The Germanwings 9525 deliberate crash in the French Alps in March, 2015 may hold relevant information & datapoints regarding chip survivability. Both trajectories were near-vertical at very high speeds. In a February 20, 2017 paper entitled "Texas A&M-Led Team Uses Mathematical Modeling to Explain Complete Destruction of Germanwings Flight 9525", the pattern of destruction of the Germanwings aircraft is studied, (link to the actual paper, behind an academic wall, is here). The BEA Report, English version, is here.

While there may be a different focus to the study referenced above, it is encouraging to know that both recorders' chips survived the very high-speed impact with solid-rock and were readable (four days' later). The WQAR, (Teledyne) was X-Ray'd and found not usable.

NOTE CAVEAT: Other than the point being made regarding chip survivability, NO comparison between these two events is implied or intended. It remains unknown why and how the China Eastern accident occurred.

A simpler view of why the recorders have been sent to the US is to apply the original manufacturers flight test data calibrations. An important activity at the time of download given that it might be a one-shot opportunity due to damage.

Many years ago the most damaged recorder analysed was that extracted from the San Louis Obispo fatal accident. Pax / engineer shot the crew, then ‘vertical dive’ suicide. NTSB had the recorder on display in the entrance of their Washington DC HQ

I wonder if this could be related.?

https://www.straitstimes.com/asia/se...light-incident

The FR24 plot for that flight (MH2774/03Apr) shows the aircraft turned back to KUL about 35 minutes after take-off. The altitude trace shows a 'blip' about 18 minutes after take-off, where the altitude changed suddenly from 25,000 ft to 23,375 ft and back to 25,000 ft. There's a second, smaller excursion about 7 minutes later, where altitude changed from 29,000 ft to 28,275 ft and back to 29,000 ft. The aircraft turned back to KUL about 5 minutes after the second altitude excursion and climbed to 30,000 ft. The remainder of the flight appears to be uneventful.

​​​​​​​https://www.flightradar24.com/data/f...h2664#2b5da11b

Yep, Looks unrelated to me. Unless of course you would continue after an "unexplained and uncontrolled" couple of descents.

Yup, not related. Theres no left roll present.

MU5735 and SJ182 is carbon copies.

SJ had a Autothrottle failure.

AAKEE...

I can certainly see why someone would say MH2774 and this China Eastern accident are very likely unrelated (though to declare "yup, not related" at this stage is not something an aircraft accident investgator is apt to say).
But to say, "MU5735 and SJ182 is carbon copies" is a giant stretch, even for an investigator familiar with these and "similar" accidents.. Care to expand on that?

The Germanwings crash was hardly "near-vertical." Perhaps we can leave the exaggeration to the media.

Descent rate was ~3500 fpm and took about 10 minutes from top of descent to impact. Roughly a 30° glide path, over a ground track of about 60-70 nm.

It did, of course, hit a steep mountainside, so the impact angle was closer to 70° or so - and was at high speed.

Airbus cannot be put in attitude beyond -15° pitch. So no question of vertical dive. It was just a planned descent.

Yes.

I should have written that the initial upset is carbon copies.

If the FR24 log is viewed in 3D both these accidents show a reduced groundspeed during the initial upset. There is a not insignificant left movement that can only be explained by a left roll. The height loss and the parallell movement during the first half roll is the same. The dive angle is not that different and the continued roll is very like each other. MU5735 began the upset with higher speed, and also did loose a little more altitude during the roll/tunnel roll.

Descending at 3500fpm taking 10 minutes (around 30,000 to 35,000') covering 60-70NM CERTAINLY ain't a 30 degree descent, more like closer to 6 degrees.

So in your own words, "Perhaps we can leave the exaggeration to the media".

pattern is full, vilas, thanks for your responses regarding my characterization of the descent - I agree with you both and my over-eager post describing the Germanwings descent has been corrected, (strike-out of "near-vertical").

The point being made in the post, "...it is encouraging to know that both recorders' chips survived the very high-speed impact with solid-rock and were readable (four days' later).", stands of course and remains encouraging, at least until we hear otherwise.

It has been a while & there is a silence on the contents of the CVR & FDR.......(Definitely more than 4 days later now)

CAAC is having a PR hard time. So you know the reason…

Any information release would need to come from the Chinese investigative authority. While the NTSB, Boeing, CFM, etc. are all involved, they are basically under a gag order prohibiting them from any information release, regardless of what they may know.

Seriously ?
 

An intermediate report will be forthcoming within 30 days.
This has already been stated, and is the industry norm.

TR

Leeham in a subscription article states that investigators are leaning towards suicide as the cause of the crash.

https://leehamnews.com/2022/04/11/ai...ce-widebodies/

Suicide with two other (much younger and fitter) pilots in the cockpit seems a bit unlikely to me.

That last bullet with the suicide claim seems weirdly disconnected from the rest of the article ?

If it was the Captain I’m sure that he could have found a reason to get them both out of the cockpit.

Just sensational.

Lookup the original Leeham article in March and you'll see, they did just copy the last (bullet) line from that article. The other bullets in that article refer to other options also passed on PP.

...and there are no directives requiring early attention or urgent action on the part of B737 operators

If it was anyone that committed suicide, it was the captain demoted to first officer.

I know that it's happened before, but any pilot who deliberately crashes his plane is also that very rare bird, a mass murderer.

The case is not silicon.

With only one person in the Cockpit, a suicide mission would not have had a successful (!) recovery halfway down to the ground.

For a directive, it would need an understanding of what happened and why it happened. For something happening on automation, that might need some time to figure out.

really? it would take just a few moments, very little two 'fitter' pilots could do to stop it.

Of course it can. Just not in normal law.

From Chinese TV channel CGTN on twitter today: (can't post link yet)

Yes, understand that, thanks. Likely you know this, but sometimes in the course of early investigation, the DFDR may reveal something that could be of an immediate airworthiness concern. If readable or partially-so, I suspect by this time they may have an inkling of what happened and that it is, again likely, that it is not a type or fleet-wide matter. Also, I am taking into account who is in charge of the investigation in terms of the relative silence but the Chinese have indicated a 30-day report will be issued. That nothing by way of implementable procedure or technical change has been issued may be construed as a positive sign. Just grasping at straws like everyone else, believing that, "no news is 'good' news", I suppose.

Yep, an initial pretty plausible indication of what has happened, may trigger these early notifications, etc.

For now, we do have the situation the MU B737 fleet is still not returned to service. Suggesting, IF the CVR is indeed read out, the cause of this tragedy is not a deliberate human aspect, but more technical (weather and all kinds of other fantasies can be ruled out, I think).

TBH, I don't expect China to release anything from the CVR. China does have stringent privacy regulations (effectively, though, unless when the state itself is involved).

So, yeah, for now, we need to wait a couple of days. If no further notices are released, either the FDR can not be read, or this incident is not a copy cat of earlier incidents with a recognizable cause.

Is that confirmed? FR appears to show MU 737s operating flights today.

While preparing my post about this, I checked around 20 frames (the newer ones) and none of these did fly in April. Checking again for you, now for nearly all, none of these did fly in April. All B737-800, the crashed version.

Checking for MU B737-700, you can see, these do fly (regularly), so I assume, this is what you saw.

Partial recovery or a glitch in pressure data implying a temporary climb?

My thoughts are it was plummeting, and the FR data shows a drop in pressure rather than a physical climb. It's a long time since I did any altimetry but isn't the data fed by a static vent? Orient the aircraft body so that vent is in a low pressure air-flow (rolling and falling) and you get the apparent increase in altitude. ???

By wobbling around the aircraft nose, you can get some variations in pressure indication, though not that much as the FR24 data shows. If your theory would be valid, the wobbling would be able to distort the altitude measuring in the 5000-th feet, and that, during a steep decent. Also, when the nose stops wobbling, there is no catch-up visible in the FR24 data, so no sudden discontinuity, graphically showing a drop of 10000 feet (in the end, the recording roughly stops at terrain height). Just make a altitude graph yourself and you'll see, it's all pretty smooth. I think, theory busted.

Oh, and at those speeds, don't wobble the nose around to much, otherwise you may loose your tail-end, if not, even shred the aircraft.

Thanks Widescreen. I take your point about the no catch up; just musing that a 1200ft gain is roughly a 40millibar drop in pressure.

Guess we'll just have to see what the end report has to say.

WideScreen (and others that are fairly new to pprune)... With all due respect, you may want to click on a poster's public profile before you reply to one of their posts. And perhaps even look at some of their inputs to previous accident threads. To put it more simply, telling PJ2 what is needed for a directive to be issued is like telling a fish how to swim...

Cheers!

Yes. Even better than plotting a 2D graph is to use google earth or similar to make a 3D graph from the granular data. It is very informative.

Yep Google Earth is very valuable for the impression, but less for the analytics. Plotting a cloud of dots in just 2D in LibreOffice or so, gives the analytics part. Just take care to handle the time axis properly, and you see many more aspects than in a fancy Google 3D view.

For example: Data points like the Long/Lat give a nice straight line (especially for the first half of the down to earth trajectory) with just one small "S-curve" in it. Practically suggesting, the chance, there are 360's involved is pretty low. And, these data points being subject to aliasing with a max cycle of 30 seconds based on a 50+ ton passenger aircraft movements, is also quite unlikely.