I LIKE TO SAY THAT ROCKETRY IS ONE OF THE FEW THINGS IN THE WORLD WHERE THE OUTCOME IS BINARY: IT EITHER WORKS OR IT DOESN’T. NO HORSESHOES AND HAND GRENADES HERE. IN SHORT, THERE IS NO PARTIAL CREDIT IN THE BUSINESS OF LAUNCHING STUFF INTO SPACE WHEN LAUNCH DAY COMES.
However, as with all things, there is a longer answer that offers some room for nuance. This longer answer starts with a simple motto: if you are going to make mistakes, it is important to make new ones. This doesn’t mean that there’s no room for failure along the way. It just means that it’s important to fail as quickly as possible while the stakes are low and to always be implementing what you’re learning as you iterate to success.
A local investor recently emailed me and asked if the Starship program is doomed due to the simple fact that things keep blowing up. From the early days of Starhopper pressurization tests to the early losses of SN1 through SN4 on the ground, and then the more recent losses of SN8, SN9, and SN10 during landing operations following high-altitude flight tests, there has been a lot of twisted metal flying around the prairie. From the outside, it is dramatic and can look like a serious issue. Many people may look at this explosion and think, “well, hell, that sucks that they failed.”
And, in a way, they wouldn’t be wrong — it does kind of suck. I can tell you from the early days of working F9 first stage landings at sea, that it’s a bummer to see your work go up in a cloud of smoke. But hard things remain hard and it’s only in having a constant conversation with reality, through the language of physics, expressed with the practicality of engineering, that we can progress to success. So, all of this is to say we must look at flights like the SN9 in a different light. Let’s look at it through the lens of making new mistakes, rather than repeating old ones.
Raptor ignition of #SN8,#SN9,#SN10 prior to landing.@SpaceX pic.twitter.com/qrsNH1YIqM
— Saumit 💻🚀🛰️ (@SaumitPradhan) March 8, 2021
The video on the left is of SN8 and the one on the right is of SN9. At first glance, SN9 seems a lot worse for sure. It overcorrected by a lot on its attempt to stand back up vertical after its bellyflop and hit the ground at a 45-degree angle. Although if you look closely, you’ll notice they also started quite a bit later on the relight on the SN9 launch. The video for SN9 shows the rocket engine hitting full thrust a lot later than SN8. In short, they started the engines later this time. There are three possible reasons for this, in my mind.
The first explanation is simply that they had an engine relight issue. This seems almost assured, after all, only one engine kicked fully back on and they want two for landing. The second, less obvious explanation is that it seems they were pushing the relight to the last possible second. Why would they do this? Put simply, to save fuel. Saving fuel is really advantageous because it allows some combination of less parasitic mass to space, more usable payload, and/or a longer flight time. However, the most likely explanation is the third: that there was both a relight issue and SpaceX was trying something new.
My bet is that SN8 was set conservatively to relight at a higher altitude because it was the first flight and they simply didn’t know how much time they’d need to establish full control and slow descent. They figured out that they could do it faster than they planned, even though SN8 slammed into the ground, it was going slow. With SN9, they decided to shorten the relight window, getting them closer to how real operations will be in the future. They pushed the boundary to make a new mistake and learn something in the process.
Keep in mind that SpaceX has been making “new mistakes” for years — and to their benefit, too. Again, when I was at SpaceX, I was a part of a Falcon 9 landing that came in sideways, blasted into the side of the landing barge I helped design, and blew up. We rebuilt the barge in a few weeks. And then we blew it up again. And again. But every single time, we made progress. We iterated on the design of the barge and the vehicle team iterated on Falcon 9. I always joke with engineers on the launch and landing sites that, “there is no shortage of problems to solve or mistakes to make! Don’t get attached to this one. You’ll find more soon.”
Yes, not being able to relight is a challenge. But they are iterating through that. With SN10, we saw that they lit all three engines. That way if only two light in the process, you’re golden. We saw that vehicle landed! Of course, there was what appeared to be a landing-burn-induced ground fire that may have compromised the legs or propellant tank. Whatever happened, SN10 took back off again and did a second, shorter flight with a dramatic ending. But, again, now we have new information and we can go make the next mistake on the way to orbit.
WHILE IT MAY LOOK LIKE SPACEX IS FAILING AT LAUNCHING ROCKETS, THEY ARE ACTUALLY GIVING THE WORLD AN EXCELLENT LESSON IN TRYING TO DO SOMETHING REALLY DIFFICULT: RESPONDING QUICKLY TO WHAT THEY LEARN FROM EACH PREVIOUS LAUNCH. THE BEST MISTAKES ARE THE ONES THAT YOU LEARN FROM AND IMPROVE ON, THESE SEEM LIKE THOSE TYPES OF MISTAKES.
