Eli5: “Why do spacecraft keep exploding, when we figured out to make them work ages ago?”

431 viewsEngineeringOther

I know its literally rocket science and a lot of very complex systems need to work together, but shouldnt we be able to iterate on a working formular?

In: Engineering

41 Answers

Anonymous 0 Comments

Most rockets don’t explode. We have a formula to send things to space. However when we push the limits, and experiment in making our rockets better, we often fail.

SpaceX in particular, when testing their rockets use rapid testing models for development. They make changes and test it, see where it went wrong and improve it. So they have lots of failures by design

Anonymous 0 Comments

We do iterate and we get better.

We developed cars a hundret years ago and people still die in car crashes.

There is far less lethal accidents in space compared to 20 years ago.

Anonymous 0 Comments

They are trying to drastically cut the cost per.Ton to launch. There are very high reliable launch systems that are very expensive. If your satellite itself is relatively cheap to make, do you go with the launch system that has 5% failure for 10 million dollars or 20% failure for half a million dollars. (For example, not real numbers)

Anonymous 0 Comments

I performed final tests on rocket control systems for NASA, Northrop Grumman, etc and there are multiple reasons rockets fail. One that you’ll never completely avoid is humane error. I experienced so many corrective actions that were put in place to prevent accidents and they still happened. Some are ridiculously simple and I wonder how it could happen but it does. Humans will always make mistakes and that will never stop.

Anonymous 0 Comments

To make getting large, heavy things to orbit faster and cheaper we need to push the boundaries of engineering. The harder and faster you push them the faster you make progress, but you also have more catastrophic failures among the way.

SpaceX takes this push hard, fail hard approach to rapidly iterate their designs. By contrast, NASA and ~~big~~ established contractors like ULA prefer to spend long development cycles to avoid failures. Both approaches are valid, SpaceX’s is more materially expensive and faster and has more high profile failures, but the failures are expected in their case.

They also have “solved” rockets they use too like falcon 9 which is the most reliable launch vehicle we’ve ever had if you start counting at the human rated version (you can go back further but that’s a good goalpost).

Anonymous 0 Comments

If everything worked 100% as designed and expected, nothing would ever fail, ever.

Unfortunately, designs and expectations don’t often mesh with reality, especially when you have tiny possible issues, defects, or things behaving in ways you didn’t expect. Some of the disasters of the past were due to a faulty part–where you don’t *expect* a part to be faulty.

Anonymous 0 Comments

Rockets have a lot of constraints that make it kinda a miracle they are as reliable as they are.

You need to cram as much highly energetic fuel as possible into as little rocket as possible. If you use safer or more manageable fuel, or build more rocket to control it you lose the very small capacity you have to lift stuff (other than the rocket & fuel) into space.

We use similar technologies for ICMBs, ground to air, & air to air missiles that are perfectly reliable because it’s an easier problem that allows for some wiggle room in design.


Because it’s a hard problem with no easy compromises. Most games with this many cards stacked against you just aren’t played.

Anonymous 0 Comments

Because most companies are trying to make something better than the previous generation, using new technology and material. To optimize weight and performance, most parts are custom for each new vehicle and so, while lessons have been learned, it’s still a lot of new technology development with heavy constraints on weight and cost.

Anonymous 0 Comments

Space is hard, really hard. It looks simple, but it’s very complex. Unknowns are at their greatest with a new design, so ways to fail are at their greatest.

Take something as simple as fuel sloshing. We can model that, but not with 100% accuracy. So you launch it, the fuel sloshes anyway, and it blows up.

Takes SpaceX. They have the Falcon 9 that is extremely reliable. They’ve blown up several test articles with Starship because rockets at this scale with those abilities have never been done before. Also, SpaceX is pushing the envelope on what can be done, which makes it harder. For example, engineering says they may be able to save weight by not having shielding on the engines. Nope, blew up, put the shielding on.

Anonymous 0 Comments

Rockets don’t explode anymore. I can’t remember the last launch of a fully developed in production established rocket system. What we do hear about is explosions of in development, new and experimental rocket systems. It’s quite different.