Absolutely not suggesting it to be the case, but the lower wing skin in that condition makes me wonder about the recent pickle fork AD.

Not only it is very possible to put a 737 in such a dive, it's also impossible to pull it out from a high speed dive. It's not a protected aircraft. It was put in 65° dive in Rostov on don, in JAL incident it rolled on the back due to inadvertent rudder trim before it was pulled out. If it was mishandled then it could have easily entered into inverted dive. No surprises there.

My guess - stab trim runaway not noticed by crew. Autopilot eventually gives up an disengages leading to violent pitch down. Crew figure out what’s wrong and retrim only to overstress during the pullout.

Would that be Silkair 185 you are referring to? https://en.wikipedia.org/wiki/SilkAir_Flight_185 I recall the tail section separating from the rest of the A/C during the resulting near vertical dive.

As a lifelong professional jet pilot I make a couple of observations.

- A runaway trim that is not detected(?) will put the aircraft into an uncontrollable dive - but NOT a vertical one.
- This a/c seems to have been just fine and stable right up to the point it left cruise altitude approaching destination
- The fact we are seeing lots of panels on the surface with pulled rivets suggests design forces were exceeded in flight leading to structure break up. The rest of the structure is in a very deep hole.
- No conclusions can be drawn from the FR24 data that a recovery was being attempted. This aircraft had departed controlled flight suddenly and catastrophically

For me the circumstances are very similar to Metrojet 9268 (EI-ETJ) that crashed in Sinai. When the four corners are established we may well find that the empennage is not located with the main body of the wreckage indicating separation. The list of causal factors for that separation at the end of a stable cruise is a very short list indeed.

Runaway trim is pretty obvious in a 737 as the trim wheel makes a loud clattering noise as it turns. A prolonged turning of the wheel is not something you’d expect in the cruise. And you can stop it by putting your hand on the wheel.

China Eastern 737-800 MU5735 accident
 

China Eastern pilots’ policy is both crews must be seated 20 mins prior to TOD. This was well past that point.



This is an 6 year old aircraft, but from the data available it looks like a catastrophic failure or a midair collision with a military aircraft/drone.

Any Thunderstorm activity or CAT located close to the TOD? haven't seen any Mid / High Level Significant Weather Chart for that area.

The video appears to show the aircraft either side on with no vertical stab, or a fuselage stripped of all flight surfaces.
Either way, RIP to all souls onboard.
Very sad outcome for all concerned.

the thought occurred to me as well, something akin to the GOL mid-air - although the GOL 73 broke up at some point on the way down iirc.

I would think it is (true) airspeed, as transmitted by the transponder. If this was groundspeed, it would not correspond with a dive.

It is Ground Speed indeed. It comes in a packet called Velocity from which Ground Speed and True Track can be derived.
No further speculations on the values, could be a number of reasons.

Last bits that looked like that I examined had a flutter-type event from going deep into the buffet boundary. Shredding laminate is not a very common failure mode.

Sad day.

There had been some convective weather around that area.

Ran video(of showing an object coming down in a vertical arc trajectory) by a friend who put it through a forensic video editing program, he does this for a living.

I am pretty confident seeing the resulting video that the tail section is not present during the length of that video. Wing and engine on 1 side seems present but not sure, other side not visible.

Analysis of FR24 data indicates in last 2 minutes an abrupt extreme vertical descent followed by a steep vertical ascent, followed by a final extreme vertical descent till impact. Continuous loss of altitude with a temporary 'leveloff' after the 'climb', followed by increasing extreme rate of altitude loss.
I am not familiar with the airspeed data from FR24.
Pictures are showing popped rivets in several parts of the fuselage indicating over exceedance of design limits.

Somehow crew lost control of the vertical flightpath; likely loss of tail section, immediate or during the last 2 minutes of flight.
failure of rear bulkhead, collision in the air, runaway trim,we are just guessing right now.

Crew seems to have been fighting to get the aircraft recover from the initial steep abrupt vertical descent. Must have been horrific those last 2 minutes.

May their souls be at peace.

30 year plus airline/instructor Boeing/MD's/Airbus, lot of hand flying

The FO suggests the captain use the washroom before TOD. Captain obliges. Tries to return.

RIP to the crew and pax. Terrible tragedy whatever the explanation is.

I think it depends on whether the FR24 data is derived from ADS-B or Multi-Lateration (MLAT)… GS is calculated simply by time between horizontal fixes…

https://www.flightradar24.com/blog/h...hts-with-mlat/

AA587's loss had a number of factors going on, and for an event in Nov 2001, the term "improper" loses a bit of it's crispness as causation.

• The FO who was PF had experience in aircraft that were able to take aggressive control inputs.
• The Airbus manner of giving a rudder ration limiting effect does not act to reduce the sensitivity of the input, ti does the exact opposite; instead of the Boeing method of reducing the deflection for a given deflection of pedal input to reduce loads, the bus does the opposite, it reduces the pedal deflection available, to reduce the rudder deflection, and that means for a given aero load, the Airbus sensitivity of the rudder pedals becomes more sensitive.
• The certification of the load sequence of the rudder was not required to be accounted for in the loads analysis or in the TIA.
• the sequence of alternating deflections coinciding with high yaw angles resulted in very high bending loads, and the rudder deflection added alternating torsion loads, and they went wild.
• The vertical stabiliser was mounted by a series of pins running longitudinally along the perimeter of the composite stabiliser to fix the primary load paths. The secondary load path was by an internal yoke system.... the load required to shear the attachment lugs results in a lever arm for the structure that exceeds the failure loads for the secondary structure, as the lug/pin failure results in a lever arm acting on the secondary structure.
• AI and TBC both had guidance for jet upset that had crew pre primed (a Gary Klein sort of concept) to wiggle them pedals.
• The awareness of the structure sensitivity to alternating torsion-bnnding was not recognised in Part 25 before the accident.

We know a lot more after that event than we did beforehand. Amazingly, we were recording QAR/DFDR data of cyclical rudder imputs after that disaster, with the penny not dropped.

Sad. Interesting though; that video seems to show the fuselage going supersonic with characteristic vapour puffs. I thought at first it was disintegration debris, but review does not confirm that.

Last three minutes of MU-5735

https://cimg2.ibsrv.net/gimg/pprune....36796dcb33.jpg

The sheets parts looks like they have been torn by high aerodynamic pressure. the loss of the rudder may have led the plane to a major descent.

My first thoughts on seeing the raw video. Some of the debris, not found in the hole, seem to be winglets and other wing or tail parts. Maybe nothing to do with the cause.

https://cimg2.ibsrv.net/gimg/pprune....99cc1c747b.jpg
Right Wing tip.

Let see what the investigating team comes up with.

Ref the wing parts that appear to have separated, I suspect from their state and the descent profile, that the aircraft reached its divergence speed and the outboard wings came off. This would not be the root cause, but an indirect result of an earlier problem that produced the extreme dive angle. It has happened before - a B737-200 flown by Copa (Panama) sometime in the 90s.

Looking at the last data points: GS 376kts VS 30976 ft/min = ~305 kts. Decent angle is arctan(305/376) = 39 degrees. <please check my maths!> Based on the video evidence, it suggests that AOD increased markedly in the very last phase of the decent. Maybe up until this point, the A/C was relatively intact but suffered some significant break up close to the ground.

Your maths is intact. So something on FR24 didn't catch it - which is hardly surprising, the algo's on FR 24 are designed to be as accurate as possible in conventional flight profiles - not this sort of thing.

Somebody made a graphic here; you can see some attempt at recovery, maybe (at 06:22:16 UTC) - or, change of configuration of surfaces (non-pilot initiated, e,g; further structual failure) resulting in change of flight path. Or, again. this could be FR24's Algorithyms's interpolating unusual data and as a result plotting erroneous trajectory.
Obviously, the blue aircraft symbol is not representative of the aircraft's attitude but simply it's position in altitude and even then - only very approximately and exaggerated for illustration purposes.

So a pointless task of guessing. The only thing that will clarify what happened is the boxes.



https://cimg6.ibsrv.net/gimg/pprune....8d4046b0fb.png

The dashcam video Dominic Gates had on his Twitter feed [can't post links] does look more like 45 degrees.
I tend towards seeing a vertical stabilizer - but it's tough.

Weather at crash time, sat pic, lightning
 

Hello, since it was mentioned earlier in the thread:

Doesn't look like anything remarkable, you can click into the district or change time if needed.

Visible sat pic

IR Sat pic

Lightning detection

Any CAT or mountain wave forecast?

300 hPa wind at the time
 

Dear Sir, please can you tell me if any airliner in modern commercial aviation history has been brought down by lightning or CAT?

Looking at FR24 data, did they overshoot their TOD? Did they try and rush the descent and something happened? The investigators will be checking out the maintenance history of this bird. It's a fairly new plane but has it had any repairs recently? What about tail strikes lately?

What do you suppose the loss of a ton of weight (VS) from the extreme aft of the fuselage (long lever arm) would do to the CoG? I would expect a severe and probably uncontrollable nose-over.

https://www.pprune.org/archive/index.php/t-457960.html

What would it do to the integrity of the hydraulic lines running to the rudder? And thus the availability of hydraulic control pressure for elevator operation?

If it counts as modern, a BOAC B707 was brought down by severe CAT in Japan in 1966

BOAC 911

The investigation concluded that the vertical stabiliser came off first.

The Flightradar24 table presented earlier shows a speed of maximum speed of 590kts around the point of the vertical speed reversal. The airspeed was possibly faster. Very fast for an airliner at 8000'. As for the G loading at the point of the vertical speed reversal, my guess is it was probably very great in the positive or even negative direction.

Past cases of such occurences tend to show otherwise. Neither the B-52 which lost 3/4 of its VS nor AA587 nosed over. Problem with lost VS is loss of directional control in Yaw.
Massive nose overs have occured in the past (in Military and General Aviation) as a consequence of losing Horizontal stabilizers. If the FR24 is valid this could potentially happen due to exceeding the structural limit in the first apparent recovery (if the FR24 traces are correct). That said in one of the videos the final trajectory rather looks like 50° ND and possibly with a slight recovery path. The other video on the other hand looks pretty vertical. Simply not possible to safely conclude from the few bits we have. First important information bit to obtain woould be the four corners. After that there is a first chance to get a somewhat clearer picture.

[QUOTE=fdr;11203404]Last bits that looked like that I examined had a flutter-type event from going deep into the buffet boundary. Shredding laminate is not a very common failure mode.


https://cimg2.ibsrv.net/gimg/pprune....e2818dfce.jpeg
more parts of the right wing tip mid leading edge and inboard right logo panel

That is part of the wingtip not the rudder

Re a previous poster's comments about asymmetry, the track was remarkably constant until the 30000ft/min descent rate was well established. This implies the autopilot may have remained engaged for some time. Even an engine failure with autopilot engaged would show some track change.
Edited,
There was also clearly a recovery and even a brief climb so the initial upset event cannot have been totally catastrophic.

That was me and all I was saying was that the more unusual and "in extremis" the flight path - the more FR24 data can be discarded as inaccurate - and in this case with what seems to be the case, absolutely discarded. We know it dived, that's all we know. FR24 cannot reveal anything more at this point. Anyone chasing FR24 for facts about causal factors is chasing windmills.

Unfortunately PILOT SUICIDE is the most likely reason for this crash.

Rudder hardover, Stab trim runaway, mishandling a depressurisation WILL NOT result in a descent profile like this. I’ve tried it in the sim and the descent profiles are NOT near vertical like this.

SILKAIR flight MI 185 is a very similar example to compare. Deliberate, sustained pilot input with stab trim at full manual nose down and a high thrust setting was determined as the cause of this crash by the NTSB as an addendum to the final report.

Let's let facts and Data rule - not idle speculation
 

- Profound sympathy to the crew, pax and their families.
- More data is needed to even start any credible assessment. Like the distribution of remaining identifiable parts, any ATS communication, preliminary failure mode of identifiable pieces, etc.
- an ultimate pitch tuck can have many many root causes
- much of the above speculation is both technically way off-base, and very unfair to the families of the lost flight crew, pax, ....and even to the airline at this point.
- let's hope a credible internationally staffed accident team can be fielded ASAP, and the results honestly, credibly, and quickly reported.
- There are just too many NGs we're flying out there, to not have credible bounds being placed on the inaccurate speculation now being distributed... ASAP.
- Let's hope for usable DFDR, QAR, and CVR data... but in accidents like this, that can be long shot.
​​​​​​- Profound sympathy to the crew, pax and their families.

I'm a tad suspicious of the video. It looks very much like one of a crashing Long March booster from a couple of years ago.
https://cdn.images.express.co.uk/img...=1599550692168

In trying to ascertain what speed it was falling at, does anyone know what ther terminal velocity would have been - and how long to reach that from the start of the fall? The media are putting all sorts of numbers on the vertical speed, doubtless some from here.