Author Topic: Starship On-orbit refueling - Options and Discussion  (Read 811640 times)

Offline TheRadicalModerate

  • Senior Member
  • *****
  • Posts: 5130
  • Tampa, FL
  • Liked: 3753
  • Likes Given: 702
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2800 on: 12/15/2024 05:33 am »
The primary benefit for VLEO in regards to MMODs is the relative low density.  They just don't last very long, so there's no accumulation.
In circular orbits, yes. I mentioned that. Density is pretty low up to almost 400 km. But debris after a collision will include particles in elliptical orbits that spend only a little time at 200 km. Yes, those will circularize fairly quickly and then decay, but they're still an issue.

Neither of us is reasoning about this in a very statistical fashion (because statistics are hard and boring), but it's important to put the probabilities in perspective:

1) Depots have a high MMOD risk because they stay in orbit for extremely long periods, and they're large.  Of these two factors, the long period is the biggest problem.

2) Depots have a moderate probability of generating MMOD via explosions, but a single MMOD event, especially in VLEO, doesn't substantially raise the risk to other vehicles.  And, if they're in VLEO, any MMOD that's generated deorbits in a short time (where "short" is less than a small number of years).

3) Small pieces of MMOD generated in VLEO have the lowest ballistic coefficients, and deorbit extremely quickly (I don't know the exact number, but even eccentric pieces probability can't survive for more than weeks.)  Big pieces, with higher BCs, are trackable.

4) Tankers have a somewhat elevated risk, but only because they have a lot of cross-sectional area.  Risk is still fairly low for the amount of time they spend in orbit.

Space is still big.

Reason #1 is by far the biggest problem for depots, and may indeed require putting shielding around the tankage.  Even a pinhole in a tank is a loss of mission--and potentially the loss of a bunch of previous tanker missions.

Offline Narnianknight

  • Member
  • Posts: 39
  • Ohio
  • Liked: 34
  • Likes Given: 155
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2801 on: 12/15/2024 08:47 pm »

Maybe a Depot is a bad requirement.  Let's remove it and see what happens.

"Maybe [my favorite opinion this week is right]" isn't a real argument, I'll have you know.  ::)

If you want to claim that it is actually a bad requirement, you have all your work ahead of you.


No, the burden of proof is on the one who wants to keep the part (in this case, a custom depot Starship).

Elon isn't being flippant when he says the requirements are probably wrong.  30 years of software engineering tells me the same thing.


As for the whole "burden of proof" silliness, attempting to seize the high ground Null Hypothesis isn't a real argument, it's an epistemological sleight-of-hand trick.  ::)

Nobody is suggesting Elon is being flippant about requirements. I'm just observing that your idea of the requirements can be wrong too. Hence why the quote "cuts both ways."  ;)

Assuming "the best part is no part" is to be taken seriously, the pro-depot position does have the burden of proof. Shifting burden of proof is not invalid here; I will take the burden.

With the miscellaneous rhetoric issues out of the way, as far as actual engineering we come back to:


Quote from: Elon Musk
Everything possible must be done to ensure astronaut safety.

Elon isn't being flippant when he talks about astronaut safety.    ???

That's valid if there are crew onboard, but not a very good argument for depots overall, considering ships technically would never have to be crewed during refueling. Anyway, "let's remove it and see what happens."

1. Some money and development is saved (that's about where the advantages end).
2. Either boil-off protection is placed on the final ship (sometimes extra mass) or the prop just boils off.
3. Either the final ship or all the tankers have to have a male QD port (have fun with the GSE).
4. The final ship has to be launched at the beginning of the refueling campaign instead of the end.
    4a. Everything onboard has to be rated for an extended loitering lifetime.
    4b. Crew during refueling is absolutely prohibitive now.
    4c. Final ship outfitting, testing, and revising cannot continue throughout the refueling campaign (not good for timelines).
    4d. Months-long MMOD considerations are now relevant to the expensive final ship rather than to a barrel of prop.

MMOD risk to depots between refueling campaigns doesn't even move the needle compared to these disadvantages.

Offline TheRadicalModerate

  • Senior Member
  • *****
  • Posts: 5130
  • Tampa, FL
  • Liked: 3753
  • Likes Given: 702
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2802 on: 12/15/2024 10:36 pm »
1. Some money and development is saved (that's about where the advantages end).
2. Either boil-off protection is placed on the final ship (sometimes extra mass) or the prop just boils off.
3. Either the final ship or all the tankers have to have a male QD port (have fun with the GSE).
4. The final ship has to be launched at the beginning of the refueling campaign instead of the end.
    4a. Everything onboard has to be rated for an extended loitering lifetime.
    4b. Crew during refueling is absolutely prohibitive now.
    4c. Final ship outfitting, testing, and revising cannot continue throughout the refueling campaign (not good for timelines).
    4d. Months-long MMOD considerations are now relevant to the expensive final ship rather than to a barrel of prop.

MMOD risk to depots between refueling campaigns doesn't even move the needle compared to these disadvantages.

The HLS Starship and the depot may have more in common that the HLS Starship and an EDL-capable Starship.  In effect, HLS may be more accurately described as a depot with landing legs, thrusters, and crew module, than as a EDL-capable Starship without TPS and elonerons, plus landing legs and landing thrusters.

The reason I don't think that tankers will directly fuel HLS is because HLS is a quite expensive piece of crew-rated hardware, while a depot is pretty cheap.  It makes sense to minimize even the smallest of risks to the HLS--and I don't believe that prolonged MMOD risk in LEO is a small risk by any means.  And it's not just a monetary risk:  Even if you don't get an MMOD strike on a tank, which is an immediately obvious mission failure, I'd worry more about an undetected MMOD strike somewhere else, which could compromise crew safety once the lunar surface mission started.

Offline InterestedEngineer

  • Senior Member
  • *****
  • Posts: 2969
  • Seattle
  • Liked: 2254
  • Likes Given: 3705
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2803 on: 12/15/2024 11:07 pm »

1. Some money and development is saved (that's about where the advantages end).


The reason I brought up this tired old argument is that is not where the advantages end.

It would appear that hot gas thrusters are a requirement for docking with a very heavy depot, if I understood the math above correctly.

It's the "we have to change the mission and equipment with x, y an z" for the depot that is the problem.

Requirements drag a whole pile of consequences with them, which have their own requirements and their own consequences.

Which is why you try to get rid of requirements, as then you get rid of all sorts of down stream consequences of those requirements.

The requirement to transfer 2000t of fuel in one go is one of those things the depot drags along with it.    If you don't have depot, you don't have this requirement.

So let's enumerate all the problems with transferring 2000t of fuel in one go, instead of just ignoring it like the quote poster did.

Are there any other requirements that go away when the depot is eliminated from the plan?

Offline Greg Hullender

  • Full Member
  • ****
  • Posts: 793
  • Seattle
    • Rocket Stack Rank
  • Liked: 583
  • Likes Given: 409
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2804 on: 12/16/2024 12:38 am »
The requirement to transfer 2000t of fuel in one go is one of those things the depot drags along with it.    If you don't have depot, you don't have this requirement.

So let's enumerate all the problems with transferring 2000t of fuel in one go, instead of just ignoring it like the quote poster did.
By this I gather you mean that one cost of using a depot is that each liter of propellant gets transferred twice: once to the depot and once to the vehicle. Without the depot, each liter is transferred only once. Since the transfer process has some inefficiencies, this roughly doubles them.

I don't see any other issue with this, though. Is there something else you're getting at?

Offline InterestedEngineer

  • Senior Member
  • *****
  • Posts: 2969
  • Seattle
  • Liked: 2254
  • Likes Given: 3705
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2805 on: 12/16/2024 12:49 am »
The requirement to transfer 2000t of fuel in one go is one of those things the depot drags along with it.    If you don't have depot, you don't have this requirement.

So let's enumerate all the problems with transferring 2000t of fuel in one go, instead of just ignoring it like the quote poster did.
By this I gather you mean that one cost of using a depot is that each liter of propellant gets transferred twice: once to the depot and once to the vehicle. Without the depot, each liter is transferred only once. Since the transfer process has some inefficiencies, this roughly doubles them.

I don't see any other issue with this, though. Is there something else you're getting at?

The length of time it takes to transfer a whole tank in one shot vs. the a bunch of small transfers (which you already have to do)

The added risks of all fuel eggs in one basket while doing a risky maneuver.

I think depot drags along the requirement for hot gas thrusters.

Offline Greg Hullender

  • Full Member
  • ****
  • Posts: 793
  • Seattle
    • Rocket Stack Rank
  • Liked: 583
  • Likes Given: 409
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2806 on: 12/16/2024 02:16 am »
The requirement to transfer 2000t of fuel in one go is one of those things the depot drags along with it.    If you don't have depot, you don't have this requirement.

So let's enumerate all the problems with transferring 2000t of fuel in one go, instead of just ignoring it like the quote poster did.
By this I gather you mean that one cost of using a depot is that each liter of propellant gets transferred twice: once to the depot and once to the vehicle. Without the depot, each liter is transferred only once. Since the transfer process has some inefficiencies, this roughly doubles them.

I don't see any other issue with this, though. Is there something else you're getting at?

The length of time it takes to transfer a whole tank in one shot vs. the a bunch of small transfers (which you already have to do)

The added risks of all fuel eggs in one basket while doing a risky maneuver.

I think depot drags along the requirement for hot gas thrusters.
Why would it be better to make multiple small transfers?

Where does the requirement for hot-gas thrusters come from? The duration of the ullage burn?

Online DanClemmensen

  • Senior Member
  • *****
  • Posts: 7446
  • Earth (currently)
  • Liked: 6053
  • Likes Given: 2532
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2807 on: 12/16/2024 02:28 am »
Sorry, but I seem to have lost track.
     Has anyone seen a stated requirement to transfer propellant from a Tanker to any ship other than a Depot?
     Has anyone seen a stated requirement to transfer propellant to any non-Depot from a ship other than a Depot?
     (exception: the transfer demo does not specifically call for a Depot.)

If all transfers involve a Depot, then no ship other than a Depot needs active transfer hardware. This simplifies all of the ships at the expense of a more complicated Depot.

Offline Greg Hullender

  • Full Member
  • ****
  • Posts: 793
  • Seattle
    • Rocket Stack Rank
  • Liked: 583
  • Likes Given: 409
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2808 on: 12/16/2024 02:43 am »
Sorry, but I seem to have lost track.
     Has anyone seen a stated requirement to transfer propellant from a Tanker to any ship other than a Depot?
     Has anyone seen a stated requirement to transfer propellant to any non-Depot from a ship other than a Depot?
Nope. The thought experiment is to ask "what happens if we remove the depot requirement?" After all, it wasn't that long ago that Elon declared "depots suck" (or something to that effect).

Online DanClemmensen

  • Senior Member
  • *****
  • Posts: 7446
  • Earth (currently)
  • Liked: 6053
  • Likes Given: 2532
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2809 on: 12/16/2024 02:47 am »
Sorry, but I seem to have lost track.
     Has anyone seen a stated requirement to transfer propellant from a Tanker to any ship other than a Depot?
     Has anyone seen a stated requirement to transfer propellant to any non-Depot from a ship other than a Depot?
Nope. The thought experiment is to ask "what happens if we remove the depot requirement?" After all, it wasn't that long ago that Elon declared "depots suck" (or something to that effect).
One clear consequence is that at least one other ship type, presumably Tanker, will need active transfer hardware. This reduces the amount of prop a Tanker can lift to orbit.

Offline TheRadicalModerate

  • Senior Member
  • *****
  • Posts: 5130
  • Tampa, FL
  • Liked: 3753
  • Likes Given: 702
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2810 on: 12/16/2024 05:30 am »
I think depot drags along the requirement for hot gas thrusters.

Landing on the Moon drags along the requirement for combusting gas thrusters.  So does doing a variety of maneuvers where the main tanks might not be pressurized to the needs of the maneuver.

There's certainly a possibility that settling thrust is the only case where you'd need an extremely low thrust combusting gas thruster.  But there are all kinds of cases where you need immediate attitude control deep into a transfer orbit, where your ullage gas may have collapsed.  Long-duration spaceflight comes with all kinds of interesting requirements, not all of which are dumb.

Offline andrewmcleod

  • Member
  • Posts: 35
  • Yorkshire Dales
  • Liked: 48
  • Likes Given: 54
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2811 on: 12/16/2024 09:12 am »
[...]
Reason #1 is by far the biggest problem for depots, and may indeed require putting shielding around the tankage.  Even a pinhole in a tank is a loss of mission--and potentially the loss of a bunch of previous tanker missions.

Is that true? That Progress had a reasonable-sized hole in it for ages without the ISS suddenly depressurizing...

Let's throw some poorly-considered numbers at a random formula on Wikipedia...

First, to estimate the density of methane at its (sea-level) boiling point of ~110K and a pressure of 6 bar, using the ideal gas equation you get a density of 10.52 kg/m^3.

https://en.wikipedia.org/wiki/Orifice_plate
Then, if I plug in some inappropriate values to this equation for a 1mm^2 hole leaking methane *gas* (discharge coefficient of 0.85, methane density of 10.52 kg/m^3, expansibility factor 1 even though that's inappropriate for a compressible gas, pressure 6 bar and pressure differential of 1) then I get a mass loss rate of 260kg/day. If I used a more accurate expansibility factor (which I think is less than 1 for compressible gases as the flow becomes choked), it would be even lower. Not ideal, but not a total disaster if you are leaking gas rather than liquid.

If I use a density for liquid methane (420kg/m^3) then I get 1.6 tons per day which is rather more problematic.

Anything much more than 1mm^2 is obviously more problematic.

Major limitations:
1) I've used an expansibility factor of 1 which is inappropriate for gas (correct for liquid), but (I think) conservative
2) I've used an arbitrary but conservative discharge coefficient of 0.85
3) I've used a pressure of 6 bar; reducing pressure would reduce leakage rates (but only as the root of the pressure)
4) I've used a sea-level boiling point of 110K instead of the actual boiling point at 6 bar (and assumed this is the temperature the methane will be at, which is likely)

So a pinhole is not necessarily completely disastrous (a 0.1mm x 0.1mm pinhole would only be 16 kg/day, or ~6 tons per year, and you'd probably never notice it) but larger holes (>1mm^2) do become problematic quickly.

And obviously a hole might cause structural failure of the tank which would be Bad.

Offline OTV Booster

  • Senior Member
  • *****
  • Posts: 5477
  • Terra is my nation; currently Kansas
  • Liked: 3778
  • Likes Given: 6565
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2812 on: 12/16/2024 02:58 pm »
I found this paper from 2017 (I think) on evaluating MMOD risk. https://ntrs.nasa.gov/api/citations/20170010258/downloads/20170010258.pdf

The upshot is that the sample size is small, good numbers haven't been collected and the alignment between the numbers and models is weak.

Because of the diversity of sat builds the samples were limited to a couple of constellations plus one singleton and looked at batteries and tankage hits only. Even limiting the samples to mostly constellations had difficulties. One example of this was different sources specifying different tank wall thicknesses.

ISTM that limiting analysis to a single, large, well characterized constellation at various altitudes would give clearer numbers and better models.

Now, if only somebody had a constellation like this...
« Last Edit: 12/16/2024 02:58 pm by OTV Booster »
We are on the cusp of revolutionary access to space. One hallmark of a revolution is that there is a disjuncture through which projections do not work. The thread must be picked up anew and the tapestry of history woven with a fresh pattern.

Offline leovinus

  • Full Member
  • ****
  • Posts: 1261
  • Porto, Portugal
  • Liked: 993
  • Likes Given: 1931
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2813 on: 12/16/2024 03:46 pm »
By coincidence, I stumbled about the report N72-30800 IN-SPACE PROPELLANT LOGISTICS from 1972. PDF download via the little red icon on the top left of the page and attached as well. Sounds familiar? Like a depot? Probably.

Maybe a non SpaceX historical note on fuel depots, transfers et al is useful for the discussion. There are other volumes as well 2 Technical Report, Trade studies 3, Project Planning Data 4In-Space Propellant Systems Safety. Volume 1 Executive Summary.

Granted, materials and many things will have changed in the last 50 years but the laws of physics are the same. Conclusion 4 and 5 seem still relevant.

Offline Narnianknight

  • Member
  • Posts: 39
  • Ohio
  • Liked: 34
  • Likes Given: 155
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2814 on: 12/17/2024 06:01 pm »
In October 2020 up-thread, the best guess was that the necessary settling acceleration is on the order of 1e-4 m/s2. Also, keeping the depot pointing toward the sun to minimize illuminated surface area and therefore boil-off is probably important. Either way, using drag for the acceleration doesn't really make sense, because the ship would have to be brought up to altitude again using propulsion. Besides, boil-off will create ullage gas that will have to be vented anyway; they might as well use it for settling.

Correction: that was 1e-4 g, not m/s2, which is about 1 mm/s2.

Offline TheRadicalModerate

  • Senior Member
  • *****
  • Posts: 5130
  • Tampa, FL
  • Liked: 3753
  • Likes Given: 702
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2815 on: 12/17/2024 10:41 pm »
[...]
Reason #1 is by far the biggest problem for depots, and may indeed require putting shielding around the tankage.  Even a pinhole in a tank is a loss of mission--and potentially the loss of a bunch of previous tanker missions.

Is that true? That Progress had a reasonable-sized hole in it for ages without the ISS suddenly depressurizing...

Let's throw some poorly-considered numbers at a random formula on Wikipedia...

First, to estimate the density of methane at its (sea-level) boiling point of ~110K and a pressure of 6 bar, using the ideal gas equation you get a density of 10.52 kg/m^3.

https://en.wikipedia.org/wiki/Orifice_plate
Then, if I plug in some inappropriate values to this equation for a 1mm^2 hole leaking methane *gas* (discharge coefficient of 0.85, methane density of 10.52 kg/m^3, expansibility factor 1 even though that's inappropriate for a compressible gas, pressure 6 bar and pressure differential of 1) then I get a mass loss rate of 260kg/day. If I used a more accurate expansibility factor (which I think is less than 1 for compressible gases as the flow becomes choked), it would be even lower. Not ideal, but not a total disaster if you are leaking gas rather than liquid.

If I use a density for liquid methane (420kg/m^3) then I get 1.6 tons per day which is rather more problematic.

Anything much more than 1mm^2 is obviously more problematic.

Major limitations:
1) I've used an expansibility factor of 1 which is inappropriate for gas (correct for liquid), but (I think) conservative
2) I've used an arbitrary but conservative discharge coefficient of 0.85
3) I've used a pressure of 6 bar; reducing pressure would reduce leakage rates (but only as the root of the pressure)
4) I've used a sea-level boiling point of 110K instead of the actual boiling point at 6 bar (and assumed this is the temperature the methane will be at, which is likely)

So a pinhole is not necessarily completely disastrous (a 0.1mm x 0.1mm pinhole would only be 16 kg/day, or ~6 tons per year, and you'd probably never notice it) but larger holes (>1mm^2) do become problematic quickly.

And obviously a hole might cause structural failure of the tank which would be Bad.

Good catch (although I haven't verified your results).  Two things jump out:

1) It's non-intuitive why mass loss from ullage gas would be less than from liquid prop.  Gas has much lower viscosity.  Of course, there's only so much gas available, but if the pressure drops to near-zero, the prop will boil like crazy.  Did you take this into account?  It's possible that the pressure is still very low at equilibrium, but that's a much more complicated calculation.

2) If your hole opens up on a gas pocket, the thrust from the escaping gas will accelerate the vehicle in the opposite direction, which will cause liquid prop to settle around the hole.  Given the outcome of the boiling calculation, that may be a good thing.  Or not...

Offline Paul451

  • Senior Member
  • *****
  • Posts: 3691
  • Australia
  • Liked: 2645
  • Likes Given: 2287
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2816 on: 12/17/2024 11:19 pm »
Even a pinhole in a tank is a loss of mission
Is that true?

It's not. But impacts by hypersonic "pins" don't create pinhole leaks. So TRM's broader point is true. MMOD shielding will be vital.
« Last Edit: 12/17/2024 11:20 pm by Paul451 »

Offline OTV Booster

  • Senior Member
  • *****
  • Posts: 5477
  • Terra is my nation; currently Kansas
  • Liked: 3778
  • Likes Given: 6565
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2817 on: 12/18/2024 02:57 am »
Even a pinhole in a tank is a loss of mission
Is that true?

It's not. But impacts by hypersonic "pins" don't create pinhole leaks. So TRM's broader point is true. MMOD shielding will be vital.
This is another argument in favor of a depot. A ship hanging around waiting for tankers would need MMOD shielding. A ship or tanker spending minimum time in LEO would have a lesser need.


After discussion on the point we seem to have reached a consensus that where possible the depot should carry the heavy stuff so the tankers and outbound ships don't have to repeatedly haul it up to orbit. MMOD shielding is heavy.


Hmmm. A thought. MMOD shielding is a two layer affair with the outer layer taking the hit and turning the intruding particle into a spray of fine low energy particles that get stopped by the second layer. Just noodling here but could metallic heat shielding double as the outer MMOD shield?
We are on the cusp of revolutionary access to space. One hallmark of a revolution is that there is a disjuncture through which projections do not work. The thread must be picked up anew and the tapestry of history woven with a fresh pattern.

Offline OTV Booster

  • Senior Member
  • *****
  • Posts: 5477
  • Terra is my nation; currently Kansas
  • Liked: 3778
  • Likes Given: 6565
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2818 on: 12/18/2024 03:02 am »
In reference to safety during prop transfers, would transferring only one fluid at a time give a significant safety improvement?
We are on the cusp of revolutionary access to space. One hallmark of a revolution is that there is a disjuncture through which projections do not work. The thread must be picked up anew and the tapestry of history woven with a fresh pattern.

Online DanClemmensen

  • Senior Member
  • *****
  • Posts: 7446
  • Earth (currently)
  • Liked: 6053
  • Likes Given: 2532
Re: Starship On-orbit refueling - Options and Discussion
« Reply #2819 on: 12/18/2024 03:08 am »
Even a pinhole in a tank is a loss of mission
Is that true?

It's not. But impacts by hypersonic "pins" don't create pinhole leaks. So TRM's broader point is true. MMOD shielding will be vital.
This is another argument in favor of a depot. A ship hanging around waiting for tankers would need MMOD shielding. A ship or tanker spending minimum time in LEO would have a lesser need.


After discussion on the point we seem to have reached a consensus that where possible the depot should carry the heavy stuff so the tankers and outbound ships don't have to repeatedly haul it up to orbit. MMOD shielding is heavy.


Hmmm. A thought. MMOD shielding is a two layer affair with the outer layer taking the hit and turning the intruding particle into a spray of fine low energy particles that get stopped by the second layer. Just noodling here but could metallic heat shielding double as the outer MMOD shield?
I think you mean solar insolation shielding, not re-entry shielding. A Depot never needs to re-enter. I think a multi-layer insolation shield like the one JWST uses would make a fine Wipple shield. The problem is that the solar heat flux is not aligned with the MMOD threat.

Tags: HLS 
 

Advertisement NovaTech
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0