sitemap Donaldson's Review of the Flight Data Recorders.

The Smoking Gun!
The Flight Data Record

"So in other words there is no way on God's green earth, and above it apparently, that an explosion
from the center fuel tank could have caused this pressure reading on the external pressure sensor"

Art Bell talking with Cdr. William S. Donaldson
December 23, 1997


From the Art Bell Show, Dec 23 1997. Transcribed from Real Audio starting about 3:32 into the show. Bell is interviewing Bill Donaldson about information from TWA 800's flight data recorder.

Editorial Note from M. Hull

Click for pdf file showing tabulation of the FDR paramaters.  (Requires Adobe Acrobat Reader - Zoom to view the section at time 20:31:12 otherwise table will just show lines.) For those unable to read pdf files I have transcribed The Last Two Seconds of the FDR pdf File to aid in its examination. The table shows no data after time 20:31:13. TWA Flight 800 data is color-coded red, the data from a previous flight (TWA 803) is color-coded blue. (See Captain Mann's analysis below).

Click for pdf file showing NTSB Plot No. 8 dated February 4, 1997 of the FDR paramaters. (Requires Adobe Acrobat Reader - Zoom to view the section at time 20:31:12).  Note how the NTSB has plotted data to time 20:31:20 - i.e. eight seconds after the statement "End of TWA 800 Data" as indicated on the document by the line at time 20:31:12.

Click for pdf file  presenting an analysis of the Flight Data Recorder "12 second line" by Captain Howard Mann, a retired TWA pilot and crash investigator. TWA Flight 800 data is color-coded red, the data from a previous flight (TWA 803) is color-coded blue.

D: Now lets go on to what you say is the smoking gun and its a big one. When I got to the hearings I went as a credentialed reporter for Accuracy in Media and to classify the thing as a public hearing is really a misnomer. There were very few of the actual people off the street that were allowed into the place. Anyway, I get this mountain of paper. One of the most critical things on an air-crash investigation, particularly in civil aviation where they have these flight data recorders, you have a voice data recorder and a flight data recorder. They call them black boxes .. they're really not black. .. they're orange.

B: They're separate.

D: Right. They're separate pieces of gear.

B: One records the final voice of the cockpit crew.

D: Right.

B: The other is recording the aircraft's technical performance up until the very last second.

D: Right. Now the only reason you put a flight data recorder in an aircraft is to capture - I mean it sits there and rides - most airplanes never crash - they're sitting there in the event that it does happen. The target - the whole reason that you put that piece of gear - the expensive piece of gear in this airplane is to be able to recover it and play it back and see exactly what everything in that airplane was doing.

B: Of course.

D: Now. It's almost laughable. When I got to the data on the flight data recorder, I didn't even catch it at first. It was one of these retired captains from TWA who was working with another crash investigator. We had been looking at this together. A guy by the name of Howard Mann picked it up. The reason I didn't notice it was the last line of data was literally lined out by the NTSB.

B: Now wait a minute. We're talking about the flight data recorder. The last line was lined out?

D: Now what they did - you can see aircraft's performance in all kinds of things. It shows the altitude readout. It show the attitude of the airplane. It shows about 15 or 20 items and as the aircraft goes along every second there's a data block with a recording in all these different categories.

B: Gotcha.

D: And in between seconds there are even partial data blocks on some of the stuff. So when they get to the very end of the tape - the stuff that's the whole reason the gear is in the airplane - they literally drew a line through the last data block and made a note on the side: end of flight 800 data. When you look at it, it's hard to read through it at first - you assume as I did that this was some kind of previous recording or something and I sort of skipped over it.

B: So in other words you thought that this lined out last line was not part of the flight 800 data at all but was part of some previously recorded flight's data.

D: Yeah exactly and as it turns out this type of flight data recorder has about a 30 hour playback and it was only on the first 30 minutes of, you know, flight so there was no data that was still on the tape.

B: So they do not erase previous flights? They record over them much as we might record a tape over another tape.

D: Right.

B: So here's this lined out last line and kinda hard to decipher, but I take it you managed to decipher the line.

D: Right. And once you realize that it is a valid data line and it is the last second recorded of the flight, it's startling because for instance here's what it tells you. The aircraft was climbing out and I did a little look at the previous 10 or 15 seconds. It was climbing about 22 & 1/2 feet per second -the airplane was. Each one of these data iterations going across there. And all of a sudden the airplane went from 13,799 feet is what it should have been at that last second.

B: What was the altitude again? 13,000 ...

D: 799 is what it should have registered. Now what happened is that it suddenly dropped. The altimeter dropped 3,672 feet and registering down on the 10,000 foot level.  I went: whew - wow - what happened? Obviously the airplane didn't suddenly drop almost 4,000 feet. You look at the next column and the airspeed indication goes from 298 knots to 100 knots. In other words it lost 198 knots instantly in one second. That can't happen either. I mean the airplane didn't hit a brick wall up there. And you go across and there's a whole series - I won't - I just concentrate on these two because of time for a minute.

B: All right.

D: Now you have to, if you're working for the NTSB, it's your duty to explain this data line. These are real instruments. These are not somebody's imagination. This is hardware in airplanes that are recording actual data. The way the altimeter works is that you have a what's called a static air port on the side of the aircraft and the altimeter is just like a barometer. It senses the pressure outside the airplane, and it converts it to a reading of altitude.

B: All right - important point - so this sensor is looking out the side of the aircraft.

D: Exactly! Now all of a sudden you get this tremendous drop in the reading of altitude - what this means is that this instrument recorded a pressure that normally would be at 10,170 feet or whatever the reading was.

B: So in other words you're telling me this last data line recorded a change in pressure from 13,799 down into the 10,000 range.

D: Right. Now, what that means is that you have to go to an atmospheric table to see what the actual conversion is in real pressure - pounds per square inch - and the figure is in the neighborhood of 1.32 pounds per square inch. That doesn't mean a lot to some people but it does to me and to other people that understand about explosives because in order to get a sudden increase in pressure like that - It sounds small but remember that on the side of an airplane you've got 1 square foot is 144 square inches so it adds up when you start looking at area -

B: All right Bill - important point - how do you know that this pressure difference which was sensed and recorded that you have found came from an external source?

D: Okay. The reason being what I did - There's a series of equations that you use when you compute what an explosion will do as far as delivering pressure to a distant point and they're not all that complicated but the bottom line is what I said is okay I'm sure folks are going to say well the airplane blew up so the pressure came from the center wing tank.

B: Right.

D: So I said all right, I'm going to do the calculation except instead of using the 60 pounds per square inch in the tank like the NTSB scientists were saying was the maximum, I'm going to arbitrarily assume that it's 600 pounds per square inch in the tank.

B: Ten times more.

D: Right ten times the pressure and I'm going to compute what the overpressure would be on that static port 70 feet up the side of the airplane near the nose. The answer is .43 PSI - pounds per square inch. That's using a figure ten times what their explosive people said was the maximum capability of that tank.

B: So in other words there is no way on God's green earth, and above it apparently, that an explosion from the center fuel tank could have caused this pressure reading on the external pressure sensor.

D: Right. And see -

B: Translated. The center tank didn't do this.

D: No. In fact the point being - it's another point of common sense. The only power you get on the aircraft is through the generators that are out on the wings - on the engine. Now that wiring routes right by the center wing tank on the way up to the forward part of the aircraft.

B: Uh huh. So in other words, if the center tank had blown the data could not have been there because it would have blown that wiring to bits.

D: Sure. Exactly correct. So okay let's assume that we have the toughest wire in the world and its going to last at least a second through the middle of this explosion. It's both the airspeed indication that dropped to 100 knots down to 100 knots from almost 300 and the altitude both work off the pitot static system they call it in the aircraft. Now I'm going to switch to another sensor.

B: Now you're saying that neither one of these changes - the change in altitude and the change in apparent airspeed are real.

D: No.

B: These are simply anomalies produced by the external pressure.

D: Right. Every time there's an explosion in the atmosphere there's an overpressure that goes out at the speed of sound and when it hits something it delivers an overpressure.

B: All right. you want to talk about another sensor.

D: Okay. The other sensor is what's called the angle of attack system. Angle of attack is a system that measures the exact angle that the wind is striking the nose of the aircraft. Normally in flight and you see in the data - the airplane climbing at that speed and so on - the angle of attack was three degrees. Almost directly on the nose. Just 3 degrees below the nose was the wind striking the aircraft. All of a sudden when this overpressure wave hit, it goes from 3 degrees to 106 degrees.

B: My God.

D: Okay. So this vane on the outside of the aircraft - it looks like a little wind vane.

B: No. I understand.

D: Okay. Now it suddenly gets blown up past the 90 degree position and then you see that's one of the data blocks. It has two more hits after the main last hit and it shows that the vane goes to 106 on that last full data block.

B: So in other words the direction of the prevailing winds suddenly, utterly, completely, instantly changed.

D: Yeah changed to going almost perpendicular to the flight path of the aircraft.

B: And again the information you have just presented has come from the last line which was lined out for some unknown reason.

D: Well, they were handing it out to the media.

B: Flight data recorder. Was it lined out in such a way as they intended for you not to be able to read it? Was it lined out in such a way -

D: It's legible if you looked at it closely but remember they're handing this out to reporters and reporters would go down and say how come this last line is so screwed up? I don't think they wanted reporters to ask the question. Okay, I'm led to believe that that last line is not lined out on the Internet one. They posted this stuff on the Internet. It's in plain view. But let me finish on this angle of attack thing. It's important. The reason it's important is there's two more data hits on the angle of attack system before it shuts off and they happen a quarter second after the main data line and then a half second after it and what it shows is that this angle of attack thing goes up to 106 degrees; then a quarter second later its back down to 30 degrees and a quarter second after that it's back down to three degrees which is essentially the normal position. Now what that means is that this data is real.

B: It records the entire thing.

D: Right. And at least a half a second after the last full data block and the vane did exactly what it would have done if it had encountered an outside air explosion.

B: The only thing that can explain - the only thing - that can explain this data is an external detonation.

D: Right and you can even see it in the engine - there's two or three other main systems that all say the same thing. I mean things happened to that airplane on that last data line that can only happen they encounter a near experience of high explosiveness.

B: Bill, how can you know for certain or even nearly certain that that last line belongs to flight 800?

D: Because it's sequentially. The time is there. In other words, everything is sequentially there.

B: My God. You mean they have time hacks.

D: Yeah. In other words the previous full data line was at 8:31 and 11seconds and then the one we're talking about is at 8:31 and 12 seconds.

B: Then you've got them.

D: Oh yeah, that's what we're on the air for here. I mean to me and to an expert that's used to reading this stuff you better explain that last line. I mean before you put this puppy to bed.

B: God. You've got them.

D: I think.

B: So what do you want to happen now Bill?  Should you be on Nightline?  Should you be on the NBC ABC evening news?

D: Well, that requires an invite, but I think what Accuracy in Media is going to do is we'll probably put a press conference together sometime in the near future and what we really need is more support for the Aviation Subcommittee that's investigating this thing.

B: How can we help?

D: The simple way would be to drop a card - just a postcard - to Chairman Jimmy Duncan. He's a congressman from Tennessee but way to say it is Chairman of the Aviation Subcommittee House of Representatives, Washington DC, and he'll get it. We want the truth and the whole truth about flight 800.


Editorial Note from M. Hull

Click for transcript of final part of Cockpit Voice Recorder which indicates end of the recording at 20:31:12.   In my opinion there is a very interesting feature in the CVR at time 20:31:03 which refers to an "unintelligible" word which occurred about nine seconds before a missile detonated outside of the aircraft. Shortly after the CVR was recovered in July 1996 an "official" reported that the last words on the CVR were "uh, oh" which would lead me to wonder if one of the crew members observed a missile just prior to detonation? At this time in the Flight Data Record all systems within the aircraft were normal and if the "unintelligible" word is indeed "uh, oh",  then the observation causing the pilot to utter this expression is highly significant. (I  regret to say that I am unable at present to locate the supporting reference to the "official" who made the statement and accordingly the reader must regard this matter at present as purely speculative.)  

The sequence of events indicating detonation of a missile outside the aircraft at time 20:31:12 followed several seconds later with the aircraft's disintegration is consistent with  the eyewitness accounts.


Cdr. Donaldson's Press Conference Analysis of the Flight Data Record

Go to the Donaldson File Index