I only have a few minutes tonight, but part of what’s been keeping me busy lately has been two Space 2009 papers I’ve been involved with (one as the primary author, and another as a very minor coauthor).
First, the propellant depot paper I keep talking about: AIAA 2009-6756 Near-Term Propellant Depots: Implementation of a Critical Spacefaring Technology (also available on the ULA publications page), and the presentation that went along with it: SPACE 2009 Prop Depot Paper Presentation.
For those who’ve read most of the rest of my posts on propellant depots, there’s only a few key new concepts:
- The Single-Launch Dual-Fluid Depot concept that Frank Zegler and myself both independently came up with this year (which I’ll go into more in a later blog post). This idea holds a lot of promise because it shows how a single EELV launch with existing fairing sizes can put up a depot capable of storing 75-114mT of LOX/LH2.
- The realization that unless you have some sort of high-Isp or propellantless propulsion system (like an ED tether) for stationkeeping purposes, that Zero-Boiloff storage might not be very useful for an LEO depot. The amount of propellant you lose to boiloff is less than the amount you would’ve spent for stationkeeping anyway, so it’s effectively free.
- The realization that LEO depots really ought to be treated as “use-it-or-lose-it”, high-throughput depots, and that its L1/L2 depots that should be used for longer-term storage.
I had been avoiding discussion of the dual-fluid depot concept for a while, mostly so I wouldn’t be stealing my own thunder. Now that the paper is presented and out in the public, I hope to have the time soon to discuss the concept a bit. I also have some space transportation architecture ideas using that depot concept that I may flesh out a bit either here on the blog or in some white papers (which I’ll post on the blog).
The second paper, which was mostly written by Robert Frampton of Boeing talks about a project to use our XA-0.2 vehicle as a testbed for demonstrating autonomous landing systems for planetary landers: AIAA-2009-6571 Planetary Lander Dynamic Model for GN&C
I’m not really sure what I’m allowed to say about the project that isn’t spelled-out in the paper, but I figured I’d bring it to people’s attention, and hopefully at some point in the future I can discuss things more on the MSS blog.
Jonathan Goff
Latest posts by Jonathan Goff (see all)
- Fill ‘er Up: New AIAA Aerospace America Article on Propellant Depots - September 2, 2022
- Independent Perspectives on Cislunar Depotization - August 26, 2022
- Starbright Response to ISAM National Strategy RFC - July 2, 2022
Both papers were very interesting indeed. The most surprising was the detail and images of XA 0.2 in the Boeing paper. I assume that craft is intended for the LLC level 2 effort. I liked what was shown as it is a clean design though it was surprising to me that it shows a quad-engine arrangement. I can perfectly understand this being the next stage after the level 2 prizes have been collected though to get the control to work well enough on differential throttling with so short a period to go for this year. Having said that it is understandable that Masten’s focus is likely toward progress that moves you toward your goals rather then being side tracked to far from your planned development schedule.
Thank you for sharing.
Very nice! First author of a very interesting paper with some heavyweight coauthors. We shall watch your career with great interest. 😉
As you know I have some ideas for even simpler first generation depots (or better yet mere in flight refueling), but it is interesting to see we reach similar conclusion for different reasons. For instance, using L1/2 as the main node, with LEO only acting as a transit point. The thermal environment at L1/L2 is better, but it also leads to smaller depots, shorter burns and makes high Isp less important. No need for nukes, even for Mars. And no need for cryogenics either. 😉
The stationkeeping issue is surprising. Are you thinking of depots in very low orbits? Apparently there are plans to raise the orbit of the ISS once the Shuttle is retired so it encounters less drag. Limitations of the Shuttle are what’s keeping it in its current orbit. Might a higher orbit not be a good idea for depots as well? Especially if you want to use the ISS as a staging point for crew and as a maintenance base for the depot. Or would it make the thermal environment worse?
To what degree is the drag determined by the need to minimise the area presented to the sun? Non-cryogenic or maybe even mildly cryogenic depots (i.e. anything not containing LH2) might be permanently oriented along the orbital velocity vector. Denser propellants (i.e. anything but LH2) would also be better off in this respect.
Satellites in L1/L2 need 30-100 m/s of station keeping per year.
The same thrusters can be used to rotate a depot and perform the station keeping. LEO to L1/L2 ascent can be performed by the same thrusters.
In the depot paper it is mentioned several times that the hydrogen has over ten times the heat capacity of oxygen. I have been using sixteen for quick and dirty figuring. I checked specific heats in wiki and came up with 15.77 difference.
I’m curious as to where I went off track.
Heat capacity or heat of evaporation? Both seem important if you use subcooled propellants.
I see what you mean. I’ll do a few figures using both. Thanks.
There’s also a difference between specific heat capacities at constant volume and constant pressure. Both probably depend on temperature and pressure as well. We’re going well beyond my very limited thermodynamics knowledge now…
Hi,
I have a wild, but maybe interesting idea I don’t find anywhere. Launch vehicles have a nominal capacity that must be over their actual payload’s mass in order to be able to get the desired orbit. Supposing that the upper stage uses the same propellant and oxidizer as the spaceships, his difference between the two masses can be filled with excess prop/oxy on every single flight. Maybe a small amount, but practically for free. This excess can be collected to orbital fuel depots.
Since upper stages are ‘dumb’ tanks, a tug is needed, and this works mostly at LEO, since most launchers get there. I envision the ISS or a similar space station where the interplanetary spaceships are assembled, and filled with propellant, partially form excess fuels of the launchers go there.
> The amount of propellant you lose to boiloff is less than the amount you would’ve spent for stationkeeping anyway, so it’s effectively free.
Why is that? I’d think they were additive so any savings would be good. But you’re implying not?
Does boiloff can count as effective propulsion? Or is station keeping cost really so significantly large compared to boiloff cost, like 100x worse, such that you wouldn’t even notice the boiloff by the time it ran out of fuel from station keeping…
I see the paper does not mention dielectric settling (where a liquid propellant is attracted into regions of strong electric field by virtue of its dielectric constant being larger than the vapor phase). Is there a problem with this approach?
Paul Dietz,
Oversight on my part, AFAIK it’s a perfectly legit approach. I think I actually mentioned it in the presentation, but I was trying to put this paper together right in the middle of when I was doing most of the design work for our LLC vehicles.
~Jon
dso,
Basically warmed up GH2 (which has been used to remove heat from the LOX, intercept heat from solar panels, and electronics, and everything else) can provide a surprisingly high Isp (ie almost as good as a NTO/MMH thruster). So you take all of the hydrogen that you were losing for boiloff, and direct it out a suitable nozzle in the direction needed for reboost. You’ll need to add a little bit extra, and might be worth actually burning, or intentionally dumping some extra heat into it, but the basic idea is that the boiled off GH2 becomes part of the propellant flow for stationkeeping thrusters.
~Jon
John,
I didn’t actually run the numbers myself (I was repeating what Frank Z/Bernard K had written). But 15.77 or 16 are both “over ten times” greater than 1….
~Jon
Martijn,
Regarding stationkeeping, I’m not sure what altitude they were assuming. Remember that too high of an altitude, and you start taking performance hits for reaching and departing from it. But it’s also important to remember that depots are going to be really “fluffy”, so even at higher altitudes they’ll suffer more drag losses than a denser station. Well, the momentum they lose from drag will be the same as a station of equal frontal area, but since they are typically less dense, they’ll drop faster. I’ll have to ask Frank what altitude he was assuming.
~Jon
David,
XA-0.2 was actually an idea from two years ago that we were going to use for the LLC, and haven’t really updated since we got the tanks in a year and a half ago. Plans for this year’s L2 vehicle (XA-0.1E aka “Xoie”) are a lot simpler. And in fact, the design of XA-0.2 is probably going to get revved based on everything we’ve been learning from Xombie and Xoie. We’re probably going to make the design a lot more aerodynamic, go to the new fuel tank concept we’ve been using, and possibly revisit the engine number. The Boeing team wanted engine-out capability for their series of demonstration missions, so we’ll likely keep at least one 4 or 5 engine version of XA-0.2. But anyhow, a lot of that is likely to change as we go on.
~Jon
fascinating stuff. I’m a Mars guy myself, but I’m coming around to the view that while a Moon mission does very little to help a Mars mission, serious attempts to establish a LEO/lunar infrastructure makes a Mars mission and ultimately a Mars infrastructure massively easier. It really should be about extending our economic sphere and more and more I’m seeing that propellant depots and other LEO infrastructure are the obvious next step.
Remember that too high of an altitude, and you start taking performance hits for reaching and departing from it.
Only by tens of m/s for reasonable altitudes. Low orbits may be a smart thing to do for LH2 depots, since if you’re going to lose some propellant to boil-off anyway, you might as well get something out of it. Then again ISS staging is very nice and a low altitude orbit complicates the staging substantially. Not insurmountable, but it does reduce the benefit. And regardless, boil-off may well be the price worth paying for high Isp chemical propulsion from LEO. No need to justify it with reuse as stationkeeping propellant. Nice if you can get the reuse, but cryo depots are an excellent idea even without it.
BTW: interesting that the idea is to use GH2 as a monopropellant. Makes sense of course, but somehow I was under the impression they’d use GOX/GH2. That incidentally is what Bigelow is planning to do for one of the Sundancer propulsion systems. The reboost system will still use hypergolics.
But it’s also important to remember that depots are going to be really “fluffyâ€, so even at higher altitudes they’ll suffer more drag losses than a denser station.
Only LH2 depots will be fluffy, pretty much everything else will have decent density. As you know I’m mostly thinking about hypergolics, but the same goes for kerosene, LOX, methane, silane, ammonia, H2O2, ethanol etc. And with propellant gels density can be increased for all propellants, even LH2. Metal loading these gels will increase density even further. Fluffiness is not an inherent property of depots, whatever HLV apologists may claim.
Martijn,
While hypergolics may not kill your performance from L2 down to the surface too bad, it really sucks for the LEO to L2 leg of the trip. LOX/LH2 really makes a lot of sense, and the boiloff losses appear to be pretty reasonable. According to the numbers Frank Z ran for his ACES-derived dual-fluid depot, if you were getting 30mT of propellant to the depot per month (1-2 EELVs or a daily 1mT RLV launch), you’d lose about 3% of your yearly throughput to boiloff losses. That’s trivial compared to the amount of savings you get compared to a hypergolic system.
~Jon
Oh sure, I wasn’t trying to plug hypergolics here, although I can’t resist mentioning them. I already annoyed Rand with that and didn’t want to repeat the performance here. I don’t even think hypergolics are a desirable EDS propellant.
What I was trying to say is that LOX/LH2 is a good idea even without reusing boil-off GH2 as reboost propellant and even in higher orbits, specifically ISS orbit. Also that fluffiness only applies to LH2. In other words I was trying to separate out the issues that are specific to LH2 depots from those that are not.
I do believe LOX/LH2 depots are the best bet we have for commercial activities beyond LEO and that’s an important goal. Even though it’s not inconceivable that once you have RLVs you could use hypergolics from LEO (remember that Russian pump-fed engines have an Isp of 340s and with metal loading you could reach as high as 370s which is pretty darn decent), I’d say it’s unlikely because the stuff is so nasty. Kerolox would be more likely than that and not necessarily a bad bet. The Falcon 9 2nd stage could be a pretty good EDS. But if I were a betting man, my money would be on LOX/LH2. And the sooner the better.
Personally I don’t believe we’ll have commercial activities beyond LEO before we have RLVs. That’s why I think any kind of depot soon is more important than optimal depots. Note that LOX/LH2 depots soon is still important from that point of view because it means ULA would no longer be the sole provider of launch services for EDS propellant. But that’s a discussion for a different forum.
BTW: interesting that the idea is to use GH2 as a monopropellant. Makes sense of course, but somehow I was under the impression they’d use GOX/GH2.
If you have a source of external power, a resistojet or arcjet with GH2 can have higher Isp than GOX/GH2.
Paul,
In fact, ULA and another group did a paper on just that for Space 2009. Converting a CRYOTE system into a free-flyer that could deliver up to 4 smallsats with up to several km/s dV. They’d use the boiloff GH2 run through a resistojet style engine to get high Isp pulses (about 800s Isp).
As it is though, even just letting the GH2 warm up to room temperature gets it up above 300s Isp supposedly.
~Jon
Noting that there may be significantly more H2 on the moon than previously thought, but that O2 is still far more plentiful. What is the prospect of a high ISP engine that uses O2 and power from the thin film solar (~4.3kW/kg) being suggested for SPSs?
Should we also be looking at propellant depots for higher ISP engines? Seems almost time to develop higher ISP systems that can also operate of extra terrestrial resources.
Jon:
There seems to be some conflict between your three points.
As I understand it, if the depot is not empty, boiloff scales to depot capacity (surface area, etc) , not the current propellant load.
If you have, for example, a 100 mt LEO depot, that gets 35.5 mt of propellant a month, and every two months it send a 71 mt tanker to EML2, boiloff will be x. Call it 3% a year, or 13 mt.
Doubling the flight rate to EML2 doesn’t help that much, because you only reduce boiloff when the depot is entirely empty.
Will,
If you double your flight rate to L2, you’re getting twice as much propellant through for the same amount of boiloff. If you boil off 13mT of LH2 per year, and you’re shipping out 12x 71mT tankers per year, you lose a lower percentage than if it is 6x 71mT tankers per year or 3 or 2 or 1. Basically I’m just saying it favors increasing the throughput…that make any sense?
~Jon
Great stuff, Jon. Thank you.
Question: In “your” article, the configuration in Figure 1 looks dynamically unstable. That is, it would seem that the system would attempt to move to a lower energy state but twisting around the angular momentum vector until it was pitching end over end about the 15″ axis, instead of rotating about the 31′ longitudinal axis. How is this prevented? Counterweights along the rim of the sunshade or just inside it? Is that in any particular ULA paper?
Jim,
To be honest, I don’t know for sure. I do know they’re planning on doing a flight test of the system using the leftover propellant after the DMSP mission this fall. I can ask Frank Z.
~Jon
To maintain stability a 360 degree sun shield can be rotated in the opposite direction to the depot’s tank. This will require an ordinary electric motor and a magnetic bearing.
The images that I see of the Centaur have about the same proportions as the spinning depot in Figure 1, probably for reasons of heritage. Using a stage after DMSP deployment means that there won’t be any additional hardware to stabilize it. I don’t know who Frank Z is, but I hope he has some clever idea beyond muscling the tank until the ACS runs out. Perhaps they will intentionally pitch it end over end, although that would throw the residual LH2 to the wrong end of the tank.
Counterrotating to cancel the net angular momentum is not an answer. The sunshade, as suggested by A_M_Swallow, is a very low mass, non-rigid element, and the bearings would have to handle significant torque, and be a serious issue.
As it is, the illustration implies a spun-despun configuration to keep flat solar panels pointed at the sun. However, they might be better off with cylindrical panels, like those built into the HTV, to avoid the complexity of despining.
PS Ah. Frank Zegler, third author
On LEO cryo depots being use-it-or-lose-it, this is a great argument against HLV-based depots like the enormous depots proposed by DIRECT. Supposedly depots now need to be so big they can only be launched on HLVs. Boil-off foils that idea in LEO since large LEO depots are useless without massive active cooling and L1/L2 depots don’t need to be large at all. Heheh.
Huge depots were not *necessary* anyway, even if there were no bad sides to them.
And if the only use for a heavy lifter is lifting a big depot (remember, the depot removes all the other heavy lift needs), then that gives an awfully low flight rate, not worth the investment in a heavy lifter…
Hmm, on second thought the situation may not be as rosy as I thought.
Momentum loss through drag and the propellant mass required to counteract it are proportional to the area of the depot. Boil-off is also proportional to the area. Total mass is of course proportional to volume. And since area scales quadratically while volume scales cubically that would actually seem to support larger depots.
There are plenty of arguments left that argue against large depots, but boil-off might not be one of them. Am I missing something here or does the boil-off argument go out the window?
Large depots and boil-off – desert animals are either large like camels or small and able to hide from the sun like mice.
Good comparison. I was thinking of the explanation why monstrous insects cannot exist anymore. Insects breathe through their skins and that is why in the current atmosphere they cannot get enough oxygen if they grow much beyond their current sizes. In earlier times there was much more oxygen in the air which is why monstrous dragonflies could exist back then. Boil-off goes in the opposite direction from breathing and therefore leads to the opposite conclusion.
Martin: Going way off on a tangent here…
I’ve recently read theories that the early Earth atmosphere also had much more nitrogen. I find it plausible that back during the Carboniferous the partial pressure of O2 could have been considerably higher without the fraction of O2 in air being much higher than todays. This would have gotten around issues with flammability of plants in a high O2 atmosphere. A denser atmosphere would also help explain all those huge flying insects.
And, a thicker atmosphere with more nitrogen would help explain the “early faint sun” paradox, which asks why the Earth wasn’t permanently glaciated when the sun was 30% dimmer. Extra N2 causes the CO2 in the atmosphere to become a more effective greenhouse gas by pressure broadening of absorption lines. Just increasing CO2 itself doesn’t work, since at the levels required (with N2 held the same as at present) it starts making dry ice clouds in the upper atmosphere.
Just noticed a typo in figure 1 in the paper. It should be “gas core” and “liquid annulus”.