Reproducing an incident: British Airways B744 at Johannesburg 11 May 2009

[James H]

G'day,

I have been testing my mettle using PSX to simulate various real-world system failures. I stumbled across an incident, with a happy ending, involving British Airways BA56 (G-BYGA) from Johannesburg to London on the 11 May 2009.

In summary, during the takeoff roll, engine 3 (before V1) and engine 2 (after V1) erroneously indicated thrust reverser deployment. As the aeroplane was still on the ground, the system logic automatically retracted the leading edge group A flaps, causing stick shaker activation and significant buffet shortly after becoming airborne. The crew held off the stall long enough to retract the landing gear, causing the leading edge flaps to deploy again. The aeroplane was then able to resume normal performance and return for a safe landing.

http://www.skybrary.aero/bookshelf/books/1319.pdf

To reproduce this in PSX I assume I'll need to do two things:
1) Trigger an erroneous reverser indication.
2) Retract the leading edge group A flaps (Boeing changed the retraction logic after this incident to prevent it ever happening again).

Does anyone know how I would be able to do this?

Many thanks,

James.

[Hardy Heinlin]

Hi James,

interesting case. PSX can simulate the errouneous REV indication by a faulty reverser unlock signal; this can also be simulated on full flight sims, but I don't know if this fault should also affect the flap control units. I've seen a description of the fault on the instructor screen of a Thales full flight sim and that doesn't mention any flap effect, just a faulty REV indication.

I would consider adding this flap effect to the REV fault model in the next PSX update, but, as you wrote, Boeing has now introduced a protection anyway, so I leave it as it is.

So, in this historic scenario, you would have to retract that flap group by another trick. Unfortunately, this can not be achieved by the malfunction controls. They can only stop the flap motion.

One could edit a situ file that starts at the point in time where the malfunctions have just occured, i.e. group A already fully retracted, and flap motion suppressed, and wrong REV indication activated. However, retracting the gear will not fix this problem; one will have to push a button on the Instructor to fix it.

Only an add-on could control this specific malfunction scenario -- by injecting the respective values into PSX.


Cheers,

|-|ardy

[Hardy Heinlin]

This fault is in the QRH under "REVERSER UNLOCKED":



Condition:
REV annunciation displayed with reverse thrust not intentionally selected.

With no yaw, loss of airspeed, or buffet:
Operate engine normally.

With yaw, loss of airspeed, or buffet:
FUEL CONTROL SWITCH (Affected engine) . . . . . . . . . . . . . . . CUTOFF
etc.
etc.




There is no comment re flap retraction. So I think the flaps are not affected in the sims by the REV indication fault model.


|-|

[Avi]

My manual (from 2006) says about this:
Group A (LE) flaps retract with the aircraft on the ground and when engines 1 & 4 reversers are both deployed or engines 2 & 3 reversers are both deployed.

According to this the LE slats should not retract in flight (anymore).
In PSX the slats retract when the aircraft is on the ground and engines 2 & 3 reversers are deployed (but not when engines 1 & 4 reversers are deployed).

Cheers,

[Hardy Heinlin]

Strange. In my code I have 1 & 4 as well:

Code Select Expand


if ( ( acc1.revPhaseInd > 0 && acc4.revPhaseInd > 0 )
  || ( acc2.revPhaseInd > 0 && acc3.revPhaseInd > 0 ) ) {

  ...

}


... and it works on my system.


|-|ardy

[Avi]

Not strange at all and I'm sure it works on my system too (even without re-checking it).
I started a situation where the aircraft was ready for takeoff. I wanted to see the extended flaps position so I depressurized hydraulic system 1 and then moved flaps from 20 to 10 and then conducted the "test".

The problem was that I used an El-Al model (PW engines) which needs hydraulic to move the reverser. I didn't notice that reverser 1 never deployed because it didn't have hydraulic.

This is a second "big" mistake of mine in just 2 days (only here). That's a sign.

Cheers,

[Hardy Heinlin]

The sign that you're a human :-)

[James H]

Thanks for the reply Hardy!

Maybe it is time to try my hand at making an add-on. It would certainly be useful to trigger the precise sequence of events that happen in a real life emergency.

Best wishes,

James.

[Hardy Heinlin]

Your add-on could make a simple list where each list item has a UTC value or a countdown time, and a message string.

At the respective time, the associated text message is to be sent to the PSX server.

The list could be stored in a TXT file and could be edited for any purpose.

That's all.

I could tell you how a text message need to look like for a specific effect, e.g. for the effect of having certain leading edge flaps at a certain position.


Cheers,

|-|ardy

[John H Watson]

I would consider adding this flap effect to the REV fault model in the next PSX update, but, as you wrote, Boeing has now introduced a protection anyway, so I leave it as it is.

Precisely. The LE flap deployment/retraction is now independent of thrust reverser cowl position. It only looks at reverse lever position (10 deg from full down) and the logic in the FCUs (which includes air/ground signals). Power for the circuit comes from the respective T/R CONT CB on the P6 panel.

The previous incident with unintentional LE retraction (which involved a book/manual on the forward pedestal hitting some of the reverser levers with the forward levers advanced for takeoff) is now not possible after a modification to the levers' electrical switches. The reverse levers now have to be moved at least 10 degrees, which I recall is not possible due to mechanical locks on the thrust lever assembly (the reverse levers can be moved only a few degrees upwards with the forward levers away from idle). The mechanical locks have been on the aircraft since day 1.

Rgds
JHW

[Avi]

The LE flap deployment/retraction is now independent of thrust reverser cowl position. It only looks at reverse lever position (10 deg from full down) and the logic in the FCUs (which includes air/ground signals). Power for the circuit comes from the respective T/R CONT CB on the P6 panel.

Hi,

I'm returning to this thread because I think there is still a problem in PSX.
If Boeing changed (and apparently they did) the logic and now they look at the reverse lever position and not the actual reverser position, then in my test (raising reverser levers 1 & 4 with PW engines with no HYD #1 pressure) I should have seen the LE group A retract (on the ground with the engine's respective T/R CONT cb powered).

Cheers,

[Hardy Heinlin]

Hi,

that's right, PSX still looks at the actual reverser positions. Boeing didn't send me any mod instructions :-) When did they change it? In 2011?


Cheers,

|-|ardy

[Hardy Heinlin]

Now in 10.0.5-beta7 (item 0.5.0035):

http://aerowinx.com/forum/topic.php?id=2939

[John H Watson]

When did they change it? In 2011?

2009~2010

For reference, the signal path for PW engines is more complex and goes through the respective T/R CONT CB, Fire Handle, air/ground relays listed below, thrust reverser lever, relay in P414 and then to the FCUs.

Applicable air/ground relays for PW: R7363 (Eng 1), R7365 (Eng 2), R7366 (Eng 3), R7364 (Eng 4).

The FCU's on all aircraft have logic which includes:
1) additional air/ground logic
2) Eng Rev (1 AND 4) or (2 AND 3) logic
3) 0.5 second delay on retract/ 5 second delay on re-extend
4) No action if flaps are in Electric mode
5) other logic which I haven't figured out yet   :mrgreen:

Rgds
JHW

[Hardy Heinlin]

I'm using the same "in air" relay combination that the FCUs use; this involves 3 CBs (F22, F23, H8) and it's so safe that it's impossible to get ground mode by electrical power faults.

Each of my "lever > 10°" check requires the respective T/R CONT CB to be powered. By the way, this condition corresponds with the reverser triggered high idle activation in the existing versions. So those flaps start retracting at the same time when low idle goes to high idle.

The other conditions you mentioned are already included as well, except for the fire handle check on PW engines. Why isn't the fire handle checked on the other engine models?


Cheers,

|-|ardy

[John H Watson]

Each of my "lever > 10°" check requires the respective T/R CONT CB to be powered. By the way, this condition corresponds with the reverser triggered high idle activation in the existing versions. So those flaps start retracting at the same time when low idle goes to high idle.

The high idle (CF6) logic hasn't been affected by the flap modifications. Reverser triggered High Idle is still dependent on the old stuff (T/R Cont Eng X CB, air/ground, >10deg lever angle and reverser cowl position)

Why isn't the fire handle checked on the other engine models?

I'm still trying to figure that out. The Chaper 26 Schematics (PW) for the fire handle say that pulling the fire handle will actually stow the reversers. I don't see any time delays in the fire handle engine hydraulic shutoff valve circuits, so I don't know how hydraulics could get to the reverser to stow it. Can the ADP provide hydraulic power to the reverser on a PW4000?

[Will]

I just read the full report on this incident - very intersecting. I expected the authors would at least comment on the decision to continue the takeoff, but they did not. (The first amber EICAS reverser indication occurred at 125 knots, while V1 was 150 knots.)

I do understand the section of the QRH about the reverser deployment in flight; you look for a degradation in aerodynamic performance and only if that is present are you directed to shut the engine down. I also understand that as the aircraft accelerates, pilots will only abort a takeoff for serious issues, like a fire infection, or engine failure, etc.

But would anyone have faulted the crew for seeing the REV indication and aborting before V1? In retrospect, a timely abort would have prevented them from taking off without the leading edge flaps positioned incorrectly. And since it was before V1, there should have been enough runway to stop safely.

Real-world pilots, if you saw the REV indication on takeoff, past 100 knots but a few seconds before V1, would you fly? Or stop?

[cagarini]

I can't answer Will ... My glider doesn't even display V1 :-(

It does have Rev though, in the form of a drag chute that equips the Phoebus B :-) If it deploys, has has happened at least once in our airclub, you don't even have time to consider aborting :-)

[Britjet]

Will - that situation can't happen anymore, but even if it did - my answer would be No.
If there was no other indication, such as an EPR change or obvious loss of thrust - that is only one parameter and could be a false warning.
I know the pilot (co-pilot actually) who flew the climbout. It was a very traumatic event for all. He was (is) an excellent handling pilot (a display pilot), and many thought that the aircraft should have been unflyable. Not him..

Peter.

[Will]

So here's a CRM question. Let's say you are the captain and you are also the pilot flying. The first officer is monitoring.

V1=150 kts.

Accelerating through 125 kts, the F/O notices the REV indication on the primary EICAS. As captain and pilot flying, do you want to know about it?

We have a situation where the pilot flying may not see the indication, but has the information about aircraft handling. Meanwhile, the pilot monitoring sees the indication, but doesn't know how the aircraft is handling. There is plenty of time to talk about it if it happens enroute, but as captain performing the takeoff, would you want the F/O to tell you about it at 125 kts?