Metar OMDB 030900Z 11021KT 3000 BLDU NSC 49/07 Q0993 WS ALL RWY TEMPO 35015KT 1500
[Translated: wind from 110 degrees at 21 knots (observed a few feet above the ground), visibility 3000 meters with blowing dust, with no significant changes in the last reporting period, temperature of 49 degrees Celsius, dew point at 9 degrees Celsius, lower than standard atmospheric pressure of 993, with wind shear observed and or reported on all runways, the a temporary wind out of 350 degrees at 15 knots with 1500 meters visibility.]
Some news reports say the gear problem occurred prior to landing. But if that is so, why didn’t the crew stay airborne, attempt to get the gear down green and then land?? If the right main was problematic, such as hung up or unsafe down, as reported, why wasn’t an airborne emergency declared by the crew, with the crew requesting tower to have fire crews standing by for an unsafe gear landing?? One report indicates that the gear had been raised because the crew was performing a Go-Around, but that doesn’t account for procedures that state in essence: Add Go-Around Thrust, set flaps to approach, Establish a positive rate of climb, then raise the gear. [Go Around Thrust, Flaps 20, Positive Rate, Gear UP are the memory items for a Boeing Go Around.]
So if they were not executing a Go Around Procedure, which I suspect that they were not, if the crew raised the gear, in lieu of a gear collapse occurring, why had they not added the power to actually go around?? This could account for the landing without three gear down and locked. However, the left main and nose gear appear to be down and locked, so I would doubt that the gear was raised manually for a go around, because if they had raised the handle to raise the gear, all of the gear would have been unlocked and thus collapsed on touch down, which they were not.
If, as I suspect, the gear was down and locked with three green, I would then look to the Metar and flight aware flight log ( see above link) for clues.
Possibly the thrust levers were in auto-throttle mode and had moved back towards idle due to over performance head wind shear. The established rate of descent last recorded on the Flight Aware Log was at -800 fpm, but the airspeed had accelerated towards 180 knots.
So, without the crew clicking the auto pilot and auto thrust levers off, in order to hand fly the flight controls and the thrust levers, or at least, hand fly the power by the adjusting the thrust levers, and leaving the auto flight controls coupled to the flight director and ILS, the final descent onto the runway could have been close to -800 fpm. It could have even been higher if a lower power induced higher sink rate had developed, as I suspect it had. With a previously steady -800 fpm descent rate, high IAS above bug, the power would have been back, setting up an eventual higher sink rate in a few seconds. To keep the aircraft on glide slope, the auto pilot would start raising the nose attitude, until at or above the L/D max point on the lift curve, the drag increased so much that the aircraft speed slowed. This would lead to a higher sink rate.
Possibly the sink rate at the time was in excess of the radar altimeter cued auto pilot flare capability, with the possibility of a previous increased IAS and thus higher descent rate, as the IAS bled off as the power remained low. This could have resulted in a pretty hard landing, along with what looked like a slight left crab, adding stress to the gear, all of which could have collapsed the right main gear.
Additionally, upon the high descent rate touch down, the whip action of the wing and engine pod, could have led to an attachment point-pin failure, due to whip action aka “over-momenting” in engineering terms, resulting in the engine pod separation, with a subsequent fuel leak and fire.