Page Links: First Previous 1 2 3 4 5 6 Last Index Page
violator
2025-06-21T15:42:00 permalink Post: 11907854 |
Resubmitting following some Mod Feedback and a significant re-write. Yes, it is speculative
I have a theory that I'd like to share. It brings together various pieces of known information, along with 30+ years of my experience as an aircraft engineer that forms a plausible (IMO) explanation of what may have happened. We Know - From the Video's and the ADSB Data: That up to and for the first few seconds after take-off appears relatively normal. The AC appears to lose thrust without e.g. birdstrike or other spectacular smoke /fire producing event. That the RAT deployed. That the pilot reported 'Thrust not achieved' We can see that the AC had a relatively busy schedule in the few days prior to the accident flight, so there was no significant downtime for maintenance activities that could have caused incident. The AC flew DEL-CDG on 11 Jun with quite a racy turnaround in CDG of 1h12m. The centre tank would have been empty at CDG on arrival, and would have been partially filled for the return CDG-DEL. CDG-DEL Arrived 01:47 am IST. Again the Centre Tank would have been empty. But quite a bit of fuel in the wings. 8 Hrs later, at 09:48 am IST the AC departed DEL-AMD. For such a short-hop, Fuel upload would have been minimal, merely a 'topping up' if at all. Certainly nothing into the Centre Tank. DEL That night was fairly hot and humid - 57% at 02:30, 54% at 05:30, 44% at 08:30. That wing tank fuel could have picked up a fair amount of water. The flight DEL-AMD would have only used the wing pumps. Thus any water in that 'overnight' fuel would have been vigorously stirred and evenly suspended. At concentrations that would cause no ill-effect at all. The AC was on the ground at AMD for 2 Hrs, from 11:17am to 1:17 pm IST. The AC would have re-fuelled, first filling up the wing tanks to the top, then filling the centre-tank to whatever quantity necessary. There was enough time for water in the wing tanks to settle out. The B787 Fuel system has pumps in the wing tanks, and pumps in the centre tanks. The Centre Tank pumps are also known as 'override' pumps because they output a higher pressure than the wing tank pumps, thus ensuring that with all pumps running, the centre tank fuel is used first. Should the centre tank pumps stop, due to either filure or running out of fuel to pump, the wing tank pumps then produce the pressure. In the event that all pumps stop running (e.g. an electrical failure), the engines will suck the fuel from the wing tanks. The 'sucked' fuel comes from a dedicated pipe in each tank through the 'Suction Feed Check Valve' (so that pumped fuel doesn't just exit through the suction tube). The suction tube pickup is in a slightly different position to the wing pump pickups. It is conceivable to me that the suction tube pickup could have been immersed in water, settled out from the fuel in the wing tanks. Then - at start-up of the aircraft in AMD, The engines would have been supplied with fuel from the centre tank. Fresh Fuel. All OK. Wing pumps running and doing not much. But, I speculate, the suction pick-ups immersed in water. Waiting. Start up and taxi out was all normal. Runway acceleration up to v1 appears normal. V1 - Rotate - (positive rate - Gear up? - Not my debate). But somewhere around that time, I speculate that a significant electrical failure occurred. Enough for the RAT to deploy. Enough for the fuel pumps to stop. I'll not speculate on the cause. We know that it can occur, that's why the RAT was designed to operate. The engines at that point were at TOGA thrust. In a significant electrical failure, the engines will keep on doing what they were last told. Keep that thrust stable. So the AC climbed for a few seconds more. The pilots did what they were trained to do for a power failure, manage that, thankfully the engines were still going well... But there was only so much 'good' fuel in the lines. The engines sucking fuel themselves, the fuel would now be coming from the suction pickups, a different supply. A supply likely heavily water contaminated. It would take a few seconds for that contaminated fuel to actually reach the engines, but when that contaminated fuel hit, Thrust would have been significantly reduced. The EEC's would have been doing their best to maintain the thrust, firewalling the throttles would probably have little effect at that exact moment. The engines would have likely worked through that bad fuel in a shortish period of time, but a period of time that our crew did not have. A fully loaded aircraft producing less than take-off thrust, is not sustaining enough thrust for continued flight. The rest - is down to the skill of the crew in deciding exactly where to hit the ground within the very narrow range of choice they had. |
lighttwin2
2025-06-21T15:46:00 permalink Post: 11907858 |
TCMA continues to be one of the few (very unlikely) causes of/contributors to simultaneous shutdown of both engines. So far, though, I don't think we've seen a credible scenario explaining the possibility that TCMA was triggered in this accident. I'm not sure I understand your speculation.
In the scenario you are considering, it's clear that the air/ground state would be wrongly "understood" by the TCMA function. But we don't have, AFAIK , a credible theory for how that might happen. Surely it would have to result from either incorrect signals from the relevant sensors or a failure of the related logic in the FADEC TCMA function, or a combination of those. Indeed, I don't think we yet know exactly which sensor readings that logic depends on or how those readings are fed to the FADEC. Does your speculation include any thoughts about this? Also, the FADEC TCMA function has to "believe" that the engine is operating at high power and not responding to thrust lever operation. In your proposed scenario, is this also a logic failure — in both FADECs? Or false inputs from both TLs? Or are both engines actually operating at higher than commanded power levels? Or do I misunderstand your post?
Q: Would the a/c have enough kinetic energy a 184kts to climb to 100-150ft agl and then reach its final position if the engines had failed at, or just, before rotation? A: Theoretically possible - see calculation here . NB, the a/c actually flew 1.5km from the end of the runway and 2.3km from the likely point of rotation. Q: Doesn't the forward position of the gear mean that power failed after the pilots had selected gear up? A: Inconclusive - had hydraulic power had been lost prior to rotation, the gear could also be in this position - explanation here Q: If the throttle levers were brought to idle during take-off, would the A/C have applied autobrake, reversers and speedbrake? A: Yes, although there is a built in delay before reverser and speedbrake actually deploy - see here . Q: Is the ADS-B data consistent with this scenario? A: Yes, e.g. the Flightradar data shows the aircraft decelerating rapidly (12 knots in 4.2 seconds) from close to rotation. However, it's not clear how accurate this data is. For one, the altitude data is +/- 25 feet, second, while I was under the impression FR would have received airspeed data from the a/c sensors, this post suggests maybe not. Q: Does TCMA activation require the thrust levers to be at idle or does it function when the thrust levels are above idle, but where the actual thrust is above that commanded? A: No, the latter is true (i.e. idle is not required) - confirmed here - there are of course many protections against false activation Q. Did AI171 have the same software version / logic paths as NH-985 A. Unknown. That a/c had Trent 1000s so to some extent the software is different, but we understand the TCMA logic is broadly the same regardless of engine. I have not seen a post clarifying whether the TCMA software has been updated /changed via SB since 2019 to account for this incident. Be grateful if posters could refrain from speculative responses "e.g. I think this is unlikely because I feel x". I am not opining on how likely this sequence of events is, simply trying to summarise whether or not this theory has been ruled in or out. I also recommend this post for a summary to read before posting. . Last edited by lighttwin2; 21st Jun 2025 at 16:13 . |
AAKEE
2025-06-22T07:08:00 permalink Post: 11908310 |
The gear tilt position is not definitive evidence crew had selected gear up. I've speculated another cause for this non-normal gear tilt is that C hydraulics failed around time of rotation. This would explain the gear remaining in the forward tilt position. There are reasons why the crew may have not selected gear up,
see earlier post.
Therefore we cannot determine wow or air/ground logic from an assumed gear retraction.
Another point pointing to that the aircraft did consider itself being \x94In Air\x94 is the ADS-B data sending Altitude from the first 575 feet at 08:08:46.55 until at least 08:50.87\x85? I would think the sub systems like TCMA would use the same In Air / On Ground logic as the aircraft normally use? I come from an FBW aircraft with a Air/Ground logic that seems rather bullet proof and would guess the 787 wouldn\x92t use a less solid logic which probably, in doubt would consider it being \x94In Air\x94? It would be \x94logic\x94 for the TCMA to use this logic? 5 users liked this post. |
Icarus2001
2025-06-22T07:13:00 permalink Post: 11908312 |
Another point pointing to that the aircraft did consider itself being ”In Air” is the ADS-B data sending Altitude from the first 575 feet
|
Musician
2025-06-22T07:33:00 permalink Post: 11908325 |
https://mode-s.org/1090mhz/content/ads-b/1-basics.html : ![]() When the aircraft is on the ground (or on the water), it does not transmit altitude. 1 user liked this post. |
Innaflap
2025-06-28T14:25:00 permalink Post: 11912515 |
I suspect both recorders will contain the same data. Given the radio transmission after the loss of thrust the aircraft still had at least the emergency electrical bus powered. This should have kept both recorders online. It is however possible given the 10 minute battery backup that Boeing chose to put the the recorders on another bus but that\x92s not the norm.
Any device subscribing to the AU will be getting the same information. This would include the DFDRs and whatever is used for ADS-B Happy to be corrected |
jdaley
2025-06-29T13:48:00 permalink Post: 11913045 |
I got the idea that with no (or very little) thrust, and with the aircraft falling, the pilot (may have) realized that he was in
out of control flight
, and falling.
In a pedantic sense: if you make control inputs, and the aircraft won't or can't respond to them, you are in out of control flight . ![]() computed bearings from ADS-B positions Given the ADS-B height figures are reported to have a granularity of 25' it's compatible with the video to assume a height around 90' at the last ADS-B data point. A simple ballistic calculation - speed 88m/s (171kt), 11degrees, 30m height, results in a maximum height of 145' - given the impact of aerodynamics the observed height in the video isn't in dramatic disagreement - ie it's possible nothing interrupted the flight path after rotation. Given the clearer land 100' to the left of the flight line the "in out of controlled flight" is sadly very feasible. |
jdaley
2025-06-30T14:01:00 permalink Post: 11913650 |
1 user liked this post. |
NSEU
2025-06-30T14:10:00 permalink Post: 11913661 |
ADS-B stops 330' from the end of the runway, well before top of climb. A previous poster stated that ADS-B granularity on the 787 was 25'. As the aircraft obviously climbed higher than 71', it's reasonable to assume that ADS-B was about to transmit 96' when the electrics failed. Looking at the CCTV, subjectively, the aircraft reaches the height of a wing about 7s after rotate and top of climb about 7s later.
|
nachtmusak
2025-06-30T15:20:00 permalink Post: 11913688 |
Just read a report by Richard Godfrey on the climb and descent
https://www.dropbox.com/scl/fi/a9hhz...ioijg&e=1&dl=0 Not knowledgeable to know about a lot of this myself but interested in views. His calculation asserts that it was not a dual engine flameout to cover the distance it did but there must have been at least some thrust provided to do so. The determination of the height AGL seems quite incorrect - it completely ignores the local pressure and temperature, which very much need to be corrected for. Applying those corrections even using rough rules of thumb (~30 ft per mbar, ~4 feet per \xb0C per 1000 ft) gives a figure that's around ~100 feet AGL. The follow-up sanity check also fails even without knowing the correct math because if you match the ADS-B data (timestamp + location) to when & where the aircraft lifts off and starts to climb in the CCTV video (versus just assuming that the peak height seen in the video matches the last ADS-B data point), the aircraft is very much not 300 feet above the ground when its transponder reports an altitude of 625 feet. Also the estimation of the glide ratio with flaps 5, gear and the RAT deployed being 3.5 to 1 seemed incredibly low to me, but I'm the first to admit that I have nowhere near as much knowledge of gliding performance as I do of weather math. So I looked up the closest incident I could think of: Air Transat 236, an A330-200 (so of a pretty similar shape, wingspan and wing area as a 787). According to the final report (link: https://www.fss.aero/accident-report...1-08-24-PT.pdf ), the aircraft arrived at a fix approximately 8 nautical miles (48609 ft) from the runway at an altitude of approximately 13000 ft, at which point the crew decided to execute a 360 turn to lose altitude as well as to extend the slats and landing gear during the turn (the RAT of course had already long been deployed at this point), both to prepare for landing and to help further lose altitude. Sure there's some rounding here, and my understanding is that "flaps 1" on Airbus aircraft deploys only leading edge and not trailing edge devices, but this already suggests that their expected glide ratio was significantly higher than the raw 48609:13000 ratio (~3.74:1). They re-established themselves on final in their landing configuration at an altitude of ~8000 feet and a distance of 9 nautical miles (54685 ft), so let's say that the true distance was somewhere between 8 NM and 9 NM to account for rounding. That gives a glide ratio of between ~6:1 and ~6.83:1. But on top of that, the crew still had to execute a series of S-turns to lose enough altitude to actually make the runway, so their "dirty" glide ratio must have been even higher than that. Unfortunately I don't think it's possible to determine conclusively what the ratio was since we don't know how many track miles were added on by the turns (the flight data recorder stopped when the engines flamed out and human testimony only goes so far), and again they did not have flaps extended, but I think it's fair to say that a glide ratio of 3.5:1 is a wildly low estimate for an airliner of the 787's calibre even with the gear down. Sorry if this is off-topic or too much rambling, but considering how much speculation there tends to be both in this thread and elsewhere about real-world glide performance (especially in non-ideal configurations), hopefully these details are helpful. 10 users liked this post. |
MikeSnow
2025-06-30T16:34:00 permalink Post: 11913755 |
Just read a report by Richard Godfrey on the climb and descent
https://www.dropbox.com/scl/fi/a9hhz...ioijg&e=1&dl=0 Not knowledgeable to know about a lot of this myself but interested in views. His calculation asserts that it was not a dual engine flameout to cover the distance it did but there must have been at least some thrust provided to do so. Richard Godfrey first 7 points seem accurate as far as I can tell, but he makes a lot of mistakes at point 8, when he computes the altitude the aircraft reached. The first error is when he said "At liftoff from 180 feet AMSL, barometric altimeter would read: Pressure (ISA) = 1006.68 hPa." While that is correct, that only applies when the QNH is 1013.25. But the QNH was 1001.1, so the altimeter would actually read around 994.6 hPa. He then said: "Aircraft later reported: Pressure = 990.573 hPa (corresponds to 625 feet in ISA conditions).". That is correct. Then he subtracts the two numbers, 1006.68 hPa and 990.573 hPa and gets a 16.107 hPa difference, which is wrong, since the field level pressure he uses is wrong, as mentioned earlier. The correct number would be 994.6 hPa - 990.573 hPa = 4.027 hPa difference. Using the approximation of 30 ft/hPA, gives us an actual around 120 ft altitude over the field (+/-12.5ft since the aircraft reported altitude has a 25ft granularity). Instead, he gets a much higher altitude of 480ft over the field and, as a further mistake, he misinterprets it as altitude over sea level, so he subtracts 180ft (the field's altitude). This second error bring the value closer to reality, around 300ft, but still a long way from the correct value which would be around 120ft. I also did a quicker back of the envelope calculation with a different method, and got 80ft +/- 12.5ft, even further from his 300ft. Then, he makes a third mistake, he assumes that the last altitude reported by the aircraft is the peak altitude it reached, when in fact it is almost certain the aircraft kept climbing for a while after the last ADS-B datapoint (the last datapoint was reported when the aircraft was still above the runway). So it might have indeed reached 200ft or more after all. Finally, at point 9, he makes an estimation based on the ratio between the wing span and the altitude in the CCTV video, compared to the actual wing span. That's a good idea in theory, but you need the camera to be exactly behind the aircraft and very far away not to be affected by perspective significantly. For example, if you film the aircraft from a 45 degrees angle, the apparent wingspan would be reduced by about 30%. I didn't fact check points 10 to 19, since most of them relied on incorrect numbers computed earlier. 12 users liked this post. |
fdr
2025-06-30T23:59:00 permalink Post: 11913958 |
For the past 30 years, we have collectively bagged Boeing in increasing measure, for doing what all corporations are obliged to do; the board is answerable to the investors, and the investors have a "10-second Tom" event horizon when it comes to their earnings. It is little wonder that the race to the bottom is being won by Boeing, it is what Wall Street demands, and in the end, after the bones of the OEM are picked clean by those that do such things on the carcasses of once great US engineering giants, then they will bitch and moan about the losses to shareholders while they pocket the profit of their profession, fleecing the investors. L-M has been better in managing the milking of the public purse, and maybe that is the way it should be. 6 users liked this post. |
FullWings
2025-07-01T06:45:00 permalink Post: 11914044 |
Perhaps a result of being too dense, in these threads I have not understood whatsoever the discussions on L/D, best glide, AOA, stall speed, angles, whatever, as being relevant to this flight. I assume that the pilot flying was flying, i.e. stick and rudder. I give him/her the benefit of the doubt on account of being a pilot. Professional or not.
The real question is why this happened when engines and their associated systems are, by design and regulation, as independent as possible. The top runners at the moment are (in no particular order): pilot action, simultaneous hardware/software malfunction and massive fuel contamination. They are all very unlikely (and cogent technical arguments can be made against each of them) but so is the event that followed. |
nachtmusak
2025-07-01T12:06:00 permalink Post: 11914222 |
This is my latest attempt to square the circle using all the data points and minimal assumptions. The main shortcoming of the analysis is not knowing the maximum L/D and the speed for maximum LD with the gear down, flaps 5, and the RAT extended. However, if I use a reasonable number in my opinion for the L/D in that configuration and assume that the airplane is being flown at the speed for it, it will not get to the crash site. The distance from the runway of the crash site is from a previous graphic (1.55 km); the rotation point from fdr, permalink 314; 200 feet max height above the runway being generally accepted; crash site 50 feet below the runway elevation cited previously. An average speed of 180 knots is consistent with the dimensions given and 30 seconds flight time. A flare at 50 feet will briefly increase the L/D to 20, maybe even 30 (500 feet more than shown) but still not enough to make up the shortfall, In fact, with a head wind the L/D will be lower than assumed as well as if the speed being flown is higher or lower than required for maximum L/D in that configuration. In other words, there must have been some thrust available.
![]() As the aircraft visibly continues to climb past that height (and for a longer period than ADS-B data covers, if the camera's perspective casts doubt on that), it seems rather clear to me that it reached its peak height past the end of the runway. In light of this I find the fact that people keep calculating a glide from the runway to the crash site to be a bit strange. Wouldn't the first step of any math be to try to determine where it started descending? |
jdaley
2025-07-01T12:38:00 permalink Post: 11914245 |
Subjectively the aircraft continued climbing for ~7s after the last ADS-B value, achieving around 200'. 7s at say 160Kt is 574m. The last ADS-B point was 1.6km from the hostel so the descent was around ~1km from ~200' an 18:1 ratio. |
Tailspin Turtle
2025-07-01T13:05:00 permalink Post: 11914261 |
There is easily-correctable available data with the aircraft's altitude at pretty much the end of the runway and it is not at 200 feet (it's around 100\xb112.5 feet).
As the aircraft visibly continues to climb past that height (and for a longer period than ADS-B data covers, if the camera's perspective casts doubt on that), it seems rather clear to me that it reached its peak height past the end of the runway. In light of this I find the fact that people keep calculating a glide from the runway to the crash site to be a bit strange. Wouldn't the first step of any math be to try to determine where it started descending? |