alf,
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

alf5071h

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.

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. 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!

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

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.

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.

PWS
 

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.

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?

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...

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.

All four crew members were on the seats for arrival. It is not "policy" that I can remember seeing anywhere, but certainly convention.

Alf..


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?

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
8. TCAS RA
9. TCAS TA

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.

Here is the Telegraph's account of the story, 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."

one more opinion
 

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.