PDA

View Full Version : Why use only pitot-static system for altimeter/airspeed


mirkoni
3rd Feb 2014, 06:39
Hello all,

Yesterday I was watching AF 447 documentary on natgeo. What crossed my mind:
Why there is no backup system if all pitot systems fail - i.e. radio altimeter or at least GPS. I know this is not too precise, but it is better than nothing or even worse, false indications?
Other words, in 21st century, with computerized airplane, we still rely fully on "hole-and-pipe" technology, with no alterantive backup?

Volume
3rd Feb 2014, 06:52
Radio altimeter is not working at that altitude
I know this is not too preciseIt is so inprecise, that it is basically unusable. The difference between (GPS) groundspeed and airspeed can be much larger than the speed margin between stall and overspeed.
For good reasons we do have an alternative for about 100 Yers, it is called pitch and power. But mayby some people have forgotten it since...

welliewanger
3rd Feb 2014, 07:24
Well said Volume. I couldn't have put it better myself.

mirkoni
3rd Feb 2014, 07:27
Thanks for the prompt answer.

Regarding radio-altimeter not working on cruise altitudes, I presume it is because of the time it needs for the beam to go and return back from earth to the airplane which has already "gone" forward?

Regarding altitude what about GPS as alternative? In AF447 they had a false reading that airplane is descending, so copilot puiled up and that is how the trouble started. Although GPS is not too precise reading altitude, wouldn't it at least give a "hint" that they are not descending?

Wizofoz
3rd Feb 2014, 07:40
AFAIK, AF447 had altitude read-outs at all times- it was airspeed that was in error.

I do agree that a read out of GPS altitude might be useful in some circumstances- more so that Rad Alt, as that is actually a reading of height above terrain, not altitude.

Indeed, the smart landing "Altimeter Setting" system works by comparing Baro to GPS altitude.

Similarly, the AF flight WOULD have had a ground-speed reading from it's inertial system- which includes GPS. but ground-speed, as has been pointed out, is of little use as a flight instrument at altitude, and without an air data source, there is no way to calculate IAS or TAS.

Colibris
3rd Feb 2014, 07:47
Some Airbus aircrafts are fitted with the BUSS, you're flying with the AOA and GPS altitude (even if not precise, it's still showing the altitude change trend and is still better than nothing). It becomes active as soon as the 3 ADR's are turned off.

tvrao
3rd Feb 2014, 08:14
One Australian professor has suggested an alternative system using laser beam and Doppler principle to find airspeed as this youtube video shows: Laser speed sensor promises safer air travel - YouTube
Its reliability,accuracy and certification by Aviation authorities is not shown in the video.

awblain
3rd Feb 2014, 08:18
The pitot-static system tells you the local airspeed, which is crucial because you fly in the air around the aircraft, not with respect to the ground.

Some sort of calculated back up might be helpful, but if the temperature or wind changes, the old value won't be the right one to maintain.

A GPS would have told AF447 that they were plummeting to the ocean and their groundspeed was way down from its cruise value, but they should have recognized that already from the altimeter. If there were skeptical of the altimeter, then the GPS might have provided a reference to start to piece things together, not that things should ever have been in that state in the first place.

Volume
3rd Feb 2014, 10:20
Regarding radio-altimeter not working on cruise altitudes, I presume it is because of the time it needs for the beam to go and return back from earth to the airplane which has already "gone" forward?well, it is about the much higher power you need to receive back an echo from 30.000 ft away, and you will have too much scatter from all the different hills on the ground reflecting.

LDA is a great ide, but it works only because there are particles in the air. I am afraid it is too clean up there, to really work. Otherwise, it is absolutely perfect, as it is a truly independent source of information.

AFAIK, AF447 had altitude read-outs at all times- it was airspeed that was in error.The mach correction of static is done by computers using airspeed info, so at least the altitude information was "uncorrected", but probably much more exact than a lot of other information.

ShyTorque
3rd Feb 2014, 10:25
A Radalt can only work accurately if the aerials are pointing to the ground. Extreme bank and pitch angles will affect the reading, aerobatics render it useless.

The Radalts I've been used to only read to 2500 feet.

RTO
3rd Feb 2014, 11:38
Laser laser l a s e r .... Just had to test this, why is the a in laser replaced sometimes with @?

bubbers44
3rd Feb 2014, 13:00
Pitch and power is available in any jet for altitude and weight. This will maintain airspeed and approximate altitude. Pilots have to know how to fly to use this simple procedure. That is today's problem however.

glendalegoon
3rd Feb 2014, 13:13
PItch and Power is right. Bubbers said it all. My aircraft manual has a very nice page of pitch/power numbers just in case of missing radome after hail encounter or some other reason. And the radome area is where the pitot lives.

Talking about radar, wondering how many people remember Doppler Radar Nav which gave us ground speed/drift? I only knew about it because all of our planes of a certain age had the sticker: DOP INOP

G0ULI
3rd Feb 2014, 14:10
RTO
Check out the l@ser pointer thread.
The automatic search engines that insert sponsored ads to support the site were picking up on the word l@ser and inserting ads for high powered l@ser pointers on a thread that was complaining about their use by idiots to target aircraft.

Chris Scott
3rd Feb 2014, 14:13
Pitch and power are great tools, whether you've got a good IAS indication or not - belt and braces? But although the pitch values are easily remembered (remember the mnemonic "(DTD)585" for the B707?), the N1 or EPR settings are less so, unless you have noted the setting just before the IAS indications go AWOL. It can take a while to find the appropriate page in the QRH. On some a/c, the total fuel flow is a better tool than N1 or EPR, but it's not generally taught, and digital flowmeters don't seem to be as user-friendly as the old analogue ones.

I have a pipe dream in which the TAS would be detected by a kind of anemometer mounted at a suitable point on the fuselage. Provided the static pressure and TAT were available, the ADC could then calculate the CAS/IAS, and of course the Mach. But an anemometer might be even more prone to icing than a pitot head, and its RPM could be problematic...

Quote from glendalegoon:
Talking about radar, wondering how many people remember Doppler Radar Nav which gave us ground speed/drift? I only knew about it because all of our planes of a certain age had the sticker: DOP INOP

Suspect you may be betraying your age as sixty-plus? ;) Most Doppler-equipped a/c must have been retrofitted with INS by the late 1970s. IIRC, our B707s had their Doppler removed at the same time. As you no doubt remember, it wasn't much good over smooth sand or calm sea.

G0ULI
3rd Feb 2014, 14:31
Mirkoni
The pitot system is cheap, simple and well understood. The AF447 system iced up due to the pitot heating system not being powerful enough to cope with the prevailing conditions at the time.

As usual, it was a combination of factors that lead to the aircraft crashing. The crew apparently chose to fly through a storm system rather than route around it which made the situation worse due to extreme turbulence while trying to recover the aircraft to stable flight. The sidestick controls didn't provide visual feedback to other cockpit members that the aircraft was being held stalled in a nose up attitude by the pilot flying. The crew were over reliant on automated systems and apparently unable to fly the aircraft manually without an accurate indication of airspeed.

A tragic accident where an otherwise perfectly serviceable aircraft crashed because the pilots were basically incapable of flying the aircraft manually - just one of an increasing number of such incidents.

gums
3rd Feb 2014, 14:42
Yeah, good old doppler. And my age is posted, heh heh.

Even better than the pitch/power rules-of-thumb is an inertially- based flight path marker with a HUD. Speed is still a problem if you are close to the mach limit. Less so if close to stall, as AoA indicators are really great for that. In fact, the AoA can be used at cruise speed, but it is very senstive, like fractions of a degree for a few knots.

BTW, my first inertial system was supplemented by a doppler ( nowadays they use GPS, but still use inertial vectors for accuracy in the tenths of a knot or even less). We could even do INS alignments inflight using the doppler and then providing a nav fix. This was late sixties/early seventies. You can do the same with the GPS.

Chris has a good idea about an old-fashioned anemometer, but the lazer doofer seems to offer a good backup for the hole-in-a-pipe things we use now.

EEngr
3rd Feb 2014, 15:10
I am afraid it is too clean up there, to really work.There is a new technology called molecular tagging velocimetry. It works by tagging (triggering a chemical or quantum state change) in the medium (air) and then observing the relative movement of the marked molecules.

Molecular tagging velocimetry - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/Molecular_tagging_velocimetry)

The down side is that new technologies take time to gain regulatory acceptance.

Desert185
3rd Feb 2014, 16:40
Gouli
Mirkoni
The pitot system is cheap, simple and well understood. The AF447 system iced up due to the pitot heating system not being powerful enough to cope with the prevailing conditions at the time.

As usual, it was a combination of factors that lead to the aircraft crashing. The crew apparently chose to fly through a storm system rather than route around it which made the situation worse due to extreme turbulence while trying to recover the aircraft to stable flight. The sidestick controls didn't provide visual feedback to other cockpit members that the aircraft was being held stalled in a nose up attitude by the pilot flying. The crew were over reliant on automated systems and apparently unable to fly the aircraft manually without an accurate indication of airspeed.

A tragic accident where an otherwise perfectly serviceable aircraft crashed because the pilots were basically incapable of flying the aircraft manually - just one of an increasing number of such incidents.


I fly an atmospheric research DC-8. It is normal when entering heavy precip for the airspeed indicators to go to zero, which is most noticeable in the ITCZ. The autopilot will not trip off and we all know about pitch/power. The INS/GPS continues to chug along with groundspeed, etc. We (science) even have a radio altimeter in the back that accurately reads much higher than the standard 2,500' reading on the pilot's panels.

Not even going out on a limb here, I have to say had AF447 been a DC-8 everyone would still be alive, provided the "pilots" didn't touch anything while the airspeeds read zero. Of course, the pilots wouldn't have had those "cool" little tables in front of them. Oh, and no autothrottles or magenta lines (except for the magenta lines on the pilot's iPads, of course).

Sorry. Had to say it. Them's the facts. Period.

awblain
3rd Feb 2014, 17:31
Could the anemometer icing heater work well enough?
Iced anemometers are likely to give a zero reading?

Maybe a spinning Sidewinder type wheel to measure power from the airflow, or a piezoelectric shear plate to measure stress on a surface not prone to icing.
Someone mentioned having an emergency pop-up spare at some point too.

I'm not sure any of these would work better in bad conditions than the current standard.

HazelNuts39
3rd Feb 2014, 18:01
A backup airspeed and Mach could simply be calculated in the airplane's computer system from altitude, AoA, weight and loadfactor.

Chris Scott
3rd Feb 2014, 18:17
Quote from Desert185:
"Not even going out on a limb here, I have to say had AF447 been a DC-8 everyone would still be alive, provided the "pilots" didn't touch anything while the airspeeds read zero."

Presume you've not flown Airbus FBW? I have, and in previous years negotiated the ITCZ many times in jet types such as B707 and VC10 (both without A/THR), and the A310 and DC10 (both with A/THR). Must admit I never lost the ASIs, even on the rare occasions when the windshields were icing up. But GOULI should recognise that there isn't always a clear way through the ITF, and that the turbulence encountered by AF447 was far from "extreme".

Haven't flown one, but the A330 (even the shorter 200) probably handles fairly similarly to the DC8 in chop. It's a big, stable aircraft. Admittedly, the AP was lost because of the UAS. (The A/THR wouldn't be usable for the pitch/power technique anyway.) The AF447 PF had the advantage of Pitch Alternate law, which would make pitch control easy, but he had to keep the wings level himself. Why he chose consistently to pull up will always remain open to conjecture - I've floated ideas myself. Like you say on your a/c, the trick is to maintain the status-quo.

Now with plenty of time to read and think, I think there is a growing issue in aviation of loss of basic flying skills; partly due to automation and complacency. Airbus has to take its share of the blame, including that for making the MTBF so much longer than it was even in my youth... But so do the airlines and the regulators.

Hats off to anyone who's involved in airborne research into a/c behaviour in adverse wx, particularly in a veteran airframe!

bubbers44
3rd Feb 2014, 19:58
No difference except pilots flying the airplane. Most can, some can't.

gums
3rd Feb 2014, 20:21
Nice, Desert, and Okie was a frequent contributor on the 447 threads.

As one said on the thread, "don't just do something, just sit there".

The 'bus reversion modes do not seem to go to "attitude" hold when the air data is deemed unreliable . They just go back to "manual", such as it is in the FBW system they have. And that isn't too bad, considering the system tries to achieve 1 gee adjusted for pitch attitude. So it resembles an "attitude" control mode. But the pilots tried to do "something" versus taking a moment to analyze what they had and what the jet was doing. Still makes me cry.

Only had one pitot-static freeze and it was the static pressure. So IAS went off the charts as I descended. Fortunately, I had the super flight path marker in my HUD and knew the power settings for the descent. Below 5,000 feet the radar altimiter kicked in and finally the sensors thawed out. 'course, I had good AoA for the approach if the air data stuff remained frozen. This was back in early 70's

I cannot emphasize the value of an inertial flight path marker, and a HUD makes it even more useful. Airspeed goes tango uniform, BFD. You can see your flight path and stay where you want it - TOGA, cruise, descent on a precision approach or non-precision approach ( think Huntsville and Asiana). Sheesh. It was like cheating. With a flight director display, you don't even need the HUD. Just line up the symbols.

Inertial systems these days are very cheap compared to when I first saw them. One neat thing was the inertial vertical velocity bug next to the altitude scale on the HUD. Beat the hell outta the baro indicator, and I think about Asiana. The suckers are also independent of the baro data for short term use, but keeping the FPM on the horizon line was peachy keen when baro things went tango uniform. Then was the power settings we all knew from experience and the manuals.

henry_crun
3rd Feb 2014, 21:08
Unusual airspeed sensor

One little known airspeed sensor used ion flow to measure airspeed. As I recollect it measured electrical ion flow transverse to airflow and a combination of devices could sense airflow in any direction.

It was invented by Ed Durbin, head of Kaiser Electronics, and I had the task of interfacing it with the AH-1T electronics. That aircraft could and probably did interface with anything and everything in the inventory and then some.

As I recollect the sensor was only developed at helicopter speeds and never saw actual airframe service. Dunno if it could be extended to higher speeds. Very useful to have the full vector, rather than just forward speed.

PDF download:
http://www.aerovelco.com/papers/DurbinAirspeedVectorSensor.pdf

Patent:
http://www.docstoc.com/docs/47436556/Fluid-Flow-Rate-And-Direction-Measuring-Apparatus---Patent-4680962

galaxy flyer
3rd Feb 2014, 21:51
Agreed, OK465, the FPV is great SA device. In the Global, we have it, too. Coasting in on northern South America sometime back around 1am local, we plunged into an ITCZ storm that didn't show on radar or lightning detection--severe turbulence, lost about 1,500'. The "moose antlers" we're the clue to start down, now! Shaker fired for just a second, but I'd already started unloading. The FPV showed exactly how much pitch down it would take to stay out of the stall. Anyway, popped out into the clear.

Denti
4th Feb 2014, 03:17
OK465, you might reread your FCOM, the NG FPV is not completely inertial, it does receive baro input and is dependent on the ADC. Simulators might display it differently though as the FPV is not a critical system for operation its simulation often is quite sketchy.

Text from the manual regarding the FPV:

The FPV shows the Flight Path Angle (FPA) above or below the horizon line and drift angle left or right of the pitch scale's center. The FPA uses inertial and barometric altitude inputs. The vertical FPA is unreliable with unreliable primary altitude displays.

If one is absolutely sure that only the speed indication has failed but the barometric input is still correct he can use the FPV in those situations, if only the slightest doubt is there that barometric input might be impaired as well, it is better not to use it.

Aside from that it is of course a nice tool and can be of use in real world flying as well, even if of dubious note in non normal situations.

Volume
4th Feb 2014, 06:58
The pitot system is cheap, simple and well understood. The AF447 system iced up due to the pitot heating system not being powerful enough to cope with the prevailing conditions at the time.So maybe it is not that simple, and especially not fully understood? Pitot has always and will be vulnerable as it is exposed to anything you may find in the air, hail, rain, birds, insects, ice crystals... Pitot may fail in an undetectable way, it is not always obvious that it is not working (Birgenair is probably not the best example, as this was even detected, but ignored as it later on seemed to work).

A backup airspeed and Mach could simply be calculated in the airplane's computer system from altitude, AoA, weight and loadfactor.Unfortunately this is not that easy. As stated already, AoA is perfect at low speed but very, very sensitive at high speed when we need extreme accuracy, while having mach effects on the AoA vane as well. Calculating fails as well if your aerodynamic characteristics are not the ones you assume, e.g. due to icing of the wing. Mach effects also make it hard to calculate bach from lift to speed. So finally you are using a backup which is sensitive to exactly the same effects which may cause your pitot to fail.

LDA is a truly independent system, you may possibly even use it through the cockpit windows, so it is well protected from the environment and for example any ice blocking the light path is evidient to the crew.

On the other hand, why do we need a new system, if all it takes are old pilots...

HazelNuts39
4th Feb 2014, 07:26
LDA is a truly independent system, ...Is it? If it measures TAS, you still need ambient pressure and temperature to determine CAS and Mach. And you still need AoA for stall warning. You need ambient pressure for altitude. No system is truly independent.

Chu Chu
4th Feb 2014, 09:37
Could control effectiveness be used to calculate airspeed? For example, perhaps the FBW system could induce a small, brief, rudder deflection and measure the resulting yaw rate?

Skyjob
4th Feb 2014, 09:56
Due to the inertia on large(r) aircraft this yaw rate would not be noticeable, so I suspect this is not a good option and no real alternative.


As many have pointed out the pitot static system has its limitations but it is relatively well understood. Problems can arise in parts of the system and cause erroneous information but this can be overcome by flying technique.

The problem remains how to be made aware of such failures.
This is an area that needs attention and development, as a recently released image of a heated pitot tube shows, where the tip was actually not heated due to design, where Boeing has since advised of a retrofit by switching polarity of the connectors.

Rather then looking for alternatives, efforts should be made to increase feedback to pilots when a system may have failed, as this is not always annunciated until a feedback loop records a complete failure. Such information would be useful to crew so to know when to ignore the affected instruments and use flying techniques to overcome a downgraded system indication.

We have all been told to fly the magenta line and to trust our instruments at some point in training, as well as to ignore those flight directors and instruments under certain conditions.

It is imperative that recognition of these conditions occurs. It may not always be easy when flying highly automated aircraft to notice such conditions as a crew. A feedback loop that would indicate a problem before failure would enhance the awareness of the crew to an impending failure or equipment downgrade.

gums
4th Feb 2014, 15:57
I have to go with Okie on the utility of a flight path vector, not a flight path angle.

The two jets I flew that used the inertial system to display our flight path vector showed actual vector in space, not simply angle above/below the horizon. 'course, this was because we used the system for weapon delivery, so "drift" was automatically compensated for, as well as other parameters. No more TLBAR when bombing ( that looks 'bout right, heh heh).

The basic inertial platform was also used for the "attitude" indicator ( steam gauge). If the accelerometers went TU, then the sucker still worked like any other attitude indicator, and could be re-initialized as with the older systems, just no velocity vector. remember the J-8 from the 60's?

The beauty of our systems was their independence of air data. As long as we had electrons, we had an excellent display to use if the air data was completely gone, and our vertical position in space was excellent. Now, it is true that we used air data to smooth the vertical position in space ( system altitude), but if the air data went TU, no big deal. Inertial drift of the 70's and 80's platforms was less than one knot/hour in all reference axis.

To summarize, using an inertial platform to provide flight path data and speed and altitude is very useful if the pitot-static system goes west. Granted, the plane flies in the air mass, so actual speed thru the airmass is important. But for short periods, it's a life-saver until your air data sensors thaw out. Even if they don't, you can use the rules-of-thumb for power and attitude. Oh yeah, we had AoA sensors that might still be working if the other sensors froze up. Fine for stall avoidance, not so useful for limiting mach parameter.

Denti
4th Feb 2014, 16:28
The NG FPV has a purely inertial mode and more importantly can still be displayed even if both ADRs are 'TU'. It can function and be useful with only the IRUs to drive it. Naturally whether you use it or not is up to you. Follow your FCOM of course.

Interesting, can you point me to the way to switch it to that mode? In the normal mode it relies on ADC inputs for the vertical part of its display. It can be displayed without that, but is unreliable in that case, which is why boeing removed a note to use it from the unreliable airspeed NNC after they discovered that it can show in fact very misleading information. We had a few cases where the FPV was used in an airspeed unreliable scenario (caused by frozen AoA vanes which also invalidates altitude and GS as well as wind display) in the real world and if followed would have lead to loss of control. Basic pitch and power was of course the way to go and ultimately saved the day.

If you mean with "AoA" information the pitch limit indication which is displayed with flaps deployed or activated stick shaker, those are not pure AoA driven either, they do take inputs from the AoA vane (no surprise there), ADIRU outputs (both IR and AD part), anti-ice controls, wing configuration, thrust and FMC outputs. Quite a lot of information mixed together and computed by the SMYD computers.

I would be very cautious about simulator trials for that stuff, many level D simulators are not very good at correctly simulating some of the things that are not used with standard boeing SOPs, like for example the edge of envelope behavior of the FPV.

gums
4th Feb 2014, 22:10
I am curious about a flight path marker display/system/implentation such as Denti describes. I can understand a crude system using AoA and actual pitch attitude, but why? Inertial platforms these days are extremely reliable and accurate and cheap compared to the stuff we had in the 70's and 80's.

The inertial flight path vector we had was no sierra inertial with respect to Mother Earth. No AoA inputs, no baro inputs. I mean, why? Sheesh. The inertial patform was also used to present basic attitude reference accurate to maybe 0.01 of a degree. Add the accelerometers and get the flight path vector, maybe to 0.02 of a degree ( figure 1 milliradian or so). Add the HUD pitch lines from the basic platform and it was a simple matter to see if you were climbing or descending. Made night takeoffs and missed approaches very comfortable when cross-checking the steam gauges.

Guess I better look up all the avionics on these "new" jets, ya think?

MurphyWasRight
4th Feb 2014, 22:55
Chu Chu : Could control effectiveness be used to calculate airspeed? For example, perhaps the FBW system could induce a small, brief, rudder deflection and measure the resulting yaw rate?

Skyjob: Due to the inertia on large(r) aircraft this yaw rate would not be noticeable, so I suspect this is not a good option and no real alternative.

Reminded me of an idea I had a while back:

I believe some AOA sensors have an integral test driver (think analog meter movement coil) to move the vane, in normal use this is part of an end to end test of the system.

My idea was to use the test driver to force a small defiection (up and down) of the AoA vane, the drive current required for the deflection should be proprotional to airspeed. (actually density and speed?)

Way beyond my knowledge to specifiy other parameters that would be needed to calculate air speed but should be possible.

One nice thing about this is that it would require no additional HW, probably would need a enhanced AoA sensor (calibrated drive) but that would be a direct replacement.

Just a thought, will wait patiently for the royalty checks :-)

lifeafteraviation
5th Feb 2014, 02:49
Some good posts and discussion in this thread.

I don't think anyone's mentioned this but this discussion reminds me of a (hang on...Google it now)...B757 Aeroperú Flight 603 crash back in 1996.

The crew crashed a perfectly good airplane because they allowed a failed static port (taped up during a wash the night before and missed during preflight) to confuse them enough to lose control.

Basically this crash was caused by the crews inability to recognize a simple failure and work around it.

Back to the OP topic....why do we rely on such old technology instead of some new digital technology? Because when it works it works really well and usually it works...it's a very reliable technology.

So...what about when it doesn't work? This crew failed to recognize the problem and work around it....Probably confused because there were multiple failures caused by more than one piece of tape. there are so many other ways to verify altitude and airspeed already on board....the GPS systems being just one of them.

The pressurization controller may also have been taped up but they could have slowly depressurized the airplane in an extreme case to verify altitude. An airplane can never have a significant negative differential so as long as the cabin altitude is say...ten thousand feet...the airplane is at least ten thousand feet or higher and that reading is from inside the cabin...it wouldn't matter if every port on the outside of the plane was taped up.

But...still....what was said over and over in this thread...pitch and power...a B757 will fly just like a single engine Piper or Cessna.

Oh yes....do a thorough pre-flight walk around too.

tdracer
5th Feb 2014, 03:09
Lifeafteraviation


Aeroperu had many things going against it - aside from no usable altitude or airspeed and near zero visibility, they were being mislead by ATC telling them their altitude, with no one realizing that the altitude that ATC saw was just as corrupted as what they saw on their displays. In 20-20 hindsight, they should have used the radio altimeter, but that would have taken some out of the box thinking since they believed that, at the time they hit the water, they were well above the altitude where the RA was meaningful. In short, I doubt many pilots - even old school stick and rudder types - could have made a happy outcome from the pile of http://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/censored.gif they were given.


BTW, in the aftermath of Aeroperu and Bergenair accidents (which coincidentally happened at about the same time, both 757s), Boeing developed specific AFM procedures for dealing with bad or suspect air data indications.


I don't know if those procedures ever had to come into play...

Uplinker
5th Feb 2014, 03:30
I have a pipe dream in which the TAS would be detected by a kind of anemometer mounted at a suitable point on the fuselage. Provided the static pressure and TAT were available, the ADC could then calculate the CAS/IAS, and of course the Mach. But an anemometer might be even more prone to icing than a pitot head, and its RPM could be problematic...

Yes, probably.

What about a 'tongue' made of suitable material sticking out perpendicular to the aircraft fuselage which has no moving parts and which is bent back by the airflow? There would be no pipe to get blocked by insects or ice, and one would just measure the amount of bend electrically caused by the airflow to compute airspeed. You would heat the tongue to keep it de-iced, but this arrangement seems much more robust to me.

Any instrumentation designers or engineers care to comment?

Wizofoz
5th Feb 2014, 04:52
Uplinker- that's exactly what the likes or early Dehaviland Moths were equipped with!!

Galleries | Dehavilland Gypsy Moth Air Speed Indicator - 1929 | Flickr - Photo Sharing! (http://www.flickr.com/photos/mr38/5279127241/galleries/)

henry_crun
5th Feb 2014, 08:11
Okaaaay - let's invent a new one then.

Use three small transducers mounted in a straight line fore-and-aft. Emit pulses of sound from the centre transducer and receive them on the forward and aft ones. Measure the time difference. Compute speed.

Of course noise will be a nuisance, but by using spread spectrum sound and suitable processing this would not be a problem.

Would it attract any attention? Nope, not here. Nobody understood the ion anemometer I posted on page two and I guess nobody will understand this one, apart from Gums who will recognise it as a use of doppler.

Retires all bitter and twisted to kitchen/workshop. Hat, coat, soldering iron....

Centaurus
5th Feb 2014, 11:10
But although the pitch values are easily remembered (remember the mnemonic "(DTD)585" for the B707?), the N1 or EPR settings are less so, unless you have noted the setting just before the IAS indications go AWOL. It can take a while to find the appropriate page in the QRH. On some a/c, the total fuel flow is a better tool than N1 or EPR, but it's not generally taught, and digital flowmeters don't seem to be as user-friendly as the old analogue ones.

a
This is all very elementary knowledge. It should be taught as part of the basic type rating; otherwise you should not be certified on your licence to fly the aircraft. Unfortunately the accent in type ratings is on flying the automatics when there should an equal emphasis placed on manual handling on instruments. There is a difference between becoming a flight deck "Manager" as most pilots are and being a professional fully qualified airman in every sense of the word "Airman".

mm43
5th Feb 2014, 22:30
Shortly before the AF447 accident, a EU Commision supported project sponsored by Airbus, Dassault and Thales, and known as the NESLIE project was sufficiently advanced to be flight tested on a Citation II by the Nederland National Aerospace Laboratory (NLR). This TAS and AoA laser based system was deemed, following over 40 hours of flight tests in all climatic conditions and up to FL410, to be very promising and further technical work is underway to fine tune the results.

A technical paper detailing the NESLIE flight testing work in PDF format is available here (http://reports.nlr.nl:8080/xmlui/bitstream/handle/10921/181/TP-2010-436.pdf?sequence=1).

The DANIELA project (also supported by the EU Commission) has taken the NESLIE project one stage further, i.e. the aim of the DANIELA (http://www.danielaproject.eu/) (Demonstration of ANemometry InstrumEnt based on LAser) project is to prepare the operational use of a flush mounted ADS (Air Data System) built around a 3 axis Doppler LIDAR function as a primary air data channel on civil aircraft, and to assess optical concepts for the measurement of temperature and density to complete the ADS, in a second step.

Overall, these laser based projects are designed to acquire TAS, AOA, density and temperature, as well as being reliable in side-slip situations.

Denti
6th Feb 2014, 06:47
@OK465: what i wanted to convey with the second statement you quoted was: use it at your own risk. I guess it doesn't really contravene the first one.

Now, the FPV is a very handy tool, but be aware of the interconnections in modern aircraft systems and wary of added complexity in case of partial or complete system failures. Always be sure to keep those basic flying skills up, despite all the nice displays and computers they are still very much the basis of our job.

Volume
6th Feb 2014, 06:51
The major issue I see with LDA (obviously meanwile called LIDAR) is that you need at least two reference points (in 3D at least 3) on the aircraft. Not easy on an object which is subjected to quite some deformation. An Aircraft is not a laboraty floor. The referenze points move realtive to each other with temperature, pressurisation, air loads, acceleration, thrust setting etc. I worked with LDA 20 years ago in a wind tunnel, and even there we had issues because auf relative movement due to temperature changes and related deformation. But this is nothing which could not be solved with a few more time and money.

And I still believe it would be a truly independent system, so the overall reliability would no longer be affected by common failure causes, to add this system would reduce the probability of total failure much more, than adding another pitot would.

Tu.114
6th Feb 2014, 09:32
One option that is already proven as an alternate airspeed measurement is to use the AoA vane. On the DH8-300, the AOA is used to feed a fast/slow indicator (looking like a glideslope indication, but on the L/H side of the horizon). It does not deliver numbers but it will be centered when at 1.3 Vs. So it allows to fly an approach at an appropriate speed.

Flap extension speeds need to be deducted via the speed booklet page - when at Va for the present flap setting, one is safe to extend the gear and also a further step of flaps. Apart from that, flying on this was fairly unspectacular. Programming such an indication into a contemporary airliner should be easily done. Unlike on the DH8, it would not necessarily need to be indicated at all times; besides manual selection eg. via a P/B on the EFIS control panel, it might be coupled with the ice detectors and pop up as soon as they are triggered.

Simple, easy to install and almost intuitive to use.

Uplinker
6th Feb 2014, 11:31
Uplinker- that's exactly what the likes or early Dehaviland Moths were equipped with!!


Ah, interesting, Wiz, but not what I meant. The Moth speed gauge has moving parts which could get iced up. My thought was to have a small plate about the size of a credit card sticking out at right angles to the fuselage which is fixed at the base and therefore has no moving parts or joints. This would therefore be less affected by ice. The airflow would cause the plate to bend and this could be measured by piezo electrics or resistive sensors on each surface of the plate.

Only problems I can foresee would be airflow turbulence affecting the plate and not giving a constant reading.

Tu.114
6th Feb 2014, 12:12
Also, if the heating fails or is overwhelmed by the amount of ice accreted, the bending moment will change and it becomes just as inaccurate or useless as an iced-up pitot tube.

What about a birdstrike - if a bird manages not to take off the plate but to bend it, the cosine of the bend angle will need to be calculated out of the measurement (if the system survives).

Also, a plate sticking out in the airflow will cause drag and might cause vibrations/tones at least in a certain speed range.


Well, the idea might be crazy - but why not harness this effect? An array of plates of increasing size and therefore frequency that will start to hum at different airspeeds might be installed in the airflow. Then add a microphone that catches the sound of this contraption and a computer that translates the frequency into an airspeed for indication.

Also, why not turn the pitot tube around? Then, Ptot will no longer be Pstat+Pdyn but Pstat-Pdyn - the much smaller values are likely harder to handle but at least there is no longer an opening facing forward and susceptible to icing.

henry_crun
6th Feb 2014, 12:13
No comments on my acoustic anemometer above? Did nobody reconise it is a stand-alone direct-reading machmeter? No other air data required!

henry_crun
6th Feb 2014, 12:17
Also, why not turn the pitot tube around? Then, Ptot will no longer be Pstat+Pdyn but Pstat-Pdyn - the much smaller values are likely harder to handle but at least there is no longer an opening facing forward and susceptible to icing.

This is done on gliders/sailplanes and produces a variometer output, ie indicates the total energy. But no use for airspeed.

Tu.114
6th Feb 2014, 12:19
I may have a brain fart here, but will there be a doppler effect if both the receivers travel at the same speed and in the same direction as the sender?

Tu.114
6th Feb 2014, 12:24
Henry Crun, indeed, but could this not be used for airspeed indication by feeding the Pstat into the diaphragm and the Pstat-Pdyn into the instrument case (and adjusting the translating mechanism and the scale accordingly)?

henry_crun
6th Feb 2014, 12:53
Tu.114 - sort of double doppler for each path, the sound pulses have to travel further forward than rearward

Sailplane variometers are well described (including math) in the book by Lorne Welch, alas I do not have current access to a copy. You need to express the quantites in algebra and differentiate.

A rearward facing variometer 'pitot' as used in sailplanes is carefully designed and mounted. It looks like a tube with crosswise slots embracing 120 degrees of the rear facing surface and is mounted with its axis vertical above the tailfin. Alternatively electronic computation from pitot and static is possible.

Airspeed sensing pitots are generally pointed forward and clear of the boundary layer. In protoype aircraft they are mounted way out ahead of the aircraft to get the cleanest possible air. Static vents are always a problem because they can only see already disturbed air. It is usual to calibrate them during a series of low-level test flights.

Your arithmetic would be correct if you could find a way of supplying your rear-facing orifice with undisturbed air.

MrSnuggles
6th Feb 2014, 17:13
henry crun

The Doppler-ish air speedometer you are suggesting... thinking about it.. I do wonder if it would work when in turbulence or clouds. Soundwaves would be somewhat absorbed or distorted, don't you think?

henry_crun
7th Feb 2014, 08:31
MrSnuggles - that's where the spread spectrum comes in!

MrSnuggles
7th Feb 2014, 14:40
henry crun

Please forgive me for being thick but does this not result in another radio device? I'm just trying to imagine how to construct such a thing. It is an interesting concept, and would solve the problems with pitots, but it is important to understand what new problems would be introduced with this kind of technology.

HazelNuts39
7th Feb 2014, 15:20
Also, why not turn the pitot tube around? Then, Ptot will no longer be Pstat+Pdyn but Pstat-Pdyn I wonder what an aerodynamicist would have to say about that assumption.

AFAIK sailplanes use a venturi to produce a pressure of Pstat - Pdyn to drive a total-energy variometer.

Did nobody reconise it is a stand-alone direct-reading machmeter? No other air data required!A machmeter indeed, measuring the local Mach close to the airplane. But staying airborne has to do with pressures, so you still need ambient pressure (Pstat) to have something that relates to lift.

henry_crun
7th Feb 2014, 15:54
Hazelnuts

It was only a five minute idea, and it does answer the OP question. I don't presume to have answered the prayers of all maidens everywhere, but I feel the device could have some merit.

Yes, you'd have to compute from air data, the inverse of mach computation, to get a more useful half rho v squared term. I acknowledge that any device which senses q directly would be the ideal solution.

I rather think reading the mach number local to the airfoil surfaces could be useful in some instances, for example where mach is the limiting factor, and of course the device could also be mounted remotely to get a true mach reading.

It also could monitor in three orthogonal directions to provide a full airspeed vector, providing a useful alternative or backup for aoa vanes and sideslip indicators.

Owain Glyndwr
7th Feb 2014, 16:14
HN39

Originally Posted by TU114
Also, why not turn the pitot tube around? Then, Ptot will no longer be Pstat+Pdyn but Pstat-Pdyn
I wonder what an aerodynamicist would have to say about that assumption.An aerodynamicist would say that you won't measure Pstat- Pdyn but Pstat-base pressure on a rearward facing step - say about -0.15*(0.5 rhoV^2).differential.

A machmeter indeed, measuring the local Mach close to the airplane. But staying airborne has to do with pressures, so you still need ambient pressure (Pstat) to have something that relates to lift. I fully agree, with the emphasis on local Mach, and close to the airplane. Unless it sticks well out into the flow it will measure the Mach Number in the boundary layer. If you stick it well out you may have a devil of a job calibrating it. The device, so far as I can see, measures Mach number along its "line of sight" which in general will not be aligned with the local flow (AoA or sidewash or both). I also wonder about sound waves "bounced" off the aircraft skin and how that might interfere with the received signal

pattern_is_full
8th Feb 2014, 05:20
Let's keep in mind that in the case of AF447, the problem was not the general technology of pitot systems, but a specific model from a specific manufacturer that proved to be less-then-perfect when it came to certain icing conditions.

A few years back, an automaker shipped some SUVs with tires that were prone to tread separation and caused a number of crashes. No one took that as a sign that tires were an outmoded technology that needed an alternative.

The solution was not to back up tires with some other technology, but to refrain from building tires with engineering problems in the first place.

I certainly don't object to exploring what alternatives there are for additional backup to the pitot system, airspeed being as critical as it is. I'm just pointing out that well-designed, well-maintained pitot systems do this critical job very simply and effectively in 99.999999% of flights.

Sometimes new technology is a breakthrough in safety and effectiveness - and sometimes it is just a solution in search of a problem.

Owain Glyndwr
8th Feb 2014, 09:32
pattern is full

Hear, Hear:

Chu Chu
8th Feb 2014, 11:44
Well, one result of the SUV tire episode was a new regulation requiring tire pressure monitoring systems. . . .

Owain Glyndwr
8th Feb 2014, 12:53
And one result of AF 447 was that the icing requirements were changed ......

Tu.114
8th Feb 2014, 16:02
This is what is so nice about this place. One comes along having heard a little bit about a subject and there is always someone around to fill one or two of the many gaps in ones rudimentary knowledge. Thank You Owain Glyndwr and Henry_crun for the explanations.

Mad (Flt) Scientist
8th Feb 2014, 20:07
And one result of AF 447 was that the icing requirements were changed ......

But it is not (as of yet, anyway) retroactive, so older aircraft (which means virtually everyone) are potentially at-risk from a similar event.

There's still research going on into how to specify and address these kinds of conditions, as well, so the requirements are not frozen (ha ha) yet. (There's supposed also to be a flight test exercise out of Darwin, Australia, in the next couple of months to help with this; it might already have started)

gums
8th Feb 2014, 21:01
Besides the taped over ports on one plane, then the "mud daubers" plugging up the pitot tube on another, the incident that gets the most attention is AF447. See the thousands of posts on that elsewhere.

I only had one air data freeze in 20 years, and was the static ports. Speed went crazy, but altitude on the steam gauge was stuck. BFD. Power, pitch and and so forth. 'course, I was cheating, as I had an inertial flight path marker and knew by heart the power and descent angle for the let down. Inertial groundspeed was a good crosscheck, even though it did not compensate for actual airspeed. But I also had a valid AoA indication if I got slow. Gotta use all the tools you have, ya think?

Good heaters on the probes/ports are essential. But seems we have had very few losses of the probes due to the heaters and then losses of the plane folowing losses of the probes.

The axiom to hold dear is: "Don't do something, just sit there". The AF447 accident is worth hours and hours of training lectures, and we'll never know what was going through the pilots' minds.

Owain Glyndwr
9th Feb 2014, 06:13
@ Mad Flt Scientist

But it is not (as of yet, anyway) retroactive, so older aircraft (which means virtually everyone) are potentially at-risk from a similar event

True, but the airworthiness authorities (FAA, EASA) had and have the power to make it retroactive if they think it necessary.

One should also remember, as pattern is full said, that:

Let's keep in mind that in the case of AF447, the problem was not the general technology of pitot systems, but a specific model from a specific manufacturer that proved to be less-than-perfect when it came to certain icing conditions.

With that model removed from the supply chain they presumably think retroactivity to be unnecessary.

gums
9th Feb 2014, 22:34
While we're looking for a backup to the pitot-static systems, I could not resist a high tech backup that wasn't so high tech 40 years ago.

First, you must have a decent inertial system, GPS-aided or not.

Then you need a decent computer to provide movement across the earth - velocity and position.

If your nav computer notices that with your heading, that your course is not on the same heading, and also notices that your speed across the earth does not agree with the inertial/nav system numbers if you had zero "wind", then what's up? . Duh? We're flying in the "wind", the air mass. So a fairly simple computer uses actual velocity across the earth and heading with what it should be with no "wind"/air mass movement. You can now compute airspeed. Kinda the reverse of what we used to do way back with "forecast" winds and airspeed indicators and so forth to get course/groundspeed.

Back when the earth was still cooling, I flew a plane with that capability. It was very accurate, and ATC center controllers would frequently ask us what the winds were at our altitude/position. We didn't need actual indicated speed, but we could have easily reversed the equation. Mach could also be calculated using "system altitude" versus the pitot-static system altitude.

In short, if you have the modern systems and a few software folks, it is not that hard to provide "airspeed" thru the airmass without the pitot-static sensors AoA is a different matter.

Skyjob
9th Feb 2014, 23:18
gums, wouldn't such calculation require a known wind speed and direction, which is constantly changing, to calculate the airspeed?

tdracer
10th Feb 2014, 00:07
With that model removed from the supply chain they presumably think retroactivity to be unnecessary.

Bingo! Making a pressure probe that won't ice isn't that hard - you can put as much heat into it as you want (and pretty much anywhere on the probe you want) without significantly affecting the measurement. The only real limitation is the electrical load on the aircraft, and making sure the heater doesn't burn itself up. Boeing ties air data probe heat to 'air/ground' - switching pitot probes from 'low heat' to 'high heat' when in air to help keep from burning up the heaters on the ground when there isn't any cooling air. TAT probes are much tougher since too much heat in the wrong place will corrupt the measurement - as a result corrupted TAT measurements have remained an issue in spite of repeated redesigns.

IIRC, engine inlet P2/T2 probes (which have occasionally had icing issues, especially the temp portion) get ~500 watts heat, while aircraft pitot probes get close to a kilowatt at high heat.
Of course, that doesn't help with the mud dauber or the taped over static. Sometimes there is no substitute for basic maintenance.:ugh:

Uplinker
10th Feb 2014, 07:26
engine inlet P2/T2 probes (which have occasionally had icing issues, especially the temp portion) get ~500 watts heat, while aircraft pitot probes get close to a kilowatt at high heat.

Blimey, that's a lot. Didn't realise it was so much.

Owain Glyndwr
10th Feb 2014, 08:53
@tdracer
Bingo! Making a pressure probe that won't ice isn't that hard - you can put as much heat into it as you want (and pretty much anywhere on the probe you want) without significantly affecting the measurement.

I agree, but with one caveat - it shouldn't be hard provided that the design requirements are known. The problem has been, so far as I can see, that nobody has been able to describe the characteristics of the ice particles with enough certainty to define a foolproof requirement. Since 1996 the FAA have had a working group trying to do just that, interestingly their motivation was a series of engine power loss events not airspeed probes. A good summary of the situation can be found in AIAA 2006 - 206; The Ice Particle Threat to Engines in Flight.

gums
10th Feb 2014, 17:43
@ Skyjob, and thanks for your support, Okie...

The whole point is if your actual movement across the earth is not the same as it should be with zero airmass movement, then the delta is the wind. That's how our nav system calculated the wind values we would provide to the ATC guys. In those days, we would frequently see over 100 knots of wind at 20,000 - 25,000 feet during the winter in the Washington Center area and somewhat south. Our system also calculated true airspeed using the same algorithms, although we had an air data computer that was supposed to do the same thing. So on crossing the pond with a tanker, the tanker nav would ask us for TAS and ground speed and such. He would also ask for lat/long at a time hack. A minute later, we would notice a slight heading change and cracked up. The whole time we were looking at our steering "bug" in the HUD that was slightly off of our heading. The new heading centered the "bug", LOL. Was amazing that our system was so advanced to what the tankers had at the time.

If we knew the movement of the airmass, then it was easy to calculate our speed thru the airmass Equations upon request, but haven't done a lot of this since 1996 or so. Granted, the value we get would not be very accurate for a few seconds, but what the hell. The inertial is extremely accurate for milliseconds of velocity/flight path data. But for nav ya gotta smooth it out, and that takes a second ot two for all the Kalman filters to do their magic.

Mach is harder to calculate even if you know speed thru the airmass, and a basic inertial/GPS will provide altitude within 100 feet or so. Problem is temperature, and mach is directly related to the OAT ( outside air temp), just talk to a SR-71 pilot or nav.

Just thot I would add to all the other proposed widgets to back up the pitot tubes.

Chris Scott
10th Feb 2014, 19:10
Quote:
"Our system also calculated true airspeed using the same algorithms, although we had an air data computer that was supposed to do the same thing."

Errr, gums - I'm also confused. What sensor did your system derive its TAS from?

gums
10th Feb 2014, 19:29
@ Chris

The TAS could have been the actual velocity thru the airmass, and we didn't have the computer use system altitude to calculate the CAS. That was just my own "invention" if we wanted to do it. The winds were easy, and we could display them on our nav panel. Our nav system was better than any other plane for another ten years or so. We had a super inertial, a doppler and a nav computer borrowed from the Apollo lunar landing module - an IBM PI 3, or close to it. So we had about 5 nav modes until it came down to the primitive methods we had all learned. We could even align the INS using the doppler, then get a nav fix for navigation and press on.

BTW, I flew from Hawaii to Guam in our "airmass mode" using predicted winds aloft and only my TAS from the air data computer and attitude/heading from the basic gauges. You know, the way we used to do back when the earth was still cooling and had no doppler or computers or.... I did it because I was in a flight alongside a tanker and the weather was supposed to be good. Besides, I had an ADF for homing if I got lost ( thinking of Amelia,,,,,)

I was about 15 miles off when getting to Guam. Not too shabby. Island was easy to spot due to the clouds above, so I would have made it without help.

Denti
10th Feb 2014, 20:05
@OK465

If i remember correctly it was removed in 2012 or early 2013, but i can't really recall the date.

Haven't been lucky enough to fly aircraft with the AoA gauge which seems to be indicated where some have the round dial RA display, there is an alternate purely digital RA display location on the bottom of the ADI part of the PFD. We have both RA displays in our fleet. The AoA display is available on the 737NG, 764, 777 and probably 787 and 748.

I did understand the same as you that it displays unfactored AoA, however with a red tick for stall AoA which is dependent on mach number. Boeing published a nice article about it in their aero magazine (http://www.boeing.com/commercial/aeromagazine/aero_12/attack_story.html).

Wouldn't have helped in the majority of our unreliable airspeed cases as those were caused by frozen AoA vanes. Interestingly enough that resulted in wrong indications of airspeed, altitude, ground speed and wind info on the affected side. Luckily there was always only one sided affected and the ISFD doesn't take any AoA input and displayed correct information as well which made the crosscheck easier.

yanrair
17th Feb 2014, 15:53
The GPS is a true life saver in a loss of indicated airspeed. When AF lost their airspeeds they only had to look at ground speed that existed at the time and then keep it there. It would have been around 440 since they had a bit of a headwind at the time I seem to remember but it doesn't matter. In still air it will be 480, and in a 200kt headwind it will be 280 but all you do is keep it there.
Your forecast wind on the flight plan is dead accurate these days and so all you need to do all the way to Paris is to subtract or add the headwind to find out what GPS groundspeed should be and fly that speed. You won' t be more than five its off your target. A lot better than what happened - being 350 kts too slow!
John

yanrair
17th Feb 2014, 16:05
Hi there awblain
The GPS in AF 477 during the descent would have been reading less than 100 its. They were actually descending faster than that VERTICALLY! The pitot static system icing up can render the Altimeter and the airspeeds invalid - don' t know what the situation was with air france. The big clue is that the ground speed was 350 knots too slow. It simply amazes me that many airline pilots don't realise the value of GPS Ground Speed to stabilise the flight path Way back in the 80s, we flew the 737-200 with total failure of airspeed and altimeters back to a save landing using G/S and of course, pitch and power but the real saviour was ground speed. Say your landing speed is to be 130 and the headwind is 30 knots you just fly the approach at 100 kts ground speed and you will land at precisely 130 kts airspeed. The Tristar used to do this routinely and funnily enough, the Airbus does the same in managed speed mode (if it is working , and this is the problem).
It flies the approach using groundspeed to fine tune the approach speed.
Cheers for now
John

yanrair
17th Feb 2014, 16:38
Another thing that most folks haven't spotted is this. Once they developed a deep stall with forward speed below 100 and vertical speed in excess of 100kts the angle of attack would have been greater than 45 degrees. That means that to unstall they would have had to lower the nose from what they had (15 deg nose up) to say 50 deg nose down to stand any chance of unstalling. That is a pitch change of 60 degrees - in the dark and with them in a state of panic not knowing what was going on. It was all over by the time they reached 25000 ft. Nothing after that was going to work. So, when you get into airspeed unreliable territory, fix onto ground speed and you will stay safe.
The rules are simple
Maintain pitch and power that exists and maintain the same ground speed. All is well.