Posts about: "CCTV" [Posts: 84 Pages: 5]

Incidental, but I wonder the camera operator was actually trying to capture. Did the aircraft fly into his shot or had something occurred which made him, (sounds like a him), grab his device and start recording? The footage from X is not CCTV, its a hand held device, probably a phone, and not professional. Was capturing the final moments of this unfortunate aircraft a coincidence?

He may have noticed that the plane was flying lower than usual. Anyone living near the airport would be well aware of the normal flight path.

Although the CCTV video is of unfortunately low resolution the trailing edge discontinuity between flaps and aileron can just about be seen on the port wing at about 3/4 span, just as it can in this video at around 54 seconds in:


Looking at the original video again the equivalent discontinuity is just about perceivable on both wings, so I was wrong at first and the flaps were not retracted. What appears to be the RAT quite possibly is; at that angle the nosewheel is completely hidden by the fuselage.

in case you suffer double engine flame out or failure, what kind of rudder compensation do you expect? to me, watching the CCTV video, it was only wind correction slightly to the right

I am not a pilot so apologies if this is a basic question.

Under what conditions will the RAT deploy? Is this automated or does the PF / PM need to hit a switch? <-- [Edit] I see this was largely answered in post 369. Thanks!

Watching the CCTV from the airport, it doesn't look like there was much time to do anything. I'm surprised someone called a Mayday (if that's true). And even to me, it seems pretty clear that bird never had enough power to go much of anywhere except where it did.

All so very tragic.

Your profile says this.
How familiar are you with the Boeing 787 flight deck and its controls?

A link to the report Zoot0 referred to:
https://ad.easa.europa.eu/blob/AIR-2..._AIR-22-09R1_1

Excerpt in the spoiler


{there is more, this is just a taste, see the whole report for details}.

Would any of our 787 qualified pilots care to comment on this as it relates, or doesn't relate, to the aircraft's brief flight shown on the CCTV video?

For those still under the impression that the takeoff or rotation was at the runway end... In the video of the CCTV camera you can see rotation happens basically behind the small structure in the foreground. Unless I am severely mistaken... Knowing the approximate rotation of the camera, one can simple draw a line until it intersects with both the structure and the runway:



Yes, it is just an approximation, but I'd say they still had a decent chunk of runway left, and not in any way close to taking out antennas on the opposite end.

So... likely not at MTOW, not using the full runway, normal rotation, followed by videos of it with RAT deployed which should only happen upon loss of all electrical/hydraulical sources, which should only really happen on loss of both engines without APU running, combined with barely (if any) sound produced by the engines in said videos... It leads me to form some sort of image of what happened. Could still be completely wrong.

I have done a RAT deployment test flight on a previous type, and even though I was prepared for it, it was still a bit of a shock to see the cockpit go mostly blank before reconfiguring itself. I don't know how quickly the 787 would reconfigure, if there is any downtime at all, but that would not have been a thing you'd expect at 100-200ft AGL. I do not know how many of us would quickly overcome the startle.

From the airport CCTV video it looks to me like a normal takeoff and start of climb, until suddenly there's a loss of climb performance with no obvious upset at that point. From the picture of the wing post crash it looks like the flaps were still deployed (N.B. based on pre-accident photos that's the right wing so closest to the camera is aileron and flaps are further away, damage had me confused first time), so I'm going with loss of thrust rather than flap retraction.

From the videos from bystanders it looks like RAT deployment (both sound and zoomed in pictures) rather than thrust lever retard, and that would also explain failure to retract gear - if you are dealing with both engines out at that altitude then gear isn't your first thought. From the airport CCTV video I don't see anything that looks like bird strikes at that point in the climb i.e. no obvious flocks of birds, no smoke out of the engines, no slewing one way as one engine fails and then the other is cut by accident - plus you wouldn't cut the engine at that point, you'd climb on one engine and then sort it out.

Even fuel contamination or water build up in both tanks is likely to result in one engine failing a few seconds before the other. So I can't come up with anything other than both fuel cutoff switches that would result in loss of thrust and RAT deployment. Looking at a picture of the cutoff switches https://www.nycaviation.com/2013/08/...is-fired/30179 I don't see how they get hit by accident.

I'm confused, hope we get an FDR / CVR readout soon.

184kts - 21'AGL
179kts - 46'AGL
177kts - 46'AGL
177kts - 71'AGL
174kts - 71'AGL
172kts - 71'AGL

I can't post the image of the FR24 data, but the granular data points show a 14kt loss of speed from 21' to 71'. With all engines operating, or even with one engine failed, the speed would be increasing or at least remain constant through this period.

Additionally, the CCTV footage from beside the runway shows a ballistic flight trajectory from liftoff to impact.

184kts is a very high rotate speed (a common OPT output) and the aircraft has a lot of excess energy at that point. With Flaps 5 and no thrust it could trade airspeed for altitude at approx 14kts per 50' until about 250'AGL before running out of puff.

Again, that is consistent with the video.

It looks to me like it lost all thrust at some point in the rotation.

Such a terrible shame, condolences to all. It looks inexplicable from the CCTV.

Seems time for a visual evidence review.

There seems to be a RAT theory based on a hyper zoomed artifact and someone showing a RAT deployed on a different airframe. Not convinced about that, you might get a similar artifact from a belly antenna. The noise? CCTV doesn\x92t have noise and the other pictures I saw were from a car in traffic.

Others are are saying it climbed to 500\x92, not sure about that, the highest I have seen visually is less than 300\x92, QNH vs. QFE I suspect.

Flaps vs. Gear definitely a possibility and the AoA was increasing particularly after the descent started.

Double EF (If RAT deployment not a red herring) Fuel contamination? Would have to be deliberate as no other aircraft affected, unlikely. Maintenance or crew error, possible unlikely. Bird strike, no evidence.

MTOW error possible but it seemed to take off fine so no reason for the return to the ground.

What about the bang the survivor heard? I suspect you can treat the evidence of anyone involved in an air crash with a pinch of salt. Order of events are often out of sequence even when talking to trained observers in less stressful situations

MCP mis-setting to 100\x92. Engage AP early, often seen, thrust immediately commands to idle by ATHR, starts to sink, extreme startle and forget gear because it appears like a double EF.

I know where my money is but only time will tell, if they get the Black Boxes in good condition, the factual statement should clear it up quickly.

Your MCP theory sounds feasible but doesn\x92t stand up to Boeing VNAV engagement logic (assuming it\x92s typical Boeing) which would mean at 400\x92 the aircraft would still try to maintain V2. It just might pitch the wrong way. Interestingly, the AT system has caused more accidents on the Boeing than the Bus. Much to everyone\x92s surprise considering the TL\x92s don\x92t move on the Bus..

Most likely anything between 256 and 1024wps, individual parameters will vary with regards sample rate. Not au-fait with 787 but other Boeing aircraft may have dataframes relative to the original owner requirements (though this is unlikely to vary much given my experience with Boeing aircraft). I\x92m not sure if GE have a real time EHM feed like RR? Anyhow, and not wanting to speculate too much, if there was anything engine related it would certainly show up on the DFDR and GE could possibly even know that by now if they are equipped with real time health monitoring. In my opinion, for what it\x92s worth, the CCTV footage is consistent with flap retraction instead of gear, something that would also be easily identifiable upon review of the FDR which will be reported upon in due course.

Ive seen it happen in a previous (ground based) life fairly frequently (in relative terms) but exclusively during descent/approach. Usually the context would be expecting a flap call but when the gear call is made the flap handle is moved and quickly returned to the previous gated position followed immediately by gear selection, confirmed upon review of the data. To happen on takeoff would be most unusual, and although the CCTV footage is indicative of it being highly plausible, neither pieces of footage really show anything to draw any conclusions on. Certainly there may be numerous aspects to this, perhaps there was some catastrophic electrical failure that wasn\x92t of itself an immediate threat to the flight but the startle effect in the flight deck prompted other reactions. It\x92s impossible to know based off the information available. The human element will form a significant part of the investigation regardless of root cause(s).

Regarding the comments about who should be allowed to comment on such threads. I like the fact the mods have taken a pragmatic view of this. Theres a reason many in the accident investigation field come from an engineering background. Whilst highly experienced pilots are clearly crucial on an SME level, there can be a tendency for iwouldnthavedoneitthatway-itis. This can at times hinder investigations where an open mind free from confirmation bias is essential.

I'm pretty convinced that everyone is watching different videos, which wouldn't surprise me in the least given how much images and footage can get mangled from being downloaded/re-uploaded to different services. Personally I've seen at least two different videos (and slightly different variants of each): one that's a direct recording of the aircraft and one that's a recording of a phone playing a recording of the aircraft. The buzz of a propeller was quite audible in the former but not in the latter, and the engines are notably quiet compared to any 787 takeoff I've ever been close to.

I do think that at this point people are working backwards from the cause being retracting the flaps, rather than working forwards from the evidence towards an explanation. If everything else was as it should be apart from flaps being retracted instead of gear, you'd at the very least expect the aircraft to reach a higher speed than is suggested by how little distance there is between the end of the runway and the crash site. Coupled with the apparently parabolic trajectory seen in the airport CCTV video, the crew (allegedly) reporting engine problems on the radio, the surviving passenger testifying about flickering cabin lights, the plausible deployment of the RAT, the fact that the ADS-B data cuts out well before impact...surely at some point one must consider a lack of thrust, no matter how improbable. Recognising the what doesn't need a full thesis on the how .

I wonder if (given all the facts and rumours about the situation so far) the flaps would be so high on everyone's minds if they weren't already a hot topic from the initial, largely baseless speculation that they somehow took off with flaps retracted.

Would it be reasonable to expect multiple CCTV videos, cf Jeju at Muan? I was surprised by how many videos became available after that sad crash.

I attempted a rudimentary timeline analysis of the two most prominent videos. Folks on here are usually far quicker & more adept than me at producing this type of analysis, so approach the following with due scepticism!

Please verify the videos using the links below, so we know we are all talking about the same thing. I deliberately avoid any judgment regarding theories posted thus far - I am merely supplying information against which you can further test those theories.

1. Primary eye witness video with audio:

[ X link ending 1933089931347345596 in case the hyperlink itself doesn't work for any reason]

- Footage starts with aircraft directly overhead, give or take, (based on both sound & vision)
- 13 seconds from start until fireball clearly visible above roofline

2. Airport CCTV:

[X link ending 1933162059556159903 ]

- 49-50 seconds until impact and fireball clearly visible

Subtract the 13 secs noted in video #1 and this establishes the approximate moment the aircraft passes over the video witness's position when viewed from video #2 (CCTV):

49 - 13 = 36 secs into video 2

Estimated timeline (CCTV #2):
19s: rotate
31s: climb rate noticeably deteriorates (12s after TO / 18s from impact)
36s: estimated moment aircraft overflies eyewitness camera (17s after TO / 13s from impact)
38-40s: pitch up then descent begins (19-21s from TO / 9-11s from impact)
49s: first contact with ground (30s after TO / approx impact)
50s: fireball visible above rooftops

Conclusions from combining both videos:
- Aircraft overflies eyewitness camera roughly 13 secs before impact
- This is well after it has stopped climbing (~6 secs)
- This is also mere seconds before CCTV shows the aircraft pitch up and start to descend (which I believe I can substantiate in the eye witness video, although this is tougher to confirm with the naked eye from this viewing angle - someone may have to look more closely than I can this evening!)
- I have previously asked questions about the audio in video #1 and whether we can draw any conclusions regarding RAT deployment & engine thrust. There are strong opinions on both sides. Your personal view on this will influence your evaluation of what the videos show, but either way your theory needs to fit the timeline (or advance a different one).

So, anyone who is developing a theory ought to consider whether it is consistent with the following:-
- rate of climb decays to 0ft/m within 10-12 secs of rotation
- RAT (possibly) audible within 17s of rotation *
- pitch up shortly afterwards with no discernible increase in engine noise & unable to arrest descent
- impact with ground within 30 secs of rotation

* I recognise that the RAT deployment is not an established fact, but any theory that proposes RAT deployment needs to take into account this timeline in addition to the rest.

Alternatively, you are welcome to refute this simplistic, late night analysis of the limited video evidence.

Miscellaneous comments:
I have seen in manuals for other airliners that because the bogie tilt is by a hydraulic actuator, gravity deploying the gear means the gear doesn't tilt to landing position.

With the loss of centre-system pressure*, would you expect the bogies to tilt naturally? I.e. spring pressure holds the gear in the stowed tilt, a hydraulic cylinder pushes the gear to the landing tilt. No pressure means the gear returns to the 'stowed' tilt.

The tilt actuator is designed to be overridden when the bogie hits the ground, so perhaps it has some kind of intentional bypass and doesn't stay in place without continually applied hydraulic pressure.

If so, that would also point towards total loss of electrics and no attempt to raise the gear.

* 787 centre system is powered by two electric pumps, plus the RAT. The RAT hydraulic pump only powers flight controls, not the landing gear.


Electric loss:
Surely even total AC power loss shouldn't result in engine loss, even if the RAT doesn't come online. The FADECs have their own alternators, bare minimum flight control computers and actuators are available on battery (though probably result in some equivalent of Direct Law), and boost pumps are unnecessary at low altitude. Left/right EDPs will remain active if the engines are running at any serious speed; providing flight controls.

Poor crew reaction to ending up in direct law is possible but it's hard to see the electrical issues as a cause, not a symptom.

Anti-terrorism squad looks to be doing an excellent job (/s) and recovered a convenience-store-grade CCTV recorder, probably from somewhere on the campus it crashed into:
https://economictimes.indiatimes.com.../121823103.cms

The point you think you are making here has recurred throughout the thread. To try and put it to rest:

The first omission/error here is in describing a definite article (i.e. the video), alongside claiming that the pic you posted is 'unprocessed". It might not be processed, it just isn't very good quality. Which you might not realise if you haven't seen the original, higher resolution version of the clip.

The are now numerous social media sources for copies of the flyby/crash smartphone video. Many of them are actually a repost (or possibly multiple independently made iterations) of a second generation recording - made via a smartphone filming the original video playing on a monitor and via the monitor's (likely very poor) speakers. This version has been incorrectly assumed to be (and described as derived from) "CCTV" in some posts. The "screen recorded" nature of this particular version of the video is obvious from the camera movement (including showing the bezel and edge of the screen it is filming), moire patterns etc. The original (or rather, what *appears* to be a first generation, but compressed, copy of the original) version of the video has *also* been posted on social media (and thereafter, here) with both then being dissemanated across multiple social media accounts and at varying video resolution and compression ratios.

There are further posts of the "screen recorded" copy of the video in which AI enhancement and other filters have been used to try and "improve" the video quality. Anyone even attempting to filter the video in that way doesn't understand the way such filters operate (i.e. they are trying to make the picture "more watchable" rather than sharpen it/make it more *accurate*), and anyone trying to draw conclusions from such "enhanced" versions needs to think carefully about what they are looking at.

The framegrab you have posted above is clearly from the "screen recorded" video. The original version (with its attendant much higher video *and* audio quality) *does* appear to show an object beneath the airframe where the RAT hub would be - and also - albeit only for a handful of frames - appears to show the motion blurred impeller disc. The fact that this artifact is only visible for a few frames is explicable via the heavily compressed digital video but it is consistent.

The "full" video itself is also longer than the "screen recorded" version and starts earlier in the incident timeline. And a distinctive propeller-type beat *is* audible, both before *and* after the aircraft comes into frame, at near identical pitch to the various examples we've seen of a 787 passing a camera with the turbine deployed. One would imagine that the first generation copy of the video as it exists on the device that recorded it is noticeably better quality than *any* of the downsampled/overcompressed social media versions we have seen. And one also sincerely hopes that the Indian AAIB are already in possession of it.

The prevalence of smartphones and social media means this sort of footage is more immediately accessible and more easily dissemanated than at any time in the past. However, anyone trying to draw solid conclusions from a clip posted on social media - especiallly with limited quality, and even more, "enhancement", needs to remember the limitations of what they're seeing.

(Not a pilot, but a former aero electronics engineer with significant subsequent digital forensic experience)

IMO

In the CCTV video, the aircraft stops climbing at 00:28. 3 seconds after, it starts visibly descending. At peak altitude, using the 197ft wingspan as a measure, the altitude is around 200ft or below. The fireball is at 00:48, 17s after descent starts visibly.

Per google maps and the impact location mapped at avherald, the impact point is ~3990ft from the airport boundary road and about 4200ft from the midpoint of the runway threshold and the airport boundary road.

16:1 to 25:1 is what I could find for the 787 glide ratio range (unpowered) with main landing gear down and flaps 5. So the aircraft could cover 16 to 25 ft horizontally for every 1 ft of descent.

With a starting altitude of 200ft, that would imply it could have covered 3200ft.to 5000ft during unpowered descent.

The actual distance covered, around 4000ft, certainly seems to suggest that the descent was unpowered.

Thanks for those numbers. Note, however, the max 787 L/D, flaps and gear up , is reportedly 21 (best in class), which casts shade on the credibility of those L/Ds, certainly 25 with the gear down. What is your source? IF the actual L/D with the gear down and flaps 5 is somewhat lower than 16 and given that starting height and distance traveled are correct\x97and considering that the aircraft might not have been at the speed for max L/D with the gear down and the possibility of a headwind\x97then there must have been some thrust during the descent (it did look to me to be somewhat flatter in that video taken from one side than I would have expected with the gear down, a relatively slow initial speed, and no thrust) or the engines were spooling back up at the end.