After two starship prototypes succeeded in their high-altitude launches only to suffer last-second landing failures for unique reasons, SpaceX is preparing for a third launch later this week.
Crucially, while Elon Musk’s realism (or pessimism) seems to underestimate the real chances of success, the SpaceX CEO is substantially more confident in the third Starship launch attempt than on the first flight two months ago. At the time the serial number 8 starship (SN8) was preparing to attempt the program’s first high-altitude launch, Musk estimated the probability of a launch, free fall and successful landing at approximately 33%.
As it turned out, he was not wrong, but the Starship SN8 ended up doing far closer to total success than almost everyone – within SpaceX or not – expected it to arrive on the first try. Less than two months later, the SN9 starship suffered a similar landing failure in the last second with more than six minutes of flight, although the causes of both failures were unique.
In other words, both flight tests served their nominal purpose, discovering two failure modes that would have eventually shown their heads in one way or another. With SN8, Starship was unable to maintain enough pressure in its secondary landing fuel tank to supply two Raptor engines with enough fuel for a landing burn. The SN9 spacecraft failed a few seconds before the SN8, when one of the two Raptor engines needed for a flipped flip and landing never ignited, causing the rocket to hit the ground at an angle relative to the impact of the SN8’s tail.
As previously discussed at Teslarati, Elon Musk ended up revealing his view that the SN9’s engine failure was potentially preventable and that SpaceX would change the way future starships try to land in an attempt to add more redundancy.
“Although SpaceX obviously did not spin and fix a complex propulsion problem in a matter of days, Musk ended up revealing his opinion that he, his engineers or some combination of both” were too stupid “to explore an obvious way to mitigate the risk of engine failure during [Starship SN9’s] flip and landing. This ‘obvious’ adjustment: rekindles all three starship’s available landing engines, not just two. “
By running not just two – but all three – Raptor engines during the Starship’s flip flare, SpaceX could essentially run a static fire in the air, giving the rocket flight computer a few seconds to analyze performance and select both engines healthier for final landing firing. With this change in place, the spacecraft would theoretically have enough redundancy to land if only two of its three sea-level Raptors were rated.
Currently installed in one of the two ‘suborbital bleachers’ on SpaceX’s launch pad in southern Texas, the Starship SN10 will be the first high-altitude prototype to attempt the burning of three engines and direct selection down. Musk says his confidence that the SN10 will land successfully is now 60%, an almost double improvement over the SN8. The SN10 spacecraft can fly later this week, although the prototype has yet to complete a nominal three-engine static fire test and the launch has yet to receive FAA approval.
Down the road, Musk says SpaceX is working hard to improve the performance of the Raptor’s deep accelerator, potentially allowing future starships to burn two – or even three – engines to land for even more redundancy. However, the deep acceleration of large and complex rocket engines is extraordinarily difficult, so the upgrade is unlikely to occur in less than a few months. In the meantime, Starship SN11 is effectively completed and Starship SN15 to SN18 are being assembled to support a relentless flight test campaign while SpaceX works on orbital flights.