In certification flight test the airplane is decelerated 1 kt/second with idle thrust

Worth keeping in mind that, for the formal test programs, the TP folk run a series of deceleration rates so that the book 1 kt/sec (1 mph/sec for even older aircraft) is determined by sums done in the backroom.

No doubt JF will come into the fray at some stage but, in the interim, several of the recent posters know more than quite a bit about this stuff ... so their counsel is worth attending.

Meikleour

Dont forget there is always the very knowledgeable CliveL who was a Concorde areodynamicist on here http://www.pprune.org/tech-log/42398...-question.html :ok:

As a one time certification test pilot, as stated by John Tullamarine to achieve a one knot per second approach rate it was usual to bracket about 0.7 and 1.2 knots per second. The boffins did the correction to 1 knot per second.
The minimum speed after all corrections was the certification stall speed.
The problem with this stall speed determination was that high drag aircraft could have a significantly lower g loading at the stall than low drag aircraft - this could result in a stall speed advantage ie a lower stall speed.
Around the time of A320 certification decisions were made by various authorities to change to Vs1g rather than Minimum Speed in the Stall.
Arguments then arose about the safety factors to be applied. It being obvious that (say) 1.3 Vs was being applied to stall speeds (minimum speed in the stall) which really were not equivalent

John Tullamarine: "always one needs to check the frozen certification standard for a particular aircraft before throwing numbers into a discussion."

Most important point in this and numerous other discussions!

Zzuf has the best explanation, at least insofar as it corresponds to my memory of what went on in the Flight Test Harmonization Working Group back in the 90's. Earlier certifications were not required to demonstrate stall at a 1G loading; later certs were and will continue to be.

No doubt JF will come into the fray at some stage

.. or, indeed, zzuf - greetings, good sir, and one trusts that the nautical life still treats you well.

Hi JT residence now just S of PH. Nautical pursuits seem to involve vehicles with multiple redundancies, minimum displacement circa 80,000 tonnes, and effective anti slop devices for the champagne.
I should also have previously added that certification stall testing involves two very different test procedures.
1. Stall speed determination, 1 knot per second in the configuration which will give the highest stall speed.
2. Stall handling demonstrations, up to 4 knots per second, power settings of (IIRC) that for around 1.6 Vs, configuration to give the most unsatisfactory handling characteristics. These are highly dynamic stall approaches in both straight and turning flight. The pitch attitude for a straight stall, fairly high power and 4 knot per second can be disconcerting. The stall speeds during such manoeuvres can be very low due to the low g loading in a reducing angle climb and the high approach rate.
For those who may not be aware, high approach rates give a lower stall speed because of alpha over swing beyond the normal stall and airflow separation is delayed.
The low stall speeds contribute to reduced damping of any departure and less response to aerodynamic control inputs.
In some ways it is a pity that line pilots never see how well modern aircraft handle even when grossly abused in this manner.

Deceleration into stall - illuminating discussion.
 

Using the awareness of this 1 knot per second deceleration to find the stall during testing - about delta-g=0.5/9.8=0.05 - illuminates the answer brilliantly. Added on to the downward acceleration as the lift drops associated with the stall.

It's obvious when presented so plainly, yet the picture conjured up by "a stall" is of something happening with almost no changes leading to it, which is also strange when you think about it, since it's obviously something caused in response to changes in circumstances.

Hi guys

A few comments if I may:

Clearly the business of certification in general (not any specific test) is about getting good data about how the aircraft flies. The best (ie repeatable) data comes from measurements taken in as near steady state circumstances as possible. When steady state is not possible then one lets the least important term change – a good example being if you want to plot angle of attack against g at a particular speed. In this case inevitably when tightening the turn you will eventually not have enough thrust available to hold speed and height so you descend as required to hold the speed – in this case letting height change.

When it comes to this topic – involving stalling in some particular configuration of flaps gear etc - then just selecting idle power and holding height (flying club style) the aircraft will decelerate quite rapidly (speed not constant, AoA not constant) and so when the thing eventually ‘stalls’ you will likely overshoot the actual stall AoA and finish up at an even lower speed, an even higher AoA and really in some quite unknown post stall condition where nothing is steady. This can be unpleasant for the inexperienced student and could even become dangerous should a spin result. But the main thing is such an exercise is useless when it comes to getting good (ie repeatable) flight test data.

Hence the slow 1 kt per sec approach to the speed reduction so that the aircraft is in quasi steady state so far as the aerodynamics are concerned with nothing changing rapidly. Of course you have to accept a loss of height to do this. However such a change of height does not really affect the plot you are able to get of IAS vs AoA which is the object of the exercise. Even more importantly (especially with a new type) you are only creeping the AoA up and so if some aerodynamic event occurs (buffet, wing drop, aileron snatch, tendency for the nose to pitch up etc) then you only have to relax the back pressure a tad and you are instantly back to where you were before the event. All nice and safe and leaves you able to record the data as appropriate (well in my day you had to write it down – although today everything will be recorded for you of course).

So far as the 1 kt per sec goes I was always taught that this was a maximum acceptable rate of reduction and the slower the better. So if you took 5 sec to knock another 1 kt off all to the good.

Off topic and speaking personally if ab initio flying instructors taught stalling this way then the student would not get alarmed and actually perhaps enjoy flying slowly rather than being frightened whenever the ASI needle starts to get lower than usual.

:DAnd refresh regularly these basics by airline instructors.

Sorry to dig up this older post after the useful info above, but I figured I'd better be the tech guy here.

What computers can do (if given precise sensors), is detect deviations from the required flightpath, sometimes before they would become perceptible to the Mark 1 homo sapiens, and react more quickly to correct those deviations. This is not due to inherent superiority to human pilots, it's simply because that is all the computers are designed to do. As Tourist points out, some pilots can fly incredibly accurately - enough to give the computers a run for their money in any case.

However, regardless of whether controlled by a human or a computer, an aircraft will always be subject to the laws of physics.

What this means is that if the forces acting on the aircraft exceed the abilities of the aircraft's control authority to hold flightpath and orientation precisely, the computers are programmed to do what pilots are trained to do, namely trade off a degree of precision to maintain stability.

Also on occasion, despite the design, testing and certification processes being among the most thorough and rigorous in history, real world scenarios - usually weather-related - will demonstrate areas that were not taken into account during the design.

Computers are there to assist pilots, not replace them - and I suspect that will remain the case for many years to come.