Moon Starship

[Yggdrasill]

That is similar to sevenperforce's idea near the top of the thread. This seems very natural to me.
* pretty simple requirements and also Elon has mentioned some expendable versions for deep space robotic missions that are probably not that different from this.
* Surely this makes sense in the further future if we develop enough infrastructure, eg the ability to service and replace engines without returning to earth. Unlike earth, there will not be any wear of the thermal protection. The hull should last for ages.
* It could possibly solve some problems for initial unmanned missions landing on unprepared surfaces, eg to deploy the initial landing pads for later manned missions.

Ah, I actually missed that post when reading the thread.

I think inspections could be done on the moon fairly soon, and if the Raptors are actually capable of doing 1000 firings, many years might pass before any real maintenance being necessary. Later, being able to disconnect/unbolt Raptors and send them to earth for maintenance would be good. This would actually be easier on a moon-specific version, without the unnecessary interstage/shroud, which is only there for reentry purposes. If you don't have that, you can have easy access to the Raptors.

Just to adress some of the issues sevenperforce mention:

It also presents challenges for cargo delivery. You would have to transport your cargo out of the standard Starship and into the lunar Starship in LLO, in zero gravity. It's doable, for sure, but it's not easy. There's a big advantage in being able to design and implement a cargo deployment apporach on Earth that you lose if you are doing in-space cargo transfers.

This is perhaps the biggest issue, in my view. It adds significant complexity. You would probably have to dock twice. Once for fuel and then for crew and small cargo. Larger cargo might require EVAs.

But for regular crew replacement operations, larger cargo might not be that common. It might make more sense to send cargo versions for that.

The other trouble is the very problem with Gateway, because LLO is simply not a good place for staging due to mascons

This is pretty managable. With regular operations, you wouldn't need to stage in LLO for very long. You could have a crew returning to earth at the same time as you have a crew going to the moon. So a regular Starship and a lunar Starship arrive at the same time in LLO, most of the fuel is transferred to the lunar Starship, you swap crews and cargo, and then both return to their origin. This could be done in a matter of a couple of days (or hours, if the need presents itself), which isn't enough for mascons to really matter.

Some staging could also be done in frozen orbits, allowing operations for years.

Earth return timing

Having a Starship version designed for lunar operations doesn't preclude also having a fueled regular Starship on standby for quick evacuation. Also, LLO allows for very frequent launch windows. NRLO is a lot worse.

and the desire for polar access.

Operations in LLO doesn't preclude any kind of access to the surface. This is only an issue with a lunar station like the Gateway, as it will necessarily have a specific orbit. A lunar Starship could go into a different LLO every time it leaves the surface, depending on the mission.

[speedevil]

Operations in LLO doesn't preclude any kind of access to the surface. This is only an issue with a lunar station like the Gateway, as it will necessarily have a specific orbit. A lunar Starship could go into a different LLO every time it leaves the surface, depending on the mission.

A base ship/depot is of most banefit in LLO if you are reusing it without changing the orbit.
Removing and replacing reduces the benefit. Trying to take SS and a low mass descent stage all the way back to earth hurts - changing inclination through L1 is less painful, at ~1km/s.

Abort windows to a polar LLO are ~once every 14 days for sites not at the poles, equatorial LLO every hour or so, but only to a narrow band of the surface.

I hope that a lot of this yak shaving is eliminated by accepting fuel is cheap.
We will learn in the next six months hopefully if ~20 launches is of a meaningfully different cost to (say) 10 and a new vehicle development. If it's not - then optimising with special vehicles for tasks may be a whole lot less pointfull.

[livingjw]

Operations in LLO doesn't preclude any kind of access to the surface. This is only an issue with a lunar station like the Gateway, as it will necessarily have a specific orbit. A lunar Starship could go into a different LLO every time it leaves the surface, depending on the mission.

A base ship/depot is of most benefit in LLO if you are reusing it without changing the orbit.
Removing and replacing reduces the benefit. Trying to take SS and a low mass descent stage all the way back to earth hurts - changing inclination through L1 is less painful, at ~1km/s.

Abort windows to a polar LLO are ~once every 14 days for sites not at the poles, equatorial LLO every hour or so, but only to a narrow band of the surface.

I hope that a lot of this yak shaving is eliminated by accepting fuel is cheap.
We will learn in the next six months hopefully if ~20 launches is of a meaningfully different cost to (say) 10 and a new vehicle development. If it's not - then optimising with special vehicles for tasks may be a whole lot less pointfull.

Why do you need a depot? Might be nice to support a base, but then you know exactly what orbit to put it in around the moon. To support general operations on the moon, marrying a SS up with a lunar surface transfer vehicle seems better.

- Why would you want to take the lunar surface transfer vehicle back to earth?

The L1 gateway station, in general, does not improve support for or access to the moon's surface.

John

[speedevil]

Operations in LLO doesn't preclude any kind of access to the surface. This is only an issue with a lunar station like the Gateway, as it will necessarily have a specific orbit. A lunar Starship could go into a different LLO every time it leaves the surface, depending on the mission.

A base ship/depot is of most benefit in LLO if you are reusing it without changing the orbit.
Removing and replacing reduces the benefit. Trying to take SS and a low mass descent stage all the way back to earth hurts - changing inclination through L1 is less painful, at ~1km/s.

Abort windows to a polar LLO are ~once every 14 days for sites not at the poles, equatorial LLO every hour or so, but only to a narrow band of the surface.

I hope that a lot of this yak shaving is eliminated by accepting fuel is cheap.
We will learn in the next six months hopefully if ~20 launches is of a meaningfully different cost to (say) 10 and a new vehicle development. If it's not - then optimising with special vehicles for tasks may be a whole lot less pointfull.

Why do you need a depot? Might be nice to support a base, but then you know exactly what orbit to put it in around the moon. To support general operations on the moon, marrying a SS up with a lunar surface transfer vehicle seems better.

In short - with a depot - a very modest number of extra propellant launches let you use SS and skip the lander.

[Negan]

An expendable SS for the beginnings of a moon base could be a relatively inexpensive when compared to other options. Might even be able to bury it.

Edit: makes me wonder if having windows preinstalled in the tank part would be feasible if it was going to be used as habitable space later.

[sevenperforce]

An expendable SS for the beginnings of a moon base could be a relatively inexpensive when compared to other options. Might even be able to bury it.

Edit: makes me wonder if having windows preinstalled in the tank part would be feasible if it was going to be used as habitable space later.

Not when you can easily cut the window out and weld in a prefab pane. 301 stainless baby!

[Negan]

Not when you can easily cut the window out and weld in a prefab pane. 301 stainless baby!

I was just thinking with regard to doing this in an earth environment verses a moon environment, but the lessons learned doing this on the moon would very be helpful for future missions.

Edit: I think SpaceX would rather pay a worker (and all the associated labor costs) to install the window during manufacturing rather than on the moon.

[jebbo]

I came at this from the opposite direction: assume you want to send a lot of stuff to the moon with SS, what would a system look like?

You want to send max payload on a regular interval. If you assume Elon’s maximum launch cadence of ~3 tanker launches a day, and 2-3 tanker launches to refuel an SS for 150 tonnes to TLI, then with one tanker (+ spares) you can support sending one SS per day to LLO (your preferred orbit may differ ). I’d guess the round-trip time for one SS is ~10 days (transit time, load / unload, maintenance, etc), so you can sustain this with a fleet of about 10 SS (+ spares).

So a rough concept of operations for each SS would be:
-   Launch SS/SH with 150 ton payload to LEO
-   3 tanker launches to refuel the SS
-   SS to TLI then LLO
-   Transfer payload to landing system
-   SS to TEI then land
-   Check / maintain
-   Repeat

If you assume a working year of 40 weeks, this is 42,000 tonnes of payload in a year!

But do you need to return to Earth? With on-orbit payload transfer, you could perhaps have a very stripped down SS acting as a transfer vehicle (or even a dedicated tug), which might reduce the refuelling requirement. However, I don’t really understand the trade-off between delta-v cost for TEI to LEO versus the fuel savings for a much lighter transfer vehicle.

Also, does the refuel need to use Earth-based resources?

--- Tony

[guckyfan]

Also, does the refuel need to use Earth-based resources?

--- Tony

If they can produce just oxygen on the moon it would greatly reduce refueling requirements in space. There have been test setups to produce oxygen from SiO2 which is available everywhere on the Moon. Melt the SiO2 with a solar furnace and split it using electrolysis. The remaining Si may one day in the future be useful for solar arrays.

[jebbo]

If they can produce just oxygen on the moon it would greatly reduce refueling requirements in space. There have been test setups to produce oxygen from SiO2 which is available everywhere on the Moon. Melt the SiO2 with a solar furnace and split it using electrolysis. The remaining Si may one day in the future be useful for solar arrays.

Indeed!

My question is really whether it is cheaper. To use lunar O2, you need to transfer it back to LEO, which requires a lot of capital intensive equipment to set up a system. Which may ultimately be cheaper, but possibly isn't in the near term.

--- Tony

[guckyfan]

If they can produce just oxygen on the moon it would greatly reduce refueling requirements in space. There have been test setups to produce oxygen from SiO2 which is available everywhere on the Moon. Melt the SiO2 with a solar furnace and split it using electrolysis. The remaining Si may one day in the future be useful for solar arrays.

Indeed!

My question is really whether it is cheaper. To use lunar O2, you need to transfer it back to LEO, which requires a lot of capital intensive equipment to set up a system. Which may ultimately be cheaper, but possibly isn't in the near term.

--- Tony

I think of oxygen only for the return flight. I am confident to say that refueling in LEO will always be cheaper than bringing the fuel from the Moon.

[jebbo]

I think of oxygen only for the return flight. I am confident to say that refueling in LEO will always be cheaper than bringing the fuel from the Moon.

Okay, makes sense.

Incidentally, using Elon's minimum turnaround numbers, you only need 1 SH for this system!

--- Tony

[libra]

If they can produce just oxygen on the moon it would greatly reduce refueling requirements in space. There have been test setups to produce oxygen from SiO2 which is available everywhere on the Moon. Melt the SiO2 with a solar furnace and split it using electrolysis. The remaining Si may one day in the future be useful for solar arrays.

Indeed!

My question is really whether it is cheaper. To use lunar O2, you need to transfer it back to LEO, which requires a lot of capital intensive equipment to set up a system. Which may ultimately be cheaper, but possibly isn't in the near term.

--- Tony

Bouncing on this...

You could emplace a propellant depot in cislunar space - at EML-1, EML-2, NRHO, DRO. It would take an average 200 m/s in and out to access these places. Yet the advantages of LUNOX are so huge, maybe it could be interesting.

[BZHSpace]

Gateway will be outdated if Elon Musk succeed with Starship because of the vercatility of the system, however if Starship is the option they choice the impact will be that orbital/support station will disapear expect if they launch two starship one in orbit and one on the Moon surface. But I don't think because of polical interest which motivate Orion spacecraft and LOP-G architecture, the Starship will replace LOP-G. But can provide a great advantage in Artemis mission.

[speedevil]

I came at this from the opposite direction: assume you want to send a lot of stuff to the moon with SS, what would a system look like?

You want to send max payload on a regular interval.

Unless you are specifically volume constrained, or cannot do cargo transfer, 150 tons is not close to the maximum payload to the moon.
In the post earlier in the thread
I outline the infrasructure for a system wth depots at LEO, GTO, LLO.
The most efficient route gets something like 300 tons on the moon per 10 launches. With three depots (which can be repurposed tankers) that remain in LEO, GTO and LLO.
And ideally two tankers that move between them, as well as at least one to retank on earth.

30 tons/launch.
If your cargo is dense enough, 600 tons or more payload transshipped onto one vehicle which then moves between depots, and lands on the moon before returning is
very efficient.

You are also missing some delta-v I think - with your scheme you need four retanks in LEO to get to LLO, and to fill it basically full with eight to get the SS in LEO back, not three. And this neglects the landing propellant for the lander - at least a hundred tons neglecting its mass and return - so it's more like twelve retankings. (if you can fit this additional mass in stretched/extra tanks.
10 tons/launch.

The scheme with a tanker in GTO does mean you can either launch once a month, or so frequently that you have the tanker in GTO empty daily and can do a small manoever to change its orbit.

[jebbo]

Thanks, speedevil.

I had decoupled transfer to LLO and landing, so yes, this isn't payload to the surface ... so I need to rework things to work out a landing system. I'll also do the sums on retanking (I'm assuming direct reentry from TEI).

I like your scheme, but (ignoring the LLO to surface stuff) I was trying to minimise things like depots, payload transfers, etc.

As I have a quiet week at work (China is on holiday :-) ) I'll work things out properly tomorrow

--- Tony

[JonathanD]

Some comments from Dr. Zubrin below.  I'm not too concerned about ejecta as I'm assuming they'll eventually have prepared landing pads by the time a base exists.  And with on-orbit refueling, SpaceShip should be able to get back to Earth without lunar ISRU.

https://twitter.com/robert_zubrin/status/1178786132127039488

[Barley]

If they can produce just oxygen on the moon it would greatly reduce refueling requirements in space. There have been test setups to produce oxygen from SiO2 which is available everywhere on the Moon. Melt the SiO2 with a solar furnace and split it using electrolysis. The remaining Si may one day in the future be useful for solar arrays.

There is less than 10kg of Si in a kW panel.  The mismatch in volume between the need for O2 and Si is almost comical.

Getting any needed SI as a byproduct makes sense, but it isn't a huge win, and it still leaves a mountain of waste Si. 

You can also look at this the other way round.  If all your O2 is a byproduct of the solar panel plant you will not launch many spaceships.

This is a minor optimization barely worth mentioning, not an example of an integrated manufacturing system using every part of the pig but the squeal.

[guckyfan]

You can also look at this the other way round.  If all your O2 is a byproduct of the solar panel plant you will not launch many spaceships.

This is a minor optimization barely worth mentioning, not an example of an integrated manufacturing system using every part of the pig but the squeal.

I am not looking at it that way. I am looking at LOX production. Si is just a sideproduct that may or may not be useful.

[DigitalMan]

Some comments from Dr. Zubrin below.  I'm not too concerned about ejecta as I'm assuming they'll eventually have prepared landing pads by the time a base exists.  And with on-orbit refueling, SpaceShip should be able to get back to Earth without lunar ISRU.

https://twitter.com/robert_zubrin/status/1178786132127039488

This has been posted before.  The issue is that he is focused on flags and footprints, something SpaceX isn't interested in.  Nobody uses rafts to transport cargo to a base.