Maximum speed must not have been the sole objective.

  452. Richard Godfrey says:

    @Victor

    (Please excuse me butting in Paul.)

    Victor you asked “How do we know the ground speed was constant?”

    We don’t!

    But we do know the last recorded ground speed was 471 knots. A turn back starting where the Malaysian Report states at a bank angle of 27 degrees and accounting for the wind increasing the ground speed to 486 knots, then MH370 arrives at the first inbound radar data exactly on time.

    Please see my previous post.

    This information does not tell us exactly what happened, but provides an average measure without having to consider changes in altitude or air speed or ignore where the Malaysian Report says the turn started.

    Against this ‘yard stick’ average measure, we can now evaluate alternative turn back scenarios, but there is only a little wiggle room to allow alternative scenarios, including climbs, chandellles, slow downs, full throttle panics, etc.

  453. Paul Smithson says:

    @Victor – how do we know groundspeed was constant?

    Mean ground speed from 172438 (exit waypoint) to 173033 I made 455.5kt
    Mean ground speed 173033 to 173157 = 457.2
    Mean ground speed from the first 17 “10-step” boxcars from 173033 I made 456.6 and no discernible trend. That’s an empirical confirmation of steady GS if the exit waypoint position and time are accurate.

    If Exit waypoint was 1NM out laterally it would make no material difference to the mean ground speed 172438 to 173033. If it was 0.5NM out fore/aft it makes only 5kts difference to resultant ground speed. My independent methods for prediction of exit waypoint put it abeam to within 1s, which further confirms my believe that it is correct.

    My independent methods for estimating speed over the entire track from 172031 to 173033 predict mean groundspeed 452-456.

  454. Paul Smithson says:

    My method for predicting track solution from 172031 to 173033 goes as follows.

    1) Simplify the track to a “trombone” shape. The trombone can be slid in and out (affecting the max NE extent of turn). The radius of the turn can also be allowed to vary freely.
    2) From the track back to 173033, take a perpendicular to the position opposite 172036s (where plane should have completed turn on IGARI-BITOD). The “bar across the trombone”.
    3) Now measure from that perpendicular to the NE maximum per military trace (I make it 16.3NM +/-0.3)
    4) Now, we have “legs” as follows:-

    From 172046 to the start of the turn (free-floating, any time after 172046)
    From the start of the turn to end of 180 turn (result dependent on speed and turn diameter. From the end of the turn back to 173033

    Distance from 172046 to start of the turn is 16.3 minus the radius of the turn. That has to be travelled twice; once on the way out, once on the way back. Distance around the turn of 180 degrees is pi*r. So now we can calculate distance traveled for any turn start position and any turn diameter.

    Now allow speed and turn diameter to vary freely (I allowed from 11 to 13.2), constrained only be the requirement to start at 172046 and end 173033 and not exceed the NE turn maximum per military trace. No AOB assumption is made, although we can later “filter” the solutions by AOB feasibility.

    My solutions for mean GS from 172046 to 173033 using this method are as follows:

    Length of trombone from bar to end = 16.3NM. GS 458, +/- 2kts (for turn diameters between 11.6 and 12.8NM)
    Length of trombone 16.8. GS 464 +/-2 for turn diameters 11.6-12.8
    Length of trombone 15.8. GS 452 +/-2, diameters 11.6-12.8

    The solution is remarkably insensitive to assumption on turn diameter. Smaller diameter trades off smaller distance around the turn for a marginally later turn.

    I am encouraged that this exercise:
    a) produces a solution that predicts time and position abeam exit waypoint
    b) correctly predicts speed for turn end to 173033 that we can now confirm empirically from exit waypoint to 173033
    c) that speed is further confirmed by the first 90s of the KB primary radar from 173033 to 173150

  455. Victor Iannello says:

    @Paul Smithson: What do believe was the altitude over Kota Bharu?

  456. Paul Smithson says:

    * “From the track back to 173033, take a perpendicular to the position opposite 172036s” should read 172046 = IGARI +15s (as it does throughout the rest of the post).

  457. Paul Smithson says:

    From the previous, it should also be clear that the reason that faster speed solutions don’t work is that they go way to far beyond the NE extent of the military turn.

    High AOB (>25) turn solutions are possible, but they are all for slower speeds in the range discussed.

    It should also be clear that if you cut the corner towards Entry Waypoint instead of starting the turn from IGARI-BITOD, then then you also need a much tighter turn diameter. The result is a much shorter distance traveled and even slower speed solution!

  458. TBill says:

    @Paul
    I agree with one implication of your proposal, it is hard to imagine anyone other than ZS conducting the overall action at IGARI.

  459. Gysbreght says:

    @Paul Smithson:

    I don’t understand your assumptions.

    We have the official flight path derived from primary and secondary radar data. These radar data have been ‘corroborated’ by the NTSB, accepted and published by the ATSB, and used by the DSTG in the Bayesian analysis. “The Team” selected Entry and Exit waypoints from these data for the turn simulations reported in the Safety Investigation Report:
    “The turns were based on the recorded primary radar data that recorded a primary target conducting a left turn from where the SSR code ceased, shortly after the aircraft passed waypoint IGARI.”

    The official data show that the airplane passed overhead IGARI before it turned right to a heading that would have intercepted the IGARI-BITOD route segment from the North.

    We also have unofficial “super-secret” secondary radar data obtained from a source that canned be named. Those data are claimed to show an entirely different path in which the airplane started turning to the right before IGARI and passed south of that waypoint in the normal fly-by mode that would be expected with the autopilot in LNAV mode.

    Since the two sets of data show different paths, they cannot both be right. If you believe the unofficial data and assume the left turn started on the IGARI-BITOD route segment, you should ignore the official data including the Entry and Exit waypoints that are derived from it.

  460. TBill says:

    @Victor
    Re: Geof Thomas quote about “toy” simulators , does it look like the stolen aircraft in Seattle may have been playing with FSX or FS9? Is that aircraft available as an advanced type realistic model like PMDG B777? Easy to take-off, easy to fly, but don’t ask me to land.

  461. Victor Iannello says:

    @TBill: When I saw how he was flying, the thought crossed my mind, but who knows at this point?

  462. Victor Iannello says:

    @Gysbreght: If we had the raw military data, we might be able to do a proper comparison with the available SSR data. Otherwise we are relying on filtered speed/track data that shows unrealistic speeds, or a graphical depiction of the path that shows an unrealistic sharp turn. All the more reason for sharing the raw military data.

  463. Gysbreght says:

    @Paul Smithson: If you assume that the turnback manoeuvre was performed by the autopilot in LNAV, then the bankangle would be 26 degree as obtained in the simulations.

  464. Gysbreght says:

    @Victor Iannello: The official data are SSR up to the point where the SSR signal is lost. That part is already different from the “super-secret” data.

  465. Victor Iannello says:

    @Gysbreght: What do you believe is the “official” SSR data? We have the ADS-B compiled by Paul Sladen, and the KB SSR data Richard provided. Is there other SSR data?

  466. Gysbreght says:

    @Victor Iannello: You tell me!

  467. Neville says:

    Neville says
    @All

    Can anyone recommend a good glossary of aviation acronyms for the unenlightened? The one’s I have found so far simply don’t cut it.

  468. Victor Iannello says:

    @Gysbreght: I think the Kota Bharu SSR data ends at 17:20:30.850. I don’t think there is any “official” SSR data after that point.

  469. Gysbreght says:

    @Victor Iannello: The ADS-B Lat/Lon data end at 17:20:18 and don’t show a turn. I doubt the authenticity of the course values 28 and 40, the latter in particular.

  470. Gysbreght says:

    @Victor Iannello: SIR – “The Mode S symbol of MH370 dropped off from radar display at 1720:36 UTC [0120:36 MYT]”

  471. Victor Iannello says:

    @Gysbreght: The SSR data points showing a deviation from the track to overfly IGARI are at times 17:20:23, 17:20:27, and 17:20:31. The final ADS-B position we have is at 17:20:18, which is before the SSR data shows the deviation from the track to IGARI. I see no inconsistency between the ADS-B coordinates and the SSR positions. As for the Mode S symbol dropping off the radar display at 17:20:36, that only indicates that the system waited about 5 seconds before dropping the symbol. Considering that the radar head rotates with a period of 3.8 s, that’s not surprising.

  472. Gysbreght says:

    @Victor Iannello: I also see no inconsistency between the ADS-B coordinates and the SSR positions. That is because there are no ADS-B positions after 17:20:18.

  473. TBill says:

    @Victor
    CNN talking about apparently flight sim trained on the Seattle incident. Mary Schiavo said the module for the (Q400) aircraft is $50. No mention of the platform.

  474. Paul Smithson says:

    @Gysbreght. Since you seem to be so fond of the military “route”, why don’t you tell us what mean ground speed it requires from 172031 [your “over IGARI” to 173033?

    If you follow the right angle corner, I make it 79.5NM in 602 seconds for mean GS 475kts. If you don’t follow the right angle but cut the corner at “Entry waypoint” I make it 77.6NM in 602s for mean GS 464.

  475. Victor Iannello says:

    @Gysbreght: You are making no sense, at least none that I can discern. The ADS-B, the MLAT, and the KB SSR data are all consistent at times when multiple sources provide data. The KB SSR data shows an anticipated turn towards BITOD before reaching IGARI, consistent with an aircraft in LNAV mode. There is no other official SSR data that I am aware of.

  476. Paul Smithson says:

    @Victor. I do not know what the altitude was over KB. I have an opinion about it but it is not germane to the discussion above (which is entirely agnostic about altitude).

  477. Victor Iannello says:

    @TBill: I haven’t followed this closely. I see very little resemblance to MH370, other than an airline employee went renegade.

  478. Victor Iannello says:

    @Paul Smithson: I think it is very germane to this discussion if the drop in speed is related to a climb.

  479. Gysbreght says:

    @Victor Iannello: “@Gysbreght: You are making no sense, at least none that I can discern.”

    [Obnoxious comment deleted].

    Which data support your alleged KB SSR data points at 17:20:23, 17:20:27, and 17:20:31? Which MLAT data are you referring to? Why is “The Team”, the NTSB and the ATSB not aware of the data you present?

  480. Gysbreght says:

    @Victor IannellO: Quoting again from the Safety Investigation Report:

    “The disappearance of the radar position symbol of MH370 was captured by the KL ACC radar at 1721:13 UTC [0121:13 MYT]. The Malaysian military radar and radar sources from two other countries, namely Viet Nam and Thailand, also captured the disappearance of the radar position symbol of MH370. The Bangkok radar target drop occurred at 1721:13 UTC [0121:13 MYT] and Viet Nam’s at 1720:59 UTC [0120:59 MYT].”

    The KB radar was not the only facility that received data from MH370 in that fateful night. Did any of the facilities mentioned above receive position data after 17:20:15 that support the data from your source?

  481. Victor Iannello says:

    @Gysbreght said: Which data support your alleged KB SSR data points at 17:20:23, 17:20:27, and 17:20:31?

    As I have said before, the ADS-B data ends at 17:20:18. Other than the three KB SSR positions at 17:20:23, 17:20:27, and 17:20:31, I am not aware of any other SSR data.

    Which MLAT data are you referring to?

    The lower precision (.01° latitude and longitude) position data available from FR24 is likely either from MLAT or extrapolated by FR24.

    Why is “The Team”, the NTSB and the ATSB not aware of the data you present?

    I have no idea whether this civilian radar data was supplied to the NTSB, the ATSB, or anybody else on the investigative team.

  482. Victor Iannello says:

    @Gysbreght said: Did any of the facilities mentioned above receive position data after 17:20:15 that support the data from your source?

    As I have explained to you repeatedly, I am not aware of any other official SSR data. If you are aware of any, please produce it.

  483. Gysbreght says:

    @Paul Smithson: “Since you seem to be so fond of the military “route”,”

    Please do not twist my words.

    There are two paths. Up to the point where the secondary radar signal is lost, both paths are based on SSR data broadcast by the airplane. If the airplane followed one path, it did not follow the other.

  484. Mick Gilbert says:

    @Victor Iannello
    @Paul Smithson

    Why would a pilot with an intention to land reduce speed without descending?

    Well, who’s to say that there wasn’t a descent? There’s nothing to preclude a descent at a rate of about 2,500 fpm as the aircraft reversed course back towards Kota Bharu. The aircraft’s rate of acceleration toward and terminal speed passing Kota Bharu is much more likely to have been within the aircraft’s capabilities at around FL300 than at FL400+.

  485. Victor Iannello says:

    @Mick Gilbert: I was following Paul’s hypothesis of a constant groundspeed, with the assumption that there was neither a climb or descent.

    On the other hand, the PSR data near Kota Bharu seems to indicate high altitude. And yes, I understand that it appears the weight, altitude, and speed are beyond what Boeing defines as the standard performance envelope. That’s why I am cautious about making a definitive conclusion about the altitude near Kota Bharu.

  486. Paul Smithson says:

    @Gysbreght. I don’t even understand how you can get that confused.

  487. Gysbreght says:

    @Paul Smithson: Thanking you for your subtle hint that my terminology was incorrect. For “broadcast by” please read “of”.

  488. flatpack says:

    @all

    Just a random thought about the reboot(s) of the SDU.

    Victor covered the issue in his March 18th 2018 post.

    Now I have a marginal understanding of the crystal oscillator and it’s temperature controlled oven but not even a marginal understanding of esoteric effects such as ‘retrace’.

    So, my question is; Have cabin depressurisation and associated temperature changes actually been tested?

    It seems to me that there are plenty of reboots recorded on the ground at normal temperature and pressure but that these may not be representative of the actual event.

    A depressurised airframe that has also been low temperature soaked seems significantly different from one sitting on the ramp.

  489. DennisW says:

    @flatpack

    There has been a lot of discussion on SDU reboot behavior, and opinions vary. Little to no physical testing has been done other than collecting data from historical in service reboots. Initially I was annoyed by the lack of physical testing, but I now believe that the results would be very “dispersed” (tests would show a lot of variabilty from unit to unit and would probably not be statistically meaningful).

  490. Richard Godfrey says:

    @Victor

    The MH370 Turn Back, flight path across Malaysia, transit up the Malacca Straits and Final Major Turn (FMT) southwards were all conducted in full view of at least 16 civilian and military radars from 4 different countries:

    1. Sabang
    2. Lhokseumawe
    3. Medan
    4. Sibolga
    5. Dumai
    6. Western Hill
    7. Butterworth
    8. KLIA TAR
    9. Genting RSR
    10. Kota Bharu TAR Primary
    11. Kota Bharu TAR SSR
    12. Phuket
    13. Surat Thani
    14. Hat Yai
    15. Conson Island
    16. Tan Son Nhat
    17. …

    Below is a link to a map of the claimed coverage of the various civilian and military radar systems:

    https://www.dropbox.com/s/ppz9atwe46dh3n9/Regional%20Civilian%20and%20Military%20Radars%20capturing%20MH370.png?dl=0

    Even after the transponder was switched off, the Turn Back would have been captured by at least 2 military radars and the FMT would have been captured by at least 3 military radars.

    Below is a link to a map of the claimed coverage of only the military radar systems:

    https://www.dropbox.com/s/sgiulpkip3w5tm0/Regional%20Military%20Radars%20capturing%20MH370.png?dl=0

    It is simply not believable, that all the military radars in question, from Malaysia, Indonesia and Thailand were switched off, undergoing repairs or maintenance or otherwise unserviceable.

    The Malaysia Safety Report contains no raw radar data. There are a few isolated data points and graphics (mostly unbelievable), but no continuous path data. Even the DSTG graphic, which shows filtered data, does not align with the Malaysian Safety Report, with the ground speed dropping to 191 knots at 17:21:12 UTC and increasing back to 553 knots just before 17:30:00 UTC. The Malaysian Safety Report is even more fantastical showing speeds up to 589 knots and altitudes from 58,200 feet down to 4,800 feet.

    Where did DSTG get the 10 second radar data from? The DSTG data does not align with the 4 seconds radar data made available to the IG and does not align with the radar data points and graphics of the Malaysian Safety Report.

    Are we all being led up the garden path? Many experts have concluded, that following the failed Ocean Infinity search, the remaining options are:

    1. Further North of 25°S.
    2. Further South of 39.6°S.
    3. Between 25°S and 39.6°S, but further away from the 7th Arc.
    4. Missed by Fugro or Ocean Infinity in the searched areas.
    5. Something else, we have not thought of.

    … and that option 5 is the only logical conclusion.

  491. Paul Smithson says:

    @All. The objections to my proposed solution (path/speed profile) between 172031 an 173033 seem to gone rather quiet. Is that because you are convinced it is essentially correct or because you are still looking for reasons to be unconvinced?

  492. DennisW says:

    @Paul

    My assertion three years ago was that the radar data added little to our understanding of a terminus. I pretty much discarded it then, and my interest has not been rekindled.

    Richard’s query as to why we have so little data is more interesting to me than the data we have.

  493. Victor Iannello says:

    @Richard: Very good post. There are many questions that get asked on this blog every day that are known by some. Releasing all the raw military data would be a good start.

  494. Victor Iannello says:

    @All: Here is an image showing the civilian radar data near IGARI. The turn at IGARI seems close to a bank of 25°. The lower resolution FR24 data at 17:20:25 and 17:21:03 (assumed to be derived from MLAT) should be used with caution, as Paul Sladen advised long ago. On the other hand, the ADS-B data and the SSR from Kota Bharu seem to agree, as others have said.

  495. Sabine Lechtenfeld says:

    @Richard G, I wholeheartedly agree with your last comment re: the scarcety of military radar data. My thoughts have gone lately into the same direction. As far as the military radar data are concerned, we may well have been led up the garden path as you have expressed it. Something is very wrong.

  496. flatpack says:

    @Victor Iannello

    Raw data would be interesting but absent a ‘leak’ we are not going to get it. If they were willing to release it post government change, we would have had it by now.

    Having said that, on page 4 we are told:-

    “The only useful information obtained from the Military radar was the latitude and longitude position of the aircraft as this data is reasonably accurate.”

    Then they do on to discuss the very data that they implied was not actually ‘useful’. Odd that, maybe some points are ‘better’ than others…

    If I squint at the profile chart on page 8 with my mumpty goggles on, I see three black blobs, each with a thick red arrow sticking out of it. That tells me; ‘these are the data you are looking for’.

  497. Gysbreght says:

    Safety Investigation Report:
    “The Mode S symbol of MH370 dropped off from radar display at 1720:36 UTC [0120:36 MYT], and the last secondary radar position symbol of MH370 was recorded at 1721:13 UTC [0121:13 MYT].”

    The MLAT points indicate that that the transponder continued to operate without Mode S, probably until 1721:13 UTC (i.e. the SSR symbol was not ‘coasting’).

  498. ventus45 says:

    @Victor,

    Your image is the best I have seen yet, great detail, but I think it needs expansion “back track”, to last ACARS at least.

    When MH370 departed WMKK, it was on a SID, the “Pibos Departure”, when it was re-cleared “DIRECT IGARI”.

    If that clearance had not been given, it would have flown the SID, and thence the filed flight plan, ie, to PIBOS, thence along airway R208 via MATSU, VKR (VOR), IKUKO, and UPRON, to IGARI.

    BUT IT DID NOT DO THAT.

    The clearance “DIRECT IGARI”, meant that the aircraft was always west of R208, gradually converging on it, as it approached IGARI.

    I suggest that the best chance of possibly determining when the aircraft was “taken” would be to examine the track from last ACARS (at least) to IGARI (west of R208) in minute detail.

  499. Victor Iannello says:

    @ventus45: I don’t understand your point. Nobody is debating that the Direct to IGARI path was flown as opposed to the filed flight plan. There is no indication of any deviation from that path before IGARI. I could show more of the path, but all the detail around IGARI would be lost. What exactly are you expecting to see?

  500. TBill says:

    @Victor
    “The turn at IGARI seems close to a bank of 25°…”

    Nice graph and thank you for 25 deg bank calc. Shouldn’t it be less bank if it was set to Auto? The way I have been simulating the “rogue” sharp the IGARI turn on FS9, I set to 25 deg bank instead of “Auto”, and normally I do that ahead of IGARI.

    So this could be a hint that autopilot was possibly used to start the turn?

  501. Victor Iannello says:

    @TBill: In LNAV mode, the maximum bank angle is 25°, subject to stability margins. Before and after IGARI, the change in track angle would be about 35°. I don’t know whether this would cause the AFDS to execute a 25° bank.

  502. DennisW says:

    @all

    In the theme of Richard’s post on various terminal locations:

    1> A mechanical failure is out of the question unless you are an idiot. The SIO data on ZS’s drive completely eliminates that along with statistical data. The exploding nose wheel, exploding oxygen tanks, lithium battery issues,… are a fantasy (only Gysbreght would entertain such thoughts).

    2> The suicide issue is likewise dead. There is no plausible reason to fly the route MH370 flew in that circumstance. A nose dive into the South China Sea would be expected.

    3> That leaves a political motive. A failed negotiation. I won’t bore you with my theories on what that negotiation was all about.

    In my opinion Victor’s Cocos path makes a lot of sense. I was originally inclined to favor Christmas Island, but I no longer feel strongly about that scenario.

  503. DrB says:

    @All,

    I would like to make some comments summarizing my interpretations of recent discussions in this blog. In particular, I am trying to connect several topics and to understand the implications of doing so. If I have missed something or got something wrong, please point it out.

    1. The addition of the high-precision X-Y civil radar data is helpful in understanding the initial diversion flight path, which began within ~90 seconds after the last radio call at 00:19:30.
    2. My interpretation is that the diversion is consistent with a ~180 degree left turn-around beginning a few seconds prior to 17:21:00 (about 30 seconds after passing abeam of IGARI circa 17:20:31 and then continuing the slight right turn to reach the IGARI-BITOD track of 59 degrees true). There might have been a slight decrease in air speed during the turn back, and an altitude change could have occurred during the turn around.
    3. The subsequent speed increase to an excess of 500 knots ground speed over Kota Bharu began circa 17:32 and did not begin until well after the turn back was completed.
    4. Several actions probably occurred between the last radio call at 17:19:30, the diversion circa 17:21, and the speed increase circa 17:32. These include the following:
    4A. The Left ATC/Mode S Transponder (the one normally used for this flight) was disabled circa 17:20:31 by setting the cockpit switch to STANDBY (or possibly by pulling its overhead circuit breaker P11). This occurred approximately 27 seconds before the 180-degree diversion turn began, and it disabled the radar transponder for the remainder of the flight within civil radar interrogation range.
    4B. The Left Main AC Bus was de-powered by selecting the Left (or the Right) Bus Tie Breaker Switch to ISLN (i.e., “isolation”) and by selecting the Left Integrated Drive Generator (IDG) control switch to OFF. This had the effect of de-powering the SDU, the TCAS, some of the IFE, and some of the main cabin lighting.
    4C. On the ACARS Manager Screen, the SATCON ENABLE and the VHF ENABLE options for transmission were turned off (note: the VHF ENABLE may have been de-selected at the gate).
    4D. The SDU de-powering or the ACARS disabling, or both, occurred before the next expected ACARS transmission at 17:37:29, which did not occur.
    4E. Other actions may have occurred in this time period, including depressurization of the cabin.
    5. Approximately (and possibly even exactly) 1 hour later, power was restored to the Left Main AC Bus circa 18:24, causing a log-on request message to be transmitted by the SDU at 18:25:27.
    6. The switch settings used to restore power to the SDU at 18:24 are critical in correctly interpreting the sequence of events at the end of the flight circa 00:19.
    7. The SDU log-on request transmitted at 00:19:29 implies the SDU was powered up circa 00:18:30 and the APU therefore auto-started circa 00:17:30.
    8. The APU only auto-starts when power is lost from both Left and Right Transfer Buses.
    9. The Right Engine consumed more fuel than the Left Engine, leading to a predicted fuel exhaustion roughly 8 minutes prior to fuel exhaustion of the Left Engine.
    10. The question is, did the Left Engine flame out at 00:17:30, or was it the Right Engine?
    11. The answer to this question is determined by the switch settings used to restore power to the Left Main AC Bus at 18:24 (assuming they were not changed in the interim).
    12. If the Left IDG were simply turned back ON at 18:24 (with either BUS TIE isolated), then the APU would not auto-start until the Left Engine flamed out (some minutes after the right) and the Left Transfer Bus was therefore de-powered in addition to the Left Main AC Bus).
    13. Note that if the Right Engine flamed out circa 00:09:30, 8 minutes before the Left Engine flamed out, while cruising at high altitude, the aircraft would first slow down and then begin to descend (toward the nominal single-engine maximum cruise altitude of ~FL270) even before the Left Engine flamed out at 00:17:30. In this case, the lower aircraft TAS and the Rate of Descent must be used as initial conditions for simulating what the aircraft did subsequent to Left Engine flame out and the concomitant reversion to Secondary Mode of the auto-pilot. This might have a significant effect on potential phugoidal oscillations.
    14. An alternative method of restoring power to the Left Main AC Bus at 18:24 is simply to set the Left Bus Tie switch back to AUTO so the Left Main AC Bus is now powered by the Right IDG. This will cause the APU to auto-start when the Right Engine fails (assuming the Left Backup Generator is still OFF).
    15. For the APU to auto-start upon Right Engine flame out, both the left Main AC Bus and the Left Transfer Bus must be powered by the Right Engine (in addition to powering the Right Main and the Right Transfer Buses). In this case, the Right Engine would flame out at 00:17:30, and the Left Engine would continue to produce thrust well after 00:19:37. Perhaps this could have produced an impact more than 22 NM from the 7th Arc. I don’t know what happens to the Thrust Asymmetry Compensation in this unusual situation. I suppose the auto-throttle servo motor would also go dead, perhaps holding the Left Engine fuel flow fixed.
    16. Why were no SDU transmissions received after 0019:37? IFE transmissions would be expected to occur circa 00:21, unless the IFE were turned off after 18:29 and before 00:21. The motivation for turning off the IFE could have been to disallow any communication from the cabin, although this was not done prior to the 18:24 re-powering of the SDU. Was this an afterthought?
    17. Alternatively, the aircraft crashed prior to 00:21, or it was oriented so that the line of sight to the satellite was blocked, but in this case the aircraft does not seem likely to have traveled > 22 NM from the 7th Arc.
    18. Alternatively, the APU ran out of fuel prior to 00:21, but in this second case of Right Engine failure at 00:17:30, this appears inconsistent with the Left Main Tank still holding enough fuel then to operate the Left Main Engine for an additional ~8 minutes.
    19. If the Right Engine flamed out at 00:17:30, the Left Transfer Bus was being powered by the Right Engine then and it would seem likely that the IFE had been manually turned off after 18:29. In this case, the aircraft would have Left Engine thrust continuing until circa 00:25, and the aircraft could possibly have traveled > 22 NM from the 7th Arc before impacting the ocean.
    20. If the foregoing is indeed possible, then It would therefore seem worthwhile to search wider along the 7th Arc, especially near 31.6S, where the 181.2 degree CMT route ends and which is consistent with many of the drift studies, especially the forward drift models.
    21. Griffin continues to claim that his drift model indicates no solution north of about 30.5S. He also discounts the accuracy of the water temperature estimates based on barnacle growth patterns seen on the flaperon, and he assigns low confidence in using this method to discern the starting drift latitude.
    22. This is a key point, and Richard Godfrey has argued that the inferred water temperatures are sufficiently reliable for estimating a northern range of impact latitude.

    Switchology Addendum

    Here are the four cases of AC Bus switches: one for normal flight, one for de-powered Left Main Bus, and two possibilities for re-powering it.

    Case #1 is the normal in-flight settings for the AC power switches.

    Case #1 – Normal Flight Settings :

    L BUS TIE : AUTO, L GEN CTRL : ON, L BACKUP GEN : OFF

    R BUS TIE : AUTO, R GEN CTRL : ON, R BACKUP GEN : OFF

    In this case both L & R IDG’s are turned on and are operating, and both Backup Generators are turned off. The TRANSFER BUSES are being supplied power by the MAIN AC BUSES for each side, respectively. Both BUS TIE switches are set to AUTO, which normally would allow one IDG to power both MAIN AC BUSES (and the TRANSFER BUSES) in case one IDG should fail.

    Case #2 – Left IDG Off & one Bus Tie Open (L MAIN AC BUS is de-powered) :

    L BUS TIE : ISLN, L GEN CTRL : OFF, L BACKUP GEN : OFF

    R BUS TIE : AUTO, R GEN CTRL : ON, R BACKUP GEN : OFF

    Case #3 – Left IDG back on & both Bus Ties isolated (L MAIN AC BUS is re-powered from Left IDG) : same as normal flight Case #1 :

    L BUS TIE : AUTO, L GEN CTRL : ON, L BACKUP GEN : OFF

    R BUS TIE : AUTO, R GEN CTRL : ON, R BACKUP GEN : OFF

    In Case #3 the same two switches used to de-power the Left Main AC Bus are used to re-power it at 18:24

    Case #4 – Left IDG Off & both Bus Ties closed (L MAIN AC BUS is re-powered from Right IDG) :

    L BUS TIE : AUTO, L GEN CTRL : OFF, L BACKUP GEN : OFF

    R BUS TIE : AUTO, R GEN CTRL : ON, R BACKUP GEN : OFF

    In this fourth case, the Left Main AC Bus is powered up at 18:24 using the Right IDG by means of the Left Bus Tie Switch, rather than by turning the Left IDG back on. This case causes the APU to auto-start upon Right Engine fuel exhaustion circa 00:17:30.

  504. DennisW says:

    @DrB

    Forward drift models are fundamentally flawed, but you know that (only Ge Rijn woud argue otherwise). Basically I have a low regard for Griffin.

    I like the 30S region analytically, but those analytics have a lot of arm waving.

    Victor’s solution is appealing from a motive perpective. I think we have beat analytics to death without any resuult. It is time to consider motive.

  505. lkr says:

    @DrB: “Griffin continues to claim that his drift model indicates no solution north of about 30.5S. He also discounts the accuracy of the water temperature estimates based on barnacle growth patterns seen on the flaperon, and he assigns low confidence in using this method to discern the starting drift latitude.
    22. This is a key point, and Richard Godfrey has argued that the inferred water temperatures are sufficiently reliable for estimating a northern range of impact latitude.”

    IMO, inferring water temperature from stable isotope analysis is established science. What I see in the French report appears to have been done carefully, and I would trust the inferred water temperature over the life of the barnacles [along with indication that the smaller barnacles represent more recent establishment.

    The inferred age by size class is of course much less precise.

  506. DrB says:

    @DennisW,

    Actually, reverse drift models are fundamentally flawed.

  507. DennisW says:

    @DrB

    You are truly humorous to me.

  508. DrB says:

    @DennisW,

    I laugh at a few of the things you say, too.

  509. Andrew says:

    @Victor

    RE: “Before and after IGARI, the change in track angle would be about 35°. I don’t know whether this would cause the AFDS to execute a 25° bank.”

    I’d be surprised if the AFDS would command a bank angle of 25° in LNAV mode for a relatively track change of 35°. As I mentioned in a previous post, I would expect the bank angle to be somewhere around 15°.

    I calculated a turn radius of 10.1 nm, based on the last three points in Richard’s data set (17:20:23.250-17:20:30.850). That gives a bank angle of 18° at 471 kt.

    @TBill

    The ‘Auto’ selection only affects the bank angle in HDG SEL or TRK SEL modes. It has no effect on the bank angle in LNAV mode.

  510. Andrew says:

    Oops – delete ‘relatively’ in the first paragraph of my previous post.

  511. Richard Godfrey says:

    @Paul Smithson

    You stated “@All. The objections to my proposed solution (path/speed profile) between 172031 an 173033 seem to gone rather quiet.”

    ICYMI, I posted a detailed discussion and graphic of the flight path between 17:20:30.850 UTC and 17:30:33.310 UTC, concluding a bank angle of 27° fits the timing of the radar data, which disagrees with your findings.

    Previous comment: http://mh370.radiantphysics.com/2018/07/30/mh370-safety-report-raises-many-questions/#comment-17691

    Graphic: https://www.dropbox.com/s/c8ve5q9x64vaj8s/Combined%20Position%20Data%20Turnback.pdf?dl=0

  512. Richard Godfrey says:

    @Andrew

    You conclude the turn near waypoint IGARI towards BITOD was at a bank angle around 18°, assuming a ground speed of 471 knots. This is on auto-pilot.

    I conclude that the turn back 2 minutes later was at a bank angle around 27°, assuming a ground speed of 471 knots increasing to 486 knots as the head wind changes to a tail wind. This is not on auto-pilot.

    https://www.dropbox.com/s/c8ve5q9x64vaj8s/Combined%20Position%20Data%20Turnback.pdf?dl=0

  513. flatpack says:

    @DennisW

    RE SDU testing, you wrote

    Initially I was annoyed by the lack of physical testing, but I now believe that the results would be very “dispersed” (tests would show a lot of variabilty from unit to unit and would probably not be statistically meaningful).

    Assuming you are correct, would such a result not have been been useful in itself ?

  514. Gysbreght says:

    Starting from an old chart that I re-posted here a 10 days ago, I added:

    – the Flightradar24 positions
    – a turn at 32 degrees bank (one of the SIR simulations of the turnback)
    – turns at 26 degrees bank starting at two locations on the IGARI-BITOD route segment

    https://www.dropbox.com/s/6wunt9rgprbq4hw/3_Turns.pdf?dl=0

  515. Victor Iannello says:

    @DrB said: If the Right Engine flamed out at 00:17:30, the Left Transfer Bus was being powered by the Right Engine then and it would seem likely that the IFE had been manually turned off after 18:29. In this case, the aircraft would have Left Engine thrust continuing until circa 00:25, and the aircraft could possibly have traveled > 22 NM from the 7th Arc before impacting the ocean.

    The TAC would likely only partially compensate for the thrust asymmetry at the time of fuel exhaustion of the right engine, and the autothrottle would also disengage (no power to the servo motors). The plane would have to remain in trim as the plane descended with this very asymmetric configuration. Also, the final BFO values (the vertical speed and downward acceleration) do no support an extended phugoid with no pilot inputs.

    At the time of fuel exhaustion of the left engine, the plane would roll quickly to the left. Also, the depowering of the IFE from the cockpit should not affect the log-on of the IFE head.

  516. Gysbreght says:

    @Andrew: “I’d be surprised if the AFDS would command a bank angle of 25° in LNAV mode for a relatively track change of 35°. As I mentioned in a previous post, I would expect the bank angle to be somewhere around 15°. “

    The civil radar data show a turn rate of 1 deg/second and that corresponds to 24 degrees of bank at 484 kTAS (M.82 at FL350/ISA+10.7C).

    It would be of interest to determine the bank angle in LNAV for a turn through 34 degrees.

  517. Paul Smithson says:

    @Richard G. You referred me to your earlier comment-17691. In my view, it is not a superior solution. You match Entry Waypoint at the expense of a poorer match to time and position at Exit Waypoint. You get back to 173033 on time but at the wrong speed. You get a poorer match to the intermittent segments of miltary trace pre-173033. You employ an angle of bank that is normally precluded by bank angle protection.

  518. Gysbreght says:

    Apology – “Microsoft Print to PDF” distorts the aspect ratio.

    https://www.dropbox.com/s/5ddhpvfprnqxvnp/3_Turns.png?dl=0

  519. Gysbreght says:

    @Victor Iannello: “At the time of fuel exhaustion of the left engine, the plane would roll quickly to the left.”

    The airplane turned right in two of the 2016 End-of-Flight simulations with loss of autopilot at the first engine flame-out. Why ?

  520. Richard Godfrey says:

    @Paul Smithson

    I did not state that my solution is superior. I was only answering your complaint that no one was looking into the turn back.

    As I have pointed out before, your solution relies on ignoring the military radar data according to the Malaysian Safety Report.

    You also ignore the head wind changing to a tail wind.

  521. Victor Iannello says:

    @Gysbreght: At FE of the right engine, the right engine spools down and the left engine spools up. The TAC will apply left rudder, but the amount of rudder does not change after AC power is lost and the control law degrades to secondary. The turn to the right means the rudder position does not completely balance the yaw from the resulting thrust asymmetry. Either the transition to secondary occurs before the TAC can reach the value to fully compensate for the thrust asymmetry of no right engine thrust and full thrust of the left, or the TAC by design does not completely balance thrust asymmetry. Whatever the explanation, the Boeing simulations show a turn to the right followed by a sharper turn to the left, presumably after the left engine flames out.

  522. Gysbreght says:

    @Victor Iannello: Thank you for your explanation which makes sense.

    However, in the two simulations I referred to the airplane stayed in the right turn all the time until it ended in a spiral dive.

  523. Victor Iannello says:

    @Gysbreght: I don’t know the differences between the two simulations that rolled the right and then left, and the two that continuously rolled to the right. In both, the A/P was disengaged after the right engine flamed out. Perhaps there is a difference in the amount of trim unbalance.

  524. Andrew says:

    @Victor

    RE: ”Either the transition to secondary occurs before the TAC can reach the value to fully compensate for the thrust asymmetry of no right engine thrust and full thrust of the left, or the TAC by design does not completely balance thrust asymmetry”

    I’d say it’s the former. AC power loss occurs before the maximum thrust asymmetry is realised, so the asymmetry isn’t fully compensated. I doubt the left engine would increase thrust very much in that scenario. The autothrottle isn’t particularly aggressive in the cruise, so the AC power would probably fail before the autothrottle commanded much increase in thrust.

  525. Victor Iannello says:

    @Andrew: Yet the FCOM states

    TAC does not fully compensate for the failed engine so the pilot can recognize failure through roll/yaw clues. These roll/yaw clues are greatly reduced when compared to an airplane operating without TAC. The amount of rudder used is proportional to the engine thrust difference. Rudder movement is back-driven through the rudder pedals and displayed on the rudder trim indicator. Following engine failure, the pilot can trim the airplane using additional rudder trim, control wheel input, aileron trim, or autopilot engagement.

  526. TBill says:

    @Andrew
    “The ‘Auto’ selection only affects the bank angle in HDG SEL or TRK SEL modes. It has no effect on the bank angle in LNAV mode.”

    OK well thank you…if that is the case, then that is a fault of the PSS777 sim model.

  527. Andrew says:

    @Victor

    That only applies in cases where there is a very high degree of thrust asymmetry (eg engine failure during take-off), because the PFCs limit the TAC authority to 60% of the available rudder. Consequently, the pillot needs to apply an additional rudder input that also helps to identify which engine has failed. The same does not apply for an engine failure during the cruise, where the thrust asymmetry is less and the TAC can fully compensate..

  528. Victor Iannello says:

    @Andrew: Fair enough. In any case, the turn to the right means the rudder position does not completely balance the yaw from the resulting thrust asymmetry.

  529. Jos says:

    Hi,
    I’ve been a long-time reader of this blog.

    After following the story for 4 years, I think it will never be found unless someone will come forward with the truth.

    The final report only brings out the things that are allowed to be known (mangosteens, batteries) or to work further on it in the wrong direction (radar data and so on).

    There is one thing that bugs me in the final report, and that is the flight sim data on the cap’s PC. Although it seems that this falls in the latter category as well, did anyone notice the two southern points? Why did they not put the coordinates there so there could be speculation on that?
    Does anyone have those coordinates?

  530. Victor Iannello says:

    @Jos: A place to start learning is to read this previous post, including the references provided by links therein.

  531. Don Thompson says:

    @DrB: Your switchology scenarios describe the BACKUP GENERATOR switches as set to OFF. It’s normal, as per FCOM pre-flight procedure, that their state is ON, so as the generators are available if necessary. I’m not clear whether you are describing a scenario where the backup generator supply is intentionally disabled, or whether you expect the switches should be set on ON as a procedure in response to some non-normal event.

    As Victor notes, we’ve verified that the flight deck P5 panel IFE/Pass Seats switch does not remove power from IFE or CSS LRUs located in the MEC. The SATCOM ISO-8208 SSN connections for IFE originate from these LRUs.

    For wider consideration: it is possible that the final Log On Request, successfully received by the GES at 00:19:28, was not the first LOR transmitted by the AES after the SDU recovered from the power interruption. If the AES does not receive a Log On Confirm reply within 12-19 seconds for its first LOR transmit it retries (and retries up to five times with the preferred GES, the maximum delay for each retry extends due to a randomising function). This would not be unexpected as the Stratos Log, for the MH371 flight does show the protocol recovering from ‘lost’ SUs.

    Bank angle: the AFDS limits bank angle above 332kts TAS. Between 332 and 381kts the max bank angle reduces, linearly from 25º to 15º. Above 381kts the limit is 15º. Note this is an AFDS limit, not associated with the PFCS protections. The FMS sends its LNAV steering commands to the AFDS. I expect the intent of this bank angle limit is to ensure a margin for thrust or buffet conditions. Just as a benchmark, MAS training sets out a requirement that a turn up to 45º bank angle can be manually flown with “smooth coordinated pitch, bank, and power to maintain the specified altitude within ±100 feet (30 meters) and the desired airspeed within ±10 knots [and the pilot] rolls out of the turn (at approximately the same rate as used to roll into the turn) within ±10° of the entry or specified heading, stabilizes the airplane in a straightand-level attitude“.

  532. Paul Smithson says:

    @Richard
    “I did not state that my solution is superior. I was only answering your complaint that no one was looking into the turn back.” Thanks

  533. Andrew says:

    @ Don Thompson

    RE: ”Bank angle: the AFDS limits bank angle above 332kts TAS. Between 332 and 381kts the max bank angle reduces, linearly from 25º to 15º. Above 381kts the limit is 15º. Note this is an AFDS limit, not associated with the PFCS protections. The FMS sends its LNAV steering commands to the AFDS. I expect the intent of this bank angle limit is to ensure a margin for thrust or buffet conditions.”.

    The bank angle limiting function described above only applies to the HDG SEL/TRK SEL modes, with the bank selector in AUTO. The AUTO bank angle limit in HDG SEL/TRK SEL mode depends only on TAS. In LNAV mode, the FMS limits the commanded bank angle to ensure the thrust and buffet limits are not exceeded and will command up to 25° bank angle (30° in a hold), even at high TAS.

  534. Andrew says:

    The last sentence in my previous post should read:

    In LNAV mode, the FMS limits the commanded bank angle to ensure the thrust and buffet limits are not exceeded, independent of the AUTO setting on the MCP. LNAV will command up to 25° bank angle (30° in a hold), regardless of TAS, provided the thrust/buffet limits are not exceeded.

  535. TBill says:

    @Andrew
    “…(30° in a hold)”
    I suppose the turn at IGARI looks a little like the start of a holding pattern.

  536. DennisW says:

    Wise’s latest article in Popular Mechanics:

    Marks the last thing I will ever read in Popular Mechanics.

  537. DennisW says:

    @Andrew

    The last sentence of the article Jeff posted on his site

    Acts of wanton violence by white men bearing racial grievance have been on the rise during the Trump administration, which has embraced white nationalism and adopted openly racist policies.

    was not included in the article published in Popular Mechanics, yet Wise makes no mention of that in his blog. The fact that Popular Mechanics even publishes his stuff is reason enough to boycott the publication. I withdrew my online subscription and told them why.

  538. Andrew says:

    @DennisW

    Thanks – I gave up reading his blog some time ago.

  539. DennisW says:

    @Andrew

    I am banned. You cannot disagree with Jeff and survive on his blog.

  540. TBill says:

    @DrB
    Your above scenario with fuel remaining to glide is what I am also worried about.

    I assume the remaining fuel could run quite some time at reduced thrust e.g. idle mode which could enhance glide beyond 100-nm. Possibly well beyond 100-nm glide if there could have been some way to save more fuel e.g., starving engines via adjusting aircraft pitch/ascent (assuming that could push some fuel away from the fuel inlet ports).

    We can hope that did not happen, and search North of 25 South. If it did happen, I currently see that as an intentional attempt to put much distance after the last ping and make the aircraft impossible to find.

  541. Victor Iannello says:

    @DennisW: I guess you just are not tolerant of divergent opinions.

  542. DrB says:

    @TBill,

    I don’t think either pilot knew that some crude positional information was being transmitted by the SDU, in the form of BTO/BFO, even with ACARS and ADS-B disabled. They would not have seen the need to stretch the time and distance between fuel exhaustion and impact, even if the goal were to hide the aircraft.

  543. DrB says:

    @Don Thompson,
    @Victor Iannello,
    @Andrew,

    Thank you for addressing several of my comments.

    Some of the points made were:
    (1) The IFE messages normally sent by the SDU beginning about 90 seconds after log-on acknowledge cannot be stopped from the flight deck. A visit to the MEC to shut off a circuit breaker there is required.
    (2) If the APU started upon Right Engine shutdown, when would it fun out of fuel? Which tanks(s) supply the APU its fuel?
    (3) The normal Backup Generator switch position is ON (I had assumed it was OFF). I infer from this that the Transfer Bus gets its power normally from the IDG, but the Backup Generator is automatically connected to power the Transfer Bus if the IDG fails or is turned off. In my scenario, the Pilot would have to turn the Left Backup Generator OFF to assure the left Transfer Bus was powered by the Right Engine. Offhand, I don’t see a rationale for doing this.
    (4) Does anyone know what configuration Boeing assumed to get Secondary Mode at Right Engine shutdown? What did they assume for the APU start time?
    (5) When the Right Engine stops, the TAC sets the rudder to lessen or perhaps even eliminate (when in cruise) the expected turn to the right. If the autopilot is active, one would expect the aircraft to maintain the same track. However, in Secondary Mode, an initial turn to the right is likely, because of incomplete TAC, but it might be slight. Then, some minutes later, a sharp turn to the left would occur when the left engine failed. The question is, could the aircraft in this scenario fly straight enough to end up beyond 22 NM?

  544. Gysbreght says:

    @DrB: “The question is, could the aircraft in this scenario fly straight enough to end up beyond 22 NM?”

    That depends on the radius of the turns, which depends on the amount of mistrim remaining when the TAC function is lost. The simulator is not perfect. How certain can we be that it simulated the amount of mistrim perfectly?

  545. Gysbreght says:

    Also airplanes are not perfect, and the imperfections are not included in the simulation.

  546. DennisW says:

    @DrB

    I don’t think either pilot knew that some crude positional information was being transmitted by the SDU, in the form of BTO/BFO, even with ACARS and ADS-B disabled.

    Totally agree. In fact, I think the Inmarsat data came as a complete surprise to the Malaysian government as well. The radar data story was invented after Inmarsat made it known that the plane continued flying for some time. The initial search in the South China Sea has never been satisfactorily explained if the radar data was known and used in real time.

  547. DennisW says:

    @Victor

    @DennisW: I guess you just are not tolerant of divergent opinions.

    Like Sabine says, it is probably an age related issue. Ami, says the same thing.

  548. flatpack says:

    @DennisW

    You wrote:-

    2> The suicide issue is likewise dead. There is no plausible reason to fly the route MH370 flew in that circumstance. A nose dive into the South China Sea would be expected.

    Respectfully, I beg to differ. A while ago I was made aware of the circumstances of what was most likely a ‘suicide by plane’ although it was not classified as one and the cause was apparently left open.

    One key parallel to MH370 was that there was no plausible reason to fly the route taken in the circumstances. The crash was not direct and obvious as you apparently expect.

    I suspect that the cryptic events were planned to ‘shield’ the immediate family.

  549. Gysbreght says:

    IIRC, one day after the accident, several news agencies received an email from someone signing as Chinese Martyrs Brigade. The mail was quickly dismissed as a hoax, and it probably was. However, the person who wrote it must have been uncannilly prescient, because it said that the plane would never be found. The mail was well written, as if the text had been prepared before the accident happened.

  550. DennisW says:

    @flatpack

    Do you believe suicide was the motive for the diversion?

  551. Victor Iannello says:

    @DrB asked: The question is, could the aircraft in this scenario fly straight enough to end up beyond 22 NM?

    Two points:

    1) That did not occur in the Boeing simulations.

    2) Perhaps a better question is whether the aircraft could have flown an additional 22 NM after a descent of 15,000 fpm and an average downward acceleration of 0.7g over the preceding 8 seconds. I believe the answer is no.

  552. Gysbreght says:

    @DennisW: Thank you.

    Yes, there are several press items referring to that email. I spent some time trying to retrieve the full text of the message, without success. It seems to have been wiped from the internet.

  553. Gysbreght says:

    @Victor Iannello: “I believe the answer is no.”

    Hmmmm, …

  554. Paxlambda says:

    @Gysbreght: full text seems to be there: http://www.freerepublic.com/focus/f-news/3131469/posts but it is images, not text, so I don’t know how to have a translation…
    I have pasted the different images*** in one there (link valid only few days): https://nsa39.casimages.com/img/2018/08/14/180814024151137943.jpg

    *** It seems that 4 lines are repeated.

    PL

  555. Paxlambda says:

    About the text above: trying some translation using Internet free Chinese OCR, perhaps it is not the true text…

  556. Gysbreght says:

    @Paxlambda: Thanks a lot.

    Unfortunately I can’t read Chinese, and my OCR doesn’t either. Can anyone help?

  557. Victor Iannello says:

    @Richard Godfrey: Regarding your previous comment, I have often wondered whether there is ANY Malaysian military radar data other than what the civilian radar system has fed to the military radar system. There have been so many inconsistencies, and the raw data has never been made public.

    The claim was made that Butterworth had no quick response capability after the fighter pilots went home for the evening. What good is operating a radar system capable of detecting threats during a time period in which no response is possible?

  558. DrB says:

    @Victor Iannello,

    Regarding the question of whether it is possible for the aircraft to fly itself in Secondary Mode such that it matches the final BFOs and ends up > 22 NM from the 7th Arc:

    You said: “

    1) That did not occur in the Boeing simulations.

    2) Perhaps a better question is whether the aircraft could have flown an additional 22 NM after a descent of 15,000 fpm and an average downward acceleration of 0.7g over the preceding 8 seconds. I believe the answer is no.”

    On the first point, the Boeing simulations were limited in number and may not have sampled the actual rudder mis-trim.

    On the second point, you could be right. All we know is that none of the ~ 10 runs ended up > 22 NM away. That doesn’t mean it did not happen. It’s difficult to prove the negative, so all we are left with is opinion. I agree it seems unlikely. I also seems unlikely a conscious pilot would allow a high ROD to develop before taking control and gliding a long distance, but again that does not mean it didn’t happen.

  559. Victor Iannello says:

    @DrB said: We have to deal with probabilities, as we don’t have certainties about many factors.

    As others and I have said, if we hypothesize that a pilot provided inputs after fuel exhaustion for an extended glide, it’s possible (and I think likely) that the impact occurred +/-100 NM from the 7th arc. It’s also possible (and I think likely) that there were pilot inputs after 19:41. With that uncertainty in the crossing of the 7th arc combined with the long glide distance away from the arc, it becomes very difficult to define a search area of manageable size.

    On the other hand, if we hypothesize no pilot inputs after fuel exhaustion, likely the plane impacted close the arc, and we can define a search area along the arc and north of 25S.

  560. Richard Godfrey says:

    @Paxlambda

    I sent the original CMB message to @Haxi and asked for his help to translate.

    https://www.dropbox.com/s/d26wo74f1jf1ger/CMB%20Message.jpg?dl=0

    His response was “In my opinion, this ‘statement’ deserves no attention. This is intentional misinformation claiming that the Malaysian and Chinese govts are behind the disappearance of MH370, for political reasons. It is absurd obviously, and has no scientific value.”

    I had started to try to translate, but when I received Haxi’s response, I gave up.

    https://www.dropbox.com/s/1b8g77f4jrc4tkd/CMB%20Message%20-%20Simplified%20Chinese.pdf?dl=0

  561. Richard Godfrey says:

    @Victor

    Searching underwater between 25°S and 18°S along the 7th Arc with a search width of ±22 NM, results in a search area of around 76,000 km2.

    Searching underwater between 38°S and 25°S along the 7th Arc with a search width of ±110 NM (discounting the area already searched), results in a search area of around 766,000 km2.

    The first search area is currently doable with Ocean Infinity’s technology and the second search area is not.

    If MH370 is not found in the first area, then we give up until new technology comes along to make a much larger search area doable.

    For the second search area, you need a fleet of AUVs, which can surface, recharge batteries from solar energy, download the data from the previous search via satellite and then resume underwater operations automonously. You simply leave the fleet of AUVs in the Indian Ocean until they find MH370. A fleet of 15 super AUVs for 3 months should do the trick for 766,000 km2.

    However, you can cover the first search area in 3 months on-site with 6 normal AUVs plus one port visit in between.

  562. Victor Iannello says:

    @Richard Godfrey: Those are numbers are quite sobering.

  563. TBill says:

    Re: CMB
    I recall there was some old discussion on Mike Chillit’s defunct web site, I think it was called SeventhArc.net at the time.

    A Chinese speaking expert said he felt that the untraceable email claim was contrived by someone who does not actually speak Chinese. So I have always wondered if it could have been sent out by ZS as a timed outgoing message or an accomplice.

  564. flatpack says:

    @DennisW

    You asked:-

    Do you believe suicide was the motive for the diversion?

    A yes or no answer would be a false choice. So, to me, the suicide issue is not dead but sadly still on the table and is more probable than mechanical failure.

    Ultimately it might be difficult to determine the primary motive, for example someone who is, for whatever reason, deeply dissatisfied with their lot, depressed and has suicidal ideation may be more likely to seek to become a martyr, thus giving meaning to their actions.

    You wrote:-

    That leaves a political motive. A failed negotiation. I won’t bore you with my theories on what that negotiation was all about.

    I would welcome you, or anyone else, to set out your stall in this regard. It seems to be the 600 pound gorilla in the room here.

  565. Gysbreght says:

    @TBill: “A Chinese speaking expert said he felt that the untraceable email claim was contrived by someone who does not actually speak Chinese. “

    That certainly makes a lot of sense, and leaves open many possibilities. I never thought the message was what it pretended to be. That is not the issue. It was sent to divert attention from the real cause. By whom?

    Incidentally, how many people know how to send an untraceable email? I don’t. Is it easy, or somewhat elaborate?

  566. Victor Iannello says:

    @flatpack: @TimR has presented that case many times.

  567. Gysbreght says:

    On 9 March 2014, while the search was focused on the South China Sea, what did the person know to predict that the airplane will never be found?

  568. TBill says:

    @DennisW
    Re: Suicide/mass murder
    This NYTimes article discuss rationale and basically it mentions problems wth relationships or other “massive disgruntlement”. So you are saying political motive not mential health issue, but it is not clear to me how we separate those issues.

    https://www.nytimes.com/2015/04/07/science/the-mind-of-those-who-kill-and-kill-themselves.html?_r=0

    I am open to various possible reasons for MH370 human causes (conspiracy, suicide, stunt, etc) we just do not know.

  569. Nederland says:

    An English tranlation of the above-mentioned statement (I can’t say if this is accurate) was posted on this website, which is now offline:

    http://www.seventharc.net/2016/10/14/mh370-terror-claim/

    The text posted there was the following:

    ” Declaration and Explanation about Malaysia Airlines MH370
    Malaysia Airlines,
    Malaysian government,
    Chinese government,

    ” I am going to deliver important information about the missing flight MH370. You all must be wondering: how a flight can suddenly be lost of contact, and lost without any trace afterwards? Now everybody is trying to search for the whereabouts of people on board and the airplane itself. But here I am going to announce: all your searching is going to be futile! All the people on board had kneed down and made their confession to Allah, no exception! If you want to search for them, the only way is to see Allah yourselves! Therefore I would like to give you the following suggestion: stop searching immediately and stop doing pointless things.

    ” This is a political incident! There are two reasons why the MH370 incident happened: first of all, this is a revenge to Malaysian government. This is the revenge for the cruel persecution that Malaysian government did to me (Malaysian government should know exactly what it has done, please reflect yourself). The incident of MH370 is the revenge to Malaysian government. Secondly, there are many Chinese passengers on the flight. The reason for crashing this flight is the answer to the cruel repression and persecution which Chinese government has done to Uighur people. On March 1st, there was a uprising activity of Uighur warriors in Kunming, China. The cause of the uprising is Chinese government’s long-term cruel repression to and racial discrimination against Uighur people. But Chinese government shot them dead ruthlessly, including females. A female teenager was caught because of injury and now is undergoing torture! The incident of MH370 was the result of Kunming incident in China. This is a revenge to Chinese government. We are warning Chinese government: if you kill a single Uighur person, we will kill hundred Chinese people to compensate.

    ” The two reasons explained that Malaysia Airlines MH370 incident is a pure political incident. Here I would like to declare that Malaysia Airlines takes no responsibility in this incident. The airplane has no technical problems and pilots made no fault operation. Malaysian government and Chinese government should take the responsibility of the happening of this incident, in which Chinese government should take 60% of the responsibility while Malaysian government should take 40% of it.

    ” For the passengers with other nationalities (not Chinese nor Malaysian nationalities) on MH370 flight, I have to say sorry to and feel sorry for them. We also had hoped that only Chinese people were on board. But as you know, it is not possible. There are always passengers from other countries. But the number is limited, and this will not hurt their countries that much. Our main target is Chinese government and Malaysian government.

    ” Since Chinese government should take the main responsibility of this incident, we suggest that Malaysian government make 40% compensation to the family members of victims while Chinese government pay for the rest 60%. Furthermore, Malaysian government can also ask for compensation from Chinese government, since Chinese government is the trouble-maker.

    ” Last but not least, Chinese government should reflect its policy towards ethnic groups and human rights. Chinese government should immediately stop persecuting people of ethnic groups and immediately release the Uighur warriors who were captured in Kunming on March 1st. Otherwise MH370-like incidents will happen to Chinese people endlessly. To Malaysian government, I have to say that problem will never end! Malaysian government cannot compensate the resentment by a airplane and several dead bodies. This is a not a terror threat, but a description for the future. I have to say that Malaysia will have to face all kinds of incidents in the future. Of course, it does not have to be plane crash. To Chinese civilians, you should work together to stand up against Chinese government. You should not be the accomplice with the government! Otherwise, you are going to face more misfortune and more civilians will die in vain.

    ” China Martyrs Brigades leader 9-3-2014″

    —–

    Two details could point to knowledge of what happened: a) the plane will never be found and the current search was futile b) the plane was at one point heading towards Mecca (which it did en route from Penang to MEKAR/NILAM.

  570. Gysbreght says:

    @Nederland: Thanks.

  571. TBill says:

    @Nederland
    Yes thank you, that is the statement.
    Yes airway N571 is quite pointed to Mecca (not to say I especailly agree with that idea). Also, Ed Baker had felt the flight could have been directed towards Mecca later in the flight…note that airway L894 to 22 South also generally points towards Mecca, and in order to make BTO ping timings work on L894, it is basically necessary to envision a huge slow down (or possible maneuvers such as turn back).

  572. lkr says:

    @Ned:

    The claim that the plane would never be found is hardly surprising. Sounds standard “we did it” boilerplate to me. Especially if finding it would likely debunk a false crediting for the hijack..

  573. Don Thompson says:

    The website, http://www.seventharc.net, is a site maintained by Chillit. The page, originally serving the doc image of alleged claim of responsibility is now redirecting to a chinese site.

  574. Victor Iannello says:

    @Don Thompson: I wonder whether Chillit is aware of the redirection to this site. He might be a bit surprised about the content.

  575. mash says:

    Gysbreght says:
    August 14, 2018 at 5:36 am

    “However, the person who wrote it must have been uncannilly prescient, because it said that the plane would never be found.”

    If this were a well-organised ‘disappearance’ mission, perhaps someone could really (somehow) know how and where the plane landed, so this message could still possibly be ‘created’ by the real ‘originator’ (for whatever purpose) – given its high predictive power.

    At least, by providing such a ‘long’ message after such a ‘short’ time (after the event), the sender acted as if they were very confident that there would not be any subsequent contradictions (contradictory claims by other parties). Therefore, this is perhaps just another indication of the ‘pattern’ (of the ‘omniscient’ party) – distraction and deception.

  576. TBill says:

    @Nederland
    Do you have access to the comments on the CMB statement on Mike Chillit’s site? I am thinking one person chimed in with a pretty good analysis that it may not have been authentic from a Chinese native speaker.

  577. Andrew says:

    @DrB

    RE: “(2) If the APU started upon Right Engine shutdown, when would it fun out of fuel? Which tanks(s) supply the APU its fuel?”

    The left wing tank supplies the APU with fuel. The question of ‘when would it run out of fuel?’ is a bit more complicated. I think the APU would probably flame out several minutes after the left engine flamed out, possibly earlier if the aircraft attitude changed significantly after the autopilot disengaged. My reasoning follows:

    In the scenario where AC power is lost after the right engine fails, the forward and aft boost pumps in the left wing tank would fail and the APU would be fed by the APU DC fuel pump. The left engine would operate by suction feed via the suction bypass valve in the left tank, and would flame out when the fuel level dropped below the level of the valve.* At that point the APU would still have about 200 kg of fuel available, due to the different location of the inlet for the APU DC fuel pump. However, upon left engine flame out, the APU isolation valve would automatically open and the engine would then be fed by the APU DC fuel pump as the engine attempted an auto-relight. In that configuration, the remaining fuel would be consumed fairly quickly (several minutes?), especially if the engine relit, causing the ‘terminal’ flame out of both the engine and APU.

    However, the fuel supply relies on the aircraft attitude remaining fairly stable following autopilot disengagement. A change in attitude after autopilot disengagement might cause the small quantity of fuel remaining in the tank to move away from the fuel inlets and starve the APU and/or engine of fuel.

    *Possibly earlier at altitudes higher than FL380.

    RE: “(3) The normal Backup Generator switch position is ON (I had assumed it was OFF). I infer from this that the Transfer Bus gets its power normally from the IDG, but the Backup Generator is automatically connected to power the Transfer Bus if the IDG fails or is turned off.”

    Yes. The L & R transfer buses are normally powered by the L & R AC buses respectively. If one or both transfer buses lose power, the backup power system automatically supplies power to the affected bus(es). The backup power system consists of the L & R backup generators and a single backup generator converter that converts the variable frequency power from the generators to 115V, 400 Hz AC power for the transfer buses. The backup generators are normally selected ON and ready to supply power, however, only one backup generator supplies power at a time. If one or both transfer buses need power, the associated converter circuit breaker(s) (CCB) close, allowing the backup power system to supply the affected transfer bus(es). If the L backup generator is selected OFF, the R backup generator will supply the backup power system, but the L CCB will remain open and the L transfer bus will remain unpowered.

  578. Victor Iannello says:

    If the Chinese Martyr Brigade (or any other group that might have written the email) was involved in the disappearance of MH370 as they claimed, they would have provided information that proved that involvement. They didn’t provide that information because they had none, which is why they were ignored then, and they should be ignored now. It’s not as though they haven’t had time. By predicting that the search would be futile and therefore should be stopped, they were buying time, because they knew once the plane was found, their claims would be proven false. As it turns out, the plane was not found, but also nobody takes their claims seriously. I think even THEY stopped trying to persuade people that they are responsible.

  579. DrB says:

    @Andrew,

    Thanks for addressing my question regarding the Transfer Buses. That is very interesting, and surprising to me. You said that the Left Transfer Bus would be unpowered if both the Left IDG and the Left Backup Generator were turned OFF. What equipment items get their power from the Left Transfer Bus? How would an unpowered Left Transfer Bus affect Secondary Mode controls and the subsequent aircraft flight?

    Can you think of any scenario or rationale for turning off a (working) Backup Generator?

    Considering the scenario where the Right Engine fails at 00:17:30 and the left engine continues thrusting, and assuming the aircraft control is then secondary mode with no human inputs, do you think it is possible to have rates of descent consistent with the last 2 BFOs?

  580. Andrew says:

    @DrB

    RE: “You said that the Left Transfer Bus would be unpowered if both the Left IDG and the Left Backup Generator were turned OFF.”

    Yes, but that assumes the right engine has failed and the APU is not yet running. If the L IDG and backup generator are selected OFF, the L transfer bus will lose power when the right engine fails. The APU generator will re-power the L & R main and transfer buses once the APU is running.

    RE: “What equipment items get their power from the Left Transfer Bus?”

    The transfer buses power a number of items of important equipment; hence the backup power system. I don’t have a comprehensive list, but equipment powered by the L transfer bus includes:

    L pitot probe heat
    L radio altimeter
    L multi-mode receiver (ie GPS)
    L ILS
    L DME
    L ADF (where fitted)
    GPWS
    L weather radar
    L HF
    CVR
    L FWD/AFT boost pumps
    Some cockpit instrument panel lighting
    L FWD window heat

    RE: “How would an unpowered Left Transfer Bus affect Secondary Mode controls and the subsequent aircraft flight?”

    With the L IDG and L backup generator selected OFF, the left transfer bus would be unpowered from the time the right engine failed until the APU generator became available. The right transfer bus would also lose power when the right engine failed, causing the PFCS to revert to secondary mode and the autopilot (& TAC) to disengage. The subsequent restoration of power by the APU would not have any effect on the PFCS mode. Normal mode would only be restored if the crew were to cycle the PRIMARY FLIGHT COMPUTERS disconnect switch on the overhead panel.

    RE: “Can you think of any scenario or rationale for turning off a (working) Backup Generator?”

    No.

    RE: “Considering the scenario where the Right Engine fails at 00:17:30 and the left engine continues thrusting, and assuming the aircraft control is then secondary mode with no human inputs, do you think it is possible to have rates of descent consistent with the last 2 BFOs?”

    I think it’s possible, but there’s still a question of the timing of the BFOs in relation to the the AC power loss. I don’t think that has been resolved.

  581. Nederland says:

    @TBill

    No, I don’t have access to that site and can’t remember any comments, I simply copied the text into a personal file.

    I don’t think any particular group wanted to claim that incident. Nevertheless, it is not impossible that the text could reveal some insider knowledge. The prediction that “Malaysia will have to face all kinds of incidents in the future” is chilling.

    At the time, it was not ruled out that this was an act of terrorism and claims of responsibility were not necessarily considered a hoax.

    https://abcnews.go.com/Blotter/cia-chief-ruling-terrorism-malaysia-airlines-tragedy/story?id=22863091

  582. Victor Iannello says:

    @Andrew said: I think it’s possible, but there’s still a question of the timing of the BFOs in relation to the the AC power loss. I don’t think that has been resolved.

    In the abnormal electrical configuration, if there was a banked descent to the left caused by left rudder TAC input at the time of failure of the left engine, and the left engine failed about two minutes after the right, the timing of the BFO values can be explained. We don’t know for sure when one engine failed relative to the other. For all we know, the fuel was actively balanced for a lot of the flight.

  583. Gysbreght says:

    @Victor Iannello: “For all we know, the fuel was actively balanced for a lot of the flight.”

    Are you assuming that whoever did the balancing was inactive at the end of the flight?

  584. Gysbreght says:

    A related question: Could the fuel imbalance have contributed to the tendency observed in the simulations for the airplane to turn left at low bank angle?

  585. Victor Iannello says:

    @Gysbreght: It’s possible that fuel was balanced actively by the pilot and then that pilot became incapacitated later. We also don’t know exactly how one engine performed relative to the other. The bottom line is that the timing of the flame out of one engine relative the other is a guess.

    I suppose the fuel tank imbalance could have caused a small bank to the left. Ultimately, the direction of the bank is the net effect of a many factors, including thrust asymmetry, TAC, manual rudder input, weight imbalance, aerodynamic asymmetry, etc.

  586. Gysbreght says:

    @Victor Iannello:
    “We also don’t know exactly how one engine performed relative to the other.”

    I thought we knew that fairly accurately for the preceding flight MH371.

  587. Victor Iannello says:

    Malaysia Today is running a story claiming that the Malaysian Defense Minister, Mat Sabu, will announce on Saturday that they have found MH370.

    Recently, Mat Sabu claimed that only four of the 28 Sukhoi fighter jets are operational.

    The entertainment value of Malaysian politics never ceases to disappoint.

  588. Victor Iannello says:

    @Gysbreght: We believe the PDAs for the right and left engine were 2.25% and 0.75%, respectively. We don’t know the uncertainty in those numbers. We don’t know how those numbers might change for the specific conditions of MH370. We don’t know if fuel was manually balanced between the tanks during the flight. All these factors change when one engine might flame out relative to the other.

  589. DollarBill says:

    @Gysbreght

    “Incidentally, how many people know how to send an untraceable email? I don’t. Is it easy, or somewhat elaborate?”

    This is a challenge I have faced several times in my career: both the sending, and the tracing.

    The short answer is that the cost of any potential search to identify the target is proportional to the value of the target.

    Any email will include information in its headers about the network location from which it originated. That network location usually corresponds to a server (mail transfer agent). The internal workings of the server (logs, etc.) are, in the absence of hacking or a subpoena or other mandate from the authorities, only accessible to the administrator of the server. It is easy to create or subvert a server in such a way that any emails originating from it have no identifying information other than the address of the server. It is also relatively easy to conceal ownership information about the server.

    So for e. g. a Chinese state actor to identify individuals sending an email from a U. S. server, the only option is to hack in to the server, given the absence of the subpoena option. Even if they are successful in breaking in, the trail might only begin there, pointing to other servers (mail transfer agents).

    There are other scenarios such as emails emanating from home routers or other network non-server devices, made less likely by ISP blocking of port 25.

    So,for a moderately experienced hacker, easy.

  590. TBill says:

    @Victor @all
    There is a very deep trench (23,000-ft or more) along the south side of Zenith Plateau, perhaps it is called the Zenith-Wallaby Fracture Zone.

    >>How deep can OI seartch? is my question.

    I have long thought of that trench as a possible intentional landing spot, and previuously noted it falls almost directly on a straight-line path from POLUM to UXORA waypoints. Lately Mike Chillit is “triangulating” on the same spot, which is a coincidence.

  591. Gysbreght says:

    Victor IannellO: “We believe the PDAs for the right and left engine were 2.25% and 0.75%, respectively.”

    Is there possibly a minus sign missing in 0.75% ? This was discussed more than a year ago, ALSM did a convincing study using both fuel flow and fuel QTY. Searching through all posts between 4 and 28 July 2017 I could not find it. Apologies if my memory has failed me again.

  592. Victor Iannello says:

    @Gysbreght: Mike and Bobby studied the historical fuel consumption data, so I’ll let them answer. My understanding is an average PDA of 1.5%, and a difference of 1.5% between the two PDAs.

  593. Richard Godfrey says:

    Hi Victor,

    The DSTG state that they have radar data from 16:42:27 UTC, whereas the IG received radar data from 16:41:43.720 UTC. Boeing in their report as part of the latest Malaysian Safety Report (page 1289, appendix 1.6E), state they only have radar data from the departure in KUL at 16:42:00 UTC and the fuel report from the ACARS message at 16:41:43 UTC is “prior to take off”. This ACARS message is the so-called “OFF THE GROUND” message in the series OOOI (OUT OF THE GATE, OFF THE GROUND, ON THE GROUND, INTO THE GATE).

    The DSTG state that they have radar data up to 18:01:49 UTC, whereas the IG received radar data only up to 18:00:51 UTC. Boeing in their report as part of the latest Malaysian Safety Report, state they have radar data up to 18:01:19 UTC.

    It gets more mysterious, DSTG state their radar data is from 2 radar sources, which agrees with the Boeing report, but the IG has received data from 3 data sources.

    Even more mysterious, DSTG and Boeing claim they have 10 second radar data, whereas the IG has received 4 second radar data.

    Yet even more mysterious, DSTG claim there was only one gap in the radar dataset, Boeing claim there were several and mention 2 in particular, whereas in the IG dataset there are 6 gaps.

    A table of the the various radar datasets is linked below:

    https://www.dropbox.com/s/cuixq307cj3jh2s/Radar%20Discrepancies.png?dl=0

    Quite clearly we, Boeing and DSTG are not being told the truth, the whole truth and nothing but the truth.

  594. Victor Iannello says:

    @Richard Godfrey: I believe the civilian radar data that was supplied to us. I am suspicious about all things related to the military data, including whether it exists. I know people here are confident that the military data really does exist, but I much less confident.

  595. flatpack says:

    @Victor Iannello

    Ten of the 29 fighters are MiG-29s.

    Still, their MiG-29 simulator should be good to go as an acquaintance of mine was one of the design team (and even flew one in Poland IIRC).

    If Sabu is a canny politician he will get one fighter fixed-up, then announce a 25% increase in fighter availability.

    Simples!

  596. Richard Godfrey says:

    @Victor

    You stated “I know people here are confident that the military data really does exist, but I much less confident.”

    I can understand your scepticism, that Malaysian Military Radar installations do not exist or were not functioning, but …

    Are you seriously saying that the military radar installations at Sabang, Indonesia; Lhokseumawe, Indonesia; Sibolga, Indonesia; Dumai, Indonesia; Phuket, Thailand and Hat Yai, Thailand also do not exist or were not functioning?

  597. Victor Iannello says:

    @Richard Godfrey: I didn’t mean to imply that I don’t accept the existence of the radar installations.

    It might be that Indonesia and Malaysia radar, like the Indian radar in the Andamans, is turned off at night. If there is not a willingness to pay for a quick response capability at night during periods when there are no regional conflicts, there might also be no willingness to pay for radar at night.

  598. DrB says:

    @Gysbreght,

    The MH370 Flight Brief gives the average PDA as 1.5%. From the FI there was the engineering report saying the Right Engine consumed 1.5T more fuel than the left. I believe that is a typo, as the units should be %. That nominally gives the numbers Victor mentioned for each engine. I was given fuel quantity and fuel flow sensor data by ATSB for a previous flight. It is consistent with the Right Engine using 1.5% more fuel than the left when in cruise.

  599. Gysbreght says:

    @DrB:
    “Richard says on July 4, 2017 at 11:43 am
    @Victor, Excellent Post! Many thanks.

    We do have one cruise phase fuel flow rate information from MH371 at 05:06:41 UTC, when the aircraft was flying at 40,001 feet at Mach 0.832 and the Static Air Temperature was -56.0 °C (please see the decoded SU Log Excel: Tab “Flight 1” at 05:06:41).

    https://www.dropbox.com/s/6sajojhz3p9v8yp/ISUs%20and%20SSUs.xlsx?dl=0

    Cruise fuel flow rate is 3.51% higher in the Right Engine during flight MH371 at 05:06:41.”

    My recollection is that ALSM concluded that the right engine fuel flow was 3% higher than the left in cruise.

  600. Gysbreght says:

    Also the ATSB said in one report that based on ‘historical’ data the left engine flameout could have been up to 15 minutes after the right engine, which is consistent with 3% difference in fuel flow.

  601. lkr says:

    @RG, VI

    I’m pretty comfortable with the idea that the Malaysian military radars were not exactly on war footing late at night. As for the neighbor countries, I get the impression that the Indonesians [and probably the Thais] have taken a “not-our-problem” attitude. Might have been turned on, might not, and the one thing they care about is letting the details out.

    Further, there seems to be a tendency for militaries to spend their budget on shiny new hardware, and assume the money to train and operate will be available if there’s a conflict. Even idle, it looks impressive when the polis do a walk-through.

    Another thing that I’m comfortable with is that US assets didn’t track the flight. They might well withhold details, but I’m sure the ASLM would have gotten a pointer in the general direction long before the Fugro phase started.

  602. DrB says:

    @Gysbreght,

    The difference in right versus left fuel flow is not as large as 3%. In Climb it is a bit under 1% and in cruise it is 1.5% with an estimated error of about 0.5% 1-sigma at best. I have an entire flight’s data (hundreds of readings) with fuel flow sensor and fuel quantity sensor readings. The largest uncertainty is the errors in the absolute calibration of each of the sensors. The fuel flow sensors give a slightly different ratio than the fuel quantity sensors. If there were no fuel transfers during the flight, I would expect an error of at least 2 minutes in the difference in predicted flame-out times.

  603. DrB says:

    @Andrew,

    Still considering the same scenario, would the still-running left engine provide any hydraulic assist for flight control surfaces used by the secondary mode, either before or after the APU auto-started? If so, could this enhance the stability or reduce the roll rate?

  604. Brian Anderson says:

    Fuel flows:

    Note that the PDAs cannot provide accurate measure of the relative fuel consumption between the two engines. The PDA for any engine is a reasonable estimate of the allowance to be made when calculating the required fuel take-up for a particular flight, above the fuel burn for a [text book] new engine. It provides for the additional fuel required to meet the required engine performance while allowing for the age and hours on the engine, and any changes in performance due to maintenance. Of course the PDA for any engine will be based on recent historical data for the engine, but the actual fuel burn will be based on the nature of the flight profile and any changes to a standard profile en-route.

    The FI report [appendix 1.6B] shows the difference in fuel burn between the right and left engines during takeoff and during climb. During takeoff the right engine fuel flow was 1.9% greater than the left engine. During climb the right fuel flow was about 0.9% greater than the left. Note that these numbers are not directly related to the PDAs.

    A long time ago I tested a variety of fuel burns that might have applied during the last 15 minutes of flight. I tested a range of fuel flow differences that might have applied during say 7 hours of cruise conditions, with 0.8% being a reasonably representative number, and a wide range of fuel burns assuming that there had been no fuel transfer, and that the right engine went to full power when the left engine shut down.

    The maximum fuel remaining for the right engine under any of these assumptions was about 350 Kg, and more likely closer to 310 Kg, and this situation occurs only when the left engine fails.

    If these conditions are realistic then I think it is most unlikely that anyone, at any stage of the flight, would have moved fuel in order to balance the aircraft. My understanding is that a fuel imbalance of up to 500 Kg is considered perfectly manageable, and does not normally require intervention to cross feed fuel.

  605. Gysbreght says:

    @DrB: MH371 was an entire flight, and also the most recent.

  606. Andrew says:

    @Brian Anderson

    RE: “If these conditions are realistic then I think it is most unlikely that anyone, at any stage of the flight, would have moved fuel in order to balance the aircraft. My understanding is that a fuel imbalance of up to 500 Kg is considered perfectly manageable, and does not normally require intervention to cross feed fuel.”

    The maximum fuel imbalance is 2,041 kg when the total main tank fuel is less than 22,680 kg. The limit decreases linearly with increasing main tank fuel quantity, down to 1,360 kg when the fuel quantity is greater than 51,709 kg.

    According to Boeing, the limitation is primarily for ‘the structural life of the airframe and landing gear and not for controllability’(FCTM/Non-normal Operations/Fuel). A fuel imbalance greater than the limit also increases fuel consumption due to trim drag. However, Boeing also states: ‘routine balancing when not near the imbalance limit…does not significantly improve fuel consumption’.

    As you indicated, an imbalance of 350 kg is trivial and there would be no requirement for fuel balancing.

  607. Gysbreght says:

    @Brian Anderson: “I think it is most unlikely that anyone, at any stage of the flight, would have moved fuel in order to balance the aircraft. My understanding is that a fuel imbalance of up to 500 Kg is considered perfectly manageable, and does not normally require intervention to cross feed fuel.”

    I agree, except possibly at the end of the flight, when the “FUEL QTY LOW” appears on the EICAS.

  608. Andrew says:

    @DrB

    RE: “…would the still-running left engine provide any hydraulic assist for flight control surfaces used by the secondary mode, either before or after the APU auto-started? If so, could this enhance the stability or reduce the roll rate?”

    It was shown some time ago that the L & R hydraulic systems remain pressurised, even at windmilling RPM. With both engines failed, every primary flight control surface would have at least one actuator available. The running left engine shouldn’t make any difference.

  609. ventus45 says:

    @Victor, you said
    I believe the civilian radar data that was supplied to us. I am suspicious about all things related to the military data, including whether it exists. I know people here are confident that the military data really does exist, but I much less confident.

    I agree with you on that point at least Victor.
    I think there is now the distinct possibility, more distinct than I previously thought, that there was no Malaysian Military Radar (RMAF) active on the night, at all, i.e. none, and that fact, is the biggest embarrassment, that both the RMAF, and Malaysian Government (both the past government and the new one) are continuing to try to hide.

    My reasoning is as follows.

    The IADS is the IADS.
    To be specific, the “I” in IADS, is the “integrated” bit.
    The IADS integrates both the military radar picture, and the civilian radar picture, to create one, all encompassing, “total air situation display”.
    I think that all the civilian radars (both PSR and SSR) were operating on the night, as normal, but that the RMAF radars (both their PSR’s and their associated SSR’s) were not, they were off.
    However, the civilian radar data, was being fed into the IADS, as normal.
    Thus, the IADS would record, and display it, as normal.
    I propose, that it is therefore likely, that due to the IADS “time base” and the civilian “time base” not being perfectly synchronized, that there is the “impression”, that there are two data sets (one military and one civilian), when in fact there is only one (the civilian), but that there are two copies of it, one civilian and one military (in the IADS), but since they are recorded with slightly different in timing, they are being presented, (to the gullible), as two genuinely independent data sets.

    The fact that we know that the civilian terminal area radar heads (and their associated co-mounted SSR’s) rotate at 15 rpm, lends credibility to the notion that the data at 4 second intervals is entirely civilian.
    The fact that the alleged military radar data is at 10 second intervals, although consistent with the slower rotation rates of long range (400km/250nm) dedicated air defense search radars, it is glaringly inconsistent with the known coverage of the known RMAF primary radars, if they had been operating.
    But, the 10 second data we have, is consistent with the known Thai PSR radar sites, and their known (deduced) coverage.
    In fact, the Thais did admit to three things with respect to their own radar data:
    (a) the “not a normal return” at the IGARI turn back,
    (b) the return subsequently tracked south west (across the peninsula), and
    (c) the return entered the Malacca Strait.
    The Thai’s were pressed very – very hard (diplomatically) by Malaysia, to make those statements.
    Why would they bother, if they had their own RMAF military radar returns, from their own radar heads, which are much closer to the presumed/assumed flight path from IGARI anyway ?
    There are only two possible explanations for the Malaysian refusal to release the RMAF radar data publicly.
    (a) It does not exist at all (as above), or,
    (b) It does exist, but that “it does not fit the narrative” post the Penang FO phone log on.
    Remember again, the “Thai point charlie” above.
    The devil is in the detail.
    The Thai’s DID NOT SAY THEY TRACKED IT UP THE MALACCA STRAIT, (as they should have been able to do – LIDO white circle excluded) they ONLY, repeat ONLY said, “it ENTERED the Malacca Strait”.
    Then, there is the issue of the Indonesian radars, both civilian and military.
    The Indonesians have consistently said:
    (a) (paraphrased) “we did not see it where the Malaysians said it was”.
    (b) (paraphrased) “we did not see it in our airspace”.
    You can interpret those statements in three possible ways.
    (1) Our radars were on, and we did not see it at all, so it did not fly up the Malacca Strait or around Ache, because we have good coverage of those areas, or
    (2) Our radars were on, and we did see it, but it was not where the Malaysians said it was, but we are keeping mum about that, because it did enter our airspace, it did cross Sumatra, but for obvious reasons, that is very embarrassing to us too, so, we are not going to admit it, and we are going to lie about it, or.
    (3) our radars were off too, so obviously we did not see it, and that is a mega embarrassment for us, (just like the Malaysians) so we have to hide that fact.
    Now remember the case of the Indonesian Police General, who opened his big mouth, and was swiftly silenced for doing so.
    I would put money on (2) being the truth (as I did long ago – as I am sure you know).
    But, alas, where it went after Penang, is, in my view, indeterminate, on “the presented” radar data alone.

  610. Brian Anderson says:

    @ Andrew,

    Yes, I wasn’t entirely clear. Of course there is a fuel imbalance right from the start which is much greater. It would be most odd if someone tried to cross-feed fuel at an earlier point in the flight. To do so would risk flame-out of the right engine at a much earlier time.

  611. Gysbreght says:

    On August 12, 2018 at 6:34 pm I said:

    “The MLAT points indicate that that the transponder continued to operate without Mode S, probably until 1721:13 UTC (i.e. the SSR symbol was not ‘coasting’). “

    That observation could be important, considering that:

    Paul Onions said on July 16, 2017 at 5:44 pm:
    “But given the following facts, we know that the transponder was not manually switched off in the cockpit via the Transponder Mode Selector. There isn’t a switch position that enables the transponder without mode S.”

    I believe Mr. Onions information was essentially supported by:
    Don Thompson on July 17, 2017 at 12:18 pm
    Andrew on July 17, 2017 at 4:32 pm

  612. Don Thompson says:

    @ventus45

    On a scale of 0-5 (spectrum: unbelievable thru solid fact) I’m about 4 on the post diversion radar. The pieces fit, loosely. A full log of plot data from the radar system is necessary to take me to 5.

    We have:
    a) data from DCA’s Kota Bharu ATCR-33 PSR;
    b) track illustrations from RMAF’s joint civil-mil NEC PSR located on the aerodrome at Butterworth;
    c) a track illustration that is the ‘Beijing Lido’ image;
    d) a track illustration included in ATSB and DTSG documents described to be compiled from plots at 10sec intervals;
    e) a track illustration included in SKMM’s analysis for the LBS registration of the FO’s cellphone.

    You mention a difference in scan rates vis Thai and other PSRs. If you refer to scan rates of long range air defence surveillance PSRs, your information is incorrect: the RTADS-III FIPS-130 (= ARSR-4) rotates at 5rpm. The TUDM RAT-31DL rotation rate is 6rpm. Thai civil PSR sites are too far distant to have tracked 9M-MRO after its diversion.

    A plot log at 10sec intervals is consistent with the RAT-31DL on Western Hill. It is feasible that radar system tracked 9M-MRO from its climb out of WMKK to the 18:22Z point.

    After collating the radar plots provided in the SKMM’s cellphone registration analysis, the radiosonde data for WMKP (albeit for 00Z 2014-03-08, 6 hrs after -MRO is described to have passed), the geometric altiude derived from the RAT-31DL 3D radar is close to a corrected pressure altitude.

    The factual information presented in the Safety Investigation Report includes Fig 1.1B that depicts speed and altitude profiles derived from Military Radar. After recreating that chart and including the plots from the SKMM report, the profile appears more credible if the obvious outliers are discounted.

    3D PSR technology is an imprecise discriminator of altitude. For example, the FIPS-130/ARSR-4 is stated to be accurate to within 3000ft (with no range bounds specified). A USAF study, in 2000, of radar data from the joint FAA-USAF ARSR-4 site at North Truro, MA, showed 3D PSR radar derived altitude varying between 14K ft and 102k ft.

    The IADS acronym expands to Integrated Area Defence System. I’ve yet to find a credible source that describes the deep surveillance systems integration that some believe is inherent in IADS. Many references describing cooperations among the five powers for broad policy, planning, regular conduct of joint exercises, and hosting partner deployments. But nothing describing data sharing.

    The Safety Investigation Report describes:

    1) that at 2145UTC “ATSC Duty Watch Supervisor requested from the radar maintenance personnel to carry out radar data play-back (with permission granted by KL ATSC’s Chief Assistant Director). The SME successfully restored the desired file from the recording play-back back-up hard disc. At 2200 UTC [0600 MYT] the ATSC Duty Watch Supervisor performed radar data and voice recording play-back.” Section 2.2.8/1/p, page 347 of 495.

    2) “At 1721:13 UTC [0121:13 MYT] the Military radar showed the radar return of MH370 turning right but shortly after, making a constant left turn to heading of 273°, flying parallel to Airway M765 to VKB (Kota Bharu). Section 1.1.3/1, page 50 of 495.

    This Safety Investigation Report sets out that the Malaysian authorities, ATC personnell and TUDM ADOC staff, were aware that 9M-MRO had been diverted, back, towards Penang during the time the aircraft was still airborne. It is a scandal that the Razak/Hishammuddin regime permitted neighbours to expend resources in futile search operations over the South China Sea.

  613. TBill says:

    @ventus45
    In your terminology (IADS), in the Safety Report, Malaysia seems to be adamant that IADS successfully kept track of MH370 as it turned around at IGARI. Therefore it was classified as friendly and known commerical flight.

    I don’t know if IADS could connect the dots from KB radar from 17:20 to 17:31, but possibly it could and make up the strange IGARI-turn shape we see in the middle?

    @Gysbreght
    “…But given the following facts, we know that the transponder was not manually switched off in the cockpit….”
    What is the implication if what you suggest is true? Mechanical failure?

  614. Don Thompson says:

    @TBill,

    At this point, the conflation of IADS and the activity of TUDM’s air defence surveillance is wholly uncorroborated.

    @TBill/Gysbreght

    Probably best to refer back to this comment. Coasting is an extrapolation from last received radar return.

    Slight aside, there is no record that any SSRs employed Mode-S interrogation, only Mode A/C interrogations returning assigned ATC ident code (2157), and pressure altitude.

  615. Paul Smithson says:

    @Don. Why do you think Mode A/C and not S? I don’t see how that can be true. The ref below specifies standard SSR range resolution RMS of 230 metres, but Mode S rms 7 metres. You would need the latter to be able to see consistent speed and heading changes <1 degree over legs that are only a few seconds each.
    https://en.wikipedia.org/wiki/Secondary_surveillance_radar

  616. TBill says:

    @Don Thompson
    Confirming your “4” confidence rating, the reported flight simulator studies around IGARI could probably not be conducted without military radar data. Although presumably another country might have suppplied the radar data (if we are to postulate Malaysia was down for the night).

  617. DennisW says:

    @Don

    This Safety Investigation Report sets out that the Malaysian authorities, ATC personnell and TUDM ADOC staff, were aware that 9M-MRO had been diverted, back, towards Penang during the time the aircraft was still airborne. It is a scandal that the Razak/Hishammuddin regime permitted neighbours to expend resources in futile search operations over the South China Sea.

    I believe the radar was active, but I am not sure anyone was looking at it in real time. Another possibility is that the regime was aware of the Penang trajectory, and decided to bury it until the Inmarsat data surfaced and forced their hand. There really is no other way to explain the South China Sea activity.

  618. Don Thompson says:

    @Paul

    Per DCA’s AIP ENR 1.6 RADAR SERVICES AND PROCEDURES

    Genting Highlands, KLIA, and Kota Bharu implement MSSR, monopulse SSR. Your reference states 13m RMS accuracy for MSSR.

    Mode S extended squitter communicates aircraft lat-long derived from the aircraft navigation system.

  619. Don Thompson says:

    @Paul,

    Should’ve qualified the last sentence; Mode S, if datalink working is exploited, can communicate onboard navigation derived position and altitude. My understanding of ATC ops during the night of 9M-MRO’s final departure is that Mode-S datalink was not exploited.

  620. Paul Smithson says:

    @Don. Thanks for the response. I don’t know enough about radars to know what’s monopulse and what’s “standard SSR”. From the reading I had understood that with Mode A its ID only, Mode C its pressure altitude and the rest is down to radar reflection and attendant azimuth/range resolution limitations. Do “all” SSRs use some sort of pulse from the aircraft when it is painted/interrogated? Even so, I’m still amazed that the spatial resolution can as good as is implied by the fine-detail heading/track resolution of the data provided. We know that the aircraft was still sending ADSB, right? Is that not the same data that gets squirted at Mode S?

    Other than demonstrating my limited understanding of radar, you still haven’t explained why you are convinced that the positions didn’t come from Mode S. What leads you to believe that??

Leave a Reply