Air North Brasilia Crash in Darwin (Merged)

[Harry Cooper]

Most aircraft won't allow you to come into beta unless there is Weight-on-Wheels. I have flown some types where with age squat switches fail and you can come into beta early - not a good position to be in if beta "deploys" assymetrically, you will be well and truly screwed.

I agree totally with you in the other respect. Practicing a V1 cut or EFATO by pulling the power lever to FI increases drag and increases Vmca by a substantial margin. Obviously Zero thrust is used to simulate either the engine secured or initially at least autofeather operation. At light weights there could be only a small margin then between say V2 and Vmca. Again I use the B200 as example. The new models are certified to Part 25, meaning they run to transport category numbers. At max weight V2 is 111 KIAS. If autofeather was to fail, then Vmca will rise from 91 KIAS to 108 KIAS. Thats a 3 knot margin between the them - f**k all. Everybody here is going on about gradients, windmilling etc but bottom line is that if these aircraft are certified in the US then they do not need to meet these requirements in their basic form. If they have a shortcoming modifications such as MANDATORY autofeather can be employed and performance credits are then given, but these have to be functional at all times they are required or else the aircraft is not airworthy. Obviously some aircraft such as the Dash are so overpowered that this is not an issue. For some such as the B200 it is. The autofeather must be tested each day and the aircraft is grounded if it does not work. I believe the Bras is the same. It would be interesting to find out what a "light" V2 would be for a Bras as well as Vmca in the event of autofeather failure. I'm sure it would be tight. Obviously the only way to recover from a Vmca departure is to reduce power and lower the nose to increase airspeed. Who the hell is going to do this at a few hundred feet AGL as their aircraft is rolling and descending - I doubt I would.

[john_tullamarine]

Some thoughts .. (caveats - no free turbine piloting experience, only F27/L188 and not speaking specifically to this particular aircraft loss)..

Many have disputed this from their own hands on experience, yet some continue this "windmilling prop / no climb theory".

The certification basis will give you the starting point for figuring what might or might not happen given this or that problem.

In general, if the weight is (appropriately) low enough, there ought to be no great difficulty climbing with a more adverse than certification thrust/drag situation .. not much different, really, to the similar problems with light twins in a takeoff engine failure situation ... hot and heavy = forget it ... light and cold = far less sweaty.

The principal flight management problem with this sort of training manoeuvre lies in handling qualities and consequences of uncommanded/unexpected failures at a fairly critical stage of flight. If a failure-related problem occurs, say, around or especially a little below V1 with a near Vmcg/Vmca limited speed schedule, and the takeoff is continued, then things might rapidly get out of the pilots' ability to keep on top of the situation. The propeller beast is far more critical, in this regard, than its turbofan brother.

So far as what the failed engine's propeller should do, generally we see the certified system's incorporating either autofeather or neg torque sensing to drive the failed engine's prop to a pitch condition resulting in a predictable (reasonably low) drag so that handling problems, initially, and climb capability, subsequently, are manageable and within the certification expectations.

In my type we brief to climb to acceleration altitude (400 feet AGL) with a windmilling prop (NTSing), T/O flap and gear retracted.

which is the typical certification scenario .. where the handling and performance is predicated on autofx/NTS correctly functioning

The only time we will feather the engine straight away is if we have an NTS failure in which it will happen as soon as the gear has been selected up.

.. (or autofx failure, as appropriate), either of which can be expected to result in a critical situation generally requiring fairly rapid pilot response to maintain control of the situation

There are many school of thoughts that it's ok to fail an engine before V1

.. provided you intend to stop .. especially if you are on a near min speed schedule

There's no reason the a/c cant be configured to simulate an EFATO at 2000 feet

.. perhaps not really a useful training consideration generally for heavies ... however, a good strategy for initial failure introduction to a multi engine new chum.

Who the hell is going to do this at a few hundred feet AGL as their aircraft is rolling and descending - I doubt I would.

.. been in a similar situation in the 727 sim engine failure with the FE being a bit slow on the old "check essential" thing (long time ago so I'm stretching the memory a bit) ... either way we ended up with no rudder for a bit and the thrust had to be reduced substantially to give us time to catch up with the FE's delay.

In the real world, it is a matter of discipline and training ... leave the power on you DO die .. inverted. Pull the operating levers back a bit and you MIGHT just live in the belly landing. You might well doubt that you would .. I doubt that I would not ... when you have departed controlled flight without a preprogrammed recovery, the normal rules go out the window and you are looking to maximise the probability for a survivable accident.

So far as training in the aircraft versus training in the sim, the sim is preferable if you have a reasonable fidelity device available .. you get to have the other guys smile and chuckle at your mistakes but you also get to have another go at getting in right next time around.

Having said that, some training ends up needing to be done in the aircraft due either to no sim being available or inadequate sim fidelity for particular requirements. I did all my F27/L188 training in the aircraft but with fairly rigorous adherence to published protocols. Even then, some silly things were done eg for my L188 endorsement - it was all done at night due to aircraft availability and with non-flying pilots standing in the cockpit observing. I shudder now at the likely headlines had something gone wrong .... however, we didn't do overly silly things like training at near min speed schedules.

[ace from space]

sound asleep

There are many school of thoughts that it's ok to fail an engine before V1 OR wait until 400 feet and give yourself some safety margin.
This was an accident, no matter how you look at it. I can potentially see CASA regs being ammended re this type of training - either to mandate sim training and/or limiting simulated EFATO to a min altitude.

Sound asleep.
Not sure what you mean there.
V1 occurs whilst the a/c is still well and truly fixed to the runway, prior Vr, so if one were to suffer a failure prior to V1 then stopping on the runway is still an option, after V1 it is not. (Unless you are prepared for the potential risk of over running the R/W)
EFATO is a term more applicable to light twins or singles and such, that don’t fit into Transport Category performance criteria.
In a Transport Cat a/c the failure could actually occur before takeoff (before Vr) and with training one learns to deal with the failure and continue the takeoff. Hence the obvious benefits of being able to do this kind of training in a sim.
I do agree with your points about doing this exercise (in the real aeroplane) at a safe height.
I for one have never felt the need to do this exercise at or just after V1 but rather at a more comfortable 100-200 ft AGL so that if anything were to go pear shaped there was some margin for error. This doesn’t give the same experience as suffering a failure at V1 and controlling the a/c on the runway whilst waiting for Vr to occur before rotating and maintaining control and stabilising the a/c at V2 and positive rate.
Call me soft, but I never felt the likelihood of ever suffering such a failure is worth the risk of carrying out this exercise at 0 ft AGL; the Tamworth Metro crash brought that truth home to me.


FGD135

This is alarming. You were on the verge of being out of control. It could be said that you *were* out of control - given your continued problem with directional control.
Is this a frequent occurrence on these sorts of training flights? Is CASA aware of this?

Being on the verge of losing control, just a small hiccup (e.g. wind gust) would have put you out of control.

With the rapidly developing yaw and roll towards the dead engine, what then is the procedure? Is there an officially documented one?

You said that you "decreased power on the good engine". Is it possible that somebody could instead try to increase power on the "failed" engine - with a long spool-up time then making a crash inevitable?

In response to your questions FGD135, I only experienced this situation twice where I felt we were at the limit of controllability during a ‘critical’ engine failure and put it down to over pitching or over rotating and low speed.
I discussed it with my colleagues at the time and also with the Airframe ‘ginger beers’ and no one else had had a similar experience or reported such. I recall the ‘ginger beers doing a control rigging check and full control operation check but reported that everything was within limits, so I put it down to poor technique and the vagaries associated with different airframes. (Some just fly straighter than others)
The technique to recover from such a situation would be to increase speed if it was low by lowering pitch attitude and decreasing Tq on the live engine (which someone already pointed out is the below VMCA recovery procedure).
If control loss was imminent one could quickly advance the ‘failed’ power lever as with a PW118 series engine, from FI there is very little power lag.

[ace from space]

relax737

good story about both pilots switching the tanks...
A couple of blokes I know who were both senior pilots (Capts) went flying together on a test flight and left the gear pins in....oops!
Human factors...
Btw the Alligator analagy is an oldie but a goodie....

[Harry Cooper]

John, I agree with all your points. My last point however? Witnesses report that the aircraft had just got airborne and then there was a rapid roll to the left. I have never briefed during my takeoff safety brief for a loss of control on departure due to Autofx failure and possible subsequent Vmca departure. I think with the height they had available to them and everything else going on even if they had recognised the issue, which maybe they had, it may have been too late. A 10 tonne airplane already carrying inertia toward the ground is going to be hard to stop with inadequate airspeed let alone height and controllability. Of course this is just hypothetical scenario based on my comments and the accident at hand.

[ace from space]

JT good post, well said

[ace from space]

Redline

The crew recalled that, as the pilot in command advanced the engine power levers at the commencement of the go-around, the aircraft yawed and rolled left ‘aggressively’. The co-pilot applied right rudder and aileron but was unable to control the aircraft. He informed the pilot in command that he was unable to hold the control inputs, so the pilot in command placed his hands on the control yoke and his feet on the rudder pedals and assisted the co-pilot. Together, they were able to steady the aircraft’s flight path, which they estimated at that time to have been approximately 100 ft above ground level, with 30 degrees bank left, and 100 to 111 kts airspeed. The crew reported that the stick shaker activated twice, and that each time they slightly reduced the control yoke back pressure. During that period, the EGPWS warning ‘too low terrain’ sounded. The co-pilot called for the flaps to be retracted to the ‘flaps 25’ position. After the pilot in command confirmed that the co-pilot had control, flap 25 was selected. The crew reported that the airspeed slowly began to rise above 111 kts and the left bank started decreasing, although full right aileron and rudder inputs were still required. The crew retracted the flaps
to 15 degrees, and the co-pilot was able to return the aircraft to a wings level attitude. The aircraft achieved a positive rate of climb a short time later. The crew then retracted the landing gear.

interesting post, must have been a very uncomfortable few minutes for those involved.

[Harry Cooper]

This is the aspect of this place that I like. Regardless of the sh*t fights that always ensue in these forums after an accident or major incident, once the forum gets rolling and experienced guys such as J_T start posting a lot can be learnt. If only one person walks away with new useful knowledge then these posts have served a purpose. If subsequently that knowledge saves lives down the track then accidents such as these, although tragic, are not in vane and forums like this are worthwhile. When you look at Australian aviation regulations and the background of how they came about (such as the Coffee Royal Affair), you see that these regs are written in blood and it is unfortunate that new regulations will be written (most likely) on the basis of this accident too. I think that this certainly won't be the last either.

[ace from space]

Harry

I agree totally with you in the other respect. Practicing a V1 cut or EFATO by pulling the power lever to FI increases drag and increases Vmca by a substantial margin. Obviously Zero thrust is used to simulate either the engine secured or initially at least autofeather operation. At light weights there could be only a small margin then between say V2 and Vmca. Again I use the B200 as example. The new models are certified to Part 25, meaning they run to transport category numbers. At max weight V2 is 111 KIAS. If autofeather was to fail, then Vmca will rise from 91 KIAS to 108 KIAS. Thats a 3 knot margin between the them - f**k all. Everybody here is going on about gradients, windmilling etc but bottom line is that if these aircraft are certified in the US then they do not need to meet these requirements in their basic form. If they have a shortcoming modifications such as MANDATORY autofeather can be employed and performance credits are then given, but these have to be functional at all times they are required or else the aircraft is not airworthy. Obviously some aircraft such as the Dash are so overpowered that this is not an issue. For some such as the B200 it is. The autofeather must be tested each day and the aircraft is grounded if it does not work. I believe the Bras is the same. It would be interesting to find out what a "light" V2 would be for a Bras as well as Vmca in the event of autofeather failure. I'm sure it would be tight. Obviously the only way to recover from a Vmca departure is to reduce power and lower the nose to increase airspeed. Who the hell is going to do this at a few hundred feet AGL as their aircraft is rolling and descending - I doubt I would.

Interesting point you make about the margins above Vmca for V2.
You are so right, margins are tight so good technique with speed control
Is important for maintaining control as well as achieving the desired performance.
I don’t recall Vmca for the Bras or the margin between V2, someone else may be able to help with that?
From my recollection of performance for FAR part 121; V2 = 1.1 Vmca or 1.2 Vs and doesn’t change whether auto fx is operative or not, as the prop is assumed to be windmilling in the engine out case.

[CharlieLimaX-Ray]

Okay PPrune experts explain how you can take-off in a transport category four engine aircraft(jet or piston) with an engine shutdown or in the case of piston engine with a failed starter motor prior to commencing the take-off roll, and still maintain directional contral and still achieve the climb gradients?

[aussie027]

I agree with some of the above comments. Very valuable when people post high value data that others can think about and learn from, it may save their butts one day.

I was told during basic flight training--- "Learn from the mistakes of others, you won't live long enough to make them all yourself".

Very true, I also have found, as have many others to pre think your actions in certain super critical situations so hopefully if the situation ever arises your actions will be immediate and correct. A second or 2 could make the difference.

Illustrated by this true story--

A navy carrier pilot in a fighter felt the cat shot was "cold", ie very weak , so his aircraft was basically moving just fast enough as it left the deck to go down into the ocean right in front of 100 000 tons of carrier moving at 30kts.
He ejected just as his plane left the deck.
Later, when someone asked him when he made the decision to eject, he said, " about 10years ago".

[ace from space]

Archangel7

for such a sophisticated aircraft like that to suddenly lose control on take off with 2 highly experienced pilots for no apparent reason makes no sense. Even if they where conducting EFATO drills this shouldnt happen. The Brasilia has gauranteed performance with one engine on T/O... Obviously Something terribly went wrong.This could not be pilot error. Without a doubt, mechanical failure!!! To me it sounds like the malfunction of the engine propeller control unit and this would allowed the propeller blade angles to go below the flight idle position if they where conducting EFATO drills. I am sure there are many other factors contributing to the accident as all accident have a chain of events to lead up to it.

like i said, its only my opinion and its not a big issue! :P i guess it is easier to just point the finger at the dead pilots and say it was "human error" . That would make me less arrogant and close minded i guess. I personally believe it is a mechanical error and you can call me as narrow minded and as arrogant as you want it will not change my opinion. But a high-speed crash on take-off points the finger more towards a mechanical error then human error. Do some research on Brasilia fatal crashes and you will discover that 90% of them are mechanical errors!. they vary from propeller going into reverse thrust on final approach, failure of the horizontal stabilizer and there was another interesting one which i read up on the other day where the propeller separated from the engine assembly and became imbedded in the leading edge of the wing and the pilot was committed to make a forced landing into a paddock killing 18 passengers. do some research and see for your self. I don’t think I am being closed minded at all.

Dude, you are persistent about this...
No one is saying that it is pilot error or a mistake by the crew, but you can’t definitively rule it out by making a big assumption like that.
Recommended reading ‘crash comics’ vol 1-3 (or otherwise known as Airline Disasters by Macarthur Job)
For me personally, I have a very healthy respect for the V1 cut, after experiencing an experienced colleague/sim partner losing control and crashing the sim during this exercise. A good learning experience that we got to walk away from as it was in a sim not the real a/c.

Background info for you.
Although you attribute some of the earlier Brasilia fatal accidents to mechanical failure, several of the earlier prop or prop governor related crashes were before the electrically activated ‘FI stop solenoid’ was fitted. Why was it fitted you may care to ask?
Because it was determined that relatively inexperienced crews on the type were pulling the power levers (particularly when hot and high) back over the mechanical stop, either advertently or inadvertently with disastrous consequences. To my knowledge no further related hull losses have occurred since this mod was carried out.

[Centaurus]

Later, when someone asked him when he made the decision to eject, he said, " about 10years ago

I'll go along with that. After years of reading accident reports from around the world where high speed aborts for whatever reason on limiting length runways often ended in grief off the end, I made up my mind that, unless absolutely sure the aircraft would not fly, then from 15 knots below V1 I was going to continue the take off and sort the problem in the air. Might have been procedurally illegal but that's the way I saw it. Still do, too. No further correspondence entered into

[t303]

OK FGD

t303, your account refers to simulating the engine failure by setting flight-idle. Flight-idle is not the same as windmilling. This is twice now that I have called you on this.

I mistakenly thought the argument was concerning drag, your concern being about a windmilling prop, lack of climb performance and all. I offered my experience in the aircraft with the gear out (lots of drag). My bad, so I'll fold.
Life's too short.

[LeadSled]

CL X-Xay,
Engine out ferry is a very specific operation, with it's own limitation and performance requirements.

The conditions for an engine out ferry are very type specific.

For example, on a DC-4, the prop. on the failed engine must be removed, a DC-6/7 can just have the prop. feathered.

Engine out ferry of a B707 was available, but few airlines ever used it, because it was virtually impossible to produce a V1/Vr/V2 that could cope with a further engine failure on T/O ----- the Vmca2 was so high, compared to any reasonable speed at which you could keep the aeroplane on the ground.

In the case of some B747, Vmca2 was low enough that a "survivable" set of figures could be produced. As I remember, it was normal to remove the fan and blank the core engine, much reducing the drag of the failed engine ---- ie: the failed engine configuration was the same as if the engine was a fifth pod.

Because of the very "iffy" nature of 3-Engine (or in the case of the DC-10/MD-11 2-engine) ferry, very few airlines ever exercised to option. Given that an engine change is a straight forward operation, and carrying a fifth pod is straightforward, or there are plenty of L-382 or AN-125's around, changing the failed engine is the much preferred solution, short of WW111.

Tootle pip!!

[Capt Claret]

CLX

As best I can recall the procedure in the BAe146, was thus:

1. Obtain specific RTOW for three engined ferry from Flight Ops
2. Confirm from RTOW for given atmospheric/aerodrome conditions that RTOW is ok.
3. Line up & with feet on brakes, set take-off thrust (N1) and mark thrust lever quadrant, reset ground idle.
4. Set takeoff thrust on symmetric engines and then release brakes.
5. At 60 kias introduce thrust on asymmetric engine to have take-off thrust set by 80 kias.
6. No V1 as such, Vr is considered V1,
7. Because only tech crew allowed, it usually climbs much better than with a load of pax & 4 engines.

To be current, we had to have been checked in the sim within the last 6 months. I did at least two that I can recall, ASP-ADL & TIM-DRW (Timika).

[CharlieLimaX-Ray]

Done at Airlines of Tasmania when a Heron had a starter fail motor fail at YKII or YFLI, three engine take-off approved(crew only) approved in the ops manual.

Likewise National Jet, Southern did three engine ferries on the BAE-146.

I have seen photos of the demo pilots taking a BN-2 Islander off with one engine feathered at the start off the take-off roll.

The original Aerocommander was flown across the USA on one engine with the non operating engine prop removed and tied down in the cabin.

[remoak]

Centaurus

from 15 knots below V1 I was going to continue the take off and sort the problem in the air. Might have been procedurally illegal but that's the way I saw it.

Hmmmm well I hope you briefed your F/O accordingly... or that could have resulted in some very ugly incidents...

Better to just artificially reduce the V1 if you are going to be doing stuff like that.

[obie2]

...or better still, go and ask the Chief Pilot to step down and give you the job so that you can amend the SOPs to incorporate your own stupid theories and go and kill everybody!!

[Shed Dog Tosser]

This has been one of the most interesting threads in quite some time, some great learning from respectful speculation.

Also glad the moderators cut the peanuts out of the thread, must be great to have a "ban" function on your tool bar.

Yes Smoka 21, I'm talking about you, loved your PM, thanks:

Subject: bout you

You Sir are obviously a neophyte and a grub!

Nothing more than a grub.

Let me know your name and phone number big man, and I personally guarantee I will end your career before it starts. Regardless of under which rock you hide.

You grub!

Good luck

All from a "man" that list the following as your profile:

Current a/c Type (eg B737. Pilots only):black smoke spewing noise machine
Biography:hot ******* dude
Location:bum**** alabama
Interests:oh wanking and stuff
Occupation:dole monkey

You are not my peer.

BBBbbbaaaaaaaahhhhhhhh hhhaaaaaahhhhh