Aircraft Upset

[Niallo]

As an engineer, non-aviation, I am interested in aircraft upsets.
As I understand it, and please correct me if I am wrong, for an aircraft in cruise at high altitude and weight, the allowable range (margin) of air speed is small, to avoid stall at the lower limit or Mach overspeed at the higher limit. If the aircraft flies at a lower altitude, the margin increases; at a higher altitude the margin decreases.
It seems then, that the risk of aircraft upset would be inversely proportional to the size of the speed margin. (A margin of 5knots would be riskier than a margin of 20knots).
Who decides at what cruise altitude to fly, and hence decides what risk of upset is acceptable?
Is it the aircraft manufacturer, the certifying authority, the airline, or the captain?

[BOAC]

Basically the crew, guided by the manufacturer's figures for max alt v weight, typically based on a margin of 1.3g to allow for moderate manoeuvre or turbulence. Loads of other factors like sector length, speed and ATC, but these should answer your question.

[411A]

In the AFM, is a chart labeled 'Cruise Onset Buffet Boundry' (or similar), and this is (or should be) consulted by the pilots, prior to selecting a cruise altitude.
Many FMS units also have this information.
IE: jet transport airmanship 101

[Loose rivets]

Berlin Air Corridor comes to mind when thinking about keeping within the envelope. Sometimes over-speeding and stalling at the same time seemed preferable to an altercation with the hot and heavy things that seemed to be way within 5 miles of the wingtip.

Although I never witnessed it, one guy used to push 5k over the max height in a 1-11 if he couldn't go down, and ice was building up. So, he had the start of icing and a tricky airplane.

It was him that told the story, so not quite hearsay.

[BOAC]

Berlin Air Corridor comes to mind when thinking about keeping within the envelope. Sometimes over-speeding and stalling at the same time seemed preferable to an altercation with the hot and heavy things that seemed to be way within 5 miles of the wingtip.- apologies for an obvious 'thread spread', Niallo, but can you amplify that, LR? Doesn't fit with my corridor experiences. What level were you and when?

[PBL]

for an aircraft in cruise at high altitude and weight, the allowable range (margin) of air speed is small, to avoid stall at the lower limit or Mach overspeed at the higher limit.

A typical figure for a typically-loaded long-haul aircraft at cruise altitude (high-thirty-thousands of feet ISA) is 70-80 kts TAS.

Typically, Mach overspeed is not a problem leading to upset. During certification, the airplane is flown at Maximum (Mach) Operating Limit Speed in level flight, then the nose is pitched down 7.5° for twenty seconds, followed by a 1.5g pull-up. The speed reached is V_D (by definition), and the aircraft controls are cleared for flutter on paper to 120% of this. You can check out the criterion at CS 25.335. CS 25 is available from the EASA WWW site under Certification.

There is a huge drag increase at some transonic point. This is known to aerodynamicists as the Drag-Divergence Mach number, M_DD. For many airfoils, the difference between the critical Mach number, M_Crit, the point at which some of the airflow over the wing reaches Mach 1, and M_DD, is small or zero. Supercritical airfoils are designed for this difference to be large, so that the airplane can be flown at speeds between them.

This drag increase is sufficient to inhibit large increases in speed, even if you put the nose down. However, getting to M_D can be somewhat unpleasant, not to speak of nerve-wracking, so I am told, since there is a huge increase in buffet as more of the airflow over the wing goes sonic and supersonic. I haven't talked to anyone who has actually performed such tests.

Upsets at cruise altitude and speed are hard to come by: first, because of drag divergence; second, because the dynamic pressure is low enough to restrict the reaction in attitude of the aircraft to control displacements, although controls are sensitive - large excursions in altitude and acceleration may be caused by relatively small control inputs.

The U2 is an example of an aircraft point-designed to fly in a very narrow range between stall and M_DD: only about 7mph separates the two (John D. Anderson, Jr., Introduction to Flight, 6th Edition, McGraw-Hill 2008, p 338).

PBL

[john_tullamarine]

(a) Typically, Mach overspeed is not a problem leading to upset.

caveats - mach tuck and stab mistrim can provide some excitement and was a major upset problem duet in the early jet transport days ... leading to several hull losses. Mach tuck increases the tendency for speed exceedance and a mistrimmed stab can make recovery from the dive very difficult .. especially if the stab trim jack stalls and the pilot doesn't realise what's what.

Not so much a problem now with better design/understanding and operational procedures.

(b) U2 .. fly in a very narrow range

those folks had an interest in keeping away from the bad guys' toys. I had a course lecturer many years ago who had been a US U2 driver .. related tales of watching the bad guys start their final zoom climb .. whereupon he would turn this way or that and watch them run out of puff somewhere over there and fall back towards Earth.

Worked real well for all except FGP, I guess.

[PBL]

caveats - mach tuck and stab mistrim can provide some excitement and was a major upset problem duet in the early jet transport days ... leading to several hull losses

Point taken. I could rather say ".... is no longer a problem" although the excitement remains, I understand.

PBL

[411A]

Mach tuck increases the tendency for speed exceedance and a mistrimmed stab can make recovery from the dive very difficult .. especially if the stab trim jack stalls and the pilot doesn't realise what's what.


Ain't that the truth.
The perfect upset scenario with early B707 types, especially the straight-pipe powered models, where the engines could get you high, but the wing was far above optimum, for the weight.
Not good...at all:ooh:

[Niallo]

Thank you all for the informative replies.

So the choice of altitude, and hence the speed margin, is chosen by the flight crew.

Therefore the the flight crew is inherently accepting a risk of upset corresponding to the speed margin.

Do they also take into account the ease by which they can recover that model of aircraft in the event of an upset?

[BOAC]

Yes, but as we have said, in line with performance guidance in the Drivers' Handbook. Crews have exceeded the recommended altitudes and 99/100 got away with it. Even the 'Maximum operating altitude' of an aircraft is not a limit in reality.

In terms of 'ease of recovery' - by sticking to the guidance you should not have to unless some totally unforeseen event occurs.

[safetypee]

Niallo, re “… the flight crew is inherently accepting a risk …”, I not sure that you have fully understood some of the explanations.

The flight crew can choose the altitude, but the maximum value would still ensure a sufficient speed range for normal operations. The maximum certificated altitude is not the ultimate altitude.
The size of the speed range does not directly represent the likelihood of an upset or the extent of the risk.
For inadvertent over-speed there is a reasonable margin beyond the normal speed limit and warning; for low speed there is a smaller margin, but there is an earlier warning and often a protection/recovery device.
Provided the crew adhere to the limits, normal operations only involves minimal risk, i.e its safe, Safe is the avoidance of unnecessary risk.

For environmental hazards, including wake turbulence, the aircraft has to meet stringent certification requirements on gust response and structure loading. Any speed excursion associated with these should be contained within the ultimate safety limits, which are beyond the normal limits, i.e. a larger speed range.
For such events the aircraft has to be controllable, and/or recoverable from an ‘upset’.
In these circumstances the additional risk from the external hazards and high/low speed is contained or the effects mitigated in proportion to the expected frequency of encounter, i.e they are rare or very rare events.

The unknown in these operations is human behaviour. The certification rules which define the speed limits, assume an acceptable level of pilot performance. However, if the rules are violated then the risks can be higher. If procedures are not followed or skills deficient, then again there is increased risk.
In recent incidents, many, the large majority of events have involved human behaviour, hence the industry’s concerns about training, experience, etc.

[Gas Bags]

Niallo,

The manufacturer of the aircraft type (for example Airbus) are the ones who originally determine how an aircraft can be operated. This is formulated based on design and engineering, among many other things.

Therefore to put it very simply the OEM gives envelopes that an aircraft can be operated for any given scenario. The flight crew has discretion to operate within those envelopes.

The OEM also gives instruction on how the aircraft is to be maintained, parts that are approved to be used on the aircraft, and basically every circumstance that an aircraft would encounter during its operational life.

The regulator approves those instructions and the operator then operates the aircraft in accordance with those approved instructions.

This is a simplified basis of how it works.

GB

[Mad (Flt) Scientist]

... During certification, the airplane is flown at Maximum (Mach) Operating Limit Speed in level flight, then the nose is pitched down 7.5° for twenty seconds, followed by a 1.5g pull-up. The speed reached is V_D (by definition), and the aircraft controls are cleared for flutter on paper to 120% of this. You can check out the criterion at CS 25.335. CS 25 is available from the EASA WWW site under Certification.

Just to comment that in some (most? many? - can't speak for everyone) cases the 25.335 "manoeuvre" is, as per the suggestion in 25.335(b)(1) calculated rather than performed, in order to avoid flying an inherently hazardous manoeuvre. There are other requirements which do require flight demonstration, and to speeds approaching VD/MD, but not necessarily reaching them. 25.251 and the immediately following regulations either require flight to VDF/MDF or demonstration that this speed is not exceeded. Flight flutter testing must also, per 25.629(e), demonstrate flight at VDF/MDF - albeit with another caveat about use of analysis in lieu of test, although that is harder to achieve than for 25.335.

[BOAC]

Poor Niallo.......................only asked about level choice:confused:

[Niallo]

I realize my questions may have been naïve and I thank you all for your patience in replying. I also appreciate the discussion.
I have tried to summarize the replies. Please correct me if I am wrong.
The flight crew chooses the best altitude for the estimated aircraft weight while respecting the limits defined by the aircraft manufacturer in the Flight Manual and in the Flight Management System. The manufacturer's limits include a margin for moderate manoeuvre or turbulence.
Therefore the selected cruise altitude should ensure a safe flight unless a manoeuvre (advertent or inadvertent), or turbulence, results in the aircraft exceeding the margin.
If that should occur, the flight crew is confident, thanks to their training, firstly that they can identify what has caused the situation (eg over or under speed), and secondly that they can correct the situation before severe buffet occurs and certainly before an excursion outside the certified flight envelope, possibly leading to an upset.

[PBL]

Therefore the selected cruise altitude should ensure a safe flight unless a manoeuvre (advertent or inadvertent), or turbulence, results in the aircraft exceeding the margin.

Not quite. The certification regulations require it to be shown that an inadvertent manoeuvre or turbulence should also not result in any "margin" being exceeded. There is a question what degree of turbulence may be found at cruise flight levels, because by no means everything is known about the weather at these altitudes. But the certification regulations try to ensure that, if a disturbance ensues at cruise altitude and airspeed, the airplane will recover itself. As john_tullamarine pointed out, this wasn't always so, that the regulations have evolved as a result of experience, but I think it is safe to say that it is the case now, give or take a little bit.


If that should occur, the flight crew is confident, thanks to their training, firstly that they can identify what has caused the situation (eg over or under speed), and secondly that they can correct the situation before severe buffet occurs and certainly before an excursion outside the certified flight envelope, possibly leading to an upset.

....the flight crew is confident, thanks to the airworthiness certification of the airplane, that the airplane can recover itself (= "correct the situation") by virtue of its stability characteristics, before an excursion outside the flight envelope (e.g., upset) occurs.

This is not to say that the certification process is perfect, by any means. But I have described the overall intention and I think that is what has been achieved, give or take a bit, over fifty years of experience certifying passenger jets for flight at these altitudes.

PBL

[Niallo]

Thx PBL for clarification as below:

...the flight crew is confident, thanks to the airworthiness certification of the airplane, that the airplane can recover itself (= "correct the situation") by virtue of its stability characteristics, before an excursion outside the flight envelope (e.g., upset) occurs.By "airplane" I take it you mean the airplane including its automated control systems which provide augmented stability.

This seems to imply that the pilots should not intervene, but simply leave recovery to the airplane.

[safetypee]

Niallo, I think that most certification requirements assume some pilot intervention, i.e. reduce thrust to control an over speed, or pull-up to regain altitude.
With stalling, the assumption is that the crew will provide a recovery, and that the use of the recommended procedures will not create further problems.

Beware of the generic ‘automated’ system.
In older aircraft, ‘automation’ could be a control gearing change (sliding cam), a big spring to increase stick force, or automatic positioning of trim (Mach Trim) to do likewise.
In modern aircraft there can be a wide range of computer driven ‘automatic’ functions. Here some of the problems have been inappropriate crew intervention (knowledge, experience), and system failure, which may require alternative interventions.

I have only flown one aircraft (military) which required the pilot to wind the watch and ‘pick your nose’ as a recovery for loss of control (A7), but then most normal flights were an ‘upset’.