altimeter calibration
 

Early this year I started a discussion on another forum asking what might cause a discrepancy between altitude calculated from radalt plus ground elevation, and altitude as shown by an altimeter, correctly set on local barometric conditions. The plane in question, a Boeing 757, was travelling at very high speed at low altitude. The consensus was that radalt would not be affected by high speed but the altimeter would likely be out of its calibration envelope.

I have been told that there is a graph in the flight manual which shows errors that would be encountered under various flight conditions. Is there any way to get hold of this graph?

Why do you want to be flying a 757 at high speed (i.e. outside calibration envelope) below 2,500 ft agl?

I don't want to do it. I am looking at the FDR data file, and I see this discrepancy between radalt and pressure and would like to know the cause.

The cause is that they are two different types of instruments that operate on totally different principles and measure two distinctly different parameters.:ok:

Thanks skwinty, that has already become clear from comments on the other thread. What I am looking for now is documetary evidence of the calibration envelope. I want to see what Boeing says about speed and calibration errors. They would have done the measurements, but on their website they do not reveal such details to the public. I am hoping someone can provide the graph.

gravity32, try reading this link. It will have the same and more data that you are seeking.

Pitot-Static Instrument Calibration

Your questions kept referring to the differences between radalt and pressure altitude so I just wanted to remind you to not conflate the two.:)

Additionally, radalt measure the height above the actual ground. Barometric altitude is relative to a notional ground level or sea level and not necessarily relative to the local terrain.

There's no reason why the two should agree, and in many cases they do not, especially if you're looking at FDR data with a fine tooth comb.

As a senior ATPL lecturer of aircraft instrumentation, may I comment?

An altimeter "measures" a vertical displacement from absolutely any pressure datum you wish to choose. It has nothing to do with a "notional ground level" or indeed "sea level".

An altimeter is merely an aneriod barometer, the measurement datum of which may be chosen as you wish.

:)

Thaks skwinty, that is useful but it still does not provide me with Boeing's calibration data. I have been told there is a graph of it in the Flight Manual and that is what I am hoping for.

No problem gravity32,
I hope that some one is able to come up with the graph you are looking for. I cannot help you with that..:)

Sorry, but when calibrated as an ALTImeter, it is using pressure to sense altitude i.e. height. The reference pressure does indeed correspond to a notional ground level, in the case where you set to achieve the correct reading at touchdown, or a notional sea level, which is the higher altitude approach, using the ISA that gives a notional height above notional sea level.

The measurement data used for aeronautical purposes are to enable an instrument to provide a relative height to a reference.

Otherwise, the scale would be in Pa or Mbar, not feet and metres.

Mad (Flt) Scientist and Lightning Mate,

The FDR data file shows that the radalt and the pressure agree very well while the plane is flying at 310 knots at 2500 feet. This was shown by adding the radalt to the ground elevation at each position report and using the known local baro and temp to calculate true altitude. As the plane descends and picks up speed the two diverge, with the pressure indicating a higher altitude than the radalt. The difference between them was about 120 feet at the end of the file. This is quite a big discrepancy.

Some pilots have been telling me that the altimeter is more accurate than radalt. I am looking for Boeing's data to refute them.

Hmm.. Having no idea about the reasons for it I can report that my 737 happily calls 'twenty-five hundred' at 2300ft on the altimeter when coming in for landing over the sea... I guess it would be interesting to look at the accuracy of the RA as advertised by the manufacturer?

That is exactly why you should use your pressure altimeter at altitude and your radalt for ground proximity at low altitude.

Horses for courses.;)

Its a meaningless statement so other than by logic its not refutable.

Radalt measures distance to an object giving a radar return.

Altimeter uses pressure data to infer height above a reference datum.

You really can't compare the two. they aren't even really measuring the same thing. Depending on what you actually want to know, either may be better/"more accurate".

... another 9/11 conspiracy thing, then . :rolleyes:

If you have access to the FDR readout for analysis then, ergo, your involvement is such that you will have direct access to the relevant P/N AFM.

Something doesn't make a lot of sense here, methinks ?

STBYRUD,
Apparently you are getting an 8% error in your radalt. That seems excessive. Don't manufacturers quote 2 or 3 % max? That error would diminish as you descend, as manufacturers quote a max error of 1 foot on the ground. In the FDR file in question the error does not diminish, it increasess as the plane descends, so does not seem to be due to radalt error.

Checkboard,
Yes, this is about flight AA77 which went into the Pentagon. The pilots who are saying the altimeter is more accurate than the radalt are trying to make the case that the plane flew over the Pentagon and that the damage was done by other means, thereby casting suspicion on the authorities. The pilots have an effective website and have convinced a lot of people, which I find pretty annoying.

john_tullamarine,
No, the FDR file was released to the public by the NTSB. Anyone can get it. What seems to be hard to get is the altimeter calibration data.

@gravity32

Trying to convince a 9/11truther (or whatever they are0 is like trying to convince any other believer. The best you can hope for is to remove a piece of evidence, but you'll almost never shake their belief.

if you did manage to produce a convincing argument to show that, for example, Boeing has altimeter calibration charts above Vmo which demonstrate that the altimeter was overreading, they'd just turn round and say something like "how convenient that NOW this is available to aid in the cover up".

Unless you like tilting at windmills ....

Mad (Flt) Scientist,
This is not about tilting at windmills. You are no doubt right that it will be impossible to persuade these pilots that they are wrong, but it will be possible to reduce their influence with the public by proving that these pilots are wrong. This can be done by publishing the proof. The proof must be available in the Boeing calibration data. That is what I would like help to locate, as a public service.

Any graph in the AFM would be for the normal flight envelope. you may be able to extrapolate but the reationship would not be linear and the further away you get from VMo the worse any errors will be. As said, you won't convince the conspiracy theorists no matter how hard you try.

One thing to suggest however is that loads of people must have seen a 757 flying AWAY from the pentagon at 120 ft which I doubt.

mono,
Yes, but the graph will provide the limits to the calibration envelope, and thus prove the plane was outside the envelope. That will be something. It will also be interesting to see in what direction the error may be trending at the boundary.

The argument that the plane would have been seen by many people if it flew over has already been used.

It has been my experience that you will never ruin a good story with facts.

People who believe in conspiracy theories will not be swayed by a calibration graph, which is not surprising given that the graph will prove nothing with respect to the precision of the pressure altimeter on the day in question.

skwinty,
You are apparently not involved in debate with people about these theories so you are not familiar with how it works. As you say there is no point in trying to convince a believer to change his mind, but there is a very much larger body of people making enquiries. These people are not believers. They are doing what everyone does these days, they are surfing the internet for information. They see both sides. They make up their minds on the basis of the weight of evidence. That graph I am hoping for will add to the weight of evidence.

The weight of the evidence lay in the ruins of the Pentagon and WTC.

I have debated with enough twoofers to know that they are delusional.:ugh:

Speed doesnt have an effect on the altitude reading of an Altimeter, how ever altitude can have an effect on speed indication. At a higher altitude there is less "ambient " pressure on , so the pressure due to speed in the pitot probe will be less. This difference is compensated for by mixing in the static air pressure. Speed doesnt affect static pressure, so doesn't alter altitude readings.

An altimeter when calibrated in a lab is set to read a certain calculated standard height for a certain pressure. eg 10000ft = 1013mBar or hecta whosits.

The air pressure over an area changes, so if flying straight and level, over perfectly flat terrain, its possible that an altimeter that has been set at a reference pressure (or even zero feet at the departure point) will show ( and record) a change in height. In this instance ( hypothetical of course) the RAD alt which uses radar waves to measure its true height above the ground (and record it). On examination of the recorded data it would show a difference between the two readings.

Its not a question of accuracy of the calibration of either device , as they are measuring differnt things. the altimeter measures air pressure, and the radalt measures the time it takes for a radar pulse to return to it.

Note, that a radalt is only accurate when aimed vertically at the ground, any tilt in the aircraft axis will give an erroneous reading.

Avtrician,
Speed does have an effect on the altimeter if the plane flies outside the calibration envelope. These issues have already been covered. The plane was nearly level at the end so radalt would not have been upset by angles. The True altitude was calculated from the recorded pressure using the known baro and temp on that day. Radalt plus ground elevation should therefore have been the same as True altitude. They were the same while the plane was travelling at 310 knots at 2500 feet, as expected. They were not the same after the plane descended and picked up speed, therefore one can deduce that the plane was flying outside the calibration envelope for the altimeter. What I am hoping for is documentary evidence that the plane was flying outside its calibration envelope, and for that I need the Boeing calibration graph. Can anyone provide a copy?

gravity32,

Do you think that the terrorists flying the aircraft would have worried about resetting the reference datum of the altimeter en route to their target?

skwinty,
This is a good point to check out. Naturally any adjustment will make a difference to what the pilot will see. However in the case at hand we are using the pressure data in the FDR file. There are 12 flights on the file and in no case is there a step in the pressure, either in ascent or descent. We can assume therefore that this is unadjusted data and will give correct True altitude when adjusted using the known baro and temp at the location on the day.

gravity32,
have a look at this boeing 757 bulletin.

"
The amount of undershoot or overshoot depends on the amount of the barometric correction, when the altimeter is reset, and the elapsed time from altimeter reset to level-off. "

It may help you a bit in your quest.

Boeing 757 Boeing Bulletins :: FMC-16 : FMC VNAV Anomaly During Intermediate Level Off

Thanks skwinty, but I don't think it is relevant as it refers to errors the pilot may encounter shortly after passing through the transition level if he makes a large adjustment. It therefore applies to what the pilot may see on his altimeter, but we are looking at unadjusted data.

Its worth remembering that the earth beneath you is very rarely flat. An extreme example is flying onto a runway in the mountains or on a cliff. The Rad Alt will read every bump and quarry but the altimeter will give a much smoother change since it measures pressure changes that generally are very gradual.

felixthecat,
You are of course right. However this is not relevant to the problem as we used the ground elevation at every position report to produce a corrected radalt altitude. The result was almost as smooth as the calculated pressure altitude.

I read the bulletin as the changing of the barometric setting on the altimeter having an effect on the VNAV signals and FMC systems.

Anyway, I am in uncharted territory here, so good luck in your endeavours.:ok:

gravity32,

See link for analysis of a Boeing 757 altitude data being used in serious incident investigation. The altitude data used in this investigation is FDR QNH corrected altitude data.

Are you sure you are looking at uncorrected data? Perhaps the math you are applying to correct data is causing a problem.

I claim no experience on Boeing aircraft but I am interested in accident/incident investigation.

ETA: Here is another link about pitot static system errors wrt to Boeing.

http://www.boeing.com/commercial/aer...23/aero_23.pdf

REPORT ON THE serious incident to icelandair BOEING 757

gravity32,

Sorry for barging in, but if I understand correctly, the issue is with the FDR record.
Doesn't that show the displayed baro alt ? (Most relevant in a crash analysis.)
And does the FDR show the reference datum set on the altimeter ?

Are you saying the FDR also shows the "raw" static pressure (in mbar or psi or whatever) as 'measured' by the ADC ?
I can think of several reasons for discrepancies between "baro alt" and "rad alt" records on an FDR, depending on where exactly the FDR gets its data. Your "outside the calibration envelope" is only one of them.

In the case of a "CFIT" such as this, I would first look at the RA data, and only then see to what extent they could be correlated with the baro alt data.
RA is pretty accurate, especially low down, which is exactly why it's there in the first place.

Have you got a link to the 'moonbats' site (PM me if you prefer)? First I hear of this particular controversy, so I'm curious to read the original "arguments".

CJ

skwinty and ChristiaanJ

Yes the altitude data is unadjusted in the FDR file. We know it is unadjusted because in all 12 flights on the file there is no jump at the transition level on ascent and descent. The units of the data are feet, so two steps were used to arrive at True altitude, first converting the feet back to pressure then calculating True. On the ground at Dulles this gave the published elevation for Dulles, after adjusting for baro and temp, so we know it is based on standard atmosphere, and so should give correct altitude at destination by applying destination baro and temp. What it gave at the last position was 124 feet higher than radalt plus ground elevation.

The radalt appears to be reliable and accurate on the ground as all 12 flights show -6 feet while on the ground. If we add 16 to this we get wing height, allowing that the radalt should read zero on touchdown. The last radalt reading was 4 feet, giving a wing height of about 20 feet, which looks about right as the plane still had a little distance to go before impact.

The "moonbat" site is Pilots For 9/11 Truth

gravity32,

How can the altitude data be "unadjusted" if it's units are feet?

A raw or unadjusted signal would be measured in pressure units and not feet.

The deviations to the data at transition levels or altimeter adjustments are very small in relation to the overall signal and therefore not easily discerned.

I would say that the data is FDR QNH corrected altitude as per the investigation I linked to earlier.

The data is unadjusted in the sense that nothing has been done to it other than to convert pressure to feet, using standard atmosphere. If you can convert pressure to feet, you can convert backwards from feet to pressure.

The recorded "altitude" of Dulles was 40 feet. Using standard atmosphere this calculates to 29.88 inch Hg. Then using known actual atmosphere this converts to 312.8 feet, which seems about right for Dulles.

Similarly at the end of the flight the raw altitude, -99 feet, calculates to 180 feet. The radio altitude is about 56 feet thus the difference is 124 feet.

Hi gravity32,

Have you seen these?

AA77 FDR Data, Explained - Page 108 - JREF Forum

108 pages of FDR altitude conspiracy theory debunking


Orbitfiles - Download "D226A101-3G.zip" - Free Online storage, Free Online File Backup, Free Online File Sharing


588 pages of Boeing 757 FDR manual.

:ok: