AA77 FDR Data, Explained

[W.D.Clinger]

Hi Will,

Perhaps you could use the PITCH ANGLE CAPT parameter that is recorded 4 times per second rather than the FLT DIR - PITCH CAPT parameter which is only recorded once every 4 seconds.

I started out using PITCH ANGLE CAPT (DEG) and PITCH ANGLE F/O (DEG), but the results weren't consistent with changes in pressure altitude and radar heights. The FLT DIR - PITCH CAPT parameter would seem to have been included for calculations such as mine.

Can anyone tell me why these parameters would have such different values?

My guess is that the two PITCH ANGLE parameters were measured with respect to some reference other than the direction of flight. If I knew that reference, I might be able to extend my calculations back before t=-20 seconds.

It appears to me that the aircraft could have hit the light pole as placed in R.Mackey's graph in The Reconstruction of Hanjour's Final Approach, however it shows the aircraft still descending when it hits the Pentagon rather than having levelled off as in the DoD 5 frames video. Does anyone have any comments on why this is?

With my initial conditions, which were chosen because they gave a close match to the pressure altitude, the aircraft is still descending at about a 3% grade at end of data, which is less than 2 degrees from horizontal. IMO, the one frame of the DoD 5 frames video that shows the aircraft doesn't offer enough detail to distinguish a 3% grade from level flight.

The aircraft had been descending at about a 10% grade just a few seconds earlier. It levelled out a lot during the last two seconds, which averaged 1.8g of vertical acceleration (including 1g gravity). By comparison to its previous rate of descent, the 800 or so feet it covered during that last second would have looked pretty level.

Another possibility is that the recorded pressure altitudes were a little off, as has been speculated. I'm now inclined to disregard that speculation, because there's a pretty good match between the pressure altitudes for the last few seconds and my reconstruction from the accelerometer data.

As you noted, another possibility is that our assumption about the end of data coinciding with the generator trailer or Pentagon could be wrong. John Farmer suggested that the longitudinal accelerometer might have had such fast transient response as to fail to record impacts that didn't coincide with the instant of sampling. If that were true, then impacts that did happen to coincide with the instant of sampling would register as much larger; hence the pegging of the last longitudinal acceleration could have been an impact with a light pole. I think that's highly unlikely; to improve the accuracy of reconstructions such as mine, the longitudinal accelerometer should have been selected for a transient response that would perform some averaging over the 1/4-second sampling interval.

Without the 4 new seconds you uncovered, we wouldn't have been able to discuss these things sensibly. Thanks again, Warren.

Will

 

[R.Mackey]

Pitch Angle should be measured relative to the freestream, i.e. the direction of flight.

But keep in mind that the pitch angle primarily influences lift, and thus acceleration, not vertical velocity directly. Lift will be roughly linear in pitch angle (measured in radians) for small values of pitch. Also lift at zero pitch is probably insufficient to counteract gravity -- there will be a constant term in the equation.

ETA: "Should Be" doesn't mean it is. If they aren't following the mathematical convention it could also be measured from the horizon. I don't know how they have it wired.

 

[beachnut]

levelled off as in the DoD 5 frames video.

Flight 77 was not level in the video; the lens is NOT accurate.

 

[W.D.Clinger]

Pitch Angle should be measured relative to the freestream, i.e. the direction of flight.

I'm pretty sure they weren't, because they don't match up with (the additive inverse of) the FLT DIR - PITCH CAPT (DEG) parameter. Here are the values recorded for those parameters during the last 12 seconds:
(time in seconds), PITCH ANGLE CAPT (DEG), PITCH ANGLE F/O (DEG), FLT DIR - PITCH CAPT (DEG):

-12, -7.6, -8.1, 12.5
-11, -8.3, -7.9
-10, -7.4, -6.9
-09, -6.7, -6.9
-08, -6.5, -6.2, 11.6
-07, -5.3, -4.7
-06, -4.2, -4.6
-05, -4.9, -5.3
-04, -5.8, -5.6, 11.1
-03, -5.6, -5.6
-02, -4.2, -3.2
-01, -2.3, -1.9

ETA: "Should Be" doesn't mean it is. If they aren't following the mathematical convention it could also be measured from the horizon. I don't know how they have it wired.

Maybe. That interpretation pushes the divergence back from about 20 seconds to about 50 seconds, and makes the divergence go in the opposite direction. On the other hand, the fit between calculated and pressure altitude isn't quite as good for the last 10 seconds.

With that interpretation, and using those parameters instead of the FLT DIR - PITCH CAPT (DEG) parameter, it looks as though there may be some systematic calibration or alignment error in the pitch parameters for which I might be able to correct. Alternatively, I could use the pitch angle relative to the horizon to correct my original open-loop calculation. I'll look at this more when I have time.

 

[Reheat]

I'm just guessing here. In looking at these numbers it might be that the pitch angle is being measured from whatever is selected on the ADI. That setting is adjustable by the pilots on the ADI. The FLT DIR - PITCH likely correlates to whatever mode is set in the Flight Director i.e. Altitude Hold. I haven't looked at the previous parameters for AA 77, but I did review them for UA 93. Level flight at cruise was indicating about +3 degrees (IIRC) on UA93. If you are persistent and determined to find the reference datum you're likely going to have to review previous data to figure it out. Also, if you can determine the mode setting on the Flight Director and when it was set that might help.

I seriously doubt the hijackers adjusted either the ADI or the Flight Director, so whatever was set prior to the hijacking is likely the datum you are seeing here.

 

[wstutt]

<snip>

With my initial conditions, which were chosen because they gave a close match to the pressure altitude, the aircraft is still descending at about a 3% grade at end of data, which is less than 2 degrees from horizontal. IMO, the one frame of the DoD 5 frames video that shows the aircraft doesn't offer enough detail to distinguish a 3% grade from level flight.

OK, point taken. It just looked steeper from the graph but of course the graph is not 1:1!

<snip>
Another possibility is that the recorded pressure altitudes were a little off, as has been speculated. I'm now inclined to disregard that speculation, because there's a pretty good match between the pressure altitudes for the last few seconds and my reconstruction from the accelerometer data.

Thanks for doing the reconstruction and finding that out. That makes me more comfortable in using the final 20 seconds of pressure altitudes to determine vertical speed.

As you noted, another possibility is that our assumption about the end of data coinciding with the generator trailer or Pentagon could be wrong. John Farmer suggested that the longitudinal accelerometer might have had such fast transient response as to fail to record impacts that didn't coincide with the instant of sampling. If that were true, then impacts that did happen to coincide with the instant of sampling would register as much larger; hence the pegging of the last longitudinal acceleration could have been an impact with a light pole. I think that's highly unlikely; to improve the accuracy of reconstructions such as mine, the longitudinal accelerometer should have been selected for a transient response that would perform some averaging over the 1/4-second sampling interval.

The lateral and longitudinal accelerations are recorded in the data frame only 2 words apart so I think whatever event caused the pegging of the last longitudinal acceleration should have caused the large lateral acceleration as well. According to the data frame layout, positive lateral acceleration is to the right so this large lateral acceleration would be to the left in the same sense that the large negative longitudinal acceleration is to the back of the aircraft. There are only 16 more words of data after the last pair of lateral/longitudinal accelerations which is only 0.0625 seconds at the data rate of 256 words per second.

Without the 4 new seconds you uncovered, we wouldn't have been able to discuss these things sensibly. Thanks again, Warren.

Will

Your welcome. Thanks again for your work,

Warren.

 

[W.D.Clinger]

I'm just guessing here. In looking at these numbers it might be that the pitch angle is being measured from whatever is selected on the ADI. That setting is adjustable by the pilots on the ADI. The FLT DIR - PITCH likely correlates to whatever mode is set in the Flight Director i.e. Altitude Hold. I haven't looked at the previous parameters for AA 77, but I did review them for UA 93. Level flight at cruise was indicating about +3 degrees (IIRC) on UA93. If you are persistent and determined to find the reference datum you're likely going to have to review previous data to figure it out. Also, if you can determine the mode setting on the Flight Director and when it was set that might help.

Thanks, Reheat. Your post pointed me to this:
John O'Callaghan and Daniel Bower, PhD. Study of Autopilot, Navigation Equipment, and Fuel Consumption Activity Based on United Airlines Flight 93 and American Airlines Flight 77 Digital Flight Data Recorder Information.
February 13, 2002, NTSB #DCA01MA064, http://www.ntsb.gov/Info/autopilot_AA77_UA93_study.pdf

The Flight Director vertical mode was set to Altitude Hold for the last 6 minutes. I presume that means the FLT DIR - PITCH CAPT (DEG) parameter was recording the difference between actual pitch and the pitch needed to maintain level flight. That's confirmed by its 0.0 reading during periods of level flight.

The ADI setting doesn't seem to have been recorded by the FDR, so I still don't quite know how to interpret the two pitch angle parameters, but your information helped. The pitch angles recorded during level flight at 35000 or 25000 feet were 1.8 or 1.9 degrees.

 

[Reheat]

The ADI setting doesn't seem to have been recorded by the FDR, so I still don't quite know how to interpret the two pitch angle parameters, but your information helped.

Yes that would be a a virtually impossible parameter to record by itself, but as I stated that it was a guess. I currently have no other clue why the Capt and FO pitch angles are different (unless it's based on ADI settings. That's the only thing which might explain the difference. Perhaps if you look at the differences previously recorded we could determine. Bear in mind that the ADI setting itself doesn't need to be recorded (with reference to what?) only the difference between what's set and what's being flown.

I do think you've identified the FLT DIR PITCH issue. Bear in mind that the Flight Director doesn't actually know the Altitude IIRC, it works from a differential aspect when it's set. It will adjust if the altitude changes SIGNIFICANTLY. It's been a while, but I think that is correct.

 

[tfk]

For the record, as I'm sure you understand, I used a parabolic model in my derivation because (a) at the time we had no data for that part of the flight at all, and (b) a parabola corresponds to constant vertical acceleration, making it the simplest model that answers the question of whether the wings would fall off.

Now that we have more data, this is not the right approach; instead we can simply fit through the aircraft sensors and/or integrate the control inputs to get a much more accurate path.

Exactly.

You do the best you can with all the data that you have at hand at any given moment.

When more data appears, you use it. And there's no reason to consider the previous work "wrong".

Successively better analyses.

It's like everything else. It's the folks that'll never step up & produce anything original that are the oh-so-annoying second-guessers.

Tom

 

[Reheat]

Bear in mind that the Flight Director doesn't actually know the Altitude IIRC, it works from a differential aspect when it's set. It will adjust if the altitude changes SIGNIFICANTLY. It's been a while, but I think that is correct.

Will, I apologize as I suspect you've noted this is wrong. Obviously, the Flight Director (FMS) has altitude info from the Air Data Computer (ADC). What I was attempting to say here is that once either the Auto Pilot or Flight Director are disengaged, they no longer have an altitude reference. In attempting to say that I bunged it.

I am not knowledgeable of the specifics of 7x7 FMS, however I am familiar with the systems generically.

 

[moorea34]

Thank you Warren for your Job....


With your file:
http://www.warrenstutt.com/AAL77FDRDecoder/OutputFiles/AllSubframesFltDeckDoor.csv

You can know the 12 flights done by the plane before the crash:

Flight Time Latitude Longitude Latitude Longitude Departure Arrival

FL 00 1:20:20 38.52716 -99.11951 41.97567 -87.89560 San Francisco (?) Chicago
FL 01 2:30:40 41.97567 -87.89560 28.43708 -81.31840 Chicago Orlando
FL 02 2:37:08 28.43021 -81.31136 41.95078 -87.90144 Orlando Chicago
FL 03 4:20:00 41.97001 -87.89886 33.93951 -118.39828 Chicago Los Angeles
FL 04 4:19:24 33.94484 -118.40601 41.97464 -87.90470 Los Angeles Chicago
FL 05 4:10:36 41.97464 -87.90470 33.94020 -118.39949 Chicago Los Angeles
FL 06 3:54:12 33.94020 -118.39949 41.97481 -87.89938 Los Angeles Chicago
FL 07 3:55:08 41.97481 -87.89938 33.94089 -118.40738 Chicago Los Angeles
FL 08 4:29:44 33.93814 -118.40773 41.96777 -87.90419 Los Angeles Chicago
FL 09 4:20:48 41.97361 -87.90367 33.94810 -118.39296 Chicago Los Angeles
FL 10 4:33:48 33.94501 -118.40618 38.93984 -77.45310 Los Angeles Washington Dulles
FL 11 1:26:28 38.94001 -77.45310 38.87032 -77.06154 Washington Dulles Pentagone

The first flight is not complete...

Then you can go here ... www.bts.gov
And you try to find the plane and .... Bingo !!! The plane is numbered N5BPAA

Even for the incomplete flight, it's probably this one, from San Francisco:

AA 09/06/2001 1608 N5BPAA ORD 23:20 23:16 0240 0238 -4 23:32 0016


Time is correct for each flight, delays also!!!


Objection of truthers: the number is not correct !!!!!!

Yes, but American Airlines uses alias for its planes...
You can verify it here : http://www.airliners.net/search/photo.search?regsearch=N644AA&distinct_entry=true

For example, this picture http://www.airliners.net/photo/Amer...0194197/&sid=ef57906423646ba2c61507284d57d013
01/22/1999, Los Angeles, no N644AA registered but :
AA 01/22/1999 0144 N5BPAA IAD 12:45 12:46 0293 0293 1 13:10 0024
Again ?
http://www.airliners.net/photo/Amer...0090394/&sid=baf702b166e4519ad4811041e87c78db
Austin,Texas, May 2000, no N644AA for this month but 29th :
AA 05/29/2000 1733 N5BPAA SJC 18:50 19:10 0207 0203 20 19:23 0013
And again...
http://www.airliners.net/photo/Amer...0982095/&sid=baf702b166e4519ad4811041e87c78db
February 2001, Miami... 11 flights for N5BPAA, 0 for N644AA during this month...


So, from where this black box come from ?

 

[wstutt]

I now have version 1.4.1 of my AAL77 FDR Decoder available on my web site along with new output files.

I have added some new parameters.

You can read further notes on the new parameters here.

Now that there are more than 256 columns in the output, I have split the output files so that Microsoft Excel can open the files.

I have also added a note about my citizenship to my web site.

Warren.

P.S. You're welcome moorea34.

 

[moorea34]

OK, N5BPAA was the fleet number: it was already known for a long time ...

http://www.airliners.net/photo/Amer...90718/L/&sid=2ac915b8f461b4760ceb4907e9116537

http://www.airliners.net/photo/Amer...88152/L/&sid=7258d62c19c78f470f8cb7fa6b099802

Sorry...

 

[wstutt]

I now have a AAL77 FDR Decompressor program available on my web site along with it's output file.

The output file consists of the raw 12-bit words in binary with one line per subframe. The end of each section of compressed data is marked by a line with the text Sync Lost.

The output file allows people to try out different data frame layouts themselves without needing to modify or even download any of my programs.

Warren.

 

[beachnut]

Balsamo comes up with new phyiscs based on more woo. 11.2G was pulled out of nowhere, 34Gs were calcualted with hockey stick physics, and now 2223Gs are the new math for p4t morons on the short, very short march to madness.

(truthNAZI, like the soup NAZI but ... Balsamo the never flown left seat heavy jet not a airline pilot) Keep in mind folks, the descent rate from the top of the VDOT Antenna to pole 1 is roughly 6 times less than the new data provided by Warren, if you hypothetically tried to descend from the last altitude point in Warrens data to the top of pole 1, based on the last altitude point being 1.5 seconds west of the wall, base on speed data.

The VDOT "pull up" was 10.14 G's as demonstrated in "9/11: Attack On The Pentagon".

Warrens data would require more..

a = 2 s / t^2

a = 2(100)/0.3^2

a = 200/.09

2222 G's

Add 1 G for earth

2223 G's

Nothing this planet could perform or survive such a maneuver.

2223Gs of stupid.

The final G forces listed in the FDR support impact at the Pentagon as seen when the actual calculations are done. Balsamo is incapable of doing the math and all his experts at p4t join his ignorance.

Balsamo's 11.2G failed physics is still posted at his web site; the p4t dolts think he corrected it, I wonder why it is still posted?

Was his hockey stick 34 or 35 Gs? Now we have a 2223 G special physics. Someone needed to log in at p4t and ask him to show his work, it should be a special moment.

 

[tsig]

Balsamo comes up with new phyiscs based on more woo. 11.2G was pulled out of nowhere, 34Gs were calcualted with hockey stick physics, and now 2223Gs are the new math for p4t morons on the short, very short march to madness.



2223Gs of stupid.

The final G forces listed in the FDR support impact at the Pentagon as seen when the actual calculations are done. Balsamo is incapable of doing the math and all his experts at p4t join his ignorance.

Balsamo's 11.2G failed physics is still posted at his web site; the p4t dolts think he corrected it, I wonder why it is still posted?

Was his hockey stick 34 or 35 Gs? Now we have a 2223 G special physics. Someone needed to log in at p4t and ask him to show his work, it should be a special moment.

Seems to me he has his units all wrong. acceleration is ft per second squared but he is dividing time by time which should give a dimensionless number.

Then he takes that meaningless number and calls it gravity.

 

[George 152]

Balsamo comes up with new phyiscs based on more woo. 11.2G was pulled out of nowhere, 34Gs were calcualted with hockey stick physics, and now 2223Gs are the new math for p4t morons on the short, very short march to madness.



2223Gs of stupid.

The final G forces listed in the FDR support impact at the Pentagon as seen when the actual calculations are done. Balsamo is incapable of doing the math and all his experts at p4t join his ignorance.

Balsamo's 11.2G failed physics is still posted at his web site; the p4t dolts think he corrected it, I wonder why it is still posted?

Was his hockey stick 34 or 35 Gs? Now we have a 2223 G special physics. Someone needed to log in at p4t and ask him to show his work, it should be a special moment.

It seems to be inventanumber calculations done on the back of his prerelease form.

 

[wstutt]

Hi beachnut,

Balsamo comes up with new phyiscs based on more woo. 11.2G was pulled out of nowhere, 34Gs were calcualted with hockey stick physics, and now 2223Gs are the new math for p4t morons on the short, very short march to madness.



2223Gs of stupid.

<snip>

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.

 

[beachnut]

Hi beachnut,

I don't know if you realize, 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.

Balsamo can't fix the math, he has no clue what happen on 911 due to his extreme ignorance and paranoid mind which spews implied delusions.

The fact is Balsamo has nothing correct about 911. He spews the FDR can't be missing any seconds poppycock based on a salesman repeating the requirement that data arrives in the secure chip in no more than 500 milliseconds. The salesman cannot give us the Titanic can't sink crap in the form of the FDR can't be missing seconds, many times FDRs have been missing seconds and more.

The point is the final data you have decode proves flight 77 was 4 feet above the ground, about 10 feet between the ground and the bottom of the fuselage. The actual delay in the FDR could be 50 to 200 ms.

Balsamo fails to understand how data is stored in the FDR when he can't grasp the resolution DME is stored at, or the errors possible in the DME system; so he can't resolve the actual position of flight 77.

I don't need the FDR to know 77 impacted the Pentagon. The only fraud about 911 is Balsamo's DVD and moronic implied delusions based on his intense ignorance.

I find it ironic education found the missing 4 seconds that Balsamo has had in the raw data for years but he failed to see it, he failed to decode it, he failed! You decode the data Balsamo has had for years.

The ironic part of the decode it match very close in vertical acceleration to the study done by Mackey. Science comes though and Balsamo has to correct his delusional physics based on his failed assumptions.

Now we have 58 Gs from the dolt pilot Balsamo who can't figure out the feet in a nautical mile, has no clue what frequency range a RADALT works at and is paranoid nut case who posts as Ryan Mackey due to some insane need to act as stupid as he is.

p4t expert paranoid nut first posts 11.2G, then 35G, then 2223G, and now his final G-force based on his delusional efforts from your decode he has 58Gs. Based on Pressure altitude which is +-75 just sitting on the ground. And as shown the pressure altitude in Flight 77 was reading high, and the errors in a decent are even greater due to lag, and the errors at 483 KIAS are not even defined for the 757 airframe. The RADAR Altimeter is exact to a foot, so we have more delusions from Balsamo who thinks the Flt Crew Door was closed and recorded on the FDR.

58G, final answer? He never took down the 11.2G failed physics! What a clown; he is selling lies on DVD to people too challenged to know better.

Flight 77 impacted the Pentagon killing people; Balsamo fails for 8 years to understand this fact and he makes up moronic junk to sell on DVD.

The good part for us is Balsamo's insanity will ensure he will never be a left seat heavy jet pilot.

Balsamo now has 58Gs of stupid, down from the pull it out of his super-nano-brain 2,223Gs of dirt dumb stupid. I did not think Turbofan was that stupid and it was proved by Balsamo after he admitted he was posting as Turbofan from Tino's basement.

The FDR supported impact before your decode, and it supports impact after your decode. Balsamo his pack of delusion believers are too dense to come up with rational ideas; nothing new.

58 g lol; it was debunked a while back...

September 11, 2001.

 

[wstutt]

Hi beachnut,

<snip>

... Based on Pressure altitude which is +-75 just sitting on the ground. And as shown the pressure altitude in Flight 77 was reading high, and the errors in a decent are even greater due to lag, and the errors at 483 KIAS are not even defined for the 757 airframe. ...

<snip>

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.