Starship goes up. Starship goes down. But is the program moving forward?

So what, exactly, are we to make of the third flight of a full-scale Starship prototype?

If nothing else, Wednesday afternoon’s flight provided several minutes of first-rate entertainment: rocket ship goes up. Rocket ship comes down. Rocket ship lands. And then, with an incredible plot twist 10 minutes later, rocket ship briefly ascends again and then blows up.

It all looked remarkable. Like many of the most inspiring things SpaceX has accomplished over the last decade, this launch, landing, and subsequent explosion looked almost otherworldly. It felt like a peek into the future, a glimpse of something yet unseen, that might yet be.

SpaceX founder and Chief Engineer Elon Musk seemed pleased after the flight. “SpaceX team is doing great work! One day, the true measure of success will be that Starship flights are commonplace,” he tweeted.

So the Starship program is clearly making progress. But is it getting to where it needs to be? Three explosions in three flights are hard to ignore.

What happened

Let’s start with the technical details of Wednesday’s launch. Like the flights of SN8 and SN9, in December and February respectively, SN10 ascended to an altitude of about 10 km. This flight looked smoother than the previous ascents, but that could have been an optical illusion. Like those previous flights, SN10 then executed a belly flop maneuver and began falling toward Earth.

Unlike the earlier flights, this vehicle had no issues with relighting its Raptor engines. All three ignited as the vehicle neared the ground, and then one was shut off by design. With two engines, the vehicle completed its reorientation to a vertical position. Then, as intended, the vehicle touched down slowly on the landing pad with a single engine.

SpaceX engineers have been experiencing some difficulty with Starship’s stubby landing legs—they’re likely a temporary design fix, and at least some of them must have been crushed on Wednesday. It also looked like Starship may have briefly bounced. But the vehicle touched down, and although leaning, with a bit of a fire at the base, it made it back on the ground in one piece.

For about 10 minutes, it stood there. Suddenly, the vehicle briefly rose upward in a violent explosion and crashed back into the pad. SN10 was no more. SpaceX has yet to provide details about what happened and likely won’t. However, informed sources suggested the accident may have been caused by a leaking valve, likely methane fuel. It is notoriously difficult to operate fuel valves at cryogenic temperatures.

Takeaways

SpaceX engineers must be delighted to have figured out the vexing issues with propellant and Raptor relighting that had scuttled the two previous landing attempts. This was a consistent problem with the Falcon 9 program, and its simpler Merlin engine design, in making a successful landing. For a Starship to land after only its third high-altitude flight is notable.

SpaceX, too, will take away a lot of meaningful data from this launch, flight, and landing that it can use to refine both the design of Starship as well as its flight software.

What we don’t know is how NASA will see this. Will it be deemed progress? Or as a negative, with the third destruction of a Starship in three flights? This matters as the agency gets closer to a down-select next month for its Human Landing System contract that could see billions of dollars flow to SpaceX for its Starship program—or not. NASA may decide to go with more conventional landers under development by teams led by Blue Origin and Dynetics.

With that said, NASA is not stupid. Starship is undergoing a unique development program, progressing through rapid iterations and taking risks by design. Each failed mission buys down risk for future flights. It’s no accident that SpaceX is building a new Starship every two or three weeks in South Texas. Being hardware-rich means you can move fast, try, fail, try again, and ultimately succeed.

SpaceX first used this method to develop the Falcon 1 rocket, which made it to orbit in 2008 after several failures. And Musk now believes this method offers the straightest path toward getting Starship into orbit late this year or early next. Enough good things happened on Wednesday to believe he may just be right.

Listing image by Trevor Mahlmann