Wiki Super Heavy/Starship - General Development Discussion

[Ben W]

Unfortunately, the GEO tank farm would be untenable. You're going to spend propellant to get your tanker there, and then back again for another trip. That means you're using propellant to push the tanker and its return propellant around. The amount of propellant coming from the Earth explodes in mass. You'd be killing yourself just trying to get one load of propellant to the GEO tank farm.

True, supplying non-LEO depots really does call for a Spaceship-only design, perhaps with ion thrusters for propulsion instead of methalox. (But with methalox as the payload.) How long would it take for an ion-propulsion kick stage to boost a full methalox depot from LEO to GEO? Napkin math: experimental ion thrusters can provide 5N thrust with 100kW power. Let's attach 100 of these to a kick stage, with nuclear to supply 10MW. That would give 500N of thrust, or about 1/25,000 the thrust of Starship with six Raptors. (Ion engines are ~70% efficient, so let's imagine we can dissipate 3MW of heat somehow; this may require a truly massive radiator.) If it takes 4 minutes of thrust for an ordinary Starship to reach GEO from LEO, it would take this ion thruster about two months. Once the depot fuel is used by an outbound Starship, the kick stage can boost the empty depot back to LEO for another cycle.

Then again, if such a kick stage is feasible, it would probably end up being used instead of methalox for pretty much all space-to-space missions, with the possible exception of moving crew quickly between LEO and low lunar orbit?

I've been trying to wrap my head around an optimal way of doing round trips, but it's just so ugly without an LMO tank farm supported by a Moon base. A fully-tanked stack requires 4,600 tons of propellants. Optimistically assuming that a Starship tanker could get 200 tons to LEO, that would mean twenty-three launches to get a LEO stack loaded.

LMO = low-medium (eccentric) Earth orbit? What value would the Moon base provide? (I don't think liquid-methane ISPP is viable on the Moon?)

 

[JB47394]

Then again, if such a kick stage is feasible, it would probably end up being used instead of methalox for pretty much all space-to-space missions, with the possible exception of moving crew quickly between LEO and low lunar orbit?

It's a big 'if', but yeah, if you could put a 10 MWe nuclear reactor on it, then I think it could get from LEO to LMO in a couple months. It would give a 450 ton vehicle (400 tons of cargo) about six times the performance of the ESA SMART-1 mission. They used a tiny Hall Effect thruster to get a small spacecraft to Moon orbit in 14 months. I should probably mention that the thrusters run for less than half the mission time because the spacecraft is laboriously moved closer and closer to the Moon with ever larger orbits around the Earth. So it may still take rather longer than two months because you spend a lot of time waiting for the orbital mechanics to work out for the next period of acceleration.

As a point of reference, NASA's 10 kWe Kilopower space reactor is expected to mass 1500 kg. To get the required 10 MWe, you'd need 1000 of those, which would be excessive mass devoted to power. If the power was generated with solar panels at 400 W/m2, you'd only need 25,000 m2 of the things. Only.

So SpaceX will need to get into nuclear reactors next.

LMO = low-medium (eccentric) Earth orbit? What value would the Moon base provide? (I don't think liquid-methane ISPP is viable on the Moon?)

Sorry for the lack of an expansion on that. Low Moon Orbit.

The value of the Moon base is that it parks propellants and anything else that can be manufactured there at a place that has a low delta-V to reach (assuming that you're going somewhere beyond LEO/GEO). If we're going to the Moon, then we want to get as much in situ stuff as possible. We don't know what we can pull out of the Moon yet, so it might only be LOX supplies, but that's still the lion's share of the mass of vehicles moving around on methalox - or hydrolox. If we can't get methane out of the Moon in any reasonable way, then missions to the Moon may prefer hydrolox. Water ice gives us hydrogen and oxygen.

Or, as you suggest, ion thrusters. It only takes a couple hundred kg of propellant to move a lot of mass between LEO and LMO. But power remains a significant problem there.

A Moon base also makes trips to all points outside the Earth-Moon system more accessible. If a ship can go to LMO and tank up there, then it can continue on to other destinations with full tanks. The alternative is to tank up in LEO with everything you need to reach, say, Mars. When you leave LEO, you're going to be pushing a lot of propellant. The rocket equation is brutal. It's much better to tank up at points along the way.

The general pattern is that if you've got a low mass object somewhere in the solar system that has propellants, then you've got a natural site for a propellant farm.

 

[JB47394]

A side-by-side replay of IFT-1, IFT-2 and IFT-3. It's interesting to see how IFT-2 and IFT-3 compare. Staging was at the same elapsed time, not at the same altitude or speed. The balance of propellants was different. IFT-2's Starship was higher, but IFT-3's Starship was faster. And so on.

 

[Ben W]

It's a big 'if', but yeah, if you could put a 10 MWe nuclear reactor on it, then I think it could get from LEO to LMO in a couple months. It would give a 450 ton vehicle (400 tons of cargo) about six times the performance of the ESA SMART-1 mission. They used a tiny Hall Effect thruster to get a small spacecraft to Moon orbit in 14 months. I should probably mention that the thrusters run for less than half the mission time because the spacecraft is laboriously moved closer and closer to the Moon with ever larger orbits around the Earth. So it may still take rather longer than two months because you spend a lot of time waiting for the orbital mechanics to work out for the next period of acceleration.

As a point of reference, NASA's 10 kWe Kilopower space reactor is expected to mass 1500 kg. To get the required 10 MWe, you'd need 1000 of those, which would be excessive mass devoted to power. If the power was generated with solar panels at 400 W/m2, you'd only need 25,000 m2 of the things. Only.

If it's used for raising a depot to high orbit, the "cargo" would be 1200 tons of methalox. (Or 3000 tons of liquid xenon? One can dream!) And if the power source is continuous, it may make sense to just keep accelerating even at non-optimal times, depending whether ion fuel consumption is the limiting factor. An advantage of solar panels is that they can double as a sun shield, reducing boiloff, as well as providing shade for waste heat radiators. The ISS solar panels currently generate about 200 watts per kilogram; suppose this could be doubled to 400 watts per kilogram. (One kilogram per square meter, 400W per square meter). Then a 10MW solar array (spanning about five American football fields, 10x the area of the current ISS solar array) would mass around 25 tons, MUCH lighter than the Kilopower reactor equivalent (1500 tons). Astronomers might not be too keen!

The value of the Moon base is that it parks propellants and anything else that can be manufactured there at a place that has a low delta-V to reach (assuming that you're going somewhere beyond LEO/GEO). If we're going to the Moon, then we want to get as much in situ stuff as possible. We don't know what we can pull out of the Moon yet, so it might only be LOX supplies, but that's still the lion's share of the mass of vehicles moving around on methalox - or hydrolox. If we can't get methane out of the Moon in any reasonable way, then missions to the Moon may prefer hydrolox. Water ice gives us hydrogen and oxygen.

Or, as you suggest, ion thrusters. It only takes a couple hundred kg of propellant to move a lot of mass between LEO and LMO. But power remains a significant problem there.

A Moon base also makes trips to all points outside the Earth-Moon system more accessible. If a ship can go to LMO and tank up there, then it can continue on to other destinations with full tanks. The alternative is to tank up in LEO with everything you need to reach, say, Mars. When you leave LEO, you're going to be pushing a lot of propellant. The rocket equation is brutal. It's much better to tank up at points along the way.

The general pattern is that if you've got a low mass object somewhere in the solar system that has propellants, then you've got a natural site for a propellant farm.

All very good points. It's even possible that xenon could be mined from lunar soil; it is too heavy to escape the moon's gravity like lighter gases. (Xenon was found in Apollo rock samples.) And if lunar ice can produce liquid hydrogen, it may be possible to bring just the carbon and create liquid methane, though it may be more trouble than it's worth vs just hauling the liquid methane. So if it's possible to create hydrolox on the Moon, and use hydrolox hoppers to lift tankers of liquid oxygen (or xenon, or ?) to LMO, that may be the best case.

 

[BrockMisner]

THANK YOU for posting!

 

[mongo]

A side-by-side replay of IFT-1, IFT-2 and IFT-3. It's interesting to see how IFT-2 and IFT-3 compare. Staging was at the same elapsed time, not at the same altitude or speed. The balance of propellants was different. IFT-2's Starship was higher, but IFT-3's Starship was faster. And so on.

That fits with a lower perigee and closer target (Indian vs Pacific Oceans)... I think...

 

[scaesare]

A side-by-side replay of IFT-1, IFT-2 and IFT-3. It's interesting to see how IFT-2 and IFT-3 compare. Staging was at the same elapsed time, not at the same altitude or speed. The balance of propellants was different. IFT-2's Starship was higher, but IFT-3's Starship was faster. And so on.

Having those half a dozen Raptors out vs all of them lit really makes the rocket seem to leap off the pad rather than sluggishly struggle to gain altitude. How fast it clears the tower is significant...

 

[DarkandStormy]

https://twitter.com/x/status/1768810190718009446

Another delay.

 

[ecarfan]

In 2017 at the IAC in Australia Elon said:

We are targeting our first cargo missions [to Mars] in 2022…that’s aspirational.

In his most recent statement about the timeline, “on Mars within 5 years”, he doesn’t specify if that is just cargo or including humans. Personally I think a flight with humans could be possible 5 years from now (launch window Q1 2029). A flight with cargo seems possible in Q4 2026, the next launch window after the one at the end of this year.

Such a flight will of course require in-orbit refueling and perfecting the ship landing technique. I envision SpaceX sending multiple Starships during the launch window, spaced days apart, using each Mars landing attempt to gather data and then modifying the landing algorithms to try to improve the next landing.

 

[Max Spaghetti]

Once the cargo is at the LMO tank farm, a separate vehicle that cycles between LMO and the Moon's surface takes it down. It goes down, unloads its cargo, tanks at the surface, takes on propellant cargo, goes back to LMO and transfers propellant to the LMO tank farm. That vehicle would be custom designed for its mission as well. It would have a landing capability, and be structurally strong enough for the rigors of Moon operations.

So I guess the bottom line here is that Starship is designed specifically for Earth operations, and we're going to need new vehicles for Moon operations. But that's a ways off in the future.

Are you thinking like a NASA scientist?

Elon shook up the space-science world when he suggested building a rocket out of 316 stainless rather than the latest carbon-composites.

Heavy? Yes. Inefficient? Yes. Cheap? Yes.

Look who has changed the economics of space.

Your science is sound, but the earth-moon shuttle that actually works economically will probably be a stainless-steel Starship.

 

[JB47394]

In his most recent statement about the timeline, “on Mars within 5 years”, he doesn’t specify if that is just cargo or including humans.

Or smeared across the surface.

Personally I think a flight with humans could be possible 5 years from now (launch window Q1 2029).

We'll be lucky if we see a manned flight to the Moon in five years.

A flight with cargo seems possible in Q4 2026, the next launch window after the one at the end of this year.

I'd guess a test flight with some experiments onboard launching in 2029 - at best. Ideally, carrying some really capable flying rovers because nobody even knows where to start a colony on Mars. The planet's surface is vast, and we have no idea where to set up shop. They can start somewhere, but collecting enough information to decide where to really invest is going to take time. Imagine finding a location that sits on a buried glacier of water ice versus starting a colony in a location with no exceptional resources.

With so many flight tests needed on Starship and the booster, SpaceX should really focus on the Moon because of the faster turnaround of each test.

 

[JB47394]

Your science is sound, but the earth-moon shuttle that actually works economically will probably be a stainless-steel Starship.

I'm thinking like Elon. He wouldn't take a standard part that he has and make it work. Instead, he'd create a part that is custom-tailored for the job. Starship is engineered for launching from the Earth's surface to LEO. That's it's reason for being. It's massively over engineered for operations in space. It doesn't need to be aerodynamic, it doesn't need to devote all that mass to structure, it doesn't need six engines, and certainly no sea level engines. And so on.

As for the economics of it, nobody is going to need lots of Earth-Moon shuttles. They might need a handful. The reason for mass producing Starship is because of the need to send so much mass to Mars, where atmospheric reentry is a thing. I suspect that they'll alter the Starships going to Mars because the demands of reentry and even of landing will be different there. For now, they're probably over engineered even for Earth operations. Consider what IFT-1 went through without breaking up. The thing is a tank.*

*And that's an awesome multiple-entendre going back to WWI

 

[scaesare]

I wonder if Tom is a tad wistful:

https://twitter.com/x/status/1769049396568797406

 

[ecarfan]

Or smeared across the surface.

Hah hah! That certainly is a third possibility, in fact even the likely outcome of the first few Mars landing attempts, which is why I think SpaceX could plan to send multiple ships staggered days/weeks apart during the Q4 2026 launch window. They will simply be test vehicles to figure out how to land safely and accurately, no other purpose. SpaceX has such a huge advantage because they will learn to build vehicles relatively inexpensively.

We'll be lucky if we see a manned flight to the Moon in five years.

Well, that’s a US government program which operates on a far different and budget constrained timeline compared to what SpaceX can do on its own.

With so many flight tests needed on Starship and the booster, SpaceX should really focus on the Moon because of the faster turnaround of each test

I think SpaceX can potentially do both; fulfill its HLS contract obligations and build Mars test vehicles for the Q4 2026 launch window.

 

[bxr140]

Before getting to higher payload capacities, they need to demonstrate landing and in-orbit refueling.

That's right. We're a year if not years away from starship being properly operational, let alone trying to land anything on the moon, to say nothing of going to Mars.

I guess the bottom line here is that Starship is designed specifically for Earth operations, and we're going to need new vehicles for Moon operations.

That's the bottom line here. Starship's primary design point is to put a bunch of starlinks into LEO. Certainly that capability can be leveraged into beyond-LEO activity, but going beyond LEO will end up with unique variants. For instance, there's not a ton of First Principals in dragging a sea level raptor around the solar system.

On depots, I think the answer ends up being that you really only want it in two places: 1) orbiting the celestial body where production occurs, and 2) a final destination where the return trip is otherwise untenable on available propellant. I don't think there's much efficiency in pushing a lot of propellant to a way station; I think a more appropriate solution for the Mars case is to either a) make a bigger thing that can carry more propellant and get the payload there from LEO, or b) send more things from LEO.

Alternate propulsion solutions for pushing propellant around are certainly worth the thought experiment; the major upside of something like electric propulsion is really just the increased efficiency of the propellant vs Starship's methalox....and then there's quite a bit of downside, notably complexity and cost--both the equipment as well as the propellant itself. And from a practical perspective we're way off on technology. FWIW Big Electric propulsion thrusters in service pretty much max out at low hundreds of mN thrust. The 1000mN ones on PPE are pretty much cutting edge right now (Busek has a 1N unit in dev also, and they're supplying some of the lower thrust PPE thrusters). There's experimental stuff out there, but...it's a bit like following development in battery technology. Lots of "in the lab" results...

 

[scaesare]

That's right. We're a year if not years away from starship being properly operational, let alone trying to land anything on the moon, to say nothing of going to Mars.



That's the bottom line here. Starship's primary design point is to put a bunch of starlinks into LEO. Certainly that capability can be leveraged into beyond-LEO activity, but going beyond LEO will end up with unique variants. For instance, there's not a ton of First Principals in dragging a sea level raptor around the solar system.

On depots, I think the answer ends up being that you really only want it in two places: 1) orbiting the celestial body where production occurs, and 2) a final destination where the return trip is otherwise untenable on available propellant. I don't think there's much efficiency in pushing a lot of propellant to a way station; I think a more appropriate solution for the Mars case is to either a) make a bigger thing that can carry more propellant and get the payload there from LEO, or b) send more things from LEO.

Alternate propulsion solutions for pushing propellant around are certainly worth the thought experiment; the major upside of something like electric propulsion is really just the increased efficiency of the propellant vs Starship's methalox....and then there's quite a bit of downside, notably complexity and cost--both the equipment as well as the propellant itself. And from a practical perspective we're way off on technology. FWIW Big Electric propulsion thrusters in service pretty much max out at low hundreds of mN thrust. The 1000mN ones on PPE are pretty much cutting edge right now (Busek has a 1N unit in dev also, and they're supplying some of the lower thrust PPE thrusters). There's experimental stuff out there, but...it's a bit like following development in battery technology. Lots of "in the lab" results...

If a Starship is going to return to earth, wouldn't you want SL Raptors for the catch?

 

[JB47394]

If a Starship is going to return to earth, wouldn't you want SL Raptors for the catch?

Yes. He was referring to Starship variants that would never reenter Earth's atmosphere.

 

[bxr140]

If a Starship is going to return to earth, wouldn't you want SL Raptors for the catch?

For sure. And of course you need them to get up there in the first place (though maybe that can be done with just vac raps for some vehicles?).

It's certainly going to be interesting to see where SX goes with the concept of application specific variants. It's plausible early on (the first moon missions, maybe?) that there won't be dedicated space-only vehicles. You really only need a lander variant and tanker variant, and it does make a bit of sense to not put resources into doing anything else. But...certainly once they get to heavier moon missions and certainly Mars, it's easy to imagine significant value in very application specific variants. Brute force (or in this case, basically just one version of SS) only goes so far when it comes to First Principals.

That's also next decade, so plenty of time to iterate.

 

[scaesare]

I wonder what the larger bell size for an RVac adds to total mass.

 

[ecarfan]

I wonder what the larger bell size for an RVac adds to total mass.

Well, every kg matters. Recall that recently SpaceX has been flying F9 2nd stage engines with a cut off bell when they don’t need quite as much performance.