CONF iture;
Question : Is the FDR settable at the airline request ?
Yes, it's possible but complicated.
The JARs, FARs and CARs, (Europe, USA, Canada), require mandatory data frames which record parameters at mandatory rates, (frames-per-second) on the FDR.
The regulatory standards are uniformly, pitifully, low when it comes to recording aircraft data. The schedules for mandatory parameters in the regulations are complicated, having to do with manufacturing and original aircraft certification dates. I'm working right now with aircraft which legally require about 30 parameters and that's exactly what they have and nothing more; later certification dates require up to 88 parameters.
The A332 which was operating flight AF 447 had about 1300 parameters. My bet is most of them are engine parameters, then system parameters, then flight parameters, (instrument, ILS, 'g', aircraft performance) and the fewest, if any, will be ECAM text, FMA text, GPWS and TCAS text parameters.
One of the motivations to have more than the minimum number of parameters beside flight safety is maintenance troubleshooting. Most aircraft systems today are self-reporting but often ACARS-type messages are not sufficient for complex troubleshooting. Engine monitors record enormous numbers of parameters separately from the FDR as do some other aircraft systems. We know this was the case with AF447.
Regarding the recording of parameters, (which really means designing a large data frame into which engineering data is placed, which is later changed into readable parameters by the DFDAU), that is a very complex process and piece of software which costs an enormous amount of money to construct. The work is almost always proprietary so there isn't any sharing of information done in the industry.
The thinking behind leaving out the RHS airspeed would likely be that there were "two" such parameters and so there was no need for "three". Much of this work is the victim of "statistical thinking"...that "averages" can actually tell a rich picture, when in fact they cannot; a clear example is the one we frustratingly have before us.
For many reasons, some historical, the kind of investigative work now being done using the FDR data is not, quite frankly, the primary consideration in building data frames which go beyond the legally required minimum parameters. It is cost-savings, usually through maintenance, not investigative capability, that is the priority, but that is changing.
To illustrate the potential for the number of parameters possible if the data frames were actually programmed and installed, the C-17 has about 40,000 parameters. Military FOQA however, is mission-oriented. Reliable dispatching is the priority though flight safety is part of MFOQA.
The B787 will be a remarkable aircraft in this regard. It will record thousands of "standard" parameters and leave many thousands off the record. Should trouble develop with a system however, the recording system will shift resources and storage to that system and begin recording those parameters which are related to the system providing a far more detailed recording for investigative work.
In a word, the answer to your question about airlines being able to "program" FDRs, is "yes". Anything can be recorded for a price. Designing the data frame and buying the license to use it from the manufacturer, whether a major one like Airbus or Boeing, or corporations like SAGEM and Teledyne who retain standard data frames for most aircraft types or others who have been hired to do the work themselves, is the way such capability is achieved. For changes and new installations for aircraft already in service there is always an STC involved and that can be extremely expensive and a very time consuming process to accomplish. Most airlines are so thinly resourced and working on such thin margins that getting such work done is very difficult.
The reasons that the #2 airspeed parameter and many others, (I suspect no ECAM messages are recorded), are not present requires a more lengthy explanation but essentially the reason involves cost.
A good link for some basic info in "CAP 731" is at
http://www.ihst.org/portals/54/Attac...20L_CAP731.pdf . See especially, Appendix B.
None of this discusses the QAR, primarily used for flight safety analysis programs. These recordings are usually far richer, with many more parameters. Such recorders are not crash-protected, (the QAR unit on AF 447 was mounted in the 800VU rack brought aboard the Ile de Sein; hopefully the card survived and they can read it).
I'd be happy to exchange PMs for further, as it's a bit OT.