Gear Collapse at Schiphol: A Perfect Storm
What do you get when you combine a popular airport for plane spotters with a stormy forecast and a crosswind on the runway?
The perfect storm.
On the 21st of February 2017, the Met Office issued yellow and amber warnings for wind, snow and rain as Storm Doris began to gather steam. On the night of the 22nd and into the morning of the 23rd of February, Storm Doris underwent “explosive cyclogenesis”, when the central pressure inside an area of low pressure falls at a very rapid rate, over 24 millibars in 24 hours. A jet stream high up in the atmosphere can remove air from the column, which reduces its weight and causes the pressure at sea level to fall. This sucks in air from the surrounding regions which results in faster and faster rotation of the circulation, in the same way that ice skaters spin faster by drawing their arms in. The effect is known as a weather bomb.
Warnings of the weather bomb with its violent winds spread across the UK and Northern Europe as the centre of the storm moved through Northern Ireland, across northern England and out into the North Sea. There were delays and cancellations to flights throughout the day.
Meanwhile, Flybe flight BE1284, a DHC-8-402Q (Dash-8) departed Edinburgh, Scotland at 14:21 UTC, bound for Amsterdam-Schiphol International Airport. An hour later, it entered a holding pattern over the North Sea off of the western coast of the Netherlands. Schiphol airport had struggled with strong winds all day.
15 minutes later, the flight crew positioned for an approach to runway 22, touching down at 15:54. As they landed, the wind was recorded as 31 knots from 240º but the wind direction was varying between 210º and 270º and gusting up to 46 knots.
As the aircraft rolled down the runway at speed, the right-side landing gear folded in on itself. The right wing-tip scraped the ground and the aircraft came to a halt. The flight crew made a mayday call and evacuated the cabin immediately.
This would be pretty much a non-incident except for the spectacular coverage of the landing from all angles. And I really do mean from all angles. Within a few hours of the incident, videos from the front and back and inside were online.
Here’s the aircraft coming in (skip to 1:50 if you just want to see the landing):
Here’s the view from behind:
And another view from behind, focused right in on the collapse.
This video is taken from inside of the aircraft:
Here’s the Mayday call made after the landing:
Here’s the fire trucks rolling out:
And finally a video of the evacuation:
Pretty much the only thing that’s missing is a go-pro style video from underneath the right wing!
So what happened? A Dash08 pilot commented on the landing on PPRuNe:
The DH8D is generally considered not an easy aircraft to land. Reasons for this are e. g. the large speed spectrum between plain Va and Va+20 on short final, the differences in approach pitch both due to these speed differences and different flap settings (with Va and flap 15°, a pitch of 2-3° is not untypical while flap 35° and Va-ICE may show a pitch of -3 to -4°, both in a stabilized final approach). Also the flap setting makes a world of difference: while with flap 15°, drag is rather low and an early (ish) power reduction is appropriate, flap 35° will not only result in some background buffet but also lend the aircraft a drag coefficient of a proper Amish barn. Careful handling of the power levers is required. Also the prop RPM makes some difference; speed control is much more responsive with higher RPM than with lower (compare a car at 1st speed vs. 4th). And to top it up, a pitch at touchdown of -0,5° may result in a nosewheel landing while from +5°, you are in tailstrike territory. Not much wriggle room there.
Bleeding off the speed is therefore not so much a function of the flare height but of power lever handling; the same can be said for the length of the flare, although here the height comes into play as well. Using power to break the descent works on the type as well but is sternly frowned upon by my company at least due to the unpredictable influences on landing performance.
I find that starting a flare in 30-50ft, simultaneously raising the nose and slowly reducing power as needed, thereby assuming the desired pitch for landing and adjusting the sink rate via power, works rather well on this aircraft and results in power-off touchdowns, some soft(ish), some more noticeable.
All this is of course fine and dandy in theory, but then along comes such a weather as the colleagues encountered at AMS. The most thought-out and well flown roundout and flare can quickly go south when a nasty gust shows up at the right time.
Maybe the airlines aren’t crazy about this kind of grass-roots coverage. Let’s face it: it’s never good to have a picture of your broken aircraft going viral. But I have to admit, I think it’s amazing! It’s almost as if I was there! Although I’d prefer to be sitting on the grass with my binoculars rather than in the cockpit for an incident like this. :D
The Dash 8 is a type of aircraft I have never flown, the pilot who does gave a clear enough explanation of the techniques that work.
From the video I got the impression that the aircraft made a firm touchdown using the technique.of holding the upwind down, rather than bringing the nose in line with the runway from a wings-level “crab” at the last second. The wing-down technique works well with aircraft that are not restricted by ground clearance between the underswung engines and the hard concrete as is usual on most airliners, like Boeing, Airbus.
Keeping the wing low is more suitable for lighter aircraft. The bigger jets have sufficient momentum, in strong, gusty crosswinds a lighter aircraft may be pushed to the side of the runway unless the pilot puts it on the ground immediately after kicking off the crab. During a wing-low crosswind landing, in fact a modest side slip, the pilot can make small corrections all the way until touch-down. And at big airports there should be enough runway, that amount of latitude is available.
Looking at the various clips, it would seem that the pilot did a good job. He touched down, upwind wing down. Even though it looked like it was a firm landing, it did not look so hard as to cause the landing gear to collapse. The circumstances would have dictated a firm landing. With the storm force winds that came on “Doris Day”, a positive landing would have been mandatory anyway.
By the look of it, the collapse occurred right on touch-down. Very unusual, the undercarriage of modern aircraft is very sturdy.
I just wonder if there had not been a weakness in that landing gear already? Looking at the videos, the collapse did not make sense. I would be inclined to say that it should have been able to take the strain of a slightly hard landing.
I have made many crosswind landings myself. Not all were within “approved limitations”. Being retired, I can now own up to it: no 4, Fokker F27 Farnair Europe, Kassel, crosswind component 32 kts, 3. Aerospatiale SN 601 Corvette EI-BNY. Dublin, crosswind component 37 kts., 2. Cessna Citation 550 N121C, Prestwick component 43 kts, 1. Shorts 3-60 Aer Arann, Dublin, crosswind component 42 kts. All 4 landings “uneventful”. The undercarriages did not collapse.
Two of Flybe Dash 8s had gear collapses so it does make me wonder if there’s a fault or a maintenance issue there. I certainly agree that the landing didn’t look very hard to me although it does seem clear that the right gear touched first.
I did smile at your list. :D
Error, no. That was SAS. Due to poor maintenance.
I’m surprised by all the evacuated passengers with carry-ons.
I thought it was rather low compared to some other evacuations. Although as it wasn’t a rush, we don’t really know what they were told.
Sylvia,
I had another look at the video and you are right, the landing is best viewed on the clip from behind.
The right gear touching first is quite normal. There are, as I explained, two techniques. Larger aircraft, especially with underswung engines, come in “crabbing” and at the last moment the PF kicks the drift off and aligns the aircraft with the runway. The momentum of the big birdies will prevent it from being blown across the runway. A wing-down landing will most likely result in an engine scraping the runway.
In smaller aircraft, the wing-low technique is often used. The aircraft in fact comes in side-slipping. The wing into the wind will be held down so a component of the lift will compensate for the drift. This can be held right through the flare and touch-down. The pilot will retain full control during the landing which will be on the upwind gear initially. A strong crosswind landing in the Corvette especially was very controllable. The wheels were a bit close together but once in the flare the ground effect prevented the wing tips – virtually all were fitted with tip tanks – to scrape the deck. It was the best handling small jet that I have flown. Initially it had problems with the flaps until a modification solved it. In the end we tackled even short runways (1200 metres) flapless. Aerospatiale gave as crosswind limit 37 kts. When I asked the instructor, Capt. Briot, a test pilot, he told me that during the certification process there had not been more, “so maybe one day you will handle more crosswind, you will write to us and we will amend the manual”.
The Citation was another one that handled well in crosswinds. It had no published limits, the manual mentioned 25 kts “demonstrated”. Based at Prestwick we had frequent high winds and eventually demonstrated 43 kts in a very sophisticated simulator, with engine failure at V1 on a short (1000 m.) runway.
I lost control on take-off, my colleague managed to get airborne (N-1 !) but he could land it. I managed, so in the end we set a limit for ourselves of 40 kts. This was certified by the training institution (Simuflite) and accepted by the FAA.
This meant 40 knots directly side-on and in gusty conditions. Our instructor soon stopped the session because she started to feel sick – in the simulator.
The Shorts was on a cargo flight, at night. The airport authorities had closed runway 28 for maintenance and the wind was reported as being from direction 270, gusting up to 43 kts. The only runway available was 17. The main problem – of which I was aware and briefed – was that the ailerons had to be fully held into the wind. It required both of us, but otherwise the slab-sided 3-60 could be blown on it’s side.
Back to the Dash 8: The landing looked a bit hard, but considering the circumstances not excessively so. The wind came from the right, the pilot used the wing-down method and so he correctly touched down right gear first. I do not see why it collapsed, I don’t think that the landing was that hard. In fact it looked nearly normal to me (except for the gear collapsing of course).
I’m sorry, I laughed out loud at the instructor feeling sick in the simulator. That’s some pretty impressive flying! Your comments always make my day, Rudy, thank you.
There’s something strange about these videos!
In the second one, the aircraft FIRST touches heavily on the RH undercarriage which promptly fails.
In the third one, it touches first on the LH gear (gently) theh kn the RH gear on what appears to be a perfectl executed gentle landing!
What is going on?
OK, Maybe it’s just me that didn’t realise that these videos are of different incidents. One an obviously heavy landing, the other a perfectly satusfactory normal landing but perhaps the undercarriage had previous damage or fatigue.
Hmm, have I missed something? Sadly, I’m in the process of moving and won’t get proper internet access for another week, so I haven’t checked the videos again but I will.