This is probably the worst way a plane could go down in terms of damage caused. Maximum effect in term of damage. Cargo plane apparently reached V1 (go/no go speed) on the runway, and suffered a catastrophic engine failure. They passed V1, so they knew they were going down. Engine was shedding large debris, including the housing (!!!) which is a shrapnel shield.
They were on fire just as they reached V1.
Plane was fully loaded with 38,000 LB of fuel for 12 hour flight to hawaii. Worst case scenario.
Pilots did the heroic thing - they tried to take off instead at 160 MPH to minimize collateral damage (highway and warehouses at the end of the runway) and crash and die somewhere else, instead of go beyond the runway at that speed. Accelerating a fully loaded jet plane at ground level beyond the runway has obvious consequences. They had one choice.
Instead, they clipped the UPS factory because they were so low, they tried to clear it but did not. Plane then hit the ground port wing down, shearing it off entirely, smearing a fireball of jet fuel across half a mile (not an exaggeration) before the plane flipped. Crew were likely dead by before this, footage shows the cockpit being slammed into the ground like a mousetrap by the flip once the port wing was gone and gravity took the starboard wing over.
Physics took over. Plane flipped and rolled upon loss of port wing, smearing a rolling fireball of the remaining fuel load from the starboard wing for another half a mile.
Louisville is now a firestorm as a result.
Respect to the flight crew; rest in peace, they made the best they could out of a really shitty scenario. They flew it all the way down.
Standard procedure at V1 is commit to the takeoff and diagnose the problem in-air. Much of your comment is pure speculation until flight data recorders come back, we have no idea what the crew was thinking or what issues they were even aware of.
No, and reverse trust is not included in the calculation for stopping distance for a failure below V1. You can stop from just below V1 with only the brakes, if that's not possible you're not allowed to start the takeoff. You would have to reduce weight until the numbers fit the runway.
After V1 you must be able to take off on only the remaining engines. If that's not possible you must reduce weight until it is possible or you're not allowed to start takeoff at all.
This is why in very warm weather and higher altitude airports (lower performance) sometimes cargo/luggage or even some passengers are left behind, while in colder weather all seats could be used.
It's a few blocks of fire. I was on Tanker 4565 standing by as a backfill for units on scene. It's no where near "All of Louisville", that's a ridiculous thing to say.
"Pilots did the heroic thing - they tried to take off instead at 160 MPH to minimize collateral damage (highway and warehouses at the end of the runway) and crash and die somewhere else"
No, they were most likely NOT trying to be heores, its simply the standard thing to do. An aircraft is supposed to be able to take off with one engine inop, and at this speed the expected behaviour is to continue takeoff because that is what has been deemed safest.
There is always a grey area where the decision could go either way, stop or lift off and in this case it looks to me that trying to stop might have been better. But that is literally impossible to diagnose during the few seconds they had.
I saw Penske Truck Leasing trailers on fire at Kentucky Truck on WHAS"s helicopter footage from last night. 15 years ago, I'd be there once a week doing parts runs from the PTL branch off Newburg. The security cam footage of the plane rolling from Kentucky Truck was insane. Thanks for working the scene. That smoke looked incredibly toxic.
> they tried to take off instead of accelerate past the runway at ground level
Do runways have some sort of barrier between them and the next "important" thing. It seems like that would be prudent both for cases like this, and breaking failures following landings.
> Do runways have some sort of barrier between them and the next "important" thing. It seems like that would be prudent both for cases like this
Ha, Jeju Air Flight 2216 smashed into a barrier on the second landing attempt in Muan last year [0], and people commented "How could there be a barrier at the end of the runway, so obviously stupid, irresponsible", etc.
Now a plane does not smash into a barrier at the end of the runway and people suggest putting barriers at the end of the runway.
Don't mean to attack parent post, but may I suggest that
a) hordes of experts have thought long and hard about these issues, and it is unlikely that you can encounter this for the first time as a lay person and come up with a solution that has eluded the best engineers for decades ("why don't they attach a parachute to the plane?"), and
b) we are very close to an optimum in commercial aviation, and there are few if any unambiguous ("Pareto") improvements, but rather just tradeoffs. For example: You leave cockpit doors open, terrorists come in and commandeer the plane to turn it into a weapon. You lock the cockpit doors closed, and suicidal pilots lock out the rest of the crew and commandeer the plane to turn it into a weapon of mass-murder-suicide.
One improvement is a bed of concrete at the end of the runway that will catch the wheels and slow an airplane down to a stop. Pretty much everyone agrees it’s a good idea but it’s not always possible due to space needs or cost. https://en.wikipedia.org/wiki/Engineered_materials_arrestor_...
If you're designing a new airport, sure, you can have runways with ample safety margins and generous overrun areas at the end of the runways. If you want to make an existing airport safer, and you can't buy up and demolish buildings around it, using EMAS is actually a cost-effective safety improvement.
After 9/11, unlock the pilot doors. The passengers will revolt against any attempted aircraft hijackings.
We need members of the public ready to help in a situation where a pilot goes crazy, and they can’t help with a locked door making it impossible for them to enter.
The belief is that MH370 was depressurized which would have killed the passengers. A better example is Germanwings 9525 where the locked door allowed the first officer to crash the plane.
The solution is to "free" the perimeter of takeoff/landing. Bonus: People don't have to chose to work/live in these noisy areas. I understand some areas have challenges to come up with space but the US has tons of space and maybe the sea should be used to host these airports.
This is how almost all airports built in the last 50 years have worked. They were built way outside the cities. The cities grew to the airports.
More so, because of strong property rights it's very difficult to stop any development near the airports at all. The airport would have to buy up hundreds of square miles of land to prevent it at a staggering cost.
Lastly, one of the buildings that was hit was the UPS warehouse that stored goods to load on the plane. You want that as close as possible to the airport. Though right at the end of the runway is not the greatest place.
The ramp will need to be very long and very high in order to absorb the momentum of a fully loaded widebody jet. Not something that you'd want near a runway where planes can land in either direction.
Consider the possibility that gigantic flying aluminum tubes filled with tons of flammable fuel hurtling around at hundreds of kilometers per hour comprise a dilemma that has no trivial answers. Even defining what "important thing" means at any given instant is not straightforward.
Unless you have a berm several dozen meters high with a 100 meter base, you ain't stopping something like this from a physics standpoint unfortunately.
Many airports have this problem. The recent korean air disaster which echos this is another example. BTW, this is why most airports, if possible, point out to sea...
Newer airports usually try to have space, that's the only thing helping with the physics involved here.
Older airports might have EMAS [1] retrofitted at the ends to help stop planes, but that's designed more for a landing plane not stopping quickly enough (like [2]) - not a plane trying to get airborne as in this case.
There is a dead zone between rejection and successful take-off speeds. We see it hit too often.
I think pilot training is playing a factor. A normal rotation kills too much energy. One engine can climb when you have some airspeed and get clean, but if you lose too much energy on rotation, the inefficiency of the AoA for the rest of the short flight means that engine can no longer buy you any up. I've seen too many single-engine planes going down while trying to pitch up the whole way down.
So, less aggressive single-engine rotations and energy absorbers at the ends of runways that can't get longer. This seems like the kind of thing where we do it because it removes a significant cause of people dying.
Another crash video shows the aircraft clearly descending before colliding with anything. It manages to go up a bit, so it's fast enough to get airborne. The normal looking rotation kills too much energy. The plane is then too inefficient to maintain speed. AoA goes up while energy goes down. Power available goes negative and then it's over.
Rotation does increase drag, but you need to rotate in order to achieve the necessary angle of attack. The only way to reduce the rotation angle is by going faster than the normal rotation speed for the given weight and airfield density altitude, but doing so is out of the question in this scenario.
There might be other kinds of damage where the quicker altitude gain of a normal rotation is crucial for survival.
I'm skeptical whether pilots can realistically make this kind of decision, given that they have no more than a few seconds to make it, and in cases such as this based on very incomplete information about the state of their aircraft.
Increased thrust requirements for airliners that force planes to hit an increased v1 (or whatever it's called) sooner on the runway to allow for more time to reject takeoff.
> It manages to go up a bit, so it's fast enough to get airborne. The normal looking rotation kills too much energy.
Yes, it did get airborne for a few seconds but from the video below, it looks like the left wing was damaged by the fire and could not provide enough lift, then the right wing rolled the plane to the left causing the crash.
> looks like the left wing was damaged by the fire
The wings and aerodynamics don't really care if air or air with combustion are flowing around them.
Roll is a consequence of the loss of control due to low speed and the yaw of the good engines. Speed up, rudder works, plane might maintain positive climb.
For skin, a few seconds might be significant. For the spars, not nearly enough time to matter. It's also not at cruise speed slamming into a downdraft or anything. This is about a 1G loading. Negligible for a while. While the fire looks cool, there's a lot of free stream mixing in and the temps won't really get that high beyond the cowling.
More likely is the hydraulics on that side burst, leading to a loss of pressure keeping the control surfaces deployed. If that lead to the leading-edge slats retracting (like they did in AA 191) you'd get a massive loss of lift on that side. The structural parts of the wing didn't have time to melt, but the fire certainly could damage all the internal control materials.
Some runways have been extended with ‘engineered materials’ surfaces, often a form of porous concrete into which an airliner’s wheels will sink, absorbing a lot of energy and arresting the airplane without causing it to break up. It is very effective for landing overruns, but I don’t know about last-seconds aborted takeoffs.
Security/debris fencing yes, but that's like, orders of magnitude short of what would stop the amount of energy we're talking about here.
You also don't particularly want it to be catastrophically effective as there are real world cases where planes have clipped the fence and then NOT gone on to crash, or at least to crash in a fairly controlled manner with the majority onboard surviving. Hitting a brick wall at 180mph is going to have a 0% survival rate.
Yet a reinforced concrete wall of e.g. triangular section and anchored with "long enough" piles would be about the only not-that-expensive way to turn a short strip of "airport land" past the EMAS into a V1 stopping supermarket.
Work place related accidents always have a certain tragedy to them. Still remember when in the industrial park, my employer is located in, tanks belonging to a trash incinerator for special chemical waste exploded, taking several people with it.
>> highway and warehouses at the end of the runway
It's astonishing that this is a thing. Why aren't we building airports with enough space for a plane to remain on the ground and have plenty of room to decelerate in this situation? I can understand why it can't be retro fitted to existing airports but is it a scenario that's considered at new airports? Just seems like such an absolutely basic safety step.
> Why aren't we building airports with enough space for a plane to remain on the ground and have plenty of room to decelerate in this situation?
But that's exactly what a runway is? They're extremely long, have ample safety margins, and have "protected areas" extending out on either end, and outside of that there are regulations about what can and can't be built along the extendend runway centerlines. But jetliners are huge, heavy, fast, and designed to go long distances - the stopping distance of a fully loaded jet at full takeoff speed is measured in miles.
Yes, new major airports (rare as they are) do try to acquire large areas of land, larger even than they think they need now, in anticipation of future expansion. However, for scenarios like this, there's limited utility to making the runway longer "just in case." They already pick runway sizes "big enough and then some" as the minimum to even bring planes of each size to an airport. So there is margin.
But no matter the margin, a plane can always crash on the wrong side of any fence. And people will always build right up to wherever you put the fence as closer to the airport is more convenient for everything airport related.
Airports are usually built (originally) out in the boonies away from the major metro area. As time goes by and that land gets more valuable developers grease palms of politicians in land use commissions to allow developments closer and closer to the airports.
Airports also grow themselves. Some municipal airports sited for small aircraft extend their runways to handle larger planes.
Yes, in fact, lots of the area that is warehouses to the south, and where the larger run way to the east and some buffer zone to the east used to be neighborhoods and they were bought and torn down to make room around the airport.
It was around 250k gallons of fuel. Our CAD notes on the initial dispatch said 250k, one press briefing said 280k, and then it was changed to 220k which I think is the actual number.
It depends on whether or not, at the point in which you realize you have an engine on fire, you have room on the runway left to stop.
As I understand it, there is a low speed regime, under 80 knots, where are you stop for basically anything.
Then there is a high speed regime, where you only stop for serious issues, because you now have so much kinetic energy that stopping the plane, while still possible, will involve risk. (i.e. fire from overheated brakes.)
At a certain point, called V1, there’s no longer enough room to stop, no matter what your problem is. You’re either getting airborne or you’re crashing into whatever is ant the end of the runway. In general, getting airborne is the safer option, while obviously still not risk free.
However, this calculation also assumes that the engine fails in an isolated fashion, and its failure did not affect the other engines. If the failure of the left engine threw off debris that damaged the middle engine then we are now talking about a double engine failure. I’m sure the pilots knew there was a problem with the engine when they made the decision to continue, but it’s possible that problems with the middle engine weren’t apparent yet and that it only started to fail once they were committed.
Obviously, this is just speculation, and we will have to wait for the preliminary report at least.
Being untrained but spending a little bit of time in a full motion 737 simulator that’s used to train and certify commercial pilots, I was amazed at how quickly things happen even in a scenario with no faults.
This situation (single engine failure at V1) is something that commercial pilots are certified in at every recurrent certification since it’s one of the most difficult you can be in. The crew now need to climb and go around for a landing on one engine while simultaneously running through the engine failure (and also likely fire) checklist. I don’t know if a double engine failure at V1 on a fully loaded 3 engine aircraft is technically survivable or if it’s something that’s trained on. They were put in an incredibly difficult situation just based on what reports we’ve already seen.
I would be astounded if there was anything the pilots could have done to prevent this.
The plane was two engines out and a main fuel tank on fire, fully laden with a full fuel load. No amount of training or improvisation was going to fix that.
If anything it's lucky/professional they crashed into an industrial park and didn't have time for a go around. It would have been an even bigger disaster if they'd crashed into the town centre or a residential area.
V1 is the speed at which you can still stop the plane before the end of the runway. (It is computed for each takeoff based on runway length, aircraft mass, takeoff engine power setting, flaps, wind, runway condition, etc.)
When the plane reaches V1, pilots take the hand off the throttle: they're committed to takeoff, even if an engine fails. It is better to take off and fix the problem or land again, than to smash into whatever is beyond the end of the runway.
Isn't there any margin? Does it calculate stopping before end of runway or before causing damage?
Surely uncertainty about the situation contributes to defaulting to committing, but what if it's a passenger plane and at V1 pilots know they've lost power?
Wouldn't veering into highway at 30 mph be weighted against certain, big loss of life?
Edit: I now see that this has been partially answered by uncle comment
There is some margin in the calculations. But the training is very very clear, before V1 you must abort and after V1 you must continue. No discussion, no decision to make. You call V1, hands go off the throttles and no matter what you're going to fly.
The margin is for example that the plane must not just be able to fly, but also reach a minimum climb gradient to clear obstacles with a bit of safety margin. There is also an allowance for the time it takes from calling abort to actually hitting the brakes. And for example headwind is part of the calculation (it makes the takeoff distance shorter) but only 50% of the headwind is used in the calculations.
But all of those margins are not for the crew to use, the crew must just execute the procedure exactly as trained which means at V1 you're committed to continue the takeoff. And before V1 in case of an engine failure you have to hit the brakes to make sure you can stop before the end of the runway.
Maybe we should just stop commenting about whether or not something is AI generated. It doesn't add anything to the discussion, and is a waste of time.
An apology (in advance or afterward) doesn't absolve you of responsibility. And if you feel the need to apologize for something in advance, that's a huge clue that maybe you should stop yourself from doing the thing you've just apologized for.
> Maybe we should just stop commenting about whether or not something is AI generated. It doesn't add anything to the discussion, and is a waste of time.
Ok, sure!
As for the rest, "excuse me if I'm wrong" is a very common and valid phrase, though a bit ruined by sarcastic misuse
I attempted to show with it that I don't assume anything or default to hostility, though on a different occasion you'd yourself probably argue that others feelings are not my responsibility.
I'm not sorry for asking the question, unless t0mas88 got offended by it.
Finding oneself with the need to apologize in advance is an excellent hint to examine extra hard if you really should do what you're apologizing for. Apologizing when necessary is good, not having to apologize is much better.
It's a great level-up for characters of most alignments.
I did. By apologising and making a question, not accusation.
The commenter himself didn't seem hurt by it.
I'd recommend raising your outrage threshold.
I clearly am keeping this light hearted, while you go into fight mode because you saw the word "offence". Loosen up a bit :)
I see now that it probably wasn't, but "nothing" is an overstatement.
And knowledgeable and helpful responses can be AI, so there might be a fallacy somewhere in your offence-taking.
Are you offended when people do that in general, or only when they are wrong?
I do appreciate the effort put into writing a good comment.
> it probably wasn't, but "nothing" is an overstatement.
No, "nothing" seems perfectly accurate to me. I don't see even a single indication in tone, phrasing or punctuation that is a classic LLM giveaway.
It's offensive to standards of decency to question the authenticity of someone's speech, and it doesn't matter if you phrase it as a question or preface it with "excuse me if not". Unless there is really a strong reason to suspect something, which is absolutely not the case here. It's offensive when it's not warranted.
Absolutely not. There was genuinely nothing in their thoughtful and informative reply that seemed AI generated to me. Have you never seen people on HN write detailed, articulate answers? This was one of them.
Asking a question, getting a helpful response, and then implying it was written by AI is quite rude. Saying, "Excuse me if not, $SOMETHING_VERY_OFFENSIVE" does not make "$SOMETHING_VERY_OFFENSIVE" any less offensive! It's disheartening to see someone take the time to write a great answer only to be met with such a rude question. Please don't do that here. It's frustrating and discourages genuine contributors.
If it's absolute, then why add that?
Mine "to me" gave me a different impression, but it being a "to me", I questioned, not accused.
You perhaps recognize t0mas88 after all these years on HN. I don't. I'm relatively a new and infrequent user.
So I hope that I would be criticised just as passionately if that'd be a random user whose comments indeed turned out to be AI. Because no matter what is the fact, the question is rude, nay, very offensive.
Using AI isn't a tabu. It is fully debatable wether generating, reviewing and pasting helpful information on an informal forum is wrong, I just implied my own frustration with what I often experience.
Of course I could've phrased it much better, and I suspect you guys wouldn't bat an eye.
> Have you never seen people on HN write detailed, articulate answers?
> It's frustrating and discourages genuine contributors.
But regarding flight ability, wouldn't that be V2?
I thought there exist conditions where V1 is well below rotation speed.
Anyways,
> to make sure you can stop before the end of the runway
answers my main question, and makes sense from a procedural standpoint.
But still, hard to believe that there is no room for in-situ evaluation if runway overrun is worse than likely crash.
Of course then again, those have to be split second decisions.
A fully loaded plane is extremely likely to turn into a fireball if it hits anything on the ground, even at 30mph. It's just a thin shell of aluminum with tons of fuel sloshing inside.
V1 is the decision speed with respect to a single engine failure in a multi-engine aircraft. It's the speed below rotation speed at which the decision to abort safely can no longer be made.
Captains can make the decision to abort the takeoff in the case of absolute power loss or for 'failure to fly' (where the aircraft is clearly not going to fly, e.g. the elevator/pitch controls aren't responding). But the training is adamant: if you're uncertain what has happened after V1 you try to fly the plane away from the runway.
That's what I'm getting at. I want to abort unsafely. Imagine 400 meters of grass field after the end of the runway, and a water body. I'm asking wether such factors are accounted for, or if plane on ground beyond runway does-not-compute.
I expect pilots are trained explicitly not to do that.
If you can't abort safely, than it follows that the safer course of action is to try to fly. I'm sure there are exceptions to that, but a pilot has barely seconds in which to decide if any of those exceptions apply, so they're not going to abandon procedure unless the situation is clear.
That "extra" 400m of grass? That's for all the other things that can still go wrong even when you follow procedure. e.g., you're below V1 so you abort takeoff, close throttles and hit the brakes. You should be able to safely stop on the runway.
But now your brakes fail because maybe the reason you had to abort was a fire that also managed to burn through your brake lines, or it started to rain just as you were taking off, or...
Now that's where the 400m of safety margin comes in to save your ass (hopefully). It's "extra", you don't plan on using it.
> Wouldn't veering into highway at 30 mph be weighted against certain, big loss of life?
V1 for this plane in those conditions is nearly 200 mph. Even if they shut down all engines and applied full brakes (and assuming the brakes/tires didn't catch fire), they'd still run off the end of the runway with enough kinetic energy to kill themselves and anyone else in the way.
To avoid mass casualties at the end of the runway - on the road, or the buildings that the runway points to. Check the layout on google maps.
More specifically, V1 is the max speed at which you're about to take off, but you can still abort from. They hit that max speed and realized there was a major problem that hypothetically, they could have slowed down from, but realistically was not possible. They had no choice.
They were on fire just as they reached V1.
Plane was fully loaded with 38,000 LB of fuel for 12 hour flight to hawaii. Worst case scenario.
Pilots did the heroic thing - they tried to take off instead at 160 MPH to minimize collateral damage (highway and warehouses at the end of the runway) and crash and die somewhere else, instead of go beyond the runway at that speed. Accelerating a fully loaded jet plane at ground level beyond the runway has obvious consequences. They had one choice.
Instead, they clipped the UPS factory because they were so low, they tried to clear it but did not. Plane then hit the ground port wing down, shearing it off entirely, smearing a fireball of jet fuel across half a mile (not an exaggeration) before the plane flipped. Crew were likely dead by before this, footage shows the cockpit being slammed into the ground like a mousetrap by the flip once the port wing was gone and gravity took the starboard wing over.
Physics took over. Plane flipped and rolled upon loss of port wing, smearing a rolling fireball of the remaining fuel load from the starboard wing for another half a mile.
Louisville is now a firestorm as a result.
Respect to the flight crew; rest in peace, they made the best they could out of a really shitty scenario. They flew it all the way down.
Footage:
https://x.com/osinttechnical/status/1985845987684855969?s=46
https://x.com/faytuksnetwork/status/1985849267152699741?s=46
https://x.com/faytuksnetwork/status/1985848132500885995?s=46
https://x.com/faytuksnetwork/status/1985843126934614297?s=46