More Random Thoughts About Cis-Lunar Transportation

Ok, before I run off to work today, I wanted to follow-up on my comments last night regarding reusable lunar transportation systems.

Propulsive Braking or Aerobraking?
One of the questions that comes up a lot regarding reusable LEO-LUNO or LEO-L1 tug is how do you actually intend to get it back to LEO. The main options are to do it propulsively (ie by using its engines to decelerate and recircularize in LEO), or to use the earth’s upper atmosphere to slow down sufficiently, or to use a combination of both.

Propulsive braking is the easiest to do, has the least unknowns involved, requires the least beefing up of the tug, and requires the least new development work. On the other hand, it requires a huge amount of propellant to work correctly. This is due to the fact that with nothing to brake against other than your hot flamey stuff, you end up spending exactly as much propellant (to a first order approximation) getting back from L1 as you did getting out to L1 in the first place. I ran the numbers last night for a tug using LOX/IPA, LOX/Propane, and LOX/LH2 in order to take 4000lbs on an LEO to L1 and back round trip, and for taking the maximum amount of cargo to L1 with no cargo on the return loop. I assumed a roughly 3.7km/s Delta-V requirement for getting from LEO to L1. The good news is that even for LOX/IPA (which I assumed only had an Isp of 340s or so using space rated engines), the required stage fractions were well within what has been historically feasible (roughly 10% fuel and a few percent payload). Problem was that the stage in LEO required nearly 200klb of fuel in LEO to take 4000lb of crew or cargo on a round trip to L1 and back. For the LH2 stage, this dropped to only ~45klb of fuel, which is a lot more palatable. LOX/propane was in the middle with ~87klb of fuel required. The interesting thing though was that if you assumed a one-way delivery with the tug returning empty, the payload mass shot up to about 12000lb for the LOX/IPA, 11000lb with the LOX/Propane, and about 9000lb for the LOX/LH2 stage. The short and long of this analysis is that purely propulsive braking is likely quite expensive, but perfectly technically doable. And as Rand and other would likely point out, if you’re flying that many small flights, you’re going to see the cost per pound in orbit go way down along the way, since you can amortize your fixed costs and development costs over more flights.

Full or “100%” Aerobraking is both more tempting and trickier. With full aerobraking, you only need about 1km/s of Delta-V to go from L1 to an ellptical orbit intersecting the earth’s upper atmosphere, which does the rest of the job. But as any Heilein fan will point out, TANSTAAFL! The problem is that hitting just the right part of the atmosphere to slow down in a single pass without using any extra fuel is very, very difficult. For some background, since atmospheric heating varies strongly with gas density, you really want to do your aerobraking higher up in the atmosphere to keep the heat load down. However, particularly at the altitudes of interest, there’s a big problem–the atmosphere doesn’t stay put! As various factors such as solar radiation and the solar wind change, they cause the upper bounds of the atmosphere to contract and expand. I don’t have a good handle on what kind of timescales we’re talking about (my guess would be hours to a day or so for a fairly large change), but they’re likely fast enough to be a problem. For a vehicle designed for 100% aerobraking, you’re in a bit of a pickle, particularly if you have passengers. If you undershoot, you’ll skip off and it’ll be days or weeks before you can get yourself back into LEO (which could be a very bad day). If you overshoot, you’ll deorbit and have to land (if your vehicle can take that kind of extra heat input over and above the aerobraking). The problem is that with all the unknowns involved, it’s really hard to do this right. One solution might be to get much better and more frequent measurements of the part of the atmosphere of interest, either using some sort of sensors on some satellites, or using sounding rockets or some other means. If you could know the atmospheric properties in the target area with enough accuracy, it would be a lot easier to do a last minute course correction, and slide in right in the sweet spot.

The other option is to cheat and take the best of both worlds, and use partial aerobraking. What I mean by partial aerobraking is that of the 2.7km/s of delta V needed for circularizing into LEO, you only have the atmosphere provide part of that, say 1.5-2.5km/s. By keeping enough reserve fuel on board, you can afford to aim a little higher up in the atmosphere. Basically, you can bias your errors so that you tend to err on the side of less delta-V provided by the atmosphere, and hence more propulsive braking in order to reduce the odds of accidentally deorbiting yourself. I’m pretty sure that some combination of better information, lots of practice actually doing aerobraking, and a decent amount of reserve propellant should make a huge difference in the safety and reliability of such a scheme. And even with only about 75% aerobraking (ie having atmospheric drag contribute 2km/s out of the 2.7km/s of Delta-V), the mass savings are enormous. The IPA fueled tug drops from nearly 225klb wet mass in LEO to only about 55klb. The Propane fueld one drops from 87klb to about 37klb, and the LH2 one drops from about 55klb to about 26klb (with 4klb of that being cargo). Since you’re not on quite as steep of a section of the Delta-V versus mass ratio curve however, your one-way only cargo only goes up to about ~5500-6000lb. End result is however that you end up spending less than half as much propellant mass in LEO for given payload in L1 if you can do at least partial aerobraking.

And since you’re only trying to bleed off 2km/s, have a very favorable ballistic coefficient (ie you’re very fluffy), and are doing most of your decelerating in the upper atmosphere, the heat shield may be a lot easier to work with. I slated for a heat shield in each of those cases that was about 5% of the loaded mass of the vehicle (about 20-30% of the aerobraking mass however), which is probably reasonable.

So as you can see, there are tradeoffs, but once we can get in the practice of doing at least partial aerobraking (and a series of tech demos of such would go a long way to proving the idea out), it has some strong advantages.

Ok, I’ve ran out of time again, so I’m going to have to pick this up again later on today.

The following two tabs change content below.
Jonathan Goff

Jonathan Goff

President/CEO at Altius Space Machines
Jonathan Goff is a space technologist, inventor, and serial space entrepreneur who created the Selenian Boondocks blog. Jon was a co-founder of Masten Space Systems, and is the founder and CEO of Altius Space Machines, a space robotics startup in Broomfield, CO. His family includes his wife, Tiffany, and five boys: Jarom (deceased), Jonathan, James, Peter, and Andrew. Jon has a BS in Manufacturing Engineering (1999) and an MS in Mechanical Engineering (2007) from Brigham Young University, and served an LDS proselytizing mission in Olongapo, Philippines from 2000-2002.
This entry was posted in Uncategorized. Bookmark the permalink.

25 Responses to More Random Thoughts About Cis-Lunar Transportation

  1. Tom Cuddihy says:

    It’s an interesting concept…but it only gets relevant if you carry it out all the way to figuring out, based on a single crewed mission from LEO to the surface of the moon and back, what you need on a recurring basis:

    -assume your crew can get to LEO via whatever means available

    -assume there are fuel depots at L1 and in LEO

    -assume you can use partial airobraking as you describe

    I mission
    1. Crew in LEO-> L1
    2. L1 –> Soft Landing Lunar Surface
    3. Lunar surface –> L1
    4. L1–> LEO
    (alt: combine 3 & 4 for direct return)
    II mission support
    1. fuel delivery to L1, vehicle return to LEO
    2. fuel delivery to LEO by whatever means
    III schedule of a ‘sustainable’ architecture: 1 crewed mission to the moon every month.

    So know you’ve got three fuel related questions to answer:
    A)how much fuel do you need at L1 to accomplish #2 and #3 with the minimum crew / payload that is reasonable for a safe round trip (I’ll guess 10-12 mT for a crewed mission payload)
    B)how much fuel do you need at LEO to accomplish the crewed boost to L1?
    C)how much fuel do you need at LEO to accomplish the support mission #1?

    It’s an intriguing problem. Unfortunately, I have to go to class now & realistically I wont be able to get back to this until after my wedding / honeymoon. But if you keep it here, I’ll get back to this with numbers.

  2. Jon Goff says:

    You’re right that the devil is in the details. Right now I’m too busy with teasing out how to cram all the stuff into our vernier engine modules that we need to without impeding the hinge actuation, and trying to figure out how to do the plumbing in as simple and reliable a manner as possible. But as I get time I’m going to flesh it out a bit. The biggest problem is that nobody can really know where the technology is going to be five years from now let alone fifteen, so it’s really easy to lock yourself into premature overoptimisation.

    All that said, I’m going to go into a little more detail in my next post about the depot questions.

    BTW, if you’re about to get married, you’re probably busier than I am–I know how much time that sort of thing takes. Congrats! Hopefully your fiancee is as tolerant of space nerdlichkeit as my wife is. But even if she is, no talking rocketnerd stuff on the honeymoon. That’s just way too nerdy (not that I’m speaking from experience on that one…just being hypothetical there…whew look at the time…)



  3. Juan Suros says:

    If we’re throwing schemes around, how about trying a series of high altitude, low delta Vee atmospheric braking passes? Scrub that 3700 m/s off over tens of braking maneuvers. You spend a lot of time waiting for your next pass through the atmosphere, but so what? Sailing around the world takes weeks and months if you depend on wind power alone. Rocket power and ballistic trajectories are the wind power level of space flight, so why do we plan all our travel in space as “sprints”? Space travelers don’t have to be spam in a can military pilots. Think guys on a sofa playing D&D or watching Passions on cable as the better model. Merchant marine rather than fighter jocks.

    A 5 minute wikipedia/excel calculation gives the period of a LEO to Moon orbit as 27 days. Suppose you leave the moon in such an orbit rather than a 4 day transfer orbit. Two weeks to Earth, brake, and wait a couple weeks for your next shot. Repeat. Eventually you’re down to a 90 minute, 400 km orbit and you dock with a space station for a ride home.

    I’m guessing fuel mass from lunar surface to a 27 day transfer orbit is less than a 4 day transfer orbit, but I don’t know by how much.

  4. Tom Cuddihy says:

    “Hopefully your fiancee is as tolerant of space nerdlichkeit as my wife is.”

    she is. I’m a lucky man…

  5. murphydyne says:

    Great stuff Jon! Hopefully I can get around to the Direct Trajectory

    How about those European inflatable shields? I saw a neat adaptation at the Paris Air Show that could be used for jumping out of skyscrapers. Another thing to note is that L-1 is perfect high-ground for keeping an eye on the Earth’s atmosphere, up out of the clutter we’ve littered all the way out to GEO.

    One thing to remember is that from L-1 you have more destination options than just the Moon and LEO. Depot requirements are going to be more significant at L-1 than at either the Moon or LEO as a result. What I foresee in the intermediate near-future is that the LOX comes from the Moon and eventually trickles all the way down to LEO, and while at first all of the LH is going to have to come from Earth, eventually it may be that only LEO is supplied with terrestrial hydrogen, especially if we get moving on tapping the asteroids. Ad Astra online has a nice article in that regard, and L-1 is one of your best starting points.

    One thing that I try to keep in mind is that everything should serve at least two purposes. LEO is a halfway rest to points beyond, as well as a point of entry for returning crews. They might need new immunizations for example. L-1 gives cheap dV access to a number of destinations. An L-1 station can not only serve as a port, for free-flyer tugs, GEO missions, NEO missions, Moon missions, and so forth, but also as an eye in the sky that is less concerned with the objects around other stars and more concerned with those around Sol.

    That’s going to take a lot of fuel. (and oxidizer)

  6. Arthur says:

    If the point is to go from one orbit to another, rather than landing or launching, aren’t the best options going to be high Isp, to keep your propellant mass fraction low? Are there any options in between hydrogen-oxygen and solar electric that might make the most sense in balancing popellant mass and time (high Isp usually means low thrust and longer transit) – VASIMR maybe?

  7. Jon Goff says:

    If the point is to go from one orbit to another, rather than landing or launching, aren’t the best options going to be high Isp, to keep your propellant mass fraction low?

    Yes and no. The real thing is what are the total life-cycle costs of the system. The actual physics of the system aren’t as critical as the economics. While a LOX/LH2 vehicle may be smaller and use less propellant mass per given amount of payload, it will likely cost more to develop and to run. Whether or not it is worthwhile isn’t readily clear, and would probably depend a lot on the individual system. IOW, while it may appear obvious from a physics standpoint that LH2 should be used, once you get into the engineering and economics, it’s more of a tradeoff.

    Are there any options in between hydrogen-oxygen and solar electric that might make the most sense in balancing popellant mass and time (high Isp usually means low thrust and longer transit) – VASIMR maybe?

    The short answer to your question is: probably, but not at the moment. There seems to be two types of propulsion systems suitable for cis-lunar transportation: those than can get you there in a few days, and those that take several months. It’s usually one or the other. Of the options that fall in the first category but are “more efficient” than LOX/LH2, the only serious ones I can think of are Microwave thermal, nuclear thermal, or electrodynamic. The problem with VASIMR is that since there is no release of energy from the propellants itself, your thrust/Isp end up being constrained by your power source. There just isn’t an internal power source that you could fit on a vehicle that would make VASIMR have enough thrust at a high Isp to get to the moon any quicker than an Solar Electric system. If VASIMR does actually work, there may be something that can be done to resolve that issue, but that’s a post for another time.

  8. Jon Goff says:

    The big problem with multi-pass aerobraking is those pesky Van Allen belts. Having to pass through them almost two-dozen times during a flight is likely going to make either your vehicle design very difficult, or give crew/passengers a pretty hefty radiation dosage. For electrically hardened unmanned space hardware, it might be ok, but for people probably not.

    Plus, with 75% aerobraking you get almost all the benefit of the aerobraking with none of the costs of doing a multi-pass deceleration.


  9. Juan Suros says:

    > The big problem with multi-pass aerobraking
    > is those pesky Van Allen belts.

    I’m not so sure. I don’t see how we will ever leave low orbit again until we adopt something like the M2P2 idea for particle radiation shielding. I certainly don’t think any near term design is going to have enough mass of hull or H2O or propellant surrounding the crew capsule to keep space travelers healthy.

    > Plus, with 75% aerobraking you get almost all
    > the benefit of the aerobraking with none of
    > the costs of doing a multi-pass deceleration.

    You make a good argument. I guess I just prefer the flexibility of being able to come around for another try if the atmosphere wasn’t as thick as you expected.

  10. Anonymous says:

    Tethers in Earth orbit seem to make sense in this context. Exchanging the energy of the incoming vehicle for boosting an outgoing would be nice.

    If development costs can be kept reasonable of course.

  11. Paul Dietz says:

    About the suggestion for high specific impulse:

    How about a non-ablative heat shield that is cooled by hydrogen, with the hydrogen subsequently ejected through a nozzle? Kind of like a nuclear-
    or solar-thermal rocket, but with the heat energy being provided by the vehicle’s motion itself. You could control the deceleration by changing the orientation of the nozzle(s).

  12. kert says:

    LH as a propellant in cislunar system is pretty much a nonstarter just because of the boiloff problems, which is an order of magnitude worse than with LOX. You’d have a logistics nightmare arranging the launches in time so that your depots would have enough LH left at critical point of time.
    Another issue is that at present state of our knowledge, you’d have to ship it from earth anyway, as you’d do with hydrocarbons, at least until the presence and extractability of lunar hydrogen is eventually proven in situ.
    And of course, all the hardware development is obviously more expensive with LH than with hydrocarbons.

  13. Juan Suros says:

    Arthur, Kert –

    Here’s a crazy idea for a high Isp space storable propellant combination: Iron Carbonyl and LOX. Both are dense and space storable. Burning them probably wouldn’t give you much of an ISP directly, but the exhaust contains ferromagnetic products that can be boosted electrically with something like MHD coils to whatever Isp your power supply can support.

    It always bothers me that standard electrical engines waste so much of the available power on ionizing the propellants, so a chemical-electric hybrid might make sense.

    The real win, of course, is that making Iron Carbonyl and LOX at an asteroid is very simple. It doesn’t seem so attractive for the moon, but given the number of asteroids that have crashed into it over the ages there must be carbon somewhere on the lunar surface.

    It sounds too good to work. What did I forget about? Is the exhaust too nasty for near earth space, maybe?

  14. Juan Suros says:

    Paul –

    Use a heat exchanger instead of a heat shield for reentry? What a great idea. I do wonder how much of the hot flamey stuff would come right back at you, but I suppose you could direct the exhaust off to the side…

  15. Tom Cuddihy says:

    re: Atmospheric thickness, or uncertainty in atmospheric density

    This is actually a big deal, as it’s the major source of error in the space object catalog by the US govt and Russia, and in all orbit propogation techniques. The russians probably have the best method–rather than using extremely sensitive modeling of the atmosphere by soundings, they just use a composite taken directly from the differences in actual satellite tracks from the projected. It’s primitive, but it works.

  16. Jon Goff says:

    About the suggestion for high specific impulse:

    How about a non-ablative heat shield that is cooled by hydrogen, with the hydrogen subsequently ejected through a nozzle? Kind of like a nuclear- or solar-thermal rocket, but with the heat energy being provided by the vehicle’s motion itself. You could control the deceleration by changing the orientation of the nozzle(s).

    As usual, your idea is quite intriguing. The only issue I could think of is that the studies I’ve seen so far seem to suggest that having a rocket jet facing into the wind will cause the bow shock to become more pointed, thus greatly reducing drag. There may be ways to solve this problem (possibly have the engines around the outside edge of the vehicle like the verniers on our XA vehicles?), but the idea has enough merits that it’s probably worth looking into.


  17. Antti Toiviainen says:

    About the difficulty of (near) 100% aerobraking; how about active control of vehicle’s ballistic coefficient? And I don’t mean just rolling/trimming CG like semiballistic capsules but deploying flaps as additional ‘air brakes’.

    The vehicle might resemble Kistler K-1 upperstage with blunt heatshield, but instead of fixed conical aft skirt it’s sidepanels would be hinged near the heatshield and open up by hydraulic or electric rams. With sidepanels fully open the thing would truely look like a giant shuttlecock.

    Start aerobrake with panels partially deployed using trajectory that should do the trick without any control, given the information about the atmosphere you have at that point. When the craft enters the upper atmosphere you get realtime data and may begin to calculate whether you need trimming. If atmosphere appears thicker than it should, retract flaps, and if thinner then deploy them wider. Not sure how uniform the atmosphere is but I’d imagine that once you hit the perigee you have pretty good and valid model of the current atmosphere which you can use to compute final trimming for the rest of the aerobrake.

  18. Anonymous says:


    I believe there is a fourth option for LEO return. Aerobraking, propulsive, and tethers are already on the table.

    It will take a few minutes with sketches and calculators in person. If you have time at SAS 06, mention “zero fuel rendezvous” and I will show you a possibly new concept. It is somewhat tether based, and more important, can be developed and tested on the ground, if feasible. Mass would appear to be competative with pure propulsion by mission two.

    Trying to describe it accurately here is beyond my communications skills.

    John Hare

  19. Andrew Case says:

    If you are doing aerobraking you might as well plan for nonzero L/D so you can do plane changes as well. Once you have nonzero L/D you can accomodate atmospheric density changes by controlling attitude to increase or decrease lift at the expense of drag in order to achieve the right delta-V. You’d probably have to use the RCS for trim control, but it beats the hell out of using all that fuel for correction burns after the aerobraking maneuver.

    In addition, it’s not as if atmospheric density is going to be a complete unknown during the late phases of the exoatmospheric trajectory. Simple optical techniques based on scattered solar light can narrow the range of possible densities considerably, so that the terminal exoatmospheric trajectory can be tweaked to minimize RCS fuel consumption during the atmospheric braking menauver.

  20. Jon Goff says:

    Andrew, Antti,
    The idea of using aerosurfaces to modulate the drag coefficient and the L/D ratio is probably a very good idea. If what you say about using scattered solar light to get a good initial estimate of the atmospheric density could work as well as you imply, that could be quite interesting. Use something like that to plot your initial trajectory, and then tweak it in real time using Tom and Antti’s suggestion, since that can get you real-time, local data on what’s going on. It makes the control system a lot more crucial, and having some pretty serious propellant reserves to handle mistakes there could be a good idea…

    Hmm…this is giving me an idea for another blog post…


  21. Iain McClatchie says:

    Re: Aerobraking accuracy

    The moon-earth transfer takes a few days. When leaving moon orbit, you could fire off two or three small expendable probes, nothing more than beacons with known drag coefficients and maybe accelerometers. One day, and maybe one hour out, the beacons re-enter, and you adjust your course based on the data they return.

    Re: Lifting entry

    First, note that rocket thrust against the gravity vector instead of against the velocity vector has a great deal of leverage during the early part of a reentry, because it can keep you in the upper atmosphere and low-G, low-heating portion of the flight longer.

    During an aerobraking maneuver, you might use rocket thrust at right angles to the flight path to guide the vehicle through the portion of the atmosphere with proper density.

    Both of these ideas are kaput if the atmospheric thickness varies unpredictably from one longitude to the next. One can imagine turbulent standing waves at dawn and dusk, etc.

    And, of course, I like tethers best. This kind of delta-V seems like it’s right in line with modern material strengths. I have no sense of the logistical issues, which I suspect are real showstoppers. (ie you have to come home one day early, but the tether is not in place.)

  22. Sam Dinkin says:

    Single pass on a 4 day orbit has about a 50x capital utilization rate vs. 12 pass starting at a 27-day orbit. Don’t forget to optimize ROI.

  23. Cujo says:

    I ran some quick numbers for aerobraking altitudes with which I was comfortable, with densities near Solar Min, and came up with needing an area of roughly 10^3 m^2/kg to perform single-pass aerobraking from L1 (see URL for more numbers). Unfortunately, the thinnest “parachute” we can make now is about 15 kg for every 1000 m^2 of area, so there’s no mass break even; even if I’m off by a factor of 15. Some would say we could get this down to 0.005 kg/m^2, in which case there is some hope, but that’s a lot of heat (about 28 MJ/kg) to dissipate over such a thin surface – comes out to about about 100 W/m^2.
    The options are to go to lower altitudes (well below 150 km), which I dislike for safety and thermal reasons, or to go with multiple passes. Each subsequent pass would be in a lower period orbit and would spend more time near low altitudes, so we’re probably only talking about 4 passes to get to feasibility. The first pass could be at a relatively high (about + 30 km) altitude to minimize thermal issues and put less stress on attitude stability.
    This is probably fine for unmanned cargo transport, but is a risk for a crew. I tentatively conclude that if you’re going to build a crewed spacecraft capable of dissipating a high proportion of its excess energy through aerobraking, then you’ll have to build a vehicle that could re-enter directly anyway. It’s always very complex to operate in more than one environment.

  24. meiza says:

    So, for lunar operations, use a tether or multipass aerobraking for the empty orbital module / bigelow habitat + service module and do a flap / sideways thrust controlled re-entry with a dragon capsule. Both hardware can be reused, they’re separated on the way from the moon and maneuver to slightly different trajectories.

    What’s the solution to big impact loads with space tethers?

  25. meiza says:

    That way, all hardware is reusable, but you don’t have to launch or bring down to Earth but just the dragon. And launch some fuel to the LEO depot and L1. Even the lunar lander can wait at L1…

Leave a Reply

Your email address will not be published. Required fields are marked *