Landing into a storm

[flight191]

I'd say that, with the current knowledge, practices and training, they should have gone around at 3, or 6 the latest. Having failed to do that, 8 would have been another good opportunity with the knowledge and training that they had back then, and even 12 was still doable.

12 is highly doubtful, at this point the cell was at full intensity and had developed to that point in the previous 60 to 90 seconds. The peak of the cell's intensity at the foot of 17L coincided with DL191's arrival. See post #128 for the radar readouts. Richard Bray from NASA came to many of the same conclusions you have all already hashed out here but I think after 8 all hope was lost. Past is prologue.

[Evan]

The manner in which you attained that speed, however, could meet the definition of windshear. That is what you are really looking for. There is a difference in gusty conditions increasing to that speed, but averaging a reasonable value, and going from on speed to +25 sustained due to a thunderstorm, when you are likely to fall out of it again at an inopportune moment.

Of course they could have given it away at 800ft, and if there was lightning and heavy rain, there's pretty good reasons to do so! We also understand a lot more about what the consequence of an increase in headwind associated with a storm is likely to lead to.

MCM, I understand what you are saying but I meant this in the context of the sequence of events as Gabriel has since posted. The Capt recognized windshear but did not call go around, and after the precipitous drop in airspeed (that he had expected) still did not call go around. This seems to me like a classic case of a unstable approach that was continued due to overconfidence and get-there-itis. Granted, we know much more today then they did about the phenomena. I just hope pilots these days are more willing to throw off an approach like this one long before it gets this real.

Does an automated windshear alert necessitate a go-around?

[elaw]

Overconfidence, get-there-itis, and perhaps just a slight amount of excess optimism?

It's quite possible the pilot understood and recognized that he was caught in a windshear, but also quite reasonably thought the effect would dissipate, and that the aircraft had enough speed/energy to make it through. And the consensus seems to be that the aircraft actually *did* have enough speed/energy to make it through, it's just that the pilot wasn't quite aggressive enough on the controls to make sufficient use of it.

[MCM]

The Capt recognized windshear but did not call go around, and after the precipitous drop in airspeed (that he had expected) still did not call go around. This seems to me like a classic case of a unstable approach that was continued due to overconfidence and get-there-itis. Granted, we know much more today then they did about the phenomena.

As you acknowledge, we know a lot more about windshear than we did then. I disagree that this is the classic example of an unstable approach. The classic example of that is the pilot who 'overcooks' the approach, struggles to get the flap and gear out, lands rushed, in an inappropriate configuration. This is different - the crew were stable, conducting a reasonable approach. They didn't assess the weather conditions as being a threat - with today's knowledge, they would likely think differently. Similarly, pilots were aware that speed could increase then decrease - but they didn't likely know just how bad that can get. I believe, for example, that the Captain knew that the speed would come off - that is something that happens daily around the world. Approaches are made where a gust, or shift of wind, will have speed temporarily increase, and then decrease back to the expected landing value. Wind is changing all the time - the risk comes when we don't know what it is going to do. I do not believe, in this accident, that the Captain realised that they would be entering an area where the speed would not only "come off", but would continue to decrease as they entered large tailwinds and downdrafts. I say this, because if he had known, they would never have commenced the approach.

Lets not forget that this isn't just a bit of windshear. These pilots encountered one of the most dangerous and unpredictable meteorological phenomenons now known to aviation.

I just hope pilots these days are more willing to throw off an approach like this one long before it gets this real.

Yes, they are, because they know more about them, get far more accurate details about microburst and windshear events, and have tools on board to assist.

Does an automated windshear alert necessitate a go-around?

Absolutely.

[Gabriel]

12 is highly doubtful, at this point the cell was at full intensity and had developed to that point in the previous 60 to 90 seconds. The peak of the cell's intensity at the foot of 17L coincided with DL191's arrival. See post #128 for the radar readouts. Richard Bray from NASA came to many of the same conclusions you have all already hashed out here but I think after 8 all hope was lost. Past is prologue.

Well, no. As I've said before, not only after 12, but until the point of hitting the car, the plane had enough energy and reserve performance to fly away.

At any point during the approach and until hitting the car, applying TOGA and pulling up to 15 deg nose up or to the onset of the stickshaker (whichever came first) would have prevented the accident. Note that now this is the current windshear escape technique, but not back then.

The FO, however, first, when hit by the downdraft and tailwind, increased thrust but below TOGA and raised the nose a lot but just as needed to keep the glide slope, and then, when hit by the updraft and the split-second stickshaker activation, applied TOGA but shoved the nose way too much down and then was too cautious pulling back up. Even a couple of seconds before hitting the car the plane had a healthy airspeed that was healthily increasing too with TOGA applied, and the plane was already nowhere near stall. All the pilot had to do was to put the plane in a climb attitude.

Of course, easy to say from here. He was probably very confused by that point.

[flight191]

Well, no. As I've said before, not only after 12, but until the point of hitting the car, the plane had enough energy and reserve performance to fly away.

Connors called out TOGA at about 6:05:45 and a little over 300 feet. At that point they were losing about a 100 fps.

They touched down in the field at 6:05:52 (and were 6336 feet from the foot of 17L) hit the Celica maybe 1 to 2 seconds later and by 6:05:56 it was all over. They had at most 3 seconds before they made contact with that water tank.

http://www.youtube.com/watch?v=HY7pH3fzsvY

[Gabriel]

Fixed.

I agree with your fixing.

A man named Gabriel once said, it's almost always better to crash under control.

Yes, but "better" requires the other reference to compare against.
Crashing under control is better than crashing out of control, but no crashing is even better.

[Evan]

Lets not forget that this isn't just a bit of windshear.

MCM, this is where I am confused. Is it acceptable to continue after encountering "a bit of windshear"? If an automated warning absolutely necessitates a go-around why does pilot recognition of potential windshear not necessitate a go-around? Is there an allowable amount of windshear that is not going to trip the reactive windshear alert? I realize there is a distinction between windshear in general and downburst and it's downburst that I'm mostly concerned about.

[3WE]

The Capt recognized windshear but did not call go around, and after the precipitous drop in airspeed (that he had expected) still did not call go around....I just hope pilots these days are more willing to throw off an approach like this one long before it gets this real.

Time out...when they lost the airspeed, the primary issue was that there was a huge likely hood they would hit the ground. Maybe "not hitting the ground" was a bit higher on their priorty list than "the stabilized approach criteria and go-around process"

I thought you knew a little more about airplanes...calling go around doesn't do all that much when you get hit in the ass with a surprise 70 kt tailwind.

Calling go-around would have changed nothing!

In fact, they did call for TOGA POWER- which...excuse me...don't cha think that there might have been an implied go around?

I've said this before (in this thread no less) Go to Microsoft flight simulator- program in a 35 knot headwind that starts at 600 feet and a 35 knot tail wind that starts at 400 feet AGL.

Then fly the ILS to 17 L at DFW in a DC-10 (or an L-1011 if you can find one).

Be sure to call "go around" to yourself.

And also be sure to remember that you might know one little bitty thing that the 191 Pilots did not.

[Gabriel]

Is there an allowable amount of windshear that is not going to trip the reactive windshear alert?

Yes.

Is it acceptable to continue after encountering "a bit of windshear"?

Let me state it this way: It is almost impossible not to encounter "a bit of windshear", so rather than "acceptable" I'd say it's almost "mandatory" to continue after encountering "a bit of windshear", else you'd be often going around once and again until, eventually, you would still be landing after encountering "a bit of windshear" under the provisions of the PIC's prerogative in an emergency situation (like "either we land while we still have some fuel or otherwise we land anyway").

[Gabriel]

8- The FO manages to follow the localizer very well, which under these conditions of course implies raising the nose a whole lot.
9- The plane is hit by a strong updraft.
10- Since the AoA was already high, that updraft generates a split-second stickshaker activation.
11- The FO responds to that stickshaker activation by pushing down way too much, putting the plane in negative Gs and shoving the nose down, thus effectively putting the plane in a dive.
12- The captain says Go around.

12 is highly doubtful, [...] I think after 8 all hope was lost. [...] Connors called out TOGA at about 6:05:45 and a little over 300 feet. At that point they were losing about a 100 fps. [Note: this would be point 12]

10 seconds before that, at 6:05:35, the plane was flying at Vref, exactly on the glide slope, and the nose was pointing about 10° nose up, when the plane changed from being hit by a downdraft to being hit by an updraft.

So how do you pass from this situation at :35 to the dare situation of :45 that you described? Answer: With 6 seconds of continuous, strong, disproportionate nose-down inputs. This was the most critical mistake because it actively put the plane in a dive that was not necessary and from which it would have been hard to recover, but not impossible.

They touched down in the field at 6:05:52

Interesting that they took seven seconds to descend from 300ft. It should have taken 3 seconds at the 100 fps they were doing. But not. Do you know why? Because the pilot (FO) managed to level off, almost. And not only that, but during all these 7 seconds the AoA remained below the stickshaker threshold (except perhaps a split second at :48 ), and the airspeed remained not only well above the stall speed but also above Vref and with a strong increasing trend. This means that the plane had much more performance to give than what the FO managed to extract from it.

In particular, if you look at the last 3 seconds before touching down on the field, the airplane was flying almost level, the airspeed remained above Vref +25 (about 1.5 x Vs, enough to pull 2.2 Gs), and the AoA remained below 10°. Let's take a good margin: Instead of 2.2 Gs during 3 seconds let's take 1.5 Gs during 1.5 seconds: that's a change about 1500 fpm in the vertical speed (for example, a change from -750 to +750).

If anything, the animation in your link supports that the accident was technically avoidable even after item 12 and up to a couple of seconds before touching down in the field.

[3WE]

Tangential factoids:

Wind shear happens extremely often.

In the years after Delta 191 a lot of ground-based wind shear detection systems were installed and if it was a somewhat windy day, at a busy airport, you would hear the tower broadcast a wind shear alert from time to time.

The wind aloft is almost always stronger than the wind on the ground- so you are almost always landing into a diminishing, speed-losing headiwnd. (meant sarcastically)

The last light airplane ride I took in mundane weather, a lear jet reported shear and speed disruption on short final...sure enough, the 172 did a little flaump-da-dump at 100 feet AGL.

Several year ago, I was doing touch and gos on a windy day...every time I turned on the cross wind leg, the airspeed dropped 10 knots and I paused to build speed...then on base...suddenly there was 10 extra knots...time after time.

And one of my favorite experiences was going out on a perfectly clear, perfecly calm cold fall night (roughly 24 hours after a strong cold front) and releasing a helium-filled baloon.

The thing rose perfectly straight up for about 200 to 300 feet, and then wham- it took off going South at a brisk clip.

And recall that AA 1420, landing in a severe storm with a strong, gusting crosswind, encountered very managable wind shears.

So the art is discrimininating every-day-extremely-common wind shears versus the risk (acceptable level of risk) of a nasty-ass 70 knot Delta 191 levels of wind shears.

[3WE]

...the pilot (FO) managed to level off, almost. And not only that, but during all these 7 seconds the AoA remained below the stickshaker threshold (except perhaps a split second at :48 ), and the airspeed remained not only well above the stall speed but also above Vref and with a strong increasing trend...

There's a ton of stuff contained here.

*You could say that the plane essentially made a normal, gentle, controlled touchdown. (At a slightly nose-low attitude which- just like you say, meant they had energy to haul back and take flight again)

*We may be glossing over that the spoilers automatically deployed at touchdown, thus killing lift and the ability to regain fight (at least I think the spoilers deployed? or was there an automatic over ride if you are at full throttle?)

*I recenty realized a subtle mental hang up that I have...that is stall warning = stall imminent. Therefore you need to avoid a stall warning. And, there's nothing wrong with that until you are in a Delta 191 scenario. That's wrong- you want to avoid a stall, not the waring.

More correctly is that a stall warning (stick shaker) = a warning that your speed is getting close to- but is still several knots above stall speed.

So, if the pilot is a 'good, stall-fearing guy' he's got the safety buffer between stick shaker onset and stalling + the safety buffer between his speed and the stick shaker onset , and it's this second safety buffer that sort of did them in.

I know I'm overly hung up on this- but if I hear a stall warning- I want to shove the coal stoker forward and punch the nose down 5 or so degrees and build some speed! (Might have something to do with reading 'Stick and Rudder' and a fear of falling.)

[brianw999]

..................
............So the art is discrimininating every-day-extremely-common wind shears versus the risk (acceptable level of risk) of a nasty-ass 70 knot Delta 191 levels of wind shears.

Like I say, I'm not a pilot but I would imagine that you wouldn't know which you were in until you're in it ......by which time it's too late ?!!

[Evan]

Time out...when they lost the airspeed, the primary issue was that there was a huge likely hood they would hit the ground. Maybe "not hitting the ground" was a bit higher on their priorty list than "the stabilized approach criteria and go-around process"

I thought you knew a little more about airplanes...calling go around doesn't do all that much when you get hit in the ass with a surprise 70 kt tailwind.

Calling go-around would have changed nothing!

In fact, they did call for TOGA POWER- which...excuse me...don't cha think that there might have been an implied go around?

Well, I know a little bit about airplanes. I know pointing the nose down doesn't get you altitude. I know continuing on the localizer doesn't 'imply' go-around. I know TOGA involves moving the thrust to full power, which they didn't do when they first got into this. I know windshear recovery involves doing this. Whether or not it gets you out of it doesn't determine whether you apply the procedures in the first place. I know these guys experienced a very large and sudden drop in airspeed in the presence of convective thunderstorms at a very low altitude which is a bit different from +/- 10kts in VFR. I know any pilot worth his stripes knowing now what they didn't know then would be bugging out of there at that point, or at least trying to. I know that after applying TOGA they mostly arrested the descent and almost made it out alive. What I don't know, as usual, is what your point is.