Originally Posted by mms43
Total time for each message was approximately 6 seconds, which comprised synchronization, message header, data and receipt confirmation.
If you play the recorded VHF ACARS message you'll note that it takes a good deal less than 6 seconds to take place. So there must be either a whole lot of very slow processing that takes place somewhere or there is a maximum message rate of 10 per minute to avoid congestion. I'll vote for the latter.
I note the gap at 2:11:33-2:11:42 did not raise signal loss concerns with BEA while the roughly equal size gap at 2:13:20-2:13:45 did and was referred to as 2:13:16-2:13:41. They must have access to aspects of the ACARS satellite signal diagnostics that are not shown. It may be that there are indications something was trying to transmit during their indicated window and could not. Although, at a 6 second periodicity either they are indicating that the transaction at 2:13:14 did not complete by its scheduled time of 2:13:16 and that 2:13:41 is when the message labeled 2:13:45 started transmission OR that there are indications something was trying to transmit during that period but was garbled.
(Note: my times were built in the assumption of a 6 second maximum rate and the time stamp being the time the transmission reception started. That assumption is fragile.)
I hesitate to make assumptions beyond what BEA made about the message transmission gaps. HN39's data is interesting. But, as usual, too much data is missing to make firm assumptions about what time gaps might mean.
Let's loop at the periodicity that shows in the plot using your data in message 2411.
Times: 2:11:00 2:11:27 2:11:55 2:12:16 2:12:51 2:13:14 2:13:51
TBG: 27s 28s 21s 35s 23s 37s
(TBG - Time Between Gaps)
With a plane violently gyrating at that kind of rate what would life be like in the cockpit? Presume no antenna aim aiding that means an attitude perturbation in some direction of about 40 to 60 degrees depending on normal communications margins. (Note, that's a Scientific Wild Assed Guess. I'd have to research antenna gain and convert that to approximate beamwidth. They are generally simply related because you cannot get more power out of an antenna than you put into it. All you can do is redistribute it in the sphere around the antenna. More gain means less of the sphere is covered. I remember a low gain of maybe 6dB, which is where the SWAG came from.)
Would the plane vary its gyrations in that kind of a pattern, medium, medium, short, long, short, long? That variability bothers me as does the BEA only calling out the one gap as being one that indicated possible signal loss in some manner.
Note that if I "presume" ASR33 speeds, 10 characters per second BAUDOT, the longest ACARS message indicated would take about 10 seconds plus plane identifier overhead to transmit. (That's adding in a few figures shift characters to get numbers and the punctuation into the messages, a characteristic of BAUDOT. If ASCII was used that would trim out the overhead. The 10 cps was a hardware limitation of the punch and teleprinter.) That was "about" the state of the art in the early 70s for things that actually printed out. That's probably about where the "6 seconds" figure comes from for "average" ACARS messages, shorter than the 94 characters longest message.
(Phew, I'm getting too long winded. I plead an effort to make things clear to non-experts in my field.)