By Victor Iannello
Don’t be fooled by claims of the red tape causing the delay in the determination of the provenance of the flaperon.
Boeing and the NTSB were parties to the investigation when the flaperon was first brought to Toulouse. It is very unlikely that the Spanish subcontractor ADS-SAU did not immediately turn over all documentation when requested by Boeing. The investigators had to know soon after the start of the investigation what the provenance of the part is, whether or not that determination was made public.
I have said before and continue to believe that there was an attempt to delay the release of the results of the investigation in parallel with planting a seed of doubt regarding the provenance of the part. Just look at the series of events this week. First the claim that Spanish vacation schedules have delayed the identification of the part. Then the claim that the identification was not possible. This was followed by the claim that the flaperon was certainly from MH370.
The pattern of leaking contradictory or false information to the media from off-the-record sources continued in full force this week. I believe this is a story in its own right that should be getting a lot of attention. Perhaps when enough journalists are made to look foolish by reporting contradicting statements, their “reputation instincts” will kick in and compel them to dig deeper.
We who are following this incident should demand that more facts be fully disclosed. Technical reports should be released so that we are not parsing statements from a judge-prosecutor to understand the true meaning of what was written. And journalists should not blindly report statements without attribution.
While Adding…
Remain calm, all is well.
Brock,
Most of the classic conspiracy theories are supported by facts and logical chains by their proponents, but the issue is often the jump to the conclusion that is drawn. In this case that the twists and turns of a complex multi party, multi national search conducted with limited and ever changing information in the full glare of the media spotlight, are evidence of deliberate manipulation and a hidden ulterior motive.
I think your position is both passionate, and biased. But that’s ok, we all have our own predilections and preconceptions based on our personal histories and world views. You are after all an auditor by profession.
I would never suggest that you do not post your ideas, and this is the ideal forum for them. I note you are happy for them to be challenged, and hope you take this in the spirit that is intended!
@DennisW:
The video of the Australian journalist can be found here:
https://au.news.yahoo.com/thewest/wa/a/22235213/flight-recreated-in-777-simulator/
In this video the airplane maintained altitude after dual flame-out until it stalled, which suggests that the autopilot somehow remained engaged or was re-engaged.
ALSM posted details of his simulator tests on Posted November 10, 2014 at 3:11 PM and at 8:41 PM, then on November 15, 2014 at 5:05 PM. I kept copies of those posts because I expected more discussion which didn’t occur at the time. From the November 15 post:
A comment on the observed Mach number of 0.9 .
9M-MRO had a maximum operating Mach (Mmo) of 0.87, but other variants of B777 are certified for Mmo = 0.89. Regulations require flight tests at up to 0.07 above Mmo, so no hazardous phenomena such as flutter will be encountered up to M.96 .
There will be limited flight test data for the speed range between M.89 and M.96, particularly for sideslip due to mistrim, but no data at all above M.96, so any observation of simulator performance above M.96 is pure speculation.
Gysbreght,
This media report reflects the issue raised by the original post, i.e. “false information”. I’m not suggesting that Geoffrey Thomas, a frequent reporter on matters MH370, maliciously attempted to misguide his audience but he exaggerated the facts in his report. He wrote:
It’s not Australia’s only B777 simulator: Virgin Australia’s Level D certified FFS device was commissioned in 2008.
In fact, Mr Thomas’ experience was conducted in an uncertified fixed base procedures trainer operated by FlightCity in Perth. Its manufacturer doesn’t claim any aero/flight dynamics fidelity nor does its systems simulation appear representative.
The video reflects that after the second engine flameout the captain’s PFD annunciation, incorrectly, remains for ‘A/P’. While this procedures trainer shows the f/o’s instruments blanking it should show that everything on the flight deck except the captain’s instruments will go dark when all engine generating capacity is lost (until the APU or RAT generators come online).
:Don
@GuardedDon:
” he exaggerated the facts in his report”
Agreed, particularly when he exclaims: “we’re spinning, gyrating around the sky, this is awful” while the flight instruments show nothing of that. I think the video is mainly instructive in the earlier parts where he shows how the Autopilot is controlled via the MCP.
You also write “In fact, Mr Thomas’ experience was conducted in an uncertified fixed base procedures trainer operated by FlightCity in Perth. Its manufacturer doesn’t claim any aero/flight dynamics fidelity nor does its systems simulation appear representative.” I’m not sure that is relevant. The simulator is clearly an expensive professional training device. The lack of a motion system is only relevant with a pilot in the loop. If it lacks aero/flight dynamics fidelity and correct simulation of its systems operation, how useful is it for training?
Gysbreght,
A fixed base device is useful for procedural training in a ‘realistic’ environment: this device might serve a use to practice procedures as detailed in the Quick Reference Handbook without need to use time in a FFS. My experience of training facilities (Qantas, Boeing/Alteon, KLM, SAS, AA, & others) is that FFS are utilised 18 or 20hrs a day, 7 days a week.
While the procedures trainer might superficially look ‘realistic’, a certified FFS requires comprehensive compliance to a manufacturer supplied data pack, it must demonstrate deterministic real-time responses to all events, the outside visual scene requires a collimated display system, and more.
:Don
I’ve noticed several recent references to the B777-200 simulator tests Paul and I ran back on Nov2. Those tests clearly indicated that, following fuel exhaustion, the plane will enter a turn evolving into a steep spiral descent with airspeed exceeding 500 KIAS (the instrument upper limit) at times.
For those who find these results difficult to accept, perhaps you will be interested to know that I shared the results with ATSB last Nov., and they informed me that Boeing conducted the same type of end of flight simulations (only many more trials than we did) and their results were essentially the same. So, if you can’t accept our results, maybe you can accept the Boeing test results.
If you do accept these simulator trials as indicative of what probably happened, then the final BFO observations make sense too. They only appear to be faulty (smaller numbers than might be expected) until you understand that the aircraft Doppler values are only compensated for horizontal motion, not vertical motion. The vertical Doppler signal was calibrated during the climb out from 16:41-17:06. During that period, we had Inmarsat data, ADS-B reports and 5 minute GPS position reports via ACARS. So we know the uncompensated vertical Doppler was accurately modelled.
If all this adds up for you, then flaperon separation prior to impact is also a possibility that needs to be considered. The speeds observed in the simulator were high enough to potentially cause this type of damage. Hopefully, we will eventually get some results from the detailed flaperon inspection to confirm or dismiss this hypothesis.
JS, RetiredF4:
I have compiled a few snapshots of the aircraft and flaperon in a single image:
JS:
Re “Considering that most of the stock photos and most of the Reunion photos are taken in bright sunlight, I simply don’t see a white balance setting as a factor.”
Based on the shadows, I would say that the aircraft images were mostly taken during evenings. RGB balance of supposedly gray color is slightly shifted to the red. On contrary, “gray” color of the flaperon appears to be shifted to green-blue. Either biofouling took place, or the flaperon images were taken during such an ambient light.
Either way, I do not see any proof that the colors are different. As well as no proof that the colors are identical.
P.S. For some reason Jeff’s site always ‘eats’ my hyperlinks. Try:
dropbox.com/s/yz35l5201089a5q/aircraft_vs_flaperon.jpg?dl=0
in the previous post. Add https(colon)(slash)(slash)www(dot) in front.
@ALSM
Not doubting your sim results. I was just looking for a summary reference. Likewise your interpretation of BFO at the end of the flight. I have no doubt that you are capable of running the numbers and getting the right answer.
I just did not look at this end of flight scenario, and thought I would brush up on it while things are slow. There is also an uncompensated BFO contribution from heading sensor latency during turns. You might want to look at that.
http://mh370corner.blogspot.com/2015/04/bfo-errors-again.html
@GuardedDon;
You are only considering Level D certified FFS simulators approved for zero flight time training. I understand that those require a motion system, but I also understand that there are Flight simulators and Flight training devices of a lesser level that still must be evaluated, qualified and approved to represent the airplane to the required level of accuracy, and do not have a motion system.
It depends how you interpret “essentially”.
In its October 2014 Update, the ATSB wrote:
Anyway, the airplane was not found close the 7th arc, so that pulls out the plug under the spiralling descent scenarios, whether “low bank angle” or “near vertical supersonic”.
For those, who are tired of broken records… Several months ago I promised a technical note on a possible non-AP flight trajectory, so here you are:
TN-CTS-Rev1.0.pdf
The note might be particularly interesting for those, who, like me, used BTO&BFO residual-minimization approach to calculate AP-based trajectory, as it is essentially the same method. I apologize for the delay.
I will also write up my interpretation of the anomalous BFO values of 273 and -2 Hz and will post it soon.
Part of the link as usually eaten…
Trying again:
https://www.dropbox.com/s/tswp7pknhk9tq1t/TN-CTS-Rev1.0.pdf?dl=0
As I demonstrated earlier, at constant altitude the BFO value is only sensitive to the N-S component of airplane groundspeed. Therefore I do not understand the sensitivity to heading sensor latency derived in your paper on a heading of 180 degrees.
Gysbreght,
I think the note will be especially interesting for you as we already discussed ascending trajectory and you know its background idea.
@Gysbreght
“Anyway, the airplane was not found close the 7th arc…”
There have been several analyses, including one of mine, that show low bank angle descents can have termination points up to 25km inside the 7th arc. That area has not yet been cleared in the Northern sector of the search area so your conclusion cannot be reached yet.
Gysbrecht,
Of course appropriate certifications also exist for FTDs.
Back to my first comment, above: Thomas was using a training system to demonstrate an end-of-flight scenario; you commented that the autopilot seemed to remain engaged post both engines inop; I noted that the electrical system behaviour evident in the video clip didn’t reflect the aircraft behaviour.
Therefore, the training device in the clip didn’t perform representatively.
:Don
@Gysbreght
It is a heading error that my paper addresses. That error can either be static due to a static sensor error or dynamic. The dynamic error will always be present during a turn since the heading used by the AES is not current. So during a turn the compensation applied by the AES will be for a heading that was accurate at some previous time. The inaccuracy of course, being a function of turn rate and sensor update rate.
@GuardedDon:
“Therefore, the training device in the clip didn’t perform representatively.”
Well, I’m not convinced of that. Consider how they simulated fuel starvation:
Mr. Thomas activates a switch just behind the thrust levers that closes the high-pressure fuel cock on the engine fuel control unit.
Simultaneously, but more slowly, the pilot/instructor in the RH seat pulls the thrustlever to the idle position. At first sight, that pilot action makes no sense at all. Why does he retard the thrust lever of a dead engine (for both engines), unless it affects the operation of the systems?
@DennisW:
I’m afraid you didn’t get my point. The N-S component is equal to groundspeed times the cosine of heading. It doesn’t change for small changes of heading about 180 degrees.
@Gysbreght
Actually, I did understand your point relative to the 180 degree heading. The reality is that what you are saying is true only near the equator or for a flat earth.
In any case, the derivation is there for you to look at.
No, that is not correct. It applies at all times, all locations, and the WGS84 earth assumed in Yap’s BTO and BFO calculator.
@gysbreght
OK. I’ll take another look at it. I am perfectly capable of making a mistake. There is even historical precedent for it.
@Gysbreght
Please don’t ask if there is any other kind of precedent besides historical. I am close enough to the edge as it is.
@Gysbreght
You are mixing up BFO (which is the addition of the Doppler and the AES’s calculated correction) and the underlying trend of the individual Doppler and calculated correction terms. The (zero-lag) BFO is a slow function of heading at 00:19 but the components are not – Dennis is saying (I think) that if the correction algorithm is given old data (the time lag) then the calculated BFO is signficantly different – 5 degrees of heading difference gives a large change enough to explain the 00:19 BFO value (at certain headings).
Gysbreght,
The training device in the clip didn’t perform representatively.
If it was capable of adequately simulating an empty fuel tank there would have been no need to intervene and move the fuel control switch to cutoff or retard a throttle lever. And the other points noted.
I think we agree that Geoffrey Thomas exaggerated/misrepresented his case in the clip.
:Don
Gysbreght, Dennis, Richard,
I had a long ‘fight’ with Gysbreght about BFO calculation half a year back, and if I am not mistaken Niels has analytically shown that what Gysbreght says re S-N is a quite accurate approximation if altitude is constant. The keywords are “sensitive” and “constant altitude”. Both the exact and approximate formulations have right to exist.
This is in response to ALSM’s post “When the smoke clears…” (September 5, 2015 at 12:58 PM). I have a philosophical question, mainly addressed to him.
Either way, the flaperon has fallen into the water, agree? Attached or not to the wing at the moment of the impact. I can name the following possibilities:
1. The flaperon was detached in the air at high altitude (fluttering).
2. The flaperon was torn away when it still was in the air, but other parts of the aircraft already hit the water (ditching/gliding), i.e. at the moment of the impact.
3. The flaperon hit the water at high speed (dive) and it was smashed by water and/or other parts of the wing.
4. The flaperon was detached manually and carefully put into the water.
I am not going to discuss the last two options. But I would like to ask how is it possible to differentiate between the first two cases based on the flaperon damages only? In both the cases the flaperon would not be torn away by the water impact.
The BFO Group: Over many months, I have spent time parameterizing the BFO and can state with confidence that apart from the vertical velocity, which is understood to be uncompensated, the residual Doppler is primarily a function of the speed times the cosine of the track angle. The Doppler shift is a function of the speed times the cosine of the difference between the track angle and the “radially inward” angle, which can be rewritten as the superposition of terms proportion to the N-S speed and the E-W speed. The Doppler correction term can be also be expanded in terms of the N-S speed and the E-W speed. When the Doppler shift is added to the Doppler correction, the E-W terms cancel almost perfectly, leaving the N-S term.
Any error introduced into how the Doppler correction is calculated, be it from a sensor lag, navigational input error, or change in the assumed satellite position (remember that one?) will have a big effect on the BFO, as many here have noted.
@Olexandr, yes, that the flaperon has found it’s way into the Indian Ocean can’t be denied, considering that it picked up some hitchhikers along the way. Since the flap-clinging barnacles apparently loathed colder water, the destination – the beautiful island La Reunion – must’ve suited them just fine 😉
But you forgot to mention a 5th option: the flap might’ve separated from the plane sometime after it rode up the Strait and went out of sight for good. It could’ve gotten damaged during the plane’s apparent high speed flight back over the peninsula.Maybe not a very likely scenario but I really hope the French engineers will get eventually more specific re: the cause of separation. However, in the case of an early separation the flaperon unfortunately couldn’t tell us anything about the plane’s further route and it’s final moments.
@GuardedDon:
According to the FCOM, moving the Fuel Control Switch to CUTOFF –
• closes the fuel valves
• removes ignitor power
• unlocks the engine fire switch.
That action quasi-instantaneously interrupts the fuel flow to the engine, causes it to flame-out, starts the rundown to windmilling RPM, and the thrust changing within a few seconds from cruise thrust to windmilling drag, as with fuel starvation.
What is the purpose of moving the thrustlever to idle, how does that action or the fuel control switch moved to Cutoff affect the response of systems, to make it different on the real airplane from an empty fuel tank?
@Oleksandr: The fracture analysis would potentially show different failure mechanisms. For instance, if the flaperon was dragging on or impacted the water, you might see evidence of high bending moments, producing high tensile forces on one surface (leading to elongation and failure) and high compressive forces on another (leading to buckling). Fluttering would produce alternating stresses, which would produce elongation and buckling in a different way.
A crash investigator would be very aware of what kinds of stress would cause a particular kind of fracture, and differentiation between flutter and impact damage should be possible.
@Victor, I agree and tend to think that the French engineers have already formed a pretty good idea about the most likely cause of separation. Which makes it a little unsatisfactory to debate about it.
@Gysbreght
Notwithstanding what other commentators have said, I did look at my analytics, and I do not see a “mechanical” mistake. However, I am having difficulty locating the source for my LTP velocity (in the North East Up format) to XYZ velocity transformation matrix. Frankly, it does not look right to me. I am preparing for an extended trip to Europe (leaving tomorrow), but I will get back to it.
If anyone has a link to a correct matrix, I would appreciate it. Can’t seem to find one at the moment.
@Dennis, a question for you: have you checked what the weather and the sea was like on the morning of March 8, 2014 in the area where you think the plane was ditched?
@littlefoot: The NTSB, the BEA, and Boeing have all investigated the part. I agree with you–I suspect the failure mode is known, which is why I have not speculated. They have the superior expertise as well as the part in hand. The news reports are suggesting that the investigators believe the plane was ditched much further north than the current search area. We need to see the technical details in order to reasonably assess their conclusions, whatever they may be. As we know, news reports have been wrong many times in the past, so we can’t really be sure what the conclusions are.
Perhaps I should rephrase that last question:
On the real airplane, would moving the fuel control switch to CUTOFF and the trust levers to idle modify system response in comparison with empty fuel tank without those actions?
@DennisW: This might be more semantics than anything else. As Oleksandr has said, the sensitivity to E-W speed is in general not zero, but very weak (less than 1% of the sensitivity to N-S speed, if I recall).
@VictorI: Agreed, I should have added the caveat “for practical purposes”.
Victor,
Re: “The fracture analysis would potentially show different failure mechanisms”.
Yes, indeed. It will almost definitely show if the flaperon was detached by the water impact. But I was basically talking about the 2nd Newtonian law: suppose the wing hit water first, while the flaperon’s inertia caused its detachment. Not in opposite way. I would compare with a driver, who doesn’t fasten his seatbelt to make it clear what I mean. I am not sure that the residual deformation of the flaperon would be too much different from the deformations caused by fluttering. May be yes, may be no. Moreover, in both these cases the flaperon eventually would hit the water receiving additional deformation due to the water impact (what is free fall speed of the flaperon, btw?).
Anyhow… Let the crash investigators to comment. Meantime I would also like to have your comments on my TN re trajectory in the alternative flight mode, possibly AT – thanks in advance. I’m considering sending it to ATSB if public peer review is positive.
@GuardedDon:
Another question: If the APU was started before the engines were shutdown, would the A/P remain engaged?
@Gysbreght
You are correct. The NEU to XYZ velocity conversion matrix in my writeup is just very very wrong. Not sure how that happened. The BFO error resulting from heading errors near a heading of 180 degrees is indeed negligible. Will fix it, but not right now.
The basic idea is sound, however, and large BFO errors can result from sensor lag and sensor error.
@littlefoot
I have checked it and posted link here, it was very calm especially comparing to current search area.
A lot has happened since we researched this, so please forgive me if I’ve got anything not quite right, as I’m only going by memory…
A group of us over here in NZ last year spent a lot of time going over the manuals on the question of the electrics of a B777… including consultation with two Air NZ LAMES, and this was our basic findings:
If the pilot’s turn off the main engine driven generators – they cannot be reconnected in flight.
If these main generators are turned off (or fail) the backup engine driven generators come on line.
If (by memory) either of the backup generators fail, the APU will come on line to keep two generators in the loop.
In the extremely rare occasion of all generators failing only then will the RAT activate to supply power for the pilot’s flight essentials only.
The SDU is classed as non-essential and is taken out of the loop by the PCM a long way back… I think it was somewhere about when it was the back-up generators kicking in.
As to the interruption of power as it switches from one generator to another, everyone seems to forget about the fact that their is a battery in the circuit that maintains power in the interim, so I’ve always maintained that all this talk of the RAT power re-energising the SDU for one last “ping” (as far as I’m concerned) is just another lot of ‘miss-information’…. or maybe I’ve missed something?
The RAT circuit is not linked into the SDU and to top it off, if the jet engines run out of fuel, they continue to windmill and as the generators are still connected to the engines (Backup ones at least) THEY STILL GENERATE POWER!
I’m off to work now, so feel free to discuss this amongst yourselves… check it out, and correct me if I’m wrong!
@Gavin: From time-to-time, people erroneously claim that the RAT powered the SATCOM. Then somebody like Don posts a comment to correct them. As is clearly written in the ATSB’s June 2014 report, it is the APU that supplies power to the SATCOM after fuel exhaustion, not the RAT.
@littlefoot
I did not check the sea state for that time. StevanG has also responded above.