Self-Destruct

(Photo from Cosmic Log)

Applause is usually for a successful mission. It can also be for remarkable salvaging of a failed mission, such as bringing the Apollo 13 astronauts home alive.

It took a while for enough information to filter out for me to get a sense of what happened to Starship. SpaceX probably thought they sounded like NASA with their tweet As if the flight test was not exciting enough, Starship experienced a rapid unscheduled disassembly before stage separation.

They didn’t sound like NASA.

I didn’t want to post about the launch until I got at least a vague notion of what happened. It seems that multiple (at least three) Raptor engines in the first stage failed, the stages didn’t separate as planned, attitude control of the spacecraft was lost and around the four minute mark self-destruct became necessary. Of course SpaceX got loads of data from the launch and brief flight, and yes it flew for a few minutes, but it is hard for me to see the test outcome as reason for applause.

In case you didn’t know, separation of modules in a spacecraft is not as easy as it looks. For decades, NASA did it with explosive bolts. The explosive in the bolts had to be just enough to blow them, designed in such a way that nothing else would be damaged, able to fire at precisely the same instant as each other and able to go through the many stresses and insults of spaceflight without firing until they were told to do so. For decades, one man was the guru for all of that. He was still doing that wizardry when I was working on the Space Shuttle and Spacelab.

NASA does extensive testing of rocket engines on test rigs at Stennis Space Center in Mississippi before going into launch tests. My great grandfather’s house on the coast a few miles away sometimes shook a little during tests and we could hear the roar when a test occurred during our visits. I cannot imagine a test rig that could cope with the complete set of Raptor engines in Thursday’s launch at all, let alone do it for a burn long enough to be meaningful. Maybe that is why SpaceX has gone through so many flight tests that ended in balls of fire. This latest one was the most ambitious, but certainly not the first.

Developing a new launch vehicle often involves exploding failures along the way. Sometimes they happen at the launch pad or during ascent. It’s always better to discover such severe problems as early as possible, when both their consequences and fixing them cost less.

I’ve heard commentators dismiss Starship explosions by saying SpaceX is more willing to take risks than NASA. People are supposed to ride on that spacecraft within just the next few years. They will only ride if they feel reasonably sure they will survive the flight. I wonder whether SpaceX wasn’t ready for a full-bore integrated test flight. In computer software it’s common to roll something out before it’s really ready and remediate it later. That approach doesn’t work for space flight.

The spacecraft isn’t the limit of the damage from this launch test. Powerful rockets are very hard on launchpads and anything nearby. NASA puts great effort into managing that as well as sending rockets into orbit. An immense amount of water pours onto a launchpad during a launch to buffer the tremendous energy coming out of the engines. The launchpad is designed to channel the flow of it all. SpaceX saved money and time by not doing so much quenching and not channeling the output from the launch.

It looks to me like this rocket ship tore up its launchpad too much to be sustainable. The damage from liftoff needs to be less severe for launches to become a regular occurrence. Launch threw dangerous debris alarmingly far, generated dust that came down on the area around the site for miles, broke windows locally and peppered the Gulf of Mexico with debris.

Musk seems happy to throw large amounts of money at his projects. If it’s his money, he can spend it rebuilding what he broke. But if it is not all his money, investors could start to demand a more NASA-like approach instead of the tech industry’s “move fast and break things” approach. Virgin Orbit collapsed after a much less worrying failure. Investors and satellite owners expect rigorous quality in space-related work. They turned away from a failure that looked closer to success than the SpaceX test flight.

NASA chose to fund two development efforts for rockets that take people beyond low earth orbit again. One is based on older technology that is less powerful but better understood and another offers more power if it can fly without blowing up. What happened to the fully integrated launch on Thursday shows why. It isn’t duplication of effort. It’s one wild swing at a large advance in the field, with a more conservative fallback in case the wild swing fails.

Time will tell whether Starship will earn its name. But if it doesn’t, it isn’t our only option.