AA77 FDR Data, Explained

[beachnut]

Hi beachnut,

Where did you find your information on the pressure altitude accuracy?

When you say that it was shown that the pressure altitude was reading high, do you mean it was shown by the radio height or something else?

If there is lag during descent, would there be lag during ascent as well?

Warren.

All rational pilots know the limitations of pressure altitude. 36 years ago I knew the errors of pressure altitude when I was first trained to fly. I was a USAF pilot for 28 years; chief instructor pilot for my squadron of heavy jets. I was on active duty on 911 actively flying. I have an ATP, ...etc

It is possible for the pressure altitude to read what you decoded from the FDR with 77 only 10 or 20 feet above the ground on track to impact.

Using the RADALT, Google earth, and pressure altitude you can see 77 is reading high on takeoff.

There is lag in the pressure altitude because there are lines (pipes) filled with air and subject to the physics of air. Pilots lead level-off by 10 percent of the VVI. For the 3600 fpm reading on the last second, that would be 360 foot lead, 6 seconds ahead for no lead, and about 10 or 12 seconds to level off. Lag is up and down.






Practical example -

1. A pilot sets his altimeter to the local altimeter setting and begins an ILS approach. He begins his missed approach when the altimeter reads the decision height shown on the approach plate. According to the approach plate, the DH corresponds to 200 AGL, but how low could he actually be?
   He might actually be as low as 100 feet AGL (or even less). Since a 75 foot altimeter error is allowed for IFR operations, if the altimeter we reading 75 feet too high, this could result in being 75 feet lower than you think you are when you reach the decision height. There may be additional error if it is very cold. For example, if it is -15 degrees C, you could be an additional 25 feet lower (AIM 7-2-3). It would be better practice to adjust up the decision height for any known altimeter error and any error due to cold. http://gadoid.ices.cmu.edu/flying/answers.html

http://www.risingup.com/fars/info/part43-E-APPX.shtml

 

[W.D.Clinger]

Hi beachnut,

I don't know if you realise, however the quote you posted was actually from a thread Another Question For Warren Stutt, concerning altitude. I've been replying in that thread.

Warren.

ETA: Rob has since added another reply in that thread that fixes the math.

Contrary to Rob Balsamo's calculation of 2223g, later changed to 58g by removing a "small error", the last four seconds of FDR data recovered by Warren show a maximum vertical force of 2.2g, and are consistent with collision with the first light pole (about 80 feet above sea level, about 1016 feet from the Pentagon). See the first graph of altitude versus distance at
http://www.ccs.neu.edu/home/will/Music/Jokes/Balsamo/Software/

Note that the approximately 80-foot altitude at -1016 feet was calculated without assuming any lag in pressure altitude. Assuming a lag in pressure altitude would move the corrected pressure altitude to the left, reducing the altitude at -1016
feet. That would induce changes in the initial (really final) state that would produce a near-matching reduction in the calculated altitude as well.

In short, Balsamo's calculations are worthless. As Warren pointed out, Balsamo somehow assumed Warren had recovered only 2 new seconds instead of 4. Correcting that one mistake would change the time base from 0.3 to 2.3 seconds, so Rob's calculation would become

[latex]200/(2.3 * 2.3) = 37.8 \hbox{ft/sec}^2[/latex]

which is a little less than 1.2g. Adding 1g for gravity, we get 2.2g, which is only a little higher than the average 1.8g recorded by the FDR for the last two seconds. The small remaining discrepancy is explained by a couple of relatively minor mistakes on Rob's part, notably his insistence upon an overly large correction to the pressure altitude.

Will

 

[W.D.Clinger]

In my previous post, I gave Rob Balsamo too much credit here:

In short, Balsamo's calculations are worthless. As Warren pointed out, Balsamo somehow assumed Warren had recovered only 2 new seconds instead of 4. Correcting that one mistake would change the time base from 0.3 to 2.3 seconds, so Rob's calculation would become

[latex]200/(2.3 * 2.3) = 37.8 \hbox{ft/sec}^2[/latex]

which is a little less than 1.2g. Adding 1g for gravity, we get 2.2g, which is only a little higher than the average 1.8g recorded by the FDR for the last two seconds. The small remaining discrepancy is explained by a couple of relatively minor mistakes on Rob's part, notably his insistence upon an overly large correction to the pressure altitude.

In reaity, Balsamo's calculation was so wrong that it was just a matter of chance that correcting his time base produced a result in the right ball park. As Warren pointed out, Rob used the wrong equation: The vertical velocity was nonzero at both ends of the time interval (as we now know from the new FDR data Warren recovered), so Rob's equation doesn't apply. My remark about the last two seconds was irrelevant, because Balsamo appears to have been talking about an interval between the VDOT antenna and the first light pole.

To review: Balsamo's original calculation of 11.2g was off by a factor of almost 6. His "corrected" calculation of 10.14g was off by a factor of more than 5.

Now that Warren has recovered FDR data for the final seconds so we no longer have to guess, Balsamo calculated 2223g. That was off by a factor of 1000. After "correcting" that to 58g, Balsamo remains off by a factor of more than 25.

The FDR shows a maximum vertical acceleration of about 2.3g (including 1g of gravity), with an average of 1.8g for the last two seconds (which includes the encounter with the first light pole). If we assume the end of data comes near the Pentagon, then numerical integration using the FDR accelerations yields an altitude of about 80 feet above sea level at the first light pole, consistent with the physical damage to that pole. There is also an excellent fit between the altitudes obtained by numerical integration for the final seconds and the shape of the pressure altitude curve recorded by the FDR.

http://www.ccs.neu.edu/home/will/Music/Jokes/Balsamo/Software/

 

[beachnut]

...
To review: Balsamo's original calculation of 11.2g was off by a factor of almost 6. His "corrected" calculation of 10.14g was off by a factor of more than 5.

Now that Warren has recovered FDR data for the final seconds so we no longer have to guess, Balsamo calculated 2223g. That was off by a factor of 1000. After "correcting" that to 58g, Balsamo remains off by a factor of more than 25. ...

His corrections are worse than his errors. He has the G force given to him in the FDR decode and prefers to expose his ignorance in math and logical thinking by making up junk.

His own words when a p4t decode using stolen software was produced.

Balsamo said, "A radar altimeter presents no lag, 273 is a hard number above the ground."

The final second of FDR decode now has a "HARD NUMBER", it is 4, FOUR, feet above the ground according to the criteria used by Balsamo.

There is no p4t explanation based on evidence on how the FDR was found in the Pentagon.

 

[wstutt]

Hi beachnut,

<snip>

Using the RADALT, Google earth, and pressure altitude you can see 77 is reading high on takeoff.

<snip>

The pressure altitude recorded by the FDR shows about 40 feet prior to takeoff. The barometer corrections show 30.2 and 30.21.

Using an online altitude correction program, and using delta T = 0, the true altitude for the above pressure altitude is 297 feet or 306 feet depending on which barometer correction is used.

It appears to me that the flight took off from Runway 30 of Dulles. According to AirNav, the elevation of that runway 287.8 ft.

This is only 18.2 feet at most below the calculated value, so at this point the pressure altitude recorded by the FDR is not more than about 18 feet too high. Is this correct?

Warren.

 

[beachnut]

Hi beachnut,

The pressure altitude recorded by the FDR shows about 40 feet prior to takeoff. The barometer corrections show 30.2 and 30.21.

Using an online altitude correction program, and using delta T = 0, the true altitude for the above pressure altitude is 297 feet or 306 feet depending on which barometer correction is used.

It appears to me that the flight took off from Runway 30 of Dulles. According to AirNav, the elevation of that runway 287.8 ft.

This is only 18.2 feet at most below the calculated value, so at this point the pressure altitude recorded by the FDR is not more than about 18 feet too high. Is this correct?

Warren.

Pilots put in the local QNH of 30.21. What are the known errors? The QNH the pilots put in the baro is based on reported values for a specific location, the pressure varies. Want to guess why AIM says +-75 feet for ground check?

Runway 30 is 287.8 feet and the same runway at the other end is Runway 12 at 309.8 feet. Is that the touchdown elevation? What is the touchdown zone? What were Flight 93 errors?

You are right, you can say the error is 18 feet, but the AirNAV is an average height and I have to look up how it is calculated before I can explain it. The runway is actually sloping. What is the real error?

During the takeoff roll and 77 is still on the ground what are the errors? Just after lift off, you can use the RADALT to estimate error in the first 100 feet or so.

Studies for pressure altitude show how errors change with speed. Some aircraft have errors that show low and as the airspeed increases the error moves to showing high. Some aircraft start high and go higher.

What exact value is stored in the FDR, raw pressure altitude or corrected pressure altitude. By corrected I mean the many errors in the altimeter system are modeled and removed. Have you seen all the errors in the system?

Using your data, I found over 50 foot errors while 77 is on the ground and in flight after takeoff. I looked at many points.

The RADALT reads four feet; could be when it passed over the overpass and HIT the trees one second before impact. RADALT is 1 foot accurate for low altitude, the pressure altitude is +-75 sitting still, not moving, on the ground. (even the nut case Balsamo agrees with RADALT being perfect; 1 foot error)

USAF procedures ...

4.6.1.2. Compare. Compare the indicated altitude to the elevation of a known checkpoint. The maximum allowable error is 75 feet. If the altimeter error exceeds 75 feet, the instrument is out of tolerance for instrument flight.

FAA procedures ...

3. Note the variation between the known field elevation and the altimeter indication. If this variation is in the order of plus or minus 75 feet, the accuracy of the altimeter is questionable and the problem should be referred to an appropriately rated repair station for evaluation and possible correction. AIM 7-2-3 download at http://www.faa.gov/air_traffic/publications/ATPubs/AIM/AIM_Basic_2-14-08.pdf http://www.faa.gov/air_traffic/publications/

Why is the check +-75, yet numbers from 20 feet to 70 feet are found? Because there are many errors in the pressure system and googling the altimeter errors usually finds specific errors, not the practical end user guide of +-75 feet. In the last seconds we have pressure altitude of plane well above certification speed, descending at 3600 feet per minute, in a 2 g pull-up and we are discussing a value which can be about 75 feet off in static conditions? The final seconds of Flight 77 are outside the envelope for using the pressure altitude for anything useful.

Flying works find with the altimeter and it's many errors. We have systems in the aircraft to model the errors and correct them in planes with ADCs. I have flown with altimeters corrected but they still have errors.

If you doubt 77 impacted the Pentagon you are not rational. If you want to study errors in the systems and learn more then that is great. But there is no doubt Flight 77 impacted the Pentagon. I was interested in 911 due to the claims of Flight 77 making some fantastic maneuver. The FDR reveals a poorly flown plane with less than standard flying skills required. Charlie Sheen knows nothing about flying, just like Balsamo the failed pilot who can't hit the side of building over 900 feet wide and 70 feet tall in the safety of a simulator.


Back to errors, in AIM they discuss some and give guidance where more information hides.

a. Most pressure altimeters are subject to mechanical, elastic, temperature, and installation errors. (Detailed information regarding the use of pressure altimeters is found in the Instrument Flying Handbook, Chapter IV.) Although manufacturing and installation specifications, as well as the periodic test and inspections required by regulations (14 CFR Part 43, Appendix E), act to reduce these errors, any scale error may be observed in the following manner:

I have in good faith given information revealing why pressure altitude could be 300 feet off in certain conditions. The truthers spew off the top of their head junk ideas and dismiss facts in favor of lies and Balsamo's failed physics; what is he up to now 58 or 70 Gs. How stupid does Balsamo have to be for the cult at p4t forum to praise his stupidity?

We know exactly where Flight 77 hit, the final true track heading is confirmed in the FDR with impact damage on the lampposts. If you want the final word on the Pressure Altitude offset at and prior to impact you must figure out exactly where Flight 77 is for any one of the final seconds, and you can calculate the error in the Pressure altitude. The fact is the pressure altitude looks like it says the plane is higher.

I believe the data made it to the FDR secure chip in 50 to 200 ms. What do you think?
Having studied FDRs, I saw typical delays in the 100 ms region for various sensors. If the data was stored quickly, the final lat and long acc could be the trailer impact; if so we know exactly where 77 is to back up the time line and figure out errors. The only thing missing is actual accurate ground elevations. But the 4 foot RADALT reading would be the overpass sometime near 77 impacting the trees and passing the VDOT camera.

I never doubted 77 impacted the Pentagon, I knew UBL would kill Americans when and where the opportunity came; who didn't?

Good work decoding the data Balsamo sat on for years.


I found greater than 20 foot to 50 foot errors during take off and right after lift off.

Kind of useless to use pressure altitude in the final seconds when the FAA acknowledge the tolerance to be +-75 when the aircraft is sitting still, let alone traveling 483 knots puling 1.7 g and descending at 3600 feet per minute. The normal approach speed is less than 200 KIAS, no pulling g, and 500 to 750 feet per minute decent. We are not in the ballpark for using the pressure altitude as a source for accuracy. There is a big reason the FDR was found in the Pentagon.

 

[tfk]

His corrections are worse than his errors. He has the G force given to him in the FDR decode and prefers to expose his ignorance in math and logical thinking by making up junk.

His own words when a p4t decode using stolen software was produced.
The final second of FDR decode now has a "HARD NUMBER", it is 4, FOUR, feet above the ground according to the criteria used by Balsamo.

There is no p4t explanation based on evidence on how the FDR was found in the Pentagon.

You're right. Robby couldn't calculate his way out of the 5th grade.

One other slight correction, tho.

The radar altimeter is clearly calibrated to read zero as the wheels touch down on the runway.

You can easily see that it reads -6' while the plane is taxiing around. Apparently about 6' of compression on the wheel struts. (Perhaps they include the swing of the main gear too.)

From Boeing's scale drawings, the belly of the plane is still 6' above the tarmac at this point, while sitting on compressed struts. Slight variation due to loads, of course.

So, in order to get the height agl of the belly of the plane, you need to add 12' to the reading.

At 4' reading, the belly of the plane was about 16 feet above whatever was returning the pulse.

Tom

 

[beachnut]

I was using 10 feet below the bottom of the jet for 4 feet, but 16 for 4 feet sounds better. Zero should happen with the aircraft touches down, landing gear extended down. It worked for the KC-135 to the foot at touchdown, 3, 2, 1, touchdown.

You do have to add some to the RADALT to get the actual height of the plane above the ground since the landing gear is not down, the RADALT reading is from the bottom of the wheels before touchdown.

 

[wstutt]

Hi beachnut,

<snip>

You do have to add some to the RADALT to get the actual height of the plane above the ground since the landing gear is not down, the RADALT reading is from the bottom of the wheels before touchdown.

The Boeing data frame layout document D226A101-3G.pdf confirms this. It states that the positive sense for the RADIO HEIGHT parameters are ABOVE TOUCHDN.

Warren.

 

[tfk]

Beach,

That was a superb summary of the big picture.

The only point that I'd add is:
1. we knew professional pilots were at the controls when the plane took off.
2. we know that a poor pilot was at the controls from midway thru the flight.
3. we know that they had to open the door to swap pilots
4. we know that the FDR does not record the FD door opening
5. ergo: the FD door was not recorded in the FDR.

Runway 30 is 287.8 feet and the same runway at the other end is Runway 12 at 309.8 feet. Is that the touchdown elevation? ...
During the takeoff roll and 77 is still on the ground what are the errors?

Assume the elevations apply to the ends of the runway. Assume continuous linear slope. Close enough.

This gives about a 20' elevation rise from start to end of take off roll. (Assume he used most, not all, of runway.)

And yet, from the start of the roll until the START of rotation, the PA continuously increases in reading from 40 to 74 ft. This is a 34 foot increase, with only a 20' actual rise in the runway. (Yeah, I know there might be a hump in the middle of the runway, but the plane is going to be close to one end or the other at start & departure.)

The extra 14 feet is easily explained by the increase in speed & Bernoulli's effect. The faster you go, the lower the static pressure reading. In this case, 0 to 132 kts produced an static pressure decrease that was equivalent to a 14' increase in altitude reading.

During the rotation, but before liftoff, you can see the effects of the AoA on the PA reading.

During rotation, the pitch angle rises from 0° to about 8°, which increases the static pressure port elevation by about 50'(sin 8°) = 7', and yet the PA reading decreases by 25' (74' @ subframe 146712 to 49' @ subframe 146717). Obviously due to changes in air flow direction over the fuselage & past the static port.

There are two possible interpretation of the above. Either one of which reinforce what you are saying about errors in instrumentation. (And also destroys Balsamo's idiotic "we have to believe that the PA is precise.")

Option 1: The data recorded in the FDR is uncorrected for the various effects (airspeed and AoA) that introduce errors.

Option 2: The corrections are made for the normal portions of the flight envelope, and once you are significantly outside of that regime, your corrections are no longer valid.

Either way, at 480 kts & 50' MSL, Balsamo's "the PA must be considered to be exact" is a load of hooey.


Tom

 

[beachnut]

Hi beachnut,

The Boeing data frame layout document D226A101-3G.pdf confirms this. It states that the positive sense for the RADIO HEIGHT parameters are ABOVE TOUCHDN.

Warren.

That is the purpose of the RADALT. The Flight Director on planes uses the RADALT to show the runway coming up to you. You can read the instrument and know the instant your wheels will touch. With the pressure altitude system errors you can't use pressure altitude to gage your height above touchdown.

Your last value of 4 feet is about 16 feet below 77. IF the FDR last data was the impact of the Trailer, then the 4 foot reading was back about 800 to 900 feet, the overpass where 77 hit a tree and was knocking down the first lamppost. The time values are stored in the FDR will help figure out the final path.

 

[tfk]

Ooops, a correction to my post 4050 above...

It was sloppy to attribute the errors to Bernoulli effects. Worse than sloppy, it was incorrect.

Rather, errors in the PA reading would be due to the different flowstreams patterns that the air takes at different speeds as it curves around the body of the fuselage.

Look at this diagram:

This shows the variation in the static pressure along that line. From the static pressure distribution, you can see that, at any given speed, the static pressure can read too high, too low or just right.This is exactly the reason that they place the static port at a point where the static pressure adjacent to the body of the plane is equal to the freestream static pressure. (i.e., delta(p)/qc = 0).


If it happens that the curvature of the flowstream at the static pressure port is zero, then the static pressure will read accurately. If the curvature (perpendicular to the body of the plane) is convex (curving away from from the plane) then the pressure will read high. And if it is concave (curving towards the plane) then the pressure will read low.

We can see from photos (like this one below) that the position of the static port on the 757 is approximately at position 3 on the chart above. But it is displaced downward, level with the wing roots. At this elevation on a 757, it looks like point 3 has moved somewhat forward.

A second consideration is that this static pressure profile is technically correct for ONLY ONE airspeed and attitude. To a first order approximation, the shape of the curve ought to stay approximately constant over a typical range of flying speeds, with only the magnitudes changing. But again, this is only an approximation. At "takeoff roll" speeds, I'd expect this profile to be significantly different.

Clearly, the airspeed & attitude for which the port location will be chosen is cruise configuration. Obviously, with ADCs they could apply corrections to the reading for different air speeds & AoAs. I don't know if they choose to do that with either the data presented in the cockpit, or with the data sent to the FDR, or with both.

But from the PA data alone from the takeoff roll, it seems evident that, either:

1. they do not correct for airspeed and AoA,
or
2. those corrections are not accurate down to takeoff speeds.

If the corrections were applied AND accurate, then the PA reading would have matched the elevation rise of the runway during the takeoff roll and would have risen slightly during rotation.

Clearly, the PA values did neither.

Tom

 

[apathoid]

Clearly, the airspeed & attitude for which the port location will be chosen is cruise configuration. Obviously, with ADCs they could apply corrections to the reading for different air speeds & AoAs. I don't know if they choose to do that with either the data presented in the cockpit, or with the data sent to the FDR, or with both.

But from the PA data alone from the takeoff roll, it seems evident that, either:

1. they do not correct for airspeed and AoA,
or
2. those corrections are not accurate down to takeoff speeds.

I believe the opposite is true actually. I mentioned quite a number of times that we don't test for low altitude accuracy with high speeds. We use low speeds/low altitudes and high speeds/high altitudes. The ADC does do a number of corrections but only within its calibrated range. Off the top of my head, for the 757 we test at:

0' 120 kts
5000' 200 kts
20000 300 kts
30000 350 kts
40000 400 kts

The altitude is corrected for AoA above mach 0.55. I noticed the altitudes were all reading 120' low on one side during one my checks and found that the protractor was installed wrong. I mentioned above(probably a couple hundred pages back by now) that this fact may explain, by itself, why the altitude reading was off. But it would be foolish to ignore all the other factors, like lag and source error.

 

[tfk]

I believe the opposite is true actually. I mentioned quite a number of times that we don't test for low altitude accuracy with high speeds. We use low speeds/low altitudes and high speeds/high altitudes. The ADC does do a number of corrections but only within its calibrated range. Off the top of my head, for the 757 we test at:

0' 120 kts
5000' 200 kts
20000 300 kts
30000 350 kts
40000 400 kts

The altitude is corrected for AoA above mach 0.55. I noticed the altitudes were all reading 120' low on one side during one my checks and found that the protractor was installed wrong. I mentioned above(probably a couple hundred pages back by now) that this fact may explain, by itself, why the altitude reading was off. But it would be foolish to ignore all the other factors, like lag and source error.

Hey Ap,

It took me about 3 readings to figure out where we differed. Everything you said I was saying to myself, "yup, yup, makes sense, yup..."

I see, they do calibration at low altitude, low speed. Cool. Thanks for clearing that up.

And it makes sense, too. After all, the numbers that I was quoting from the data were 15' off on roll, 25' off during rotation, etc. Not big numbers, and I'd expect within allowable tolerances.

Thanks for the info.

Tom

 

[wstutt]

Hi beachnut,

<snip>

What exact value is stored in the FDR, raw pressure altitude or corrected pressure altitude. By corrected I mean the many errors in the altimeter system are modeled and removed. Have you seen all the errors in the system?

I've thought about whether raw or corrected pressure altitude was stored in the FDR. I notice that the FDR shows steady pressure altitudes of 29000, 33000 and 35000 feet with the barometer corrections set to 29.92 and 29.91 as the aircraft was ascending. Presumably the pilots were seeing the same altitudes on their altimeters while maintaining the altitudes they were given by air traffic control. So it would seem that at least at these parts of the flight, either corrected pressure altitude was being recorded by the FDR or the required correction was negligible.

<snip>

If you doubt 77 impacted the Pentagon you are not rational. ...
<snip>

I have not yet come to a conclusion as to whether the pressure altitude recorded in the FDR is consistent with that aircraft impacting the Pentagon. Perhaps the pressure altitude recorded in the FDR for the final few seconds has large errors as you say. I'm still working on it.

Warren.

 

[apathoid]

Hi beachnut,

I've thought about whether raw or corrected pressure altitude was stored in the FDR. I notice that the FDR shows steady pressure altitudes of 29000, 33000 and 35000 feet with the barometer corrections set to 29.92 and 29.91 as the aircraft was ascending. Presumably the pilots were seeing the same altitudes on their altimeters while maintaining the altitudes they were given by air traffic control. So it would seem that at least at these parts of the flight, either corrected pressure altitude was being recorded by the FDR or the required correction was negligible.

I have not yet come to a conclusion as to whether the pressure altitude recorded in the FDR is consistent with that aircraft impacting the Pentagon. Perhaps the pressure altitude recorded in the FDR for the final few seconds has large errors as you say. I'm still working on it.

Warren.

I wouldn't call these errors large. There are just many variables and the 757's static system wasn't certified to be accurate in that type of flight regime. The lag and source error beachnut keeps talking about is very real. I see it when I test and certify these systems, particularly the lag as I've rarely encountered source errors of 50' or greater.

I appreciate your dispassionate approach to this problem Warren, and if you ever need any other technical assistance...take what you get from Rob with a boulder size grain of salt, and look here first.

Thanks again for all the hard work!

 

[funk de fino]

I have not yet come to a conclusion as to whether the pressure altitude recorded in the FDR is consistent with that aircraft impacting the Pentagon. Perhaps the pressure altitude recorded in the FDR for the final few seconds has large errors as you say. I'm still working on it.

Warren.

I used to test, maintain and calibrate pitot static systems on aircraft at first line.

This is totally different than testing an altimeter on the bench at 2nd or 3rd line. The errors are greater and are affected by more things. Temperature and the test equipment (and operator) being only two of them. The altimeters in the aircraft we used to operate were only tested to around 75 feet accuracy during the tests and they were deemed to be in limit if they passed this test. The lag could be pretty bad as well in a lot of the systems depending on the design and maintenance of the system. The aircraft at the Pentagon was flying very low and very fast and this itself would also have introduced errors into the altimeter. If it was an insrument landing in next to zero visibility then the pilots did not use the altimeter. They used the rad alt.

 

[Caustic Logic]

I have not yet come to a conclusion as to whether the pressure altitude recorded in the FDR is consistent with that aircraft impacting the Pentagon. Perhaps the pressure altitude recorded in the FDR for the final few seconds has large errors as you say. I'm still working on it.

Warren.

Hi. I'm not even going to claim to know a sliver of the math stuff here. It's amazing stuff at first and I get much of the stuff what's in English. But when I see this much hard science and mathematics and this much (apparent?) disagreement coexist I sense something's wrong. This comment might identify it. I can attest with non-formulaic basic reasoning that IF this FDR had ad anything to do with the plane we've been calling Flight 77, it could only be one altitude in the last frames and that's the one consistent with where the plane and its black boxes wound up a few hundred feet into the Pentagon's ground floor.

Just be sure you aren't taking the PFFFT CIT Twoofer path of trying to re-write reality with poor reflections rather than correct their reflections to match reality. Either way, enjoy your forays, whatever the heck it is you're up to.



ETA: And I speak as someone who's not even all current on this discussion lately, so apologies if I seem horribly off-base here.

 

[BCR]

Just be sure you aren't taking the PFFFT CIT Twoofer path of trying to re-write reality with poor reflections rather than correct their reflections to match reality. Either way, enjoy your forays, whatever the heck it is you're up to.



ETA: And I speak as someone who's not even all current on this discussion lately, so apologies if I seem horribly off-base here.

Warren is a data-driven guy. There really is no 'disagreement' that the PA does not respond to rapid changes in altitude at high speeds very well. That is clearly demonstrated by an simple evaluation of RA and physics parameters in the final seconds of flight. The 'disagreement' is the why. All of the techno-stuff which is outside my area of expertise is indeed interesting to watch. I went through this years ago when I confirmed the FDR outputs were missing some data. The same was the case then, there was no doubt that it was missing data. The only remaining question was why. Similarly, the battle of the 'experts' throwing around techno-stuff.

In the end, it is the data that rules. The data indicates solidly that the PA was not responding to rapid changes during the low altitude/high speed conditions that existed in the moments before impact. From where I sit, I really don't care why that is, but it is an interesting debate between the 'experts' and an opportunity to learn something new.

 

[stewieg]

In the end, it is the data that rules.

PfT/CIT concerned with the data?