First, thanks for all the typo gets, and pedalogical weaknesses. I'll fix them.
Also, poring over the .csv file, it looks to me like each item is recorded just once in each one-second subframe. In a previous thread you referred to the FDR having a minor frame rate of 8 Hz and a major frame rate of 1 Hz, but here you call them out as 1 Hz and 1/4 Hz respectively, which seems to agree with the NTSB report's reference to a four-second frame containing four one-second subframes.
Well, my original guess was 1 second major frames, and 1/8 second minor frames. That was due to the fact that 1 second was the lowest period of sampling, and 1/8 was the highest. After looking at the file more carefully, some values are, in fact, sampled as low as 1/4 HZ (or once every 4 seconds), so 4 second major frame seems plausible.
Minor frames, strictly speaking, are just the distance of synch words in your data. Effectively, it is the "quantum" or error-checkable data. If a single bit goes, the most data you will lose is a minor frame. I guessed they would do it at 8hz due to 8hz sampling, but I was wrong. They only do it at 1hz.
I've sat and counted fields in the .csv file until my brain hurt, but bless me if I can see an easily identifiable quantity, like vertical, longitudinal and lateral accelerations, which are recorded at the top of each subframe, being recorded more often than once per second. That has me kind of confused; even if I were a CTer I can't see how I would get to claiming 1/8 second precision.
V.Acc is sampled 8 times per second according to the CSV file I have. It's in the first column. Maybe we have different ones?
Weedwacker--
I guess the NTSB went through all the trouble to make an animation in error along with the csv?
Short answer:
To show laypeople what the flight looked like within 99.9% accuracy? Why do you demand 99.9999% accuracy?
Long answer:
What's the data source? You seem to think it was the CSV file. If that' sso, it is already flawed in terms of to-the-millisecond-and-foot accuracy. Maybe it was done that way for expediency? Maybe 99.9% is good enough? What's the precision of this animation? What are the error thresholds? What was the metholdolgy?
The animation is a reconstruction done by an engineer. He made an animation that was reasonably accurate to show the big picture, not be precise down to the millisecond or the foot. If he wanted to make a foot-and-millisecond precise animation, he'd need to correct for ALL the errors I've explained, and that would be virtually impossible for someone to do without a massive scientific effort justifying all sorts of assumptions, and even then, it would be assailable on a number of scientific fronts.
As far as I'm concerned, the animation is of little value beyond the actual data given, because it's non-repeatable. I don't know any of the parameters it was constructed with. It has built-in assumptions that aren't stated. I've already pointed out a number of errors in the underlying data for to-the-millisecond-and-foot precision, so I see no reason to expect that the animation, based on that data, would be any more precise.
1. What is the True Altitude at end of data recording :44. How did you come to your conclusion.
What does "end of data recording :44" mean? What time, specifically? The end of the 44th second? What was the pressure altitude at :45.00? Is that the question? Let me ask a different question? Why is that important? Are you assuming that's the time of the crash? That would be a mistake.
I'll answer, anyway...
The presssure altitiude recorded during the :44 frame is 173. In the worst case, this was measured at around :43.50 or :43.00 (depending on the sampling rate from the ADC to DAU). In the best case, this was measured at :44.99.
That gives us a measured pressure altitude somehwere between 173 and 173 - 2seconds * descent rate (in foot per second). JDX likes to quote '66 feet per second' as the descent rate. He gets this by subtracted the pressure altitude from the :43 and :44 frames. Using the same logic, in the worst case, these were actually measured at :42.00 and :43.00. We cannot use this as the average descent rate between :44.00 and :45.00 for obvious reasons, unless we assume no acceleration (a false assumption). The moment you try to introduce V.Acc to compensate, you need to now account for simultaenity errors to over come this.
This answer is intractably difficult to answer to with any certainty given the time-error introduced by basing our calculations on the CSV file.
The only definitive answer to this questions is that, given only the CSV data, it's intractibly difficult to answer this question definitively.
Once you've made all your assumptions about simultaenity and acceleration, and come up with a calculation, you can then begin to think about correcting your answer for instrument error in the altimeter, due to lagging because pressure differentials and all that.
2. What is the vertical speed at end of data recording :44. How did you come to your conclusion.
3. What is the Absolute Altitude and end of data recording? How did you come to your conclusion.
See #1. I need the answer to #2 to be able to figure out #1, anyway. I have too many variables, and not enough data, to answer these questions definitely.
4. Why does the csv file show the altimeter being set in the baro cor column on the descent through FL180, but the animation altimeter does not show it being set?
"Appeal to animation" fallacy.
5. Why do the current G Forces for the last minute of data correspond to the changes in vertical speed...
First of all, they don't. I've already provided the graph to show that the two pieces of data ARE NOT simultaenous as shown in the file.
http://img134.imageshack.us/my.php?image=jdxdelayfn9.png
Therefore, the premise of your question is already false, but I shall proceed anyway...
yet at end of data :44-:45 it shows an increase in vertical speed never accounting for any type of level off to be level with the lawn as shown in the DoD video?
Your assumption that the time frame :44-:45 represents the final second of flight 77 is false. :44-:45 represents the last completed frame in the data. The G forces on the plane during the trip across the lawn are
most certainly not the ones shown in the CSV final's data frame.
6. Do you have any video showing a clear impact and/or of the plane on its approach to impact?
Nope.
7. Why does your animation show a flight path north of the reported flight path?
My animation? I assume you mean NTSB's animation. I can't answer that question, you should ask them.
8. Why are there no system indication of any impact with any object up to and after :44?
Since the :44 frame is the last completed frame, the CSV data "runs out" before the plane actually hits the pentagon. We don't know how many more frames (and partial frames) were recorded, but at least 1 partial frame was. Therefore, the data you are looking for didn't happen in the time range of the data given.
9. Why does the csv file and animation show a right bank when the official report requires a left bank to be consistent with physical damage to the generator?
You are assuming an impact time and simulatenity issues that you have no justification to be assuming. The data you are looking for isn't in the CSV file, to the best of estimatation. The CSV file ends before this data would have been there.
10. How did you come to the conclusion of 09:37:45 as the official impact time?
09:37:45 to what precision... 09:37:45 in reference to whom? The onflight data recorder? Was it GPS synched? When was the last GPS synch? If so, which IRIG GPS time was it synching to? IRIG-B? H? G? (look em up) If it wasn't GPS synced, then in reference to who? The FDR? Greenich Mean Time? The Naval Observatory?
In fact, I'd say it's highly unlike that 09:37:45.00 is even close to the impact time, according to the FDRs clock. More than likely the actual impact time was at some point in the next two seconds.
The NTSB set the time of impact to 09:37:45.00 because within a second or two, it's correct. I doubt very seriously they spent time trying to properly synch the FDR ex post facto with the Naval Observatory, in order to give an exact time so that amatuer forsensic analysis of the FDR could be successful.
11. What is the exact chain of custody of the FDR? What date/time was it found? Where exactly was it found? Please provide documentation and names.
Answered above.
12. Why does the hijack timeline show a 3 min interval for hijacking to take place? Why did Capt Burlingame not follow protocol for the Common Strategy prior to 9/11?
Outside of my expertise and knowledge. I can't answer that.
