from PPRUNE Bulletin Board

Author Topic:   QF1..... What the numbers say.
Clocka
Experienced PPRuner
posted 30 November 1999 03:17   

The following should be read with the knowledge and context that the author is employed by QF:
Now that the BASI interim factual report has been published, we can stop speculating and start getting into some hard data about QF1. The report make makes for very interesting reading. For it seems that whilst the journalists fall over themselves to blame cost cutting for the QF1 accident, the real culprit may be getting away with it.
I am not talking about the crew or the now, much maligned management. I am talking about runway 21L at BKK.
The sequence of events that lead to the accident are clear.
The F/O was flying.
The crew had been informed that the braking action on 21L was "good"
During the approach both he and the Captain agreed that enough visual reference was present to visually complete the landing.
In the flare, heavy rain with resultant loss of cockpit visibility so concerned the Captain that he ordered a go-round.
As the go-round was commenced, the aircraft touched down and the aircraft flew out of the shower.
The Captain judged that sufficient runway remained to land and cancelled the go-round in order to achieve a landing.
The aircraft touched down at 168 knots.
The aircraft subsequently left that runway still travelling at 79 knots.

Nobody has made much mention of the fact that the aircraft even though it touched further down the runway than normal, should still have been able to comfortably stop in the runway remaining.

As a former –400 pilot I remember that the stopping power of the 747-400 Brake/antiskid system is awesome. Just asked any pilot who has looked up at V1 at places like BKK or LA during a Max Weight Take-off. The amount of runway left looks pitifully inadequate but when an abort from such a situation is performed in the simulator, the braking system handles it with aplomb. It is with that knowledge that I offer the following:

The following can be gleaned from the BASI report and the –400 performance manual:
Landing weight was approx 267 tonnes. (reference approach speed with Flap 25- 154kts)
Headwind component was approx 3 kts
Temp was not given but lets call it a balmy 25 degrees.
Touchdown was approx 3100 ft from the displaced threshold and 7400 ft- (2250 metres) from the far end of the runway.

Using the above figures on the 747-400 Flap25 Landing Distance Required chart yields a LDR figure of 2060 metres for a dry runway. 2060 metres on a dry runway equates to a LDR of 2370 metres for a wet runway.
On the face of it the aircraft at touchdown had 120 metres less runway remaining than it required to stop.

Not so.

The Landing distance required charts are predicated on the use of Max Auto Brakes and no Thrust Reversers. The landing distance required is lengthened by a factor 1.67 as a safety requirement. QF1 did not deploy its thrust reversers but it is clear that both pilots were standing on the brakes applying Max Manual braking which is more than Max Auto.

The aircraft should have stopped on the runway.

Boeing publish landing distance guidance charts which, as the name suggests give pilots guidance as to the length of landing roll to expect with various methods of retardation. The flap 25 landing guidance chart give a brakes and spoilers only landing distance from a height of 50 to end of landing roll of on a dry runway of less than 1250 metres. Add another 200 or so metres for a wet runway and you still have a figure far less than the 2250 metres remaining. Note that is not ground roll, but total runway used from a height of 50 feet.

The aircraft should have stopped on the runway.

It is interesting to note that referring to the same graph and entering with brakes, spoilers AND reversers cuts the total runway used roll by about 90m.

Boeing also publishes an Autobrake chart that gives predicted ground roll with various Autobrake settings. The graph notes that “Maximum effort braking should achieve shorter braking distances than Max Auto.” The BASI report notes that Autobrake 4 was selected prior to touchdown but manual braking was carried out after touchdown.
The Autobrake chart reveals that with Autobrake 4, on a dry runway at 168 knots, ground roll should be approx 1800 metres. With Max Auto the ground roll is approx 1300 metres. Full manual braking would be somewhat less than 1300 metres. Once again if you add approx 200 metres for a wet runway you can see that the aircraft should have stopped on the runway.

I accept that Flap 30 and reverse thrust would have reduced the ground roll.
I accept that with 20/20 hindsight, the go round should not have been cancelled.

But I do not accept that the above were the sole main contributing factors that caused the aircraft to leave the runway at a comparatively high speed.

So why did it do so? Either the aircraft suffered an unannounced failure of its braking /antiskid system, or, more probably, it aquaplaned on the wet, rubber laden runway.

It is more than possible that the Thai authorities, who have access to the above information seem intent on pre-empting the final report into blaming the aircraft and crew as they will not authorise a final accident report that in any way blames the airport or its facilities.

It is also more than possible that some journalists will not care for the above information as they have axes to grind that will not sit comfortably with the above. The idea that an accident was caused by rapacious cost cutting by the management of a national icon probably does make better headlines than blaming the condition of a far off runway. I leave it up to the good readers of Pprune to make up their own minds. Comments welcome.

IP:

aviator
Experienced PPRuner
posted 30 November 1999 06:35    

Are your your landing distances based on the computed flap 25 ref speed (154 kts) or the actual touchdown at 168 kts?
Is the runway grooved - was there standing water on the runway?
One has to be careful to not base decisions flying in inclement weather on numbers obtained under the best of circumstances by test pilots....
That said, I realize we have the luxury of analyzing the crews actions after the fact and they do not. It serves us all well to give ourself all the margins we can.

IP:

Clocka
Experienced PPRuner
posted 30 November 1999 08:06   

The distances obtained from the Autobrakes graph used 168 kts with appropriate temperature corrections. The LDR graphs work on Gross Weight and therefore are difficult to factor in increased touchdown speed. This probably explains why the Autobrake graph gives a longer stopping distance required.
21L is ungrooved. My theory is that the cell that caused QF15 to go round 90 seconds before QF1 landed, dumped significant amounts of water on the runway, which did not have sufficient time to drain away before QF1 landed.
The QF1 crew were not aware of the go-round and had been told that the braking action on 21L was good.

IP:

Crisp
Experienced PPRuner
posted 30 November 1999 08:18  

Clocka,

So in summary, do you think that the reasons for the QF1 accident were:
- the pilots aborting the go around
- the poor communication to the pilots of the RWY conditions by ATC
- and the actual RWY conditions at the time?

------------------
Stay Crisp!

IP:

Clocka
Experienced PPRuner
posted 30 November 1999 09:51   

A combination of those 3 plus more. There is no one single cause. Any one broken link in the "event cascade" might have prevented the accident.
My interest though, is in seeing a proper investigation take place and not be side tracked by journalists and politicians who have their own agendas. I also do not wish to see a crew unfairly labelled with the lion share of the blame due to expediency.
Yes, the crew landed long for various reasons. But I contend that the Captains' instantaneous judgement to continue the landing, with the information he had at hand was not unreasonable. The aircraft, as far as he knew, could stop in the runway remaining. The reasons why it didn't stop must be identified.

IP:

HotDog
Experienced PPRuner
posted 30 November 1999 11:04  

At the risk of divulging my age, I would like to quote D.P.Davies, author of "Handling the big jets" on the subject of aquaplaning. Especially the dynamic sort where due to standing water on the runway, the tires are lifted off and completely supported by water. Under those circumstances, no matter how efficient your carbon or otherwise brakes are, the only thing that will give you retardation is full reverse. In spite of the fact that stopping distances were certified without the benefit of reverse thrust on a dry runway, application of full reverse thrust will result in 25% less landing distance on a wet runway, providing you don't float for 3,000 feet before touchdown. The simple formula for aquaplaning speed, which they don't seem to teach anymore at performance lectures is 9 times the square root of the tire pressure in pounds per square inch. Gentlemen, please note I am not passing judgement on the unfortunate QF01 crew.

IP:

sparrow
Experienced PPRuner
posted 30 November 1999 11:23    

I see no-one has taken into account the effect of four big engines spooling up in forward thrust before the decision to cancel the go-around was acted upon.......

IP:

CCA
Experienced PPRuner
posted 30 November 1999 11:28    

I just read the BASI report from Dunnunda....
it read's approx 6sec from touch down to manual braking at which point it was at 155kts how much runway did that use up? and it remainad on the runway for the next 25 sec.

IP:

pterodactyl
Experienced PPRuner
posted 30 November 1999 00:55    

CCA,
Deduced from BASI report.
Touchdown on the runway at 155 kts (262 feet/sec)
6 seconds delay before braking at 262 feet/sec = 1572 feet.
25 seconds reducing from 155 kt to 79 kts, say average speed is 117kts (198 feet/sec) gives 25 * 198 = 4950 feet.
Average deceleration is 5.76 feet/second^
So we have;
6 sec braking delay = 1572 feet
25 sec deceleration from 155 kts to 79 kts = 4950 feet.
Working backwards total distance from touchdown to runway end = 1572 + 4950 = 6522 feet.
Runway length 3500 less 305 displaced threshold = 3195 metres = 10482 feet.
This suggests touch down was 10482 minus 6522 feet along the runway at 3960 feet.
All very hypothetical and does not allow for engine thrust inputs.
As we are all aware the greatest rubber deposits are in the touchdown zone and in the vicinity of runway turnoffs which in this case was near the end of the runway.
During and for some time after heavy rain heavy water concentrations on a runway cancel out any benefit of grooving, if there is sufficient water aquaplaning will occur. This runway was not grooved anyway.

[This message has been edited by pterodactyl (edited 30 November 1999).]

IP:

Akwah_Plain
Experienced PPRuner
posted 30 November 1999 13:37  

A couple of inputs to add to the calculator - firstly, from the Jepps, 21L runway remaining from glide-slope touchdown is only 2791m (9158 ft). Also, from the dark ages of Performance A, I seem to remember a very ball-park figure of 1% increase in LDR for each knot above V touchdown.

However, if visual references were lost in the flare, Clocka probably has got it spot on about the cell dumping loads of water, causing the runway to be flooded(any wind gusts reported -hence correction to V Ref??).
Would have been nice to have been told about the other QF going around!

Incidentally, at Butterworth (Malaysia)a couple of months ago, a Nimrod (RAF) went into the over-run with a few scalded tires and very hot brakes. The crew had been told by ATC that the runway was just "wet" but it was very flooded! The on-coming crew had tried to get into ATC to warn the landing aircraft, but the only controller on duty didn't hear them.

IP:

Sir Les Patterson
Experienced PPRuner
posted 30 November 1999 15:46  

Who aborted an Aborted landing?
Sorry but it does not compute.Eyeball Mk1 may work in a C206 but it does not work in a 256 ton craft. An Abort is an Abort is an Abort.
Bugger the calculator and eyeball, if a Go-Around is commenced, then GO AROUND.

IP:

Gulf227
Experienced PPRuner
posted 30 November 1999 18:30    

I will start by saying that I only fly the Saab, and not the whale and won't pretend to know how to fly it, but I agree, If you have a feeling that you need to go around...then go around...and stick with that decision. Again, just my two cents worth...I wasn't there, and really can't say what I would have done in that situation with the information provided...
----------

Fly Safe!

IP:

Kaptin M
Experienced PPRuner
posted 30 November 1999 18:40   

Just a slight correction to the aquaplaning formula, Hot Dog, [yes, Handling the Big Jets is getting on, but nevertheless has good gen]. The aquaplaning speed is now recognised as occurring at 7.5 times the square root of the [lbs] tire pressure.
But I'm sure that a Qantas Captain would have been well aware of factoring this into any of his considerations, during the approach brief.

IP:

gas path
Experienced PPRuner
posted 30 November 1999 21:16   

Do all these figures relate to a new set of tyres or part worns????
just a thought!!!!

IP:

Kaptin M
Experienced PPRuner
posted 30 November 1999 23:37   

Does anyone have any real doubts that QANTAS and BASI will not cover ALL aspects of this accident ?
But one question that needs to be asked,
"Was Captain and/or F.O. circumcised ? "

IP:

SeldomFixit
Experienced PPRuner
posted 01 December 1999 00:48    

I was lucky enough to grab a look at some ( not all of ) the main wheel tyres from the subject aircraft and whilst severely damaged ( some, but not all of )showed NONE of the characteristic damage suffered by an aquaplaned tyre.
Perhaps the ones I didn't see had such damage ?

IP:

HotDog
Experienced PPRuner
posted 01 December 1999 03:32   

Out of the three types of aquaplaning i.e. Dynamic, Viscous and Reverted rubber, only the latter will show the characteristic, tacky, uncured rubber look on the surface of the tyre. By all accounts, if QF1 did aquaplane, they experienced dynamic aquaplaning due to standing water on the runway. I have operated into BKK RW21L on numerous occasions. Yes, it is ungrooved, rough as hell and there is a lot of rubber in the touchdown zone. But QF1 didn't touch down there and is begging the question; if he didn't aquaplane, why didn't it stop?

IP:

Thrust
Experienced PPRuner
posted 01 December 1999 06:01   

Fascinating stuff!!

On a lighter note, QF1 is to be re-numbered QF"FOUR!!" Something to do with the golf course.

Sorry if I offended anyone. Tee Hee.

IP:

CCA
Experienced PPRuner
posted 01 December 1999 10:54   

QF tyre pressure is 215psi for the -400 if anyone want's it

IP:

tired
Experienced PPRuner
posted 01 December 1999 22:01  

Does anyone have any idea why the reversers were never used?

IP:

pterodactyl
Experienced PPRuner
posted 02 December 1999 00:41   

Seem to recall that if there is enough water present, exceeding the tyre treads ability to disperse, then aquaplaning (dynamic) will occur at a speed in excess of 9 times the square root of the tyre pressure(psi) and will not cease till below 7.5 times the square root of the tyre pressure.

For 215 psi this gives 132 kts initiation speed and 110 kts cessation speed. So at touchdown at 155 kts it would be necessary to reduce 45 kts to 110 kts before cessation of aquaplaning and during this time ONLY aerodynamic means of retardation would be effective. Even so, below that speed on a wet surface, friction is reduced without considering rubber deposits. We are now in Viscous and Reverted rubber territory.
Following on from the previous posting QF reduced from 155kts to 79 kts = 133fps on the runway surface and a ground roll of 1509 metres = 4950 feet averaging 5.76fps^ deceleration and reduced from 79kts to zero in 200 metres = 656 feet.

Acceleration = V squared divided Distance Multiplied by 2
133*133/2*656 gives 13.48fps^
Ergo soft mud has good stopping qualities. Fortunate there was nothing more substantial there.

[This message has been edited by pterodactyl (edited 02 December 1999).]

IP:

Billy the Kid
Experienced PPRuner
posted 15 December 1999 10:42   

Is that the tire pressure empty and cold or hot and MTOW?

IP:

Kaptin M
Experienced PPRuner
posted 15 December 1999 00:19  

Please define cold.

IP:

RamGee7
Experienced PPRuner
posted 16 December 1999 05:26    

My ten cents worth.
The chain begins in the office of the Boy Wonder who deceided to push for flaps 25/idle reverse as S.O.P. Forgive me if I am wrong but I have always held the opinion that the prime purpose of the landing roll was to stop the aircraft, preferably on the runway and before the end.
Boy Wonder thinks differently! he has a cunning idea! if we use flaps 25 only, we have less drag and therefore we burn a poofteenth less fuel and geewhiz if we multiply this fuel saving by the number of sectors in a day by the number of days in a week by the nunmber of weeks in a year by the number of years in a decade by the number of decades in a century.....but wait! there's more cunning to come,we now cross the fence going just a tad faster, 6~7kts, so lets all forget that this is a VERY BIG aeroplane and that pesky, mass times velocity squared stuff, because we have "awesome" carbon brakes and there is nothing they like more than a good work out and if we don't use reverse thrust those fab discs will get an even better work out and as we all know this will make them last just a bit longer,infact about a poofteenth longer per sector and if we multip.....BUT WAIT!! there is still more cunning to come, by not spooling up those BRT's we reduce engine wear and save even more fuel! "But how much do we save?" I hear you cry, well the Wonderous One reckons about a poofteenth less engine wear and a poofteenth less fuel on every sector and if you multiply.... and then if you add up all of these wonderous cost saving initiatives you save.....less than the cost of a runway overrun in Bangkok. D'Oh!
The hardest part of the 400 conversion for me was unlearning a life time of training (to instictively apply maximum reverse thrust after touch down) and not pull the reversers past the interlocks. I find after a long sector,especially when landing at night at what might be 3/4am body time I rely a lot on trained responses,Clocker you and I know the crew involved and they are good men, they did exactly what they were trained to do, as we both also know the company is big on CRM/Human Factors...until it comes to saving money that is.

IP:

mcrit
Experienced PPRuner
posted 16 December 1999 00:59   

RamGee7 the points that you make are to me the most decisive yet re. QF1. Particularly with regard to the Boy Wonder that you describe. Sadly in aviation most of us have met them. However despite the fact that most of the sums above are way beyond me I must make a point regarding goarounds. Not long ago, during a night, offset VOR app. in CB’s with a decision height above 1000’ to a badly lit airfield with high terrain I very nearly made a similar poor decision. The runway could be seen for most of the app. but just before decision we entered a patch of cloud. The FO called decision and I called goaround. Then almost immediately we were in the clear again but power had been applied and we would have been just a tad high and fast. Although tempted to land I continued the goaround. On landing after the second app. it became apparent that the runway was under some inches of water in places and braking was very poor. Had I continued the first app. I might now be unemployed or worse. The point is that at the point of goaround in real time we do not normally have the time or mental capacity or all the information available to make an informed change of hart. If you decide to goaround then goaround.

IP: