View Full Version : CX Windsheer G/A! post removed

Five Livers
20th Sep 2004, 17:09
What's happened to my post about the 118kt bunted recovery from a predicted windshear g/around?

No response from the moderator so I\'ll re-post.

Has anyone heard about the CX 747-400 go-around following a predicted wind-shear warning. They apparently recovered by bunting over at 6000\' and 118kts!

20th Sep 2004, 18:14
You seem to know a great deal about this incident already.

What does "bunting over" mean ? (I haven't heard this term before).

What was the Landing Mass and Configuration ? Which airport was it ?

Was it 6000' Pressure Altitude, Density Altitude or AGL ?

Boeing recommended windshear recovery technique permits flying to the stick shaker (which should always be respected). I suspect that 118KIAS is approximately 10 knots above the stall speed at Max Landing weight (in approach config). I also suspect that 118 knots may well be close to Vref for a light B744. Once established safely in the missed approach with positive rate of climb, positive indication that windshear is no longer an influence and terrain conflict no longer a consideration, configuration may be changed.

I suspect that a tired crew at the end of a long sector at a light landing weight may well have taken the overpowered beast to 6000' before levelling off to accelerate and clean up. Most heavy pilots fly Windshear recovery a little unpolished and not as smooth as we may operate in other profiles, usually due to the urgent need to get away from the ground. If "bunting over at 118Kts" means that the aircraft went to 0G for a moment, then that makes the recovery safer still as it would now be 118 knots above the stall speed...(Devils advocacy).

I'm interested to know how you know the windshear was predicted ? If it was a report from a previous aircraft on approach, then it is a PIREP, which is NOT a prediction. If we all threw away every approach where windshear advisories were effective, we'd never get into Manchester or Anchorage in the winter 30% of the time.

TDK mk2
20th Sep 2004, 18:53

Just curious to know: are you a pilot? As Basil says one would expect a pilot to be familiar with the term 'bunt', and 6000' generally means 6000 feet above sea level, or 6000 feet QNH or 6000 feet MSL as you may say in the U.S.

I haven't heard 'Pressure Altitude' since I was studying for my PPL so could that be about where you're at my friend?

20th Sep 2004, 19:36
Jeez Basil and TDK, what's wrong with a simple answer to a simple question to wit

'What does "bunting over" mean ? (I haven't heard this term before).'

Seems some prefer to be obscure, question another's authenticity, create a scrap, or whatever, than to give a straightforward answer.

Just an internet thing I guess.

I don't know either, I am likewise curious, I am but simple SLF but I have pretty good language skills.

It might be pertinent to mention that while most prooners speak some variety of the language broadly known as English, it is actually quite different in many details for a Yankee, a Pommie, a Brit, or an Indian; and lord help you if you go to, say, Grenada, where upon arrival one of my party asked the Customs people (in her English) what language were they speaking. This did not improve their already poor attitude, as they believed they were speaking English.

Joyce Tick
20th Sep 2004, 20:52
just to put half the readers out of their agony - to bunt is to push less than 1G....

..and Skibeagle, what planet are you from? You post like an outsize irriot rather than a pilot..

20th Sep 2004, 21:17
I stand to be corrected, but I always thought that a "bunt" was a "forwards push on the stick/control column to give a negative G effect ", hence plus or minus G for the a/c type cert.
we aim to please, it keeps the cleaners happy

21st Sep 2004, 12:19
Back to topic:


My guess is they got a predictive windshear warning from the onboard system, a feature on semi-modern boeings that works in conjunction with the wx-radar and areas of windshear is shown as magenta on the screen..

Ranger One
22nd Sep 2004, 14:52
skibeagle etc:

If you push, push, until the crap that normally lives beneath the pedals is on the overhead panel, along with the charts, the Sun, the FOs Racing Post, and the remains of breakfast - congratulations! You have successfully executed a bunt.

Tea, no biccies on the menu. And the CC will piss in your coffee for the next 28 sectors.


22nd Sep 2004, 21:15
Great, I feel really enlightened now. I did find a nautical definition of "bunt": The swollen or baggy part of a sail", nothing about bunt's in aicraft though.

Is it a British thing...???

22nd Sep 2004, 21:58
Oxford Dictionary; one amongst a few definitions of bunt: A push with the head or horns, a butt. Chiefly N. Amer. & dialect . Similar to a hammerhead stall, if you are familiar with that.

23rd Sep 2004, 00:18
I understood a bunt was an outside loop, the term used to be used for that quite a bit. Not heard it used as a verb but would expect it to mean pushing forward into at least 1 Neg. G

23rd Sep 2004, 09:26
I found one reference to a 'Bunt':


23rd Sep 2004, 10:15
I'm not sure but I don't think for it to be a bunt per definition it has to be negative G's.

I've heard the parabolic zero-g flights being refered to as "bunts"..

Joyce Tick
23rd Sep 2004, 11:24
From a History of the Royal Air Force:-

"The Messerschmitt's Daimler-Benz DB 601A engine had the advantage of fuel injection which enabled the aircraft to bunt (push negative g at the top of a manoeuvre or climb) without losing power. The Merlin engine of the Spitfire had a float-type carburettor which necessitated the aircraft performing the longer manoeuvre of rolling inverted before diving to maintain positive g, thus preventing the engine from cutting out as a result of fuel starvation"

Five Livers
25th Sep 2004, 02:27
Let's try to get back to the point.

The matter for discussion is that a CX 747-400 LHR/HKG carried out a windshear Go Around from the approach into HKG.

Somehow, during the G/A, the autpilot disconnected AND nobody noticed. Some of the divergences from the planned profile were noticed but it took a long time for the penny to drop with the 4 man crew that the A/P had disengaged.

The aircraft recovered by pushing over, bunting, shoving the control column forward, whatever [not the point] at more than 6000 feet and about 118kts [the point]

25th Sep 2004, 03:02
Where did you get your info from? And what do you want to discuss? The MSA out of HK is 5000, so a windshear G/A to 6000 is standard. Admittedly, 118kts, if true, is slow, however I would say that prior to getting that slow, the stickshaker/pusher would have something to say about that. At G/A power, there is a fair bit of grunt coming out the back, so the AOA would have been pretty steep. I also find it hard to believe that no one on that crew heard the A/P disconncect, its pretty loud and hard to miss. CLK is well known for having quite bad windshear, so pilots are aware of it, its often on the ATIS, so the pilots would have been ready for a possible G/A. I am not sure what you are after, but all things considered, it sounds doubtful that ALL the cues could have been missed. Not impossible, but highly unlikely.

Five Livers
25th Sep 2004, 03:12
The information came direct from 2 CX pilots at different times and the story is that the crew did not realise that the auto pilot was NOT controlling the aircraft. It's hard to believe [especially witha 4 man crew], which is why I'd like some more information so I don't make the same mistake.The CX corporate safety people and flight ops management are keeping a very tight lid on it at the moment.

25th Sep 2004, 04:37
It sounds more like you are trying to stir the sh1t to me. You say you want more info so you don't make the same mistake? IF, and its a big IF, this is true, whats to learn as it seems a pretty straight forward mistake, no doubt helped by the end of a long overnight easterly longhaul. Don't let the Autopilot disconnect and not notice, hows that for a lesson.:rolleyes:

I think your reasoning for stirring this up is not for your benefit, but as I stated above.


25th Sep 2004, 06:03
NoseGear, if I may...

The MSA in HKG is 4300. Standard missed approach alt from RWY25R (the runway they were attempting to land on) is 3000' initially. If tower was on the ball, you MAY expect a clearance to 4100' (had it once), or 5000', but in light of the attention grabbing display on the PFD, I doubt too much was done in the way of radio calls, or changing of MCP alt until after the 3000' was kicked well into touch.

Where does the
"so a windshear G/A to 6000 is standard"
come from?

I believe the AP disconnect aural warning was over-prioritised <yes I did make that up>by the PREDICTIVE windshear warning they received.

Of greatest concern is the allegation that they ended up on the Tung Chung side of the 25L (LEFT) extended c/l.

The go-around is said to have commenced at approx 1100'.

"Admittedly, 118kts, if true, is slow"

I disagree. Ref-5 is slow.
118kts is terrifying in an AC that should be doing about 30kts more. <Yes, it seems the speed is legit>

25th Sep 2004, 15:59
Five Livers, and other Pruners, you appear to over concerned by the low airspeed, which if I recall was also the focus of the original thread.

When an aircraft is manoeuvred harshly nose down (a bunt), it reduces the angle of attack, which if this is less than the stall AOA, the aircraft will not stall irrespective of the airspeed. There are limits to this, particularly attitude, altitude, and the duration of the manoeuvre, but if the nose is lowered (AOA reduced) the airspeed will quickly increase – the same principle as stall recovery.

Similarly, people appear to be over concerned with less than 1 ‘g’, or for the passengers wellbeing in such conditions. It is very difficult to achieve and sustain reduced ‘g’. The force to move a control system usually increases proportionately to the ‘g’ required, this applies to reduced ‘g’. In order to spill the drinks (or the john) then the limit value – normally zero ‘g’, has to be both achieved and sustained, and even then passengers may not be any more concerned than they would in severe turbulence; turbulence at +/- 0.5 g may be much more violent.

The point of this incident is why the crew were not aware of the state of the autopilot.

25th Sep 2004, 16:34
Take whatever it is you fly, and picture yourself on approach at a normal landing wt.

Now take your Vref/Vapp or whatever you chose to call it, and subtract 25-30kts.

This is not alarming to you in a pax carrying widebody?

[edited to add]

To put another slant on it, if the same (non) reaction, pitch change, speed reduction, etc had occurred at 1000' on departure 12 hours earlier out of LHR, it is my uneducated guess that this would have been front page news.
-By admission....this is an extreme comparison

Five Livers
25th Sep 2004, 16:35

What a load of nonsense!

If you disregard most of the first page's red-herrings about what bunt means, this post is trying to establish how a 4 man crew can initiate an auto-pilot wind-shear go around and not realise that, at some stage, the auto pilot has disengaged. The references to speed, altitude and recovery action are only mentioned to illustrate how far the situation developed before a recovery was effected.

25th Sep 2004, 19:14
jtr one of the primary characteristics of a professional pilot is not to be alarmed; - concerned, yes.
With reference to your speeds, a normal approach is flown at approx Vref +10, this is roughly 140% of the stall speed (Vref=1.3Vs). During a Windshear recovery most flight procedures require, and some auto-recovery systems aim for, but respect stick-shake speed (1.1Vs, 110% of stall speed). Thus for a stall speed in the region of 100 kts, a speed reduction of 25-30 kts in a 747 during Windshear conditions should not be alarming, it should be expected. When flying a Windshear procedure the crew should have enhanced awareness and vigilance due to the reduced safety margins. If altitude was not a primary concern, then a nose over maneuver is acceptable, increased speed may enable quicker penetration of the windshear conditions.

However, in this incident the nose over maneuver appears to be the result of crew action to mitigate the low speed and a previous error – the autopilot was not engaged when they though it was. This is ‘alarming’. Why did such an error occur and why was it not detected earlier in high vigilance situation and with a large, and presumably well trained crew; but of course these are the usual issues of human factors that may never be established.

An example of a very professional crew is here: Windshear an accident. (http://www.fly-safely.org/story_pf.asp?id=32) The normal approach speed for a BAe146 is Vref+5 and the stall speed for this approach was approx 90 kts. Thus as a %, the loss of 15-20 kts is similar to those figures stated above.

Five Livers your point: “this post is trying to establish how a 4 man crew can initiate an auto-pilot wind-shear go around and not realise that, at some stage, the auto pilot has disengaged”, exactly!

My point - “why the crew were not aware of the state of the autopilot”.
The other ‘nonsense’ is basic aerodynamics that appeared to have been overlooked in the development of this thread.

We are bunting in formation!

26th Sep 2004, 01:07
alf, I have just realised there is one point that hasn't been made clear enough.

There was no windshear.

The reaction was to a predictive windshear warning, and in the subsuquent manoeuvre there was no windshear experienced.

This comes from a member of the crew.

To put some likely figures in your scenario...

The landing wt would probably have been around 245T.

Lets assume they were flap 30.

Vref 30 @245T is 142kts

Vs must be about 109kts.

So there you are, 9 kts above 1g stall speed, at or about 1g, pointing towards a 3000' bit of dirt, and you are still merely "concerned"???

Thats a big set of cahones you've got

26th Sep 2004, 10:48
alf5071h ..... and others here,

I realise that the concern about this incident is more about the particular aircraft and its low speed but I couldn't help noticing alf5071h's description of the aerodynamics (alf5071h - bear with me):

"When an aircraft is manoeuvred harshly nose down (a bunt), it reduces the angle of attack, which if this is less than the stall AOA, the aircraft will not stall irrespective of the airspeed."

I thought the lack of a stall here would have more to do with the bunt (i.e. negative G) effectively reducing the aircrafts weight (its mass remains the same). It weighs less for the period of the bunt which is why it doesn't stall at its normal level flight speed - it is partially ballistic for the period of the bunt.

26th Sep 2004, 16:13
jtr thanks for the timely reminder. However, assuming that the crew commenced a ‘windshear’ mode go-around, then they or the aircraft system would (should) have allowed the airspeed to reduce in proportion to the demanded climb. This reduction in is independent of an actual windshear being present. The required reduced airspeed (sometimes AOA) is taught in training or programmed by the windshear mode of the flight guidance system.
In the aircraft that I am familiar with, when the crew select go-around and windshear has been detected (reactive warning), the flight guidance system will automatically select the windshear go-around mode (low airspeed), including increased thrust. I assume that a predictive system will work in a similar manner.

An additional concern would be if an onboard predictive Windshear system gave an unwarranted alert, or if the warning was given by a ground system and it was not applicable. Whichever scenario applied the crew should still have flown a ‘windshear’ go-around.

If the aircraft was not in a ‘windshear’ go-around mode, just a normal go-around, then I agree the loss of speed is more than a just a concern. However, in either circumstance it appears that the reduced airspeed was due to the lack of autopilot control, and that the root concern is why the crew failed to realise this.

MrBernoulli you are correct, but ... a reduction in ‘g’ (less than 1, but not necessarily negative) and AOA are complimentary in lowering the stall speed. In order to reduce ‘g’ the lift demanded from the wing has to be reduced, which is achieved by reducing AOA.

26th Sep 2004, 18:27
I do not know whether they got a PWS caution, or warning.

The AFDS will respond to application of the TOGA pitch (two pushes on the go buttons) mode by commanding "the lower of 15 deg nose up, or 1 degree below the PLI*", and maximum volume on the noise makers.
:confused: (I am sure I have seen more than 15 nose up in the sim in this type of situation)

i.e. at an absolute maximum, the FD/AFDS/AP or whatever manner you choose to perform the move, should not result in more than 15 deg nose up.

The -400 doesnt have any mode for this situation other than TOGA mode. <Well none that we are told about in the somewhat limited manuals.>

Where I am going with this is that max noise TOGA in a light a/c with no apparent windshear should have you sweating about getting the flaps up quick enough when you decide you've had enough... NOT hovering over the stall. Which in turn leads us back to your point alf... Why didn't anyone respond to the AP situation? Had that been reacted to, then the speed/pitch/navigation/altitude issues would likely not occurred.

*PLI is a PFD representation of the pitch attitude where the stick shaker will occur.

26th Sep 2004, 20:40
It is most unlikely that we will determine the reasons why this crew did not detect the state of the autopilot. These ‘crew error’ incidents usually have deep underlying causes. One of which may relate to my first post, that crews should not be ‘alarmed’. I may be playing with words, but this is to make the point that in unusual circumstances crews have to control their emotions, have knowledge of what is expected to occur, and to be extra vigilant. The surprise and stress of the situation may lead to tunnel vision (over focus) on less important items. In addition, the senses are reduced, hearing is the first to degrade, and thus this is one possible reason for not hearing an autopilot disconnect warning. Another more obscure reason may be in the design in the priority of warnings; usually windshear takes precedence over GPWS, TCAS, etc. Can a 744 driver educate us as to how a 744 system works? An active PWS warning and then the autopilot disengages, – do you get an audio warning for the autopilot disconnect, if so when? Does the autopilot disconnect at stick shake?

jtr I think that you have a common misunderstanding about the operation of the PLI. The principle of most systems (I do not know the 747 in detail) is to provide an indication of the stick shake angle of attack (AOA) on the attitude indicator. Thus, the ‘pitch’ attitude shown by the PLI is actually AOA relative to the local airflow (or flight path in some systems).
In severe a windshear this ‘pitch’ angle can be very low; if a downburst causes the aircraft to descend at 12 deg (see the previous link example), the PLI is referenced to -12 deg, thus if the stick shake AOA is 17 deg, the PLI will only be at + 5 deg attitude. An upper limit (15 deg) is logical but in gusty windshear, the AOA could be very dynamic and depending on the sensor and damping there could be some over swing.

For some aircraft during windshear the stick shake may activate, this may be normal – this is ‘respecting’ the stick shake; not a failure in that it must not occur. Therefore, it is important for crews to be trained to know their system and not to over react to an alert that in specific circumstances could occur naturally.
An hypothesis in this incident may be that if the stick shake operated the crew pushed forward harshly causing a bunt; the stick shake was due to low speed, which was due to the lack of autopilot … and we are back to the beginning … why?

27th Sep 2004, 01:11
Umn, as far as I can recall, on the boeing you will only get a "normal" GA mode until an accual windshear has been encountered. Might be different on different types but the predictive-windshear feature is awareness only as far as I know...

27th Sep 2004, 02:04
Just wondering - do CX require all four crew on the flight deck for T/O and landing? When I flew the 744, (with another operator), it wasn't mandatory for the deputy crew to be there though usually the deputy F/O would stay to help with paper work, company calls, weather etc.

27th Sep 2004, 12:59
All four crew members were on the seats for arrival. It is not "policy" that I can remember seeing anywhere, but certainly convention.


I said..
"PLI is a PFD representation of the pitch attitude where the stick shaker will occur"

You said...
"jtr I think that you have a common misunderstanding about the operation of the PLI."

and then went on to say...
"The principle of most systems (I do not know the 747 in detail) is to provide an indication of the stick shake angle of attack (AOA) on the attitude indicator."

Aren\'t we saying the same thing, or am I missing something?

27th Sep 2004, 15:42
An update on PWS audio warning priorities, this refers to the RDR4 radar:
Basic Audio Prioritization
1. Reactive Windshear System (RWS)
2. Predictive Windshear System (PWS)
3. Ground Proximity Warning System (GPWS)
4. Traffic Collision And Avoidance System (TCAS)

When the RDR-4B is interfaced to the EGPWS the message/display prioritization is:.
1. Reactive Windshear
2. GPWS Mode 1 Warning
3. GPWS Mode 2 Warning
4. EGPWS Terrain Warning
5. Predictive Windshear Warning
6. GPWS Mode 1-5 and EGPWS Caution
7. Predictive Windshear Caution

But when and how is an autopilot disconnect warning given in any of the above circumstances? Could an autopilot disconnect ‘get lost’ in all of the aircraft’s logic, or is it more likely that a crew, having just received a ‘stressful’ warning would mentally lose (not hear) a lower priority warning?

jtr it appears that we are saying the same thing.

14th Oct 2004, 07:00
Here is the Telegraph's account of the story (http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2004/10/14/wplane14.xml&sSheet=/portal/2004/10/14/ixportal.html), from today's electronic edition (free subscription required to view the article). Excerpts:

The aircraft flew uncontrolled for three minutes, veering almost 180 degrees off course toward mountains and coming within seconds of a catastrophic stall.

No one in the four-person crew took any intervening action because they believed the Boeing 747-400 was being directed by the autopilot. They assumed the aircraft's unusual movements resulted from a local windshear effect, which their weather radar had warned them about.

A normal go-around would involve a climb to 3,000ft and a long right turn to take the aircraft back to begin another approach. But with no one controlling the aircraft, it failed to make the right turn, drifted to the left and ended up perilously close to a 3,100ft peak on the island of Lantau.

The jumbo then carried on climbing above its authorised altitude to 6,000ft, bringing it into potential conflict with other traffic.

Despite its steep rate of ascent, the uncommanded 240-ton aircraft was travelling at the dangerously low speed of 130 mph. It was saved from a stall, which analysts believe could have been fatal, because the left wing had dropped by 30 degrees - just enough to bring the nose down fractionally and enable a recovery. Worried air traffic controllers issued urgent instructions to bring the jet back on course.

But this had no immediate effect, as the crew - still under the impression that the autopilot was working - were not attempting to fly manually and simply keyed the navigation commands into the lifeless machine.

Investigators will be anxious to establish why the crew did not react to the aural warning that should accompany an autopilot disconnection. One theory is that it could have been drowned out by the windshear warning if they occurred simultaneously.

The aircraft's failure to comply with instructions may have been viewed by the crew as confirmation of the windshear warning's accuracy. Senior 747 pilots have been angered by what they regard as the low-key attitude taken to the incident by Cathay Pacific executives. In response to inquiries, the airline initially said only that a windshear event had taken place.

It was not until after Hong Kong Civil Aviation Department last week told the airline to issue a safety notice to all its pilots that Cathay officials admitted that "use of the autopilot" had been a problem.

One pilot said: "This was an extremely serious incident. Yet flight crew need to learn the lessons in case a similar situation arose. The fact we don't know exactly what happened makes it more likely that a repeat could occur."

19th Jan 2005, 00:12
When you fly long hauls, few landings and lot of hours in the cockpit, a crew that is already tired by the end of the flight, remeber your body is working with less than min O2 as the cabin even pressurized at FL370 or above will be above 5000ft what by FAA you should be using mask to fly at night as your vision start to loose some accuracy.
But the main point is it's very easy for people to judge and whoever flies long hauls know the huge difference between domestic I would even say that is not only two different aviations but two different worlds the important thing is the recover they didn't crash and even after a little while they manage their lag to catch the A/P out. So what's the point?
Try to say CX pilot are no good, or try to say something bad about CX because you are, maybe, one of those who failed on the interview and can't accept it? Or you might be one of those who think US pilots are the best in the world?
When you learn to fly long hauls you can talk about it.