3 ms·

I disagree with both. > One explicitly, that Starship will work in anything resembling its current design and promised operational paramters. We have a test f

by avmich 3y ago

I disagree with both. > One explicitly, that Starship will work in anything resembling its current design and promised operational paramters. We have a test flight behind us, which showed many tens of seconds of the stack flight. That's a big achievement validating the rocket architecture and general parameters. Frankly I'd assume we crossed equator in rocket development here, we're in completion stage more than developing. > This is by no means as guaranteed as you make it out to be (it's very much an if, not a when). You know, there are no physical laws which prevent Starship from working. SpaceX shows economics on their side. At this point it's not guaranteed that Starship will fly - because it's hard to predict the future - but it's a pretty safe bet that Starship can be made to fly. > The promises of rapid reusability are very much up in the air, and the whole many-engine design is also not proven in the slightest yet. There are many successful many-engine rockets. Soyuz has 5 engines but 32 nozzles firing at launch. Falcon Heavy has 27 separate engines. Saturn-I had 8, Falcon 9, Black Arrow 8 - the last one just 4-5 times less engines than Super Heavy. To say it's not proven in the slightest is bending the reality a bit too much. > Secondly, you're assuming that reusable rockets are better than one-time use ones. While this may be true for certain kinds of missions, it is not at all clear for many others. It's important however what those cases are. By now we have a good history of Falcon-9 launches, where for a number of years we have a substantial reduction of cost to orbit for a wide variety of payloads. So maybe - maybe - there are many other kinds of missions where reusability is a disadvantage, but those kinds are clearly in substantial minority. > Reusable rockets fundamentally will have lower payload for the same thrust, for the simple reason that they need fuel for the return stage. They will also need to be made from sturdier materials to survive reentry, which again limits their capacity. That's technically true, but arguing this is a bit like arguing that a car made for a single travel would be lighter and cheaper that the one made for repeated use. True, but quite irrelevant for most practical cases.

tsimionescu 3y ago
> We have a test flight behind us, which showed many tens of seconds of the stack flight. As became apparent in the post mortems, the rocket lost its chosen flight path the moment it left the pad, probably due to the engine failures. It flew of course, but it was a failed launch from T+3 or so. The safety systems also failed to self-destruct the rocket on command - they thankfully did enough damage for the atmosphere to take care of the rest. Basically the launch only proved that the parts hold together and the engines produce enough lift. The engines themselves are still far too prone to failure (apparently this was using an older generation, so maybe that part is already fixed). > There are many successful many-engine rockets. Soyuz has 5 engines but 32 nozzles firing at launch. Falcon Heavy has 27 separate engines. Saturn-I had 8, Falcon 9, Black Arrow 8 - the last one just 4-5 times less engines than Super Heavy. To say it's not proven in the slightest is bending the reality a bit too much. You're right, I was exaggerating a bit. Still, my layman's understanding is that there is a fundamental difference between Falcon Heavy's multi booster design with 9 engines per booster, and Super Heavy's 33 engine design on a single booster. Maybe I'm wrong on this. > So maybe - maybe - there are many other kinds of missions where reusability is a disadvantage, but those kinds are clearly in substantial minority. As I understand it, the disadvantages start once you want to put large payloads beyond orbit. Starship's main goal is to put a payload on the moon (this is what the NASA funding is for) and for this purpose it is much worse than a traditional single use rocket: instead of a simple trip to the moon, the design is to launch to orbit, then do an additional 6 or so launches to refuel it while in orbit, and only then will it have enough power to fly to the moon. SpaceX is banking on the launch being so cheap with their hopes of rapid reusability that they can do 7+ launches of Starship for less than the cost of a more traditional mission. Falcon 9 definitely is not that cheap compared to other rockets, but perhaps Starship will indeed be.
- avmich 3y ago
  Let's focus on the last part, I think it's more interesting. > As I understand it, the disadvantages start once you want to put large payloads beyond orbit. Starship's main goal is to put a payload on the moon (this is what the NASA funding is for) and for this purpose it is much worse than a traditional single use rocket: instead of a simple trip to the moon, the design is to launch to orbit, then do an additional 6 or so launches to refuel it while in orbit, and only then will it have enough power to fly to the moon. American plans to go to the Moon in 1960-s included a variant with LEO docking. That would allow using a smaller rocket, as the payload needed to go to the Moon could be assembled on low Earth orbit in several launches. NASA decided to go with one rocket of the size which was very ambitious at the time - about an order of magnitude bigger than the biggest existing one. The Starship goal is to bring even bigger payload to the Moon. Sure, it could be done by scaling up the Starship. But the fate of big rockets so far - Saturn-V, N-1, Energiya, even Space Shuttle to an extent - is that they suffer from the lack of use. The more they are used, the better and cheaper they get. So, no, the SpaceX goal isn't actually to bring even bigger payload to the Moon - it just happens that this scale, which Starship has, is good to land on the Moon a lot of payload in one launch. So, given that you want to use the Starship which is being created for many different missions, how would you use Starship to fly to the Moon? You'd refuel it on LEO. Noe we can remember the actual SpaceX goal - it's flying a lot of people and cargo to Mars, and SpaceX decided to create the biggest rocket they could reasonably plan, which is still not that large given that Mars wants thousands and thousands tons of stuff, so even this big rocket should be used with refueling on LEO. Once again - Starship is the biggest rocket SpaceX can build, from technical and economical standpoint (this reminds me the Soviet engineers chain of thoughts when they were planning N-1). It's still small enough for the task (flying to Mars), so LEO refueling is still needed. If it's still needed, it's logical to use it for Moon flights as well. Yes, instead of one flight (like Saturn-V) Starship needs several, for refueling. In exchange it brings larger payload to the Moon, so since we're planning to go there and stay, it could be justified. So, we're using a universal reusable super-heavy rocket for many different tasks instead of building a system for each task separately. We're losing in time - since we need multiple launches from Earth for single flight to the Moon - but gaining and payload and - crucially - in developing bespoke solutions for different tasks. I think overall it's an advantageous approach.

by avmich 3y ago

tsimionescu 3y ago

> We have a test flight behind us, which showed many tens of seconds of the stack flight. As became apparent in the post mortems, the rocket lost its chosen flight path the moment it left the pad, probably due to the engine failures. It flew of course, but it was a failed launch from T+3 or so. The safety systems also failed to self-destruct the rocket on command - they thankfully did enough damage for the atmosphere to take care of the rest. Basically the launch only proved that the parts hold together and the engines produce enough lift. The engines themselves are still far too prone to failure (apparently this was using an older generation, so maybe that part is already fixed). > There are many successful many-engine rockets. Soyuz has 5 engines but 32 nozzles firing at launch. Falcon Heavy has 27 separate engines. Saturn-I had 8, Falcon 9, Black Arrow 8 - the last one just 4-5 times less engines than Super Heavy. To say it's not proven in the slightest is bending the reality a bit too much. You're right, I was exaggerating a bit. Still, my layman's understanding is that there is a fundamental difference between Falcon Heavy's multi booster design with 9 engines per booster, and Super Heavy's 33 engine design on a single booster. Maybe I'm wrong on this. > So maybe - maybe - there are many other kinds of missions where reusability is a disadvantage, but those kinds are clearly in substantial minority. As I understand it, the disadvantages start once you want to put large payloads beyond orbit. Starship's main goal is to put a payload on the moon (this is what the NASA funding is for) and for this purpose it is much worse than a traditional single use rocket: instead of a simple trip to the moon, the design is to launch to orbit, then do an additional 6 or so launches to refuel it while in orbit, and only then will it have enough power to fly to the moon. SpaceX is banking on the launch being so cheap with their hopes of rapid reusability that they can do 7+ launches of Starship for less than the cost of a more traditional mission. Falcon 9 definitely is not that cheap compared to other rockets, but perhaps Starship will indeed be.

avmich 3y ago
Let's focus on the last part, I think it's more interesting. > As I understand it, the disadvantages start once you want to put large payloads beyond orbit. Starship's main goal is to put a payload on the moon (this is what the NASA funding is for) and for this purpose it is much worse than a traditional single use rocket: instead of a simple trip to the moon, the design is to launch to orbit, then do an additional 6 or so launches to refuel it while in orbit, and only then will it have enough power to fly to the moon. American plans to go to the Moon in 1960-s included a variant with LEO docking. That would allow using a smaller rocket, as the payload needed to go to the Moon could be assembled on low Earth orbit in several launches. NASA decided to go with one rocket of the size which was very ambitious at the time - about an order of magnitude bigger than the biggest existing one. The Starship goal is to bring even bigger payload to the Moon. Sure, it could be done by scaling up the Starship. But the fate of big rockets so far - Saturn-V, N-1, Energiya, even Space Shuttle to an extent - is that they suffer from the lack of use. The more they are used, the better and cheaper they get. So, no, the SpaceX goal isn't actually to bring even bigger payload to the Moon - it just happens that this scale, which Starship has, is good to land on the Moon a lot of payload in one launch. So, given that you want to use the Starship which is being created for many different missions, how would you use Starship to fly to the Moon? You'd refuel it on LEO. Noe we can remember the actual SpaceX goal - it's flying a lot of people and cargo to Mars, and SpaceX decided to create the biggest rocket they could reasonably plan, which is still not that large given that Mars wants thousands and thousands tons of stuff, so even this big rocket should be used with refueling on LEO. Once again - Starship is the biggest rocket SpaceX can build, from technical and economical standpoint (this reminds me the Soviet engineers chain of thoughts when they were planning N-1). It's still small enough for the task (flying to Mars), so LEO refueling is still needed. If it's still needed, it's logical to use it for Moon flights as well. Yes, instead of one flight (like Saturn-V) Starship needs several, for refueling. In exchange it brings larger payload to the Moon, so since we're planning to go there and stay, it could be justified. So, we're using a universal reusable super-heavy rocket for many different tasks instead of building a system for each task separately. We're losing in time - since we need multiple launches from Earth for single flight to the Moon - but gaining and payload and - crucially - in developing bespoke solutions for different tasks. I think overall it's an advantageous approach.