Announcement

Collapse
No announcement yet.

Air India Express 737 skids off runway and breaks in two after landing at CCJ

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Gabriel View Post
    http://avherald.com/h?article=4daf960f&opt=0

    Utterly confused by this:



    And this (note: the bottom of the picture is forward, the top is back):
    Is that from the accident aircraft? Ground spoilers are not deployed.

    Comment


    • #17
      Originally posted by Evan View Post
      Is that from the accident aircraft?
      Yes, taken from where the instrument panel should have been and facing to the back of the plane. This is a zoom of a larger picture of the cockpit. Take a look at the AvHerald link in my post.

      Ground spoilers are not deployed.
      Correct. And thrust levers are about fully forward and reverse levers are closed and flaps not sure but seems to be a low flap setting.

      Where they attempting a go around or what?

      --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
      --- Defend what you say with arguments, not by imposing your credentials ---

      Comment


      • #18
        Originally posted by Gabriel View Post
        Yes, taken from where the instrument panel should have been and facing to the back of the plane. This is a zoom of a larger picture of the cockpit. Take a look at the AvHerald link in my post.


        Correct. And thrust levers are about fully forward and reverse levers are closed and flaps not sure but seems to be a low flap setting.

        Where they attempting a go around or what?
        I suppose the thrust and reverser lever positions could be due to impact force (the reverser levers are staggered), but not the speed brake lever.

        I thought about a botched go around, but if they had not slowed down considerably the wreckage would have been much more widespread and catastrophic. By the looks of it, they weren't moving very fast when they came down that hill. Perhaps there was some indecision... a momentary decision to go-around followed by a decision to stop...?

        However, if they reset the spoilers on the first go-around and failed to repeat the checklist...

        We know that ground spoilers make a big difference on contaminated runways. Can you estimate how much distance that would be here? How much combined with late reversers...?

        Comment


        • #19
          Originally posted by Evan View Post
          I suppose the thrust and reverser lever positions could be due to impact force (the reverser levers are staggered)
          Not as easy as it sounds. Under the hypothesis that they were doing everything possible to stop, they would have been using reverses.
          The reverse levers cannot be pulled until the into the reverse idle position until the thrust levers are all the way back in the idle position, and cannot be moved further back to accelerate the engines in reverse until the reverses are physically fully deployed. Conversely, with the revere activated (even in the idle reverse position) the thrust levers cannot be moved forward. So, while anything is possible in a hard impact, the thrust levers should in principle not be able to move forward if the reverse thrust is in use.

          I thought about a botched go around, but if they had not slowed down considerably the wreckage would have been much more widespread and catastrophic. By the looks of it, they weren't moving very fast when they came down that hill.
          Agree. Which is exactly the reason why I said I was utterly confused.

          Perhaps there was some indecision... a momentary decision to go-around followed by a decision to stop...?
          Maybe... but the thrust levers.....

          However, if they reset the spoilers on the first go-around and failed to repeat the checklist...
          That would not be too surprising (unfortunately), but that doesn't explain the flaps (or the thrust levers).

          We know that ground spoilers make a big difference on contaminated runways. Can you estimate how much distance that would be here? How much combined with late reversers...?
          This seems like a simple question but it is not (unless you accept just a big IT DEPENDS for an answer).
          To talk about the effects of the spoilers (or lack thereof) in a contaminated runway, first we need to talk about the concepts involved in slowing down an airplane on the ground.

          There are 3 longitudinal forces on the airplane while slowing down after landing:
          1) Aerodynamic drag, which goes with the square of the speed and is not affected by runway conditions.
          2) Thrust, reverse thrust hopefully, which is not affected by runway conditions.
          3) Friction, which is friction coefficient time WoW. The coefficient of friction is highly affected by the runway conditions. The WoW (which here means the weight on the wheels with brakes, so it is airplane weight - lift - weight on the nose wheel) is affected by several factors.

          Pulling back on the elevator creates a down lift which reduces the overall airplane lift (or makes it more negative) thus increasing the weight on wheels, and because this downward lift is way aft of the CG it also reduces the weight on the nose wheel which in turn again increases the weight on the main wheels. Flaps increase lift thus reducing the wight on wheels, but they increase aerodynamic drag. In most conditions, unless the coefficient of friction is ridiculously low, retracting the flaps would have more effect on the braking action that the loss in aerodynamic drag, but the flaps take too long to move so by the time they get to the retracted position you would hopefully be going slow enough anyway where it will not make a significant difference neither in lift or drag (both will be very low already) and that plus the increase in an already heavy workload is why they are not retracted upon touchdown. Braking reduces WoW and hence reduces braking action. Because the point of contact between the tires and the pavement is well below the center of mass, braking creates a nose-down pitching moment that increases the weight on the nose wheel wheel and reduces the weight on the main wheels (i.e wight shifts forward during braking, as it is normally and incorrectly said). Finally, spoilers decrease lift and hence increase WoW, and also add drag. The addition of aerodynamic drag is almost negligible when the braking action is good.

          To understand the relative importance of each factor we need to consider these things:

          First, leave drag and thrust aside for a minute, and also asume that 100 of the airplane weight is applied on the braking wheels....
          Having a better coefficient of friction doesn't mean breaking more. The coefficient of friction (multiplied by the WoW) is a cap on the max possible breaking force.
          As an example, autobrakes 2 (which is typical). This setting will attempt to provide a deceleration of 0.15G or 3 kts/second. This would require a coefficient of friction of just 0.15 (with the above assumptions). That is quite low. The dry coefficient of friction easily exceeds 0.5, and most runway conditions will still be above 0.15. So if you are good to land and stop with autobrakes 2 on a dry runway, your stopping distance will still be the same in almost any runway condition except the very worst ones.
          But the other side of the coin is also true: If, given the runway condition, you cannot achieve a coefficient of friction of 0.15, selecting autobrakes 3 or MAX will not help either. No matter the autobrake setting, you cannot get more braking that the traction between the tires and the runway (in practice, selecting a higher setting can actually help because the runway condition is never super homogeneous so you can get more than .15 in some parts and compensate for other ones where you cannot achieve 0.15).

          Now, returning to real WoW, drag and reverse.

          Then, the worse the runway condition, the more of the slowing done is done by drag and reverse thrust and the less is done by the brakes. As an example, on a dry runway and applying max brakes, the drag and reverse will be a tiny fraction of the slowing down. But if the coefficient of friction is zero, then this tiny fraction (drag and reverse) is all that there is left and becomes 100% of the slowing down.

          The auto brakes don't apply a fixed amount of brake pressure. They apply as much as needed to achieve the desired deceleration. It may happen that, just after touch down, the drag and the reverse thrust are already giving you the 0.15g of deceleration so the autobrake will not command brakes. As the plane slows down and hence the drag and reverse force is reduced, it will add more and more brakes to keep the 0.15g.

          Spoilers add some drag, but their main function is to spoil the lift to increase the WoW. And they are excellent at that. Just before touchdown, the lift was about equal to the airplane weight. When you lower the nose the AoA goes down some 5 degrees to about 0 degrees, but the zero lift AoA is negative several degrees, especially with flaps extended, so the plane will still be doing maybe 60% of the airplane's weight, leaving only 40% of WoW. The spoilers really destroy the lift which becomes almost zero or even negative in some planes so you get a WoW that is even higher than the airplane's weight). This is very important because the first seconds of braking, were you are going faster, are the ones that count more (they consume more runway). SO without spoilers you get a low braking action when you are fast and the braking action increases as you slow down, not because the coefficient of friction increases but because you get the whole WoW (or even more) since the beginning.

          But the question is to know how much the spoilers affect braking is... did you need that braking? Were you going to use it in the first place? Again, you may be getting most of the required deceleration just with drag and reverse so having a strongly reduced traction in the beginning may not affect an inch the stopping distance. And when you slow down, the spoilers are not that effective anyway because you naturally lose drag and lift (and gain WoW) since they go with the square of the airspeed. Spoiler or no spoiler may not make a difference. Zero.

          In a situation where you were expecting the brakes to contribute with say 0.2g of braking (on top of whatever drag and reverse gives you), if the friction coefficient is 0.2 you absolutely need the spoilers to get this deceleration, but if the coefficient of friction is 0.5 you can still achieve the 0.2G with only 40% of WoW, so the stopping distance will not change with or without spoilers.

          In another situation where the conditions of the runway are super poor that the coefficient of friction is almost zero, the spoilers will a bit of drag and will aslo help extract whatever little is there to be extracted from the friction, but the reverse thrust is the king in these hyper slippery conditions (think aquaplaning or wet ice). So spoilers will help and my be critical if you are with little margins, but the stopping distance will not be affected by a lot.

          Now, in a situation where the runway is so short or the conditions of the runway mildly slippery and you need a braking traction that can only be achieved with enough WoW, then spoilers.

          For this particular accident, there was plenty of runway ahead of the plane even after it touched down long. So unless the speed was ridiculously higher than it should, or ifthe runway was not just wet, it looks to me that the plane should have been able to stop even without spoilers.

          Let's do "reverse engineering": When the plane touched down, it had more than 6400 ft to the edge of the cliff. Here you have some descelerations rates that would have been required to stop in that distance in function of the touchdown ground speed:

          130 kts: 0.12g = 2.2 kts/sec
          180 kts: 0.22g = 4.2 kts/sec

          And remember that, initially, if the drag and reverse provides say 0.1g of deceleration, you only need a coefficient of friction of 0.02 (equivalent to no-brakes rolling friction) or 0.12 (equivalent to autobrake 1) respectively. These coefficients of friction should be easily achievable with a wet runway, amusing that that you have full WoW. Now, no spoilers, 180 kts (= LOTS OF LIFT, event if the airspeed is "just" 168 kts due to the tailwind) and you may very well have say 40% of WoW which would require a coefficient of friction of 0.3 for the same deceleration which may be not there on a wet runway. Ans remember, we are depending on reverse and drag providing 0.1g. Delay in the reverse or less than full reverse and the situation would be more critical. But it requires not only a long touchdown but also a very fast one.

          --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
          --- Defend what you say with arguments, not by imposing your credentials ---

          Comment


          • #20
            Originally posted by Gabriel View Post
            Delay in the reverse or less than full reverse and the situation would be more critical. But it requires not only a long touchdown but also a very fast one.
            Thanks for the detailed answer. One thing you didn't touch on is antiskid. I think on a slick wet runway (not ice or standing water) WoW in combination with anti-skid is going to make a significant difference at higher rollout speed.

            As a reference, recall AA2253 a 757 that overran at Jackson Hole in light snow. The final report concluded that the lack of ground spoiler deployment had a significant effect on landing distance. The most recent runway friction report was .43 .43 .36. Those are in friction MU units in sections of runway (first third / middle third / final third). MU ranges from 0=lowest possible conditions to 1=best possible conditions.

            Two scenarios came to me:

            - After landing and slowing down somewhat, they became concerned about runway remaining and tried to lift off again. That cliff at the end is probably a great motivator. Perhaps they got airborne but immediately stalled and pancaked it a la Turkish 1951. This would explain the passenger report of acceleration and the cockpit thrust settings. The speedbrake lever will automatically reset to the down position when the plane becomes airborne again. Also, there doesn't appear to be a track down the hillside left from the plane sliding down.

            - Perhaps the 737NG is prone to the same fault that caused the AA2253 overrun: an interaction with the air-ground sensors that, in rare conditions of timing, can cause the reversers to lock out. The action to take in this case is to first fully stow the reversers and then redeploy them. The result is a late reverser deployment, but in this case, perhaps they were not redeployed, or perhaps the reverser jam led them to initiate a late go-around that went south. Or perhaps the reversers failed to deploy for some other reason and the crew restowed them while working the problem.

            Food for thought (buffet).


            Comment


            • #21
              Originally posted by Evan View Post

              Thanks for the detailed answer. One thing you didn't touch on is antiskid. I think on a slick wet runway (not ice or standing water) WoW in combination with anti-skid is going to make a significant difference at higher rollout speed.

              As a reference, recall AA2253 a 757 that overran at Jackson Hole in light snow. The final report concluded that the lack of ground spoiler deployment had a significant effect on landing distance. The most recent runway friction report was .43 .43 .36. Those are in friction MU units in sections of runway (first third / middle third / final third). MU ranges from 0=lowest possible conditions to 1=best possible conditions.
              You are right, I intended but forgot to mention antiskid.

              Antiskid as you know is basically the same than ABS in a car. It helps exactly zero in normal circumstances when you don't need the maximum brake performance that can possibly be extracted, it doesn't maximize the braking action that can be obtained in any circumstance (meaning that you could brake as much with it as you can without it), but it makes extracting that max brake performance much easier than without it. Without antiskid or ABS, you need to manually gauge the brake capacity to brake as much as possible but without skidding, which is very difficult especially in a moment of startle or stress. With antiskid or ABS, if you need to brake as much as possible, the procedure defaults to simply stomp on the brakes as hard as you can, in other words the "panic brake" becomes the most effective emergency braking technique.

              But, the antiskid cannot extract traction that is not there to begin with. If for a given landing you calculated that you need autobrakes 2 and so set it up and the antiskid starts to do its trick, then you are in deep sh*t because you are not getting the deceleration rate commanded by autobrakes 2, and changing to autobrake 3 or MAX will not make things better (or worse), so you better apply full manual brakes in case you extract a good patch of runway and you can extract more than autobrakes 2 to compensate for the lacking patches, apply full reverse (normally less than full reverse is applied in normal landings) and keep the full reverse there all the way to the stop or until the point where stopping without reverses is guaranteed (instead of moving to idle reverse at 60 knots).

              I agree with you assertion that, in otherwise normal conditions, the mid-range of friction coefficient is where spoilers make the most difference (if you are tight on stopping margin). Extremely low friction and the braking will be extremely poor with or without spoilers, and in the opposite extreme in a clean dry runway you are not going to use all of the available traction anyway (landings are NEVER calculated with full braking authority on a dry runway because that's too much braking, too hot brakes, deflated tires, risk of wheel fire, etc...) so you can compensate a lower WoW with using up more of the available friction.

              Now, did I say otherwise normal circumstances? Land way too long and way too fast (beyond the variation taking into account when calculating landing distance) and you may need to use more brakes than anticipated, and there the spoilers can come very handy even in a dry runway (and yes, be ready for overheated brakes and deflated tires).

              But was that the case here? I don't think so.
              You brought the Jackson Hole accident. Jackson Hole runway is 6300 ft long and that includes 1500 ft of "airborne" distance (descent from 50 ft and flare).
              If your required landing distance was exactly 6300 ft, that means that Boeing demonstrated that you can descend from 50ft, flare, touchdown and stop in 60% of that, which is 3780 ft. Since 1500 ft of that is airborne, that leaves 2280 ft of actual landing roll distance to stop.

              But Air India had 6400 ft of tarmac+overrun distance AHEAD of the point where they actually touched down. That's why I think that not having the spoilers deployed is not enough by itself to explain not being able to stop. Something else must have gone on there. Something to do with go-around becomes a candidate given the pictures of the thrust, reverse, spoilers and flaps levers. But what? And aborted go around seems unlikely since, at minimum, you would retard the thrust levers and apply reverse. The opposite? Trying to stop and seeing that for some reason you are not going to make it initiate a go around too late for it to be successful? As you said, I would expect a much more disastrous outcome.

              So, as I said before, I am confused. I don't have a satisfactory speculation at the moment.

              Two scenarios came to me:

              - After landing and slowing down somewhat, they became concerned about runway remaining and tried to lift off again. That cliff at the end is probably a great motivator. Perhaps they got airborne but immediately stalled and pancaked it a la Turkish 1951. This would explain the passenger report of acceleration and the cockpit thrust settings. The speedbrake lever will automatically reset to the down position when the plane becomes airborne again. Also, there doesn't appear to be a track down the hillside left from the plane sliding down.
              I don't think so (but then, at this moment I don't think anything). I would have expected either a much greater destruction and death or a longer slide. Turkish slid hundreds of feet (while breaking apart). This plane stopped just next to the clift.

              And.... there are tracks, at least in the lower portion of the hillside. If the plane had stalled and pancaked, I would expect the landing gear to be ripped off at once upon touchdown, not leave a track. http://avherald.com/img/india_expres...e_200807_7.jpg

              - Perhaps the 737NG is prone to the same fault that caused the AA2253 overrun: an interaction with the air-ground sensors that, in rare conditions of timing, can cause the reversers to lock out. The action to take in this case is to first fully stow the reversers and then redeploy them. The result is a late reverser deployment, but in this case, perhaps they were not redeployed, or perhaps the reverser jam led them to initiate a late go-around that went south.
              I have that problem with them crashing while intending a go around.

              Or perhaps the reversers failed to deploy for some other reason and the crew restowed them while working the problem.
              Restowed them and lowered the spoilers and selected a small flaps setting? (and assuming that the thrust levers moved forward by themselves during the crash)

              --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
              --- Defend what you say with arguments, not by imposing your credentials ---

              Comment


              • #22
                Originally posted by Gabriel View Post

                (landings are NEVER calculated with full braking authority on a dry runway because that's too much braking, too hot brakes, deflated tires, risk of wheel fire, etc...)
                etc as in brake fade. I wonder if that played a role here. I also think this crash is going to be due to a number of converging factors.


                Restowed them and lowered the spoilers and selected a small flaps setting? (and assuming that the thrust levers moved forward by themselves during the crash)
                I can't make out the flaps lever position. As I said, the spoiler lever will reset to DOWN when there is no longer weight on wheels. That might occur as a result of the crash. Or perhaps the levers moved from their lower mechanisms being impacted.

                Comment


                • #23
                  Originally posted by Gabriel
                  Blah, Blah, Blah * 10^X (For large X)
                  Are you saying they botched the landing AND a go-around attempt.

                  Almost all landings are perfect for typical pilots...until someone botches something.
                  Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                  Comment


                  • #24
                    Originally posted by 3WE View Post

                    Are you saying they botched the landing AND a go-around attempt.

                    Almost all landings are perfect for typical pilots...until someone botches something.
                    I am saying that I don't understand what might have happened (even at the speculation level) but it doesn't seem to be your garden variety overrun.

                    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                    --- Defend what you say with arguments, not by imposing your credentials ---

                    Comment


                    • #25
                      I spoke to a coworker who came from India, I don't know where he got the info, but he said he heard the plane was hydroplaning once they touched down and couldn't get enough control to take off again. He was also saying based on the weather conditions the pilots did not have a good view of the runway and were likely affected by some sort of "tunnel" phenomenon. They were doing a visual approach. He also said the airline policy was that they had to make 3 attempted landings prior to diverting to a new airport.

                      He also said it appeared that the pilot intentionally turned the plane sideways just before going over the edge which broke the front part off (sacrificing business class and pilots) and saved the rest of the plane as it more gently fell down the incline.

                      I have no idea what that is all based on but he was pretty confident in it.

                      Comment


                      • #26
                        Originally posted by Schwartz View Post
                        I don't know where he got the info...
                        I'm guessing from someone's imagination.

                        he said he heard the plane was hydroplaning once they touched down and couldn't get enough control to take off again.
                        I'm not sure what this could possibly mean. Lateral control was obviously there or they would have run off the side of the runway. Vertical (pitch) control is always there as long as the elevators are responding. Thrust is all that's needed beyond that to abort a landing.

                        He was also saying based on the weather conditions the pilots did not have a good view of the runway and were likely affected by some sort of "tunnel" phenomenon.
                        "Tunneling" is a term used to describe a phenomenon where the pilot is so concentrated on completing a task (a landing, for example) that other vital areas of concentration are lost.

                        They were doing a visual approach.
                        In poor visibility and challenging weather conditions with ILS available? Why would they do that?

                        He also said it appeared that the pilot intentionally turned the plane sideways just before going over the edge which broke the front part off (sacrificing business class and pilots) and saved the rest of the plane as it more gently fell down the incline.
                        This makes zero sense. The forward fuselage wreckage is directly ahead of the surviving section. Nothing turned.

                        I have no idea what that is all based on but he was pretty confident in it.
                        That wouldn't surprise me.

                        Comment


                        • #27
                          Originally posted by Schwartz View Post
                          He was also saying ... they were doing a visual approach.
                          If true, this would likely be a critical factor in understanding this accident. Looking for confirmation. Any evidence that key equipment was out of service at CCJ at the time?

                          Comment


                          • #28
                            Originally posted by flashcrash View Post
                            If true, this would likely be a critical factor in understanding this accident. Looking for confirmation. Any evidence that key equipment was out of service at CCJ at the time?
                            http://avherald.com/h?article=4daf960f&opt=0
                            Kozhikode's runway 10/28 is 2845 meters/9330 feet long and features ILS approaches for both runway 10 and 28 as well as VOR approaches to both runways. In addition a NDB approach to runway 10 is published. Neither of the approaches have been NOTAMed unavailable.

                            --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                            --- Defend what you say with arguments, not by imposing your credentials ---

                            Comment


                            • #29
                              NDB approach! Now that is a f*****g emergency right there!

                              Comment


                              • #30
                                Originally posted by flashcrash View Post

                                If true, this would likely be a critical factor in understanding this accident. Looking for confirmation. Any evidence that key equipment was out of service at CCJ at the time?
                                This is what my colleague said:

                                "It had come out in the inital DGCa finding after speaking to ATC personal which was then told to reporters 2 days after the crash ... they said they still needed to confirm the configurations at time of landing and listen to voice recording."

                                Comment

                                Working...
                                X