Starship On-orbit refueling - Options and Discussion

[DanClemmensen]

On further reflection, I am even more convinced that all of the specialized refuelling hardware should be in the depot and the SSs should be at most minimally modified: the best part is no part.

All SS already have a mating connection for thrust, namely the big nine-meter ring that connects to the SH. The depot should implement the SH side of this connection beneath a disposable fairing at the nose. After the fairing is discarded the depot is a cylinder whose top surface looks like the top surface of an SH. The depot will dock nose-to-tail with the SS. The fuel transfer will be via an extendable QD mechanism that is stored beneath the mating ring and is also covered by the disposable fairing. It extends out, up, and around to mate with the SS QD connection. Fuel transfer is done while under a small amount of thrust provided by the depot. Exactly how docking and locking will be accomplished will require actual engineering instead of hand-waving and will depend on the current design of the existing interface. Worst case: use a nine-meter version of the "soft docking" hardware from the IDSS that provides six degrees of freedom through a limited range of motion to complete the soft docking. I strongly suspect this is overkill and real engineers will come up with a simpler system.

I don't think there's a prayer of scaling up an IDSS-like soft-capture ring to 9m.  You'd have soft-capture petals covering up the engines on the tail.  The alignment, translational, and rotational velocity errors would also have to be much, much smaller than those supported by the existing IDSS.  And getting all the hard-capture latches to engage would be harder, too.

That's a lot of force on a docking system.  So any hard-capture probably has to be able to align the ships to use the launch load points.

I should not have mentioned IDSS: it was intended as an analogy.  It's an example of what some CNC folks call a "hexapod" design. You use six actuators to get six degrees of freedom. There is no effective limit to the size and strength at which they can be implemented except for mass. This SCS does not use a upscaled set of IDSS petals. It uses the same hard aligment system used for the SH/SS mateing.

We know that you can mate an SS with an SH: we have seen it done by the chopsticks.

Like the IDSS, the hexapod soft capture system (SCS) only corrects small errors. The depot must maneuver very carefully to get into the capture window of the SCS. velocities are measured in millimeters per second. This is just like the tolerances for an IDSS docking or for the chopsticks mating the SS and SH.

Yes, the system is intended to use the launch load pins. It's basically doing the same thing that is done to mount the SS atop an SH at the pad.

[Greg Hullender]

Maybe they can do without additional bonding points and I fully expect this to be way they'll first try it. My gut says that between the changing CoM and the unpredictable impulse from fluid flowing into the receiving tanks, its dicy. A slow transfer rate will keep fluid flow more controlled but this might not be viable when moving from proof of concept to operations.

If they're not rigidly bonded together, why does center of mass matter? Each one would be adjusting thrust so as to have the same acceleration (as near as possible) and to stay close enough not to put tension on the hoses, but that's not dependent on the center of mass of the combined system. Are you thinking of some other factor that would change the center of mass of each vehicle separately? Also, if we're talking about an ullage burn at 1 milli-g, it's not like the system needs microsecond response times.

[OTV Booster]

On further reflection, I am even more convinced that all of the specialized refuelling hardware should be in the depot and the SSs should be at most minimally modified: the best part is no part.

All SS already have a mating connection for thrust, namely the big nine-meter ring that connects to the SH. The depot should implement the SH side of this connection beneath a disposable fairing at the nose. After the fairing is discarded the depot is a cylinder whose top surface looks like the top surface of an SH. The depot will dock nose-to-tail with the SS. The fuel transfer will be via an extendable QD mechanism that is stored beneath the mating ring and is also covered by the disposable fairing. It extends out, up, and around to mate with the SS QD connection. Fuel transfer is done while under a small amount of thrust provided by the depot. Exactly how docking and locking will be accomplished will require actual engineering instead of hand-waving and will depend on the current design of the existing interface. Worst case: use a nine-meter version of the "soft docking" hardware from the IDSS that provides six degrees of freedom through a limited range of motion to complete the soft docking. I strongly suspect this is overkill and real engineers will come up with a simpler system.

LOL. Dan, what you describe is except for the orientation of one ship, the original SX plan. Butt to butt instead of a fore end adapter, but otherwise the same.


I think the last Elon comment on the matter said side by side. It's been a while and it's hit its 'use by' date, but it's all we have.


Here's what I expect is happening. A preliminary design study has been done and has not been looked at in months. They have more immediate things to deal with. Like us, they noodle these ideas among themselves but with tighter assumptions. Like us, they wish they had better numbers. When they have a solid (sub) orbital flight under their belt all that work will be gone over with better numbers, and maybe they stick with side to side or they pivot in another direction.


Till then, side by side is what we got. Oh, to be a fly in the SX wall.

[TheRadicalModerate]

Yes, lots of complex plumbing, but It's all in the depot. That's the whole idea. Move the entire design problem into the depot so all the other SS variants are minimally affected. Although this configuration can probably be used as a pusher, that's a side effect, not a design goal. Yes, this depot is a whole lot more complicated than a simple tank, because it must implement the entire fuel transfer system with no change to the other SS.

The tradeoff for this is that you can't send an EDL-capable tanker to cislunar for refueling ops.

If you're going to be doing enough stuff in cislunar to support a depot, then that's fine; the tanker boosts up to NRHO or LLO or wherever, transfers the prop to the depot, and goes home.  So the extra cost is that you've got a depot in cislunar that's never coming home, and you have ops costs associated with it long-term.

That may be a perfectly reasonable trade--and it also gives you the operational flexibility to ship the prop to cislunar long before the crewed mission.

However, note that you don't need a depot for an Option B LSS refueling, because for all intents and purposes the LSS likely has all the same boiloff management tech on it in the first place.  And you don't need it if you've progressed to the point where you're landing crews with an EDLC-LSS, because it's either refueled in HEEO or it's pushed to HEEO energies before immediately burning the rest of the way to TLI.  It'll have enough prop on board to return to direct EDL without refueling in cislunar.

The cases where a cislunar depot is interesting are the medium-term ones where you have a (non-EDLC) LSS that's staged with crew from LEO, and then returns the crew after the mission to LEO, propulsively, where they transfer to a D2 for EDL.  All of those cases require cislunar refueling, and a depot is handy to get your ducks in a row before committing a crew to a mission.

The fuel transfer QD is separate from the depot's standard QD, which is still there at the tail. I'm not a rocket engineer, so I don't know if you run the pipes up the outside of the depot or if you run them inside the tanks, but they extend from the "bottom" to the "top" to connect the tanks to the refuelling QD. Since the depot does not EDL, they can probably be on the outside.

So this is actually an important question that I'd like to pose for those of you who understand hypersonic flow:  How big a bump / chine / raceway / whateveryoucallit can you put on the dorsal surface of a Starship without having to substantially rework the max-q launch control system?

Quite a ways up-thread, I tried to draw a picture of a depot "kit" that was a payload, which could get deployed on the outside of any Starship, turning it into the active side of a depot.  It had in it:

1) Two active grapples, spaced as far apart as possible, to provide berthing.
2) The male-male adapter that allowed two female QD's to interconnect.
3) A cryocooler, which tapped off boiloff from the depot, via the QD, to re-liquify.
4) A big bag o' solar panels, which provided the power for the cryocooler and anything else that required extra power for long loiter.
5) Radiators for the cryocooler.

The kludgy part was that, because it lived inside the payload bay, it sorta had to crawl out on the dorsal side during deployment and make its way down to the tail to hook onto the QD.  If you can package all that stuff into a chine that launches on the outside of any Starship, then you have almost all of the advantages of the "kit" with a considerably less far-fetched deployment strategy.

If you can do this so that it runs all the way from the tail of the depot to the top of the cylindrical portion of the payload bay, then minor extensions of this idea could work for a pusher as well.

Extra credit would be if the kit could survive EDL, but fallback strategies where you'd still come out ahead would be:

a) Attach it only to depots, i.e., ships that can't do EDL.
b) Jettison the whole thing before EDL.

But you have to be able to cram all of that stuff into something that can survive all the flow regimes without having a boo-boo.

[OTV Booster]

On further reflection, I am even more convinced that all of the specialized refuelling hardware should be in the depot and the SSs should be at most minimally modified: the best part is no part.

All SS already have a mating connection for thrust, namely the big nine-meter ring that connects to the SH. The depot should implement the SH side of this connection beneath a disposable fairing at the nose. After the fairing is discarded the depot is a cylinder whose top surface looks like the top surface of an SH. The depot will dock nose-to-tail with the SS. The fuel transfer will be via an extendable QD mechanism that is stored beneath the mating ring and is also covered by the disposable fairing. It extends out, up, and around to mate with the SS QD connection. Fuel transfer is done while under a small amount of thrust provided by the depot. Exactly how docking and locking will be accomplished will require actual engineering instead of hand-waving and will depend on the current design of the existing interface. Worst case: use a nine-meter version of the "soft docking" hardware from the IDSS that provides six degrees of freedom through a limited range of motion to complete the soft docking. I strongly suspect this is overkill and real engineers will come up with a simpler system.

Doesn't this result in a lot more plumbing, though? (At least, on the depot.) I've been assuming that the depot is little more than a Starship with extended tanks, meaning it fills/drains through a QD port at the bottom. That port would get replaced with a gender-swapped hose of some kind (I'm fuzzy on exactly how this would work, but something a lot like the hoses in the existing ground support equipment that fills them up on the pad) but the interior of the depot would be almost identical to that of a Starship. That means they have to be side-by-side to refuel OR the hose needs to be really long, which (I'm told) introduces a lot of problems.

I do kind of like the idea of the depot securely attaching to the vehicles it refuels. I'm just wondering how the plumbing would work.

The approach that makes the most sense to me is to modify the QD on the depot to be an exact copy of the GSE QD plate on the tower. This means it can mate with any ship just as the towers GSE QD can. Unfortunately, this means it can't interface with the tower GSE QD plate for its own launch. The proposed solution that I prefer is to mount an adapter on the towers GSE QD plate that will allow interface and remove it after launch. Adapter used for depot launches only. The extra adapter mass stays on the ground.


Some think that it would work 'as is' and some think the depot QD plate should extend some short distance out, fixed or actively extending, to keep a bit more space between the two ships. From this came discussion of additional bracing.


Some disagree with this architecture entirely. The opinions and views expressed above are mine and do not reflect yada, yada. 

[LMT]

...modify the QD on the depot to be an exact copy of the GSE QD plate on the tower. This means it can mate with any ship just as the towers GSE QD can.

...mount an adapter on the towers GSE QD plate that will allow interface and remove it after launch. Adapter used for depot launches only.

...some think the depot QD plate should extend some short distance out, fixed or actively extending, to keep a bit more space between the two ships.

Or separate and simplify a depot's two-fluid cryo connection, much as Eta Space has done with Cryo-Dock. 

They animate the depot atop an F9.  Auditioning?

A Starship depot adaptation would retract further, to pull hw out of slipstream exposure.

[OTV Booster]

Yes, lots of complex plumbing, but It's all in the depot. That's the whole idea. Move the entire design problem into the depot so all the other SS variants are minimally affected. Although this configuration can probably be used as a pusher, that's a side effect, not a design goal. Yes, this depot is a whole lot more complicated than a simple tank, because it must implement the entire fuel transfer system with no change to the other SS.

The tradeoff for this is that you can't send an EDL-capable tanker to cislunar for refueling ops.

If you're going to be doing enough stuff in cislunar to support a depot, then that's fine; the tanker boosts up to NRHO or LLO or wherever, transfers the prop to the depot, and goes home.  So the extra cost is that you've got a depot in cislunar that's never coming home, and you have ops costs associated with it long-term.

That may be a perfectly reasonable trade--and it also gives you the operational flexibility to ship the prop to cislunar long before the crewed mission.

However, note that you don't need a depot for an Option B LSS refueling, because for all intents and purposes the LSS likely has all the same boiloff management tech on it in the first place.  And you don't need it if you've progressed to the point where you're landing crews with an EDLC-LSS, because it's either refueled in HEEO or it's pushed to HEEO energies before immediately burning the rest of the way to TLI.  It'll have enough prop on board to return to direct EDL without refueling in cislunar.

The cases where a cislunar depot is interesting are the medium-term ones where you have a (non-EDLC) LSS that's staged with crew from LEO, and then returns the crew after the mission to LEO, propulsively, where they transfer to a D2 for EDL.  All of those cases require cislunar refueling, and a depot is handy to get your ducks in a row before committing a crew to a mission.

The fuel transfer QD is separate from the depot's standard QD, which is still there at the tail. I'm not a rocket engineer, so I don't know if you run the pipes up the outside of the depot or if you run them inside the tanks, but they extend from the "bottom" to the "top" to connect the tanks to the refuelling QD. Since the depot does not EDL, they can probably be on the outside.

So this is actually an important question that I'd like to pose for those of you who understand hypersonic flow:  How big a bump / chine / raceway / whateveryoucallit can you put on the dorsal surface of a Starship without having to substantially rework the max-q launch control system?

Quite a ways up-thread, I tried to draw a picture of a depot "kit" that was a payload, which could get deployed on the outside of any Starship, turning it into the active side of a depot.  It had in it:

1) Two active grapples, spaced as far apart as possible, to provide berthing.
2) The male-male adapter that allowed two female QD's to interconnect.
3) A cryocooler, which tapped off boiloff from the depot, via the QD, to re-liquify.
4) A big bag o' solar panels, which provided the power for the cryocooler and anything else that required extra power for long loiter.
5) Radiators for the cryocooler.

The kludgy part was that, because it lived inside the payload bay, it sorta had to crawl out on the dorsal side during deployment and make its way down to the tail to hook onto the QD.  If you can package all that stuff into a chine that launches on the outside of any Starship, then you have almost all of the advantages of the "kit" with a considerably less far-fetched deployment strategy.

If you can do this so that it runs all the way from the tail of the depot to the top of the cylindrical portion of the payload bay, then minor extensions of this idea could work for a pusher as well.

Extra credit would be if the kit could survive EDL, but fallback strategies where you'd still come out ahead would be:

a) Attach it only to depots, i.e., ships that can't do EDL.
b) Jettison the whole thing before EDL.

But you have to be able to cram all of that stuff into something that can survive all the flow regimes without having a boo-boo.

Any maxQ problems would depend on the size of the whatever. Any imbalance, up to some unknown point, could be overcome by thrust vectoring or differential thrust. There's a lot of propellant flexibility because in worst case the depot would reach orbit with no propellant of its own to kick off the campaign.


Why not build depots to stay in orbit? The only downside is a ship that can have no other use and may have a short operational life. The lifetime will only become more robust as the SS matures. After all, a major design goal is fast turnaround. With only one launch and atmospheric transit, combined with non coking engines, a one campaign lifetime at the beginning will eventually extend to more as the design matures.


They're planning on turning these things out like sausages. A few special builds that are force multipliers and never come back will have little impact on production. A self deploying universal conversion kit is more of a complication than a help. If a depot is needed cislunar, send a depot. Let it stay there. If it lives long enough it'll be there for future use. If not, well, it did its job. The next one will last longer.


There is the issue of need. Consensus is that space traffic will only get heavier. Having a depot on orbit, LEO, cislunar or elsewhere, incurs little expense. In exchange, it offers flexibility and opportunity. Infrastructure of any sort is a force multiplier.

[DanClemmensen]

If you're going to be doing enough stuff in cislunar to support a depot, then that's fine; the tanker boosts up to NRHO or LLO or wherever, transfers the prop to the depot, and goes home.  So the extra cost is that you've got a depot in cislunar that's never coming home, and you have ops costs associated with it long-term.

That may be a perfectly reasonable trade--and it also gives you the operational flexibility to ship the prop to cislunar long before the crewed mission.

The cases where a cislunar depot is interesting are the medium-term ones where you have a (non-EDLC) LSS that's staged with crew from LEO, and then returns the crew after the mission to LEO, propulsively, where they transfer to a D2 for EDL.  All of those cases require cislunar refueling, and a depot is handy to get your ducks in a row before committing a crew to a mission.

Even though this depot is fairly complicated, it's still not a huge expense. Just build it and park it in cislunar space forever. Even if depots are not yet in series production to support Mars, building a second one alongside the Artemis one is not a large extra expense.

The fuel transfer QD is separate from the depot's standard QD, which is still there at the tail. I'm not a rocket engineer, so I don't know if you run the pipes up the outside of the depot or if you run them inside the tanks, but they extend from the "bottom" to the "top" to connect the tanks to the refuelling QD. Since the depot does not EDL, they can probably be on the outside.

So this is actually an important question that I'd like to pose for those of you who understand hypersonic flow:  How big a bump / chine / raceway / whateveryoucallit can you put on the dorsal surface of a Starship without having to substantially rework the max-q launch control system?

For a non-EDL depot, these whatists should pose a lot less of a challenge than the EDL control surfaces on an EDL SS. They do not need to be on the dorsal side. I think you have two of them, one on each side, and they are aerodynamically identical even if the actual plumbing is different.

[OTV Booster]

...modify the QD on the depot to be an exact copy of the GSE QD plate on the tower. This means it can mate with any ship just as the towers GSE QD can.

...mount an adapter on the towers GSE QD plate that will allow interface and remove it after launch. Adapter used for depot launches only.

...some think the depot QD plate should extend some short distance out, fixed or actively extending, to keep a bit more space between the two ships.

Or separate and simplify a depot's two-fluid cryo connection, much as Eta Space has done with Cryo-Dock. 

They animate the depot atop an F9.  Auditioning?

A Starship depot adaptation would retract further, to pull hw out of slipstream exposure.

One plan calls for interconnecting ullage spaces for pressure equalization while pumps handle the transfer and another calls for venting the receiving ships ullage and letting the other ships ullage pressure force the transfer. N2 may be used used for a while despite the goal of getting rid of it. And argument can be made for power interconnection. Some advocate for an 'in line' arrangement as you show. It can handle more connections if necessary. One school of thought is that whatever is used for transfer should interface with the connections that are already there.


The cryo-dock is targeted at transfers quite a bit smaller than what SX is looking at. I'm not sure what scaling issues might be driving the design considerations. Maybe none at all.


I'm picturing a maybe half meter deep doghouse on the depot, and maybe all ships, with a roll up door that drops into place to protect the QD plate after launch disconnect. The depot QD would extend maybe another half meter out of the doghouse when it gets to work.

[LMT]

N2 may be used used for a while despite the goal of getting rid of it.

SpaceX on-orbit cryo demo targets "autogenous pressurization and propellant transfer".

N2 isn't mentioned in the NASA doc.

[OTV Booster]

N2 may be used used for a while despite the goal of getting rid of it.

SpaceX on-orbit cryo demo targets "autogenous pressurization and propellant transfer".

N2 isn't mentioned in the NASA doc.

That looks like 12/21, 10 months ago. It states aspirations and most probably contract requirements. Some aspirations may not be contract requirements. N2 isn't mentioned either way. It has uses other than ullage so as I said, there is a 'use by' date on public SX aspirations. We have surmises but less than perfect knowledge. It'll be interesting no matter what they do.

[redneck]

https://selenianboondocks.com/2008/01/an-insane-but-interesting-idea-fleet-launched-orbital-craft/

If  you think refueling in space while under acceleration is tough, try this one from Jon Goff.

[TheRadicalModerate]

https://selenianboondocks.com/2008/01/an-insane-but-interesting-idea-fleet-launched-orbital-craft/

If  you think refueling in space while under acceleration is tough, try this one from Jon Goff.

That's not as insane as refueling under acceleration, and it's not nearly as insane as refueling under acceleration without rigid, load-bearing couplings.

[redneck]

https://selenianboondocks.com/2008/01/an-insane-but-interesting-idea-fleet-launched-orbital-craft/

If  you think refueling in space while under acceleration is tough, try this one from Jon Goff.

That's not as insane as refueling under acceleration, and it's not nearly as insane as refueling under acceleration without rigid, load-bearing couplings.

And yet if you think about it, the Shuttle orbiter was 'refueling' from the external tank on every launch.

[OTV Booster]

https://selenianboondocks.com/2008/01/an-insane-but-interesting-idea-fleet-launched-orbital-craft/

If  you think refueling in space while under acceleration is tough, try this one from Jon Goff.

That's not as insane as refueling under acceleration, and it's not nearly as insane as refueling under acceleration without rigid, load-bearing couplings.

And yet if you think about it, the Shuttle orbiter was 'refueling' from the external tank on every launch.

Shuttle was rigidly mounted to the ET. The connector and structural mating was done on the ground. For the tank, it was a one night stand. Spending a hundred tech hours (a total WAG) on shuttle side inspection and refurb would not even have been a rounding error on shuttle turnaround. Other than that...

[TheRadicalModerate]

https://selenianboondocks.com/2008/01/an-insane-but-interesting-idea-fleet-launched-orbital-craft/

If  you think refueling in space while under acceleration is tough, try this one from Jon Goff.

That's not as insane as refueling under acceleration, and it's not nearly as insane as refueling under acceleration without rigid, load-bearing couplings.

And yet if you think about it, the Shuttle orbiter was 'refueling' from the external tank on every launch.

Yeah, and that was, if not insane, at least kinda dumb, as it turned out.  More importantly for this discussion:  It required a large team of humans in a shirtsleeve environment to align the bipods, connect them, tension them, and--last but hardly least--arm their pryrotechnic frangible bolts.  That's a silly--and likely impossible--design for in-space automated operations.

Let's make sure we're all talking about the same things here.  In terms of accelerating while refueling and docked / berthed / tethered / formation-flying / whatever, we have the following:

1) Two spacecraft boosting independently to the same HEEO so that RPOD (rendezvous, proximity ops, and docking) is relatively quick once in the HEEO, allowing refueling that should take less than an orbit if everything goes right.  This is what we've mostly talked about.  No coupled acceleration involved. 

Note that the most general case of this is what we've been calling "laddering", which involves n ships refueling m ships in orbit 1, then the m ships refueling p ships in orbit 2, etc. It's similar to the thing Jon Goff was describing.  It can make sense for achieving extremely high delta-v, but its practical utility for cislunar or even martian ops is effectively zero.

2) Coupled ullage accelerations, on the order of 1mm/s˛ - 1cm/s˛.  Most on-orbit refueling schemes require this to settle the prop, and if you're doing side-by-side refueling, the accelerations need to be coordinated on the two vehicles, or transmitted from one vehicle to another via something able to take loads of a few newtons.  I assert that this requires at least semi-rigid docking, although this is a debatable point.

3) Coordinated boost accelerations while connected, anywhere from 3-12m/s˛.  This is insane for a side-by-side configuration that's docked on orbit.  It's doubly insane because it's unnecessary.  Prop consumption for boosting two ships to an HEEO in a coupled system is the same as prop consumption for boosting the two ships separately.ą  If your objection here is, "But what about cross-feed?" the only thing cross-feed would get you is higher thrust, which will only save you single-digit m/s to make an orbital burn slightly more impulsive.

4) Pusher configurations, where the tanker/depot acts as a first stage for the target Starship, then separates, allowing the target Starship to continue the burn, with the whole staging process costing low-single-digit m/s in non-impulsive losses.  This is not a case of accelerating while refueling; it's just a staging operation.  The pushing depot/tanker may or may not have previously refueled the target Starship, but that's really a separate operation.

#1, using #2 to transfer prop, is a viable option, as is #4, which can either transfer prop using ullage acceleration from the pusher, or where the pusher and target can separately use #2 in VLEO from separate tankers/depots.

I propose we stop talking about #3.

___________
ąProof:

A particular delta-v requires a fixed mass ratio MR. 
MR = exp( Δv / Isp / 9.8 ) = (inert + propInitial) / (inert + propFinal)
This is just the rocket equation.

We want to find the prop consumed in executing the delta-v maneuver.

propFinal = propInitial - propConsumed
MR = (inert + propInitial) / (inert + propInitial - propConsumed)
MR(inert + propInitial) - MR*propConsumed = inert + propInitial
MR*propConsumed = (inert + propInitial)(MR - 1)
propConsumed = (inert + propInitial)(1 - 1/MR)

Now suppose we have two spacecraft, with masses inert1+propInitial1 and inert2+propInitial2.  It should now be obvious that the prop consumption is the same whether they're boosted separately or together:

(inert1 + propInitial1)(1 - 1/MR) + (inert2 + propInitial2)(1 - 1/MR) =
(inert1 + propInitial1 + inert2 + propInitial2)(1 - 1/MR)

Remember, MR is a constant, dependent solely on the delta-v needed to reach the refueling orbit and the Isp.  Things would get weird if you had two different Isp's, but that's not what we're talking about here.

[Twark_Main]

You can argue, possibly successfully, that I've over-constrained the problem.  If you're willing to do two refuelings on the LSS, one in VLEO and one in HEEO, you can boost to pretty much any energy you want.  But NASA made a fairly big deal about only requiring one refueling against the LSS itself.  This actually came up in the BO lawsuit, where NASA responded to their complaint on this topic.  So going back on that opens up a can of worms that would be better left closed.

Frankly, this is why I like the 1500t LSS.  All of this complexity falls away, and all refueling is done in the lowest-possible energy orbit.

Ironically, with my harmonic refueling it also happens in the lowest "possible" orbit too. That's the driving principle behind the optimization, in fact! 

I believe that laddering makes sense when extreme C3 outweighs the need for mission simplicity.  However, for both lunar and Mars missions, that's simply not the case.  In the case of an Artemis mission, you'll want to limit the LSS to a single refueling, simply to reduce risk.  That requires limiting the maximum refueling energy to something that the LSS can reach with its launch prop.

I missed this particular bit of silliness the first time around.

If you are only 'permitted' one filling, then the math says that VLEO is the only sensible choice.

So essentially your argument is "if we assume no laddering, then there's no laddering." Imagine that... 

In all my laddering calculations I assume the vehicle is fully fueled in VLEO first (ie the sensible approach), so apparently we were talking/calculating past each-other the entire time.

[Twark_Main]

If a tanker had 12-hour turnaround, and it delivered 1/5 of a load, it could fill an inflatable depot for ~ 300 crews each synod, by itself. 

Otherwise, a fleet of tankers would be needed.  E.g., given a 14-day window and preloading of depot tankers at synod start, at least 46 tankers would be needed. 

Scale up ~ 10x for settlement cargo flights.

The obvious idea is to use the Starships themselves as the "depot." This eliminates the entire depot fleet, which is a massive savings.

Instead you just transfer the passengers at the last minute. You can use 500 passenger P2P-derived vehicles to do that. This further increases the efficiency of launching within that 14-day window, because each launch serves five Starships instead of one.

Delete depots, get taxis.

No, you wouldn't want to haul all depot hw (ruggedized ZBO and propellant transfer hw) on every ship, to Mars and back

Maybe, maybe not. It depends on how much mass.

Many people seem to be [over-]engineering elaborate complex systems for both. This is a mistake IMO.

[Twark_Main]

Since both mission architectures (boosting the depot vs. boosting the tanker) are readily available, it seems like we should at least compare and contrast the two options.

It's a good idea, but I fooled with it a bit and it doesn't really do anything meaningful.

Going to an EDL-capable tanker saves you some return delta-v, but it doesn't save you huge amounts of prop, because the tanker's wet mass is small after it's transferred the prop to the LSS.

Which is cheaper than developing a custom depot stuck in Lunar orbit.

Who said anything about a custom depot?  If you need prop in cislunar and your logistics work out, you send a regular lift tanker, refuel, and the tanker returns to EDL.

So which one is it, TheRadicalModerate?

[Twark_Main]

1) Option A and B missions have to be as simple as possible, and the LSS itself can't be refueled more than once per mission, nor can it be refueled with crew on board.

This is a silly constraint.  By the time a crew certified LSS leaves LEO there will be there will be an order of magnitude more experience refueling than launching SLS or reentering Orion or a great many other parts of the planned mission.  If that's not enough refueling experience, you can get a lot more fairly cheaply and quickly.  Since refueling is an essential part of any BLEO use of Starship you have to do that eventually, so the best plan is to expedite testing refueling and not place unneeded constraints on the eventual mission.

It's a fair criticism, and one I expected.  Here's my reasoning:

1) Pez-dispenser Starships are probably unsuitable for refueling RPOD and transfer tests, so SpaceX is limited to however much hardware they decide to take out of Starlink service to perform the development.

2) Things will be more schedule-bound than you think.  They always are.  I'd also expect depot and refueling R&D to lag launch, EDL, and re-use R&D, simply because refueling testing is expensive without reusability being fairly far along.  And NASA is going to be fairly impatient with delays.

Delays with Artemis??? Inconceivable!

I don't see anyone nipping at SpaceX's heels, so NASA can get as "impatient" as they want. There's no faster alternative.

3) An LSS will be close to an order of magnitude more expensive than a vanilla Starship, simply because it has to be crew-certified.  It's an asset for which you want to aggressively minimize the risks you can.  Limiting LSS to a single refueling is a simple way to minimize risks.

It's a "simple" to be sure. But what really matters is the cost/benefit tradeoff.

4) NASA made a fairly big deal about the refueling complexity in the source selection statement, and the Blue Origin lawsuit picked up on that, forcing NASA to defend the conops as having almost all refueling off the critical path.  I'd expect that anything that reduced that complexity has very nice CYA properties for NASA.

I prefer when rocket scientists design rockets, not politicians.

Requiring the LSS to have multiple refuelings clearly isn't a complete deal-breaker.

Given that admission, it's weird that you said you "need" only one refilling about eight times...