A question about airspeed sensors

[Phil Bunch]

I have a question about airliner airspeed sensors, which are exposed to the external atmosphere and which are apparently able to become blocked by insects, ice, protective tape covering the openings during some maintenance procedures, etc.

Would it be simple and useful to build a basic test mechanism into the external port so that a small burst of air could be injected into the port?  This way, if the injection-related indicated pressure changed in a way that implied mechanical blockage of the external opening, an "air speed sensor opening blocked" warning could be displayed, alarms sounded, etc.  If the pulsed injection of air were done say once per minute, there would be fewer situations where the pilots would not be confident about the basic integrity of their air speed sensors.  I realize such a test is not a comprehensive test for air speed sensor accuracy but at least it would eliminate most cases where ice or insects or externally applied protective tape made the sensors worse than useless without the pilots knowing they had a problem.  The speed monitoring system would ignore indicated air speed for a short time while the test pulse was injected.

Just a thought.  Obviously, I don't have enough background to make a serious suggestion to the aviation community.

[Hardy Heinlin]

That "small burst of air" to test the exact calibration – or just to clean the tube?

|-|

[Qavion]

Where would you find such a finely tuned source of calibrated air pressure? And surely a burst of air would change the cockpit displays or even damage the ADC/ADIRU sensors. And would this burst of calibrated air not be susceptible to most of the things the pitots are susceptible to?

One particular aircraft I came across recently was suffering from a 6 month history of sometimes unreliable TAT indications on the ground. The TAT sensors use air from the pneumatic bleed manifold to create a venturi in the TAT probe, which draws air across the wire temperature sensing element when the aircraft is not moving. Without a source of bleed air, the airflow is stagnant and direct sunlight on the TAT probes can result in high TAT readings.

It was found that the TAT bleed air plumbing wasn't hooked up at the pnuematic bleed manifold end after a pax to freigher conversion. I'm guessing that your calibrated air source would be just as susceptible to maintenance errors as any other system fitted to the aircraft.

Actually, I get the impression that the A330 has a design issue. Perhaps hearsay, but I hear rumours that the heating elements are not powerful enough to eliminate ice in all conditions. In this case, perhaps we should be building better mousetraps, not designing mousetrap activation sensors.

The 747-400 pitot/static system has anti-ice heating element failure detectors. I'm sure that the Airbus has something similar (and there would have been a message on ECAM to alert the pilots). However, these systems only detect the loss of electrical current to the heating element, not the presence of icing. Also, heavy rain has been known to bring down aircraft. Removing ice is simply a matter of melting it and letting the water drain away. If the probe has more water in it than can drain away... then you have problems.

CHeeRs
Q>

[Phil Bunch]

"I'm guessing that your calibrated air source would be just as susceptible to maintenance errors as any other system fitted to the aircraft."
--------------------------------

I fear this is the a really bottom line issue re why one must be VERY averse to changing something as reliable as airliners.  "Better is the enemy of good" to quote an often-cited aphorism.  

Their airline accident rate, per year or per zillion passengers is now so low that one would not be able to statistically prove that one has lowered the accident rate by making most types of improvements, I suspect.  As you describe, engineering something like this to work in real life under a severe range of operating conditions and situations, and essentially never interfering with or confusing the air speed subsystem is quite a challenge.  

Actually, I have never understood how airliners can possibly exist, applying these and so many other standards to so many, many components and systems that must operate simultaneously with such a low failure rate and with great resilience.  Clearly, it is impossible to design, build, and operate a safe and reliable airliner, from a probability standpoint!! (grins)  

BTW, I wasn't thinking of a calibrated air test pulse - just something to verify that the port is open as a pass/fail test.  However, since the port might not be completely blocked at the instant of the test, a crude test wouldn't pick up a partial but significant blockage, and a crude test that falsely indicates that the air speed sensor is OK might be just as dangerous as a complete blockage now that I think about it.  False air speed readings might be provided with a partially blocked port.  And, as you mention, if you add an independent test component to the subsystem, that component would need testing and maintenance too.  The more interacting parts you add, the greater the chances for a failure, and the potential for a cascade of failures becomes larger.

Intuitively, I would guess that self-testing the integrity and functionality of air speed sensors has not been done for lots of reasons - as I understand it (not very well, of course), there are plenty of self-tests applied to the airliner's many subsystems already, and the critical importance of the air speed measurement systems would surely make them a prime candidate for self-test and self-monitoring if it were safe and practical.

Thanks for your comments - as always, they are very informative and very interesting to those of us who enjoy these and related topics.

[Zinger]

In "my time", there was a ground test set used to verify the pitot and static system integrity which utilized pressurized air.

I have written a few articles regarding cockpit automation problems. In an aircraft with 4 pitot tubes and 6 static ports such as the Boeing 747-400, it is not straightforward for the crew to sort out an airspeed display error, due to the various probabilities, but once done correctly, there are alternate sources of information for them to fly the right airspeed. The human problem is logical fault isolation when under multiple piloting task pressure. It has been demonstrated during accident investigations, that the problem also exists in simulator sessions.
The critical characteristics required for candidates to command complex aircraft is personbality plus training oriented.
a. You need to select pilots with proper cognitive skills (e.g. creative decision making in unusual circumstances)
b. They need prior experience flying the boundaries of the flight envelope.
Remember the Hudson River captain?

Last, cockpit automation must utilize enhanced cues to display the multitude of information presented, in a more comprehensible way.

[Jeroen Hoppenbrouwers]

The whole issue is that cockpit automation probably is more geared towards less stress on pilots, so that you can reduce their number (currently down to two) and, possibly, training/experience level... because of money...

[Dominic Manzer]

From and engineering point of view an air puff self test makes a lot of sense. It would find many of the common (ground) problems, covered/taped over ports, bugs (insects,) detached tubing. It might be best implemented as a peazo electric acoustic transducer/pump mounted at the back of the tube. The static sources would need the test as well.

Most in flight problems, electronics failures, detached tubing, transducer malfunction/calibrations and even bird strikes could be quickly detected and unambiguously resolved. But Icing is more subtle. Sure it would detect complete blockage at the the aperture and do a great job detecting a build up within the tube that restricts the flow or resonance properties. But Icing builds up on the exterior disrupting flow around the probe changing the response. Ice on Flt 744 would have built up slowly, nearly as a thin film.

The static sources are particularly sensitive. A small amount of ice just behind the static port increases the upstream pressure anding an airspeed component to the static signal. These subtle changes in response are difficult to diagnoses. If the pitot problem is not failing heaters but inadequate heaters Flt 744 would have had ice on all the pitots causing all to shift in the same direction, by differing amounts.

Although a self test probably would not have helped Flt 744 there have been plenty that would have been. I'm confident that 744's problem will be found, fixed and not recur. I've been on many failure review boards, and this is the definitive solution type. But random failures, bugs and taped over ports will keep happening. Self test is defiantly worth a try.

[Phil Bunch]

That "small burst of air" to test the exact calibration – or just to clean the tube?

|-|

I hadn't thought of the test as an exact calibration test, but only as something to crudely test for complete blockage.  

However, I just realized a puff of air might force ice particles, dirt, or insect bodies into the port, permanently blocking it and making things worse!  

I fear that my new "career" as an avionics engineer isn't getting off to such a good start!!!

[Dominic Manzer]

I was thinking of a puff from inside the tube out word.  It could remove debris, although this was not my intent. New debris could not be introduces.

When a closed back port (the port is not connected through tubing to remote equipment ) is taped over, flexing of the tape provides enough volume flow so that the operation of the port appears normal at first. A puff from the outside would fail to detect the blocked port in this case.

In an internal test with a volume of gas introduced of only 1% of the volume of the tube, a permanent offset in attitude of 600 ft would be created in a completely blocked tube. A similar large offset in air speed would be created, but I need to get out the text books to calculate the size. The point is that any deviation from the normal profile of a test would be flagged as showing an unreliable air data system. This test would show internal blockages before they have opportunity to cause a crash.

The required volume of gas is so small that a single CO2 cartridge, as used in air guns, would supply enough gas for tens of thousands of tests. If a problem is detected in flight, a port clearing could be performed with a CO2 system that is strong enough to remove covers or tape.

This is a fertile area to investigate. The test could be implemented in many ways. Assessing the benefits and potential new risks will determine witch if any test methods should be implemented.

[Zinger]

Pitot static tubings incorporate a drain facility for maintenance action, such as clear humidity, dust snd other foreign objects.
A foreign object which constitutes a constriction can be detected by use of the system test sets.