Malaysian investigator Aslam Khan (left), Blaine Gibson (center), and Consul Zahid Raza (right) at the Ministry of Transport in Antananarivo, Madagascar, in December 2016
On August 24, Zahid Raza, serving as the Malaysian consul to Madagascar, was killed in the Malagasy capital of Antananarivo. According to reports, he was shot multiple times while seated in the driver’s seat of his car. Mr Raza is reported to be of French-Malagasy nationality.
Last December, Reuters reported that Mr Raza assisted Blaine Gibson in transferring the custody of pieces believed to be from MH370 from Madagascar to Malaysia. At that time, six pieces were transferred. This has raised questions as to whether there was a link between those MH370 parts and Mr Raza’s death.
What makes a possible link to MH370 even more suspicious is that in the time period surrounding his death, Mr Raza was expected to visit the Malagasy Ministry of Transport, retrieve additional recovered pieces, and deliver those pieces to Malaysia. In a private communication from Blaine to me, he writes (repeated here with his permission):
On August 16 possible MH 370 debris was handed over to Madagascar authorities, and authorities in Malaysia were notified. Under the agreement between the two countries, debris is supposed to be collected by Hon. Zahid Raza, the honorary Malaysian Consul in Madagascar, and delivered by private courier to Malaysia.
On August 24 the Hon. Zahid Raza was assassinated in Antananarivo. At first we did not know if he had picked the debris up before this tragedy. We just learned that the debris is still safely in the hands of Madagascar authorities. However new arrangements must be made for the collection and delivery of debris.
Our thoughts and prayers are with the family of the Hon. Zahid Raza.
In the aftermath of Mr Raza’s death, there seems to be conflicting stories about his background. He appears to be of French-Malagasy nationality, with family living in France and in the French Reunion Island. In one report, he is referred to as the “former” consul, but other reports imply he held the title of Honorary Consul at the time of his death. There is a report linking Mr Raza to a group associated with the kidnapping of residents of Indo-Pakistani descent that are living in Madagascar. (Madagascar does not grant automatic citizenship to those born on Malagasy soil. As a result, some Indo-Pakistani residents are from families that have lived in Madagascar for over a century.) The association of Mr Raza with the kidnappers has not been confirmed, and could be disinformation. Hopefully, the facts surrounding this will surface.
Surprisingly, the assassination of Mr Raza has been met with stony silence from both Malaysia and France, despite his ties to both countries.
I join Blaine in expressing my sincere sympathy to Mr Raza’s family.
@Victor
I am just a humble pilot with some navigation experience.
Taking what I was told I just simply put myself in Captain Zaharie’s situation and have attempted to come up with a plausible flight within the constraints outlined to me.
There are two major possibilities after Cocos Islands.
Turn up towards Christmas Island and on to Java or assuming disruption on the flight deck around Cocos Islands carry on across towards West Australia.
We know they made it to neither.
Good post Victor. It was confirmed to me by a Malaysian official that Raza was the current Malaysian consul to Madagascar. He had a wife and 3 children vacationing in Malaysia at the time of his death. Malaysia arranged to fly them back to Madagascar yesterday. From what I have heard so far, it does not look like there was a connection between the assassination and the (potential) MH370 debris handover in-process. The debris is still with Captain Damasy at the Madagascar Civil Aviation Dept.
@Andrew
You are right, the plane must have come to a violent end. The mangled internal partition fragment from Rodrigues is testament to that. I will admit the more I try to fit the debris to a plausible impact scenario, the more the doubts creep in. It’s frustrating in the extreme. I’m hoping maybe sometime in the future, some clever person will develop a computer program that convincingly replicates the impact, from the debris. This might not be too difficult a task.
I will also confess I am uneasy about the final BFOs. I can’t persuade myself to accept the neat explanation. During an 8 second snapshot of the unpowered descent, the time stamp of which was arbitrarily determined by the APU and SDU run-up times, the aircraft’s descent rate accelerated from 5,000fpm to 15,000fpm ? Call me a doubting Thomas if you like, but to me it’s just too simple an explanation to be true. I may turn out to be wrong, but I just can’t accept it without question. If the descent rate is real, the aircraft was actually dropping at that rate (that’s a 0.6g acceleration. In the subsequent 8 second period, was the acceleration continuing at the same rate? Could it have been 25,000fpm after a further 8seconds? If that’s the case, It would have hit the deck rather quickly. The aerodynamicists will say that’s a gross over simplification and doesn’t consider the increasing air density.
But the plane wasn’t found close to 7th arc, neither at S38 nor at S35. Something has to be wrong with the ATSB end of flight scenario. It’s just too damm straightforward.
@Ge Rijn
I’ll ask you the same question that I put to Rob and ErikN;
What high speed airplane impacts with water are you using for comparison?
And here are a couple of other questions to consider;
Do you think that the wreckage found to date is representative of all the wreckage created by the impact?
What effect do you think the “buoyancy filter” has on our interpretation of the wreckage? Put another way, do you think that there could be a negative correlation between items that have been subjected to extreme compressive forces (ie crushed) and items that retain sufficient integrity and buoyancy to drift for 17 – 31 months?
When contributors say that the wreckage is not consistent with a high speed impact I suspect that they do so using a mental model of what they believe the wreckage from a high speed impact would look like. That’s not an unreasonable approach if you have the right mental model; the problem is that mental models are often wrong.
There have been a handful of actual high speed, near-vertical impacts with water involving commercial jet airplanes; for example, EgyptAir Flight 990, Indonesia Air Asia Flight 8501, AdamAir Flight 574 and Flash Airlines Flight 604. All the accident investigation reports provide some discussion of the wreckage but the Egyptian Civil Aviation Authority accident report on the crash of Flash Airlines Flight 604 provides by far the most detailed list of floating wreckage recovered. The list is dominated by trailing edge lift device (flaps, including an essentially intact left outboard flap), control surface (aileron, elevator and rudder) and trim tab segments and fragments together with spoiler segments and fragments. The list also includes two fuselage frame segments, a wing-to-body fairing fragment, an access panel and a number of interior items (mainly escape slides and life vests). Apart from the number of spoiler segments and fragments, the list is not markedly different from the MH370 list. At the very least the list of floating wreckage recovered from an actual high speed, near-vertical impact of a commercial jet airplane with water is not inconsistent with what we have for MH370.
@Rob
I agree with you Rob. The final dive answer to the final BFO’s is obviously too simplistic and utterly illogical when you look at the flaperon and flap. If you accept the final dive, they had to come off in the air (flutter or overload or both).
Simply put, I don’t buy it.
There must be a more rational answer, and I have searched long and hard for it.
I now think the answer is staring us all in the face, and has been from early on.
I think that answer is best shown in Yapp’s spreadsheet, in my modified version, in Cells N25 and N39 in the image below.
My explanation is as follows.
After the second engine quit, all normal electrical power is lost, and it goes over to battery, with load shedding a plenty, left right and center.
The RAT drops and powers some systems.
The APU eventually started up, and some electrical power is again available to some systems, but only a few systems are powered, and they all take different times to “boot up” and come on line.
The SDU is a relatively simple system and boots up more quickly than AIMS.
So the SDU came back up, but something in the information chain from AIMS to the SDU either had not yet finished booting, or it did not come up at all.
Now, specifically, the SDU needs velocity information to calculate the Aircraft Frequency Compensation.
Without the aircarft’s velocity information, it can not do that, so, the logic must assume it is stationary, so it must default to zero.
That is the same situation, as the first entry in cell N25 when the aircraft was stationary on the ground, but with everything working, and the velocity information was being fed to the SDU, and it would, of course, be zero, since it was stationary at that time.
So, we come to “end of flight”.
I have assumed, that if the velocity information was not yet available to the SDU (for whatever reason) it would default to zero.
So I added a couple of lines to Yapp’s calculator and did some “what if’s”.
If you set N39 to zero, Set ROD to zero (at ditch), and set velocity to 50 knots (decelerating after splashdown), after a bit of “fiddling around”, you get:
BTO 18404.191 = good match
BFO 182.2517 = good match
Heading 182.3705
Position 38.3 South 87.65 East.
Now, there obviously is a “range” of solutions available under this scenario.
It would be of some use if some programming genious could develop a program to test all possibilities of positions near there, for all heading and ditching speed possibilities, from say 200 knots down to zero.
https://www.dropbox.com/s/4rnkglkfm2solco/For_Rob.png?dl=0
https://www.dropbox.com/s/nbudo33wecnemio/For_Rob.gif?dl=0
@Ventus45
RE: “The SDU is a relatively simple system and boots up more quickly than AIMS.
So the SDU came back up, but something in the information chain from AIMS to the SDU either had not yet finished booting, or it did not come up at all.”
“Without the aircarft’s velocity information, it can not do that, so, the logic must assume it is stationary, so it must default to zero.”
That’s an interesting theory. I’m not saying you’re wrong, but the AIMS and ADIRU both have multiple power sources, including the main battery. Loss of power from the IDGs should not affect their operation unless the battery goes flat. The RAT and APU should start producing power long before that happens. Consequently, the supply of IRS data to the SDU should not be interrupted. I can’t think of anything in the ‘supply chain’ that would stop that supply.
@Ventus. Thanks for setting that out. I am of the same view – that the BFO values at 0019 are anomalous due to SDU booting up without all of the inputs required for correct frequency compensation. Re your theory that “these are the values you would get if lacking aircraft velocity compensation”… is this not what Gysbrecht had previously demonstrated? I have tried to re-locate his paper without success. Hopefully somebody with a better memory/orderly reference management can put us right.
@Ventus 45. He needs to get to -2Hz at 19:37, 8secs later?
@Ventus45
RE:
1. “The SDU is a relatively simple system and boots up more quickly than AIMS.
So the SDU came back up, but something in the information chain from AIMS to the SDU either had not yet finished booting, or it did not come up at all.”
2. “Without the aircarft’s velocity information, it can not do that, so, the logic must assume it is stationary, so it must default to zero.”
In addition to Andrew’s comment about the AIMS and SDU power (actually, applies to all of the AES power), which is correct, the second quote is not correct. There is no “…default to zero…” mode. The SDU BITE logic prevents any logon unless and until all the boot up criteria have been met, which includes availability of IRS data via 429 bus.
Thus, there is no reason…at least, none following that logic…to suspect there was anything wrong with the final two BFO values.
@Ventus45 & Paul
Power supply to ADIRS and AIMS is highly redundant: three 28VDC sources to AIMS and the ADIRU, including the Hot Battery Bus. The SAARU is powered off the Centre FCDC power supply assembly, a unit that also contains its own battery.
For the reasons above it is highly, highly, unlikely that AIMS and ADIRS experienced any power interruption, and interruption to processing, even considering the conditions during and after fuel exhaustion.
Further, the AMS(R)S specification for performance of the AES states that the “maximum rate of change of the frequency of the transmitted signal when compensated for aircraft acceleration in the direction of the satellite shall not exceed 15Hz per second.”
The difference in measured BFO, 00:19:29 to 00:19:37 (8 seconds) is 184 Hz, 64 HZ outside this defined maximum rate of change. Uncompensated aircraft vertical speed is most definitely a contribution to the measured BFO (the dominant contribution).
In anticipation of a response concerning GES performance and the exceedance of that rate of change: there is nothing in any of the parameters recorded by the GES at 00:19:37 to indicate that its demodulator was stressed to maintain lock with a varying carrier frequency while it decoded the burst: RxPwr, C/No, and Est. BER, all OK.
@Ventus45, Paul, Don, ALSM
I find it still hard to imagine a manoeuvre which explains for a roughly 10,000 ft/min change in vertical speed in 8 seconds (unless by coincidence the onset of a near free fall was registered). With a failed Doppler compensation mechanism you could explain the change of 184 Hz with a relative small change in course, for example as part of a turn from southerly to westerly direction being in progress.
(If I remember well it was Oleksandr who first proposed this possible explanation)
I understand such failure in the compensation mechanism is difficult to understand and perhaps not very likely, but IMO good to keep the brainstorming open on this point.
@Victor
When you met with Blaine recently, did he say why he stopped looking for debris? Obviously he already made an important contribution and may have been to the point of diminishing returns. Is he still involved in some ways?
@Niels said, “I find it still hard to imagine a manoeuvre which explains for a roughly 10,000 ft/min change in vertical speed in 8 seconds (unless by coincidence the onset of a near free fall was registered)”
@Rob said, “During an 8 second snapshot of the unpowered descent, the time stamp of which was arbitrarily determined by the APU and SDU run-up times, the aircraft’s descent rate accelerated from 5,000fpm to 15,000fpm ?”
It may be that the restart of the SATCOM and the steep descent are related, and that’s why the BFO snapshot registered the descent. For instance, if there was an engine restart, as that engine spooled up in speed past idle, the associated IDG might have brought the AC busses to life, and powered the SATCOM. Within a minute, the asymmetric thrust might have caused yaw and induced a roll, leading to the steep descent.
It’s also possible that after the failed restart, the dihedral effect of the wings tended to level the aircraft as it transitioned from a steep, banked descent into a phugoid with a shallower bank.
A dialog with Boeing would help us to better understand what end-of-flight sequences are possible. Any entity wishing to determine the width of a search area along the 7th arc should demand access to Boeing’s end-of-flight models.
@Ventus45
Thank you. What you suggest as an explanation for the final BFOs is certainly worth further consideration.
@Rob,
@Ventus45,
@Paul Smithson,
Re: “During an 8 second snapshot of the unpowered descent, the time stamp of which was arbitrarily determined by the APU and SDU run-up times, the aircraft’s descent rate accelerated from 5,000fpm to 15,000fpm ?”
It is very unrealistic that the plane was subjected to the average downward acceleration of 0.6g over the time interval of 8 seconds, and that the ping coincidently happened exactly at this time. The IG has a difficulty to grasp this. What was the acceleration a few second before, and a few seconds after? What was the drag? One also needs to remember about horizontal velocity component. As far as I know, none of the simulations were able to reproduce such a scenario. Thus, it would be silly not to doubt that the last pair of BFOs “as is” is representative of a rapid descent.
There are at least two explanations of -2 Hz separated from preceding 182 Hz by only 8 seconds:
1). Two years ago I noticed that the log-on sequences 18:25 and 00:19 are absolutely identical. Both the second pairs of BFO and BTO in these sequences were abnormal. This suggested that the reason could be the same. So I did an experiment: I assumed that the BFO abnormalities were caused by the absent Doppler compensation. Some interesting results were outlined in my paper:
https://www.dropbox.com/s/r551bp495n2juoc/TN-ABFO-Rev1.0.pdf?dl=0
2). “Inverted ALSM/DrB”. ALSM has proposed that the log-on 18:25 was affected by the ‘overshot’ caused by OXCO warm-up process, while logon 00:19 was not affected due to a relatively short power interruption. This theory, however, has failed to explain the correctness of the BFO of 142 Hz (18:25:27). Ian Holland simply suggested ignoring this value as unreliable. Furthermore, all the curves shown by in Holland’s paper exhibit monotonic behavior, what also works against theory proposed by ALSM. However, if you invert ALSM’s assumption by applying it to 00:19 logon instead of 18:25 logon, then the ‘real’ 00:19:29 BFO value could be much lower than 182 Hz. It could be as low as -2 Hz. In other words, the plane could be descending at 15,000 fpm already by 00:19:29. In general, you can get solution for any accelerations between 0 and 0.6g. But the problem is still there: the BFO of 142 Hz remains unexplained within the frame of this theory.
@Don Thompson, @Andrew, @All
Iro the final BFOs, I don’t particularly take exception to the positing of a high descent acceleration at some stage in the descent, what I do take exception to is being asked to believe that such an acceleration could have just happened to coincide conveniently with that 8 second snapshot, arbitrarily sampled, when the APU runs up and prompts an SDU. The explanation is simply too convenient to be real. Simple as that. There must be another explanation for the BFOs.
@TBill: Blaine and others have been harassed, and this has forced him to take a less visible approach.
When we met in Virginia on Aug 3rd, he showed me photos of the debris that he was working to get from Madagascar to Malaysia. He explained to me that safety of the individuals involved required that this transfer be done carefully and quietly. That debris was delivered to the Malagasy aviation officials on Aug 16th. By virtue of a negotiated agreement, Mr Raza was specifically named as the responsible party for collecting the debris from the Malagasy authorities and sending them by courier to Malaysia. The death of Mr Raza has certainly complicated this transfer.
Mr Raza’s assassination could be unrelated to the debris he was helping to transfer. However, the timing and circumstances of his death are very concerning.
@Rob said, “There must be another explanation for the BFOs.”
As I proposed above, a failed engine restart could explain the timing of the SATCOM reboot and the steep descent.
@Mick Gilbert
On your comment:
@Ge Rijn
I’ll ask you the same question that I put to Rob and ErikN;
What high speed airplane impacts with water are you using for comparison?:
Reply; Mainly SilkAir flight 185, SwissAir flight 111
Do you think that the wreckage found to date is representative of all the wreckage created by the impact?
Yes, I think it is representative of the floating debris that remains drifting after a ditch-like impact. There must have been (or still are) many more pieces floating around which we still don’t know of.
But IMO the pieces found so far are virtualy all representing a relatively low speed ditch-like impact.
Those found pieces are all we have to draw certain conclusions from.
I drew mine though I know it’s not settled yet.
Also the ATSB does yet not exclude the possibility of a glide (and consequently a ditch) after a (steep) descent. In the latest report this is specifically mentioned as such for the first time (as far as I remember).
On your ‘buoyancy filter’:
What effect do you think the “buoyancy filter” has on our interpretation of the wreckage? Put another way, do you think that there could be a negative correlation between items that have been subjected to extreme compressive forces (ie crushed) and items that retain sufficient integrity and buoyancy to drift for 17 – 31 months?
No, I don’t quite think so. Retaining sufficient integrity and buoyancy sure. But with a high speed dive impact all found MH370 pieces should also show (extreme) compression forces/damage. Which is not the case at all with any found piece. ~99% tension-induced damage. Hardly any compression damage and intact leading edges on flaperon and flap section.
Regarding Flash Airlines Flight 604 you sure make a strong point considering the kind of floating pieces that were recovered.
But IMO the kind of damage to those parts is quite different to what we see with most of the recovered MH370 debris so far.
It also specificaaly mentions the leading edge of the found outboard flap section was crushed.
And when you look at the pictures of some of the recovered Flight 604 floating (flap related) debris the damage shows quite different too compared with MH370 debris.
All twisted sheared and torn in a way we don’t see in most MH370 flap/wing-related debris.
Page 759 and further:
https://www.tailstrike.com/030104.pdf
But I acknowledge your Flight 604 example is remarkable considering it went in the water with ~740km/h and 24degrees nose down.
Not near a vertical dive but anyway maybe this AoA had something to do with the seperation of all those flap related pieces.
@Andrew, @Mick Gilbert
And just to get back to the debris for a moment, and what it might suggest about the impact: Yes, agreed the impact was violent, but I have to bring up those two clapeyron closing panels, item nos 9 and 15 on the Malaysian Investigators listing. These panels come from identical locations on the right and left wings respectively, and are in a remarkably similar condition, as the photos demonstrate. The trailing edges with the seals are largely intact. There are other panels on the upper wing, but only these two made it to land. For just these two panels to have been separated when the air plane nosed in at high speed, defies belief imho. The assemblage of items recovered relating to the RH outboard flap and it’s fairing, the RH flaperon itself, panel no 9, but far fewer parts from the LH wing trailing edge, suggests damage largely confined to a localised area of the RH wing trailing edge, rather than a general breakup of the airframe, even when you take into account the fact that only objects made out of composite and therefore float, will make it to shore. Even when you take this fact into account, the lack of parts from other areas of the plane presents a real problem for the high speed, nose first unpiloted impact scenario. The Rodrigues interior panel cold have been released when the nose broke away on impact, an impact with a high ROD with low forward velocity. If the pilot was relying on RAT power, it’s quite possible he lost control in the final moments, and made a hash of the crash. It would explain panels 9 and 15 much more plausibly than a nose first, high speed impact.
I fail to understand why anyone would question the 0.68 G acceleration. That is not unusual at all, given the vertical speeds that have been documented in previous upsets. There have been many documented cases. See for example ATSB June/Aug 2014 Report, Appendix C (.pdf pg 53) for a few cases, some with descent rates over 30,000 ft/min. On Feb 12, 1963 N724US lost 19,000 feet in 30 seconds. That does not happen without extreme vertical acceleration. A 60 degree bank angle produces a 2 G turn, common for soaring pilots coring a thermal. Really…it is not all that extreme.
@Victor
If any of us lot disappear mysteriously, it might be because we’re getting too close to the truth for comfort.
@ALSM, I don’t object to 0.6g acceleration. I just object it to it happening precisely between 00:19:29 and 00:19:37.
@ALSM: “A 60 degree bank angle produces a 2 G turn, common for soaring pilots coring a thermal. Really…it is not all that extreme.”
I bet you would never win a soaring contest if that is your ‘common’ technique for flying thermals.
@ALSM”“That does not happen without extreme vertical acceleration.”
It does not happen without a confused pilot mishandling the controls.
@ALSM
As others have indicated, the 0.68g is not necessarily extreme. I would like to ask if you have any evidence (from calculations or simulations) that this negative vertical acceleration could have persisted for, say, a minute or more. It would make it more easy for me to accept that it was captured during the 8 sec interval.
@Victor
“If any of us lot disappear mysteriously, it might be because we’re getting too close to the truth for comfort.”
I volunteer, being old and degenerating rapidly. It would be simpler than walking out behind the barn with a Glock.
Another relatively “off the wall” diversion I am considering is investigating the flight path using geometric algebra which was not commonly used in my generation, but which has experienced a resurgence in the last couple of decades. I believe it will eventually replace linear algebra in the physical sciences. The reference link below is a start. I have checked it, but remain suspicious of math errors. It is also unclear how this method could work with satellite velocities other than near zero (such as the case at 19:40).
http://tmex1.blogspot.com/2017/08/geometric-algebracalculus-mh370.html
The CAB report for Flight 705 (N724US) does make for very interesting reading.
This refernce may be easier to read each individual page. Click the right arrow to go through the 30 pages of the report.
@Mick
Re: McMurdo
In FS9, it is the 180S CMH that gets close to the McMurdo path from 1090E, with the Fair Skies weather setting which gives a fairly strong 25 nm wind from the West to push the aircraft. In order to see this close match, you have to know how to manually update the FS9 magdec.bgl control file from the original file to 2005 standard magnetic data. According to ALSM, ATSB previously acknowledged that MH370 is believed to have used 2005 magnetic headings tables.
@Ventus45
You will have to explain how a Boeing 720 crash in 1965 is relevant.
@DennisW: To be clear, the quote in your comment was not mine. If people in Madagascar associated with debris are unsafe, my guess it is either because local thugs are opportunistically trying to make money, or there is an effort to squelch publicity surrounding new finds in attempt to end discussion about continuing the search. Just a guess.
@All
I would like to ask who has studied 180TT (/waypoint NZSP) paths, and if possibly refs exists or details could be shared. My path generation tool indicates a path following the 93E meridian from S12 (@21:20) onwards which must give a very good fit to BTO/BFO with at first sight a reasonable speed profile. It ends at S34.8 (00:19). One obvious ref which gives an example path which is close is the Inmarsat paper (JoN), but I remember others (Victor?) discussing this option.
@Niels: Here was my thinking back in Aug 2014. At that time, I had not fully incorporated meteorological data and autothrottle modes into the flight model, so the speed model (which was a simple parabolic fit) is simplified. Nonetheless, I think you’ll find the results useful.
@Niels
I also embrace a track of near 180 at 19:40. My own work both awhile ago and recently suggests a ground speed of 420 knots or so is a better fit to the BFO data, but that is an assumption based on a stable AES oscillator frequency. This assumption has weakened considerably over time.
@Niels: I should add that the path I proposed in the July-August 2014 time frame was based on a BEDAX-SouthPole geodesic (great circle) path.
A bit more on the debris, then it’s goodnight. Problem: how is it only the outboard flaperon seal/closing panels got forced away from the wings and so made it to land? Possible clue: these panels, 9 and 15 are the outboard of the three panels above each flaperon leading edge. The outboard side of each flaperon is thicker and has a greater chord than the inboard edge.
To me, the best explanation to account for panels 9 and 15 separating is that they were forced away from the wing when each flaperon was itself forced to rotate upwards beyond its stops, by a powerful upward force when the aircraft hit the water in a flat attitude with a high ROD. Evidently the outboard panels received more of an upward kick than their neighbours, because the flaperon is beefier outboard. Also, to explain why the RH flaperon made it to shore but apparently the LH one didn’t: when under RAT power alone, the LH flaperon actuators are both in bypass and the flaperon is free to rotate about its hinges. whereas the RH flaperon is powered by its primary (outboard) actuator, alone with the inboard actuator in bypass. Come the impact, the RH outboard actuator offers resistance to upward rotaton, and causes the hinges to fail. The LH flaperon, with both actuators in bypass, is free to rotate upwards, offloading the pressure on the hinges.
S
All possible evidence for the RAT being the only source of hydraulic power when the aircraft hit the water?
@Victor
Thank you for the link, that’s indeed interesting info to compare with.
@Victor, Dennis
The path I am looking at would have a slight westerly component to roughly Cocos latitude and then goes straiht south along the E93 meridian.
19:41 GS indeed a bit low,around 420 knots. Average GS after 19:41 450 knots. 19:41 latitude close to equator.
Concerning the BFO error characteristic: I’m planning to ask ATSB/DSTG for more info. For example knowing the average BFO error for all 20 previous analyzed flights would already give some useful insights.
@Niels
Don’t get me started on average and variance relative to BFO error. It will only piss off sk999. As far as BFO to fit at 19:40 near the equator, 420 knots is actually a bit on the high side for a track of 180.
@TBill
Re: McMurdo
180S CMH, thank you for running that down.
@Ge Rijn
On the basis that the SilkAir B737-300 that operated as flight 185 was roughly four times longer than the Musi River is deep at the point of impact, I’d argue that that crash is not a valid exemplar for high speed airplane impacts with water, particularly with regards to floating debris. Swissair Flight 111 was really a nose down cartwheel into water; 20 degrees nose down and between 60 – 110 degrees right bank at about 300 knots at the time of impact. Moreover, there was no inventory of floating wreckage compiled for SR111 so what are you using for comparison?
I’m sure that one of the many qualified engineers who contribute can explain the differences between failure in compression and failure in tension but I’d suggest that the 3 confirmed debris items that show signs of failure in tension can’t be used to dismiss the possibility of a high speed impact; all it demonstrates was that a (probably largely unidirectional) force was applied to the material and it failed in tension.
What the inventory of floating wreckage recovered from Flash Air Flight 604 clearly demonstrates is that the floating wreckage recovered from MH370 is not inconsistent with a high speed impact with water.
Niels:
No, obviously, with only two bfo values, we only know the quasi instantaneous acceleration at one time. From the simulations, I would say there is a good chance we happened to sample part of a phugoid. If so, it could have actually climbed a few thousand feet 20 seconds later. That would be quite common. But that would be followed by one or two more cycles ending in a high speed impact. You simply cannot impact the ocean at the top of a phugoid.
@Don.
Re the B-720, ALSM raised it to support the dive theory – so I looked it up.
The way I would sumarise the report of N724US is that it was a severe weather event induced loss of control, into a severe 9,000 feet per minute updraft, decreasing IAS, then into a sevvere downdraft, into a dive, to high speed, with an atempted recovery, causing in flight breakup, due to the tail coming off, resulting in a violent pitch down into negative G, resulting in the forward fuselage seprating upwards with respect to the remaining centre section of the fuselage with wing, which (due to the wing quarter chord axis pitch moment) rotated downwards so fast that all four of the engines were thrown off upwards with respect to the wing, and the outboard wing sections separated from the inner wing sections in a downwards direction due to negative AOA overload and/or aeroelastic induced twist. The wreckage components then free fell to the ground.
I agree that it is hardly a good comparison to what has been suggested for MH-370’s end of flight, in benign met conditions, but a good read none the less.
Don, Ventus45
I never intended to suggest that N724US was a good example of how 9M MRO ended. I was only respodind to the question raised about the possibility of a 0.68 g accelleration. There are other cases in Appendix C that might better match the 9M MRO case.
@Niels
I have a preliminary path proposal 180S CTH from ISBIX. I suppose a ghost flight scenario requires a True Track heading to meet the Arc5 to Arc6 BTO distances. I get around that by assuming a live pilot fiddling with controls after Arc5 when lots of things are happening (twilight, sat call, sunrise, high winds, etc)
Ventus45. In passing, what can explain how the SDU would see zero knots at 19:29 is a powered ditching with little damage. The APU can start on the water, wheels up, on AC loss.
I think ELT non-transmission is explicable. The APU would have fuel so IFE non-setup would be due to its failure for other reasons or pilot switch off.
However what is left to explain is how some sort of SDU corruption could result in the -2Hz BFO 8 secs later, the prospects of a successful ditching* and how the aircraft would have got so low under power without a descent being apparent at the 6th arc.
(* btw a more successful jet transport ditching than Sullenberger’s some may know about (the aircraft was towed ashore replete with crew and passengers) is at:
https://www.flightglobal.com/pdfarchive/view/1964/1964%20-%202230.html)
As to current interpretation of the BFOs, these do not require a 0.68g vertical acceleration as many believe.
At 250 knots IAS, 30,000ft, ground speed is 40,000 fpm. At 0:19:29, nose down of 13˚ will give about 9,000 fpm ROD. With a further nose down over 8 secs leading to a total of 27˚ the ROD would then become 18,000 fpm. During that an average acceleration in the direction of flight of 0.3g could be expected, allowing for glide drag. TAS would increase 4,500 fpm thereby. That would add 2,000 fpm to vertical speed. Thus the total vertical speed at 0:19:36 would be about 20,000 fpm.
The 8,000 fpm and 20,000 fpm are consistent with the ATSB calculated 0:19:29 and 0:19:36 BFO descents, being mid-way between minima and maxima. Note the nose down is < 2˚/sec; and also that to meet the minimum criteria the nose down angles would be markedly less.
The point I am again at pains to make is that 90% of increase in ROD over the 8 secs in the above example comes from nosing down the speed vector. The vertical component of the 0.3g is a measly 0.1g, not 0.68.
It appears that high bank angles would be needed for those nose drops if unmanned, though perhaps the simulations show otherwise.
Also I note that assuming SDU compensation is based on the ground velocity vector, the nose drop will affect this. In the extreme of a vertical descent even at Mach 1, speed compensation by the SDU again should be zero.
@Rob. The timing of the BFOs is crucial and I have raised this in this forum and with the ATSB. They will not release the details, not just the timing you mention but also details of how close Boeing simulations are to the BFOs. By that I mean not just where they fall within descent minimum and maximum limits but how close they did so to a 8 second spacing. As you know the ATSB has indicated that information is proprietary to Boeing, so all we have to rely on are generalised assertions plus some simulator tracks without timings or descent rates.
@ALSM. I doubt any part of a phugoid would lead to descent of this rate and amplitude, though Gysbreght may be able to add something there.
@Rob et al
RE: “During an 8 second snapshot of the unpowered descent, the time stamp of which was arbitrarily determined by the APU and SDU run-up times, the aircraft’s descent rate accelerated from 5,000fpm to 15,000fpm ?”
A couple of further observations regarding the BFOs:
1. I don’t believe it’s correct to say the timing of the final SATCOM transmissions was ‘arbitrarily determined’. As we all know, the ATSB believes the final SATCOM transmissions were caused by the APU auto-starting and then running for a short time on residual fuel after both engines flamed out. The APU is believed to have run long enough for the SDU to complete its log-on sequence, as determined by the last two SATCOM transmissions. That theory was based on Boeing’s analysis of the aircraft’s likely behaviour at the point of fuel exhaustion and is hardly ‘arbitrary’.
2. It’s not a given that the aircraft’s rate of descent increased from 5,000 ft/min to 15,000 ft/min over the course of the ‘BFO snapshot’. The ATSB suggested a range of possible rates of descent in its report MH370 – Search and debris examination update [p.11]. The derived rates of descent were a minimum of 3,800-14,600 ft/min on southerly headings and a maximum of 14,200-25,000 ft/min on northerly headings. If the aircraft was established in a tight spiral during the final SATCOM transmissions, it could have easily turned through 90 degrees or more during the 8 second snapshot. That leaves a huge range of possibilities for the aircraft’s actual rate of descent, depending on the aircraft’s heading at the time of each transmission.
3. I fail to understand comments that state it’s too ‘convenient’ that high rates of descent occurred during the BFO snapshot. The end-of-flight simulations conducted by Boeing suggested a range of possibilities, including scenarios where “the increase in descent rates across an 8 second period (as per the two final BFO values) equalled or exceeded those derived from the SATCOM transmissions” [ibid, p.14].
Under the ATSB’s scenario, the APU is estimated to have taken approximately one minute to have begun producing power after the second engine flamed out. The SDU is estimated to have taken a further minute to boot up after power was restored. That leaves a period of approximately two minutes before the first of the final SATCOM transmissions, during which the aircraft could have been uncontrolled, assuming no human input.
The fact is we don’t know the aircraft’s end-of-fllght behaviour with any certainty. The end-of-flight simulations were not conclusive, because the aircraft’s configuration (ie rudder trim, etc) when the second engine flamed out is not known. Nevertheless, it is likely the aircraft was descending before the second engine failed and it is possible that it entered a turn shortly thereafter. With no human input, it is also possible the turn could have tightened considerably, with increasing angle of bank during the two minute period before the first of the final SATCOM transmissions.
How is that too ‘convenient’ an explanation of what might have happened? It seems quite logical to me, based on the aircraft’s likely fuel state at the time of the final SATCOM transmissions and an analysis of the aircraft’s likely behaviour. It shouldn’t be ruled out simply because it doesn’t fit with the preferred theories of some others. The simplest explanations often turn out to be correct.
Regarding comments about the debris suggesting a ditching as opposed to high speed impact, I’ll say the same thing I said before: I don’t believe that enough evidence has been recovered to say what happened during the aircraft’s final moments, except that it came to a violent end.
@ventus45
Thanks for posting the accident report of Northwest 705. It’s a sobering reminder of what can happen in a jet aircraft upset. It’s also sad that some of the lessons learned from those earlier accidents seem to have been forgotten in more recent times. It took accidents such as Air France 447 and Indonesia AirAsia 8501 to refocus the attention of the airlines and regulators on the importance of upset recovery training for airline pilots.
The first sentence of my previous post should read: A ‘few’ further observations…
It started out as two and then grew!
@Victor,
Blaine was completely selfless in his search efforts and just shows how much compassion he had for the NOK and their immense suffering. There are very few people who would have scoured the beaches months on end, on their own dime and time. If not for his perseverance, vital pieces of debris may have been lost forever. He did a splendid job! I do not understand people who besmirch his good name and reputation. If anything, these people should have been by Blaine’s side and followed in his footsteps for the good of others. I feel lots better now 🙂
@Mick, Ge Rijn. For my part I hope we can look forward to less perfunctory inspection reports than those to date, in particular of that item from the fin, evidently crushed and buckled, and flaperon. I note though that such reports are unusual so apparently there has to be a specific reason for them.
The ‘selected’ Flash Airlines 604 Exhibit E Attachment 6 photos, include FW36(elevator) & FW45(rudder) with carbon fibre skin clearly ripped apart. No MH370 item looks like that. Other photographed control surfaces are all spoilers, metal skinned, and there are two of flap parts, one of which at least is of metal skin. If the selection is typical one is left head scratching as to why spoilers should be so prevalent. Perhaps they were deployed. These control and lift surfaces look badly mangled in comparison with MH370 items but then carbon fibre does not resist deformation like that.
The Exhibit 5, item FW26 which they describe as a flap looks too short (7ft) to me so more likely a part. All other flotsam is shorter than that and generally the large items are much smaller and more damaged than their MH370 counterparts. One wonders how many would have floated/endured over thousands of miles and a couple of years. Not many if any I reckon.
To me this is evidence that a high speed crash is less likely to result in flotsam being found. Thanks for the reference Mick.
@Andrew: “If the aircraft was established in a tight spiral during the final SATCOM transmissions, it could have easily turned through 90 degrees or more during the 8 second snapshot. “
Have you seen an unpiloted simulation where the airplane was in a tight spiral two minutes after the loss of the autopilot?
@Andrew, I think yours crossed mine, again!
@David,
You made an excellent point regarding how the last two BFOs can be satisfied without an extreme vertical acceleration. I’ll run some cases with my BFO model and check it out.
@David: “I doubt any part of a phugoid would lead to descent of this rate and amplitude, though Gysbreght may be able to add something there.”
The simulations typically show a combination of a phugoid motion superimposed on a gradually increasing average rate of descent caused by the bank angle. The phugoid motion alone is unlikely to result in rates of descent greater than about 4000 – 5000 fpm. To add 10,000 fpm requires a large bank angle and an elapsed time after loss of autopilot greater than two minutes.
As Rob rightly observes, the timing is essential.
BTW you need to think a bit longer about “As to current interpretation of the BFOs, these do not require a 0.68g vertical acceleration as many believe. “. They do require that vertical acceleration.
@David: Also the phugoid component cycles in sinusoïdal fashion between minimum and maximum values where the rate of change is zero. The rate of change is maximal when the phugoid component is zero.
@David
Perhaps I misunderstand your way of reasoning, however if you say:
“The point I am again at pains to make is that 90% of increase in ROD over the 8 secs in the above example comes from nosing down the speed vector. The vertical component of the 0.3g is a measly 0.1g, not 0.68”
The nosing down of the speed vector would need a transverse acceleration which in this case is downward oriented. Do you take it into account?
@David,
Here’s what my BFO model shows at 35S,93E:
BFO varies by only 3-4 Hz as horizontal speed changes from 250 kts to 400 kts.
BFO increases by ~18Hz as true bearing changes from 180 degrees to 0 degrees.
At 180 degrees and ~400 kts, you need -4400 fpm to get +182 Hz and -14,850 fpm to get -2 Hz.
At zero degrees true bearing, you need -5400 fpm to get +182 Hz and -15850 fpm to get -2 Hz.
So the BFO is very insensitive to horizontal speed and rather insensitive to the track. The vertical speed dominates.
I do like the “nose-down” approach at 00:19. It may have been discussed previously, and I calculated a “nose-up” BFO curve in 2016 to see if that might fit the 18:25-18:28 BFOs. I also calculated a steep descent (nose-down) at 00:19, but at the time I didn’t know how much the airspeed would increase due to gravity being somewhat offset by increasing drag. I recall that the angles I found were similar to the numbers you provided, and they didn’t seem out of the question.
Now you have demonstrated that the vertical acceleration can be much smaller than 0.68g. In my mind this strengthens the case for interpreting the final BFOs as a rapid descent.
@David,
Gysbreght and Niels are making the point that the nose-down path can only be achieved by significant downward force (gravity minus lift = downward acceleration). I don’t see this included in your calculation, so there may not be a large reduction after all.
[Comments here are closed. Please continue the discussion under the new post.]