AAIB recommendation on periodic air tests

[Genghis the Engineer]

AAIB report on PA28-140 fatality (http://www.aaib.gov.uk/cms_resources/Piper%20PA-28-140%20Cherokee,%20G-AVRP%2010-08.pdf) (Last page, last recommendation.)

When EASA eliminated the periodic air tests which had been normal, at-least to us in the UK, I did a bit of digging to see what their evidence was. It appeared at the time that there was no evidence that air tests were providing any safety benefit because, by and large, maintenance shops only released paperwork to the CAA once the aircraft was found to be compliant. Thus, it appeared to the central authority that air tests were an expense with no safety benefit and there was no central source of information to contradict them.

At the same time, my experience, albeit with sub-ICAO / permit to fly aeroplanes is that about 1 time in 3 I've conducted an air test (for such aircraft, they're required annually for such aircraft in the UK), I've found some performance or handling deficiency which needed correction. Similarly in a former existence I quite often sat in the back of Tucanos for post maintenance air tests, and whilst as you'd expect the maintenance was generally very good, it wasn't unknown to find faults during the air test.

How does anybody else feel on this? For my money, AAIB is absolutely correct in recommending that a periodic "performance check" (air test!) should be required, particularly for aircraft flown by on non public transport operations where this is less likely to get spotted in the course of normal flying.

G

[Shawn Coyle]

Genghis:
I believe the aim of the air test was mainly to prove that the performance was still meeting the FM posted performance.
Off the top of my head, I can only think of one accident where the aircraft didn't meet the performance published (and this appeared to be a generic, as opposed to airframe serial number specific problem).
And I suppose the authorities are going to say that Safety Management Systems will take care of this as well.
It would be nice to see something like this written in to the ops regulations somewhere so that we could be sure that the aircraft were meeting the certification requirements they were originally approved under, but I'm not overly worried if the annual air test is not something the authorities have to approve on an annual basis.

[Karl Bamforth]

At the opposite end of the scale Diamond require a "flight check" after every scheduled maintenance event. IE. 100 hrs.

OK not as in depth as a CofA flight test but still checks most things.

In private hands this might be once a year.

Here we do 23,000 hrs a year almost all on Diamond aircraft, I will let you do the math but its a lot of hours.

[Pilot DAR]

The report raises a number of important elements, and is, in my opinion, adequately objective in the discussion of the cam lobe wear issue, and the lack of understanding of the actual effect of that wear on engine performance. I cannot assert the affect of this amount of cam lobe wear, but I've seen engines in service (which certainly should not have been) with more! Checking performance is not a good measure of this, checking for metal in the oil filter or screen is. There is a reference to metal resulting from this wear being detected, an that's where to hang your hat on this issue.

However:

The report completely overlooks an important characteristic of the PA28 series of aircraft, which I'm certain would have been a major contributing factor in this accident...

PA28's have a stabilator rather than the conventional stabilizer/elevator combination. In "normal" operations this is a completely benign difference. Where it does show up, however, is in very low speed, nose high operations. Here's what happens: Pilot commands more nose up at low speed, takeoff being the most likely occasion. Where the conventional stab/elev combination would have a greater deflection of the elevator, the camber of the horizontal tail as a single flying surface is increased (as well as a change in effective pitch angle). As such, it's capacity to create more "down" lift before stalling, or experiencing a large drag increase is better. (okay purist aerodynamicists, have at me, I've got experience with the result, not the theory). On the other hand, the stabilator, has only the opportunity to change angle of attack to create more "down" lift. Like any flying surface, it will reach an angle of attack, where the Cl max has been passed, and the increase in drag is no longer proportionate to the increase in (down) lift. Yes, I'm sure that Piper gave this a lot of thought and test all those years ago, and their engineers know more than I do about this, but...

I was the right seat check pilot to a junior pilot in an Arrow PA28R-200 many years ago. He rotated prematurely out of a 1000 meter runway. It was a hot day, but we were quite light, and it was a 200HP version of the aircraft, so power was not a problem. The plane was stuck in ground effect, with the pilot holding the controls 'way back. The stall warning was sounding. The aircraft would not accelerate, or climb away. Landing back might have worked, other than the runway end was approaching, and he (well I suppose "we") had allowed the aircraft to drift off the side, so there was not a runway under us any more. I was shocked at the "poor performance" I was experiencing in this aircraft I thought I knew so well. We were nearing the obstacles. We did not wreck the plane simply because I "locked off" and retracted the gear. The resultant reduction in drag was all that was needed to allow the plane to slowly climb away.

Shocked by what had just happened, I went to experiment. I flew a PA28-140, which I believed to be in good condition, off a very long frozen lake (runway and width length no longer a concern). Sure enough, I could get the plane very nose high, with lots of aft control, and it would not accelerate or climb out of ground effect. My only option was to land back (fixed gear). I repeated this configuration enough to satisfy myself that this is a configuration to be avoided in PA28's (and probably Cessna 177's as well) So I do! The thick wing has great lift, and resists stalling well. The stabilator tail is not ideally matched in this attribute.

It is my opinion that a stabilator on a PA28 can be taken to an angle of attack which is just enough to stall it, not perhaps to the point where it looses all lift, but to the point where it's being stalled just creates so much drag, that the power of the aircraft cannot overcome it, so as to accelerate and climb away.

A pilot, particularly with only a hundred plus hours on type, who was rapidly becoming concerned about running out of runway, would be likely to over rotate the aircraft at low speed, and enter this phase of flight. This would explain how the aircraft, at the same (or heavier weight) had taken off with no problem earlier in the day. This phase of flight is very stable, and not in itself alarming, until the pilot realizes that there is nothing left, and more is needed. The lower power PA28's are the most vulnerable. You don't hear of this with the higher powered variants, or the Aztec's, simply because they have the power to mask this characteristic.

So, it is my opinion that the AAIB report has overlooked this characteristic of these aircraft, and it's potential to create huge drag, and loss of performance during takeoff. That said, gathering data on this characteristic has great potential to be dangerous, if not conducted with ample runway. Perhaps analysis (which is well beyond my capabilities) would be a help, but I am not the one to say...

As to the original question, it is my opinion that performance testing as a routine element of maintenance activity, in the realm of this type of aircraft, is probably not adequately repeatable, so as to produce meaningful results. There is too much opportunity for variation in condition, and pilot technique. These aircraft and engines have a half century of operation, and their strengths and weaknesses are well understood. The accumulated maintenance procedures for these engines, when undertaken correctly, produce consistently airworthy engines. Changing the requirements of maintenance on them now will have little effect on aviation safety. The other airframe related defects which could affect performance should also be easily spotted during a proper inspection.

I'm not saying that a flight test is a bad idea following maintenance, and I've certainly found my fair share of defects during such flights, but they were not often performance related, more things which just did not work the way they were supposed to. Perhaps pilot awareness of characteristics affecting takeoff performance in PA28's would be a more worthwhile objective. (Uh oh, I'm backhandedly suggesting a type endorsement, for one of the world's most common aircrat types! I'd better not do that!)

My thoughts for the evening...

Pilot DAR

[idle stop]

Speaking from Europe, or thereabouts...
Within at least some JAR(and EU-Ops?) Commercial Air Transport operators, and certainly the one with which I am associated, the CofA Airtest has become a Quality System item:the Check Test Flight. So it continues to be carried out, under another guise. Helpfully, the UK CAA has replaced the former CofA schedules with CTF schedules.
The question is then: should all Part M providers (now that Private and Aerial Work has gone Part M also) put a CTF profile requirement into their Quality Systems, so that all aircraft have such a periodic review?

[stevef]

As a point of interest, we once had a PA28 in the hangar for an annual inspection. We found the stabilator to be badly mis-rigged and adjusted it in accordance with the Piper MM. The aircraft was subsequently test flown by a very experienced ex-RAF flying instructor/civilian formation display pilot and was reported as handling correctly.
The owner collected his aircraft, took off and immediately returned (or maybe aborted the t/o, I forget after 20 years) with a 'pitch problem.
He'd been used to flying a mis-rigged aircraft...