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
More generic comment on making new things. Just because we can make one thing really well, doesn’t mean that if we try to make it better or different it will work the first time.
It just so happens that for 99.9% of products on the market the companies that make them get to do it behind closed doors and not in front of the whole world.
So when the slightly improved gadget you just designed breaks spectacularly you only have to answer to your boss, not the media.
TLDR: stuff breaks all the time when you’re first designing it, rockets stand out because they’re big.
It’s pretty easy to make a rocket that works with a large enough budget. The really hard part is making a rocket that not only works, but is useful, as in, can carry a payload and launch for a competitive amount of money. That requires pushing the limits, cutting weight wherever possible, and cutting a lot of corners. Rockets are constantly operating at the very limit of their capability and that makes it very easy for a failure to happen if anything goes slightly to far, like a valve that isn’t opening properly and causes pressure to build up slightly too high.
To make it worse, rockets are forced to handle some of the most extreme environments of any machine. Liquids at -150°C quickly combust into gases at 2000°C, with exhaust velocity measured in kilometres per second, hundreds of tonnes of cryogenic propellant in a tank that’s never more than a centimetre or two thick, and aerodynamic forces to rival supersonic aircraft.
The bottom line is, rockets have almost no margin for error because if they did, they wouldn’t be profitable, and that means it’s very easy to push them too far. If that happens, there’s a small nuclear warhead’s worth of energy in the fuel tanks ready to blow the whole thing up.
The same reason we can’t make a car that doesn’t fall a part. We are building a mechanical structure, moving parts have a ton of variables that create failure points some of which are environmental (that we can’t control).
Because the forces are massive, the margins of error are small, and there are always events you either can’t predict or can’t do anything about even if you see it coming. Every single thing has a failure chance.
We’ve been making cars for longer and those still fail, sometimes in spectacular fashion.
One of the big problem with rockets is that most of them are only flown once.
Which means that any manufacturing defect will only be found out on launch, by which point it’s hurtling skyward and will either reach orbit, or fail.
It’s no coincidence that the most reliable rocket in history, the Falcon 9 Full Thrust, is also the only one that’s been meaningfully reused. SpaceX haven’t actually built that many of them, they just fly them a dozen times each. Plus, because they get many flights out of each first stage, they can spend more on building each one, and therefore it’s economical to perform a more thorough inspection.
When cars or aeroplanes are built, they are taken for short test drives or test flights. This allows faults to be diagnosed in a safe environment. If this didn’t happen, those vehicles would be a lot less reliable (which sometimes happens when this step is skipped, for example in wartime).
As for exploding vs other faults (like just crashing), rockets are fitted with flight termination systems, so they don’t crash while intact and level an entire neighbourhood. So any major fault, and the rocket is blown up.
Rockets require an absurd amount of power:
The largest jet engine produces around 100,000 lbs of thrust.
The Raptor engines that SpaceX use produce 600,000 lbs of thrust, and there are ~~9 of them on the falcon and~~ 30+ on starship.
(Edit for the correction there, Falcon uses 9 smaller engines that have around 200,000 lbs thrust each)
The Shuttle solid rocket boosters produce 3,300,000 lbs of thrust each.
Containing that amount of power in a small space where you are trying to make everything as light as possible is very difficult no matter how long you’ve spent working on it.
I think one could assert that the current iteration of the Space X Falcon 9 is relatively successful now. But freezing the design in its current state doesn’t seem typical for them.
When we build things of earth, we generally work to a tolerance of “this needs so much strength, make it ten times stronger”. This makes everything very reliable.
A rocket that is more than 1.5 times heavier than it needs to be will simply never fly. Or require so much fuel and complex engineering that it will make the cost prohibitively expensive.
This forces engineers to work to fine margins. If you’re guessing you need a strength of 10, so build it with strength 100, and you were wrong and it actually needed 11, no one will ever know.
If you did the same but built it with strength 11, your rocket just blew up.
Because human errors occurs, both in main assembly and sub-contactor assembly including during delivery. No complex system ever has 0% bugs. It’s just a question of what the bug affects and time. Try reading up on “Murphy’s law”. The higher the complexity the higher the probability of bugs/glitches
Space x have not had one of their rockets fail outside of test flights in ages which I think is really impressive.
but either way rocket science is hard. everything is all high temperatures and high pressures the materials science is still catching up.
Making rockets is expensive. so getting a chance to “mess around with stuff” is not possible. SpaceXs great idea was to have enough money that allowed them to fail loads of times and get it right
Latest Answers