Thank you thf, for bringing this up, and airman1900 for linking the NTSB reports. It's a very intersting reading about smoke detection system and related procedure.
Concerning MS 804, beside using optical detectors (see Airbus statement), I would add that the smoke warning(s) were triggered several hours after take off. At this point, such warnings would be very unlikely resulting from prior ground contamination of smoke detectors, which is what happen in most cases.
Also, ACARS of MS 804 are reporting two warnings at 0:26 and 0:27 (Lavatory and Avionics), which make it even less likely that they were related to nearly simultaneous faults of two detectors reporting spurious alarms. They are related to different compartiments and systems.
Moreover, both warnings were possibly triggered by cockpit adjacent areas (Lavatory A door is only a few centimenters away from the cockpit door while the avionics bay is right below the cockpit's floor) ; on top of it, the avionics vent system is using cockpit air to refresh both instrument panel and the whole bay, where the avionics smoke detector check the exhaust ducts for particles.
Finally, the two system faults reported at 0:26 and 0:28 are telling something related to the cockpit right windows heating sensors (sliding first, then fixed one) ; both window sensors faulted... but not the right windshield which is also heated by the same WHC-2. Hence, one could think that it wasn't related to the state of WHC-2 in the bay which was still working on the right windshield and triggered no fault on its own. In case of such a single fault reported, one would suspect the sensor first, the computer next and the wiring last. Now, if the computer is not at fault, we are still left with two faults with different wirings, sensors, and circuit breakers...
So far, and without any further evidence, I would rather suspect that those smoke warnings reported were actually real ones ; also, it seems to be more likely due to an external system cause, probably in cockpit, like a sudden surge of temp close to the right sliding windows ; the original event would have to last for at least 1-3 minutes (0:26 to 0:28) in order to trigger the second sensor fault on the fixed right window.
Considering that the investigation is in possession of the full ACARS content, they would have more details on each time stamping; they would also be able to tell if any probable delay happened during the string of ACARS between 0:26 and 0:29. But based on what we have, it's still possible to make some sense out of the first faults reported.
Of course, it's not the only scenario as one could also suspect a more complicated sequence, like a wiring combustion or electrical issue contaminating one and second right window heating system. It's only less likely because it would alert the crew of something wrong (like smell) in the cockpit, and probably well before this point.
Now, looking at any previous report of incident at cruise involving cockpit windows overheating in A320 might be more interesting than spurious smoke warning events, in order to understand if this could make more sense than an external surge of temp in cockpit.
Last edited by takata; 28th May 2016 at 23:09.