Simulator path (red) and recent MH150 flights (colored circles). (Click on image to enlarge.)
Introduction
Within weeks after MH370’s disappearance on March 8, 2014, news stories revealed that Captain Zaharie Shah deleted data from his home flight simulator in the weeks prior to the disappearance, and FBI investigators at Quantico, Virginia, were assisting in recovering the data. Months later, new reports surfaced stating that the captain had used his home simulator to practice flights to the Southern Indian Ocean, and then deleted those flights from his home computer. Despite these reports, there was no official confirmation from Malaysia that suspicious data was found on the captain’s home computer.
We now know that the recovered data sets from the captain’s computer were included in a report compiled by the Royal Malaysian Police (RMP). The report was eventually obtained by several French media organizations, and large portions of the report were later made public.
On August 14, 2016, the MH370 Independent Group (IG) released a preliminary assessment of the simulator data contained in the RMP report. The data sets were created by Microsoft Flight Simulator (MSFS) software, and are in the form of fragments of “flight files”, which a user may create during a simulation session to record the state of the session for future reference or to resume the session at a future time. The particular flight files of interest were among hundreds of others found on the computer on several drives; however, the files of interest were deleted and recovered by investigators from a “volume shadow” on a single drive that was found disconnected from the computer. The grouping of the files in the volume shadow makes the set unique among the all flight files that were found.
The recovered flight files include flight and navigation parameters that are “snapshots in time”, and are associated with six unique coordinates. If these coordinates were all from a single simulation, it suggests that a user simulated a flight of a B777-200LR aircraft with a departure from Kuala Lumpur International Airport (KLIA), a climb and cruise over the Malacca Strait and then the Andaman Sea, followed by a turn to the south, and a termination from fuel exhaustion in the Southern Indian Ocean (SIO) near 45S 104E. The IG found that if a great circle path that connects the final points is extended past the final point, the great circle would cross McMurdo Station, Antarctica. (McMurdo is the largest and most populated research station in Antarctica.) This raises the possibility that McMurdo was used as a final waypoint for navigation with the expectation that fuel would be exhausted in the SIO, well before reaching Antarctica.
The simulator data was found on what the Malaysian investigators refer to as drive MK25, where Microsoft Flight Simulator (MSFS) 2004, also known as FS9, was installed. (The main drive of the computer is referred to as MK26, where Microsoft Flight Simulator X, also known as FSX, was installed.)
The fragments of the flight files were found in a volume shadow dated February 3, 2014. A volume shadow is periodically created by the Windows operating system so that files can be restored to a particular date. This means that the file fragments found in the volume shadow were created on or before February 3, and then over-written on or after February 3.
Of the eight total flight file fragments that were recovered, three were for an aircraft on the ground at Kuala Lumpur International Airport (KLIA), and five were for an aircraft flying. Based on fuel levels and other indicators, the data sets in which the aircraft is flying can be related to one of the data sets for the aircraft at KLIA, i.e., six data sets appeared to be related to a single flight. The remaining two data sets at KLIA show fuel levels that are not consistent with the other data sets. These fuel levels may reflect intermediate values before the final takeoff fuel levels were selected.
There has been much discussion as to whether the six data sets are from a single simulation session, and whether the simulator was functioning normally when the data sets were saved. On November 29, 2016, I co-authored a paper with MSFS expert Yves Guillaume that investigated these and other aspects. To replicate the software used by the captain, Yves and I independently installed FS9 with the B777-LR model upgrade from Phoenix Simulation Software (PSS). After substantial testing, we discovered that:
- The data points in the Andaman Sea share some of the same unique values as the data points in the SIO, suggesting the flight files came from the same simulated flight.
- The parameters related to fuel and flight dynamics show that the position of the aircraft was manually changed along the flight path.
- The data files were manually created and saved after certain parameters were manually changed.
- The simulator appears to have been fully functional, and the rates of climb and other flight parameters after fuel exhaustion can be explained and repeated.
After this paper was published, I created a simulation that was initialized using the recovered values from the data set just after fuel exhaustion. I was able to show that it was possible to manually fly the plane for about 27 seconds and achieve the position, altitude, heading, pitch, and bank that were close to the values contained in the subsequent, final data set. This was further confirmation that the captain’s simulator was properly functioning when the data sets were created. Here’s a video of this short flight.
With the recent release of the ATSB’s report on the operational search, we have learned more about the simulator data found on the captain’s computer. Now three years and seven months after the disappearance, for the first time, the existence of the simulator data has been officially acknowledged in a public document. Although the ATSB does not speculate about whether the captain was responsible for the diversion, the report does reveal that in April 2014, the ATSB considered the simulator data as relevant evidence in defining the search area.
Date and Duration of the Simulator Session
From the ATSB’s report, we learned that on April 19, 2014, the Australian Federal Police provided the ATSB with the recovered simulator data. The report states that The simulator data was a partial reconstruction of a flight simulator session from 2 February 2014. Based on the February 3 date of the volume shadow, we already knew that the simulation was created on or before February 3. However, I was curious to know how the ATSB determined that the exact date of the simulation session was February 2 since the information about date and time was not included in the file fragments that were provided in the RMP report.
After some email exchanges with the ATSB, I was surprised to learn that the ATSB has additional data values from the recovered flight files that were omitted from the data sets that were provided in the RMP report. In particular, there is a section of the flight files in which the date and time of the simulation session are stored. These data values tie the date of the simulation to February 2, 2014. Also, the time values show 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. This confirms that the position and fuel levels were modified during the simulation, just as Yves Guillaume and I had concluded in our paper.
The newly released information regarding the date and duration of the simulator session means that it is almost certain that the recovered data sets were from a single flight session.
However, as the signed date of this part of the RMP report is May 15, 2014, it remains a mystery as to why the Malaysians omitted important data values that were already available to the ATSB on April 19, 2014.
Simulated Flight Path
Using our previous notation, we refer to the identification of each data set by its latitude, i.e., as 2N, 3N, 5N, 10N, 45S1, and 45S2. The position of the first four coordinates are shown above in the figure, where the position for data set 2N is Runway 32R at KLIA.
After takeoff, the next coordinate is 3N, which is in the Malacca Strait on airway R467 between waypoints AGOSA and GUNIP. The altitude for 3N is 23,247 ft.
The next is coordinate 5N, which is between GUNIP and TASEK near airway B466. The exact position suggests the aircraft has already left airway B466; the track, which can be calculated from the components of the velocity presented in the 5N data set, is 317 deg, which is exactly towards waypoint VAMPI on N571. (The value for the heading is 315 deg, which differs slightly from the track due to wind.) The altitude at coordinate 5N is 32,246 ft.
The final point for the northwest portion of the path is coordinate 10N, at an altitude of 40,003 ft, and to the west of the Andaman and Nicobar Islands. The plane is at a heading of 256 deg, and is banked at 20 deg and turning left towards the south. The coordinates for 10N place it near airway N877 between waypoints LAGOG and DOTEN. The position suggests that the aircraft left the airway about 30 NM before DOTEN.
The final two coordinates, 45S1 and 45S2, are in the SIO near 45S, 104E, and are separated by a distance of about 2.5 NM. Despite the short distance, the altitudes at 45S1 and 45S2 are 37,561 ft and 4,000 ft, respectively. The data set for 45S2 shows unmistakable evidence that the altitude was manually changed from 37,654 ft to 4,000 ft just before the data set was saved, which is consistent with the large change in altitude over a short distance.
The track of 5N towards VAMPI on N571, and the proximity of 10N to airway N877, together suggest that in the simulation session, after taking off and flying to GUNIP, the aircraft joined N571 and followed the waypoints VAMPI, MEKAR, NILAM, IGOGU, LAGOG, and DOTEN. The great circle alignment of 10N, 45S1, and Pegasus Field (NZPG) at McMurdo Station, Antarctica, means the session might have simulated a great circle path towards McMurdo Station until fuel exhaustion was reached in the SIO. (Aircraft that are flying in LNAV mode follow great circle paths between waypoints.)
Similarities to Flight MH150
We know that the recovered simulator coordinates describe a flight departing KLIA that is not consistent with a flight to Beijing: the initial path of the simulation is to the northwest, flying over the Malacca Strait, and the initial path to Beijing is to the northeast, and flies above the South China Sea. I searched for whether there might be a scheduled flight that better matched the simulation coordinates 2N, 3N, 5N, and 10N. In particular, I searched for scheduled flights that had a departure that followed a route from KLIA to GUNIP, and then to either TASEK or VAMPI, and would require a fuel loading similar to the 68,424 kg that is inferred from data set 2N.
After considering the list of Malaysia Air’s departing flights from Kuala Lumpur, I discovered that MH150, from Kuala Lumpur to Jeddah, Saudi Arabia, met the fuel and path criteria. Next, I checked whether the captain had flown any MH150 flights in that time period. I was surprised to learn that he commanded MH150 on February 4, 2014, which appeared to be very close to the date that the simulation session was created. In November 2016, I privately communicated my findings to other independent investigators, although I thought it was still too speculative and controversial to present in a public forum. Months later, on April 30, 2017, I noted in a blog comment:
The flight might have been a simulation of a flight from KLIA to Jeddah with a diversion, as the fuel loading and SID are consistent with that. Also, Zaharie Shah was scheduled to captain MH150 to Jeddah on Feb 4, 2014. The Shadow Copy Set containing the deleted flight file fragments was dated Feb 3, 2014.
Now that we know that the exact date of the simulation session was February 2, 2014, the link between the simulation session and MH150 is less speculative. The authors of the recent ATSB report had similar thoughts:
On the day the simulation was conducted the PIC was on a rostered day of leave. The following day the PIC was rostered to fly from Kuala Lumpur to Denpasar, Bali and return the same day. On 4 February 2014 the PIC was rostered to fly from Kuala Lumpur to Jeddah, Saudi Arabia. The first three data points recovered from the simulator were consistent with the route from Kuala Lumpur to Jeddah. In the weeks between the Jeddah flight and the accident flight the PIC was rostered to fly return flights from Kuala Lumpur to Denpasar, Beijing, Melbourne and then Denpasar again.
I was curious to see just how close the path suggested by the simulator coordinates resembled previous paths taken by MH150. For the comparison, I used the ADS-B data available from FlightRadar24 for eight consecutive flights between Aug 17, 2017 and Sep 21, 2017. (The available data for all the flights end around waypoint LAGOG, probably due to the range limitations of the receivers in the vicinity.) The MH150 flight paths are plotted above in the figure.
For seven of the eight flights, the aircraft departed KLIA and flew over waypoint GUNIP. For the seven flights that flew over GUNIP, six followed airway N571 towards VAMPI and continued to follow N571 towards LAGOG. One of the six that reached GUNIP, one flew towards TASEK on airway B466, but rejoined N571 near IGOGU. (Some of the flights show deviations from airways, which are probably weather related.). The similarity between typical routes taken by MH150 and the simulator data is evident.
We can only speculate as to why, just two days before he commanded MH150 from Kuala Lumpur to Jeddah, and five weeks before the disappearance of MH370, the captain used his home computer to simulate a flight of MH150 that was diverted to the SIO.
In Part 2, we further investigate the simulator data and how it may give us clues about MH370.
Updates on October 13 and 15, 2017. In the figure, the dates in the legend were corrected so that all refer to year 2017. Also, the label on the airway between LAGOG and DOTEN was corrected to be N877, and the label of the airway between GUNIP and TASEK was corrected to be B466. Thank you to Twitter handle @BKKDiet and blog contributors @Mick Gilbert and @TBill for finding these errors.
@Victor
Awesome forensics. Can you estimate the latitude of MH370 at 19:41 that corresponds to your best guess fight path?
@Don Thompson
The Hilton you stayed at is no more.
http://www.express.co.uk/news/world/864531/california-fires-damage-update-hilton-santa-rosa-burnt-to-ground-sonoma-napa-wine-country
@DennisW: Thanks for your comment.
The quick answer to your question is one of the candidate paths is the McMurdo path that begins around 8.5N – 9.1N at 19:41. More details on this later.
Excellent post as always! Many thanks Victor.
My best fit NZPG scenario is using LRC mode.
I get a very good match to a 19:41 UTC location of 9.00°N 92.97°E, initial track of 168.4099°T and starting at 0.815 Mach, reducing to 0.780 Mach by 00:11 UTC.
The 00:19 UTC End Point location is 26.84°S 100.69°E.
This is very close to the constant mach 0.798 scenario in the paper we wrote dated 24th August 2016.
@Richard: Thank you for the kind words. As you know, I get essentially the same results for LRC speed (26.87S latitude) for a great circle path aligned with NZPG. The BTO fit for LRC seems to be substantially better than for constant Mach number.
@Victor
The quick answer to your question is one of the candidate paths is the McMurdo path that begins around 8.5N – 9.1N at 19:41. More details on this later.
Thx. I’ll be patient. I am having difficulty reconciling that 19:41 latitude with the graphic at the top of this post.
@DennisW: The MH370 path to McMurdo at 19:41 is much closer to VOCX than DOTEN. I’ll explain more in the next article.
@Andrew,
Thanks for your comments and references.
Re your comment: “No sense to you perhaps, but a whole lot of sense to those who design such systems. The point of error correction is to improve the accuracy of the angular & linear acceleration inputs and in turn improve the accuracy of the outputs.”
My point is that it does not make sense to use the GPS data to correct accelerations only. If the accelerations are corrected, then velocity and position would also be affected in some way. Otherwise such a correction would result in inconsistency. My current understanding is that the GSP data affects output ADIRU’s position through refinement of the angular and linear accelerations, correct?
Re: “Sorry, I can’t find the diagram. If it’s related to the SAARU, then I suspect the SAARU checks the validity of the data it receives from the ADIRU. You need to ask an electronics engineer for the answer to that one, not a simple pilot like me.”
You certainly have a lot more technical knowledge than a simple pilot, and I appreciate your comments a lot. I extracted the diagram I was referring from the AMM:
https://www.dropbox.com/s/7j0ve2n768v4ev7/CAL_Training_Manual_B777-200_WB371_p1865.pdf?dl=0
quick unrelated but interesting side-note:
“Metal Scandal Triggers Safety Probes Into Planes”
http://archive.is/rVoCc
@VictorI,
You said: “@DrB: When you balance the savings during the descent to FL200 (which would be near the optimal holding altitude) versus the additional fuel required for the climb back to FL350, for instance, what is the net fuel consumption compared to if the plane had been at cruise speed and altitude during that same time interval (forgetting the actual time at hold)?”
Using the parameter values now in FMV5.3, there is only a slight net fuel cost of a descent from FL340 to HOLDING at FL200 followed immediately by a climb back to FL340. It is approximately 150 kg or 2 minutes of cruise fuel. That fuel cost must be more than offset by a fuel savings at reduced speed when at FL200 if you want to increase the endurance of the flight. HOLDING in a racetrack at FL200 burns fuel at only an 8% reduction compared to no-turns MRC at FL340, so it takes a while to accumulate enough savings to make the descent/ascent maneuver worthwhile. If there are no turns when in HOLDING, the savings is 5% larger (~13%).
@Paul Onions,
You said: “One observation: At 1706:43 the aircraft was stable in cruise at FL350/M0.821/GWT218t. The Cost Index 52 speed according to your data is about M0.836. So either your data needs revising, or the aircraft was flying at a different Cost Index, or the crew had selected a Fixed Mach Number cruise.”
Regarding the 17:07 speed, the aircraft had not been long at FL350 and perhaps the airspeed might not have yet quite reached its nominal value. Yes, it is possible another speed mode, such as MRC, was used instead of ECON 52, but I can’t think of any reason they would not have used the Flight Plan Cost Index then. As I said: “I am also working to refine the CI = 52 speed schedule using the MH370 Flight Plan and the MH371 ACARS data. I think my current equation slightly overestimates the speed at low weights, and I may update that later.” I trust the MRC and ECON 52 speeds derived for the conditions of the Aero Figure 1 fuel mileage plot, but for different conditions both the MRC and CI 52 speeds are quite sensitive to errors in the fuel model even down to a fraction of 1% fuel flow. I will likely modify my ECON 52 speed schedule once I have completed my comparisons with the available data. These comparisons are somewhat complicated by our lack of equations for the headwind/tailwind Mach corrections applied by the FMC.
I am not aware of any reports of turbulent conditions near IGARI. I can say that the last fuel report seems quite consistent with the previous fuel reports the and the observed fuel flow ratio between engines.
@Victor – Interesting that the September 21 track has an offset similar to your prediction for MH 370
@Ge Rijn,
Re your post October 11, 2017 at 7:49 am.
As I see you are misunderstanding the numerical drift studies, and also making some wrong conclusions basing on the distribution of found fragments and their discovery dates.
Re: “What I really like to see is a same kind of animation you produced but than based on the amount of pieces found so far. How many virtual drifters you need to produce this outcome? 10.000? 5000? 1000? And where would they land from any given latitude along the 7th arc?”
Unfortunately this is impossible. To simulate drift of the individual fragments one would need to know detailed flotation characteristics of all the fragments, their decay/sinking properties, and accurate forcing (flow, winds). The whole idea of the numerical drift studies is to analyze statistical characteristics. The more particles (drifters) are used in simulations, the more accurate is statistical result. Jansen used 5,000 particles per an ensemble; Pattiaratchi and myself – 50,000; Durgadoo – 2,000,000 if I recall correctly.
Re: “I think we have a quite accurate window of dates when debris started arriving primary after the crash and when it probably ended due to the barnacles still attached on the flaperon when found and on the ‘RR-piece’.”
Absolutely wrong. We don’t know time window and locations. Your mistakes in this regard are:
1. The reporting dates are not arrival dates. For example, the flaperon was reportedly half-covered by sand when it was found. Given several storms during a preceding month, I think it is very likely that the flaperon arrived 2 weeks before it was discovered. What about the latest fragments found by Blaine? Perhaps they reached Madagascar one year earlier? Who knows. Late arrival would be a problem, but not early arrival. In terms of the modeling, a good indicator is the arrival of a group of particles at a given time. In this regard, I see a spike in the arrival to the proximities of the Mossel Bay in March – May 2016 predicted by my model, subject to the origin. I think this is a very good agreement with the discovery time of RR fragment.
2. For some mysterious reason everyone considers only 9M-MRO fragments, but not its content. Do you really believe that only 30 fragments of the aircraft were washed ashore, but not various plastic bottles, shoes, suit-cases, etc.? Where is all this stuff in your opinion? Do you really believe that a plastic bottle washed ashore in SA back in the end of 2014 or early 2015 would attract any attention?
3. A lot of objects associated with MH370 are predicted in Madagascar. In my estimation about 50 more fragments of 9M-MRO, which is consistent with overall Dennis’ estimates. How can you be sure there is nothing if nobody searched there, and how would you estimate the arrival times of these fragments if/when they are found?
Re: “In my view all much to early and not reflecting the reality of what has been actualy found and where.”
On contrary. The agreement is surprisingly good. This is because you need to look at statistical characteristics rather than individual particles.
Re: “And most of all imo not reflecting the original debris field which must have been very much smaller than 50.000 regarding the reality of pieces found to date.”
The number of particles in an ensemble is irrelevant to the number of floating fragments after the crash.
Re: “Nothing still found/confirmed from WA”
The keywords are “found” and “confirmed”. As I said many times, I believe the towelette could be from MH370 if the crash site is at 28-30S. This is because the modeled arrival time is the early July 2014. Can there be more objects, say at the Windy Harbour? Certainly yes. Would they be fragments of 9M-MRO? Would they be found and reported?
I think you may also consider Dennis’ and Victor’s comments in this regard, which I trust are in-line with mine.
@DrB: What were the times and speeds that you used for the descent and climb?
@Oleksandr said: You [Andrew] certainly have a lot more technical knowledge than a simple pilot, and I appreciate your comments a lot.
@Andrew: We are extremely fortunate to have you contribute here. I appreciate not only the technical information and knowledge you possess, but also the way you attack problems.
@Lauren H: The offset manoeuver at MEKAR and the continued offset past NILAM also caught my attention, albeit the offset is greater than 15 NM. I asked Andrew about it, and he thought it was probably a deviation due to weather.
@Andrew,
@ALSM,
@Don Thompson,
@Dennis,
@Victor,
Citations from AMM, pdf p. 3227-3228:
“The SATCOM system also connects with the airplane information management system (AIMS) for these functions:
– Data communication management function (DCMF). The DCMF supplies data communications control for airplane air/ground datalink over the SATCOM system.
– Data conversion gateway function (DCGF). The DCGF changes inertial reference system data into ARINC 429 data for high gain antenna (HGA) beam steering. The DCGF also supplies international civil aviation organization (ICAO) addresses.”
“The low gain antenna system receives and transmits data signals. This system operates only when a high gain antenna system fails or class A high power amplifier fails.”
Citation from p.3240:
“The low gain antenna system is for data only”.
1. If the INS data invalid, would the LGA be engaged instead of the HGA?
2. Are ground speed, GPS coordinates and track angle (required for Doppler compensation) supplied via DCMF or DCGF?
3. If the LGA does not require INS data, why Andrew was unable to use SATCOM after the ADIRU shutdown on the ground?
4. May the fact that the LGA is designed for the data (not voice) explain failed SATCOM calls 18:40 and 23:15?
5. Is it possible to establish whether LGA or HGA was used? I did not find any discussions on this topic in the formal reports.
@Oleksandr: It’s explained in the 2015 FI. The class of service for the log-ons at 18:25 and 00:19 was Class 3 (voice and data), which requires the HGA.
Oleksandr,
1) On the ground, without the ADIRS aligned, the log shows that 9M-MRO’s LGA was employed between power up at 12:50 & flight prep at around 15:55.
2) DCMF is the Data Comms Mgmt Function, DCGF is the Data Conversion Gateway Function. The SDU requires an ‘IRS’ data connection and a, separate, comms data connection. The ‘IRS’ data is communicated via DCGF, the comms data via DCMF.
3) I’ll leave that for Andrew to answer. Do note that not every aircraft has both LGA & HGA, many types have one or the other.
4) LGA cannot support the data rates required for the C-channel (circuit switched). A Log-On for LGA will reflect that, no C-channel possible. The SATVOICE C-channels were established, signalling progressed for a minute without answer, & the ground cleared the calls.
5) The 18:25 Log On was established using the HGA, the SU exchanges between the AES & GES emphatically demonstrates that. The 00:19 Log On also.
@Victor,
Thanks. How did they find out that
“The SATCOM link becomes available (for both voice and data – Class 3)”
for the logons 18:25 and 00:19?
@Oleksandr:
As noted many times here, all the communications during the MH370 flight were conducted via the HGA. See unredacted Inmarsat GES log.
Also, as noted many times, a/c position, track angle and ground speed are required by the SDU for Doppler correction, regardless of the antenna used.
Thus, wrt your question: ” If the INS data invalid, would the LGA be engaged instead of the HGA?”, the answer is NO, because no transmissions are allowed if there is no 429 nav data to correct for Doppler.
Thus, wrt your question: “May the fact that the LGA is designed for the data (not voice) explain failed SATCOM calls 18:40 and 23:15?”, NO. Those call attempts were via the HGA (as are all voice calls). The AES did not fail to receive the call request. It’s just that no one answered the calls. That is why we have BFO data from those attempts.
@Oleksandr: The class of service is contained in one of the SU fields at log-on. Don knows the details.
@Don,
Thanks. Re:
>1. On the ground, without the ADIRS aligned, the log shows that 9M-MRO’s LGA was employed between power up at 12:50 & flight prep at around 15:55.
So, the inability to use SATCOM on the ground experienced by Andrew has nothing to do with the ADIRU, right?
>2. SDU requires an ‘IRS’ data connection and a, separate, comms data connection. The ‘IRS’ data is communicated via DCGF, the comms data via DCMF.
AMM states that the IRS data is required only for the HGA. LGA does not seem to require INS data according to AMM. But transmissions through the LGA still require data for the Doppler correction, right? If yes, does this indicate that the ground speed, track angle and coordinates are not supplied via the INS channel?
>5. “the SU exchanges between the AES & GES emphatically demonstrates that”
Just my curiosity: how does it demonstrate this?
@ALSM,
Re: “Thus, wrt your question: ” If the INS data invalid, would the LGA be engaged instead of the HGA?”, the answer is NO, because no transmissions are allowed if there is no 429 nav data to correct for Doppler.”
Don’s comment above:
“On the ground, without the ADIRS aligned, the log shows that 9M-MRO’s LGA was employed between power up at 12:50 & flight prep at around 15:55.”
Without ADIRS aligned, there would be no valid INS data. Assuming both your statements are correct, how could the LGA be employed?
Oleksandr,
To be clear, the IRS data is transmitted to the SDU regardless of the antenna selected by the SDU. Data words will be transferred, even if padded w ‘NCD’.
The Log On Request SU defines data fields to specify the class of operation, & so on, required by the AES for the Log On. I have to pass on explaining that down to bits & fields, it’s late.
Oleksandr,
I specifically noted “on the ground”, the SDU gets a discrete for WoW (weight on wheels).
Think about it: boot up, NCD on IRS data, WoW, got an LGA. 600bps better than none.
BTW, I try to use AES & SDU terms appropriately. The AES for the entire SATCOM system, and SDU if appropriate for functions specific to it.
@Victor
I think that if we are ever going to find MH370, given the structure of the BTO arcs, there is one point, that we have to nail down fairly precisely, and that, is where, and when it crossed the equator.
Could you post all details for your flight path’s crossing of Latitude=0, specifically, Time, Longitude, Altitude, Speed, Heading(s), Track(s), Gross Weight, and Fuel Remaining.
@Don,
Re: “To be clear, the IRS data is transmitted to the SDU regardless of the antenna selected by the SDU. ”
AMM states that IRS data is required for HGA.
“The DCGF changes inertial reference system data into ARINC 429 data for high gain antenna (HGA) beam steering.” (p. 3228).
“The HGA receives beam steering information from the beam steering unit (BSU).” (p.3239).
“The BSU receives direction signals from the SDU” (p. 3243)
“The low gain antenna operates only if the high gain antenna (HGA) system fails.” (p.3248).
Note that the sets of parameters required for Doppler compensation and beam steering are different. Bearing in mind Andrew’s experience, the HGA cannot operate when the ADIRU is off. Bearing in mind your comment and citations from AMM, it looks like the LGA still can operate when the ADIRU is off.
@Victor
As always, a very thorough write-up, Victor. The question that immediately springs to mind is, do you now have the timestamps for the data points?
@VictorI,
You said: “@DrB: What were the times and speeds that you used for the descent and climb?”
In the case where there is just a descent followed immediately by an ascent, I used 17:07/M0.835/FL340 for Leg #1. For Leg#2, I used 18:29/HOLDING/FL200. FL200 is reached at 18:37. For Leg #3, I used 18:37/LRC/FL340. FL340 is reached at 19:05.
I am using 2,000 fpm ROD for a significant descent to a slower HOLDING speed (and also using idle fuel flow). For the ascent I use 500 fpm ROC. In my model the fuel cost of the climb does not depend much on the climb rate, since the excess fuel flow during the climb is scaled linearly with ROC. Thus, the duration of the ascent depends on the ROC, but the “extra” fuel quantity consumed to ascend (relative to the cruise fuel) is essentially independent of the ROC used.
@Don,
Re: “I specifically noted “on the ground”, the SDU gets a discrete for WoW (weight on wheels).”
1. Previously I have already asked whether the SDU knows/assumes that the plane is “on the ground”. You may search for the answers I received. Anyway…
2. Through what channel/function does the SDU receive this info? INS, GP1, …?
@Andrew,
@Gysbreght,
@VictorI,
@ALSM,
When you have the time, would you mind having a look at my end-of-flight model? In the second worksheet, the one called “MEFE Model”, at the top right there is a plot of the critical parameters during the end of flight period. If needed, I can also explain in words the sequence of events I am using that affect airspeed, fuel flow, and altitude.
@Don,
P.S.
In case the SDU gets a discrete for WoW when on the ground, your comment is still inconsistent with ALSM’s comment, who stated that “no transmissions are allowed if there is no 429 nav data to correct for Doppler”.
Oleksandr:
WoW may be deemed equivalent to the required nav data for Doppler corrections. WoW~0 kts, right? If the plane was in the air, the same nav variables (a/c position, track angle and ground speed) would be required for the HGA or the LGA. Although WoW may be sufficient for a Doppler correction when the plane is on the ground, it is not sufficient for HGA beam steering, thus the reversion to the LGA prior to INS alignment.
@Mick Gilbert: No, I don’t have the time values from the recovered data sets.
@ALSM,
“WoW may be deemed equivalent to the required nav data for Doppler corrections. WoW~0 kts, right?”
I think no because of the following reasons (at least):
1. Logically: the plane can be moving on the ground (at speed <30 knots with the ADIRU off), or at higher speeds during takeoff/landing with the ADIRU on.
2. Formally: there is a formal list of the required data you cited earlier. Where is WoW? How does it reach the SDU (channels are listed at AMM pdf p. 3263).
3. What if WoW sensor provides wrong reading? Will it invalidate the ADIRU data?
Btw, when I suggested an explanation of the abnormal BFOs as a result of the absent Doppler compensation, you have discarded it as impossible because there would be no transmissions. If WoW comes into the equation, such an explanation becomes possible again, right?
@Victor
Thank you for the post.
I have also recently been re-studying the simulator data with FS9, and also checking FR24 flight paths.
Here are some questions I see:
A) How could a flight path go exactly thru N10 at the big turn to McMurdo?
I see two main solutions:
Solution-1: 8x speed (wide) Turn to NZPG is at DOTEN/-46 from LAGOG
Solution-2: 1x speed normal Turn to NZPG at BIKEN/-165 DOTEN/-36
For Solution-2 the turn to DOTEN is 227-165= 62 nM past LAGOG on the way to BIKEN. Solution-2 requires either that (a) Z manually moved the aircraft beyond LAGOG on N571 and cancelled the LAGOG waypoint, or interestingly (b) an alternate path DUBTA to DOTEN could be the actual path. DUBTA to DOTEN represents the approx. angle of attack needed for 20 deg turn to go through N10, except for the 8x speed Solution-1.
Note that in either case above the trajectory to NZPG is offset slightly west from N10. with something like 10N (Solution-1) to 10.5N (Solution-2), 90E as the focal point for the top of the path. This in turn implies the final waypoint could range from NOBEY to NZIR to NZPG to best match 45S1. I still cannot rule out 180S magnetic instead of McMurdo LNAV, which also fairly closely matches 45S1 using 2005 magnetic tables and FS9 Fair Weather winds.
So the above observations are why I started looking at FR24 to see how the Andamans airspace is used.
B) Flight Radar24
I saw that late night flights to DOTEN often short cut IGOGU to DOTEN, cutting through the VOCX controlled space. They usually do this at FL300. Andrew has mentioned FL280-FL300 flight rule is often used on N877 to DOTEN.
Therefore I now theorize (Option-1) MH370 might have descended to FL280-FL300 toward DOTEN from IGOGU (per 1840 BFO). Perhaps there was then an IGARI-style U-turn with climb to FL350-FL400 to return to Arc2 looking like a higher altitude incoming flight. The U-turn may or may not have gone as far out DOTEN.
Option-2 I have not explored much, but alternatively MH370’s 1840 descent was towards DUBTA from SAMAK, cutting underneath EK343 which normally takes N571 all the way out.
PS- minor point but the apparent airway diversion to N571 VAMPI from TASEK seems to be a good round number TASEK/-25 if my memory serves.
Oleksandr:
I give up. I get the impression you are more interested in diverting the discussion down multiple rabbit holes than actually understanding how the system works. (“What if WoW sensor provides wrong reading? Will it invalidate the ADIRU data?” Really?!! What if the earth flipped over on its axis? Let’s track that down too.)
If I am wrong, please remind me: What is your point? What are you trying to understand?
@TBill: I’m not sure what the descent to FL280 might accomplish. Any primary radar sources wouldn’t easily detect the altitude.
@Victor
I can only presume descent to FL280-FL300 outbound from IGOGU looks like a regular flight at that hour of the night, in case somebody unexpectly has a radar or eyeball working (eg; Singapore, VOCX or other) or to other aircraft.
Likewise inbound return at FL400 ( or as low as FL350) just like the simulator case shows, looks to the eye or radar like a normal inbound flight on N877 or N571.
@Victor
(a) “I’m not sure what the descent to FL280 might accomplish.”
(b) “Any primary radar sources wouldn’t easily detect the altitude.”
(a) From a planning perspective, if he remained high, “after leaving airways post FMT) he was entering UPR Airspace (FL245 to FL460), first (Medan West UTA) then (Jakarta UTA), then, after crossing the FIR boundary into Melbourne YMMM FIR/OCA, he would be in RVSM Airspace, all of which is FL290 to FL410 (per Jeppesen Indian Ocean Plotting Chart 16th Edition 21 Mar 14).
He knows all other flights would be complying with “the rules”.
He may have wanted to ensure that there was no possibility of an unhappy chance meeting with any other aircraft, thus fly below the floors of the upper airspaces.
From an operational perspective, given the stormy weather on the night off the west coast of Sumartra, the need to fly low would be even greater.
He was unable to “climb above”, so his only safe option, would be to descend below all other possible traffic.
Indeed, the storms could have forced him to go even lower than he had planned, or really wanted to, which would have given him more fuel issues to deal with as well, which may have forced subsequent change(s) to whatever the original plan was.
(b) True enough, but “loss of paint” at bearing X at range Y can quite easily give you a pretty good approximation of height.
Scanned a bit of the chart:
https://www.dropbox.com/s/h7hlvjva41mkxnx/Upper-01.PNG?dl=0
To see what I mean, revisit Iin Vasquez’s animations of the weather satellite images at http://www.weathergraphics.com/malaysia/
@Gysbreght
@Oleksandr
@Victor
Thank you for your kind words, I enjoy contributing to the discussion. Let’s hope that the collective thinking of so many individuals helps solve this mystery.
@Oleksandr
Re: ”My point is that it does not make sense to use the GPS data to correct accelerations only. If the accelerations are corrected, then velocity and position would also be affected in some way. Otherwise such a correction would result in inconsistency.”
I’m not sure what the problem is, so I’ll start at the beginning. The ADIRU, like any inertial system, integrates acceleration to compute velocity and distance. Small errors in the measured acceleration are magnified by the integration process, causing much larger errors in the velocity and distance calculations, and consequently the calculated position. Those errors accumulate with time and cause the calculated position to ‘drift’ further and further from the actual position.
The aim of error correction is to improve the accuracy of the acceleration inputs that are used to calculate velocity and distance. The ADIRU does that by comparing the GPS data with the data the ADIRU has calculated. The ADIRU then determines an error correction that is used to ‘correct’ the acceleration inputs and improve their accuracy. That process in turn improves the accuracy of the calculated velocity, distance and position, and reduces drift.
Re: ”My current understanding is that the GSP [sic] data affects output ADIRU’s position through refinement of the angular and linear accelerations, correct? “
Yes.
Re: SAARU backup navigation
Thanks for the diagram – perhaps I can help with your question after all.
In normal operation, the SAARU ‘tracks’ the ADIRU outputs that it receives. In other words, the SAARU essentially outputs the same data that it receives from the ADIRU (ie the data shown in the input memory buffer). The SAARU monitors the ADIRU data for validity by checking the ‘validity bits’ that are included in each data ‘word’ sent by the ADIRU.
If the SAARU detects invalid data from the ADIRU, the backup navigation logic function enables the SAARU to use and output its own backup navigation data, independently of the ADIRU. The SAARU uses its own data to calculate the backup navigation data when the following conditions are met:
1. Any element of the ADIRU data is NOT valid, AND
2. The SAARU TAS is greater than 80 knots, AND
3. Backup navigation has been enabled for less than 3 minutes.
The last valid magnetic heading data received from the ADIRU is used as the SAARU’s initial magnetic heading. After 3 minutes, the logic function disables the SAARU’s backup navigation until the pilots enter a magnetic heading in the FMC POS INIT page.
Re: ”If the LGA does not require INS data, why Andrew was unable to use SATCOM after the ADIRU shutdown on the ground?”
As Don noted, not all aircraft have both LGA & HGA. Our aircraft do not have an LGA.
@Oleksandr
Re: “Formally: there is a formal list of the required data you cited earlier. Where is WoW? How does it reach the SDU (channels are listed at AMM pdf p. 3263).”
The AIMS data communications management function (DCMF) uses air/ground signals (ie WOW) for communications management. I assume the air/ground data is sent to the SDU by the DCMF, along with other communications data.
Re: “What if WoW sensor provides wrong reading? Will it invalidate the ADIRU data?”
Aircraft systems that use air/ground data typically default to a ‘safe’ condition (ie ‘air’ or ‘ground’) in the event that the data becomes invalid. I don’t know about the SDU specifically, but I assume it would default to ‘air’ mode if the air/ground data is invalid.
@All,
HERE is a plot and my notes on a new MH370 endurance analysis using my latest Fuel Model V5.3.
The acceptable endurance solutions I found are as follows:
1. MRC (or possibly ECON with a low Cost Index ≤ 52) at FL350-400
2. HOLDING or 250 KCAS near FL400
3. HOLDING or 250 KCAS near FL75
4. 1-Hour HOLD at FL21, followed by LRC at FL300-400, both with Air Packs OFF.
#1 and #2 are essentially the same solution. MRC at cruising altitude matches the MEFE time for FL350-400 (with Air Packs ON and no loiter).
#3 is interesting but it may not lead to any route consistent with the BTOs/BFOs because of the low speed.
#4 is also interesting because it is the extended (1-hour) loiter case. I found that it is possible to match the observed endurance using LRC, but only with both the extended loiter and the Air Packs OFF.
@Oleksandr
When I referred to the WoW, I described it as a ‘discrete’. A discrete is not a databus signal, it is a single wire. In fact, the Honeywell/Racal MCS interconnect provides two WoW discrete lines.
@victorl
Thanks for posting this new information!
These findings add a lot more weight to the scenario everyone has feared for since the beginning. I think this neatly explains Malaysia’s reluctant behavior from the get-go. Clearly they panicked, went immediately into damage control mode. Releasing only the minimally necessary, but withholding key details and pressuring search&rescue partners to keep silent.
After all, how emberrasing is that, explaining to the world that one of their prized Captains went ape?
For those not convinced, turn the arguments around. Suppose it was not the PIC, but some other, devasting electrical crisis. The plane then just happens to follow the same, freak divergent route the PIC simulated on his free day a month earlier. What are the odds?
@Don Thompson
Oops – I missed the bit about it being a discrete!
@Oleksandr
On my request about your drift-simulation I tend to think you missed my point.
Ofcourse also I don’t think only ~30 pieces beached. And I don’t exclude WA either. Could well be ~200 pieces (or more) beached and also some in WA that are not found still.
I like to see what kind of amount of simulated drifters to start with would reflect the reality of current amount of found pieces and their locations better.
Starting with 50.000 drifters results in many hunderds of drifters beaching on African shores and islands and many in WA. This does not reflect the current reality of found and estimated amount of beach pieces based on this found amount of pieces within the timeframe of ~3 1/2 years.
I still think it could be usefull to look for a turning point based on the real and (from them) estimated amount of beached pieces to make a better prediction of the initial debris field.
And still imo the barnacles on the flaperon and the RR-piece give an indication of the time window and range in which many of the found pieces arrived. Not exactly ofcourse.
@DrB: Your #4 is the case that I have maintained is possible. I have been assuming a 2.5% improvement with packs off and packs on during the loiter. That said, my guess is we can probably only predict fuel flow within a percent or so.
@Victor Iannello:
Did you do a simulation on your computer that reproduced the fragments of flight files found on Z’s computer, with all the manual changes of flight parameters that were made before saving those files?
@Gysbreght: Yves and I have performed enough testing so that we were confident drawing the conclusions we did. The new information we have gained from the ATSB confirms that our conclusions about changing the position and fuel levels during the flight session are correct.
@Victor
Victor, in your accompanying diagram you show LAGOG – DOTEN as being on the N571. I don’t think that is correct; the N571 runs to the south of DOTEN, LAGOG – DOTEN is on the N877.
Also, can you explain your rationale for showing the simulator path as running GUNIP – 5N – VAMPI rather than GUNIP – 5N – MEKAR? 5N is pretty much bang on the direct track from GUNIP to MEKAR whereas it falls about 10 – 11 nm south of GUNIP – VAMPI (aka N571).
@Ge Rijn
I like to see what kind of amount of simulated drifters to start with would reflect the reality of current amount of found pieces and their locations better.
I do not think there is any way to model what you are asking for.
@Victor
Re: “The new information we have gained from the ATSB confirms that our conclusions about changing position and fuel levels during the flight session are correct.”
Do you have new information from the ATSB beyond what has been published in their final report? There is no reference to the disparities between the fuel levels for the various data points nor is there any attempt to explain them in the final report. There is only one reference to “user input” and that relates to the altitude value for 45S2?
Interestingly, the ATSB touch on the possibility of there being an “intermediate data point not captured” between 10N and 45S1.
@Mick Gilbert: I have no additional information from the ATSB other than what I said in the article. The session lasted an hour of flight time. The only way to explain the fuel levels is manual changes.
You are correct about the N877 airway label between LAGOG and DOTEN. Thank you for noticing. I will fix the figure and the associated text.
The logic for GUNIP-5N-VAMPI is that the track at 5N points directly towards VAMPI, which you can calculate using the velocity components in the data set.
Update on October 13, 2017. In the figure, the dates in the legend were corrected so that all refer to year 2017. Also, the label on the airway between LAGOG and DOTEN was corrected to be N877, and the text changed to reflect this. Thank you to Twitter handle @BKKDiet and blog contributor Mick Gilbert for finding these errors.
@Victor Iannello:
You earlier stated with confidence that the only data on the shadow volume were the recovered flight file fragments. Now you write:
“After some email exchanges with the ATSB, I was surprised to learn that the ATSB has additional data values from the recovered flight files that were omitted from the data sets that were provided in the RMP report. In particular, there is a section of the flight files in which the date and time of the simulation session are stored. These data values tie the date of the simulation to February 2, 2014. Also, the time values show 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. “
My earlier question about the size of the recovered file fragments remained unanswered, so I have to assume the size corresponds to one block of disk space. That means that the new data sets containing the date/time stamps, having the same size, must contain a large amount of new data. For example whether or not the autopilot was used and in which mode. The ATSB Final report also states that with one or two exceptions they had complete flight files of the simulations. Have you asked the ATSB what prevents them from providing access to the complete information they have been sitting on since April 19, 2014 ?
@Gysbreght: You have my words wrong. In the past, based on the words in the RMP report, I said that the eight file fragments were the only flight files recovered. I was responding to claims that there were many flight files found in the Shadow Volume, and the selected fragments were cherry-picked among those other flight files. We now know that there were some more data values that were related to those same flight files. The number of flight files remains as eight, of which six can be related to the same flight. That’s what I’ve said all along.
If you believe that the ATSB has been sitting on a large amount of other relevant data, you can pose those questions to them. I relayed what I know.
Overall the MH370 flight path situation might be summarized as follows:
MH370 Flight Path Possibilities:
(1) MH370 Headed South at 18:40 at IGOGU
(2) MH370 Descended at 18:40 either towards (a) DOTEN or (b) DUBTA
It seems obvious to me descent towards DOTEN is logical compared to what other flights do at that time of night. Not to mention ATSB’s nearly 100% focus on Option #1 came up empty.
Option #2 explains a lot to me: Why would MH370 cut right thru Indonesia’s back yard with FMT at 18:40? Indonesia said they did not see it, but they thought it was picked up by radar in the Andamans. Makes sense that Indonesia is however now cooperating with Malaysia and allowing focus on ghost flight cutting through Indonesian air space at IGOGU. The only other option is to say intentional flight path which is unacceptable to Malaysia. To me this is the likely Halloween skeleton in the closet no one (eg; ATSB) is allowed to disclose.
@DennisW @Oleksandr
On your anwser about drift-modeling:
‘I do not think there is any way to model what you are asking for.’
To me it seems easy enough. F.i. put in 1000 drifters in the model to start with instead of 50.000 and see how many of them arrive on shores after time and within a time frame that represents the amount of found debris (and the amount predicted from them), their finding dates (especially the flaperon and RR-piece) and their locations best.
If this results in too less arrivals within the time frame than scale up the input-amount in the model or scale down if too many arrive.
I think this can result in a more realistic prediction of an initial debris field of floating aircraft debris.
In AF447 ~700 pieces of floating debris were recovered around the crash site. Maybe this could serve as a guide-line.
Ofcourse you can not include all possible bottles, towelletes and personal items in such a model which are beyond any possible identification.
Anyway 50.000 is much too high and 30S gives arriving times on Reunion and Madagaskar way too early imo.
Many drifters in @Oleksandr’s model start to arrive at Reunion already in may 2014 and by the end of july 2014 numerous drifters landed on Madagaskar. By the end of 2014 hunderds of pieces already landed in Mozambique and SA and Madagaskar’s east coast is covered with them by the hunderds also by then.
It’s hard to imagine pieces only got found a year later (flaperon) and 1 1/2 later (RR-piece) according to such a model based on so many drifters.
Impossible I would say.
To be short; 30S is too far north according his model and the amount of drifters in the model is way too high regarding actual reality.
In a way it’s mistifying.
That’s why I suggest an experiment like stated.
I think it could lead to a more realistic initial debris field, starting latitudes and time frames than the one presented in @Oleksandr’s animation.
@Ge Rijn
There is no way to estimate the probability of a beached piece of debris being found. Is it 10%, 1%, 0.1%,… who can say? Most of the shore line in the Indian Ocean area is inaccessible. We are not talking about SoCal beaches here.
The other factor is awareness. How many people would even relate a small piece of debris to the missing MH370? My guess is that if I asked random people in California about it the majority would not know what I was talking about and we have TV, radio, and newspapers (mostly liberal democratic rags) here.
@DennisW
My point is taking actual reality more into account after 3 1/2 years.
Ofcourse we can not compare African shores to Californian beaches.
But on the other hand many more people on those African Indian Ocean shores live directly with the sea, fishing, roaming around, not only sun bathing like in California.
They’ll probably notice something strange much earlier than a random American from California.
And believe me many of them have television and radio and are as much (or as little) aware of what is happening in the world.
@Ge Rijn
My point is taking actual reality more into account after 3 1/2 years.
Your point is looking for evidence of a small debris field to reinforce your opinion of a soft landing.
@DennisW
No. I would be glad to know there was proof of a large debris field.
And also of a (high speed) nose down impact. There isn’t any yet confirmed on both.
And most indications do not support this imo.
The ATSB-FR did not conclude one way or the other also.
Reality till now indicates a rather small debris field and not a (high speed) nose down impact.
You cannot reject the reality of the amount of found pieces after ~3 1/2 years, where and when they were found, their kind of damage and their positions from the plane.
Imo that’s close to denying reality in support of wishfull thinking.
Any conclusive proof of a (high speed) nose down impact and/or a large debris field would be welcomed by me. No one has given any yet.
@VictorI,
You said: “@DrB: Your #4 is the case that I have maintained is possible. I have been assuming a 2.5% improvement with packs off and packs on during the loiter. That said, my guess is we can probably only predict fuel flow within a percent or so.”
That is why I ran it. It is almost the same. My case #4 has the Air Packs OFF for both the 1-hour loiter and the subsequent MRC path southward. If the Air Packs are ON during the loiter and OFF afterwards, the predicted MEFE is at 00:15, and this is still consistent within the errors with the observed MEFE at 00:17.
I am estimating 1% for the 1-sigma error in fuel flow, so on that count we are on the same page. I hope you will soon present a plot showing your best-fit route compared directly to the simulator data points.
What airspeed are you using after the loiter?
@Victor Iannello:
Do you really think you answered my simple question?
@DrB asked: What airspeed are you using after the loiter?
Fuel calcs are based on LRC at FL350. Temperature offset = ISA+10K. FF adjustment for packs off = -2.5%. PDA = 1.5%. Fuel at 19:41 = 27.1 MT.
@Gysbreght
Thank you for your final comment of “yes” at the end of the last thread. I am interested that you feel the hijacker may have been alive in the cockpit at the end of the flight, if I understand correctly.
@ventus45
Thank you for the link to the Vasquez’s MH370 weather, I had lost track of that, and was trying to find it again. re: Arc7 end point, looks like solid cloud cover except E Baker thinks may the pilot stayed north of the cloud line at 22S. re: Arc1-Arc2 region, not sure except to say if you look at Flight Radar24 long enough, you tend to see a lot of funny flight paths apparently due to weather in the region.
@Victor
Simulation questions:
(1) why go off N571 to DOTEN? DOTEN is the pathway to London.
(2) why go to FL400 at DOTEN?
(3) Do you expect that the 45S1 and 45S2 points are southerly deviations from the LNAV path to McMurdo? In which case we are looking for a McMurdo path that goes a little above 45S1 and 45S2.
(4) Big why? Why did MH370 not follow the exact simulator path? Arc2 suggests the FMT was closer to 93E or 94E, rather than DOTEN which is all the way out to 90E.
@TBill:
1) WMKK’s latitude is 2N. OEJN’s latitude is 21N. I think DOTEN at 10N might be flown over on some WMKK-OEJN routes, which would be optimized with consideration to wind.
2) Stepped climb to reduce fuel consumption.
3) I’m not sure how far 45S1 is from the great circle McMurdo path, but it shouldn’t be far.
4) I can only speculate.
@Andrew,
If cruising at FL350, what airspeed and altitude would you select for an emergency descent?
With no ATC communications, what altitude would you select for an extended HOLD?
@DrB: There is a good video of how crew of a B777 would respond to a depressurization. At 1:22, the MCP altitude is set to 10,000 ft, FLCH is chosen, the thrust is set to idle, the speedbrakes are deployed, and the speed is set to 330 KIAS. (The values are not easy to see, but I am fairly certain that is what is occurring.)
@Andrew,
Thanks for your comments.
Re: “The aim of error correction is to improve the accuracy of the acceleration inputs that are used to calculate velocity and distance. The ADIRU does that by comparing the GPS data with the data the ADIRU has calculated. The ADIRU then determines an error correction that is used to ‘correct’ the acceleration inputs and improve their accuracy. That process in turn improves the accuracy of the calculated velocity, distance and position, and reduces drift.”
It seems finally we are on the same page: dynamic assimilation of the data from the GPS by the ADIRU affects its output parameters, including position, but through the refinement/calibration of the accelerations.
Re: “The SAARU monitors the ADIRU data for validity by checking the ‘validity bits’ that are included in each data ‘word’ sent by the ADIRU.”
That is exactly about my earlier question: are these validity bits set by the ADIRU itself? What criteria is used to set these validity bits, and why they can be different for various parameters?
@VictorI,
Interesting video. FL370 to FL100 in ~5 minutes with speed brakes (ROD 5400 fpm). I missed the speed setting. At what time was that? Would the speed be M0.87 (Mmo) down to 30,500 ft and then 330 KIAS (Vmo) below that altitude?
@Don,
@Andrew,
Re: “When I referred to the WoW, I described it as a ‘discrete’. A discrete is not a databus signal, it is a single wire. In fact, the Honeywell/Racal MCS interconnect provides two WoW discrete lines.”
Thanks for clarifying this. I see this is mentioned in Honeywell manual MCS-4200/7200, Table 5-5.
May I ask some additional questions:
1. How is the WoW parameter specifically used by the SDU?
2. Is WoW a simple bit “in the air / on the ground”?
3. Can the LGA be engaged when the INS data is invalid (the ADIRU is off) regardless the status of the WoW?
4. Does the SDU has ARINC-600 connector, through which it also receives INS data (I refer to Table 5-5), besides ARINC-429?
5. What are Primary IRS A, Primary IRS B, Secondary IRS A, Secondary IRS B (also mentioned in Table 5-5) SDU ARINC-600 connectors for?
6. Why does AMM states that the LGA is used only when HGA fails, given that the “ADIRU off” condition is not something unusual on the ground?
@ALSM,
Re: “I give up. I get the impression you are more interested in diverting the discussion down multiple rabbit holes than actually understanding how the system works.”
Please stop it. I will pretend I did not read this comment from you.
Re: “If I am wrong, please remind me: What is your point? What are you trying to understand?”
Please see above discussions with regard to the LGA-HGA use.
@Ge Rijn,
You misinterpret the results of the drift modeling. The drift modeling gives probabilistic results, not deterministic. Each modeled particle (drifter) is subjected to a random forcing component in addition to the deterministic forcing. The “random walk” is similar to Brownian motion; it governs dispersion. You cannot associate a particle with the a real fragment. To model the dynamics of the latter, you would need to go to extremely fine resolutions, know exact shapes, account for the flow-object interaction, etc. – impossible task.
Specifically:
Re: “I like to see what kind of amount of simulated drifters to start with would reflect the reality of current amount of found pieces and their locations better.”
Drift models cannot answer your question with regard to the initial number of fragments/objects. If this helps: for the origin at 99.09E, 28.84S my model predicts 18 and 25 percent of washed ashore objects by the ends of 2015 and 2016 respectively (assuming no sinking, no decay).
In my opinion the locations agree rather well. You need to look where the peaks of the concentrations of the beached particles are; not how many particles landed.
Re: “Anyway 50.000 is much too high and 30S gives arriving times on Reunion and Madagaskar way too early imo.”
On contrary. One would argue that 50,000 is insufficient. I also have to remind you that there were quite a few possibly relevant objects found in Reunion; who knows how many in Madagascar. When did they arrive if they were from MH370?
Re: “It’s hard to imagine pieces only got found a year later (flaperon) and 1 1/2 later (RR-piece) according to such a model based on so many drifters.”
Read my earlier comments. In addition. The more drifters you take, the more earlier one of them arrive. You need to look at the statistics… Once again, significant arrival to the Mossel Bay is predicted by my model in March – May 2016; only sporadic arrival of some particles before that. This timing agrees well with the discovery time of the RR fragment. I don’t care about arrival of 1 particle of 50,000. This only indicates such a possibility, but probability of such an event is very low (1:50,000 x original number of floating fragments).
Re: “To be short; 30S is too far north according his model and the amount of drifters in the model is way too high regarding actual reality.”
No. On contrary. Also the number of drifters in the model has nothing do to with the number of found fragments. Try to understand this.
Re: “Any conclusive proof of a (high speed) nose down impact and/or a large debris field would be welcomed by me. No one has given any yet.”
There is no way judge about this based on the results of drift modeling or based on the photographs of found fragments. Analysis of residual deformations might somewhat be helpful.
@DrB
Re: “If cruising at FL350, what airspeed and altitude would you select for an emergency descent?”
MMO/VMO (ie M0.87/330KIAS – crossover at FL305) down to lowest safe altitude or 10,000 ft, whichever is higher.
Re: “With no ATC communications, what altitude would you select for an extended HOLD?
Cruise altitude +/-500 ft.
@DrB
Firstly, fantastic work on the fuel model, and fully appreciated by all of us, I am sure.
Secondly, I now have a request.
The issue of “single engine endurance” has not been tackled professionally yet.
Hopefully, with a few minor additions, you may be able to model the following conditions.
If we assume that “endurance” may have been the initial post FMT requirement, and knowing that both engines had different PDA’s, could you now model the best strategies, and results, for single engine operations, for both engines, individually ?
@DennisW
@GeRijn
@Oleksandr
Dennis, you said to GeRijn “Your point is looking for evidence of a small debris field to reinforce your opinion of a soft landing.”
That kind of reasoning works just as well the other way round; the uncontrolled impact proponents are assuming, hoping for, expecting a large debris field to reinforce their presumptions of a hard, nose-down impact.
The large amounts of small sized, unidentifiable debris said to litter the shores of the Western Indian Ocean have no basis in reality. They exist only in the imaginations of the uncontrolled impact brigade.
You can tinker as much as you like with esoteric mathematical processes and obscure statistical hypotheses in an attempt to bolster your arguments, but it’s all smoke and mirrors. The debris itself, the actual debris tells a different story.
To illustrate my point I would like to draw your attention to Fig82 in the ATSB FR: 1) the preponderance of items from the RH wing trailing edge, and right engine cowl, contrasted against the lack of similar items from the LH wing. 2) The apparent progressive increase in the degree of damage to wing parts, as you travel from outboard to inboard (as evidenced by the corresponding photos in the Malaysian debris listings). 3) the remarkable fact that two of the debris items, namely Nos 2 and 27, were mating parts when on the wing but were found dozens of miles apart on the East African coast, both with their outboard sides missing, begging the question “where are the corresponding parts, any corresponding parts from the other wing?” Answer: nothing, except one small section of outboard flap with a hole punched through it (item 10) that appears to have broken neatly away from the flap trailing edge, and one piece of flap fairing tailcone (item 8) itself probably linked to the hole in item 10.
Are you still with me?
4) The infamous pair of panels numbered 9 and 15, identical panels from correspondingly identical locations on the wing upper trailing edges, directly forward of the flaperons, and in an almost identical condition, but found hundreds of miles apart on the African coast, by different finders. When on the wing, each panel comprises one of a set of three that close off the space between the wing and the flaperons, and are mounted on a common metal frame question, where are the other panels?, answe, probably still on the plane.
4) Very few items are from the aircraft interior, and none are from the cargo hold.
Now, more than three and a half years later, all we have, probably all we will ever have, is a highly restricted set of debris parts that reveal how the plane hit the water. It hit the water in an unusual and unexpected way, because it was a way chosen by the pilot with a specific purpose in mind.
@DrB asked: Would the speed be M0.87 (Mmo) down to 30,500 ft and then 330 KIAS (Vmo) below that altitude?
That would be the fastest way to descend (wth speedbrakes). In the video, at one point, the captain sets the speed very high (it looks like 377 KIAS). It really doesn’t matter what speed is set above Vmo or Mmo, because the speed will be limited to Vmo and Mmo. Earlier in the video, it looks like the speed is set to 340 KIAS. As Vmo for the 777-200ER is 330 KIAS, that would be the maximum speed (more or less).
@Rob
You seem to be working with a couple of flawed assumptions, specifically that;
1. High speed, nose down impacts produce relatively more floating wreckage than lower speed, lower angle of incidence crashes. They don’t, in fact, the opposite is true.
Adam Air Flight 574, Flash Airlines Flight 604 and EgyptAir Flight 990 all produced relatively small floating wreckage fields; very much less than a couple of hundred items per field. AF447, the mother of all belly flops, produced over 700 items.
2. The types and quantities of wreckage recovered thus far from MH370 are inconsistent with a high speed, nose down impact. They aren’t, in fact, the opposite is true. The MH370 wreckage is dominated by trailing edge components with the odd fuselage and interior item thrown in. It is exactly the same sort of pattern that Flash Airlines Flight 604 produced; a predominance of training edge components with some fuselage and interior items thrown in. Compare that pattern to AF447 which produced floating wreckage that was dominated by light interior items cabin and hold fittings (bulkheads, galley, ceiling or floor panels, seats, overhead baggage bins, cabin and hold lining) with only a very limited number of trailing edge components. In the case of AF447 trailing edge components and wing and body fairings accounted for significantly less than 5 percent of the recovered floating wreckage.
You seem intractably wedded to your Goldilocks (not too hard, not too soft) belly flop landing. How, on God’s green earth, could anyone have determined the correct altitude to initiate the requisite manoeuvre? Three B777s (British Airways Flight 38, Asiana Flight 214 and Emirates Flight 521) have been variously flown into, slammed into and dropped onto terra firma and the take away is that Boeing screw those hulls together very, very well. The only hull rupture was Asiana 214 and that only came about by flying the rear of the airplane into a concrete and stone seawall.
@Victor
@DrB
For what it’s worth, the simulator in the video is a 777-300ER, which has a higher Mmo/Vmo than the -200ER (M0.89/350KIAS reducing to 330 KIAS at about 5,000ft). The equivalent figures for the -200ER are M0.87/330KIAS, as already noted. As Victor indicated, the overspeed protection will limit the speed to Mmo/Vmo even if the pilot sets higher speed on the MCP.
@Oleksandr
Re: “That is exactly about my earlier question: are these validity bits set by the ADIRU itself?”
I think so, but I don’t know enough about ARINC 629 to be sure.
Re: “What criteria is used to set these validity bits, and why they can be different for various parameters?”
I don’t know how the validity parameters are set in ARINC 629. Perhaps Don can help.
@Andrew: To confuse things more, it may be that KEAS is displayed. I hope this video isn’t too confusing, but I thought the order of operations of setting altitude, selecting FLCH, establishing idle thrust, setting speed, and deploying speedbrakes would be helpful.
@Victor
“1) WMKK’s latitude is 2N. OEJN’s latitude is 21N. I think DOTEN at 10N might be flown over on some WMKK-OEJN routes, which would be optimized with consideration to wind.”
Really? I never see UAE343 head over to DOTEN. It would be unusual for Z to key in a sim route to Jeddah for case of a rare weather diversion. I do accept your theory that the sim study was probably based off the Jeddah route. But I would see DOTEN as the point of departure from the normal path. For what it’s worth, DOTEN seems to be in the FS9 waypoint database, but it is called TEDOD.
@Rob,
You can tinker as much as you like with esoteric mathematical processes and obscure statistical hypotheses in an attempt to bolster your arguments, but it’s all smoke and mirrors. The debris itself, the actual debris tells a different story.
I really do not have a dog in this fight. The analytics are not obscure. They are very mainstream – both Weibull and Poisson. My only point is that you cannot determine the debris field size from the amount and timing of the debris collected. As far as the condition of the debris is concerned, I have no opinion. It is not in my domain of experience or qualification.
@Victor
Re: “To confuse things more, it may be that KEAS is displayed…”
According to the books it’s actually CAS that is displayed. EAS is computed in the background and used in various calculations, but is not displayed.
Best to just call it ‘IAS’!
Re: “I thought the order of operations of setting altitude, selecting FLCH, establishing idle thrust, setting speed, and deploying speedbrakes would be helpful.”
We train the mnemonic FL-FL-S-S-S:
FL – Set a lower altitude on the MCP
FL – Press FLCH
S – Thrust levers to idle
S – Extend full speedbrake
S – Target speed Mmo/Vmo
@Victor: thanks for the reply.
Re: PDFs of scans of documents which some here claim were authored by the RMP: I am happy to follow your advice, and attempt to determine all by myself whether any such claim can be corroborated by the RMP themselves. But both common sense suggests that the burden of proof rests with those – like you – who are making such claims.
I – for the record – would never make so irresponsible a claim, because I’ve no idea who authored these docs. Until we see any actual confirmation of authorship, I respectfully submit that none of us actually has any idea.
If you – or anyone else reading this – comes across any actual evidence of authorship, please present it to this forum.
Thanks again for your time and energy, Victor – I know I’m a stickler for direct, verifiable evidence from independent sources – I’m sure this gets tiresome.
It is certainly my hope that THIS snippet of MH370 “evidence” will turn out NOT to be of murky and dubious origin.
I mean, there’s a first time for everything.
@Florence de Changi, “Inmarsat’s software had been designed when its ground bases were only in the northern hemisphere, and had not been fully updated to account for the ground base in Perth. In three years of investigating, I have not come across a single person – even in the satellite industry – who felt confident in the extrapolations based on the satellite’s “handshake pings”.
Not recommended reading.
http://www.scmp.com/week-asia/politics/article/2114940/malaysia-airlines-flight-370-search-why-give-hope-when-there-was
> Victor Iannello says:
> RE: “But ACARS has no ON/OFF switch”
> It can be turned off by a software switch via the CDU.
> You simply deselect options to communicate by SATCOM, VHF, and HF.
ATSB final report, Appendix B, page 3:
« Apart from a period between the last ACARS message at 17:07 hrs and the handshake at 18:25 hrs on the 7 March 14, the SATCOM link was available during the flight. This interruption of the SATCOM link occurred after ACARS had stopped transmitting messages and may have occurred for a number of reasons such as cycling of the electrical power, the aircraft’s antenna losing sight with the satellite or the resetting of the aircraft’s Satellite Display Unit (SDU). There is no record in the satellite Earth Station log of the link having been logged-off from the cockpit through the Control Display Unit (CDU); such an activity would have been automatically captured in the Earth Station log. The reason for the loss of the SATCOM link is currently being investigated by the aircraft and equipment manufacturers. »
@Victor Iannello:
Is your quote essentially talking about the same thing as the ATSB quote ?
IIUC you explain that ACARS is not really turned off, but just deprived of its means of communication, right ?
When the ATSB says ” the [SDU] link having been logged-off from the cockpit through the CDU” are they talking about the same thing, namely the process of disabling the SDU for ACARS transmissions? Or are they talking about a separate process, unrelated to the ACARS disabling you explain above ?
They say disabling the SDU link would have sent a log-off message to the ground.
Would the procedure you outline above (in case it is not the same), namely deselecting SATCOM as an ACARS communication channel, also send a log-off message?
But if so, how can ACARS send a log-off message if you have just deprived it of the communication channel (SATCOM) it would need to send it ?
And is that the reason why the perpetrator had to depower the SDU beforehand, because without depowering the SDU the disabling of ACARS would have sent a log-off message (or other telltale sign) ?
@David
Well that was cheery reading!
@Andrew. I could say it was superficial, junky, vitriolic, parochial hindsight but best not.
@David
That too…
@Andrew,
Thanks for answering my questions regarding descents and HOLDs.
To follow up, you said: “ “Re: “With no ATC communications, what altitude would you select for an extended HOLD?
Cruise altitude +/-500 ft.”
1. For what reason(s) would you enter a HOLD at cruise altitude +/- 500 ft?
Obviously, the primary impact would be to keep the aircraft in the same general area. It would not save fuel, because the turns would offset the small fuel flow reduction (if any) at cruise altitude. It might reduce pilot workload so attention could be diverted to other issues.
2. For what reason(s) would you enter a HOLD at a MUCH lower altitude (again assuming no ATC communications)? Depressurization? Fire? Smoke?
3. Under what circumstance(s) might you enter a HOLD in the vicinity of FL200 from cruise altitude?
Thanks again. Your expertise in this area is much appreciated.
Florence de Changi. New section stuck on top. This replaces the earlier URL.
http://www.scmp.com/week-asia/politics/article/2115300/gaping-holes-malaysia-airlines-flight-370-search-report
@ventus45,
Thank you for your kind remarks. To address your questions regarding INOP, I will see if I can modify my MEFE model to incorporate INOP conditions both in cruise and in HOLD. Roughly speaking, there would be inconsistent (excess) endurance in INOP HOLDING at all altitudes. For cruise INOP the range of flight levels for consistent endurance would be very broad, possibly even FL100-250. Detailed modeling is required to be more specific, but a quick look at the fuel flow tables indicates that endurance until 00:17 is possible in this case, albeit at reduced speeds (230-280 KIAS).
@David
Re Florence, good grief! Back in the day when it cost real money to print newspapers we wouldn’t be reading tripe like that. Sadly the words “quality” and “journalism” rarely appear in print in the same sentence these days (CVs being a probable exception).
@Rob @DennisW @Oleksandr
Your adressing me also in your previous comment. I believe also statistical models can be usefull. Not to predict precise outcomes in this case but to predict margins of probability.
This is what @DennisW has done with his Weibull-model also and it kind of works till now. Also this model depents on input (i.e. amounts of initial floating pieces).
@Oleksandr chose to start off with 50.000 ‘drifters’. His model (animation) shows it does not predict the reality of only ~30 pieces found so far and the timeframe, and some locations (WA).
So his statistical model is usefull in the sence that it indicates the debris field was very much smaller than 50.000 pieces and the latitude chosen is too far north regarding arrival times.
And still no pieces are found/confrimed from WA.
Quite usefull information I would say from @Oleksandr.
He kind of proves with his similation 30S was not the crash latitude (too far north) and the debris field was very much smaller than 50.000 floating pieces imo.
Scaling down the amount of ‘drifters’ to start with in his model at least ten-fould would give a more realistic and clearer picture imo.
Then also using different latitudes (not only a preferred one) as a starting point would make comparison possible to this animation.
We still don’t know what kind of impact it was for no conclusive proof has been given by the ATSB or anyone else.
To me it’s clear it was no nose-down impact but I cann’t prove this either ofcourse.
@Dennis
@Oleksandr
@GeRijn
Did you spot the embarrassing slip in the following sentence?
“The apparent progressive increase in the degree of damage to wing parts, as you travel from outboard to inboard”
It should be the other way round. It should have read “progressive DECREASE in the degree of damage to (RH) wing parts, as you travel from outboard to inboard”.
The RH wingtip hit the water first, so taking the brunt of the impact, energy was absorbed progressively by the RH wing, the aircraft rolling to the left in the process. Consequently, by the timethe left wingtip hit the water, much of the impact of energy had been absorbed, leaving only enough to knock off the outboard tailcone and flap fragment, items 10 and 8.
@TBill: Re DOTEN, I don’t know. You could be right.
@Andrew, @DrB: Wouldn’t the choice of altitude for the hold depend on how long you expect to hold and what clearance you have? If the pilot wanted to maximize endurance and expected to be holding for some time, and wasn’t seeking ATC clearance, why wouldn’t a descent to the optimum hold altitude make sense? I know this is a different question than @DrB asked.
@David: The last paragraph was left off, which would explain the rest: “And for these reasons, we know the plane was shot down over the Gulf of Thailand.”
@Peter Norton: Isolating the left bus, which is what I believe occurred initially, is different than deselecting the SATCOM link options using the CDU, which I believe occurred before the left bus was re-powered. I didn’t explain this because we have been over this many times.
@DrB
“1. For what reason(s) would you enter a HOLD at cruise altitude +/- 500 ft?”
To be honest, there aren’t many reasons for holding at cruise altitude. En-route holding is usually only ever required for ATC delaying action, but that’s obviously not a consideration in this case. Another reason might be a desire to hold somewhere while sorting out some kind of problem, technical or otherwise. That said, technical problems would normally be sorted during the cruise towards a diversion airfield and if holding were required it would probably take place during descent close to the airfield, rather than enroute. A non-technical problem that might require enroute holding is a hijack, but again, that doesn’t seem very likely unless Zaharie Shah’s simulator exercises were purely coincidental.
The reason for holding +/- 500 ft from the cruise altitude is simply to avoid any traffic that might be cruising at standard levels, given that an ATC clearance is not available. The same would apply to a hold at lower levels.
“2. For what reason(s) would you enter a HOLD at a MUCH lower altitude (again assuming no ATC communications)? Depressurization? Fire? Smoke?”
The only reason that springs to mind is holding to sort out additional problems before commencing an approach to land. Emergencies such as smoke or fire would normally require an immediate landing, so holding isn’t likely unless the fire was under control and time was needed to remove smoke before landing.
“3. Under what circumstance(s) might you enter a HOLD in the vicinity of FL200 from cruise altitude?”
FL200 provides the best endurance at weights around 200-220T. You might decide to descend to FL200 and hold if you wanted to stay airborne as long as possible or weren’t sure about how long you might need to hold. As for circumstances, I can’t think of any that might apply in this case, except a hijack scenario perhaps, but as I said above, that doesn’t seem likely.
@Victor
“Wouldn’t the choice of altitude for the hold depend on how long you expect to hold and what clearance you have? If the pilot wanted to maximize endurance and expected to be holding for some time, and wasn’t seeking ATC clearance, why wouldn’t a descent to the optimum hold altitude make sense?”
Yes, it would make sense to descend to the optimum altitude if the intention were to maximise the endurance or if the expected holding time were unknown. But what circumstances would justify such a course of action in this case?
@Peter Norton
From your quote above:
“The reason for the loss of the SATCOM link is currently being investigated by the aircraft and equipment manufacturers.”
Frankly, I would be embarrassed to put something like the above in a final report after 3.5 years. I call it BS. The event is simply not complex enough to “fuel” a 3.5 year investigation. An “investigation” missing anything resembling a progress report.
As Victor says, the IG (and DrB) think they have bagged it with left bus power cycling explanation because they are smarter and more experienced than the aircraft and equipment manufacturers. Of course, one has to wonder if the PIC was equally smart. Why would a PIC cycle power on the left bus? From the point of view of the PIC the aircraft was already “dark”. Why would the PIC restore power at 18:xx, if he indeed switched it off earlier. Why was the first BFO logged at 18:25:27 virtually perfect when it was a random power up value (and discarded as such by Holland)? Valid questions that no one has a credible answer for.
This whole event stinks, frankly. What makes it stink even worse is the fact that most people here are perfectly happy with what amounts to a fairy tale narrative.
@Andrew: I can only speculate, as others have. But we have to consider that a northerly track at 18:40 that coincides with a descent, followed by a loiter, followed by a flight to the SIO, will lead to a crash site north of what was searched, and also satisfy the fuel constraints. It checks a lot of the boxes (but maybe not all of them) even though we can only speculate about why a loiter occurred. On the other hand, there’s no sense speculating about scenarios that don’t check boxes.
@DennisW said: Calm down. I never said I am certain the SATCOM was powered down. However, it certainly makes more sense to me than anything else I’ve seen proposed. As for claims that we are smarter than the OEMs, I don’t think I ever said that. And I certainly accept Holland’s work on the BFO. A power up at 18:25 is what he believes occurs. The debate centers on whether or not to accept the first BFO value. He says reject it. Others try to explain it. That’s not a major divide. As for fairly tale narrative, I encourage you to write up your thoughts so that we can all discuss it. I assure you that what you consider to be a scenario grounded in logic will be considered by others to be a fairy tale narrative.
@Andrew. If one were flying the aircraft in Heading Select, what is the granularity with which one can select the heading? Is it a digital selection to nearest degree, nearest decimal degree, or an “analogue” selection on a dial? Apologies in advance for the neophyte question.
@Victor
I do not have an explanation.
I did not mean to imply that you said you were smarter and more experienced than the OEM’s. Others in your camp have said that, however.
As far as I know, no one has explained the 18:25:27 BFO value in the context of a power up. It is in the serendipitous category, and it might well be serendipitous and should be discarded as Holland states
My only point, and I should have stated it, is that the 18:25 logon is not in the “we discussed that ad nauseam, so move on” category. I am not at all comfortable with what is commonly regarded here as a “put to bed” issue. People even respond with “have you not read my paper?”. Of course, I read your paper, and carefully at that.
@DennisW: My comment about having discussed this before referred to the differences between deselecting all link options by CDU and powering down the left bus. The commenter was confused on this matter, and asked for an explanation.
@Victor
“@Andrew: I can only speculate, as others have. But we have to consider that a northerly track at 18:40 that coincides with a descent, followed by a loiter, followed by a flight to the SIO, will lead to a crash site north of what was searched, and also satisfy the fuel constraints.”
Victor I think it was a brilliant deduction on your part a long time ago that there was descent and Hold. But why not just descent and track out to DOTEN just like the sim cases? A little like your LAGOG out-and-back to BEDAX cases, but it actually towards DOTEN (probably via IGOGU to DOTEN), the sim cases would suggest. Also a possibility it was heading out to DOTEN via DUBTA, whereas DUBTA is definitely on the track to Jeddah in some route finder programs.
@TBill: I think those are possibilities worth exploring, and I appreciate your proposing them. Either way, the McMurdo waypoint (which I know you are lukewarm about), puts the plane just south of VOCX at 19:41.
@Paul Smithson
Re: “If one were flying the aircraft in Heading Select, what is the granularity with which one can select the heading? Is it a digital selection to nearest degree, nearest decimal degree, or an “analogue” selection on a dial?”
The heading can only be selected in single degrees, eg 090°, 091°, 092°, etc. See the following video from 0:28 onwards:
https://m.youtube.com/watch?v=YpgKhUo3uqw
Hi @Andrew. Thanks very much. Do you have a notion on how accurate the heading tolerance is +/- 0.x degrees?
By which I mean, if you set 090 and have zero wind, how much might the actual heading differ from target heading – whether as a one-sided error or as “noise” +/-? Would you, for example, expect better than +/- 0.5 degree vs. target heading?
@VictorI
Do I understand correctly that 45S, 104E could be close to fuel exhaustion for a real Jeddah flight after diversion from 10N?
I estimate a flight time of 8h37m based on 480 kts GS.
Another question: is it correct that the Febr 4 MH150 flight was scheduled to leave at 3:45 am UTC?
@DennisW
re “frankly, I would be embarrassed to put something like the above in a final report after 3.5 years. I call it BS. The event is simply not complex enough to “fuel” a 3.5 year investigation. An “investigation” missing anything resembling a progress report.”
Dennis you’re right on the money, it is complete bullshit, and disingenuous bullshit at that. They previously stated, in the Definition of UW Search Areas that the consensus was it was due to a power interruption. So why the backtrack? I think the Malaysians told them to change it, no less. Power interruption implies deliberate action from the cockpit, which implies pilot hijacking. The Malaysian alternative narrative machine is cranking up.
@Niels asked: Do I understand correctly that 45S, 104E could be close to fuel exhaustion for a real Jeddah flight after diversion from 10N?
No, that’s not what I meant. I calculate the loaded fuel at 2N to be 68,424 kg, based on the fuel levels for each tank, and the tank capacities. Let’s use your number of 8h37m of flying time. If the fuel flow averages around 6,600 kg/hr, that means there are reserves of 20%. Meanwhile, other arrival airports such as Chennai and Amsterdam would have either too much fuel or not enough fuel.
[Is] it correct that the Febr 4 MH150 flight was scheduled to leave at 3:45 am UTC?
I don’t know.
@Andrew
You are not alone. I’ve never understood where a descent and hold would possibly have a place in Shah’s itinerary, either.
And as for the maximum endurance mantra…
@Victor, I do not want to seem argumentative do the sake of it, or pig headed, or overly persistent but sorry, that’s just the way I am.
@Rob
Don’t have a comment on the descent part, but a hold is part and parcel of an ongoing negotiation with co-conspirators in KL.
@Victor@All
Aside from a negotiation (and presumably stalling for time) or the PIC having second thoughts, can you think of any other possible reason for a loiter?
I am only interested in scenarios that have the PIC in command of a perfectly functioning and controllable aircraft and then deliberately loitering. I’m interested in any pure speculation you may be entertaining.
Lets set aside the small matter of the condition of the pax and crew for the sake exploring the ‘myriad’ possible reasons for a loiter, aside from negotiation and reservations/second thoughts (which of course is patently absurd).
IMHO any time spent on a loiter is time wasted…unless a sensible reason as to why can be offered up?
Thanks Victor.Your work is tremendously valuable.
@Rob
Name your preferred ditching style (I’m personally partial to your version), but the idea of either a deliberate nose-down, vertical impact at high speed OR purposeful death by hypoxia pre (or I suppose post) fuel exhaustion and the ensuing uncontrolled descent are sadly lacking when one truly understands the deviousness and determination of the action. The need to feel in absolute control of events and outcomes does not end until the task at hand is satisfactorily completed.
Pointing the nose down would have been an ending that Zaharie would have never given any serious consideration to…it’s not a final act remotely worthy of the ego we are all witness to.
@All
Didn’t see the other posts also questioning la loiter. Sorry for the redundancy.
@Mick Gilbert
I know Boeing have a very solid plane in the 777.
I also know that AF447 was the mother of all belly-flops, with a lot of interior parts left floating.
What I gave you earlier was my theory on what happened, based on the debris. I stand by every word of it.
I believe that Shah was in control until the end, that he was an exceptional aviator, that he planned the flight so as to have just sufficient daylight when the time came to ditch his aircraft. I also believe he studied previous accidents when planning this flight.
As for Goldilocks, well fairy tales are strictly for the birds.
@Donald: I have nothing useful to offer. If there was a failed negotiation, all parties have certainly done a great job keeping things quiet.
@Victor
The Malay response in the hours after IGARI speaks for itself IMO. The Malays did not initiate any emergency action for over four hours into the diversion. It suggests that they knew full well the nature of and reason for the diversion.
@Niels
On 4 February 2014, MH150 was scheduled to depart at 15:15 MYT (UTC+8), actual departure time was 15:14 MYT.
@Peter Norton
Re: “When the ATSB says ” the [SDU] link having been logged-off from the cockpit through the CDU” are they talking about the same thing, namely the process of disabling the SDU for ACARS transmissions? Or are they talking about a separate process, unrelated to the ACARS disabling you explain above?”
Logging the SATCOM off and disabling its transmission path(s) are two distinct processes. The former generates a log-off script, the latter does not.
Re: “They say disabling the SDU link would have sent a log-off message to the ground.”
No, they do not. They say, “There is no record in the satellite Earth Station log of the link having been logged-off from the cockpit through the Control Display Unit (CDU); such an activity would have been automatically captured in the Earth Station log.” They are talking about logging the SATCOM off, not disabling the link (transmission path).
@DennisW
@Rob
Re: “The reason for the loss of the SATCOM link is currently being investigated by the aircraft and equipment manufacturers.
A little bit of context goes a long way. That quote is from Identification of the most probable final location of flight MH370 (Issue 2) – Joint Investigation Team paper – 26 April 2014. It is prefaced in the ATSB final report with the following statement;
“Note: This appendix provides a copy of the analysis as provided post a Joint Investigation Team (JIT) briefing to AMSA on 2 April 2014. The paper therefore describes the analysis and advice to 2 April 2014 and represents only a snapshot of the JIT on-going analysis to early April 2014.
The paper provides the background on BTO and BFO analysis during the Australian-led surface search in March-April 2014. The reader should consider the contents of this paper in this context only.”
Could that possibly have been any clearer?
And just by the bye, the ATSB final report relates to the search, it is not an accident investigation report.
@Mick
The reason for the loss of the
SATCOM link is currently being investigated by the aircraft and equipment manufacturers.
You will find the above statement on page 180 of the final report as I stated in my initial reference.
@Paul Smithson
My apologies, I misunderstood your original question. I don’t know the accuracy of the B777 ADIRU specifically, but my understanding is that the ARINC 704 specification requires an accuracy of ±0.4° (2-sigma).
> Victor Iannello says:
> @Peter Norton: Isolating the left bus, which is what I believe occurred initially, is different than
> deselecting the SATCOM link options using the CDU, which I believe occurred before the left bus was
> re-powered.
@Victor Iannello:
Thank you for your reply.
Seeing your answer, I think I phrased my comment in a way that didn’t make much sense to you.
I didn’t talk about the left bus.
I’ll try to rephrase:
— begin quote —
Victor Iannello says:
RE: “But ACARS has no ON/OFF switch”
It can be turned off by a software switch via the CDU.
You simply deselect options to communicate by SATCOM, VHF, and HF.
ATSB final report, Appendix B, page 3:
« Apart from a period between the last ACARS message at 17:07 hrs and the handshake at 18:25 hrs on the 7 March 14, the SATCOM link was available during the flight. This interruption of the SATCOM link occurred after ACARS had stopped transmitting messages […]. There is no record in the satellite Earth Station log of the link having been logged-off from the cockpit through the Control Display Unit (CDU); such an activity would have been automatically captured in the Earth Station log. »
— end quote —
1) When you say “[ACARS] can be turned off […] via the CDU […] by deselect[ing] options to communicate by SATCOM”, you are saying that ACARS is not really turned off, but just deprived of its means of communication, right? So it is essentially generating status messages, but cannot send them, correct ?
2) The ATSB says “the [satellite] link [can be] logged-off from the cockpit through the CDU”. Is this the same process you describe above, namely the process of deselecting the SDU for ACARS transmissions (“[ACARS] can be turned off […] via the CDU […] by deselect[ing] options to communicate by SATCOM”) or is this another pair of shoes ?
3) The ATSB says disabling the satellite link would have sent a log-off message to the ground.
If the answer to 2) is “another pair of shoes”, would the procedure you describe above, namely deselecting SATCOM as an ACARS communication channel, also send a log-off message?
4) If the answer to 3) is “yes”, how can ACARS send a log-off message if you have just deprived it of the communication channel (SATCOM) which it would need to communicate ?
5) Did the perpetrator have to depower the SDU beforehand, because without depowering the SDU the disabling of ACARS via CDU would have sent a log-off message (or other telltale sign) ?
@Donald
Re: Loiter
First of all I don’t exactly like the word “loiter”, I prefer maneuvers. Two theories on the purpose would be: (1) to decoy a flight north or west, to disguise actual intent go south to hide the crash site(not realizing the Inmarsat pings would give away direction); and (2) to try to evade or confuse (Indonesia) radar detection.
So one hypothesis of mine is MH370 went just like simulator case out towards DOTEN which is the waypoint to London/Europe, then cuts back at higher altitude, just like the simulator cases, to appear like a different flight coming in at a different altitude. Cutting south FMT somewhere above ISBIX. Note when I say towards DOTEN, there is not enough time to get all the way pout there for most paths, so perhaps not much further out than above LAGOG at 92E, which is the edge of the Indonesian FIR space.
@ Mick Gilbert:
Thank you for your clarifications. (we have been writing at the same time)
If I understand you correctly, the answers to the 5 questions above are:
1) correct
2) another pair of shoes (logging-off SDU -VS- deselecting SATCOM as ACARS transmission channel)
3) No. Deselecting SATCOM as ACARS communication channel would not send any message to ground.
4) invalid (answer to 3 is no)
5) No. The disabling of ACARS via CDU would have sent neither a log-off message nor any other telltale sign.
Is this all correct now ?
(thank you!)
I’m wondering about 5) now.
If the perpetrator didn’t need to depower the SDU because the procedure outlined by Victor (“[ACARS] can be turned off by a software switch via the CDU. You simply deselect options to communicate by SATCOM”) would have sufficed, then why was the SDU depowered ?
Thank you, Andrew, for answering my questions regarding holds and descents.
At the risk of becoming a target, I will try to summarize what I think is known, and what is unknown, regarding the critical period between 18:22 and 19:41. Please feel free to provide EVIDENCE to the contrary or to add other points.
(1) The Lido slide position data are most likely valid, including the last 18:22:12 position. The ATSB Final Report appears to adequately address concerns regarding its validity.
(2) The Lido slide is consistent with a constant high ground speed (507 ± 5 kts.) from 18:00 to 18:22:12, although the position errors are large enough that a slow-down could have begun during the last couple of minutes.
(3) The positions of the last two radar contacts in the Lido slide more closely follow a linear extension of the route from VAMPI through MEKAR rather than a slight turn to the right at MEKAR toward NILAM, following N571. Besides, if 9M-MRO were following N571, it would have turned right BEFORE reaching MEKAR so that it would meet the MEKAR/NILAM track without overshooting it. It would not have passed through MEKAR, as the Lido slide indicates. Based on the scatter of the radar positions along the Lido track, it doesn’t seem likely to me that the last position could have a position error so large as to be consistent with the required location then (on or slightly to the east of the MEKAR/NILAM line) if it were truly following N571. This conclusion implies that the next waypoint after MEKAR was not NILAM, but some point farther to the west.
(4) The 18:25-18:28 BTOs/BFOs are consistent with a 15NM right lateral offset beginning at or shortly after 18:22.
(5) There is a procedural explanation for the SLOP, perhaps accompanied by a small altitude change.
(6) Accepting the 18:22:12 radar position, single turns then or shortly thereafter do not appear consistent with the 18:25-18:28 BTO/BFO data. One can match the BTO data alone reasonably well with a single turn at 18:22, but matching the BFOs is problematic. In particular, matching the 18:28 BFOs requires a track of about 285-310 degrees true (assuming no climb or descent). It is important to understand the sensitivity of the BFO at this time. A 100 kt change in ground speed only moves the BFO by 4 Hz, so the BFO is quite insensitive to ground speed. A 10-degree track change causes a 6 Hz BFO shift, so this effect matters more. The highest BFO sensitivity is to ROC/ROD, where 1000 fpm causes a 22 Hz shift. Based on this high degree of sensitivity, it is very hard to match the 18:28 BFOs with any significant ROC or ROD, regardless of ground speed or track.
(7) I call attention to the 18:28 BFOs in particular, because there seems to be no debate about the validity of those data or the conclusion that the OCXO transient effect should have decayed to be quite small by that time. In other words, the debate regarding the 18:25:27 BFO is largely immaterial since using it or not using it makes no material difference in interpreting aircraft maneuvers near that time. The 18:28 BFOs (plus all the BTOs and the 18:22 position) are the key constraints.
(8) A rapid decrease in ground speed near 18:22, to about 350 kts, can match the subsequent 18:25-18:28 BTOs without a turn, but this also requires a descent to lower altitude to maintain stable flight. That descent must be accomplished between 18:22 and about 18:27, and be completed by 18:28. This descent is possible to achieve with speed brakes in an emergency descent, squeezed into a narrow time window, but then one is left wondering why a second, separate descent might subsequently occur at 18:40.
(9) Thus, it seems there could be two explanations for the 18:22-18:28 data (BTOs + BFOs + Lido radar position), but both require a major maneuver to occur at 18:22 (or within a very short time thereafter), and they must be completed before 18:28. I would judge the SLOP to be a slightly better fit to the data than the emergency descent.
(10) There is no obvious reason to enter a HOLD at cruising altitude when not under ATC control, except for hijacking, which seems very unlikely in this case. There is also no direct evidence for a “negotiation”.
(11) Other than depressurization, there also doesn’t seem to be any obvious reason to descend to a low altitude when not part of a landing approach, much less to enter a HOLD at low altitude, and even less so for a 1-hour HOLD. To quote Andrew: “Yes, it would make sense to descend to the optimum altitude if the intention were to maximise the endurance or if the expected holding time were unknown. But what circumstances would justify such a course of action in this case?” I can’t think of any circumstances that are also consistent with the flight up to and after that point. Why would anyone choose to descend and fly around in a racetrack for an hour, just to gain a few minutes of flying time? If the goal were simply to maximize endurance, then why leave the HOLD at all? If the goal were to travel a long distance south, then why go NW first, descend, fly in circles for an hour (which does nothing to avoid airspace if you want to go there), and then ascend before heading south? So far, I have not heard a rational explanation. It can’t be that this was done simply to follow a prepared plan. The recovered route doesn’t match the satellite data. Victor found one that apparently does match the sat data in the general vicinity, but it doesn’t seem to be all that close to the recovered route.
DennisW says, “The Malay response in the hours after IGARI speaks for itself IMO. The Malays did not initiate any emergency action for over four hours into the diversion. It suggests that they knew full well the nature of and reason for the diversion.”
One must always first rule out incompetence. The record is replete with examples on both sides of the FIR boundary.
@DennisW
Dennis, the statement “The reason for the loss of the SATCOM link is currently being investigated by the aircraft and equipment manufacturers.” appears once and once only in the ATSB final report. You’re correct in that it appears on the 180th page of the report but that is page 3 of the paper “IDENTIFICATION OF THE MOST PROBABLE FINAL LOCATION OF FLIGHT MH 370 (ISSUE 2)” which is included as Appendix B. That paper is dated 26 April 2014!
Have a bo peep at the 177th page of the report (Appendix B >B<) it provides the context for that paper.
@DrB: Those points are helpful. I have the following comments to make:
1) Starting at 18:22:12, I think there are other speeds and tracks that fit the BTO and BFO data. For instance, a track towards VOCX at 306 deg and 225 KIAS (about 389 kn TAS at FL340 and ISA+10K) seems to work with a BTO RMS error of around 29 μs and a near exact match to the BFO. I believe 225 KIAS is above the minimum maneuvering speed at 209 MT.
2) Clearly, if the pilot entered into a hold and later flew to the SIO, the objectives changed and the plan switched from A to B. After the diversion at IGARI, I see three possible phases: a) Flight at cruise speeds and altitude until out of Malaysian FIR and beyond Malaysia radar; b) Decelerate, descend and hold for some reason not yet identified with any degree of certainty; c) A change leading to a flight to the SIO. Unless new evidence surfaces, we can only guess about why it might have occurred. That doesn’t mean it didn’t occur. In my opinion, it remains on the table as a possibility.
3) Yes, of course the sim data doesn’t exactly match the satellite data. If it did, we’d have found the plane. But there are some interesting similarities that have not been fully explored, in my opinion.
4) There is now no denying that weeks before the disappearance, the captain simulated a flight up the Malacca Strait and into the SIO. This would be an extraordinary coincidence if the simulation was somehow not related to the disappearance.
@VictorI,
Based on a quick look, I concur regarding the 306 degree track at 389 KTAS. The BTO error I got was a bit large at 18:25:34, but I think that can be resolved by beginning the slow-down slightly before 18:22. So now we have three possible solutions for a maneuver at ~18:22.
As I understand it, in your proposed scenario the following six events occurred between 18:22-19:41:
1. 15 NM R SLOP, or a turn to VOCX and a speed reduction to 225 KIAS
2. Descent to ~FL210
3. HOLDING pattern for ~1 hour
4. Air Packs turned OFF
5. Ascent to cruise altitude
6. Acceleration to LRC
As you say, it may be possible, but that’s a lot of activity in a relatively short period of time, and some of the actions are not easily explained.
@DrB
Re: 11)
I propose that normal late night flights to DOTEN short cut from IGOGU and usually go at FL300 outbound on N877. Andrew has mentioned FL280 to FL300 is a common outbound flight rule used on N877. I suggest MH370 descended and behaved like a normal outbound flight on N877, until it turned back, when it may have ascended to look like a normal in-bound flight. This is one explanation for descent, there may be other flight path rule explanations, if paths other than N571 or N877 were followed (eg; SAMAK to DUBTA is another poss.
Slow down does not fit my expectation except possibly after ISBIX/Indonesia air space. Those favoring FMT south at IGOGU like a slow down as that makes the 18:40 BFO and later points fit better.
@DrB
Nice summary. A couple of initial observations/comments:
Re: “(2) The Lido slide is consistent with a constant high ground speed (507 ± 5 kts.) from 18:00 to 18:22:12.” As previously discussed, while I think that Lido is consistent with an average high ground speed, I think that there is evidence that the ground speed was not constant. And why should we expect that the airplane’s speed up the Straits of Malacca was constant when there’s evidence (DSTG 10 second radar data and leaked RMP report) suggesting that the airplane’s speed was not constant as it flew back across the Malay Peninsula?! There’s also evidence (FI and leaked RMP report) to suggest that the variations in ground speed may have been associated with variations in altitude. Whatever was causing variations in the airplane’s speed across the Malay Peninsula most likely persisted up the Straits of Malacca.
Re: (3), that the airplane followed “a linear extension of the route from VAMPI through MEKAR rather than a slight turn to the right at MEKAR toward NILAM … , if correct (and I think it is) then (4) “a 15NM right lateral offset” to N571 must be less likely. Why fly an offset to an airway that you are no longer on? Particularly when achieving the offset means crossing the airway that you’re trying to avoid traffic on!
Re: (3) continued, “This conclusion implies that the next waypoint after MEKAR was not NILAM, but some point farther to the west.” The only reasonable fit to passing from VAMPI through MEKAR on 286T is ANOKO but why assume that there was a next waypoint? The airplane may not have been in LNAV; in fact, all the evidence suggests that apart from VAMPI – MEKAR (less than 70 nm out of nearly 490 nm from the turn back or less than 15% of the observed deviated flight) the airplane was not navigating by waypoints.
@Mick
Have a bo peep at the 177th page of the report (Appendix B >B<) it provides the context for that paper.
I did. You are correct that Appendix B is historical which I also mentioned in the context of the earlier flight path predictions and BFO anomalies noticed by both sk999 and myself. Do you suppose the Holland paper is the final answer to the 18:25:27 event being look at by the aircraft and equipment manufacturers? There is another paper cited by Holland’s paper (no fewer than four times) that has never been put in the public domain. I asked Holland about it, but he did not reply although he did reply to an earlier query I made about his paper.
@Mick Gilbert,
You said: “Whatever was causing variations in the airplane’s speed across the Malay Peninsula most likely persisted up the Straits of Malacca.”
I don’t know about speed variations before 18:00, but from 18:00 to 18:22 the radar positions are very consistent with a constant ground speed (as I have previously demonstrated in a note posted here). I have not seen any evidence of speed variations after Penang that is not fully explainable by the apparent (small) errors in the radar positions.
I agree that it makes no sense to perform a SLOP if you are not following an airway, and that logic would seem to make it more likely that either the slow-down/descent or the slow-down/slight turn occurred near 18:22 rather than the SLOP.
I assume LNAV was used after MEKAR simply because LNAV was used from VAMPI to MEKAR (regardless of what may have happened before VAMPI).
I note that if VOCX were entered in the active route legs page to follow MEKAR, the turn would also have commenced toward VOCX before MEKAR was reached, even sooner than a turn to NILAM would commence because VOCX is 10 degrees more to the right from MEKAR than NILAM is from MEKAR. If you want to believe that VOCX was the next destination after MEKAR, then you have to accept that VOCX was entered into the active route just as MEKAR was overflown, not before, so that the turn to VOCX began after MEKAR, not before.
On the other hand, if the next waypoint after MEKAR were ANOKO, then there would be no visible turn at MEKAR in the military radar data since the track deviation is only 1 degree. In this case a deceleration (and possibly a descent) that is begun near MEKAR can match all the data without a turn. This scenario requires an even slower airspeed, however, and the match to the BFOs is not great but is marginally acceptable.
@TBill,
You said: “Those favoring FMT south at IGOGU like a slow down as that makes the 18:40 BFO and later points fit better.”
Actually, that is not the case if you use the 15 NM SLOP. In that case you really can’t go quite fast enough to make the turn before 18:40 and pass 15 NM to the right of IGOGU, even at M0.84. Without the SLOP, the speed can be normal or a bit slower to make the turn before 18:40 using IGOGU.
@DrB said: “I assume LNAV was used after MEKAR simply because LNAV was used from VAMPI to MEKAR (regardless of what may have happened before VAMPI).”
I dont’t think you can be sure that any autopilot mode was used at any point point after IGARI. The airplane can be navigated between waypoints by reference to the Navigation Display in any autopilot mode other than LNAV, and also while flying manually with autopilot off.
@Mick Gilbert
Thank you. Based on the “Timetable 1 December 2013 – 29 March 2014” a departure time of 11:45 local was announced for MH150, from Jan. 28th onwards. Would there be a way to check when this new departure time was actually implemented?
@Andrew. No apology, please. You have been most helpful and patient with diverse and trying enquiries. You had interpreted my original question correctly. a) I wanted to know whether you can select a heading in fractions of degree or only whole degrees b) I wanted to know how accurately a selected heading can be held. You have answered both. Many thanks.
@DrB
Re: “in your proposed scenario the following six events occurred between 18:22-19:41:
1. 15 NM R SLOP, or a turn to VOCX and a speed reduction to 225 KIAS
2. Descent to ~FL210
3. HOLDING pattern for ~1 hour
4. Air Packs turned OFF
5. Ascent to cruise altitude
6. Acceleration to LRC”
It strikes me that such a scenario fits DennisW’s theory quite well…
@DennisW, sk999
It is hard to imagine the diversion went unnoticed. MH370 passed close to RMAF Butterworth which is the IADS HQ. Understanding the lack of response especially from Butterworth could hold one of the keys we are looking for. Note also the role of Australia in the IADS.
https://en.m.wikipedia.org/wiki/Five_Power_Defence_Arrangements
@Niels
It looks like MH150 had departed at 1535 MYT up until 30 October 2013 when the departure time was brought forward to 1515 MYT. I’ve done a quick scan of departures from November 2013 to August 2014 and can find no sign of it being scheduled to depart at anything other than 1515 MYT.
@Dr B. Thank you for your summary. I’ll also stick my head above the parapet.
As I’ve noted previously, you can obtain BTO fit 1825-1828 without a SLOP or a turn if the aircraft has slowed down. The difficulty is in reconciling this with the LIDO slide.
For some time I became interested in an early FMT that starts around 1825, then via ANSAX and WITT. This also required a slow-down, which I explained by transition to Econ descent speed after passing top of descent for WITT approach via ANSAX 2-C. However, I ruled out this early turn on basis of BFO. Notwithstanding the transient, it’s pretty clear that it is tending towards an asymptote of about 144, indicating heading of ~291 +/- 8 degrees at 1828, ie no turn has occurred.
I have now established that you don’t need to make a two-stage turn that early. You can continue along your track of ~292 until you meet the 011 Radial that marks the initial lead to ANSAX. Heading 191M down that radial puts on you on 190T +/- 0.5 degree. If you crossed BTO=12480 at 182815, the turn onto that lead radial would begin only 15 seconds later, at 182830.
In order to do this without a SLOP, you still require a slow-down. I have gone back to the LIDO slide to try and see how this might be compatible. Looking at the original, highest resolution version of the picture, I cannot make out more than 2 of the time stamps. How confident are we of the timing labels applied to selected positions along LIDO?
If you continue down the ANSAX radial on 190T and keep on going for ever, you can obtain a very nice fit to the BTO and BFO through to predicted fuel exhaustion (first engine 00:06, second at 00:14 – so within 0.7% of “actual”) if your aircraft is at Econ Descent speed (IAS 272), FL340 and flying Constant True Heading. Question for experts: might he have flown in HDG select for a VOR approach along radials due to some problems restricting ability to program & fly the full STAR? Could that explain how you end up on constant true heading?
The scenario seems to have quite a lot to offer:-
– A reason for SDU reboot: Powering up left AC bus shortly before commencing planned approach to WITT [question to experts: what might he have needed so badly for landing that he risked re-powering damaged circuits that had been previously isolated?]
– Slow down that is required to obtain BTO fit 1825-1828. Passed TOD and speed transitioned automatically to Econ Descent in VNAV ALT (or manually set to similar)
– A reason for the FMT. To begin an ANSAX 2-C approach to WITT
– A reason for flying quite a bit slower than normal cruise speed along this “ghost flight path” –
because that was his Econ Descent speed.
– The path model produces an uncanny match to both the heading (190 +/- 0.5 degree) and the speed (272 IAS at FL340) that you require in order to produce best fit to BTO and BFO. ie both heading and speed match prediction of a planned approach at WITT that was never executed.
– Using Dr B’s fuel model, the path model also produces an excellent match to fuel endurance (within 0.7% of second flame out at 001730).
– And this path model just happens to end smack dab in the middle of DSTG’s fried egg.
If the area hadn’t already be searched, I’d be yelling Eureka from the rooftops.
The principal difficulty with this scenario is that the plane wasn’t found there. This is why I continue to ask is there some way we might be wrong about the end of flight and the impact point lies inside/outside the searched width of the 7th arc at 38S?
ps: for the drift modelling doubters, at 38S (as at 35S) there was another ocean surface current anomaly pushing NW.
@Mick Gilbert
Thank you for checking. The change of departure time was announced in combination with the introduction of an extra daily service to Jeddah (mh168). Apparently they changed plans.
http://www.routesonline.com/news/38/airlineroute/214191/malaysia-airlines-increases-jeddah-service-in-feb-mar-2014/
@DrB said: As you say, it may be possible, but that’s a lot of activity in a relatively short period of time
True. That’s one of the reasons I was skeptical of the Lido radar data. Together with the satellite data at 18:25, it requires some manoeuvers between 18:22 and 18:25. But we now have higher confidence that the radar captures shown really were MH370. I think that means we are stuck with a manoeuver between 18:22 and 18:25.
So we have a final radar capture at 18:22, a possible power up of the left bus around 18:24, a manoeuver of some type before 18:25 to satisfy the BTO sequence at log-on, and another manoeuver before 18:40 to satisfy the BFO values for the sat call. That’s what our interpretation of the data says. I think this is collectively telling us that there WERE a lot of things occurring in this time interval.
The manoeuver before 18:25 could have been a SLOP or turn and deceleration. (I am putting aside the possibility of a very steep decel and a descent.) The manoeuver before 18:40 could have been a turn or an extended descent. I don’t see how the turn/decel followed by the descent is any less likely than the SLOP and the turn.
I do agree with the previous comments that if the plane remained in LNAV and was following the waypoints MEKAR-VOCX, the radar captures near MEKAR would have shown this. That doesn’t mean there wasn’t a turn towards VOCX after passing MEKAR.
and some of the actions are not easily explained
As I have said many times, there is no scenario in which all actions are easily explained.
As for the scenario of @DennisW with a negotiation, I continue to have an open mind about it. I certainly don’t dismiss it. It could explain a lot of things. But as I said, it is strange (but not impossible) that no evidence of the negotiation has leaked out from somewhere.
@DennisW
Dennis, I’m not sure what the “final word” was on the SATCOM interruption from the airplane and equipment manufacturers. Like just about everything with MH370 there’s a few possibilities with some being deemed more likely than others. As Victor has noted many, many times the disappearance remains conjectural because there is no single scenario that completely, logically and unequivocally explains all the observations; if that weren’t the case whatever would we be doing with our spare time?
Gysbreght said: The airplane can be navigated between waypoints by reference to the Navigation Display in any autopilot mode other than LNAV, and also while flying manually with autopilot off.
There is an important qualifier. Waypoints further than 40 NM will not appear on the ND unless they are part of the active route, or a navigational aid, or an airport. Waypoints closer than 40 NM will appear on the ND if the range is set to 10, 20, or 40 NM and WPT mode is selected. Therefore, unless the waypoint is entered in the FMC as part of a route, it would be difficult to use the ND to fly towards a waypoint (such as VAMPI) that is more than 40 NM away. You’d have to somehow get within 40 NM and then fly towards it.
That’s got some interesting implications. First, I do believe that after the turn at IGARI, the plane was in HDG (or TRK) SEL mode and the pilot was using the ND to fly towards (but offset from) Kota Bharu and Penang Airports. That would be possible even when more than 40 NM from either airport.
As for flying towards VAMPI in HDG SEL mode, that would only be possible when within 40 NM. However, looking at the Lido radar data, the track of captures before the “hole” is different than the track of captures after the “hole”. Perhaps there was a change in course when VAMPI was within 40 NM, and HDG SEL and the ND could be used to navigate towards it.
That said, entering VAMPI as a waypoint in the route and using LNAV certainly seems a lot easier.
@VictorI
‘As I have said many times, there is no scenario in which all actions are easily explained.’
Then my very speculative (old) scenario comes up in my mind again:
The plane was shot at by the military at ~18:22 taking out the right engine forcing the APU to start initiating the re-log-on (while left IDG was isolated).
The plane had to descent to ~25.000ft with one engine operating.
During this descent the SAT-call at 18:40 catched the plane in the descent.
Maybe only to check if the plane was still in the air or not.
Then the flight continued on only one engine around ~25.000ft till the end.
@Ge Rijn: The endurance is possible only if fuel was drawn from both tanks. How did the tank feeds get re-aligned?
@DrB
“@TBill,
You said: “Those favoring FMT south at IGOGU like a slow down as that makes the 18:40 BFO and later points fit better.”
Actually, that is not the case if you use the 15 NM SLOP….”
OK I take it back for the moment. I have not really studied 18:25-18:28 BFO thinking Victor’s slop move fit the data, and all I have on it is your 9-Jan-2017 data sheet where you give speed and many great data details.
We need move proposed flight routes with details, in the manner you showed on 9-Jan. Maybe Victor can have a blog article discussion on 18:22 to 19:41 now that we are taking Lido as real data.
@TBill
My Google Earth log-on screen has a big white circle with a question mark label in the area to the West of Penang. My habit of starting flight path analytics at 19:41 reflects my belief that the flight path between 18:22 and 19:xx cannot be inferred from the information we have.
@Dennis
I agree with you. I am not even sure 19:41 is the end of maneuvers…19:41 might have been close to the end of any maneuvers.
@Victor
In the article, B466 right? (vs. R466)
whereas Andrew had mentioned earlier the A & B airways are older and not used too much.
@TBill: Thanks. I fixed it.
@Andrew,
@VictorI,
I am working on including LRC INOP and Holding INOP in my fuel model. Is it possible to transfer fuel from one wing tank to the other wing tank? Any idea of the transfer rate?
@VictorI,
If the SDU were depowered for a considerable period of time before the 18:25 log-on, the time when power was reapplied may be closer to 18:22 than 18:24 because of the longer warm-up time.
@Andrew,
Regarding those six steps fitting a negotiation scenario, I agree. Unfortunately, we have no clear evidence to confirm that any of the six steps actually occurred.
@DrB: You can’t transfer fuel from one tank to another. All you can do is balance by crossfeeding, i.e., open the crossfeed valves, turn on the pumps for the heavy tank, and turn them off for the light tank.
Regarding the SATCOM power up time for the log-on at 18:25, yes, I should have been more exact. The IFE log-on activity probably means it took longer to thermally stabilize than at 00:19.
@DrB said: If you want to believe that VOCX was the next destination after MEKAR, then you have to accept that VOCX was entered into the active route just as MEKAR was overflown, not before, so that the turn to VOCX began after MEKAR, not before.
There is an interesting possibility that MH370 was in LNAV mode with VAMPI to MEKAR as the active leg. If at some point VOCX was added as the destination airport, a discontinuity would be inserted after MEKAR, which would cause MEKAR to be overflown. If subsequently the speed was reduced to 225 KIAS and the discontinuity was cleared, you would get a path on a track of about 306° that satisfies the BTO and BFO at 18:25 – 18:28.
If you constrain the calculated BFO value to exactly match 144 Hz at 18:28 and then find the track and speed of a path starting at the last radar position at 18:22 that minimizes the BTO error, strangely the speed of that path is right around 225 KIAS (389 KTAS, FL340, ISA+10K) and the path points to VOCX. If you find the LNAV path that satisfies the satellite data after 19:41 with LRC speed and on a great circle ending at NZPG, again that great circle passes VOCX. VOCX is also close to the 2nd arc, and VOCX-NZPG defines a great circle that is close to tangency of the 2nd arc. So VOCX is the intersection in space of the 18:22-18:28 path extended forward in time, and the McMurdo 19:41-00:19 path extended backwards in time. What connects the paths in time is a holding pattern at VOCX.
@DrB
If the SDU were depowered for a considerable period of time before the 18:25 log-on, the time when power was reapplied may be closer to 18:22 than 18:24 because of the longer warm-up time.
If the SDU were depowered my bet would be near or shortly after the turn at IGARI. If Shah really was that “tuned in” to the SATCOM operational details he would know that calls (which he would well expect to be made to the aircraft) would indicate the aircraft had gone down near IGARI. Repowering at 18:xx could logically be associated with expecting contact via the SATCOM from co-conspirators. I know – thick smoke, mirrors in every direction, and pure speculation.
Still, it seems unlikely that a pilot would be that familiar with the operational nuances of the ISAT system. Simply no reason for any training in that domain.
@Victor
If you constrain the calculated BFO value to exactly match 144 Hz at 18:28 and then find the track and speed of a path starting at the last radar position at 18:22 that minimizes the BTO error, strangely the speed of that path is right around 225 KIAS (389 KTAS, FL340, ISA+10K) and the path points to VOCX. If you find the LNAV path that satisfies the satellite data after 19:41 with LRC speed and on a great circle ending at NZPG, again that great circle passes VOCX. VOCX is also close to the 2nd arc, and VOCX-NZPG defines a great circle that is close to tangency of the 2nd arc. So VOCX is the intersection in space of the 18:22-18:28 path extended forward in time, and the McMurdo 19:41-00:19 path extended backwards in time. What connects the paths in time is a holding pattern at VOCX.
I like it.
@VictorI,
Suppose one engine shut down some time after 18:22 but when there was considerable fuel remaining in both wing tanks. Am I correct in concluding that human actions would be required, at least once, at a later time, in order for the remaining engine to use all the fuel on board?
@DennisW,
Regarding the pilot’s’ knowledge of the satcom system, I would think they would know it needed to be turned on for the satphone and IFE to function. I doubt they knew anything about periodic log-ins, much less BTOs and BFOs.
@VictorI,
Why would inserting VOCX after MEKAR create a route discontinuity?
@DrB: I suppose opening the crossfeed valves would draw from both tanks and keep things relatively balanced without intervention. @Andrew would know if this is practical. I suppose it wouldn’t matter if one tank went dry before the other, so balance might not be critical.
But that raises another question: With two engines operating, can the crossfeed valves be opened to ensure one engine doesn’t flameout well before the other? @Andrew?
As for you other question, if the route is VAMPI-MEKAR, and if VOCX is chosen as the destination airport, then after an approach and runway are chosen, there will be a discontinuity created after MEKAR because MEKAR is not part of the standard arrival for VOCX.
@Victor
@DrB
Re DrB’s question: “Am I correct in concluding that human actions would be required, at least once, at a later time, in order for the remaining engine to use all the fuel on board?”
Yes, either crossfeed switch would need to be selected ON and the boost pumps on the low tank selected OFF to crossfeed fuel from the high tank.
“I suppose opening the crossfeed valves would draw from both tanks and keep things relatively balanced without intervention.”
In practice, the boost pumps have allowable variations in output pressure, so one side tends to have a slightly higher pressure than the other. Consequently, the fuel tends to feed from one side if all the boost pumps are left on. That’s why the Fuel Imbalance checklist directs the crew to turn off the pumps for the tank with the lowest quantity.
“I suppose it wouldn’t matter if one tank went dry before the other, so balance might not be critical.”
If one side goes dry before the other, then the boost pumps on the dry side would obviously start to lose pressure. As soon as that happened, the pumps on the other side would supply fuel to the operating engine. There are fuel balance limitations, but they are mainly for structural life and landing gear considerations, not controllability. A large fuel imbalance could also result in increased trim drag and higher fuel consumption.
“With two engines operating, can the crossfeed valves be opened to ensure one engine doesn’t flameout well before the other?”
Yes. That would probably cause a large fuel imbalance as discussed above.
@DrB @Victor
I confused if you are questioning the 15 nm Slop or just saying what other explanations are possible? I like the Slop.
According to the JoN paper, 18:25:27 corrected BTO is 12520. At 18:28 three minutes later we are still at 12500, so we appear to be heading north parallel to Arc1, exactly as shown graphically by DrB in his 9-Jan-2017 and 27-Feb-2017 papers. The BFO’s shift from 170 to 144 just like we might expect for a turn in progress that is pointing closer to 360 degrees, with some minor warm-up decay yet in the SDU), then curves over to about 300 degrees for a 144 BFO reading, to use the words of the Jon article.
I have not calc’ed speed but I am not thinking I have to slow down too much. JoN paper I think says 480 speed out to IGOGU, and that speed goes up a bit if you add the slop in there.
We do have UAE343 on our tail on N571, so that is one logical explanation for the maneuver.
@TBill: What prompted the alternative to the SLOP is that the Lido radar image shows no evidence around MEKAR of a turn towards NILAM on N571. Rather, it looks as though the plane continued on the same track past MEKAR. (It is hard to know for sure as we have few captures around MEKAR, and what we have is of questionable accuracy.) So, I looked for the speed and track after 18:22 that would satisfy the BTO and BFO. What I found is the BFO would be exactly matched at 18:28 and the 18:25-18:28 BTO sequence would have the expected error if the track pointed to VOCX and the speed was reduced to 225 KIAS at FL340, resulting in 389 KTAS with an ISA+10K temperature. VOCX-NZPG is also a great circle path that fits the satellite data after 19:41.
Victor: KTAS?
[Yes. Fixed.]
@Victor
OK but we had this data all along, right?…and that speed seems so slow to me. I fly that route fast because I figure I gotta go out and back before I get to where I am going at 19:41 or shortly thereafter.
@dennisw – sdu depowering
I suspect he received instructions from that avionics engineer he called just before takeoff.
@MH
Re: “I suspect he received instructions from that avionics engineer he called just before takeoff.”
And I suspect that you are butchering the facts. The Captain spoke to Malaysian Airlines Licensed Aircraft Engineer, Zulhaimi Bin Wahidin, well over a month before MH370 and he didn’t call Zulhaimi, Zulhaimi called him.
@VictorI @DrB
@Andrew already your question; ‘Is it possible to transfer fuel from one wing tank to the other wing tank?’
I thought it would be nice to illustrate the process with this chort vid:
https://www.youtube.com/watch?v=ViGgDWA306A
@Mick Gilbert: The inconsistency about who called whom is one of the many mistakes in the report. The captain called the engineer according to some of the graphics that look to be computer generated and less likely to be wrong.
@TBill: The speed 225 KIAS is above minimum maneuvering speed and is not “too slow”. The timing is also fine as a loiter is still required at VOCX.
TBill,
BTO at 18:25:27 is an R600 channel packet, and even after correction, it can be in error by 100 to 200 microsec (as seen in the MH371 logs).
R.E. turn at MEKAR, I doubt the data on the Lido image are good enough to say anything definitive. My digitized last radar point (from the Underwater Search Fig. 2 image) falls between the MEKAR-IGOGU and MEKAR-ANOKO tracks, being closer to the former, but the error is large enough that it could be compatible with either.
@Victor
Thanks Victor, the report can be a tad abstruse on who called whom, however, it is very clear when the call was made and it wasn’t “just before takeoff” as MH claimed.
@Mick Gilbert: I agree that there is no evidence that the captain called the engineer just before takeoff. What we know is the two talked on Feb 2, likely the captain calling the engineer, and just before takeoff the captain’s cell phone logged into WeChat.
This news from AirlineRatings on the OI offer. I’ll now say that I have heard the same news from a very reliable source:
MH370: Malaysia tipped to announce search deal
Malaysia could announce as early as this week that it is resuming the search for missing Malaysia Airlines Flight MH370.
The Malaysian government has been looking closely at search offers from private companies and an announcement is expected soon, according to sources linked to Kuala Lumpur.
A “no-find, no-fee” offer by US company Ocean Infinity is understood to be the winning candidate although Dutch company Fugro, which was involved in the original sweep, is believed to have countered with a low-fee proposal.
Ocean Infinity champions a system that uses six HUGIN autonomous underwater vehicles capable of operating at depths of up to 6000m to collect high-resolution data at what it says are “record-breaking speeds’’.
Malaysia’s Department of Civil Aviation was asked recently by Transport Minister Liow Tiong Lai to comb through the recently released Australian Transport Safety Bureau final report on the search for “credible evidence” on the crash site.
A two-year sweep of the original 120,000 sq, km. search area failed to find any signs of the wreckage and was ended by the Malaysian, Australian and Chinese governments in January amid criticism by experts it was abandoned too soon.
Since the decision to suspend the search, new findings from drift modeling and satellite imagery have led Australian scientists to believe they have pinpointed the probable site for the wreckage of the plane with unprecedented accuracy.
The studies have significantly boosted confidence that the wreckage of the Boeing 777, which went missing in March 2014 with 239 souls on board, is in the southern half of a 25,000 sq. km. search area identified by experts in 2016.
The scientists used satellite imagery and drift modeling to identify a location at latitude 35.6°S and longitude 92.8°E, near the seventh arc defined by satellite data, as the most likely location for the missing plane.
The 440-page ATSB report being examined by the Malaysians chronicled the extraordinary efforts by investigators and scientists to find the missing Boeing 777-200ER and noted that the understanding of where MH370 is located was now “better than ever”
“It is almost inconceivable and certainly societally unacceptable in the modern aviation era with 10 million passengers boarding commercial aircraft every day, for a large commercial aircraft to be missing and for the world not to know with certainty what became of the aircraft and those on board,’’ it said.
@Victor
Who will play the role of “Search Area Advisor” i.e. SSWG for the Ocean Infinity effort? Is it Ocean Infinity’s intent to make these decisions internally?
@DennisW: If they are smart, and I think they are, they will solicit advice from a variety of sources, including the ATSB.
@Victor: Re. “What connects the paths in time is a holding pattern at VOCX.”
Do you have any thoughts on why the pilot might have entered into a holding pattern?
It might be prudent to be cautious about Shah’s intentions. He could have left the McMurdo great circle data on his hard drive in order to deliberately lay a false trail. What if he had initially set up the route VAMPI, MEKAR, ANOKO then shortly after passing MEKAR, inserted IGOGU in place of ANOKO? Would that satisfy the BTO and BFO at 18:25 and 18:28?
@sk999
Re: BTO OK thank you…how did you ever feel about the proposed SLOP move?
Seems to me on the LIDO image we have data scatter on the VAMPI to MEKAR track, so to say that same degree of scatter, once past MEKAR, is a sign of no turn is a stretch.
Also we are trying connect distant radar data, presumably with some error, with distant satellite data, with presumably some error, and then trying to come to conclusions as if the data were perfect. But there may be an understandable discontinuity at that 18:22 – 18:25 juncture. It’s amazing the data matches as well as it does.
I suppose the pilot could be doing manual line up of waypoints too. Victor’s graphic above for MH150 certainly shows the pilots seem to bob and weave along N571.
@Rob: Regarding the reason for a holding pattern, I have nothing more to offer than what others have said. I note that a holding pattern near an airport gives the pilot options, although some of those options are limited by nightfall and the lack of an ILS, albeit runway lights are available. By my calculation, sunrise would have been 23:40Z (5:10 am local).
MEKAR-IGOGU is essentially the same track as MEKAR-NILAM. On that track, you need either the SLOP or a deceleration and descent (to allow a stable, lower TAS) in order to satisfy the BTO/BFO.
Of course we have to be careful about the captain’s intentions. It is an error to place either too much or too little weight on the sim data in deriving possible paths.
Peter Norton,
As it was my comment that prompted your question regarding ACARS I’d like to put this into context. I was referring to the initial stages of the flight and specifically a thirty-minute window from Top of Climb to when the next routine ACARS condition report failed to send. This would have been when the aircraft was established on its track towards Penang.
ACARS has no ON/OFF switch. It becomes operational when electrical power is established. On power up, VHF and Satcom are available. The system uses VHF when in range, and when out of range switches to satellite. This system logic, however, can be overridden. As an example, some countries’ VHF relay costs are higher than when using Satcom, so pilots deselect VHF, and the system uses Sat.
At the stage of flight I was referring to MH370 was in VHF range. Deselecting VHF would have left Satcom operational. In this configuration, the Line Select Key corresponding to Satcom is deactivated. If the other way around and VHF selected and Satcom deselected, then the LSK for VHF is deactivated. You can’t deselect both, or in simple terms, you cannot disable ACARS by pressing both VHF and Satcom Line Select Keys on the CDU.
This was directed more at contributors who seem to think disabling ACARS is as simple as flicking a switch. The system is far more complex with many layers and if Zaharie, or even his father, was the lone pilot in the two-crew cockpit, as some claim, it would require some deft and time consuming work on the circuit breaker panel, while manoeuvring the aircraft on a radically new track.
It was a small point but when a routine flight loses comms, SSR, and a complex system such as ACARS, turns right, and then left, and manoeuvres to track towards a suitable emergency airport, it would be a perfect fit for a six monthly LOFT session.
what is a “LOFT session”?
Line Oriented Flight Training (LOFT). LOFT is conducted as a line operation and allows for no interruption by the instructor during the session except for a non-disruptive acceleration of uneventful enroute segments. There are two types of LOFT: (1) Qualification LOFT.
thanks @ALSM. I was guessing it must be that an not:-
Living Our Faith Together
or
Lack of Freakin’ Talent (polite form)
@Paul Edgley wrote “you cannot disable ACARS by pressing both VHF and Satcom Line Select Keys on the CDU”.
True, you cannot use a CDU. One uses the ACARS Manager screen presented on the MFD & the CCD to uncheck & deselect the unwanted datalink. But to be really sure, just dial in an incorrect frequency on the VHF radio selected for datalink.
At worst two actions, both within barely a stretch from either flight crew seat.
The full satcom log records an ACARS Media Advisory message at 15:54 that shows ‘lost VHF’ and only SATCOM remaining available:
LV155453S
with subsequent Media Advisory messages showing ‘established SATCOM’ and, again, only SATCOM available, eg:
ES155607S
If VHF was available but not used, the trailing characters in these msgs would read ‘VS’ not simply ‘S’.
@Victor
That’s good news about the search continuation chances. I guess that changes my approach a little bit. If all progress were to stop, then I’d be looking for more serious criticism of the accident and the search stoppage.
@Mr Edgley: I think the screen you refer to is not correct for the version of AIMS on 9M-MRO. In the MFD/COMM/MANAGER/ACARS screens, in earlier version of the AIMS, you had to select between AUTO, VHF, and SATCOM, and could not deselect all (although even with this version there are ways to disable VHF and SATCOM). In later versions of the AIMS, you could individually select/deselect SATCOM and VHF links for ACARS.
The documentation I have seen specifically refers to 9M-MRO, so we don’t have to guess about AIMS versions.
So actually, disabling the ACARS function really is as simple as deselecting the links in a single screen. VHF was probably already deselected, as Don said. That means just deselecting one option. It would not “require some deft and time consuming work on the circuit breaker panel”, as you claim.
@Paul Edgley
What type of aircraft are you referring to? Clearly not a B777. As Don noted, on the B777 ACARS is managed via the ACARS Manager pages on the MFD, not via a CDU. I also note the following:
1. It is possible to deselect both VHF and SATCOM via the ACARS Manager page 2/2. The following is a quote from the FCOM:
This page allows the operator to select/deselect VHF or SATCOM transmission of data. ACARS is set to auto mode (both boxes selected) at power-up or during a manual data communication system reset. Normally, this permits ACARS to automatically use VHF or SATCOM (if VHF is unavailable). If both boxes are deselected, ACARS loses the capability to send downlink messages, but can receive and display uplink messages.
2. On the B777 it is not necessary for pilots to deselect VHF when overflying countries with high VHF relay costs. That function is performed automatically in the background by the ACARS manager software.
@Andrew: There is an earlier version of AIMS (pre-2003) where on that ACARS page (MFD/COMM, not CDU, as you say) you can only select between AUTO, SATCOM, and VHF, so one is always chosen. However, in later versions, you could individually de-select the VHF and SATCOM links, as you describe. The documentation we have is specific to 9M-MRO, and describes the version where you can de-select all links.
@Victor
I’m aware of that Victor. The previous version was AIMS I. The ACARS Manager page I described is found in AIMS II. My understanding is that 9M-MRO had the newer version of AIMS, as you noted.
@sk999,
If you will post a link to your digitized radar positions, I can overlay some example pre-turns and post-turns at MEKAR to various post-18:22 bearings (to ANOKO, NILAM/VOCX, SLOP along N571, etc.). I have a 1-second resolution turn model that utilizes bank angle limits. That overlay may allow a better assessment of the compatibility of the radar track data with the turn options we have identified.
Ge Rijn,
Re: “To be short; 30S is too far north according his model and the amount of drifters in the model is way too high regarding actual reality.”
Why do you think 30S is too far north?
On contrary, I believe it might close to the southern limit based on the drift analysis. The most likely segment is centered at:
– 28S based on the spatial distribution only;
– 30S based on the flaperon-barnacle analysis assuming the drift angle of 18deg to the left (DGA);
– 27S based on the flaperon-barnacle assuming the drift angle of 32deg to the left (DGA);
– 31S assuming zero drift angle;
– 26-27S based on the failure of the aerial search, or around 29-30S if the observed (but not recovered) objects were related to MH370;
– 28.5S if the towelette discovered at the Thirsty Point was from MH370;
– 34S based on the arrival time of RR fragment.
@Andrew,
Re: “FL200 provides the best endurance at weights around 200-220T. You might decide to descend to FL200 and hold if you wanted to stay airborne as long as possible or weren’t sure about how long you might need to hold. As for circumstances, I can’t think of any that might apply in this case, except a hijack scenario perhaps, but as I said above, that doesn’t seem likely.”
In my opinion this may also fit into the technical failure scenario. On that day the sunrise in Malaysia was around 7:20 AM (11:20 UTC), and crew’s original intent could be to stay airborne at least till the sunrise.
@Dennis,
Re: “As far as I know, no one has explained the 18:25:27 BFO value in the context of a power up. It is in the serendipitous category, and it might well be serendipitous and should be discarded as Holland states My only point, and I should have stated it, is that the 18:25 logon is not in the “we discussed that ad nauseam, so move on” category. I am not at all comfortable with what is commonly regarded here as a “put to bed” issue. ”
I am also not convinced by any existing explanation. There is a total mess with regard to the SDU: LGA vs HGA; one dataset is needed for Doppler, the other dataset is needed for HGA steering; WoW supplied to the SDU for unknown purpose; ground speed marked as optional; missed track angle in the lists of required data; SDU ignoring GS1 channel data except GPS time; mysterious conclusion that the HGA was used based on Inmarsat logs… I suspect the dataset needed for Doppler is supplied independently on the dataset needed for HGA steering. I suspect the former is sourced from the FMC while the latter – from the SAARU. Obviously, neither ground speed, nor track angle is needed for the steering, and that might explain the mark ‘optional’ and omitting the “track angle”. Can the LGA-HGA issue help in explaining 18:25:27 BFO? I don’t know, but this would be something interesting to explore.
@Oleksandr said: 26-27S based on the failure of the aerial search
Why do you say this? If I look at Fig 1.1 of CSIRO IV, an impact at the part of the 7th arc between 26S and 27S has a high probability of detection by the surface search.
Bobby Ulich,
I have posted various digitization files here:
https://docs.google.com/document/d/14hleZyx1pUPL44yaeHKt6jnSQ3DbgRq2zibbKkFLq2c/edit?pref=2&pli=1
Look in the file labeled “Data Source” (updated Jan 8, 2017). It has links to “radar-map.dat”, which is my digitized version of “Underwater Search” Fig. 2, and “lido.dat”, which is my digitized version of the Lido hotel image. However, the latter is probably not good enough to evaluate the turn (or lack thereof) at MEKAR.
I have measured the position of the last radar point off multiple versions of the radar path, and they all differ by a bit. The position I use for modeling flight paths is at 6.593, 96.289. The last point in radar-map.dat is at 6.600, 96.302. They are separated by 3 nm, and probably give a good indication of digitization accuracy. The first one falls between the MEKAR-NILAM and MEKAR-ANOKO tracks. The second is squarely on airway N571.
@Victor
Reports appearing on the Net speculating that Malaysia about to announce resumption on search, as early as Wednesday. Steve Creepy and Geoffrey Thomas reporting for the West Australian. They say Fugro have come in with a low cost offer in attempt to undercut Ocean Infinity. Also say “global experts” have been pressing for a resumption, and ATSB are said to be keen for search to resume.
We’ll see, as they say.
@Victor
Correction: that’s Steve Creedy, not Steve Creepy! The autocomplete making a Freudian slip perhaps?
@Oleksandr
Re: “In my opinion this may also fit into the technical failure scenario. On that day the sunrise in Malaysia was around 7:20 AM (11:20 UTC), and crew’s original intent could be to stay airborne at least till the sunrise.”
I’m curious to know what kind of technical failure would force the crew ‘to stay airborne at least till the sunrise’.
TBill,
R.E. SLOP maneuver, I am highly skeptical of any complex solution invented to explain one data point, but there is no real way to prove it right or wrong. Further, it doesn’t impact the derived path unless it provided a mechanism to get the aircraft into a state that it would not ordinarily be in.
R.E. scatter in Lido data points, I agree with your assessment. Also, regarding whether the plane was in LNAV mode and following waypoints, at one time I thought it was, particularly while it was crossing peninsular Malaysia, but now I am skeptical as well. While the plane passed close to several waypoints, it missed others (e.g., LOSLO) that would seem obvious targets, and the turn around Penang would have required a fine grid of points that we don’t know about. Having said that, one can make a rough calculation of whether a random path would have followed an airway such as N571 by chance as closely as it did, and the answer is it would be very low probability. A pilot flying waypoints manually would explain many oddities about the route that was flown. It would be an interesting simulator exercise to see if it could be flown in practice.
All,
Regarding the version of AIMS on 9M-MRO, according to the Final Report (p. 13), the aircraft was manufactured in May 2002. According to Honeywell, “AIMS-2 entered service in Q4 2003.”
https://aerospace.honeywell.com/en/products/cockpit-systems/airplane-information-management-system
So that would seem to say that 9M-MRO had AIMS-1. Can we get further evidence? Yes. According to Factual Information, AIMS is described as having, “Four Input/output modules (IOM)” amd “Four Core processor modules (CPM)”. That’s eight by my count. According to this document:
http://www.aviationtoday.com/2003/06/01/boeing-integrated-avionics-takes-another-step-forward/
“The new AIMS-2 system actually packs 10, rather than eight, modules into each cabinet …”
If Victor’s description of the ACARS screen on 9M-MRO is correct, it would imply that the change from the old-style screen to the new-style screen happened within the AIMS-1 product line and was not specific to AIMS-2. Feedback welcome.
@Victor
I’ll now say that I have heard the same news from a very reliable source:
Looking at this news in more detail, it is sketchy. Every single article uses phrases like “it is expected”, “it is reported”, “could announce”, and so on. Not a single reference to any person in a position of credibility and authority in any tripartite role. Most of the stories seem to link back to “The Western Australian” – a publication with a questionable reputation. At the moment, I will put this news in the rumor category.
@sk999,
Thanks for posting the link, but none of those files appear to be what I want. I want the lat/lons of the points shown on the Lido slide – not somebody’s interpretation of a continuous path, based on some assumptions or preconceived notions (like was done in ATSB’s Figure 2).
Way back when, someone overlaid the Lido slide on a SkyVector map and marked the points with X’s. I have a picture of this, but have lost track of its source.
Does anyone have the coordinates of the raw Lido slide military radar locations?
@Victor
Re: “The ACARS Manager page I described is found in AIMS II. My understanding is that 9M-MRO had the newer version of AIMS, as you noted.”
My apologies, 9M-MRO was an AIMS-I aircraft. I believe it was an earlier block point version of the AIMS software that required at least one ACARS transmission pathway to be enabled. Later block point versions deleted the ‘Auto’ selection and allowed both VHF and SATCOM to be disabled. The AIMS software load on 9M-MRO must have been one of the later versions, given that the FCOM applicable to that aircraft states that both VHF and SATCOM can be disabled.
@Victor,
Re: “Why do you say this? If I look at Fig 1.1 of CSIRO IV, an impact at the part of the 7th arc between 26S and 27S has a high probability of detection by the surface search.”
The segment from 26 to 27S corresponds to low-coverage based on my drift model. I took a look at Fig 1.1 of CSIRO IV: their model shows similar ‘gap’ for low-windage objects slightly northward, between 24 to 25S. It also appears that CSIRO’s “high-windage results” are very sensitive to the exact location between 25S to 27S (if you click link in their pdf; see non-flaperon items).
I would guess we have this difference because of the wind. CSIRO used ECMWF while I used GDAS; in my experience ECMWF often predicts about 20 percent lower wind speeds compared to GDAS, which could be a reason.
Interestingly, CSIRO’s model also shows a narrow ‘gap’ at 30S.
@Andrew,
Re: “I’m curious to know what kind of technical failure would force the crew ‘to stay airborne at least till the sunrise’”.
You know my line: what about damaged nose landing gear + failed or malfunctioning ADIRU? Would not it be better to stay airborne till daylight?
@Andrew: No need to apologize. It doesn’t matter whether 9M-MRO uses AIMS I, AIMS II, 2003 AIMS, NOT 2003 AIMS…we have documentation specific to 9M-MRO that says that we can independently select/deselect the SATCOM and VHF datalinks. I think we all can agree on this.
@Oleksandr: Thank you for your comments about the surface search. Mostly, I am trying to better understand the various claims about the efficiency of the search along the 7th arc. As we have discussed before, if as efficient as claimed by AMSA, impact above 33S is unlikely. However, I question these efficiency estimates.
@DennisW: Let me be clear. When I said I had heard the same report, I was referring to reports saying the announcement of a deal is expected this week, just as reported in AirlineRatings (and later in the West Australian). I don’t think anybody has said that the deal is done. Rather, there is reason for optimism. I would assign a probability of greater than 50% of sealing the deal this week.
@Victor,
I’ll take that bet for a beer at 50:50.
The deal is too complex – exclusivity term, wreckage verification, escrow arrangements, governing law,…
A time and material contract to Fugro would make more sense if indeed Malaysia were sincere in resuming the search.
BTW, I truly hope you are correct.
@Victor
Re: “It doesn’t matter whether 9M-MRO uses AIMS I, AIMS II, 2003 AIMS, NOT 2003 AIMS…we have documentation specific to 9M-MRO that says that we can independently select/deselect the SATCOM and VHF datalinks. I think we all can agree on this.”
Absolutely. I wanted to correct the garbage I posted at 2.41am (my time) on a sleep-deprived night! My earlier quote came from the same document.
@Victor
Just returning to the matter of who called whom with regards to the 2 February 2014 telephone call between Zulhaimi and Zaharie, the following from the leaked RMP report, Folder 4 SKMM Analysis, are pertinent:
“The analysis on the phone call made and received by MH370 Pilot showed that he received noticeably long call duration from 019-3394874, registered under Zuihaimi Wahidin, an aircraft engineer who works for Malaysian Airlines System Berhad. The call was made on 2 February 2014 at 9:49 am for 45 minutes.”
Executive Summary, Page 1 of 24
“The longest call received by MH370 Pilot is from 019-3394874, registered under of Zulhaimi Bin Wahidin (IC No:
671028-07-5029) on 2 February 2014, which lasted for 45 minutes.”
3. ANALYSIS ON MH370 PILOT’S BACKGROUND, 3.1. Phone records, b. Longest calls, Page 5 of 24
The graphic at APPENDIX J-11, Page 6 of 11, clearly shows the call path for the 2 February 2014 Zulhaimi call as being from Zulhaimi to Zaharie. Similarly, the arrow annotation on the graphic on Page 4 of 8 shows the same thing with regards to the direction of the call (viz from Zulhaimi to Zaharie).
There is only one inconsistency with those four separate references to Zulhaimi having initiated the call and that appears on Page 18 of 24 of the main report;
“Further checking on Zaharie phone call detail records, we noted that he made a call to Zulhaimi Bin Wahidin, whom we later confirmed is a Licensed Aircraft Engineer.”
I think that on any reasonable assessment it would be safe to conclude that the 2 February telephone conversation between Zaharie and Zulhaimi was most likely initiated by Zulhaimi.
@Mick Gilbert: Yes, you are right. I read one of the graphics incorrectly. Despite the inconsistencies, I agree that likely the call was initiated by the engineer.
@Victor
No problemo, I’ve misread or missed any number of matters in those folders. One thing that we can probably definitely agree on is that the RMP report was either poorly written or poorly translated, possibly both.
@Mick
“The call was made on 2 February 2014 at 9:49 am for 45 minutes.”
So what did they talk about for 45 minutes? Does it not seem strange that the engineer’s recall of the call’s subject was not included in the RMP report?
@DennisW
There are no records of interview at all in the leaked report so, no, not strange (annoying but not strange). What we have appears to be the technical and forensic report rather than the investigative report.
As to what they spoke about, that’s anyone’s guess. In terms of common background, apart from the obvious MAS connection, both Zaharie and Zulhaimi attended the Penang Free School, a very well respected (some might say “elite”) English-medium school of currently only 1800 students. Maybe the time honoured application in education systems of alphabetical ordering had brought them together at some point in their school days.
@Mick
Well, my guess is that Zulhaimi was not getting more info on Shah’s DIY videos. The details of that conversation could be a hugely important piece of information.
@Paul Edgley,
@Don Thompson: thank you for picking up on the ACARS issue
@Mick Gilbert,
@Victor Iannello: could you confirm if I understood you correctly here. thank you
@Mick Gilbert
On page 1 Of 24 of the main report we have this:
“The longest call received by MH370 Pilot is from 019-
3394874, registered under of Zulhaimi Bin Wahidin (IC No: 671028-07-5029) on 2 February 2014, which lasted for 45
minutes. Zuihaimi works for Engineering Department of
Malaysian Airlines System as Licensed Aircraft Engineer.
However,
further analysis on mobile phone records from
January to March 2014 showed no indication that the two
have cailed Bach
other.
Record also showed that Zulhaimi
made an attempt
to call Zaharie on 8 March 2014, when the
aircraft was announced to be missing”.
So it is incorrect to say that the only inconsistency is the one you point to on page 18.
I would not say that ‘any reasonable assessment would be safe to conclude’ that Zulhaimi was the one to initiate the Feb 2nd call.
@Victor,
If it’s true, am glad I was wrong in saying before that MY would never entertain a new search.
http://www.perthnow.com.au/news/western-australia/hunt-for-missing-flight-mh370-to-resume/news-story/ce350aff3a290e2476004eb04e7a82af
A recent article
“We wont be deciding anything now on whether we are embarking on a new search or not,” Liow told reporters on the sidelines of an event in Kuala Lumpur.
“We have to discuss with the companies. It will take some time as it’s some detailed discussions,” he said.
http://mobile.reuters.com/article/amp/idUSKBN1CM0A3
@Donald
Page 18 of 24 of the main report is the only inconsistency with regards to the 2 February 2014 phone call. If you want to submit more evidence of Zulhaimi initiating calls to Zaharie as some sort of refutation of Zulhaimi initiating the 2 February call, knock your socks off, it only serves to bolster my argument.
@Peter Norton
Sorry Peter, I must have missed that earlier post. Yes, you’ve got it correct; disabling the transmission path for ACARS and logging the SATCOM off are two different processes, only the latter produces a log-off to the ground station.
With regard to the phone calls between MAS Engineer Wahidin and Captain Zaharie:
It is well known that Captain Zaharie was a tech-nerd, and most significantly, that he was a meticulous note taker. This later fact is proven beyond doubt, (at segement (G) ) in the
4 Corners Program (45 minutes 37 seconds) where the former Chief Pilot of MAS, (Captain NIK HUSLAN) is interviewed. Nick is in the following video time slots.
(minutes:seconds)
(A) 11:33 – 11:44 – 11 seconds
(B) 12:44 – 12:57 – 13 seconds
(C) 12:44 – 12:58 – 14 seconds
(D) 13:53 – 13:59 – 6 seconds
(E) 13:47 – 13:59 – 12 seconds
(F) 15:26 – 15:40 – 14 seconds
(G) 16:57 – 17:05 – 8 seconds
(H) 23:41 – 23:52 – 11 seconds
(I) 24:15 – 24:30 – 15 seconds
(J) 44:32 – 44:46 – 14 seconds
I have always taken all of NIK HUSLAN’s statements to be both specific and direct, and thus, significant. The most significant, in my view, is segment (G). We must not ignore NIK HUSLAN’s comments, particularly segment (G).
The fact that the first, and most of the subsequent calls, were initiated by the enginner, suggests the possibility, that at some time previously, Zaharie may have asked Wahidin to “dig into the engineering and maintenance manuals” to find out some inner or finer detail(s) of some sytem(s) that were not in the “flight crew manuals” routinely available to him.
Teasing out such requested details, may have taken Wahidin some doing, indeed, some time. It is likely that Wahidin reported his findings to Zaharie, and that Zaharie either requested further information on more systems or deeper system information than Wahidin had “initially” found and provided. The fact that these phone calls were long (45 minutes) implies to me, that the discussions were “detailed”, perhaps to the extent of engineer Wahidin “reading segments of manuals” to Zaharie over the phone, and him taking “detailed noted” in the process.
Given the time frame of these calls (in terms of dates) if we assume that the simulator was functional (although it has been assumed that it was not) it is possible that Zaharie may have fiddled with the simulator to see if the information “worked” in the simulator.
So, the issue of the simulator’s functionality or not in this time frame is an issue itself. There were multiple hard drives. I have multiple hard drives (all bootable) myself. It is not inconceivable that although his “upgraded” system may have crashed, that he could have put an earlier version back in to do his checks, then, satisfied or not, removed that disk and put the upgraded one back in, and continued to attempt to debug it. This possibility needs to be examined. I have done the same thing myself in the past.
Clearly therefore, the “content” of these phone calls between Wahidin and Zaharie “is critical”.
Obviously, the Malaysian Police would have interviewed Wahidin, probably multiple times. A leak of a “record of interview” is what we need now.
@VictorI @Andrew
About the ACARS. Do I understand you well when I conclude ACARS transmissions will be disabled both via SATCOM and VHF, when the left IDG/bus is isolated without sending a notification?
Or does it switch automatically to VHF when that happens?
.. Which implies VHF stayes available for ACARS after isolating the left IDG/bus..
> Mick Gilbert says:
> Sorry Peter, I must have missed that earlier post. Yes, you’ve got it correct;
> disabling the transmission path for ACARS and logging the SATCOM off are
> 2 different processes, only the latter produces a log-off to the ground station.
@Mick Gilbert:
Many thanks for the confirmation. That’s very kind of you.
@Ge Rijn
Re: “Do I understand you well when I conclude ACARS transmissions will be disabled both via SATCOM and VHF, when the left IDG/bus is isolated without sending a notification?
Or does it switch automatically to VHF when that happens?”
The SDU loses power if the L Main AC bus is de-powered. There is no SATCOM log-off message and ACARS via SATCOM is obviously disabled.
ACARS can use either the C or R VHF, depending on the selection in the VHF Manager. If the L Main AC bus is de-powered, both the C & R VHF remain powered. ACARS is still available by VHF if the aircraft is within range of a VHF ground station. The system should use VHF automatically, provided VHF is enabled in the ACARS Manager and the applicable VHF is in DATA mode.
“Which implies VHF stayes available for ACARS after isolating the left IDG/bus.”
Yes, but only if:
1. VHF is enabled in the ACARS Manager,
2. The C or R VHF (depending on the VHF Manager selection) is in DATA mode (ie displays ‘DATA’ in the active frequency window), and
3. The aircraft is within range of a VHF ground station.
>> VHF stayes available for ACARS after isolating the left IDG/bus.”
>
> Yes, but only if:
> 1. VHF is enabled in the ACARS Manager
… which – if I understand Don correctly below – was NOT the case on MH370, correct ?
Don :
« The ACARS Traffic Log, published in the Factual Information report, shows that VHF data link was down. A related Media Advisory msg was sent at 15:54:53 (ACARS terminal timestamp). As a consequence of VHF Datalink unchecked in that same ACARS Manager screen on the MFD.
[…]
at 15:54:53z the ‘VHF ENABLE’ check box was unchecked, resulting in the ACARS Data Comm Mgmt function sending the Media Advisory ACARS ‘Lost VHF’ msg via SATCOM to the SITA ACARS Processor. SATCOM remained the operational datalink.
At a time after 17:07 but before the diversion was initiated the ‘SATCOM ENABLE’ check box was unchecked.
The configuration of which VHF radio may be used for DATA at this time was moot as there was no intention to use it. »
@Ventus45. “Obviously, the Malaysian Police would have interviewed Wahidin, probably multiple times”.
Conjecture as to why there is no evidence of concern about the call should allow the possibility of the comparatively mundane: the engineer called to discuss an aircraft problem he was diagnosing, either because Z was approachable and interested in all things flying or because it was a defect Z had recorded, maybe from his previous flight.
@Oleksandr + whoever is interested in technical failure scenarios:
maybe this is of interest to you:
http://www.iasa-intl.com/folders/mh370/An_MH370_Analysis-of-Likelihoods.htm
http://www.iasa-intl.com/folders/mh370/SDU-query.htm
But can a technical failure not be excluded simply by the fact that
– at first ACARS was silenced + SDU depowered
– and later the SDU was repowered but ACARS remained silent (see below)
?
Gysbreght says:
« it shows that ACARS was silenced, i.e. unable to address a message. ACARS remained silent after the reboot, otherwise it would have sent the messages that were waiting in the pipeline. »
@Andrew
Thank you.
Than when at 17:07 the final ACARS data transmission was send and recieved and the following at 17:37 was not (and none after), while the plane was still in VHF range passing over Malaysia, I suppose your option 3. can be excluded?
@Andrew
I mean your option 3 excludes VHF was enabled (and DATA mode not selected) for the plane was in range of a VHF ground station at 17:37 and 18:07.. Correct?
@Peter Norton
Re: “… which – if I understand Don correctly below – was NOT the case on MH370, correct ?”
Correct. The ‘Lost VHF’ message that Don mentioned seems to indicate that the VHF ENABLE check box in the ACARS Manager was unchecked, so ACARS was unable to send downlink messages by VHF.
@Ge Rijn
Re: “Than when at 17:07 the final ACARS data transmission was send and recieved and the following at 17:37 was not (and none after), while the plane was still in VHF range passing over Malaysia, I suppose your option 3. can be excluded?”
The ‘Lost VHF’ message sent by SATCOM at 1554:53 UTC (when the aircraft was on the ground) seems to indicate that VHF had been de-selected in the ACARS Manager. ACARS transmission by VHF would not have been possible, even if the aircraft was within VHF range.
@Ge Rijn
@Peter Norton
@Don Thompson
The other possibility is that DATA mode had been deselected on the applicable VHF, possibly because it was used for voice communications with MAS Operations or some other station. The crew might then have forgotten to reselect DATA mode. I don’t know if the ‘Lost VHF’ message would be sent in that scenario – Don??
A new Reuters story says there has been no decision on the OI proposal, quoting Malaysian Transport Minister Liow.
> Andrew says:
> The ‘Lost VHF’ message sent by SATCOM at 1554:53 UTC (when the aircraft was on the ground)
> seems to indicate that VHF had been de-selected in the ACARS Manager.
As a B777 captain, can I ask you whether
– the action itself
– and/or the timing [47 min before takeoff]
is expected or in any way suspicious ?
@Peter Norton
The timing seems to indicate the selection occurred during the pilots’ pre-flight preparation. However, there is no requirement to make selections on the ACARS Manager page during the pre-flight. It would certainly be unusual for VHF to be deselected – I’ll leave the ‘suspicious’ bit for you to decide.
@Andrew
Than anyway; if VHF was selected and DATA mode was selected after 15:54 a ACARS data transmission should have been recieved also at 17:37 regardless if the left IDG/AC-bus was isolated or not.
As I understand it the 17:07 ACARS transmission data was send via SATCOM so I think at least by that time the DATA mode was not selected for VHF.
And while no VHF ground station recieved any ACARS transmission within VHF max. range of ~200 miles around the plane (as far as we know) VHF and/or VHF-DATA mode must have been disabled/not selected.
I’m particularry curious about this for I would like to see proof the left IDG/AC-bus was not isolated.
Since the VHF or VHF-data mode was not enabled or selected at 15:54 and 17:07 we still cannot exclude the left IDG/AC-bus was isolated between 17:07 and 17:21 according my reasoning..
> Andrew says:
> The timing seems to indicate the selection occurred during the pilots’ pre-flight preparation.
> However, there is no requirement to make selections on the ACARS Manager page during the pre-flight.
> It would certainly be unusual for VHF to be deselected
@Andrew: thanks.
If there is no chance that this is a routine procedure at MAS, this looks like a red flag to me.
Do we have the pre-flight phase of MH371 included in the ACARS log, so that we could check whether the same (VHF deselection) also happened on MH371 ?
@Peter Norton
I think you asked an important question.
This sure could be a red-flag if it not happened on MH371 (or other flights).
I’m very curious about this too.
@Ge Rijn: I have seen no proof that the left bus was not isolated. On the other hand, there IS evidence that the left bus was powered just before the log-on at 18:25:
1) The BFO sequence of the 18:25 log-on is consistent with the sequence for other SATCOM power ups.
2) The lack of the second Log-on Request about a second after the first one, which seems to be related to the boot-up of the IFE, which is also powered by the left bus. For all the logs we have, including the ones related to MH371, the second Log-on Request is missing only at 12:50:19, 18:25:27, and 00:19:29. We know the SATCOM (and consequently, the left bus) was powered up just before 12:50:19. We suspect that is also the case for 18:25:27 and 00:19:29.
@VictorI
1) The BFO sequence of the 18:25 log-on is consistent with the sequence for other SATCOM power ups.
But then what about the OXCO warm-up BFO explanation at 18:25?
@Ge Rijn: I don’t understand your question. The oscillator warm up is expected with a SATCOM power up that occurs after the SATCOM has had time to cool.
@VictorI
Then what you suggest as I understand it; the IFE log-on was missing at 18:25 and 00:19.
Doesn’t this clearly indicate the IFE was switched off in the cockpit before that time?
@Ge Rijn: The log-on of the IFE eventually did occur during the 18:25 log-on sequence, including the “Pet shop boys” message, which I believe was a weak password exchanged during the log-on to the IFE server. This means that the IFE was not switched off. It just took some time to be ready to initiate its log-on. This was also seen for the 12:50:19 log-on. The log-on to the IFE server did not occur after 00:19:29, probably because the SATCOM was no longer operating.
@VictorI
Your remark:
1) The BFO sequence of the 18:25 log-on is consistent with the sequence for other SATCOM power ups.
I understood there probably has been a OXCO warm-up responsible for the high BFO fluctuations at 18:25.
I just think this is not consistent with other/normal SATCOM power ups.
@Mick Gilbert: thinking of your windshield heater theory, you might want to look at the iasa-intl.com links above
@Ge Rijn: Please read Ian Holland’s paper. His contention is that a power up that occurs after the SATCOM has cooled induces a BFO transient just as was seen at 18:25.
@Peter Norton wrote “If there is no chance that this (disabling VHF datacomms) is a routine procedure at MAS, this looks like a red flag to me.”
The Media Advisory ACARS messages (label SA) transmitted during the MH371 sector show only SATCOM available, an example below:
(soh)2.9M-MRO(nak)SA4(stx)S61AMH03710ES010740S(etx)(chksm)(del)(nul)(nul)the key portion of the string being…
ES010740SEnabled SATCOM at 01:07:40 only SATCOM available.
As I stated above, had VHF also been available then the trailing characters would indicate that condition (as ‘VS’)
VHF coverage over China is good, with VDL Mode 2 RGS deployed, but not so good with plain old ACARS over Analogue VHF in many territories. The latter is slow, exacerbated by growing traffic. Red flag? Possibly, possibly not, you might it fruitful to mine the web for hobbyist ACARS tracker logs with examples of MAS traffic over VHF (there are plenty of sites).
@VictorI
The SATCOM was operating till 00:19:37 when the plane responded with a ‘log-on acknowledgement message’. The IFE log-on was not recieved after.
It was concluded by several the plane crashed in between for this reason (within ~one minute or so).
If the IFE was switched off after 18:25 this IFE log-on could not have taken place after 00:19 and therefore could not have served as an indication the plane crashed in a high speed dive.
@Ge Rijn: You are mixing too many things. Sure, it’s possible that the IFE was allowed to log-on at 18:27 and then turned off after 18:29. However, the high speed descent theory is based on the BFOs at 00:19, not the absence of the IFE log-on. There’s no evidence that the SATCOM was powered down for any significant time before 00:19. In fact, it had to be less than eight minutes, and was probably closer to two minutes. And the BFOs were abnormally low, not high, as you would expect for the warm up transient.
@Peter Norton
Thanks Peter, I’m familiar with the iasa-intl.com links.
@Don Thompson:
Thanks for looking into the MH371 log.
My red flag comment was solely based on Andrew saying “there is no requirement to make selections on the ACARS Manager page during the pre-flight. It would certainly be unusual for VHF to be deselected”.
I would find it interesting if you could discuss this directly with him.
— begin quote —
Don Thompson says:
« The ACARS Traffic Log, published in the Factual Information report, shows that VHF data link was down. A related Media Advisory msg was sent at 15:54:53 (ACARS terminal timestamp). As a consequence of VHF Datalink unchecked in that same ACARS Manager screen on the MFD.
at 15:54:53z the ‘VHF ENABLE’ check box was unchecked, resulting in the ACARS Data Comm Mgmt function sending the Media Advisory ACARS ‘Lost VHF’ msg via SATCOM to the SITA ACARS Processor. SATCOM remained the operational datalink. »
Andrew says:
« The ‘Lost VHF’ message that Don mentioned seems to indicate that the VHF ENABLE check box in the ACARS Manager was unchecked, so ACARS was unable to send downlink messages by VHF.
The ‘Lost VHF’ message sent by SATCOM at 1554:53 UTC (when the aircraft was on the ground) seems to indicate that VHF had been de-selected in the ACARS Manager.
The timing seems to indicate the selection occurred during the pilots’ pre-flight preparation. However, there is no requirement to make selections on the ACARS Manager page during the pre-flight. It would certainly be unusual for VHF to be deselected. »
— end quote —
@Andrew,
@Don Thompson:
Rereading your statements, it’s not 100% clear to me whether the ‘VHF ENABLE’ check box was
– unchecked (status), in the sense of “not checked” (all the time)
– or unchecked (action), in the sense of “removing checkmark from checkbox”
?
I lean towards the latter.
So before 15:54:53 UTC the ‘VHF ENABLE’ check box was checked
and at 15:54:53 UTC one of the pilots on MH370 unchecked the checkbox, correct ?
Similarly, I would like to clarify for MH371:
Does ES010740S only mean that VHF was not available (status),
or that someone took an action at 01:07:40 UTC so that the enabled VHF –> became disabled ?
Btw, when was time of take-off for MH371 ?
@Victor
Please read Ian Holland’s paper. His contention is that a power up that occurs after the SATCOM has cooled induces a BFO transient just as was seen at 18:25.
The fact that the BFO “transient” seen at 18:25:27 is consistent with the aircraft speed, track, and position should be mentioned.
And the BFOs were abnormally low, not high, as you would expect for the warm up transient.
I think the BFO error during warm-up recovery (especially for short off periods) is a function of where the SDU temperature set point is relative to the turnover point of the OCXO. In fact, I suspect that this variability between SDU’s is the reason why several representative SDU’s were not tested (or not reported on) by Holland. The shape of the OCXO temperature transfer function is not addressed by Holland nor by anyone else commenting on this issue.
It is also true that in the Holland paper the 18:25:27 event is the only example of a warm-up where the first BFO was lower than subsequent BFO’s during the settling period.
@VictorI
I understand the SATCOM was not powered down again after 18:29 and not at 00:19.
At 23:13 the SAT-call to the plane proves this and the messages send till 00:19:37.
There was only not a IFE log-on after 00:19:37.
This was earlier explained by the plane crashed before this could happen.
Am I wrong here? Please explain.
The BFO’s at 00:19 show only a snapshot of a 8 seconds high descent rate. They tell nothing about the IFE switched off or on at that time.
It could possibly be the plane did not crash before the expected IFE log-on, for possibly the IFE was switched off before.
And this could have been done shortly after 18:29.
@Victor
“A new Reuters story says there has been no decision on the OI proposal, quoting Malaysian Transport Minister Liow.”
Aargh…well Malaysia are between a rock and a hard place. If the search is stopped, that will open up some discussion/articles about what’s going on here. What I would say is going on here, is that it is possible the flight followed the sim cases quite closely, but the investigation totally ruled out that option for political correctness. This has many implications.
If the search is continued, it extends the search, and probably delays the criticisms to a later date.
One important re-hash question I see, was MH370 seem in the Andamans or not? Indonesia said they did not see it on their radar, and they heard it was seen seen in the Andamans. I am inclined to suspect someone may have seen MH370 in the Andamans, which is where the sim cases would put it.
@VictorI
Want to remind you the SATCOM link to the pilot-phones in the cockpit are independently working from the IFE (switched on or off) as @Andrew stated some time ago (as I remember well..)
@TBill
If OI is granted this search they are not likely going to stop until the plane is found on a no cure no pay base. Their name and future will depend on such an effort. Failure is unacceptable. If they succeed they win a fortune in money and status for many years to come.
The ~35S CSIRO latitude would be a great point to start this new search. Then expanding north-east and westwards if necessary till ~32S.
The deep trenches south of Broken Hill will be the greatest challenge.
But it’s there where I expect the plane has found it’s grave.
@GeRijn
If my memory serves, at 18:03 MAS sent an urgent ACARS message to the plane, requesting that the crew make contact with the ground. There was no response from the SDU. If ACARS was disabled but the SATCOM was powered up, the SDU would have sent a signal back acknowledging receipt of the message.
So from this we can conclude that the SATCOM was depowered at 18:03. As has been explained many times, the SATCOM can only be depowered from the cockpit by isolating the LH Main AC bus. So the most likely scenario is that the pilot isolated the AC bus some time between 17:07 and 17:37*. ACARS was deactivated some time after 17:07, and never reactivated. VHF data link was also off.
The LH AC bus was reconnected at 18:24.
*I think it’s most likely that the pilot isolated the LH AC bus shortly before 17:19, to make sure no further routine ACARS messages could go out, and no IFE SMS or email messages could go out when it was realised that the first officer had been locked out, and the captain had gone rogue. Apparently, the pilot didn’t bother to switch off the IFE from the overhead panel, because when the AC bus was reconnected at 18:24, the IFE was able to power up and issue it’s logon request a few minutes later.
Peter Norton:
“As a B777 captain, can I ask you whether
– the action itself [VHF had been de-selected in the ACARS Manager]
– and/or the timing [47 min before takeoff]
is expected or in any way suspicious ?”
Early on it has been suggested that deselecting VHF is normal practice on MAS flights to Beijing. MAS’s ACARS Service Provider SITA has no VHF receivers in China. Receiving ACARS messages through a third party (ARINC or a state-owned Chinese company) would incur roaming charges.
P.S. The ACARS messages during flight MH371 were also routed via SATCOM.
DennisW wrote: “The shape of the OCXO temperature transfer function is not addressed by Holland nor by anyone else commenting on this issue.”
Come on! You know that is not true. I’ve been trying to explain this topic for 3 years. Bobby has also gone to great lengths to document it. Some just don’t understand the nature of thermal control loops and crystal physics. Others do, but don’t like the facts because they don’t help a particular theory.
@Rob
The LH AC bus was reconnected at 18:24.
Why?
@Rob
It seems clear by now the VHF data link for ACARS was not available after (and before) 17:07.
By 17:37 and 18:07 they should have recieved an ACARS report by SATCOM or VHF. They did not. They send a message to the ACARS at 18:03 without respons. They could have waited for 4 minutes for the aircraft to send the message.
VHF data link was already disabled at/after 15:34 and never re-established.
I think this is a red-flag if this was not common practise.
I also think isolating the left-bus was the most efficient way to handle things around 17:19. Especially when VHF was disabled/deselected before.
It would shut down all SATCOM at once; ACARS, transponders and IFE.
And it would depower cabin lights too.
The plane would go completely dark in an instance.
Only visible for primary radar who maybe did not know how to deal with it right away.
@ALSM
You know that the shape of the warm-up transient is dependent on which side of the turnover point the temperature set point is located. The set point cannot be exactly at the turnover point for practical reasons as well as aging.
Yes, I know you (and me) have been at this for some time. I do think you are correct, BTW. I am just not certain of it because of the loose ends.
@Ge Rijn: There was only not a IFE log-on after 00:19:37.
This was earlier explained by the plane crashed before this could happen.
That’s one explanation. Another is the APU ran out of fuel. It’s also possible that a left engine restart would have consumed whatever little fuel was available to the APU.
DennisW: Actually, we do know the setpoint was slightly below the turnover temperature, and not slightly above it in the case of 9M-MRO. If it was above, the transient shape would have been inverted. Anyway, these details are not so important. What is important is that we know the SDU power was off for some extended time prior to the 18:25 logon, thus the “unreliable BFO values” in the first few minutes after the logon, and off only a minute at 00:19, giving us relatively small BFO errors during the 00:19 logon.
@DennisW and @sk999: Out of curiosity, starting with the 18:22:12 radar capture, what do you believe is the path that explains the BTO and BFO sequence during the 18:25 log-on?
@DennisW
“The LH AC bus was reconnected at 18:24.
Why?”
Oh Dennis, you are a tease. I’ve explained that as nauseum. You know your theory is on the ropes, so you have to resort to filibustering. I almost feel sorry for you 😉
Victor, Dennis, SK999…
Re path 1825-1828
This chart: https://goo.gl/yHS9f5
…does not indicate the path per se, but it does clearly indicate an a/c to s/c radial speed reduction occurred. Remember, the BTO values are not subject to any warm-up issues, so all 7 are good (within jitter limits). That could mean a left or right turn, or a significant speed reduction between 1825 and 1828. The BFO values after the OCXO settles should help determine if the BTOs indicate a speed reduction or a turn.
@Ge Rijn
“I also think isolating the left-bus was the most efficient way to handle things around 17:19. Especially when VHF was disabled/deselected before.
It would shut down all SATCOM at once; ACARS, transponders and IFE.
And it would depower cabin lights too.
The plane would go completely dark in an instance.
Only visible for primary radar who maybe did not know how to deal with it right away.”
Right on pal, you’ve got it. Come on, you knew it all along, didn’t you. Red flag after red flag.
Now that we’re on the same page: Leaving the IFE to log on at 18:28 was another taunt imo. I think it was telling the ground “I’ve dealt with the passengers, they won’t be trying to get any messages out now”.
@ALSM
@DennisW
My money’s on a speed reduction. Once he was beyond the reach of the Penang radar, he could ease back the throttles and reduce his mpg.
@ALSM
@Dennis
I meant increase his mpg for the journey south.
@ALSM: Sure, but the challenge is to satisfy the BTOs AND the BFOs, ideally with level flight. After passing MEKAR and turning towards NILAM on N571, a SLOP at a constant speed of 500 knots satisfies the criteria. Also, a turn towards VOCX along with a decel to 389 knots (FL340, 225 KIAS, and ISA+10K) does it. I’m curious what others like @DennisW and @sk999 (or anybody else) think occurred.
Without a SLOP (which is really two automated turns), both a turn and a decel are required if the BFO is to be closely matched. So you need two turns, or a turn and a decel. A decel alone would not match the BFO.
The lack of response from the SATCOM at 1803 (aircraft near Pulau Perak) doesn’t necessarily mean it was depowered at this time. It could have been repowered but was unable to connect because the Left High Gain Antenna is inoperative. The Right High Gain Antenna is obscured by the fuselage at this point, as according to the antenna coverage chart.
The 1825:27 log on occurs when the aircraft rolls left at Nilam and exposes the serviceable Right High Gain Antenna. At this moment the aircraft is heading ~290 degrees. Maybe this BFO could be correct because it has had extra time to prepare. The next BFOs then follow the power up decay sequence as proposed by DrB and Ian Holland.
The aircraft was tracking Nilam-Sanob-Banda Aceh (~FL350/M0.82). Indonesian Primary Radar or Fariq’s phone log with Banda Aceh could confirm this, but both are not available.
What is not known is how the BFOs react following double engine failure with load shedding occurring and APU on? There aren’t any previous examples of this.
The aircraft is most likely within 100 nautical miles of Bayesian Hotspot.
@Victor
I have the aircraft at ~6.8N ~95.9E, track of 296 and GS of 510 kts at 18:25:27.
@DennisW: Something has to give. Are you not considering all the BTO values? The first one is least accurate.
@Rob
I believe you are correct (not about the slow down) but the emergency ACARS message at 1803 or thereabouts. And that message automatically repeated every 2 minutes for some 30 minutes or so. So I was going to say, I wonder why it did not go thru at 1825, but I suppose ACARS was turned off.
@ALSM
Actually, we do know the setpoint was slightly below the turnover temperature, and not slightly above it in the case of 9M-MRO.
That is my take on it as well based on the other settling transients in Holland’s paper.
@Paul Onions: There are not separate right and left HGAs. There is one HGA with separate right and left arrays. What would cause one of the arrays to work and not the other?
@Victor
I am considering the 18:25:27 BTO. Not to say that I did not screw it up. I do recall some controversy a while back on Jeff’s site relative to the 18:25:27 BTO (range ring position). It has been a couple of years since I looked at it.
@Paul O.
…also I believe that 1803 ACARS message was continuously repeated for some time. So ACARS was off, I believe is that answer.
@DennisW
I am thinking if you go with the 1825 BTO (12520 corrected), the aircraft had to fly quite fast to get there from 1822. If you go with 1828 BTO (12500), the aircraft had to travel slow to get there, or make a SLOP. But we are putting a lot of emphasis on expecting radar to seamlessly match satellite pings.
@DennisW: I think your result would be different if you tried to match the expected collective error of all the BTO values, together with the last (or first) BFO value. That’s why the complexity of a SLOP or a turn/decel was introduced. Or, find a way to justify ignoring all the BTO values but the first, which is what you are doing now.
TBill:
Re “But we are putting a lot of emphasis on expecting radar to seamlessly match satellite pings.” In the chart I posted above, it is easy to see that even a 3 mile error in the last radar position can’t begin to explain the reduced radial speed. Yes, the radar could be in error, but nowhere near enough to explain the BTO data without a slowdown toward the s/c.
@Victor
You are probably correct. I got frustrated after 18:25, and could never make anything work. So I simply gave up.
@VictorI
It is hard to make a scenario that matches the BTO/BFO all the way: if you believe the BFO should be closely matched then you should probably opt for an early turn south, because you would need a 19:41 latitude close to the equator (at least for “straight” path scenarios after 19:41) to match the later BFOs.
A simple scenario I have been considering is approx. 360 kts GS from 18:22 all the way to 0.8N @19:41, with a turn south on N571 close to 95E. The problem is then in the BFOs before 19:41.
@Niels: I was really only considering the radar at 18:22 and the BTO/BFO cluster sequence starting at 18:25, which can be reconciled with either two turns (such as the SLOP) or a turn/decel. After that, there are numerous ways to make the BTO and BFO match (within some allowable error).
@DennisW
I apologise for being cheeky. It wasn’t really justified.
I remember the JW BTO ring fiasco. I was the one who pointed out to JW that he’d got his ring position wrong. He had postulated a turn at 18:22 as a reason for the radar return. I carefully drew the radar plot onto a 1:2,000,000 scale map of the area, plus the ping ring. I found that a deceleration shortly before 18:22, from M0.87, then a continual slowdown to M0.81 at FMC at IGOGU (time 18:37:15) fitted perfectly with the N571 track. speed at 18:25:27 was M0.85.
For what it’s worth.
@TBill
Yes, ACARS would have been off at 18:25, and for the rest of the flight as well.
@VictorI
I’m afraid it comes back to the unresolved BFO error issue. In any case: if you take a position regarding allowable BFO errors you have to apply it consequently.
My current position is, based on the example flights from sources I trust, that BFO errors are limited and that therefore the “McMurdo” path is really on the edge.
Furthermore, the near constant BFO sequence during the first phone call would require a very constant descent rate in your scenario. I’m not sure how likely that is.
@Niels: I wasn’t talking about the McMurdo path in particular. That’s one of many possible. There are constant Mach, constant level, great circle paths that have BFO RMS errors (19:41-00:11) as low as 3 Hz with no descent at 18:40.
@Victor. A slow down without turn will satisfy both BTO and BFO 1825-1828.
At FL340 slow to M0.7 from 181700. I get BTO residuals as follows: 80, -5, -32, 8, 5, -18, -4. All perfectly allowable if we consider the larger uncertainties on the first two. Predicted BFO (with true azimuth around 290 and speed M0.7) is 141 (vs 144 that we presume would have been observed without transient)
Taking all 7 of the BTOs and their 2-sigma uncertainty limits per DSTG and crossing the 7 arcs on this heading and speed, you have a time window that is +/- 30 seconds either side of the “min RMS” that gets you across without busting the error bounds on any BTO.
@VictorI
I see; I’m sorry for this misunderstanding.
“There are constant Mach, constant level, great circle paths that have BFO RMS errors (19:41-00:11) as low as 3 Hz with no descent at 18:40.”: I agree. These typically end south of 32S.
@ALSM
But the pings could be off too? I certainly would not mind moving the pings further west, I might even try that sometime to see if it helps.
We need a flight path like DrB’s documentation of 8-Jan path with the SLOP, to show what the slow down group is saying. I am not automatically on board. We are suddenly over-turning the JoN paper. We would need a new blog article on this from Victor.
@Niels: No apologies necessary. For instance, the BEDAX-SouthPole path has been one of my favorites for a long time. That crosses the 7th arc around 34.3S, which is as you say, south of 32S.
Yes, BEDAX – South Pole has been my favourite for a long time too.
@ulric: I became less enthusiastic about 34.3S when the 7th arc was searched at that latitude and came up empty.
@Victor,
Re: “Mostly, I am trying to better understand the various claims about the efficiency of the search along the 7th arc. As we have discussed before, if as efficient as claimed by AMSA, impact above 33S is unlikely. However, I question these efficiency estimates.”
– In my opinion 30.5S to 34.5S is very unlikely segment: the coverage of debris field would have been excellent regardless various leeway factors. And this result appears to be consistent with CSIRO’s result, adding some confidence. I think it is very unlikely the AMSA could miss debris originated from this area.
– There is a small ‘gap’ at 30S (both my model and CSIRO’s), where results are very sensitive to the leeway factor and, hence, wind. There were some suspicious objects spotted, but never recovered (in particular the white rectangular object photographed from RNZAF Orion P3K).
– Now we have some discrepancy with regard to the next gap: my model shows the gap at 26-27S, CSIRO’s – at 24-25S. My model shows the peak daily coverage for the origins in this segment on April 18-20, which is more than one month after the crash; CSIRO’s model exhibits sensitivity to the leeway factor and exact location. I think the AMSA could miss debris originated from this area.
Re: “The BFO sequence of the 18:25 log-on is consistent with the sequence for other SATCOM power ups”
I would say the opposite is more likely: the BFO sequence of the 18:25 log-on appears to be inconsistent with the sequence for other SATCOM power-ups based on Holland’s paper, and also 00:19 logon. Furthermore, the paired BTO 18:25:27 is normal (consistent with Lido) in contrast to the BTO paired with the abnormal BFO 18:25:34. In other words the pair of BFO and BTO 18:25:27 looks absolutely normal; the only signal strength was low. Summarizing:
1. The BFO profile during 18:25 logon is inconsistent with any other presented power-up event;
2. The first pair of BTO and BFO appears to be absolutely normal, but the second pair is corrupted.
Re: ” I’m curious what others like @DennisW and @sk999 (or anybody else) think occurred.”
There are many ways to fit 18:25-19:41 BTO and BFO if you consider some kind of a holding pattern with descent, e.g. spiral descents, triangular patterns, etc. It turned out that repeated triangular patterns are like Lego constructor – there are many possible combinations that can fit BTO and BFO. An intriguing feature of the triangles is that a wrong number of legs or wrong exit direction may result in the final heading of approximately 172 deg. I have suggested the radar alerting maneuver (in the frame of mechanical failure scenario). Perhaps Andrew or anyone else can suggest another reason for repeated triangular patterns; I believe they can also be ‘glued’ with the subsequent magnetic heading or track at FL200-FL250.
https://www.dropbox.com/s/h7g0754a83s4jpz/trajectory-v1.png?dl=0
https://www.dropbox.com/s/494mzxlovhvnnbx/trajectory-v2.png?dl=0
https://www.dropbox.com/s/1d5emi36jvadh7u/trajectory-v3.png?dl=0
@Peter Norton,
Re: “But can a technical failure not be excluded simply by the fact that
– at first ACARS was silenced + SDU depowered
– and later the SDU was repowered but ACARS remained silent (see below)
?”
No. Try to google on Swiss Air 111.
@TBill
Richard Cole’s plot of Fugro Equator’s 6-January 2017 scan shows the 19:29 7th arc crosses -37.6 at longitude 89.0
If you do the calculation for the 7th ping ring with Ashton’s 18400 microsecond BTO you get a ping ring that crosses -37.6 at longitude 86.8 nearly 80nm west of the Bayesian hotspot…
Explain that discrepancy and you might have the answer why 9M-MRO was not found…
@George Tilton
That’s interesting…I was not aware of it. I know Mike Chillit puts the rings further west, but I am not sure why he thinks that. I know the altitude plays a secondary, westerly role in the calc.
Do HGA and LGA rely on two different OCXOs?
@Andrew,
Just curious: would the left transponder be usually selected as default?
Victor,
R.E. reconciling the radar data with the BTO/BFOs, I’m basically with DennisW on this one. There are multiple possibilities, none of which is entirely satisfactory. Basically, I don’t use the last radar position as a timing constraint. More critical is when/where the plane turned South, and at what speed. More variables than constraints.
@Oleksandr
Re: “Just curious: would the left transponder be usually selected as default?”
It’s not specified in the MAS FCOM. Where I work, the L or R transponder is selected depending on which pilot is nominated as PF.
Oleksandr: The OCXO is in the SDU. There is only one SDU associated with the HGA and LGA.
George wrote: “If you do the calculation for the 7th ping ring with Ashton’s 18400 microsecond BTO you get a ping ring that crosses -37.6 at longitude 86.8 nearly 80nm west of the Bayesian hotspot.” This statement is not correct. Actually, -37.6, 89.0 is 1.4 nm inside the 35K 7th arc and 3.5 nm outside the 0k ft 7th arc. Thus, the arc Richard Cole was using was correct.
TBill: Chillit’s arcs are total garbage. His math was ridiculous nonsense. Ignore them.
@Gysbreght
@Peter Norton
Re: “Early on it has been suggested that deselecting VHF is normal practice on MAS flights to Beijing. MAS’s ACARS Service Provider SITA has no VHF receivers in China. Receiving ACARS messages through a third party (ARINC or a state-owned Chinese company) would incur roaming charges.”
Such a procedure is not documented in the airline’s manuals. The Operations Manual states “Ensure VHF and SATCOM are both selected…during ACARS usage”. There are no exceptions mentioned for areas with high VHF data charges.
Where I work, the airline programmable part of the ACARS software is set up so that ACARS automatically uses SATCOM in areas with high VHF data charges. There is no requirement to manually deselect VHF.
@George Tilton, @TBill: You may or may not realize that the coordinates of the 7th arc are based on an accurate satellite orbital model and the most accurate representation of the oblate ellipsoid that is available, known as WGS84. That’s the same model of the earth used by GPS to calculate position. The calculations have been independently checked many times, including by many of the contributors here.
There are many possible explanations for why the wreckage was not found. However, errors in the calculation of the 7th arc is not one of them.
@sk999: Yes, it is possible that the timings of the radar captures are wrong. In fact, that would explain a lot.
TBill asked: “But the pings could be off too? ” No, they can’t. We’ve been over this question ad nauseum for 3 years. The pings (arc’s) are not subject to any drift whatsoever (unlike BFO values), or random error other that what was meticulously analysed and documented by Thales. Thales measured the BTO jitter and found that the jitter was the result of a repeatable pattern, probably caused by software driven clocks. The bottom line is that the max arc position error is ±5.3 nm (95% confidence). https://goo.gl/bqxDNk
@ALSM
Actually, we do know the setpoint was slightly below the turnover temperature, and not slightly above it in the case of 9M-MRO.
Yes. The BTO jitter (insignificant as it is) appears to be the result of using a numerically controlled oscillator to provide the Doppler compensation. Typical source of very small pseudorange errors for GPS receivers as well.
@ALSM
Quote above is a residual from an earlier post. Ignore the italics.
Dennis:
The NCO resolution and noise does have a tiny impact on the BFO values, but not the BTO values. Thales and Square Peg could not determine where the BTO jitter was originating. It could be in the demod software or the SDU software or both. Thales reported:
“The jitter on the BTO is due to a number of components within the return path between the Ground Earth Station (GES) and the Airborne Earth Station (AES).
»» The noise in the AES receiver when detecting the time of the P-channel unique word.
»» The quantisation of the AES transceiver internal time reference.
»» The noise in the GES receiver when measuring the time of arrival of the AES transmission.
»» The quantisation of the BTO measurement.”
Bottom line:
“These test results are in agreement with the understanding of the error components and the flight test experiments and supports the assertion of a 50μs effective accuracy. In the case of MH370, there are very few samples and so we cannot treat any particular measurement other than individually. Hence the assertion that a particular BTO has an accuracy of 50μs (at a 99% confidence level). For reference, a 50μs error equates to a +/-7.5km error in satellite range. This equates to approximately +/-10km horizontal error on the Line of Position. This concurs with the reference flight measurement
analysis.”
@Ge Rijn
@Rob
Ge Rijn wrote: ““I also think isolating the left-bus was the most efficient way to handle things around 17:19. Especially when VHF was disabled/deselected before.
It would shut down all SATCOM at once; ACARS, transponders and IFE.
And it would depower cabin lights too.
The plane would go completely dark in an instance.”
Rob wrote: “Right on pal, you’ve got it.”
Far from “right on“, the notion that depowering the Left AC Bus is some sort of one-stop shop for sending the airplane “dark” by taking down airplane systems such as VHF radios, transponders, IFE, lighting, etc is simply wrong.
Comms are highly redundant. The system is designed so that at least one radio (viz left VHF, the radio normally used for ATC comms) will be available, even if the electrical system is only running on the batteries. Accordingly, depowering the Left AC Bus has very little impact on the comms system in general and no impact on the VHF radio used for communicating with ATC.
The left and right transponders are powered off the AC Standby bus and the Right AC Transfer bus respectively. Depowering the Left AC Bus has no impact whatsoever on the transponders.
Some IFE components are powered off the 115V AC R SEC 1 Bus while others are powered by the 115V AC L SEC 1 Bus. Similarly, the IFE in some seat rows is powered by the 115V AC R AC SEC 2 Bus and others by the 115V AC L SEC 2 Bus. Depowering the Left AC Bus has no impact whatsoever on those components.
Most of the cabin lighting is powered by the 115V AC Ground Service bus, which is powered by the Right AC bus in flight. Depowering the Left AC Bus has at best a very limited impact on cabin lighting.
Similarly, depowering the Left AC Bus has virtually no impact on the external lighting either; the upper and lower beacons, the right wing navigation lights, and the logo lights that illuminate the vertical stabiliser would all remain operational.
@Victor
I used Dr B’s equations for the satellite model: X=18179, Y=38051, Z=393 km @ 00:19 utc.
I did the bias calibration at the gate and agreed to within 8 microseconds on the bias. I did the N*7820 correction on the logon ack at 00:19 and got 18380 for the readjusted BTO…Ashton’s 18400 is good enough…
Slant Range to GES came out to 39197 km, Range to AES 37860 km, satellite elevation 39.06 degrees and arc radius 4920 km @ 00:19…With a sub-satellite point of 0.5347,64.4638 @ 00:19 and arc radius of 4920 km I get a longitude of 86.8E for a latitude of 37.6S. Where did I run astray?
@George
Using your circle radius (4920km) and center point (0.5347, 64.4638) I get a longitude of 88.85 at a latitude of -37.6. A two degree difference. Something is amuck with your circle conversion. Are you using a WGS84 ellipsoid? That is my guess as to where the error lies since everything else is x,y,z.
I did not check your radius or center point values since I stop at 00:11.
@George
BTW, at 00:19:39 I get satellite position of:
x = 18180.6 km
y = 38055.7 km
z = 391.2 km
using sk999’s orbital model which has been excellent. So your values there seem OK.
@DennisW
It has to be something stupid…
Glad I didn’t put a pin in the map 😉
@George Tilton: Your GES range looks close, but something is wrong with your A/C range and geodetic coordinate calculation. Using your latitude of 37.6S, I calculate a longitude of 88.8816E at sea level. The A/C range is 37,861.9 km, which can be calculated from (18400+495679)x.299792/2 – 39196.5. Meanwhile, for your coordinates of 37.6S,86.8E, I calculate an A/C range of 37780.7 km, and a BTO of 17858 μs, for an error of 542 us.
If you are using a spherical earth, that might account for some of the error.
@George Tilton: At this point, we have cross-checked our calculations so many times that it is nearly impossible that we have the arc wrong (assuming the bias for the BTO values have not shifted, which nobody thinks is very unlikely).
@DennisW: I wouldn’t have bothered checking his calcs if I knew you were doing the same.
@Victor
@DennisW
Thanks I appreciate the help from both of you…
I think I see where I may have screwed up and will take a whack at redoing the spreadsheet tomorrow.
@ALSM
“TBill asked: “But the pings could be off too? ” No, they can’t….The bottom line is that the max arc position error is ±5.3 nm (95% confidence)”
OK thank you, I’ll take +-5.3 nM if that’s all I can get.
Jeff Wise has solved the MH370 enigma!
http://jeffwise.net/2017/10/03/mh370-mission-accomplished/comment-page-3/#comment-212088
@Rob
“…at 18:03 MAS sent an urgent ACARS message to the plane, requesting that the crew make contact with the ground. There was no response from the SDU. If ACARS was disabled but the SATCOM was powered up, the SDU would have sent a signal back acknowledging receipt of the message.”
Per Page 66 of the Factual Info, that ACARS emergency message was sent every 2 minutes until about 18:45, so if you are correct that means the SDU would have sent acknowledgements every 2 minutes after 18:25.
@Perfect Storm
Wait, what? So he travelled to Baikonour with a shovel!
@Ge Rijn
@Rob
A correction to my 8.03pm on the impact of depowering the Left AC Bus on aircraft systems, because some IFE systems components are powered off the
AC L SEC 1 Bus, depowering the Left AC Bus will shut down the IFE. However, VHF and HF comms, cabin lighting and exterior lighting will be largely unaffected and transponders will be entirely unaffected.
> Peter Norton:
> But can a technical failure not be excluded simply by the fact that
> – at first ACARS was silenced + SDU depowered
> – and later the SDU was repowered but ACARS remained silent (see below)
> ?
> Oleksandr:
> No. Try to google on Swiss Air 111.
You say this because on SR111 after AES reconnected no ACARS messages were subsequently exchanged, right ?
@Mick Gilbert
Clear about the IFE correction.
And you are right about the transponders not being depowered after isolation of the left AC-bus it seems to me also after looking at it again.
@Mick Gilbert
I’m glad you corrected yourself. Obviously, I didn’t read GeRijn’s reply thoroughly enough to spot he had included the transponders as well. When I said right on pal, I was being facetious, anyway. GeRijn should have known the transponders weren’t powered off the LH Bus. This has all been discussed her as nauseum, like just about everything else on the subject.
To reiterate, isolating the LH Bus depowers the SATCOM, TCAS, RH HF radio, CVR, cockpit door lock power, some IFE head components, some cabin and galley lighting at a stroke. Because it depowers the SATCOM, it also renders ACARS inop. ACARS can then be deselected via the CDU (or is it the MFD?)
at leisure, before the the LH Bus is reconnected and wakes the SATCOM up again.
Here we go round and round and round and round….. until we all die of boredom
@Mick Gilbert
BTW, I nearly forgot, the LH Bus also powers the fuel transfer pump on the 200ER, as I remember.
TBill “Per Page 66 of the Factual Info, that ACARS emergency message was sent every 2 minutes until about 18:45, so if you are correct that means the SDU would have sent acknowledgements every 2 minutes after 18:25.”
There’s something very wrong there then, isn’t there. I would suggest you look at the INMARSAT Communication Logs to see exactly how long the ground kept sending the ACARS message. The SDU was depowered from the takeover until 18:24.
@Rob
Re: “To reiterate, isolating the LH Bus depowers the SATCOM, TCAS, RH HF radio, CVR, cockpit door lock power, some IFE head components, some cabin and galley lighting at a stroke. “
At least two of those items are wrong, but I won’t bore you with the details…
BTW, what’s a fuel ‘transfer’ pump?
@all,
HERE is a note on how the Lido radar positions limit the possible turns made by 9M-MRO near MEKAR. A significant turn at or before MEKAR at ~18:21:00 appears to be ruled out by the Lido radar positions. Slight turns between 18:21-18:22 are allowed, as are all turns after 18:22:12.
AirlineRatings is reporting no deal yet to start the search.
The US company pitching to undertake a “no find, no fee” search for missing Malaysia Airlines flight MH370 has reported good progress with its bid but says it cannot yet confirm it has a contract.
@Rob
Well I’m positively delighted that your glad that I corrected my 8.03am post. Needless to say I await your doing likewise with regards to your 4.20am.
@DrB: Your new figure showing the radar data around MEKAR is helpful. How did you determine the position of the last radar capture at 18:22:12? I have not been able to find a photograph that includes that point. The clues we have are the arrow and the timestamp label. When Ron Belt in his report digitized the data, he estimated the last radar point after MEKAR to be very close (a hair north) of N571 (see his Figure 3).
@DrB: I should add that after performing some hand calculations, I ran a PMDG 777 simulation of following airway N571 past MEKAR towards NILAM using LNAV. As you say, the path is almost right over MEKAR and stays very close to the line connecting MEKAR and NILAM (less than 100 m deviation).
@DrB
Nice. Thank you.
Bobby: Thanks for that detailed radar analysis of the turn possibilities at MEKAR. Which way and when MH370 turned or descended after MEKAR remains a mystery, but it is certain that the radial speed toward the s/c dropped significantly between 18:22 and 18:28. Note that first of the 7 18:25-18:28 cluster BTO values was consistent with the average 18:02-18:22 speed (~505 kts). But all of the next 6 BTO values starting about 100 seconds later are consistent with a lower speed toward the satellite. This suggests a turn started circa 18:25:30.
Again it seems to me the final radar capture at 18:22:12 must be somehow related to the re-log-on.
The coincidence of those two events happening at the ~same time with turns, descents or/and decelarations after 18:22 just cann’t be coincidental imo.
What happened there at the edge of Western Hill radar range?
@Dr B. Thanks for yours. How do you account for the rather different interpretation of speed from Ron Belt’s paper (485 -> 508 ->511kts GS)? It looks particularly strange when we also consider (from 1800Z GDAS wind) that there should have been a tailwind of 18-20kts near Penang, declining quite quickly and <<5kt tailwind near 1822.
I come back to the same question posed earlier. How confident can we be of the timestamps? It seems to me that there are only about three that are clear enough to be unequivocal (of which one is a clear outlier in both time and position) and the others are educated guesses.
If we were not constrained by assumption of constant speed along LIDO trace of ~M0.84 then a slow down would be a perfectly reasonable explanation for BTO/BFO at 1825-1828 (see earlier post).
ALSM: The first BTO at 18:25:27 is on an R600 channel. The DSTG advises that the SD of the error for an R600 channel is around 62 μs, versus the 29 μs for the R1200 channels. There are straight paths with turns before 18:25:27 that satisfy these error criteria and also satisfy the BFO error criterion.
@DrB
Thank you for the data work up and documentation. I am thinking your 8-Jan path has the point slightly north of N571.
Victor: Yes, I agree with your point, but it does take an assumption that the BTO was off at one extreme at 18:25:27 (12580 vs. 12520 usec recorded). Maybe it was off as much as 60 usec. If so, then a turn before 18:25 is possible, and in that case, all 7 BTO values indicate an average speed of 332 kts after 18:22, down from 505 kts for 18:02-18:22. Average speed in this context means the speed continuing along the previous track. Of course, a turn away from that track while maintaining the same ~500 kts speed is perhaps more likely than a speed reduction along the same path.
@Perfect Storm
“Jeff Wise has solved the MH370 enigma!”
I put my conjecture over on JW site
@ALSM said: Of course, a turn away from that track while maintaining the same ~500 kts speed is perhaps more likely than a speed reduction along the same path.
I think the requirement of satisfying the BTO and the BFO requires a deceleration and a turn. Part of the challenge is knowing what was the position and time of the last radar capture. I’m not sure we even know the measured position because the point falls off the Lido radar slide.
@TBill: There’s no mystery as to where he thinks the plane went. Except for a couple of short periods where he’s wavered, he’s consistently put the blame on Russia, which he believes tampered with the SATCOM to alter BFO values to masquerade a flight to Kazakhstan. (As some of you know, I thought that was a possibility before all the debris started showing up and the simulator data became known.) His efforts have been consistently focused on casting doubt on the validity of the debris, the simulator data, and anybody associated with this evidence.
@Andrew
Hi Andrew, I’m Ron’s friend. Re “At least two of those items are wrong, but I won’t bore you with the details…
BTW, what’s a fuel ‘transfer’ pump?
@Andrew
Hi Andrew, god knows what a transfer pump is. He gets most of his dodgey facts from Smart cockpit. I know, I’ve told him before not to rely on that stuff. Anyway, he’s not answering his phone now, so I can’t much help. His doctor has told him to give this up, for his health’s sake.
Bottom line is I don’t think he will be bothering you again with is bluster, braggadocio and fake news.
BTW, you are doing a great job.
@Victor
The collegiality and respect on this blog are refreshing…
@Rob: I don’t know who Ron is, but you are the one that in a previous comment introduced “transfer pump” into this discussion. I have little tolerance for people playing games with multiple names.
@Victor. I have set out why I feel that a deceleration without a turn is sufficient to be in compliance with BTO and BFO 1825-1828. Could you please explain why you insist that BOTH a turn and a deceleration are required?
@ALSM
I think sk999 is telling me the 18:25 BTO has more error:
“TBill, BTO at 18:25:27 is an R600 channel packet, and even after correction, it can be in error by 100 to 200 microsec (as seen in the MH371 logs).”
@George
Yes, a good group for sure.
BTW, with respect to your earlier query. If I use a spherical earth and your starting coordinates of 0.5447/64.4638 and your ending coordinates of -37.6/86.8 I get a distance along the surface of the sphere of 4821km which is 100km short of your 4920 distance. If I use ending coordinates of -37.6/88.8 I get a distance along the surface of the sphere of 4922km which matches your 4920 distance. So your error is not related to ellipsoidal effects as I speculated earlier.
BTW, the link below is a handy tool for spherical earth calculations.
http://www.movable-type.co.uk/scripts/latlong.html
@Paul Smithson: How does your deceleration satisfy the BFO without a turn to the right? Please state your 18:22:12 position, and then whatever decel and track you propose, and I’ll check it.
@George
BTW2, if you look at the code fragments in my link above you can see that the author uses an earth radius of 6371km which is the radius of the earth at the equator.
@Rob
I’m not pretending I’m right on the pal. That are your words.
To me it’s only still the most logical view the left AC bus was deliberately isolated around IGARI. Disabling ACARS and IFE at once.
It would have been easy done in the cockpit from the captains chair for the panel to do it was right above him.
Switching off the transponder would only have taken a second more.
I still don’t rule out completely an accident either like @Oleksandr proposed; an exploding nose wheel. But very, very unlikely given the events after IGARI.
To be short; I’m not at the same page with you the way you pretend it sometimes. In cases you also take my views and words but into your own context.
I’m not as far convinced as you express yourself.
Please be more thoughtfull when including me in your views.
@DennisW
Yes I have used many of his (movable-type) equations.
The other source that I have found extremely useful is Michael Geyer’s “Project_Memo_DOT-VNTSC-FAA-13-08.pdf”
@Andrew,
Re: “It’s not specified in the MAS FCOM. Where I work, the L or R transponder is selected depending on which pilot is nominated as PF.”
Thanks. Would it be correct to assume that the left transponder was more likely in use on 9M-MRO at the moment of disappearance?
@ALSM,
Re: “The OCXO is in the SDU. There is only one SDU associated with the HGA and LGA.”
Thanks for your comment. What I meant is whether there is only one OCXO in the SDU?
@Peter Norton,
Re: “You say this because on SR111 after AES reconnected no ACARS messages were subsequently exchanged, right ?”
In part. For some reason almost everyone assumes that any technical failure/malfunction will be quickly resolved by a crew. Everyone expects ideal solutions in accordance to the handbooks. But almost everyone forgets that the Captain and FI were also humans, and they could do mistakes under high stress conditions dealing with unknown problems, despite training and experience. We are chewing this for 3.5 years; the crew perhaps had less than one hour to make right choice of actions, or even seconds for some key decisions.
Oleksandr: There is only one OCXO for the entire AES.
@Victor.
The tool that I’m using is a modification of Barry Martin’s path modelling tool with 1s resolution, true/magnetic heading hold and altitude FL340. As I haven’t mucked around with the BFO and BTO calculations those should be fully compliant with IG method.
At 18:22:12 I have position 96.467E 6.516N. Travelling at M0.700, FL340 with TAS 415.1kt true heading 290, GS 415.6 azimuth 289.46. Predicted BTOs (residuals in brackets) as follows:-
18:25:27 12600 (+80); 18:25:34 12595 (-5); 18:27:04 12528 (-32 and +8); 18:27:08 12525 (+5) 18:28:06 12482 (-18); 18:28:15 12476 (-4).
Predicted BFO throughout 140-141Hz.
All I’m saying is that it is perfectly possible to cross the 1825-1828 ping rings on the same track as the LIDO slide, but having lost some time by slowing down. In this case, we have slowed from M0.84 to M0.70 at 18:17:00.
Of course this does mean that the 182212 position per LIDO has to be wrong by >10NM and I agree if you take that as your starting point then there is nearly no amount of reasonable deceleration without a turn that is going to avoid “hitting the buffers” on the 1825-1828 BTOs. However, slowdown along with no turn will do the job if you begin your slowdown earlier (as illustrated in the case above). And very easy to accomplish if you make no firm assumption on constant speed beyond 1802.
@Ge Rijn,
Re: “I still don’t rule out completely an accident either like @Oleksandr proposed; an exploding nose wheel”.
More accurately I suspect the L/H tire rim. Do remember the missing bolt? One possible reason is a micro-crack in the rim, which appears to be made of aluminum. The environment in KL is very humid, and it is easy to imagine that some condensate or rainwater could enter such a crack after previous flight. Besides aggressive oxidation and weakening the tire rim, the water in the crack would eventually get frozen during subsequent ascent. In addition, the metal becomes less elastic as temperature drops. To make things worse, the nitrogen inside the tire remains warm for a while due to the low heat conductivity of rubber, so that I would expect the highest pressure differential upon reaching the cruise altitude. When the rim fails, L/H tire explodes. It is just a couple of meters from the ADIRU, and other essential electronics, and a less than a meter from the emergency oxygen supply line to the cockpit. Draw a line between the L/H nose landing gear in stowed position, the ADIRU and electronics at the left rear side of the EE-Bay.
@ALSM,
Re:”There is only one OCXO for the entire AES.”
Noted; thanks for confirming this.
Beware the waypoints!
It is easy to become enamoured with waypoints and airways and piece together routes that seems reasonable as to where the plane went when, in fact, they are spurious.
The so-called “Second refinement surface search area” (Final Report, p. 29) suffered from such a failing. It was at this time that the so-called “19:12Z” “northwest point” was identified. The actual position was 8.595N, 92.586E, placing it 35 nm east of LAGOG. The JIT latched onto this fact, and for Scenario 1B, it assume the aircraft did, indeed make the FMT at LAGOG. Why? Because the JIT was apparently enamoured with waypoints.
JIT Report, Issue 2 (Appendix B of Final Report):
“Scenario 1B required the aircraft to fly a northern route through waypoint IGREX and waypoint LAGOG, before flying south. The LAGOG waypoint was included as it was close to a supplied, high confidence position labelled ’19:12Z’.”
“The use of stored waypoints in the aircraft Flight Management System (FMS) suggests that the navigation was carried out using the Automatic Flight Control System (AFCS) in the lateral navigation mode. When the aircraft departed LAGOG the next major waypoint to the south was COCOS, which could be linked to the M641 air route. This would have taken the aircraft towards Perth. Once this route was set it would need no further input from the pilot for the aircraft to maintain altitude and route using the AFCS and auto throttle system.”
“The area that was crossed by air route M641 was classified as red (Priority 1).”
This hypothetical route was backed by fits to both the BTO and BFO, endurance calculations, and an independent flight plan analysis by Malaysia Airlines. The Ocean Shield search area was close to where M641 hit the 7th arc.
I have previously indicated that the BFO calculations must of been in error. Also, a route forced from LAGOG through COCOS has rather large BTO errors. If one relaxes the constraints on the route to better fit the data (as was done for Scenario 1A), one is no longer flying via waypoints or airways.
By June, 2014, we found out that the “1912 NW point (80d 35.719, 92d 35.145) [was] reassessed as invalid and no longer used by flight path reconstruction groups.”
Goodbye LAGOG.
———————-
Note – DennisW recently said that, after a failed project, he would try to explain what he would have done differently. To do so one needs to understand what caused the failure in the first place. The above is part of my attempt to understand why the JIT and ATSB kept moving their priority search areas so much.
@all,
In response to comments on my 18:22 turn analysis:
1. The location of the last radar position is determined using the constant horizontal offset and alignment of the timestamp. Fortunately, this particular timestamp is clearly legible because it is not overwritten by other timestamps or waypoint labels. I did not use the arrow in the slide. I carefully measured the locations of other timestamp labels with respect to their radar position crosses. The relative locations are constant. In the vertical direction, the cross is exactly in the center of the label text. In the horizontal direction, the offset distance to the left is also constant with respect to the timestamp. Thus, one can locate the 18:22:12 position quite precisely based on offsetting a measured distance to the left of the label.
2. Using only a turn, it is impossible to match all the 18:25-18:28 BTOs with a turn starting after 18:25:27. After all, in the 7 seconds between that BTO and the next one the aircraft can travel at most 1 NM, so you can’t expect any turn to change the BTO significantly in only 7 seconds. That first BTO at 18:25:27 has a much larger error than all the other nearby BTOs, as Victor has reminded those who have forgotten. A turn beginning after 18:22:12 and before ~18:25:00 can provide a BTO that is in between the 18:25:27 and 18:25:34 values and is also consistent with their error bars. Figure 4 in my February 27th paper on the OCXO transient shows the BTO data with 2-sigma error bars. That figure also shows how a 15NM R SLOP matches the BTOs. There are other possible solutions involving single turns and decelerations, as we have been discussing for the past week.
3. I don’t think there is any connection between the last radar contact and the SDU log-on, other than the fact that they occurred within a few minutes of one another. The last radar contact position appears to be driven by the 250 NM maximum range of the radar at Butterworth.
4. Predictions of ground speed based on noisy radar positions require rather long periods of time to provide useful estimates of speed. For instance, suppose the (along-track) position errors are 1.5 NM 1-sigma. That position noise results in a speed error of 6 kts 1-sigma with a 20 minute track. For a 10 minute track the speed error is 12 kts 1-sigma, and this seems consistent with Ron Belt’s calculated speed variations. In other words, the radar data are not precise enough to conclude that the speed actually changed by 10-20 kts on a timescale of 10 minutes or less. The tailwind changes are also too small to be seen in the radar except in an average sense over a long track.
5. Very few of the timestamps are fully readable. The only timestamp I used in this analysis 18:22:12.
6. My January 9th paper used the ATSB’s interpretation of the last radar contact being 10 NM past MEKAR along N571. In the future I will use the 18:22:12 position which is several miles west of N571.
7. In response to a specific question from me, ATSB has told me there have seen no indication of any timing errors in the military radar track.
8. It is possible a deceleration (or even a descent) began within several minutes before 18:22, but the ground speed reduction by 18:22 could not have been very large, although it could have been large by 18:25.
@DrB: Thank you for the clarification. Some comments/questions:
1. Can you provide the exact position you derived for 18:22:12? That’s helpful in comparing analyses. I agree with your assessment of using the timestamp as an indicator of the actual radar target.
2. If the actual position was really that far from N571, then N571 was not being followed in LNAV. Either the plane was tracking west towards SANOB in LNAV, or it reached a route discontinuity at MEKAR, or the plane was not in LNAV after MEKAR. For any of these cases, the SLOP along N571 could not have occurred. So what’s your best estimate of how the BTO and BFO are satisfied?
3. It is easy for a pilot to create a fix with a selected radius, which appears on the ND as a dotted circle. A pilot would therefore have a visual clue as to when radar coverage ended. As you say, the distance between Butterworth and the last radar point is about 250 NM. (It is a little closer to the actual radar installation at Western Hill, but we don’t know if the pilot knew the distinction between Butterworth and Western Hill.)
4. It’s possible that the re-powering of the SATCOM could have been timed with either passing MEKAR or being out of the 250-NM radar range.
@Oleksandr
Re: “Would it be correct to assume that the left transponder was more likely in use on 9M-MRO at the moment of disappearance?”
I found the information in the airline’s Ops-A manual. MAS policy is to select the left transponder and weather radar systems on outbound flights from KUL. Accordingly, it’s reasonable to assume that MH370 had the left transponder selected.
@Victor
“4. It’s possible that the re-powering of the SATCOM could have been timed with either passing MEKAR or being out of the 250-NM radar range.”
My thoughts have always centered on the last part, ie, “getting out of radar range”. The question however, is out of “which” radar’s range.
Simply put.
Getting outside the 250 Nm radar range ring of PSR Butterworth is the most obvious scenario (for a deliberate act), since it is well known by just about everyone, that in general, the effective / useful radar range limit is 250Nm at 40,000 feet for a radar at sea level (which Butterworth PSR effectively is).
MEKAR is 241 Nm 285 Deg True from PSR Butterworth.
So, 10 Nm past Mekar thus conviently fits the plan / story, (if you beleive it – which I don’t).
However, Western Hill PSR is up on a hill at 2,670 feet.
It’s effective radar range at 40,000 feet is about 300 Nm.
Assuming it was a deliberate act, and assuming he knew that, he would have had to wait at least 6 more minutes to get outside “it’s effective radar range of 300 Nm”, in other words, he had to go another 50 Nm or so, until he was about half way between Mekar and Anoko.
Now draw a track line Mekar to Anoko.
Notice that the 18.4515475925926 UTC and 18.4695566666667 UTC ping rings now intersect that track smack dab in the middle, precisely where you would expect them to, (if he was smart) ie, they “fill the bill” quite nicely.
All of that makes sense, if he was actually out there.
The trouble is, it does not fit with the 18:22 radar point(s).
So, do we accept the 18:22 radar point as genuine or not, that is the crucial question.
@ventus45:
it is well known by just about everyone, that in general, the effective / useful radar range limit is 250Nm at 40,000 feet for a radar at sea level (which Butterworth PSR effectively is).
Information is publicly available that says for the civil radar at Butterworth, the PSR range is 50 NM and the SSR range is 250 NM. Perhaps he was confused about the location of the military radar installation.
Assuming it was a deliberate act, and assuming he knew that, he would have had to wait at least 6 more minutes to get outside “it’s effective radar range of 300 Nm”,
The range determined by the horizon does not necessarily determine the range of the radar. It is a limiting factor. The actual range could be less based on time gating or power.
Notice that the 18.4515475925926 UTC and 18.4695566666667 UTC ping rings now intersect that track smack dab in the middle, precisely where you would expect them to, (if he was smart) ie, they “fill the bill” quite nicely.
It would be very difficult to power up the SATCOM and precisely time the log-on, as you are proposing.
@Oleksandr. The missing nosewheel bolt.
You say of a nosewheel explosion, “One possible reason is a micro-crack in the rim, which appears to be made of aluminum”.
Many moons ago you may recall I raised the possibility of a nose wheel failure caused by successive failures of remaining bolts joining the two wheel halves, their having been weakened.
See, FI Appendix 1.6A p4, Recent Technical Log entry 47, “..– 28 February 2014..found one of tie bolt missing at nose L/H tire”.
While having no knowledge of this aircraft or wheel, the first thought I had on seeing that was the possibility that the missing bolt had failed in tension. A quite likely cause for that would be nut overtightening. As I mentioned back then, overloading for this reason has caused loss of aircraft and I personally know directly of two. Overtorque can be from; torque wrenches not being calibrated, including periodically; the torque requirement not being looked up in the manual; or not using a torque wrench (unaware that there is a torque limit or else in a hurry).
Commonly this assembly work will be done in a tyre bay dedicated to this though there can always be mistakes of supervision and checking. I do not know what action might have been taken to check that all the bolts had not been overtorqued at the time the failed bolt was secured or whether a check was done as to the possibility that bolts in other wheels in the fleet had failed or had been overtightened – which would require replacement of all that had been. Besides, failure of one for any reason would place added load on those adjacent.
The corrective action taken by the airline is entered as, “28 February 2014 Inspected wheel found satis. No leak pressure c/o tire press chk 198 psi with in limit.Transferred to MR2 per MEL K 70327” The MR2 transfer is to a list of defects deferred for future action. The deferral remained uncleared as of 8th March.
Having posted on this subject, I have not pursued it even though a nosewheel explosion could damage the main electronics centre electronics cabinets and wiring and possibly wreck the battery. Oxygen lines could be severed and regulators or cylinders ruptured. A general or local concentration of oxygen in the centre would increase prospects of fire for a while at the same time denying oxygen to the cockpit, at high altitude, at night. An oxygen bottle rupture most likely would lead also to more physical mayhem.
My reason for not pursuing this was that I could not tie that together with subsequent navigation and lack of descent, except through prolonged mild hypoxia of one conscious crew member who kept flying, even though such an explosion could explain the aircraft going dark and its lack of communication (AC and DC all gone and the APU and RAT delivery being limited due to shorts).
If there were evidence of any individuals who might persist in a mild hypoxic state for hours, particularly at high altitude, this theory would have legs IMO. I did post that I had consulted with an ex Air Force doctor. However he doubted the possibility and I have seen no evidence supporting it. Having been in a decompression chamber myself, as have others here, I have no recollection of a hazy period before passing out or on recovery though of course that is totally subjective and not much of a sample. The only instance I have read of which might be relevant is the cabin attendant who took to the Greek 737 controls after the crew has succumbed to hypoxia but that is tenuous at best.
However since you still have nose wheel explosion in mind as a possibility you might like to add this possibility, though you might have (and had) a good reason for not doing so.
@David
Do you suppose Shah, being the conscientious guy he was, simulated such an emergency? Please, the simulator data, IMO, makes all mechanical scenarios completely untenable.
Likewise the debris finds make Northern diversions untenable.
@Oleksandr
Re: “Would it be correct to assume that the left transponder was more likely in use on 9M-MRO at the moment of disappearance?”
One further point, I believe the FO was the pilot flying (PF), not the Capt. The analysis of the radio transmissions from MH370 in sub-section 1.5.10 (Air-ground Communications, p.21) of the FI report indicates the FO made all the transmissions before take-off and the Capt did likewise after take-off. That pattern would be consistent with the FO acting as PF.
MAS policy is that the aircraft must be taxied by a Commander or Command Trainee in the left hand seat, although right hand seat pilots may be allowed to taxi when operating with a TRI/TRE in the LHS. It is also MAS policy that the Commander or Command Trainee in the LHS performs all engine starts. In this case, I believe the radio transmissions show the Capt was taxiing the aircraft even though as a TRI/TRE he could have allowed the FO to do so. The FO would then have taken control of the aircraft once it was lined up for take-off. The FO acting as PF would be consistent with him operating his last training flight before he was checked-out, as stated in sub-section 1.5.3 (First Officer, p. 14) of the FI report.
@Olekansdr. Clarification.
FI p28, deferred defects outstanding, does not list this as being either cleared or outstanding though the log goes back to 2013. It appears it was never transferred to the MR2, nor apparently was it cleared subsequently in the Recent Technical Log entries.
@David
@Oleksandr
Re: “The corrective action taken by the airline is entered as, “28 February 2014 Inspected wheel found satis. No leak pressure c/o tire press chk 198 psi with in limit.Transferred to MR2 per MEL K 70327” The MR2 transfer is to a list of defects deferred for future action. The deferral remained uncleared as of 8th March.“
The applicable MEL item states that the affected wheel must be replaced after a maximum of five subsequent landings. The list of Deferred Defects (MR2) at sub-section 1.6.3.7 of the FI report does not show a deferred defect for a missing nose-wheel tie bolt, which seems to indicate the wheel had been replaced and the deferred defect cleared. However, there is no record of that action in the list of Tech Log entries. That does seem unusual, but may be related to the way MAS documents the clearing of deferred defects. There is another defect at Item 47 of the Tech Log entries for seat 4K IFE inop; that defect was also deferred and there is no mention of it having been rectified, however, it does not appear in the list of deferred defects.
@VictorI,
1. My derived coordinates for the 18:22:12 Lido position are: 6.547 N 96.344 E.
2. I think that at 18:22:12 the aircraft could have been (a) enroute to ANOKO (or with a lower probability to SANOB), (b) flown by the autopilot in constant magnetic heading after a route discontinuity at MEKAR, or (c) flown manually. I don’t know why a pilot would revert to manual flying if he had just flown in LNAV between VAMPI and MEKAR, especially since we think the autopilot was functioning perfectly. The possibility of a route discontinuity occurring at MEKAR, instead of at IGOGU, as we assessed previously, is intriguing. Why fly to ANOKO? The only reason that comes to mind is the ANOKO-2C approach to WITT.
3. I also agree that, if the probability that N571 was being followed then was very low, it is also very unlikely that a 15 NM SLOP was used. Based on my understanding, the SLOP is normally only used when one is on an airway.
4. Regarding matching the 18:25-18:28 BTOs/BFOs, I will work on this some more and put out a note after I finish the INOP endurance study.
5. You said: “It’s possible that the re-powering of the SATCOM could have been timed with either passing MEKAR or being out of the 250-NM radar range.” It is difficult to prove the negative, but I am skeptical of this because turning on the SATCOM again is antithetical to staying hidden. What “harm” could occur if the SATCOM were on before the maximum radar range were reached? Also, what “benefit” would occur when out of radar range that would not also be of benefit while still within radar range? My gut feeling is that the two events are not directly related.
@David
@Oleksandr
Re: “It appears it was never transferred to the MR2, nor apparently was it cleared subsequently in the Recent Technical Log entries.”
That’s another possibility; if it wasn’t transferred to the deferred defects then there would not have been a trigger in the maintenance records to replace the wheel. If that occurred, it would be extremely sloppy maintenance practice, especially as the aircraft flew a number of subsequent sectors where the missing tie bolt should have been picked up and properly documented.
@DennisW
Re: “Do you suppose Shah, being the conscientious guy he was, simulated such an emergency?”
MAS policy expressly PROHIBITS the simulation of abnormal/emergency situations while passengers and/or cargo are being transported.
@DrB
What “harm” could occur if the SATCOM were on before the maximum radar range were reached? Also, what “benefit” would occur when out of radar range that would not also be of benefit while still within radar range?
Yes, turning the SATCOM off/on makes little sense other than to arm wave the BFO values after 18:25.
@Andrew
I was referring to his home simulator. The coordinates found there are simply impossible to ignore. I put people in the mechanical failure scenario camp in the “whacko” category.
@DennisW
Whacko!? Come on now, Dennis, not so soon after George was complimenting Victor on the collegiality and respect on this blog.
@Mick
I respect whackos. I live in California.
@Andrew. “…..if it wasn’t transferred to the deferred defects then there would not have been a trigger in the maintenance records to replace the wheel. If that occurred, it would be extremely sloppy maintenance practice….”
Exactly so. On the face of it that aircraft was flying from 28th Feb, at the latest, to 8th March with the bolt missing and that would have continued indefinitely.
I see the IFE entry, number 48 (K70328), was of the same 28th Feb date as 47 (K70327). As it appears, neither was transferred: it could be the same person was involved and both possibly had a common cause though there should be careful and experienced management and cross-checking of such documentation so that such slips are caught.
I had noticed earlier that 2 defects deferred from 5th March had been transferred properly, 25 (K70330) and 14 (K70331), which is what to me added to the likelihood something was amiss on 28th Feb.
It could be that they have a parallel system and I hope they do but it is pretty esoteric if so given the serial numbers are all K and in sequence.
As currently it looks this is potentially more serious as a maintenance failure than the expired battery, even though it might not have been of no consequence.
@DennisW. “Do you suppose Shah, being the conscientious guy he was, simulated such an emergency? Please, the simulator data, IMO, makes all mechanical scenarios completely untenable”.
Well I think I made clear that I could not draw this failure together with other evidence, though that does not mean in my book that that could not ever be done.
So,”completely untenable” is for me a conclusion which is premature though I do appreciate that is your opinion and your scenario is certainly on the table still while this other currently is not.
@DrB @Victor
I like N571 SLOP, but alternately I like MEKAR to DUBTA…as I feel the simulator N10 point is possibly from the trajectory DUBTA to DOTEN
@DennisW. I think the simulation would be more conclusive if a good reason for it were apparent. I have put forward as have others that it might have been some sort of get-the-feel but that is a presumptuous reading of his mind and grasping at straws all the same.
Even accepting the whole simulation was continuous, its sections having been on the same date, of particular curiosity was in how much of it did he fast forward and also the way the fuel quantity was shuffled back and forth without apparent trend or reason. Likewise, why the simulation take off with the fuel he did then fiddle like that, to end up way away from where MH370 could with its load?
Those who suppose that in the event he had an end site in mind overlook that the final flight fuel quantity apparently was spontaneous. He had not simulated the trips across the Peninsula and yet took fuel which was suitable just for a Peking flight and used a deal of that before joining the simulation route, having apparently chosen this flight where he would have nothing like the fuel needed for the simulation. I mean, the MH370 fuel exhaustion was random.
That does not disqualify your scenario but sensibly it does at least those avowing he was aiming for a particular hole and it leaves what the hell he was doing shifting fuel about in the simulation high and dry.
Moreover I am puzzled why he would be rehearsing in some way a future event where by many accounts that was a spontaneous reaction to a court decision that day. In early February would he have known of the court hearing, its announcement date and outcome? Did he then ensure he was flying that day or was that by chance as was the court announcement?
I feel still he might have been testing his software, for instance after it all had stopped working as his brother-in-law plainly says (thanks Ventus45 for the video), though I have no evidence at all for that.
@DennisW
Re: “I respect whackos. I live in California.”
Very good, Dennis, you can always be counted on to lighten (and enlighten) proceedings.
@David says: I feel still he might have been testing his software, for instance after it all had stopped working as his brother-in-law plainly says
The logs show that the FSX/PMDG777 installed on MK26 was crashing. I see no evidence that the FS9/PSS777 installed on MK25 was crashing. Relatives also said the simulator was broken for a year and could not be used. That’s false.
@Victor
@TBill
Gentlemen, a couple of flight sim questions if I may;
– Under the “fair” weather setting can you ascertain what the winds would have been at FL400 at 10N (101059N0901328E)?
– What date/time does the sim use/recognise? PC system time as local time? UTC? Something else?
According to this and other stories, the Malaysian Government has confirmed it has chosen a company to begin a new search for MH370 and is now negotiating the terms of the deal.
The Malaysians change the story day by day.
Another source that suggests also the search will be resumed by OI:
http://www.dailymail.co.uk/news/article-4996262/Ocean-Infinity-resume-search-missing-flight-MH370.html?ito=social-twitter_mailonline
@Victor. “I see no evidence that the FS9/PSS777 installed on MK25 was crashing”. Fair enough thanks.
One of the NOK (Jimmie Wan) posted this notification that OI has been selected: https://goo.gl/KuLK91
Does anyone know the size in bytes of a flight file saved by FS9/PSS777?
@ALSM: The deal is not yet signed to re-start the search with OI, but I will increase the probability from my previous “greater than 50%” to “approaching 70%”. Things are certainly moving in the right direction.
@DrB @Victor
Could be VAMPI to DUBTA as a way to keep on LNAV setting…can you do that with ANOKO? I suppose but not as well.
@Gysbreght
The FLT file size seems to vary from about 60 kb up to 160 or more. The 60 is from start of flight take off from KLIA with Victor’s suggested simulator path.
@Mick Gilbert
It seems the wind stays constant at 25 knots from 271 degrees up to FL430 as far as I can see.
The user can set time and date. I can see twilight start at approx. the right time on 8-March-2014 in the SIO. I assume ATSB is talking about file save time stamp. There is much sim info we do not know from the truncated files, like waypoints, time of day, etc that would help a lot if we knew it.
@David
The fact that MH370 did not have as much fuel as the sim study could explain why the sim study path was not exactly followed. In my mind the actual flight might have gone 180S to get into deep water past Broken Ridge at 93-94E instead of McMurdo, but I am not ruling out McMurdo waypoint either.
A statement from Darren Chester says that Malaysia has accepted Ocean Infinity’s offer.
He also implied that OI will search the area designated by CSIRO to be high probability.
@Oleksandr
To come back on the nose wheel tyre exploding at altitude due to a damaged/weakend rim. I also said I think this is very unlikely regarding the events after IGARI.
If this happened the EE-bay had to be penetrated by fragments causing the electronics damage that silenced the transponders and SATCOM/ACARS and would have caused decompression of the cabin while the EE-bay is a compressed area.
It seems no emergency descent was made after the ‘going dark’ around IGARI. As it seems the plane kept ~cruising altitude at least till 18:22.
Nothing you would expect after such an emergency. No communication too while it’s very hard to imagine all radios and means of communicating where wiped out in such an event. And if it was, no sign of any other attempt to try and show someone the plane was in distress.
By flying low near an airfield or something like that. Just flying circles above the sea near an airport long enough would have alerted people.
Or even try to reach someone by mobil phone on low altitudes.
Nothing like this seems to have happened regarding the data.
Then also the found nose gear door piece. Won´t you expect those doors to be blown away by a exploding nose wheel?
@TBill
25 knots from 271 degrees, thank you for that.
@VictorI
Congratulations. This is very good news first to the NoK.
Your relentless efforts and of many others have contributed to this brave decision of Malaysia and Ocean Infinity I’m sure.
Hopefully they are in time to start searching this winter (summer in the SIO).
@Victor
Re: “A statement from Darren Chester says that Malaysia has accepted Ocean Infinity’s offer.”
Terrific news! And it seems that our Minister for Infrastructure and Transport is up late (or at least his OIC-Twitter is). Good to see he’s already hosing down any notion of “troops home by Christmas”.
@Ge Rijn: I appreciate the sentiment, but congratulations are only in order when the plane is found.
The next question is going to be where to search. I think it is wise to start with CSIRO’s recommendations. After that, I would prioritize a number of hot spots based on path reconstructions, drift models, previous underwater search results, and previous surface search results. I think it would be hard to recommend any one spot with a high level of confidence.
@TBill said: It seems the wind stays constant at 25 knots from 271 degrees up to FL430 as far as I can see.
I believe it is more accurate to say 25 knots from 270° True at high altitudes. The magnetic direction would vary according to the local magnetic variation.
@Mick @TiBill @Victor. If you are talking about wind profiles along a track WNW from VAMPI at altitude FL340, then I think you will find there is a diminution of wind associated with the cyclonic feature lying north of the tip of Sumatra (easy to visualise https://earth.nullschool.net/#2014/03/07/1800Z/wind/isobaric/250hPa/equirectangular=96.82,4.54,1538/loc=90.500,-35.200).
Using 1800Z GDAS weather at 250hPa I’m seeing a tailwind of 4kts at 1825 and 2 knots at 1830.
@VictorI
It’s a big leap which deserves due congratulations in a way, whatever the outcome (sentiment involved I know).
Most important is OI going to start. At CSIRO’s recommendations to start with is of less importance. It’s a good point to start imo.
I don’t expect much of this area but it has not been searched yet and sure falls in probability area.
From there things will evolve and you and everyone else involved will be carefully on notice and providing any support and critical evaluations along the way.
Probably even some more pieces will be recovered in the meantime also.
Ocean Infinity won’t stop easily based on such a contract. They’ll do the utmost and seem to be quite confident.
I’m optimistic. I expect them to search at least 25.000km/2 till ~31.5S and include a possible glide after a steep descent within ~50/60 miles of the 7th arc on each side.
@PauL
We are talking about the Microsoft Flight Sim 2004 (also know as Flight Simulator 9 – FS9) where the user can set weather and winds. One of the ZS saved cases indicates “Fair Weather” was selected, which is a 25 Knot wind from the West 270deg. Mick has previously pointed out that is such a strong wind that 180S mag heading from DOTEN ends up about the same place as NZPG waypoint at the 45S1 poinbt of the simulation.
@Ge Rijn: I have already had many communications with Ocean Infinity and I know they read this blog. I also support starting with the area recommended by CSIRO.
@Victor
I also support starting with the area recommended by CSIRO.
A CSIRO model supported the original high priority search area.
From the Conclusion section of a 4 August 2015 ATSB publication:
https://www.atsb.gov.au/publications/2015/mh370-drift-analysis/
The drift analysis undertaken by the CSIRO further supports that the debris from MH370 may be found as far west of the search area as La Réunion Island and is consistent with the currently defined Search area.
The reality is that there is no new information to inform the search. A CSIRO model also predicted debris to be found on the West Coast of Sumatra and Indonesia (same reference as above).
If the 19:41 arc position was above 5N (and I personally think it was), then 35S is too far South.
@TBill. Ah, ok. Sorry for the misunderstanding.
@DennisW: It’s a place to start the search, not to end it. I’d rather start to search a location where there is not total agreement rather than not search at all. There will be opportunity to search elsewhere.
@Victor. Glad to hear that you haven’t assigned all eggs to that particular basket. I’m hoping that you also have an appetite for looking “short and long” along the max probability path according to just about everybody (until the plane wasn’t found on the 7th arc there).
@Paul Smithson: I have an appetite for searching lots of places. But it’s not my money, and it’s not my decision. The good news is that Ocean Infinity will be very motivated to find the plane.
@Victor
I see what you mean…if the path was VAMPI to DUBTA to DOTEN and DUBTA was deleted after MEKAR then the track would be towards 306 deg. DOTEN (VOCX for your case). Do we know altitude of UAE343? as we maybe gotta go under it in that case.
@TBill: That’s not exactly the case I considered, but close enough. Word to the wise: don’t get too attached to any one scenario.
@All: of all the people publicly identified as receiving any portion of these [pdfs of scans of pieces of paper with the RMP logo on it], who was the first?
Thanks in advance for the name.
@Andrew,
Thanks for your comments with regard to the transponders. The reason I asked is as follows. This is in my understanding that:
1. The Left ATC/Mode S transponder gets 115v AC power from the AC Standby bus;
2. The Right ATC/Mode S transponder gets 115V AC power from the Right AC Transfer bus;
3. The dual transponder panel gets 115v AC power from the AC Standby bus.
The standby bus can be powered either by the L transfer bus, or by the static inverter, and therefore the left transponder can be considered as more reliable compared to the right one.
This is in my understanding that in case of a severe mechanical failure, the transponder would automatically transmit a message. One of the peculiarities of the mechanical failure scenario I am pursuing is that the inverter is installed at the E3 rack, just left-above to the ADIRU. If the standby bus power is completely lost due to a short-circuit of the left transfer bus and damaged inverter, then the transponder panel would not be functional, correct? Would it be possible to switch to the right transponder in this case? Also, does the transponder automatically transmit a message when the main AC bus is deliberately isolated from the cockpit?
@Victor: correction: congratulations are not in order until NoK have the whole truth of MH370’s fate. Finding seabed debris is not the same thing as learning the whole truth.
@All: I’ve asked twice before in this forum whether search leadership’s obligation to recover, examine, conclude and report on any “found” debris is any different under this new proposal than it would have been had the official search not been suspended. I think this is an important question, but unless I’ve missed it, I’ve received no response to date.
Is the JIT’s legal obligation just as strong either way?
Thanks in advance for all referenced answers.
@David,
@Andrew,
It was actually David’s question about a possible cause for the bolt to fall out, which ‘inspired’ me to look into the possible outcome of the tire explosion. Improper tightening, defect of the bolt, or defect of the rim?
It was quite surprising to me when I realized that the bolt was missing at L/H wheel; the ADIRU is at the left; the emergency oxygen pipe to the cockpit is routed at the left side; the oxygen tank is at the left side; the inverter powering the AC standby bus is at the left side; VHF-C and VHF-L are at the left side; TCAS; Intercom, the left TRU…
I have not heard anything with regard to the missing bolt since this issue was mentioned in FI. Though, I recall some MAS maintenance logs were burnt.
@Ge Rijn,
Re: “Nothing you would expect after such an emergency. No communication too while it’s very hard to imagine all radios and means of communicating where wiped out in such an event.”
Except the right VHF, not too hard. One high-energy fragment of the tire rim, followed by the isolation of the main buses (to prevent suspected fire spread).
Re: “And if it was, no sign of any other attempt to try and show someone the plane was in distress.”
How do you know this? Narita bound flight? Flying close to the Butterworth military airbase? Not talking about a possible archaic radar alerting maneuver, which fits all the BTO and BFO data 18:25-19:41 and the last radar position 18:22?
Re: “By flying low near an airfield or something like that. Just flying circles above the sea near an airport long enough would have alerted people.”.
Perhaps that was intended, who knows. Not too close to avoid air collision, but within reach of a few runways.
Re: “Or even try to reach someone by mobil phone on low altitudes.”
FO’s cellphone?
@Dennis,
Re: “Yes, turning the SATCOM off/on makes little sense other than to arm wave the BFO values after 18:25.”
Add on top of it that the following also makes no sense:
– Crossing the Malay Peninsula just to negotiate instead of staying in the SCS;
– Hiding instead of inviting foreign media;
– Not answering SATCOM calls;
– Flying into the SIO only to crash.
Is your negotiation theory in the “whacko” category or not? Try to convince me that it is not.
@Tbill
http://de.flightaware.com/live/flight/UAE343/history/20140307/1710Z/WMKK/OMDB/tracklog
Please note anything after 02:07 is a predicted flight path.
Here are some predicted locations for EK343 March 8th 2014.
Please note that this is if EK343 travelled following directly along the N571 to 02:2212 at the same speed as seen at last known postion for EK343 at 02:07.
EK343 predicted 02:1336
5 53 3.00N
97 55 50.00E
EK343 predicted 02:15
6 00 45N
97 46 55 E
28.7km for EK343 from predicted 02:15 to VAMPI
82.7799km Distance from VAMPI to 02:2212
EK343 predicted 02:2212
6 23 29.26N
96 52 6.31E
Predicted path for EK343 is approx 33NM behind MH370’s predicted path as shown in recent ATSB report
@Oleksandr
Re: “If the standby bus power is completely lost due to a short-circuit of the left transfer bus and damaged inverter, then the transponder panel would not be functional, correct?”
Correct.
“Would it be possible to switch to the right transponder in this case?”
No.
“Also, does the transponder automatically transmit a message when the main AC bus is deliberately isolated from the cockpit?”
No!
Just received three separate emails which, when pieced together and making logical inferences, suggest the first person was Mick Rooney, who had something on or before Nov. 11, 2016.
1) is this the same Mick who contributes frequently to this forum?
2) if so, were the private emails to me accurate?
3) if so: who gave you the pdf(s), and why to you?
Many thanks in advance for your transparency.
Oleksandr:
@Dennis, Is your negotiation theory in the “whacko” category or not? Try to convince me that it is not.
@Oleksandr:
Dennis is not in the “whacko” category.
A “hostage taker” always has an “ultimatum” and a “deadline”, and a “threat” of them not being met.
The mistake that the Malaysian Government made, was to “reject the ultimatum”, and then, to “ignore the deadline”, on the assumption that they could stonewall, because to their reasoning, the aircraft would have to land somewhere, then they could move to a hostage negotiation on the ground scenario.
As has been postulated by me and others long ago, the “PLAN B” if the Government refused to meet the “deadline” (which was around 19:00 UTC) was to move directly to implimentation of the “threat”.
That is what happened, pure and simple, and that is why the Malaysian Government does not want the aircraft found.
Minister’s full statement for those who have not seen it all.
Hardly enthusiastic. ‘Acknowledge, not ‘welcome’. “No new information has been discovered….”
Quite properly IMO does not talk up the prospects.
http://minister.infrastructure.gov.au/chester/releases/2017/october/dc315_2017.aspx
Brock,
You ask, “Is the JIT’s legal obligation just as strong either way?”
From my reading of the Final Report, the JIT is not a legal entity and thus has no legal obligations. It is (or was) effectively a working group of experts reporting to the Malaysian government.
“Early in the surface search the Malaysian Government convened the Joint Investigation Team (JIT) comprising experts from the People’s Republic of China, France, Malaysia, United Kingdom, United States and Malaysian Government officials. Soon after, a satellite communications working group (SATCOM WG) was also formed and included experts from Inmarsat and Thales. These groups of experts worked together to provide advice to the Malaysian Government on the surface search areas.”
For all we know, the JIT may have been dissolved ages ago.
@Victor,
Re: “It’s possible that the re-powering of the SATCOM could have been timed with either passing MEKAR or being out of the 250-NM radar range.”
What is to do with 250-NM Butterworth radar range, what could be achieved? There were 3 more radars, which were supposed to track MH370:
– RTADS-III Phuket;
– Lhokseumawe TRS2215R, and
– Sabang TRS2215D.
Furthermore, I have little doubts that the second half of the LIDO image was sourced from one of these 3 systems, not from the Butterworth radar:
– The gap in the LIDO image.
– Dispersion characteristic of the blips (compare beginnings and ends of the 2 segments).
– Malaysians, who mentioned that they were attempting to acquire the radar data via diplomatic channels from neighboring counties, which was consistent with what Thai and Indonesian military claimed.
As a matter of fact, currently I lean to suspect nothing was provided by the Butterworth radar at all. I think the first segment (including a single blip) was provided by Kuala-Lumpur radars (Fig 1.1C in FI). Visually it appears that the gap is close to the range of KL radars. Also non-detection of MH370 when it was near Penang support this assumption.
@ventus45,
Re: “That is what happened, pure and simple”
Sorry, you have not convinced me in your story. To many facts against it.
@Oleksandr
Is your negotiation theory in the “whacko” category or not? Try to convince me that it is not.
I can’t point to any diversion theory that would be helpful in finding the aircraft so I regard it as pointless to evangelize a diversion theory.
However, I do regard it as odd that a wheel would fail after retraction (weight and rotational stresses removed) as opposed to failure on the takeoff roll (which corresponds to what is typically reported). It would be similar to having a blowout in your vehicle spare tire.
@Bobby,
Re: “I don’t think there is any connection between the last radar contact and the SDU log-on, other than the fact that they occurred within a few minutes of one another. The last radar contact position appears to be driven by the 250 NM maximum range of the radar at Butterworth.”
Another coincidence of extremely low probability? Don’t forger that we are dealing with the coincidental occurrence of the 3 events:
– Disappearance from the radar;
– SDU logon;
– Beginning of some maneuver (as suggested by the BTO and BFO 18:25-18:28);
Why not to consider something more realistic, for example: the second half of the LIDO blips was provided by Phuket or Lhokseumawe radar, and the disappearance was caused by descent?
@Andrew,
Assume both engines are operating normally when cruising in ECON. Then one engine runs out of fuel. What speed mode will the FMC revert to? LRC INOP or HOLDING INOP? I would guess LRC INOP in this case, and HOLDING INOP if the previous mode was HOLDING. Is this correct? What happens if the speed was previously set manually using the MCP (not ECON)?
Darren Chester confirms Malaysia’s (yet imminent) announcement.
@Brock McEwen
I am not nor do I have any connections with Mick Rooney.
@Dennis,
The missing bolt was at the left wheel, just a couple of meters away from the electronics bay. Draw a line, and you will see what a single fragment of the wheel can damage.
Re: “However, I do regard it as odd that a wheel would fail after retraction (weight and rotational stresses removed) as opposed to failure on the takeoff roll (which corresponds to what is typically reported). It would be similar to having a blowout in your vehicle spare tire.”
I am looking forward to reading details of this:
http://www.dailymail.co.uk/news/article-4993358/Pilot-lands-passenger-plane-tyre-blew-out.html?mwv_rm=als1
Note that water would become ice as temperature drops. Its expansion and loss of elasticity could trigger the failure of the rim. Rather trivial explanation.
@Oleksandr
Note that water would become ice as temperature drops. Its expansion and loss of elasticity could trigger the failure of the rim. Rather trivial explanation.
The time from takeoff to the turn at IGARI was only 40 minutes. Even if the entire 40 minutes were at below freezing temperature warm water and a warm wheel (from takeoff friction) would not freeze in 40 minutes. It takes almost 4 hours for my freezer to generate ice cubes in the ice cube tray, and water is cold when I put the tray in the freezer.
@Dennis,
Have you ever observed frost on a window of an aircraft during ascent? Ambient air temperature at FL350 is around -50C, density being approximately 1/3 of the sea level density: a lot more freezing capability compared to your fridge. Also your tray might be made of plastic, which has low heat conductivity.
@Oleksandr
Trying model the time for water to freeze in a wheel well is next to impossible so I will only say that I doubt freezing of warm water in warm wheel will occur in what amounts to a closed compartment (big factor in the freeze time) in the amount of time the aircraft was actually at a below freezing altitude.
@DennisW. Nose wheel bolts. “I do regard it as odd that a wheel would fail after retraction (weight and rotational stresses removed) as opposed to failure on the takeoff roll…”
The tension in the tie bolts will alter with decreased temperature, depending on the difference in contraction of themselves and the wheel halves they bind together. Tension in steel (of various constituents) or titanium bolts in an aluminium alloy or magnesium alloy wheel will reduce, making failure at altitude less likely than when the wheels are hot as you say though I think for a different reason.
@Joseph
Thank you! I get FL340 for UAE343.
Wow the estimated UAE343 flight path, if true, looks like MEKAR to HITAS which is offset from N571 just like we thought MH370 was. That puts it in the Andamans and VOCX control space and towards DOTEN and near N10. Hmm…
@David
Tension in steel (of various constituents) or titanium bolts in an aluminium alloy or magnesium alloy wheel will reduce, making failure at altitude less likely than when the wheels are hot as you say though I think for a different reason.
Good point. It is common practice to take a blowtorch to a wheel hub to remove press fit bearings. The steel bearings expand less than the aluminum hub.
I was only considering the freezing of water.
@David
Anecdotal related story. The machining of the aluminum hub relative to the diameter of the press fit steel bearing is referred to as the “crush”. The early F800GS motorcycles which I had one of miscalculated the “crush”, and bearing failures were common. I had one at speed and ended up in the hospital for awhile. When replacing wheel bearings I put the bearings in the freezer and the wheel in the hot sun. The bearings slip right in, but are almost impossible to remove when the bearing and hub temperatures equalize.
Hey, excuse me, but old people like to reminisce.
@DennisW. If you and your bike had been at FL350 it would have been doubly difficult.
Moral: if at your day of judgement you are found sinful a consolation is that you will find it easier down there to replace those bearings.
@David
Yes, it is my destiny. I am sure of that. I spent a number of years racing motorcycles both on the track and in the dirt. Just about everything has been broken once. It was my only relief from talking smart with colleagues under the flourescents (which was often more painful). My kids were into it as well resulting in a divorce on the grounds that I was maiming my wife’s babies.
Back on topic. I am very concerned about the latest official ideas of where to search. I feel like I am watching a movie I have seen before. I am also concerned about OI’s pedigree. I can find nothing in the “what have they done” category. I would feel a lot more comfortable with a time and materials funding effort using Fugro. Oh well, it will be interesting and fun to watch (assuming OI gets started).
Update on the search:
https://www.theguardian.com/world/2017/oct/20/mh370-malaysia-in-deal-with-private-search-firm-to-find-missing-plane
@Mick Gilbert: thanks for confirming what was self-evident from your handle. Silly question, I admit.
@Mick Rooney: if you read this blog: questions 2 and 3 (above) are thus for you – the floor is yours.
With the ATSB now proudly broadcasting details of this same (?) report, the need for secrecy would seem to be waning.
Thanks in advance for your transparency.
@all,
I just finished modifying my Endurance Model to accommodate LRC INOP and Holding INOP (flaps up) speeds and fuel flows. Using either of these modes, beginning before 19:41, results in significant excess endurance at all available altitudes. I am making the assumption that all the fuel on board is made available to the single operating engine, and this requires a pilot to switch fuel valves and pumps during the flight so the fuel in both wing tanks can be fed to one engine.
Holding INOP is the most fuel-efficient speed schedule, and MEFE is predicted from 01:38 to 2:24 UTC on 8 March 2014 depending on flight level (FL100-240) and the exact time the Holding INOP speed begins (between 18:29 and ~19:30). It is interesting that Holding INOP has greater endurance than Holding, so you can fly longer by shutting down one engine (which then operates with improved efficiency at a higher thrust).
Long Range Cruise INOP uses more fuel, of course, than Holding INOP, but MEFE is still predicted to occur between 00:36 and 01:02 UTC, using the same ranges in start time and altitude given above.
My conclusion is that it is rather unlikely that either of these speed modes were used from a time prior to 19:41 until MEFE at 00:17.
These INOP speed modes will be incorporated in the next release of my Fuel Flow and Endurance Models, which will happen after I update my CI = 52 speed schedule based on air speed data in the MH370 Flight Plan and the MH371 ACARS data set.
@sk999: ok, then let me amend my question, by replacing “JIT” with “whoever had the legal obligation to properly retrieve, analyze & report on any suspected seabed debris Fugro, et al may have spotted while under contract to search for MH370”.
As was likely clear to you from context, I am hoping to focus on what, if anything, CHANGES under an OI arrangement.
(For the record: I agree with you re: possibility of JIT having disbanded years ago. Throw the composition, activities and longevity of the JIT on the pile of things we seem to be meant never to know. But this is old news, and beside the point.)
@DrB
Re: “Assume both engines are operating normally when cruising in ECON. Then one engine runs out of fuel. What speed mode will the FMC revert to? LRC INOP or HOLDING INOP?”
In the cruise, the FMC won’t automatically change to EO mode following an engine failure. The FMC will remain in ECON mode (all engines) until the crew activates EO mode by selecting ENG OUT on the VNAV page and pressing the EXEC key. The commanded speed will then change to the EO speed, which is essentially the minimum drag speed.
“What happens if the speed was previously set manually using the MCP (not ECON)?”
The commanded speed would be that selected on the MCP, regardless of any FMC selections.
Ouch…
@Victor
Ouch, indeed. Although it does throw up a new trivia question, “What do John Wayne and Mike Chillit have in common?”
@Mick G: That’s an easy question for Americans.
@DennisW said: I would feel a lot more comfortable with a time and materials funding effort using Fugro.
And who would provide the funding for that effort?
@DrB
Thank you for looking more closely into the one engine INOP endurance.
It suprises me the endurance is even longer than with both engines operating.
I would have expected more drag by the rudder compensating, with higher fuel consumption overall.
Now I understand endurance does not tell much about specific range.
Can you tell more about the maximum range with one engine INOP at let’s say ~20K/ft?
I thought it was well known and understood that operating on one engine extended endurance and range. Is this not what long-range air-sea search operations sometimes do deliberately?
@DennisW
I tend to trust they have looked closely into the credentials and possibilities of this company with the goal of being capable to find the plane.
If you want to think bad you can also think Malaysia knows they won’t find the plane and with them trying no money would be lost.
@Paul Smithson
What I understand about it, is the plane has to fly on far lower altitudes with one engine and at considerably lower speeds.
Endurance can be longer but range would be shorter in my view now.
I’m just curious about the max.range; the same, longer or shorter.
That’s my question to @DrB.
@Fuel Modelers
Years ago (probably a 747) we flew Qantas from Melbourne to Sydney. The pilot quickly went to high altitude (I recall at least FL400) and then “glided” down to Sydney. This was explained to me the most fuel efficient path.
Is this still true?
@TBill: The most efficient descent will be one at idle thrust and the most efficient climb will be at or near full thrust. The fuel planning algorithm will find the position of the top of descent (T/D) for the glide, which will depend on the CI and the winds. A low CI translates to a lower airspeed, resulting in a steep climb and a longer descent, and the opposite for a high CI. Depending on the location of the top of climb (T/C) and the T/D, it’s possible that for a short flight, the aircraft spends little or no time at cruise.
@Victor
And who would provide the funding for that effort?
Anticipating the preferred result, the finding of the aircraft, the funding source really becomes mute.
My guess is that Chester’s less than enthusiastic comments are related to the fact that Australia is not fond of Malaysia’s choice. It is an unusual way for a sovereign state to do business.
As Chester said, “No new information has been discovered to determine the specific location of the aircraft…” The tripartite group’s previously stated position was that the search would not be resumed in this circumstance.
@Brock,
You wrote:
a) “…scans of documents which some here claim were authored by the RMP” alluding to a report attributed to the RPM/PDRM.
b) “With the ATSB now proudly broadcasting details of this same (?) report, the need for secrecy would seem to be waning.”
The ATSB wrote:
“MH370 PIC Microsoft flight simulator data analysis provided to AMSA/ATSB by Australian Federal Police (19 April 2014).“
@Andrew,
You said: “In the cruise, the FMC won’t automatically change to EO mode following an engine failure. The FMC will remain in ECON mode (all engines) until the crew activates EO mode by selecting ENG OUT on the VNAV page and pressing the EXEC key. The commanded speed will then change to the EO speed, which is essentially the minimum drag speed.”
Thanks, Andrew. I have several follow-up questions:
1. What happens in that case with no pilot actions? If the engine out flight level is higher than ~FL240, depending on the current weight, presumably the aircraft will slow down to a minimum speed and at some point begin to descend to maintain that speed, perhaps leveling out when the remaining engine is capable of maintaining level flight.
2. Is the default EO speed when cruising the same as the LRC INOP speed table from Boeing?
3. Is the default EO speed when Holding the Holding INOP (flaps up or down) speed?
4. What happens if the speed setting prior to loss of one engine is set by the MCP as a fixed Mach or KCAS, again with no pilot action after EO?
DrB said @Andrew: “What happens in that case with no pilot actions?”
Back on March 17, 2017 at 9:24 pm Andrew reported on a simulator exercise that answers most of your questions.
@Ge Rijn,
I ran some comparisons of speed modes after 18:29 at FL240. The longest endurance is for Holding INOP, followed in order by Holding, LRC INOP, MRC, and LRC. MRC and LRC INOP are close together in range, and they have the longest range. LRC and Holding INOP have about 1-2 % shorter range. Holding has the range reduced by ~9% compared to MRC.
@Brock McEwen
@sk999
Re: “I’ve asked twice before in this forum whether search leadership’s obligation to recover, examine, conclude and report on any “found” debris is any different under this new proposal than it would have been had the official search not been suspended.”
The short answer is ‘No’, nothing changes.
The long answer:
The ‘legal entity’ that is responsible for the MH370 investigation is ‘The Malaysian ICAO Annex 13 Safety Investigation Team for MH370‘, commonly known as the ‘Joint Investigation Team‘, or JIT. According to various Interim Statements released by Malaysia, the team was established under Regulation 126(1) of the Malaysian Civil Aviation Regulations 1996, in compliance with Annex 13 (Aircraft Accident and Incident Investigation) to the Convention on International Civil Aviation. The JIT, headed by a Chief Inspector/Investigator-in-Charge, comprises 19 Malaysians and an accredited representative from each of the following safety investigation authorities:
– Australian Transport Safety Bureau (ATSB),
– Civil Aviation Administration of the People’s Republic of China (CAAC),
– Bureau d’Enquetes et d’Analyses pour la securite de l’aviation civile (BEA) of France,
– National Transportation Safety Committee (NTSC) of Indonesia,
– Transport Safety Investigation Bureau (TSIB) of Singapore,
– Air Accidents Investigation Branch (AAIB) of the United Kingdom, and
– National Transportation Safety Board (NTSB) of the USA.
The JIT has not been disbanded and Annex 13 requires that it continue to investigate the disappearance of MH370 until a Final Report is completed and the investigation is closed.
The following paragraphs of Annex 13 are relevant:
3.3 The State of Occurrence shall take all reasonable measures to protect the evidence and to maintain safe custody of the aircraft and its contents for such a period as may be necessary for the purposes of an investigation…
5.3 When the location of the accident…cannot definitely be established as being in the territory of any State, the State of Registry shall institute and conduct any necessary investigation…However, it may delegate the whole or any part of the investigation to another State by mutual arrangement and consent.
5.4 …Where feasible, the scene of the accident shall be visited, the wreckage examined…
5.6 The investigator-in-charge shall have unhampered access to the wreckage and all relevant material, including flight recorders and ATS records, and shall have unrestricted control over it to ensure that a detailed examination can be made without delay by authorized personnel participating in the investigation.
5.7 Effective use shall be made of flight recorders in the investigation of an accident or an incident. The State conducting the investigation shall arrange for the read-out of the flight recorders without delay.
5.13 If, after the investigation has been closed, new and significant evidence becomes available, the State which conducted the investigation shall reopen it…
Note. – Where an aircraft which was considered missing following an official search is subsequently located, consideration may be given to reopening the investigation.
Annex 13 to the Convention on International Civil Aviation – Aircraft Accident and Incident Investigation
Andrew,
According to the Final Report, the JIT was created “Early in the surface search”, and was certainly providing advice by Mar 18, 2014, up through April 7. The JIT paper in the Final Report is dated April 26. The Annex 13 Safety Investigation Team was not created until April 25 (many references). So it may be that the JIT was superseded by the Annex 13 team, and, indeed, the Annex 13 team seems to comprise a superset of the JIT (Singapore and Indonesia being added). I do not know how Malaysian law or government works, but my guess would be that the “legal entity” (i.e. against whom you would file a lawsuit) is the MOT.
@DrB
Thank you for the finding that one engine inop is on the cards / table again. Can you publish the updated analysis / the spreadsheet please ?
@Andrew, sk999. JIT becomes SIT.
There has been confusion with titles.
On 29th April 2014, there was reference to the, “International Investigation Team”, presumably also known as the JIT.
http://www.mh370.gov.my/index.php/en/34-mh370-press-briefing-by-hishammuddin-hussein-minister-of-defence-and-acting-minister-of-transport-29-april-2014
However that included Singapore and Indonesia.
No, I see from re-reading sk999 this would have been the SIT.
@DrB
Thanks for those range comparisons at FL240.
Then one more question if you don’t mind..;
You say MRC and LRC INOP are close together in range, and they have the longest range (at FL240).
Then what would the difference be in range between MRC(or LRC) at FL350 and LRC INOP at FL240?
David,
At the end of that article, HH states, “This investigation will be carried out in accordance with Annex 13 …” so it is the Annex 13 team that he is announcing (and presumably its first action was to figure out what to name itself.)
Having said that, I think Andrew’s reponse to Brock’s somewhat inartfully expressed original question is correct. If you read through the Final Report, the ATSB’s dealings with debris (in this case surface) were always coordinated with the Annex 13 team.
@sk999
You are correct, the JIT and the Malaysian ICAO Annex 13 Safety Investigation Team for MH370 (SIT) are two separate entities; please excuse the error in my previous post. The JIT was set up by Malaysia early in the search, and was tasked with defining the search area from the satellite data. It was comprised of specialists from Malaysia, China, US, UK and France. The SIT was established at a later date and is the body that is responsible for the accident investigation under Annex 13.
By ‘legal entity’ I simply meant the body that is responsible for the Annex 13 investigation, ie SIT, not the entity against whom a law suit would be filed. For what it’s worth, I believe the respondent in such a case would be the ‘Government of Malaysia’ rather than an individual department.
Brock asked ‘what, if anything, CHANGES under an OI arrangement’? The answer is ‘nothing changes’. According to Annex 13, the SIT is obliged to examine any recovered wreckage, where feasible. It doesn’t matter if the search is conducted in 1, 2 or 10 parts, or by whom the search is conducted.
Andrew,
Very good – I think we are all in agreement now.
As an aside, I do become very focused when terms like “legal entity” or “legally obligated” get raised. Over the years I have been part of reviewing or writing or getting signatures on a slew of documents with terms like “Agreement”, “MOU”, “Proposal”, “Statement of Work”, etc, some of which mean that I am responsible for providing some deliverable (and that’s another term), so I am very sensitive when certain words or phrases are used. They have real meaning. I am no legal expert, but I do recognize that such-and-so entity is listed on the signature form is because that entity can enter into contracts, while this other entity – no matter how prominent it may seem – is not authorized to do so. Law is an infinitely tangled topic. Give me BFOs and BTOs any day.
@Andrew, @David, @sk999: thank you.
@All who’ve done rigorous BTO/BFO fitting: the ATSB is now claiming the JIT told them that, if 19:41 start points are varied across a line from the infamous “NW point” to the tip of Indonesia, with southward legs then modeled across a variety of speeds, the path which best fit ONLY (…) the last three BFO records started from exactly the NW point, and progressed in a trigonometrically perfect curve through the BTO arcs at 400 KGS, to arrive at 23s.
I am paraphrasing, because I’m going by memory, but I’m pretty sure I’m characterizing the NEW new new new story reasonably accurately. I will not bother commenting on the wisdom of this approach to search area specification, which is self-evident.
My question is simply whether anyone has replicated this conclusion yet, using their own models?
Thanks in advance.
@DrB
Re: “1. What happens in that case with no pilot actions? If the engine out flight level is higher than ~FL240, depending on the current weight, presumably the aircraft will slow down to a minimum speed and at some point begin to descend to maintain that speed, perhaps leveling out when the remaining engine is capable of maintaining level flight.”
Yes, that’s correct. The autothrottle will increase thrust to the CRZ or CLB thrust limit (according to airline setting) in a bid to maintain the ECON speed. If the aircraft is above the one engine inoperative maximum altitude, the thrust will not be sufficient to maintain speed. The speed will slowly decay until just above stick shaker activation, the envelope protection will then lower the nose and the aircraft will descend. It should stabilise at some lower level once the operating engine is capable of producing sufficient thrust to maintain level-ish flight.
Re: “2. Is the default EO speed when cruising the same as the LRC INOP speed table from Boeing?”
No, the EO speed will be the minimum drag speed. It should be close to the engine INOP (flaps up) holding speed.
Re: “3. Is the default EO speed when Holding the Holding INOP (flaps up or down) speed?”
Yes.
Re: “4. What happens if the speed setting prior to loss of one engine is set by the MCP as a fixed Mach or KCAS, again with no pilot action after EO?”
The autothrottle will increase thrust in an attempt to maintain the MCP speed.
Brock,
Your “NEW new new new story” was reported by the ATSB over three years ago in the original Underwater Search report. The only thing new is more detail. I posited back on Oct 9 (and DennisW concurred) that the BFO calculations were in error. OLD old old old news.
“… trigonometrically perfect curve …” Never heard that phrase before, and I have no clue what you mean, but it sounds cool.
Malaysian statement refutes the media interpretation of the Australian Minister’s. Click on ‘Statement’.
http://www.bbc.com/news/world-asia-41691794
Also,
https://www.msn.com/en-us/news/world/new-twist-in-search-for-missing-mh-370/ar-AAtM6Qy
@victor
I think we may be talking about different South Poles when we talk about favourites. I like the Magnetic South Pole 😉
@David
To me, the most interesting element of the recent announcements is Chester’s statement in his Malay/OI press release:
“No new information has been discovered to determine the specific location of the aircraft…”
It does not relate in any way to the Ocean Infinity and Malay relationship. It was chosen for some other reason.
It essentially tosses CSIRO under the bus relative to their claims of a “new” very likely small area based on the French satellite imagery.
It highlights the fact that Malaysia is ignoring the tripartite agreement to suspend the search until new data refining the terminus is found.
In particular, ignoring the CSIRO “news” amusing to me. I sensed that CSIRO was not playing nice with the Aussie government with their rather fanciful claims which serve to reinforce the position of the people opposed to suspending the search.
My guess is that the CSIRO PR head will be looking for a job soon.
Seabed Constructor, the ship named by Swire Seabed as being under long-term contract to Ocean Infinity, is working in the North Atlantic, as noted by Victor after his discussion with Oliver Plunkett some weeks ago. It was in the Azores a few weeks ago and early this morning it was nearly stationary at 37.4N, 56.3W. I have not been following the ship via AIS, but the Fugro boats moved at slow speed when they were deploying their own Hugin AUV unit, so Constructor might be doing the same. Online data suggests the seafloor depth there is around 5000m, so this area would allow tests consistent with the MH370 search area.
Clearly it would take some time to move the ship to Australia if a search contract was agreed.
@Brock, sk999
I’m not sure the calculations leading to 23S were very much in error. If I start from 5-6N at 19:41 I typically find end locations in the lower 20s based on minimal BTO/BFO errors. Note that these are strongly curved paths. If I understand correctly, the additional criterion leading to a best fit at 400 kts was a constant GS. I haven’t tried to replicate this.
@Richard Cole: OI recently told me they would be testing to 5500 m, which I’m told would be a depth record for seabed scanning. OI is doing amazing work. Real experts that I’ve talked with (not Twitter personalities and fanciful bloggers) acknowledge this.
@DennisW. Re the Minister’s statement, “It highlights the fact that Malaysia is ignoring the tripartite agreement to suspend the search until new data refining the terminus is found”. It sure has that flavour ie the Minister or someone advising is unconvinced.
Or else Australia was keen to go before and the others vetoed and now there is pique.
Hard to say.
@Neils
The BFO values sk999 referred to were in error by 12Hz. No doubt about that. My guess remains that the early modeling had the math correct, but the Doppler corrections associated with the earth station were still evolving. By the time we had access to the corrections they had stabilized.
BTW, my own calculations from higher 19:41 latitudes (5N-8N) end closer to a 26S-27S terminus at 00:11. Typical ground speeds in the 400-450 kt range. Similar to the Iannello and Godfrey McMurdo path.
@David, @DennisW: Australia is trying to balance the optimism that is required to re-start the search with the realism of the challenges of success and the requirement to properly set expectations (which they have failed at doing in the past). Australia seems reluctant to spend any more money on the search. They’ve already spent more than their fair share, especially in consideration of what China has contributed. Perhaps they don’t have absolute confidence in their new proposed area. I’m not sure that matters at this point.
I don’t understand the reluctance to recognize that the OI offer represents an extraordinary opportunity. Nobody else in the world with their capabilities has agreed to take on the financial risk of re-starting the search.
The attacks on OI are both illuminating and amateurish. Such as trying to link them to a downstream construction company in Houston with a similar name. (Do people know what a downstream construction company does?) Or claims that their technology is a scam. The lack of common sense and the ignorance of this industry is evident.
It seems to me Australia by the words of Darren Chester wanted to scale down expectations that were lifted by CSIRO/ATSB on the 35S ‘pin-point’ area.
It could be a starting point for OI but nothing more. They already wash their hands if nothing gets found there it seems to me.
Griffin also already scaled down expectations about this (his) area previously so bluntly stated as a ‘pin-point’ area.
But on the other hand they probably releive stress on OI (and others) to start at a specific area (35S) and give them ‘cart blanche’ to start wherever it suites them best.
It could be this was the reason Darren Chester made this announcement this way before things were settled completely.
If so, it’s a good intervention imo.
35S is oke to start but should not be a burden to start in any way imo.
@Victor
(Do people know what a downstream construction company does?)
Yes, I made a 44% cap gain on Valero (VLO) over the last year, plus they pay over a 5% dividend relative to my basis.
@DrB
Maybe I should explain my questions about one engine INOP and making an estimate on range between both engines operating at FL350 and only one engine at FL240 after ~19:29 (LRC or MRC).
In my logic this could define a max. range with one engine INOP and with that a max. reacheable latitude on one engine at FL240.
Related to the drift studies, knowing this could exclude or include the possibility of a one engine flight after ~19:29 (in my logic).
@DennisW: Then perhaps you can explain to the less informed that downstream construction companies don’t engage in seabed surveying, scanning, and exploration.
@Victor
I don’t understand the reluctance to recognize that the OI offer represents an extraordinary opportunity.
How can anyone be expected to recognize the “extraordinary opportunity” that this offer represents when the details of the offer have not been disclosed? I have no idea what the offer is.
@VI: “The attacks on OI are both illuminating and amateurish”
But at least they’re quick and imaginative. Already figured that OI is a Russian asset inside a mail drop in the Bahamas who are feinting a test run in the Azores whilst they infiltrate Iceland and glue barnacles onto pieces of MH17 in Kazakstan. And Blaine Gibson is lizard people.
@DennisW: We know that OI made an offer to assume all financial risk for a re-start of the search in exchange for a reward if found. We know that no other company has made a similar offer. We know that the tripartite countries have stopped the search. We know that OI has the most advanced scanning technology using best-in-class sensors and a team of underwater drones. We know they are testing their technology today at amazing depths. We know they were selected after Malaysia, Australia, and China conducted due diligence.
What we don’t know are the details of the offer, or who is providing the financial backing of OI. Because of those unknowns, the deluded and those not interested in re-starting the search have attacked the credibility of their technology and the credibility of their organization, even trying to tie the company to Russia! One even has tried to encourage others to actually bet AGAINST OI’s success. To me that’s like betting against your family member’s fight with cancer.
From my vantage point, I hope Malaysia and OI are able to resolve remaining issues, OI is able to start the search in January, and the wreckage of the plane is eventually found, either in the area recommended by CSIRO or somewhere else. (OI has never said it would only search where CSIRO recommended.)
lkr: Very funny. Unfortunately, your comedy hits too close to the truth.
Of course, by trying to bring people back to reality, I will soon be accused of being part of the master plan of deception. Ego, delusions, and obsessions are not easily combatted with reason.
@DennisW
Would you have a time, position, GS and track for the 12 HZ BFO error calculation? I would like to find out what causes the difference in our results.
@Victor,
I think your post captures the essence. Probably more worthwhile for to me explain that federal and state taxes on cap gains in Cali come to some 33.3% (highest in the nation, BTW). However, the basis of your equity investments reverts to fair market value when you die. So your estate is some 50% more valuable if you don’t sell appreciated equities (100% versus 66.6%).
I thought about funding a chair of computer science at Cal Poly SLO where my daughter teaches. Cal Poly would not have to pay a cap gains tax on such a gift, but daughter said she would rather have the cash. The apple did not fall far from the tree.
@Neils
from k999’s post on October 9:
On pdf p. 190 of the Final Report, there is an account of how, on Mar 31, 2014, Inmarsat calculated the BFO for a point close to what became the Ocean Shield position. For a speed of 400 knots, the position is -23.4, 102.8. No heading is given, but my attempted reconstruction of their route gives a heading ~157 deg. Their calculated BFO has a “mismatch” of 1 hz with the measured value (252), so it is either 251 or 253 hz. Figure 9 also shows that. However, my calculated value is 240 hz. That’s a huge difference – 12 hz away from the measured value
I found the exact same values as he did. In fact, I remarked earlier that the entire appendix was historical not factual, and did not even include the information from the Inmarsat JoN paper.
@Victor
From my vantage point, I hope Malaysia and OI are able to resolve remaining issues, OI is able to start the search in January, and the wreckage of the plane is eventually found, either in the area recommended by CSIRO or somewhere else. (OI has never said it would only search where CSIRO recommended.)
Totally agree. I do wonder, however, if the OI offer will ever be made public.
Chester seems to think that OI will start where CSIRO recommended. My weak preference is further North.
@DennisW
Ok, thanks. Yes I agree with the 12 Hz error for those parameters. I found/considered a more strongly curving path ending S23, with TT 83 degrees and GS 385 kts at 00:11.
@Neils: thanks much for the reply. Would you be so gracious as to please flip me a copy of your model? I think you have my email address. Huge thanks in advance.
@Mick Rooney: you responded to my questions in another forum; I tried to thank you, but appear to have been barred from so doing by its host. So, with apologies for switching forums: profuse thanks for the reply.
And bravo for respecting the privacy of both your sources and next of kin.
When and if it is convenient, could you be so kind as to flip me the emails/texts/etc you exchanged with the folks you spoke of who helped you corroborate the authenticity of the PDF files you received? My email address appears on all my MH370 search decision audits, and I believe you already follow me on Twitter. Absolutely no rush – just trying to ramp up on my ability to authenticate a PDF file.
Also: Jeff Wise’s same-day reply to your Nov. 11, 2016 tweet claimed many people had been “sitting on” these PDFs “for months”. If his claim is accurate, and if you hadn’t received your copy until Nov. 9, 2016, then you could not possibly have been the first member of the general public to receive anything. Do you agree it would be helpful if those who had copies much earlier would disclose their receipt dates, and as much as they can of their sources?
Related: in the SCMP last week, Florence de Changy seemed to dismiss the upstream SIM analysis cited in this PDF (and cited in whatever conduit the ATSB received, if not the same as yours):
“The report mentions that six weeks before the flight, the pilot, Captain Zaharie Ahmad Shah, had flown a route on his home flight simulator “initially similar” to the one supposedly taken by MH370. The document the report bases this on has been dismissed as a clumsy fabrication based on several simulator routes flown by Shah, not one.”
I have an email in to Florence seeking clarification (dismissed by who, and why?), but welcome your thoughts. Thanks in advance for your time and consideration.
Full article:
http://m.scmp.com/week-asia/politics/article/2115300/gaping-holes-malaysia-airlines-flight-370-search-report
@Brock McEwen said: “The document the report bases this on has been dismissed as a clumsy fabrication based on several simulator routes flown by Shah, not one”…I have an email in to Florence seeking clarification (dismissed by who, and why?), but welcome your thoughts.
I am anxious to see what the basis is for Florence’s dismissing the simulator data. If the source is what I suspect, it is based on a very flimsy and provably incorrect analysis. But we can wait for her response.
@Victor. “I don’t understand the reluctance to recognize that the OI offer represents an extraordinary opportunity. Nobody else in the world with their capabilities has agreed to take on the financial risk of re-starting the search.”
You imply about the lack of enthusiasm of the Australian Minister, that, “I’m not sure that matters at this point”, though if it is from reluctance to contribute to the reward or lack of confidence in the new search area his attitude might matter still.
The Malaysians referred recently to keeping Australia and China informed, observing “the spirit” of the tripartite agreement, implying to me they may be prepared to go it alone, which would sideline the Australian attitude, at least for the search phase. (By the way, someone would, as before, have to bear wreckage recovery costs and provide other support for that, presumably from Perth. What was agreed about that and whether that still has force is unclear).
Then again that attitude might derive from a reluctance to get too publicly enthusiastic about prospects, with the NOK in mind. All murky. See further below.
But I suspect also you are alluding to discussions on the JW site also in your last paragraph, with its sceptics and cynics, though bear in mind that you are much closer to the OI bid than many unversed in the foundations behind it.
You might be frowning too on discussion between DennisW and me about what might be prompting continuing delay in the Malaysians taking up the OI offer. I think that is interesting and for one can think of legitimate contractual issues which could lead to this; in particular how far OI should be expected to continue if the wreckage is not found in the ATSB/SSWG/CSIRO search area, or if there are long technical or weather delays taking the search into the following season, or there is a change of heart for any reason by OI management or its owners. It would be nice to know they are not depending of the ATSB etc assessment of the likelihood of success too (Is David Mears involved?). Also, what is the outcome if they find part wreckage but from an in-flight break up? Just examples.
Normal prudency would entail the Malaysians approaching this with the above sorts of issues in mind, though more with care than distrust if they have already selected the OI proposal in principle. Care would ensure that enthusiasm does lead to downstream disappointment or misleading of NOK.
Returning to the murk above, that too could be what is leading to Australian caution.
@Victor @DennisW
On the reference to downstream construction, obviously the oil industry calls refining and petrochemicals industry the “downstream” side of the oil business, and the “upstream” side of the business is the exploration and producing of the crude oil from the ground, including off-shore. But I am confused who made the reference and exactly what was said.
@David
@Victor. “I don’t understand the reluctance to recognize that the OI offer represents an extraordinary opportunity. Nobody else in the world with their capabilities has agreed to take on the financial risk of re-starting the search.”
As i replied to Victor, I support the notion of a renewed search by OI. However, commenting on their offer would be ridiculous not knowing anything about what their offer actually was. People are caught up in the notion of free. It is not free when a negotiated fee is in the mix. Without knowledge of the offer details it is just plain dumb to categorize it as an “extraordinary opportunity”.
@Victor. Second last line. “…..enthusiasm does NOT lead to…” please.
DennisW: It is an “extraordinary opportunity” in that it transfers the financial risk from the 3 governments to the contractor, and thus gets the stalled search underway again. There is a big difference between risking taxpayer money (and political capital) and spending taxpayer money to learn a confirmed, certain 370 crash site. Who else was willing to do that?
@DennisW. Inadvertently left you off the distribution. Yes maybe the extent of the reward and its circumstances are the nub of the problem but the delay is now considerable.
@Victor. You said also of OI, “We know they were selected after Malaysia, Australia, and China conducted due diligence”. I had not appreciated that all three were involved with that but suppose that “due diligence” would not extend beyond the firm being sound.
@ALSM
DennisW: It is an “extraordinary opportunity” in that it transfers the financial risk from the 3 governments to the contractor, and thus gets the stalled search underway again
That statement is just plain stupid without knowing the financial obligations associated with a successful search. I had the impression that you had some business experience.
DennisW: I’m not going to get into a pissing contest with you about who has more years as a CEO, founded more companies or negotiated more multimillion dollar deals.
The only stupidity here is your inability to acknowledge the obvious fact that no one else has made a similar offer, which certainly makes it unique, if not extraordinary. That fact has nothing whatsoever to do with the financial obligations associated with a successful search. Zero.
@David: I was not speaking of the reluctance of the Australians. I believe they are supportive of this opportunity, and are simply trying to properly set the level of expectation. On the other hand, I am surprised at the hostility and false statements leveled against OI by outsiders they have only thinly veiled their opposition to the search re-starting.
As for the negotiating delays, I have no idea what details are currently under discussion. However, I’d be surprised if the success criteria are not a part of this discussion. Success if the main debris field is found? Fuselage? Engines? I can only say that I hope Malaysia and OI are both negotiating in good faith, and are able to resolve their differences in short order so that a vessel can searching in January.
@ALSM
“Similar offer”?? We don’t know what the offer was/is. Are you claiming to be a clairvoyant? If so I acknowledge your superiority to my CEO tenure.
I am about to put the whole IG in the whack category.
@ALSM
The only stupidity here is your inability to acknowledge the obvious fact that no one else has made a similar offer, which certainly makes it unique, if not extraordinary.
Do you have any idea how stupid this statement makes you look? Especially when we have have idea what the offer is.
I do now put the whole IG in the whacko category.
@Victor: I haven’t heard back yet. To save time, why don’t we assume the “who & why” of he dismissal was as you suspect: can you please tell me who is the who, and what is the why? Your adjective-heavy comment didn’t help me understand your point. Thanks.
Victor. “I can only say that I hope Malaysia and OI are both negotiating in good faith, and are able to resolve their differences in short order so that a vessel can searching in January”. Yes.
Just to illustrate this is not a routine good faith contract, another example of scope is that the ATSB’s Operational Search report p136, about wreckage recover planning, stipulates, “Part of the requirements for the underwater search were for the contractor to positively identify the debris field, map the debris field and produce a photographic mosaic of the entire field. Using this photomosaic, and prior to the commencement of a recovery operation, areas of particular interest could be identified which would form the basis for the initial priorities of the recovery operation.”
I imagine that task would be included within the reward umbrella.
(That Operational Search report section makes clear that while the Seabed constructor could to lift from 6000m there are other requirements, including extensive acccommodation. The availability of suitable vessels has been investigated.)
Separately you have said, “OI has never said it would only search where CSIRO recommended.”
The Australian Minister did say, “Ocean Infinity will focus on searching the seafloor in an area that has previously been identified by experts as the next most likely location to find MH370.” I guess the difference is that the CSIRO pinpointed more and that has not been endorsed by the SSWG ‘experts’. Can you disclose what OI actual search intentions are?
@Brock: No, I’ll wait to see what she says, if anything.
@David: OI can speak for itself regarding its search intentions. However, they would be foolish to limit themselves to where CSIRO recommends, and they haven’t said they would.
@Victor. If they find the costs and difficulties are greater than they expected, their efforts de facto could be curtailed. I now hazard a guess that the maximum extent of the search which can be expected of them is under negotiation.
This proposed Ocean Infinity search area combines the 4 CSIRO Hot Spots with the 3 IG LNAV Hot Spots based on the ultimate waypoints of NZSP, YWKS and NZPG.
Please see, linked below, the table of possible LNAV end points calculated independently by Victor and myself:
https://www.dropbox.com/s/quhfygiw29edfs2/LNAV%20Tracks.png?dl=0
@ALSM has also proposed a search width of +/- 18 NM centred on the 7th Arc at 18,000 feet.
The resulting proposed Ocean Infinity search areas are as illustrated in the link below:
https://www.dropbox.com/s/y2m6fcx879a4r33/Ocean%20Infinity%20Search%20Area.pdf?dl=0
In the illustration linked above, the light blue line is the 6th Arc, the blue line is the 7th Arc at 0 feet and the magenta line is the 7th Arc at 35,000 feet.
The four possible end points from the 3rd CSIRO Report are marked C-3-1 to C-3-4 and cover an area of 20,948 km2 of which 6,074 km2 remains to be searched.
The three possible end points from the IG LNAV waypoint analysis of NZSP, YWKS and NZPG each represent a search area of around 10,000 km2 (150 km long by 66.7 km wide).
The total length of the 7th Arc from 26°S to 40°S is 2,234 km. Searching either side of the 7th Arc +/- 18 NM gives a total search area of 148,965 km2.
Of this, 63,458 km2 has been searched by the ATSB and 88,507 km2 remains unsearched.
The proposed Ocean Infinity search area, as defined above, represents a total of around 36,000 km2.
The original ATSB search area was 120,000 km2 based on a search width of up to +/- 40 NM.
I have included the illustration below to show where the LNAV tracks to NZSP, YWKS and NZPG start on the 2nd Arc (marked in orange).
The red circle is the limit of where MH370 could have flown from the last radar point at 18:22:12 UTC up until 19:41:03 UTC.
The three start points for a Great Circle path to NZSP, YWKS and NZPG would appear to fit start points following the waypoints BEDAX, IGOGU or VOCX respectively.
https://www.dropbox.com/s/h8olw7w4moaauux/2nd%20Arc%20Limits.pdf?dl=0
@Richard/Victor.
1. Thanks for sharing those. If we are in LNAV, speed declines slowly over time, does it? If M~0.82 is your initial speed, where does average/final speed end up and how confident are you that there is adequate fuel endurance for these routes?
2. With RMS BTO as good as this, I am puzzled why these locales did not show up a secondary peaks in the DSTG’s probability distribution. Can you explain that?
“@Rob: I don’t know who Ron is, but you are the one that in a previous comment introduced “transfer pump” into this discussion. I have little tolerance for people playing games with multiple names.”
Victor, Ron happens to be Rob accidentally mis-spelt. Andrew rightly caught me out on the transfer pump.
@Rob: Unless you have multiple personalities and you refer to yourself as another person, you are playing games. I told you that I have no tolerance for this. Knock it off or go elsewhere.
@Victor,
I will knock it off, then
@Richard,
@VictorI,
I have some questions regarding your route fitting results:
1. What are the flight levels?
2. How do you handle the effects of headwinds/tailwinds in computing ground speed?
3. Are you using the GDAS temperature data in computing the TAS?
4. Have you figured the fuel consumption for any of these routes?
5. Are you assuming a constant Mach from 19:41 until fuel exhaustion?
6. What is your justification for fitting a 3-digit Mach number? After an extended Hold, why would a pilot select a 3-digit Mach using the MCP?
7. How many waypoints are entered to define the course from 19:41 until MEFE? Which ones?
8. Which BTO times are you using to compute the RMS of the residuals?
@Paul
I am puzzled why these locales did not show up as secondary peaks in the DSTG’s probability distribution.
The three discrete areas shown in the OI search graphic correspond to three different starting points on the 19:41 range ring. They are not “probability peaks”. If a continuum of 19:41 range rings points were used one would get a continuum of area on the 7th arc.
@DrB: I’ll let Richard answer for himself. I plan to present more articles on these paths, but to quickly answer your questions:
1. FL350 is my main focus.
2. Aircraft velocity is the vector sum of wind and velocity relative to the wind. Heading is changed to maintain track if there is crosswind.
3. Of course.
4. In progress. I am mostly interested in routes with a loiter ending north of 36S. Back of the envelope says air packs off after 19:41 and I get to the required endurance.
5. I consider constant Mach and LRC. I have not considered ECON, although I think the results won’t be much different than LRC.
6. A three digit Mach number is allowed via the MCP. As for why, using your words (paraphrased), in this analysis, I am not limiting the speeds to those I think a pilot might choose. Of course, considering that GDAS data is not perfect, and considering that I only considered one flight level, I don’t worry too much about finding round Mach numbers. Also, KIAS might have been chosen rather than Mach.
7. For NZPG, it’s VOCX-NZPG. For NZSP, it’s BEDAX-NZSP. For YWKS, it’s VOCX-YWKS. In all cases, there is an exit from a holding pattern, so the alignment with the northern waypoint is close but not exact. The northern waypoint is passed before 19:41.
8. BTO RMS errors are computed using 19:41, 20:41, 21:41, 22:41, and 00:11. (I don’t force zero BTO error at 19:41.) BFO RMS errors are computed for the above plus 23:14. Level flight is assumed.
@VictorI,
Thanks for responding. You give the Mach at 19:41 but not for other times. Are the Machs constant at the 19:41 value for each of the three routes listed, or do they vary with time/weight (as in LRC)? If they vary, how is that determined, if different from LRC?
Could you list the BTO times again? There seems to be a typo.
How far are the starting points off the line between the waypoints, and how far off are the initial bearings from the way-point-to-waypoint initial bearings?
@DrB: Many of your questions will be answered in an new article that will post tonight or tomorrow. But here are again some short answers:
Yes, in the list, I forgot to include 20:41 and repeated another. (I fixed my comment above.)
Yes, speed varies with weight in LRC.
I estimate the offset from the waypoint to be within the length of the holding pattern at that waypoint. There will be little change in the results if the exact waypoint is used, especially for VOCX-NZPG and BEDAX-SPOLE. You can try these paths for yourself if you want to run your own tests.
I have posted a revised Fuel Model (V5.4) HERE .
This new version has several changes:
1. I added LRCINOP and HOLDINGINOP speed modes.
2. I modified the L and R tank calculations in the Endurance Model to allow extended INOP flight using all fuel on board.
3. I fixed a bug in setting Delta SAT for the 3 legs in the Endurance Model.
4. I updated the CI = 52 speed schedule based on matching the MH370 Flight Plan and MH371 ACARS data.
5. I added a worksheet with a plot of endurance versus range for various post-18:29 speed modes and altitudes.
@all,
HERE is a paper demonstrating MH371 was flown with Cost Index = 52. My new Fuel Model V5.4 is in good agreement with the MH371 ACARS data and the MH370 Flight Plan for both air speed and fuel flow.
@all,
HERE is a new paper on MH370 endurance and range, including the INOP speed modes I have added to my Fuel Model V5.4. This will answer your previous questions, Ge Rijn.
HERE is a list of all my MH370 publications with links for downloading them. I will keep this list updated so in the future you can use this one link to access everything I have done.
@DrB: Your new paper on MH370 endurance and range has interesting results. I was not expecting CI=52 to have acceptable endurance. Can we assume that since you have not mentioned anything about air packs, the calculations were performed with air packs on?
Bobby Ulich,
You first state, “I updated the CI = 52 speed schedule based on matching the MH370 Flight Plan and MH371 ACARS data.”
You then state, “HERE is a paper demonstrating MH371 was flown with Cost Index = 52.”
Is there there circular reasoning here?
On a separate topic, have you tried reproducing any of the early endurance limits derived by the JIT and documented in the Final Report?
@DrB
Excellent work on Cost Index 52 calculations.
Based on your findings and fuel on board at last ACARS report, what is the expected fuel remaining at Igari?
Is this higher than the estimate in the Factual Information report?
If the aircraft maintained CI52 and was 6nm south of Penang at 1752:35, where would the aircraft be at 1825:27?
Would it be starting its turn at Nilam towards Sanob and exposing its Right High Gain Antenna?
And if it flew Nilam-Sanob-Banda Aceh, and then maintained HDG True, where would it end?
Would it be the Bayesian Hotspot?
Is there fuel remaining at 7th arc?
If yes, could APU have been on, resulting in the aircraft being past hotspot?
Or the left engine was still operating with inoperative Left GCU and Backup Converter and the aircraft is within 100 NM radius of hotspot?
@DrB
Many thanks for the new fuel model and the single engine analysis.
[Comments here are closed. Please continue the discussion under the new article.]