AA77 FDR Data, Explained

[jaydeehess]

Cap'n Booby is now pretending he never said it may have been a MOAB, and he desperately tried to cover his tracks by deleting this claim when he could.

Unfortunately for him, Google cache knows all! He's "johndoeX" on the old LC forum: http://74.125.155.132/search?q=cache:EXY20R0TP-YJ:z15.invisionfree.com/Loose_Change_Forum/index.php%3Fshowtopic%3D16736+%22Its+very+possible+the+attack+on+the+pentagon+was+some+type+of+bombing+run+with+some+type+of+MOAB.%22&cd=8&hl=en&ct=clnk&gl=us

Well to be fair to Cappy Bobby he was just trying to sound all sexy and such by using the term MOAB.

He actually meant the not quite so massive MOAB
He states in that same post.

This is a MOAB that is obviously declassified. It is impossible to use this type due to the physical damage being alot larger. But this is just to give an idea of what these look like.

So what he meant to say was that it was a bomb dropped by a plane. Some type of bomb. A bomb just big enough to cause the damage seen and he was trying to sound knowlegable by using the term MOAB.
There is a word to describe people like this, "poser", or for those old enough to have enjoyed the sitcom "Cheers" he is pulling off a "Cliff Claven" style of pontification pretty well.

 

[Reheat]

Cap'n Booby is now pretending he never said it may have been a MOAB, and he desperately tried to cover his tracks by deleting this claim when he could.

What's even funnier to me than the MOAB comment in that thread is that someone else mentioned that the aircraft lowered it's landing gear.

In the next comment "Meat Head" replied with a (paraphrased) Yep, Bombing Run.

I actually laughed out loud for some time after reading that comment.

 

[stewieg]

Well to be fair to Cappy Bobby he was just trying to sound all sexy and such by using the term MOAB.

He actually meant the not quite so massive MOAB
He states in that same post.


So what he meant to say was that it was a bomb dropped by a plane. Some type of bomb. A bomb just big enough to cause the damage seen and he was trying to sound knowlegable by using the term MOAB.
There is a word to describe people like this, "poser", or for those old enough to have enjoyed the sitcom "Cheers" he is pulling off a "Cliff Claven" style of pontification pretty well.

I noticed that as well. It does not change the fact that he lied when he tried to claim he never said that. What he said in a way does not make much sense. It really would not have been a MOAB considering the damage done.

Notice over at ATS that as soon as he started feeling the heat one of his pets, Turbofan, shows up to help him out.

 

[wstutt]

I now have version 1.2 of my AAL77 FDR Decoder available on my web site along with new output files.

I have added several new parameters.

You can read further notes on the parameters here.

I looked through the data for the following parameters in case they were stored in a different location than in the data frame layouts but was unable to find them. The parameters are:
GMT DAY (Date)
RADIO HEIGHT F/O
VVI (Vertical Speed)

I noticed that the ENG EPR-ACTUAL - L (RATIO) and ENG EPR-ACTUAL - R (RATIO) parameters have values of 0 for the last two subframes. The MAINTENANCE REQUIRED - L but not the MAINTENANCE REQUIRED - R parameter also has a value of TRUE for the last two subframes. Does the ENG EPR-ACTUAL parameter indicate engine thrust?

It would appear to me that the aircraft was still accelerating and had not lost engine thrust in the last two subframes since the COMPUTED AIRSPEED (KNOTS) and the TRUE AIRSPEED (KT) parameters have increasing values and all but the very last LONGITUDINAL ACCEL (G's) values are positive. Perhaps the ENG EPR sensors failed or can someone give me another explanation?

Hey Warren,

Thanks for your hard work.

Someone is wondering "who" you are. I presume "Hi, I'm Warren" is insufficient for their purposes. They'll want to know which branch of the CIA/NWO etc. you work for. Who is your handler, and how long since your last "Disinfo agent of the month" award.

http://www.abovetopsecret.com/forum/thread512723/pg12#pid7474565

Tom

Thanks Tom for your kind words. I have added a page about my credentials, affiliations etc here.

Warren.

 

[apathoid]

I noticed that the ENG EPR-ACTUAL - L (RATIO) and ENG EPR-ACTUAL - R (RATIO) parameters have values of 0 for the last two subframes. The MAINTENANCE REQUIRED - L but not the MAINTENANCE REQUIRED - R parameter also has a value of TRUE for the last two subframes. Does the ENG EPR-ACTUAL parameter indicate engine thrust?

It would appear to me that the aircraft was still accelerating and had not lost engine thrust in the last two subframes since the COMPUTED AIRSPEED (KNOTS) and the TRUE AIRSPEED (KT) parameters have increasing values and all but the very last LONGITUDINAL ACCEL (G's) values are positive. Perhaps the ENG EPR sensors failed or can someone give me another explanation?

You can think of EPR as thrust, but its really the ratio between inlet air pressure and exhaust pressure. You can dry spin an engine(that is, spin the engine at high rpm using only muscle air pressure) and register EPR as well as most other engine parameters, even though there is no fuel combustion. Even in the event of an engine flameout, the windmilling of the engines will register a small EPR value.

I haven't the foggiest what the "maintenance required" values are all about.

Many, many pages back we were discussing the fact that the last frames of data saw many parameters change suddenly(like some door warnings and other things I cannot recall). Its a possibility that *maybe* the engine ingested some debris that caused a momentary glitch of a main AC power bus. At that moment there would have been three busses supplying power to the aircraft's electrical system; the Left and Right Main AC Busses and the Standy Power bus. It's possible that the left or right bus momentarily dropped offline causing some FDR parameters to change suddenly or go fullscale or to zero. I'm not 100% happy with this theory, but its a possibility.

I now have version 1.2 of my AAL77 FDR Decoder available on my web site along with new output files.

I have added several new parameters.

You can read further notes on the parameters here.

I looked through the data for the following parameters in case they were stored in a different location than in the data frame layouts but was unable to find them. The parameters are:
GMT DAY (Date)
RADIO HEIGHT F/O
VVI (Vertical Speed)

Thanks again Warren, the VVI will be interesting because it should show the lag in the pressure altitude. The 757's VVI(VSI/IVSI) is an instantaneous indicator, its designed to eliminate the lag that we are always talking about(the lag that PfT contends doesn't exist ). If you worked out the descent rate based purely on PA, then compared it to VVI and graphed the two values on top of each other, the lag would be evident. It also comes to mind, that you could do the same thing will the vertical G's.

 

[wstutt]

<snip>
Many, many pages back we were discussing the fact that the last frames of data saw many parameters change suddenly(like some door warnings and other things I cannot recall). Its a possibility that *maybe* the engine ingested some debris that caused a momentary glitch of a main AC power bus. At that moment there would have been three busses supplying power to the aircraft's electrical system; the Left and Right Main AC Busses and the Standy Power bus. It's possible that the left or right bus momentarily dropped offline causing some FDR parameters to change suddenly or go fullscale or to zero. I'm not 100% happy with this theory, but its a possibility.

I believe you are referring to this post. Unlike the NTSB CSV files, out of those five parameters, the only one that changes value in my decode is TCAS FAILURE. I don't know why I got different values from the NTSB.

I notice that the final TCAS FAILURE value that shows FAILURE is stored in the third to last word in the data for that flight. When I looked at previous flights stored in the FDR, up to the last nine words show sudden changes, so IMHO this final TCAS FAILURE value can not be relied on.

<snip>
Thanks again Warren, the VVI will be interesting because it should show the lag in the pressure altitude. The 757's VVI(VSI/IVSI) is an instantaneous indicator, its designed to eliminate the lag that we are always talking about(the lag that PfT contends doesn't exist ). If you worked out the descent rate based purely on PA, then compared it to VVI and graphed the two values on top of each other, the lag would be evident. It also comes to mind, that you could do the same thing will the vertical G's.

Thanks again, however unfortunately I was unable to find VVI within the data.

ETA: VVI could perhaps be calculated from the vertical acceleration and compared to VVI calculated from the PA as you suggest.

Warren.

 

[wstutt]

I now have version 1.3 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 parameters here.

I noticed that the ACMS S/W P/N CODE parameter value changes twice in the data recorded in the FDR. Their are 11 flights recorded in the FDR prior to the final flight. Perhaps the FDR software was upgraded twice within that time?

I have also added a link to download just the American 77.fdr file and also added links to download an ISO image of the complete CDROM I received from my FOIA request to the NTSB.

Warren.

 

[wstutt]

I now have version 1.4 of my AAL77 FDR Decoder available on my web site along with new output files.

I have added some new parameters mainly to do with roll.

You can read further notes on the new parameters here.

Warren.

 

[W.D.Clinger]

I have added some new parameters mainly to do with roll.

Thanks again, Warren.

Here's a question about the raw binary format:

The longitudinal and lateral accelerations are sampled at 4 Hz, but appear (in the CSV file) with the data for the first half of each second. If the raw FDR file is laid out similarly, then those accelerations had to have been buffered for at least one full second before being written to flash memory. (That's consistent with the flash memory technology available when the SSFDR was built.) Does the raw FDR format imply that 1-second buffer time, or was this just an artifact of how the CSV files were produced?

Will

 

[wstutt]

Thanks Will,

Thanks again, Warren.

Here's a question about the raw binary format:

The longitudinal and lateral accelerations are sampled at 4 Hz, but appear (in the CSV file) with the data for the first half of each second. If the raw FDR file is laid out similarly, then those accelerations had to have been buffered for at least one full second before being written to flash memory. (That's consistent with the flash memory technology available when the SSFDR was built.) Does the raw FDR format imply that 1-second buffer time, or was this just an artifact of how the CSV files were produced?

Will

You have raised an important point. The longitudinal accelerations are recorded in words 33, 97, 161 and 225 of each 256 word subframe. The lateral accelerations are recorded in words 35, 99, 163 and 227 of each 256 word subframe. One subframe is recorded per second and since these parameters are recorded at intervals of 64 words in the 256 word subframe, they are recorded at equal intervals of 0.25 second. Other parameters are likewise recorded at equal intervals.

The way they appear in the CSV files is an artifact as you suggested.

Warren.

 

[W.D.Clinger]

You have raised an important point. The longitudinal accelerations are recorded in words 33, 97, 161 and 225 of each 256 word subframe. The lateral accelerations are recorded in words 35, 99, 163 and 227 of each 256 word subframe. One subframe is recorded per second and since these parameters are recorded at intervals of 64 words in the 256 word subframe, they are recorded at equal intervals of 0.25 second. Other parameters are likewise recorded at equal intervals.

The way they appear in the CSV files is an artifact as you suggested.

Warren.

Thank you!

Will

 

[tfk]

Warren,

Once again, first rate work. Congrats.

Regarding vertical speed.

Is this not the "Vert Speed Selected" (uid: Vert_spe) data?

The only reason that I ask is the units ("Ft/min"), which would be a speed. (No idea why they'd add the term "selected".)

Tom

 

[apathoid]

Warren,

Once again, first rate work. Congrats.

Regarding vertical speed.

Is this not the "Vert Speed Selected" (uid: Vert_spe) data?

The only reason that I ask is the units ("Ft/min"), which would be a speed. (No idea why they'd add the term "selected".)

Tom

That would be an autopilot function. I've looked for the VSI and cannot find it either.

 

[W.D.Clinger]

I have used Warren's new data to compute a crude reconstruction of the flight trajectory for the final seconds:
http://www.ccs.neu.edu/home/will/Music/Jokes/Balsamo/Software/

I'll be happy to correct any mistakes reported to me.

Will

 

[R.Mackey]

I have used Warren's new data to compute a crude reconstruction of the flight trajectory for the final seconds:
http://www.ccs.neu.edu/home/will/Music/Jokes/Balsamo/Software/

Wholly unnecessary, but good analysis.

For comparison, here's the picture from my old parabolic analysis, back when we thought the last few seconds of FDR data were lost forever:



Graphs are to scale. Each square on the rulers represents 100 feet. The parabola in the lower curve was the limiting case of the various initial conditions I studied here, and obviously much tighter than we really need. Your new findings, with new data, are nicely consistent with this prediction.

Another victory for science.

 

[beachnut]

... studied here, and obviously much tighter than we really need. Your new findings, with new data, are nicely consistent with this prediction.

Another victory for science.

The best part is Balsamo/Turbofan (super-nano-sock-puppet-joined-at-the-hip-until-divorce-yesterday) came up with super stupid 11.2G and 34G delusions while Hani's poor flying skills matched science.

Terrorist do science in practice, only dolts do Balsamo physics.

Just a review, Hani's last 4 seconds of G force roughly match the textbook solution.

Balsamo - 0, Science 1,237,234; science better than Oregon vs. Oregon State.

0.725, 0.659, 0.92, 0.858, 0.94, 1.121, 0.828, 0.783
0.982, 0.986, 0.927, 0.776, 1.25, 1.037, 1.231, 1.721
1.604, 1.781, 1.762, 1.964, 1.879, 2.264, 2.044, 2.181
1.675, 1.744, 1.65, 1.504, 1.785, 1.655, 1.861, 1.946

Last 4 seconds of G sampled each 1/8 second. Balsamo thinks he is using science and fails, Hani almost smacked the overpass (30 to 40 feet) but practiced better science than the hockey stick pull-up tripe Balsamo feeds his cult members.
Math and physics smash ignorance.
http://www.ccs.neu.edu/home/will/Music/Jokes/Balsamo/balsamo2.html#finalseconds

 

[tfk]

A Flight Path Model Calculated from the data

Here's my take on the flight path: I'm believing the raw data.

You can calculate the vertical velocity (second by second) by taking the difference in PA. Since it's sampled once per second, then the simple difference between each successive data point (in feet) is the vertical velocity in ft/sec. I've assigned that value to an arbitrary time 1/2 way between each time increment.

You know the plane's calibrated air speed.

Then ø = sin^-1(Vv / Vc)
where Vv = vertical velocity (ft/sec)
Vc = calibrated airspeed (ft/sec)
ø = the Glide slope over each interval.

From here, you can either start stringing together displacement vectors, or calculate the horizontal distance between points using:

∂x = Vc * ∂t * cos(ø)
where Vc = computed air speed (ft/sec)
∂t = 1 second intervals
ø = glide slope over interval (Deg)
∂x = horizontal distance traveled over each interval (ft)


Now it is easy to plot horizontal distance traveled [= ∑ ∂x] (ft) vs Pressure Altitude (ft). Simply set the aspect ratio of the plot to 1:1, and you've got the true flight path of the plane.

Here's what the curve looks like for the last 68 seconds.

[imgw=700]http://www.internationalskeptics.com/forums/picture.php?albumid=176&pictureid=1991[/imgw]

You can see that it is a succession of 5 virtually straight line segments. Meaning that the g forces aren't going to stray much from a nominal around 1g, with small excursions above & below. Especially at the transition points between segments. This is borne out by the vertical acceleration data in the FDR output.


The following graph shows a blown up portion of the same flight path, focusing on the last 14 seconds before the End of Data. This encompasses two straight line segments, with very high linear regression coefficients.


[imgw=700]http://www.internationalskeptics.com/forums/picture.php?albumid=176&pictureid=1992[/imgw]

Note that the vertical offset is still arbitrary, and the distance to actual collision is unknown from this data. We do know that Hanjour was pulling positive Gs over the last 2 seconds or so. Which would shallow out the 4° terminal glide slope segment shown in these graphs.

My conclusions are that the data shows that neither a circle nor a parabola are the best curves to model his flight path. But instead, it is best modeled by a succession of 5 virtually linear flight path segments over the course of the last minute of flight.

Tom

 

[tfk]

Errata: Both axes on the second graph above should read "(thousand ft)". Not "(ft)".

Here is the corrected graph.

[imgw=700]http://www.internationalskeptics.com/forums/picture.php?albumid=176&pictureid=1994[/imgw].

Tom

 

[R.Mackey]

My conclusions are that the data shows that neither a circle nor a parabola are the best curves to model his flight path. But instead, it is best modeled by a succession of 5 virtually linear flight path segments over the course of the last minute of flight.

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.

 

[wstutt]

Hi Will,

I have used Warren's new data to compute a crude reconstruction of the flight trajectory for the final seconds:
http://www.ccs.neu.edu/home/will/Music/Jokes/Balsamo/Software/

I'll be happy to correct any mistakes reported to me.

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

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?

Warren.

ETA: I presumed here that the end of the data as at the pentagon wall. Perhaps this presumption is incorrect