Toronto place crash 4

[LittleInch]

A Delta plane appears to have touched a wing tip during landing, ripped the wing off then promptly flipped over onto its back.

As they were on the airfield and this time didn't run into anything or catch fire, everyone is alive, though not surprisingly some injuries.

All passengers survive crash landing as plane flips at Toronto airport

Eighteen of the 80 people on board were injured after the Delta plane overturned on landing.

www.bbc.co.uk

This video https://www.dailymail.co.uk/news/article-14407855/delta-plane-crash-toronto-fireball-footage.html makes it look like a very hard landing - no visible flare

 

[LittleInch]

I think you're into the right sequence myself. Pretty clear the wing started detaching before the gear collapsed in my view.

It was all over in less than a second but the gear collapse wasn't the start point. IMHO.

 

[Sym P. le]

The right wing tip remains visible for six frames including touchdown. That's 1/5th of a second from touchdown until the fuselage punching action catapults the wing tip upwards at a rate the video can't capture well. You can practically see the right MLG compress through those frames until it bottoms out at the sixth one and the wing tip gets launched. We already know it's over at that point.

Animated GIF

Video analysis (Youtube)

TSB report (PDF)

TSB report (Youtube)

CRJ-900 Promo (Youtube) View attachment 6976

 

[Sym P. le]

There's the bolt . . .

 

[Sym P. le]

An end view of the punched down rear wing spar in 3D (cross eye). The trailing edge spar, which is supported from the distal wing structure can be seen on the right, just below the spoiler root. A bracket spanning the landing gear bay is also visible attached to the trailing edge spar.

They look flummoxed. I hope they read this thread and stop wasting my tax dollars.

 

[TugboatEng]

Just a thought. The report now says the impact with the ground produced 3g of acceleration significantly exceeded the design loading of the aircraft. United flight 1722 from Hawaii pulled 3g due to an autopilot setting error and the wings stayed on. The pilots didn't even return to the airport!

 

[Sym P. le]

That would be 3g distributed evenly across 2 wings, 4 spars. This "occurrence" was the entire weight of the jet punching on one spar, so 4 times the load issued against a shearing device. I don't think its an apples to apples comparison. Having said that, the landing gear took a bunch of the load so not the full 3g was required to shear the wing.

I can just imagine the pilot conversation. "How much did the wings flex?" "20 feet" "Carry on"

 

[Alistair_Heaton]

Human performance wise looking at the Captain and thier background there is something nagging me.

Relatively low experience less than 4k hours since 2007.



The time line from the performance drop and reaction of the PF just doing nothing isn't normal. The trainer should have at least said go around. Personally I would have taken control, I believe but there could have been other things going on.

I feel a bit sorry for the young lady to be honest. This is a normal experience gaining event which should have been dealt with by the training captain.

 

[TugboatEng]

Juan Browne made an interesting observation that the pilot flying didn't have any critical observations on their record. It's as if their training wasn't very rigorous.

I brought up the 1722 story because 3g is apparently a serious incident. I do understand that in this case an single strut and wing had to provide all of the forces associated with the 3g deceleration. In the case of 1722, perhaps the pilots should not have continued the flight over an ocean.

 

[Alistair_Heaton]

That's called the maneuvering limit load tug.

Flaps up it's -1g to 2.5g

Flaps out 0 to 2g

This is into certification requirements.

There is a load of other limits in relation to turbulence. As pilots we don't as such have any knowledge we have bust them until an email turns up.

I have learned though to always penetrate cu clouds with wings level otherwise you will get a lateral g exceedance email. Once the air density, temp, humidity, top of cloud icing have been flown through you can turn.

Never really thought about it before on other types.

 

[Alistair_Heaton]

As for the comment on the pilot training record.

He seems to have a gut feeling as well about the training picture being shown being a factor. Personally I think it's just as likely a training department big picture issue than the individuals.

Again this is coming back to the oversight by the FAA being lacking.

 

[Sym P. le]

I obtained my private license when I was younger but didn't carry on with it. I recall later on doing some go-arounds with a friend who kept up his hours and gaining new insight on how a plane is handled during landings, just by watching him. The rigors of training regimes can create a stiffness but not necessarily a flow. Putting in my minimum time and achieving adequate skills to go solo or achieve a rating was just a start.

 

[Alistair_Heaton]

Yep spot on.

It's the flow of working with energy not against it.

The plane is much better at flying than pilots are. The trick is not to annoy her.

 

[LittleInch]

At the end of the day though I'm still astounded as to how a professional pilot with over 1000 hrs flying just didn't flare. At all.

The plane wasn't built for landing one leg with that sort of bang.

I don't fly aircraft, but I landed a whole load of flying wing canopies and if I got it that badly wrong I would have not been able to walk. Self preservation does a lot of good for you I can tell you.

 

[Alistair_Heaton]

They get fixated with one parameter

Also if the training environment isn't the best they just overload and freeze.

 

[Sym P. le]

Any thoughts on the deployed spoiler / rapid descent rate?

 

[TugboatEng]

As an off-road motorcycle rider, one of my favorite advices was, if in doubt throttle out. Whatever bad situation you're in, a handful of throttle and some acceleration is likely the best solution. This applies to all situations where your front wheel isn't already elevated.

Maybe that's the dirt bike equivalent of a go around.

 

[Alistair_Heaton]

Spoilers to me are a relatively new system.

Turboprops we used to use drag off the propeller to adjust energy state or rate of reduction. But it was all done through your hand on the power levers. And those levers were always manually operated.

Jets have autothrust and it will vary the thrust to maintain the set speed the pilot has targeted. This gets adjusted depending on the conditions as detailed in the report.

They will even automatically reduce the power to idle. The airbus retard call out.

If your speed goes high to maintain stabilise approach and thrust is min your ment to use spoilers to keep it max +20 knots from Vref.

So now people have two sets of controls to remember to use.

I suspect the PF was fixated on the speed being wrong, couldn't work out the fix. The fight,flight or freeze human reaction kicked in. And they hit the ground.

The instructor Captain should be fully aware of the normal mistakes trainees do, how to spot them developing, and how to break the error chain in a constructive manner to aid learning.

Problem with putting a hand full of power in tug is that you end up over energised and people go off the runway. Over the years the stabilised approach concept has been proven to work.

In this case they should have abandoned the approach.

 

[LittleInch]

They get fixated with one parameter

Also if the training environment isn't the best they just overload and freeze.

A part of me wonders if the Instructor forgot he was in a real aircraft and not the Sim where it didn't matter if you hit the ground at 3G.

3.5 hours flying in the last month isn't exactly current in my view.

It does look like the gust of wind caught them out, reduced power and never put it back on even though airspeed declined by nearly 20 knots below VRef.

For information the report says the crew "set the speed bug to 144kias". What does that mean to the rest of us?

 

[Sym P. le]

My question being, is the automatic deployment system capable of fully deploying the spoilers within 0.6 seconds after MLG ground contact or does the departed wing have a last breath of life as it slid down the runway?




Edit: This video (Youtube) shows they deploy rapidly upon contact. I'll download it and analyze it. This video (Youtube) suggests deployment seven frames after ground contact which could be after the rear spar was punched but prior to complete wing detachment.

A quick analysis of the first video shows the spoilers deploy at 5/30sec after td and requiring 16/30sec to fully deploy for a total of 21/30sec. Very fast indeed. Compare with incident footage, rear spar detachment at 6/30sec and complete wing detachment by 19/30sec. The second video suggests deployment at 7/30sec and 12/30sec to fully deploy for a total of 19/30 sec.

Interestingly, only one of the two "Ground Spoilers" is deployed. The right MLG is still attached to the wing.

Edit Too: The spoiler doesn't appear deployed as the detached wing sat on the runway.

 

[Alistair_Heaton]

The speed bug is the speed they are trying to target.

I don't know what automation systems the crj has.

You completely correct 3.5 hours is not sufficient to be current.

To be honest his whole total hours since 2007 rings alarm bells to me.