Lunar Depot-Enabled Multi-Sortie Missions Part II: Centaur-Derived Landers

As I mentioned in the first post in this series, I wanted to paint the overall picture first, and then flesh out the details as time and interest permits. For this post, I want to discuss an interesting lander concept that could work well with the mission model I discussed in Part I. I may discuss some thoughts about how to do lunar lander reuse in a future post in this series.

By way of introducing the concept, I wanted to point out some material that LM/ULA came up with two years back that got me thinking in these directions. While many have read the AIAA paper ULA published in 2006 about various Centaur-derived manned lander schemes as alternatives to the ESAS LSAM, there was also some less-well-known material they had developed for Centaur-derived robotic landers that I found interesting. I just noticed today that a paper containing the information I had previously seen about this concept is up on the ULA site, here, so I figure it’s now ok to talk about this idea.

Basically, the second paper goes into some work LM/ULA had done for the Lunar Precursor Robotics Program back in the 2006 timeframe. They had looked into converting an existing Centaur into a lunar lander for robotics payloads, by adding a “Extended-Duration Mission Kit” and a “Lunar Lander Kit”. These kits, which the Centaur team has already detailed to some extent, would add things like better passive cryo insulation hardware, sunshields, solar panels, upgraded avionics and batteries, landing gear, landing propulsion systems, etc. The concept was based on launching the whole Centaur lander stack into LEO on an HLV.

Centaur-Derived LPRP Robotic Lander

Centaur-Derived LPRP Robotic Lander

The paper also went into a 4-person lander using the same Centaur-derived concept but extending it a bit further. A version of this concept was further discussed in the first paper. The manned lander would be two-stage with a hypergolic biprop system for ascent, and the lander would include hardware for supporting at least two-week lunar surface stays.

Centaur Derived Manned Lunar Lander

Centaur Derived Manned Lunar Lander

What I was interested in was what those concepts could do if they were used with in an architecture that included a LLO depot/waystation. In the case of the robotic lander, the lander itself also performs the TLI and LOI burns, which means that most of its propellant is used up before it gets to LLO. For the human lander, while they assumed the use of another stage to do the TLI/LOI burns, the system was constrained to be launchable with an Atlas V HLV, which meant that a full Centaur-load worth of propellant couldn’t be used for it either. Plus, with the use of a hypergolic ascent stage, the ascent fuel weighs a lot more than it would in a reusable scenario. Fortunately, this paper gives a mass budget, so we can do some number crunching.

For the robotic lander, it used a Centaur dry mass of 2500kg, a Extended Duration Mission Kit mass of 800kg, and a Lunar Lander Kit of about 1000kg, with 1500kg of LLK propellant, and 21000kg of Centaur propellant.  Now assume a mission concept where you tank the whole Centaur stage up in LLO, the Centaur propulsion provides most of the delta-V except for the final touchdown/hover, the hypergolic landing engines provide landing/hover thrust as well as enough ascent thrust to get the vehicle up a couple hundred meters before relighting the RL-10s for ascent.

Depot-Enabled Centaur-Derived Manned Landing Missions
Factoring in some extra hypergolic propellant for both a long-duration hover (>90s) during landing, and enough propellant to get the vehicle up to a decent altitude before lighting the RL-10s, I estimate a payload in the 7500-8850kg range (you can download a copy of the spreadsheet I used here).  The lower number was assuming a 2500m/s ascent delta-V (ie ascent DV plus some plane change propellants), while the higher payload was for a 2200m/s ascent delta-V, which is probably closer to what you would nominally need (when you have backup systems like tugs, depots, and a second lander in orbit, you don’t need as much in the way of contingency margins on any individual flight). 

By way of comparison, the mass of the Apollo LEM minus main propellants was 4200kg, and most of that was stuff that would already be provided by the lander.  So, it’s pretty safe to say you could haul at least four people up and down in such a system.  For another comparison, at the higher end of the hauling capacity, you could haul a full Bigelow Sundancer module to and from the surface. Lastly, comparing it to the two-stage lander that they analyzed, if you ditched the ascent propulsion system and propellants and used the Centaur stage, it looks likely that you could haul 6-8 people and several tons of cargo for a two week stay without much difficulty. I can check on that if I can get better numbers from somewhere of the mass breakdown for their concept–the numbers given in paper #2 aren’t very clear on which weight in the ascent stage is for stage and propulsion mass, and which is for crew accomodations, pressure shells, etc.

Note, that at that point, this is anything but a “light scout lander”.  Such a system would likely provide a substantial increase in capability compared to the ESAS LSAM, while only requiring a marginal 25-26 metric tonnes per mission worth of propellants/consumables.   I don’t have the latest ESAS numbers, but for a sortie mission from them, you’re talking at least 65 metric tonnes, for only 1/2 to 2/3 as many people.

Also note that all of this is based on the existing Centaur design, not the Wide Body Centaur/ACES stuff that ULA has been investigating over the past several years.

Depot-Enabled Centaur-Derived Cargo-Dropoff Lander Missions
Now, what if instead of hauling a crew module up and down from the lunar surface, you were just hauling one-way cargo down to the lunar surface? For that scenario, I’m getting about 23,900kg of landed mass. Which is probably enough to deliver a full Bigelow Nautilus module to the surface. Or just about any piece of equipment you could imagine. Once again, this is with a stage based on the existing Centaur stage, not anything fancy like the ACES stage.

Admittedly, getting a payload that big to lunar orbit is actually the bigger challenge. You would need to use something bigger than a single Centaur derived transfer stage like I had talked about in Part I. Solar electric tugs, multiple Centaur stages in series, or a WBC/ACES derived transfer stage would be required. Or just finding a way to offload a decent amount of weight from the module itself, and outfitting it in lunar orbit before landing it.

Anyhow, I think this concept shows that using propellant depots in lunar orbit can greatly enhance a lunar exploration/development program, while also making the transportation phase of the program much safer. This performance benefit is not just with tiny sortie missions, but also with missions much more capable than what could be done with the planned ESAS architecture. Depots just make too much sense.

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.
Jonathan Goff

About Jonathan Goff

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 Bigelow Aerospace, Commercial Space, Lunar Exploration and Development, NASA, Propellant Depots, Space Transportation. Bookmark the permalink.

36 Responses to Lunar Depot-Enabled Multi-Sortie Missions Part II: Centaur-Derived Landers

  1. Tim says:

    I realise drawing like this are only representative, but are those wheels on the landing gear? Sorry to go off on a tangent so early in the discussion, but has anyone seriously considered putting wheels on a lunar lander? It might make handling on the lunar surface easier (say to reposition the solar cells if your staying for a while), or if powering them didn’t cut into the weight budget too much, give you a limited roving capability (which might be useful for say picking up injured astronauts among other things).

  2. kert says:

    but has anyone seriously considered putting wheels on a lunar lander?
    CMU early IceBreaker design had rover and lander in one unit.
    http://www.ri.cmu.edu/publication_view.html?pub_id=447

  3. Exploration Fan says:

    Also a little off topic, but for those that don’t believe in the enabling nature of propellant depots how do they propose missions beyond the moon. Yes, with enough money one can build a sufficiently large rocket to support trips to the moon. But how do we go to Mars and beyond. Does anyone truly think developing a 500 T + rocket makes any sense? NASA is already struggling with storing propellants in the Ares V for more than a few days, so launch 5 of these to support a Mars mission really isn’t an option.

    It seems to me that development of depot technology is darn near manditory for robust lunar or any crewed exploration of Mars. Why not develop the depot technologies now and skip the huge Ares V expense.

  4. Will McLean says:

    Too good to be true. By landing with tons of ascent fuel in the tank and a greatly increased payload, you more than triple the landed weight from the original design. This requires a similar increase in the hypergolic landing system thrust and landing gear weight, and generally beefed up structure.

  5. Will McLean says:

    Also, a single stage reusable version of their manned lander would have less payload, not more. Although hydrogen has higher Isp than hypergolics, not discarding the descent stage doubles the burnout mass on ascent, which more than wipes out the savings.

  6. Karl Hallowell says:

    Also a little off topic, but for those that don’t believe in the enabling nature of propellant depots how do they propose missions beyond the moon. Yes, with enough money one can build a sufficiently large rocket to support trips to the moon. But how do we go to Mars and beyond. Does anyone truly think developing a 500 T + rocket makes any sense? NASA is already struggling with storing propellants in the Ares V for more than a few days, so launch 5 of these to support a Mars mission really isn’t an option.

    I understand the actual number would be more like 3 launches of the Ares V. And the cryo portion of the propellant could be burned on the initial acceleration out of Earth orbit. Finally, there’s nothing magical about storing cryo propellant for long stretches of time, you just need more insulation. It takes mass, but there’s room for it.

    Having said that, I favor the propellant depot. Launching directly to Mars is feasible though very suboptimal.

  7. Eric Collins says:

    I haven’t had a chance to read through the papers yet, but there are two things which jump out at me when looking at the sketch of their crewed lander. If these questions are answered in the papers, just tell me so and I will look them up for myself.

    First, when and how does the ascent stage separate from the descent stage? The ascent stage doesn’t appear to have any independent landing gear or support, so it would have to remain attached until the ascent engines come up to full power. It just seems that the actual separation would tend to be a bit traumatic (shearing forces as well as repulsive).

    Will they have to make their final approach flying backward? I guess this all depends on the altitude at which the RL-10 shuts down. If there is sufficient descent time remaining, the whole ship could be rotated before the final selection of landing site is made. I’m just wondering how much margin would they have to make that decision. Is 90 seconds enough time to both reorient and pick a safe spot to land?

    Alternatively, they could wait to shut down the RL-10 until they have canceled all of their forward momentum somewhere over the landing site. Then they would only have the final vertical approach to worry about.

  8. Aaron Williams says:

    What if you were to build a separate little hypergolic(or even solid) emergency evac module, whos sole purpose is to sit on the pad at the outpost ready to go in case of emergency. Then the landers would land, offload cargo, change crew and leave immediately rather than stay on the ground. Added cryo storage insulation and hypergolic propellants would no longer be needed on the lander and it could tank only LOX and LH(or other fuel) reducing lander dry mass.

  9. Jonathan Goff Jonathan Goff says:

    Will,
    Too good to be true. By landing with tons of ascent fuel in the tank and a greatly increased payload, you more than triple the landed weight from the original design. This requires a similar increase in the hypergolic landing system thrust and landing gear weight, and generally beefed up structure.

    Well, I don’t know how beefed up the structure needs to be, but yes you would likely need more thrust and heavier landing gear. Even if you triple the dry mass for the LLK, you still get a system that is very capable. I’m still getting 7.2 mT for the round trip lander, and about 17mT for the cargo dropoff lander. It was a first order SWAG, but even adding that in doesn’t seem that much of a loss. We’re still talking about something that can haul almost as much as a cargo LSAM in cargo mode, and could still probably carry 4-6 crew down for sortie missions.

    ~Jon

  10. Jonathan Goff Jonathan Goff says:

    Will,
    Also, a single stage reusable version of their manned lander would have less payload, not more. Although hydrogen has higher Isp than hypergolics, not discarding the descent stage doubles the burnout mass on ascent, which more than wipes out the savings.

    Nope, it actually doesn’t. By eliminating the ascent stage engines, propulsion systems, separation systems, etc. and by going to higher Isp systems, you more than make up for not staging. If you read the LM papers that this was based on, they came to the same conclusion. That for lunar delta-V requirements, staging often doesn’t get you that much performance, and in many cases a single-stage LOX/LH2 lander can indeed compete with a two stage lander that uses something other than LOX/LH2 for the ascent stage.

    ~Jon

  11. Jonathan Goff Jonathan Goff says:

    Eric,
    First, when and how does the ascent stage separate from the descent stage? The ascent stage doesn’t appear to have any independent landing gear or support, so it would have to remain attached until the ascent engines come up to full power. It just seems that the actual separation would tend to be a bit traumatic (shearing forces as well as repulsive).

    The papers go into it. But for my case I’m not proposing a two-stage system. I’m proposing eliminating the ascent stage propulsion system, and going with a purely single-stage concept.

    Will they have to make their final approach flying backward? I guess this all depends on the altitude at which the RL-10 shuts down. If there is sufficient descent time remaining, the whole ship could be rotated before the final selection of landing site is made. I’m just wondering how much margin would they have to make that decision. Is 90 seconds enough time to both reorient and pick a safe spot to land?

    Probably a lot more than enough. 90s is a *long* time. Also remember, there are downward facing windows on the lander, which means that by the time the RL-10 has nulled-out all the orbital velocity, you can actually see a decent distance “behind” you as well as ahead. For a landing site that has decent orbital imagery data, that should be more than adequate to find a good landing site.

    Alternatively, they could wait to shut down the RL-10 until they have canceled all of their forward momentum somewhere over the landing site. Then they would only have the final vertical approach to worry about.

    Ah, should’ve read further ahead. Exactly, bingo.

    ~Jon

  12. Jonathan Goff Jonathan Goff says:

    Aaron,
    Your suggestion would be possible, and might increase the performance by a little bit. OTOH, it also reduces your flexibility a lot. Which way you would go would depend a lot on how important performance ends up being vs. flexibility.

    ~Jon

  13. Will McLean says:

    Jon at #9:
    It’s not hard to calculate the maximum payload of the cargo lander. Their manned lander delivers 2 mt of lunar cargo on the descent stage and a 10.6 mt ascent stage, so the maximum is 12.6. However, the ascent stage provides more than half of the final landing thrust and attitude control, as well as guidance, so you’d need to move those systems to the descent stage. Probably a mt for the propulsion and avionics alone, reducing landed payload by that amount or more.

    And comparing it to the LSAM is comparing apples and oranges. The LSAM could deliver a lot more payload if it didn’t deliver itself and the CEV to lunar orbit.

  14. Will McLean says:

    Jon at #10:
    Not so. You can’t eliminate the ascent engines and propulsion, because both are essential for final landing in this design. You could make the hypergolic tanks a bit smaller, but only by making the hydrogen tanks larger. And tanks holding a ton of hydrogen are a lot heavier than those holding a ton of hypergolic.

  15. Will,
    It’s not hard to calculate the maximum payload of the cargo lander. Their manned lander delivers 2 mt of lunar cargo on the descent stage and a 10.6 mt ascent stage, so the maximum is 12.6. However, the ascent stage provides more than half of the final landing thrust and attitude control, as well as guidance, so you’d need to move those systems to the descent stage. Probably a mt for the propulsion and avionics alone, reducing landed payload by that amount or more.

    Look at the other part of the paper. The robotic section. Compare the main propellant loads for the robotic design compared to the manned design. Notice that the manned lander carries a much lighter LOX/LH2 load. If you look a bit longer, you’ll see the reason is they’re trying to fit the fully fueled manned lander on a single Atlas V Heavy launch. Notice that that constraint doesn’t exist for a lunar depot based system. Instead of only 13mT of LOX/LH2, you can carry 21mT of LOX/LH2 in the tank (a full Centaur load). That makes a huge difference. Also notice that for the robotic lander condition, that the LMK dry mass is only 1mT. That’s the full landing system. The cargo dropoff system I was looking would weigh about 3.25x as much on landing, so naively bumping that up to 3.25mT should give a conservative estimate of the LMK dry mass. That’s still giving me a landed mass in the ~18-19mT range, depending on the delta-V assumptions you make. The cargo dropoff system isn’t going to stay long on the surface. Pretty much long enough to unload, clear the area, and then leave. So it doesn’t need as many features as the manned system does, not by a long shot.

    And you can base the propulsion requirements for the manned lander in a similar manner. Sure, the manned lander is going to have additional systems (the solar panels, heat rejection systems, life support stuff, structures, consumables, maybe regenerative fuel cells, etc). But most of those count toward the actual payload mass, not toward the propulsion system mass. I think you’re looking at the numbers wrong.

    And comparing it to the LSAM is comparing apples and oranges. The LSAM could deliver a lot more payload if it didn’t deliver itself and the CEV to lunar orbit.

    I was comparing the Centaur cargo version to the LSAM cargo version (the ones used for missions where it doesn’t have to brake the CEV into orbit). And this solution has a payload that’s pretty similar. If you went for a higher performance landing system (and/or had a terminal landing guidance system that didn’t require as much hover time), you could pretty much match performance.

    ~Jon

  16. Will,
    Not so. You can’t eliminate the ascent engines and propulsion, because both are essential for final landing in this design.

    So you’re saying the robotic lander can’t land because it doesn’t have several more metric tons worth of propulsion systems? If the robotic lander would work, it’s probably safe to say that a scaled version of the system (possibly using fewer engines but of larger scale) should be able to do a similar function.

    You could make the hypergolic tanks a bit smaller, but only by making the hydrogen tanks larger. And tanks holding a ton of hydrogen are a lot heavier than those holding a ton of hypergolic.

    Nope, they aren’t. Because the hydrogen is being held at a much lower pressure than the hypergolic system, since it uses pumps. Start with the robotic lander numbers, and work from there. You’ll find that there’s a real penalty for going with a two staged system in this case, and that shifting the bulk of the ascent/descent propulsion to LOX/LH2 makes a *huge* difference.

    ~Jon

  17. Brad says:

    Re: multiple types of propellant

    I see that the landers you propose aren’t pure LH2/LOX; they also use a set of hypergolic propellant engines. So the real issue is how much of the total propellant should be non-cryogenic? I still think a manned lander should use storables for the ascent phase.

    Re: manned lander ascent

    A manned lander used for support of a lunar base would probably need to remain 6 months before ascent. A sortie mission might last only 2 days to 2 weeks. In either case boiloff of the cryogenic propellant would be an issue, the question is how much?

    At what point does a cryogenic systems performance lose more from boiloff issues than it gains in increased ISP when compared to storable propellants? Again this is only a question for the ascent phase of a manned reusable lander mission. I would be very surprised if liquid hydrogen fuel was still the best choice at 6 months delay before ascent!

    Re: one-way cargo missions

    This might be the least favorable mission to compare to the ESAS plan, as the Ares V + Altair combo is optimized for one way lunar surface cargo missions. I figure that’s why NASA insisted on liquid hydrogen fuel for the Altair descent stage.

    Unlike a lunar depot architecture, the ESAS plan doesn’t have to split propellant into bite sized chunks and store it for very long periods of time in space, it just burns it all up within a few days. And the cargo has that nice fat 10 meter payload shroud to fit into too.

    I am a big critic of the ESAS plan, but I have to admit that if you are using chemical propulsion and trying to land the biggest possible payload of lunar cargo, than the ESAS plan really shines.

  18. Brad,
    1. The hypergols were left over from the original LM proposal. Their point was that they had off the shelf systems they could use for the lower thrust that were known reliable. It could just as easily have been a LOX/alcohol system like XCOR, Masten, Armadillo, Paragon, and others have used. Or it could even by GOX/GH2 using boiloff from the main tanks. It needs to be reliable, but there isn’t a requirement for it being hypergolic, especially if you’ve designed it to handle engine-out situations.
    2. Why hypergolic for the ascent? Doing that forces performance losses all the way back through the system. Since your ascent system now weighs more, it forces your landing and your descent systems to weigh more. If you have an architecture where you have rescue capability, do you really need the hypergols? Also, the manned design was based around a Dual Engine Centaur configuration, so you have some engine-out capability built in. I just don’t think that having a completely different propulsion system is really going to be that good of a deal. It adds a whole additional list of failure modes, and since most hypergolic systems are pressure fed as well, adds a lot of extra weight in tanks and pressurization systems. Using it just for landing when you only need like 150-200m/s dV is one thing, using it for another 2000m/s after that is another entirely.
    3. If you’re going to do visits of longer than 14 days to just about anywhere on the moon you need nighttime power. Which implies that the manned lander would have a regenerative fuel cell system. Once you have the cryocoolers you need to make that work, boiloff is no longer a concern. Cargo landers, that probably wouldn’t have that kit installed, would be limited to shorter flights. But if you have a method of chilling the cryogens down in the first place, keeping them from boiling off is doable. At night, you use the boiloff gasses to run the fuel cell.
    4. It may be what the ESAS architecture was made for, but even in that situation a depot-centric option still can compete. Depending on assumptions, you can almost get as much payload to the surface (~19mT vs 21-23 for Altair), and unlike the Altair, that payload isn’t 10m off the ground, needing a crane to get down. Sure, it’s limited to payloads 5m in diameter or smaller, but have you ever measured that out?! 5m diameter is huge! And it’s right there near the surface once you land. At a level where lowering it do the ground is easy. Possibly as simple as letting some gas/hydraulic fluid out of the landing gear until the cargo is on the surface, letting it go, and then reinflating the landing gear and maybe backing away a tad before takeoff. You’re already talking about a payload in the range of a Bigelow Module (possibly stripped down of now deprecated hardware like docking adapters and propulsion systems).
    5. Another important consideration is that a depot-enabled system is scalable. All of these numbers were for a lander derived from the standard Centaur stage. ULA has been working towards implementing a new ACES stage (which would likely be done if a non-ESAS approach were taken) that can grow up to 4-6x the propellant loading of a stock Centaur. That would mean being able to land payloads all the way up to 150mT chunks if you really had the need!
    6. Lastly, the ESAS lander already does split up the cryogens into lots of little tanks. Have you seen how many they’ve been talking about! Talk about a CFM nightmare!

    ~Jon

  19. Tom D says:

    Jon,

    These numbers look pretty good. Thanks for putting the time into it. The possibilities for the near-term utilization of space are looking better all the time, if you are willing to do some assembly in space. The assembly of ISS has shown that this is not too tough.

    NASA is actually putting some money into long term (6 months) storage of cryogens on the moon’s surface. It is not really that big of a problem with insulation and cryocoolers. Also, there are milder cryogens like methane and subcooled-propane that may actually be better fuels.

    More and more it looks to me (an experienced aerospace engineer) that building an infrastructure between the earth’s surface and the moon is not as difficult as was once thought. Unfortunately, as long as it is relegated to NASA’s bureaucracy, it will not be done.

    Jon, the systems engineering trade studies that you have detailed over the years could make a good book, maybe even a text book for how to do system-level engineering trade studies. Keep up the good work!

  20. Will McLean says:

    Jon at # 15:

    The manned lander can’t carry the full of a the robotic lander because it has shorter tanks and correspondingly lighter structure.

    If landed mass is 17mt, payload to the surface is a *lot* less than the LSAM cargo version, once you net out ascent propellant and the mass of the lander. And the LSAM is expected to perform LOI for itself

  21. Will,
    The manned lander can’t carry the full of a the robotic lander because it has shorter tanks and correspondingly lighter structure.

    The difference in mass between the shorter tank and the normal length one is 200kg. Once again, since the lander doesn’t have to be launched fully fueled, there’s no reason for it to use the shorter tanks. Using the longer ones makes more sense.

    If landed mass is 17mt, payload to the surface is a *lot* less than the LSAM cargo version, once you net out ascent propellant and the mass of the lander. And the LSAM is expected to perform LOI for itself

    The 17mT was the net payload to the surface in a dropoff mission, not the total landed mass. If you did just a one-way cargo mission (without bringing the lander back to LLO), it could land almost 29mT. Which is quite a bit bigger than the ESAS LSAM can land in a similar mission. And part of the point of using a depot-centric architecture is that it doesn’t matter if something has to perform its own LOI maneuver. It tanks up after that before moving on to the next stage of the mission.

    ~Jon

  22. Exploration Fan says:

    Karl,
    “I understand the actual number would be more like 3 launches of the Ares V. ”

    I’m sure that there are ways using pre-deployment of a base/supplies, advance propulsion and insitu resources that would allow a Mars mission with the equivalent of 3 Ares V launches, ~420 T. However, NASA currently believes that they need >170 T to support a few day lunar mission with 4 people. Trying to support a 3 year Mars mission with a much greater energy requirement with who knows how many people for 2.5 times the LEO mass seems unlikely. I do truly hope that we make the technical progress that allows Mars missions for this sort of launch mass investment. The cheepest way to reduce launch costs is to reduce the required launch mass.

    “Having said that, I favor the propellant depot. Launching directly to Mars is feasible though very suboptimal”

    I fully agree for a number of reasons, not the least being if Ares V is only flown 3 times every two years this will be an extremely expensive rocket. Although I hope that we can afford concurrent lunar and Mars programs I doubt that will fit any future NASA budgets. Thus once we finally are off to Mars I think it likely that we have only a single Mars mission per cycle unless launch costs are dramatically reduced.

  23. Exploration Fan says:

    To finish my thought, depots offer the potential of reducing both launch requirements as well as launch costs.

    I a Mars mission starts from L1 than advanced propulsion can be used to fuel the L1 depot and place the Mars bound space station in L1. The crew is launched to L1 just prior to mission departure. The use of advanced propulsion to move the majority of the mass out of Earths gravity well significantly reduces the total mass required.

    As many have said, competition for refueling the depot is very likely to result in significant launch improvements, substantially reducing launch costs.

    I believe that this competition and advanced propulsion will be required before we have a significant crewed exploration effort.

  24. Will McLean says:

    Jon at #21:

    I don’t think the LM papers justify anything like a 29mT one way payload for a Centaur derived lunar lander.

    The manned lander has an initial mass in LLO of 31. 6 mT and 12.8 mT cryogenic propellant, and landed mass of 18.3 mT.

    Increasing the propellant to 21 mT and the other values in proportion we get 51 mT in LLO and landed mass of 30 mT. Does the inert mass of the lander equal one mT? I doubt it.

    Robotic Centaur is 2.5 mT, plus .3 for two more RL-10s to preserve the same t/w ratio as the manned lander, plus .8 mT for the EDMK, plus 4.5 mT for LLK scaled to landed mass, plus residual and reserve main and secondary propellant scaled from the LM manned lander to 3.9 mT.

    Subtracting these from the landed mass gives a payload of 18 mT from LLO to the lunar surface. Nothing wrong with 18 mT on the lunar surface, mind you. But it’s a lot less than 29.

  25. Will,
    I think you’re making some incorrect assumptions.
    Let’s start with the rocket equation: DV=Isp*g*ln(MR)
    Rearranging the terms: MR=gross mass/dry mass = e^(DV/(Isp*g))

    Using 451s (the Isp of the RL10s used on Centaur), and 2000m/s of delta-V, you get a mass ratio of 1.57.

    MR=(propellant mass + burnout mass)/(landed mass) = 1+ propellant mass/burnout mass

    Therefore, prop mass/ burnout mass = .57, or prop mass/.57 = burnout mass. If you start with 21mT of propellant you get 36.8mT of landed mass. Not 30mT. Even using your assumptions of dry mass requirements, that puts you up to 24.8mT, which is comparable to the ESAS cargo-only lander.

    As for the last little bit, even granting the need for more RL-10s (I don’t have numbers readily available on much T/W ratio you really need on the lander main propulsion to avoid gravity losses), the 3.9mT of residuals is crazy for a one-way lander. The only reason why they had as much residuals planned into the manned lander was for the fuel cells. Sure you have to plan in some residuals and some margin and such, but typical residuals for stages like the centaur are less than .5% (ie .1mT). And landing residuals for the secondary propulsion system can also be factored in by just providing more delta-V than you think you need…I think based on an understanding of what they were trying to do, you could keep these two numbers in the .2-.4mT range for a one-way lander. If you only budget .4mT for the residuals/reserves, you’re add back 3.5mT to the lander. Which puts it up to 28.3mT.

    So yeah, depending on what assumptions you make it should be either better than the ESAS cargo-LSAM or much better.

    ~Jon

  26. Will McLean says:

    Jon at #25:

    If you are correct, the guys at LM that designed their lander don’t know their business, since their landed mass is much too low for 2000 m/s on the manned lander.

    Perhaps they are assuming something more than 2000 m/s? Maybe a teensy bit extra for seeing a boulder where they planned to land and needing some margin to deal with that? I supose it could even be as much as 2400 m/s. What did they allow for on the Apollo LM descent stage?

    The residuals are surprisingly high in th LM design. I wonder how much of that is driven by the necessity of keeping the balloon tank pressurized.

  27. Will,
    Or maybe they were designing to different constraints? By trying to fit the whole mission into something that could be launched by a Atlas V Heavy, they had to go with a smaller, less efficient Centaur tank. By going with a full-sized Centaur tank, topped up in lunar orbit, you’re totally changing the dynamics of the situation.

    I went back and checked the numbers for the two landers, and it turns out that both of us were off. If you run the numbers on Tables 1 and 2, you get both of them having a landing DV of about 2250m/s give or take. Historically from the Apollo missions they used an average of 2050m/s for descent and landing. That extra 200m/s provides over 2 full minutes worth of hover time. While I could quibble with you on the need for that much hover time, rejiggering the numbers using the higher delta-V requirement *and* the heavier 3-RL10 lander (as per your suggestion) *and* the much bigger LMK (5x bigger because of 5x the landed mass), you still get over 25mT worth of cargo on the surface.

    So, no the LM guys aren’t stupid, and yes even if you make pessimistic assumptions a Centaur derived cargo lander could put about the same payload to the lunar surface as the Cargo LSAM.

    ~Jon

  28. Will,
    I apologize if I’ve been coming off kind of rude. I still need to work on disagreeing without being obnoxious. I still think I’m right on this one, but that doesn’t excuse rude behavior.

    BTW, here’s a good link on statistical numbers for the Apollo lunar landings:
    http://www.retro.com/employees/gherbert/Space/LunMil2k/Lunar_DeltaV_PlusStats.html

    ~Jon

  29. Will McLean says:

    Jon:

    Please show how you arrived at the 25 mT payload estimate.

    Also, note that several sources give the theoretical deltaV of the Apollo descent stage as 2470 m/s, which suggests that they indeed wanted much larger reserves on a landing mission than a typical satellite launch.

  30. Will,
    I have the spreadsheet at home, I’ll upload it tonight if I get a chance. But basically, I just plugged in the 2.8mT (for 3 RL-10 engines) centaur weight, 21mT centaur propellant, .8mT EDMK, 5mT LLK with 5mT of LLK propellant, and an overall delta-V of 2250m/s, which is worse than 5 sigma past the mean used for the Apollo landers. I’m sure if you make the reserve requirement high enough, and add enough weight penalties you can make it look a *little* worse than the ESAS LSAM. What delta-V does the ESAS LSAM use though for descent? Digging around on google they were actually using only ~1900m/s of Delta-V. So, should we do an apples-to-apples comparison, or should we use the most pessimistic numbers for my concept and the most optimistic numbers for the ESAS LSAM?

    ~Jon

  31. Will McLean says:

    Jon:
    Apples to apples, by all means. The latest figure I’ve been able to find on the ESAS descent stage is 7.4 mT dry mass, not including payload, and 55 mT at TLI for the cargo version. What does that give you for payload to the surface and total mass in LLO? (I’m guessing about 19 and 45)

  32. Will,
    The numbers I’ve been able to find for the LSAM indicate that the LOI burn is about 1100m/s, and the descent orbit insertion/descent/landing delta-V is about 1900m/s. That would indicate (using 451s again for the Isp) a mass ratio of about 1.97. That means that of the 55mT, only 27.9mT actually make it to the surface, and of that only 20.5mT is actually cargo.

    If you use a 1900m/s base delta-V requirement for landing, with another 90s of hover time (140m/s delta-V) from the secondary propulsion, the Centaur based one-way cargo lander could land over 29mT.

    The big difference in this situation is that the Centaur lander isn’t doing the LOI for that big of a payload, and it starts in lunar orbit with a full tank of propellant. That payload would either need to be delivered in smaller pieces using smaller Centaur based transfer stage, or would need a Wide Body Centaur/ACES or EDS-based transfer stage to deliver it to lunar orbit.

    If you could tank up the LSAM in lunar orbit, it could deliver a much bigger payload. But that’s kind of my whole point with this article–having a lunar propellant depot gives you far more flexibility and capability than an HLV only non=depot solution.

    ~Jon

  33. Exploration Fan says:

    In the above discussion on delta V’s and lunar performance one is over looking one of the main benefits of the Centaur derived lander concept, granted the focus this thread is on the benefit of a depot not the type of lander. Centaur’s propulsion system already exists and is flight proven:
    – Centaur’s light weight tank set is already in production, although I’m guessing that for the lunar lander the tank walls could be much thinner, further reducing weight. Altair isn’t even at SRR, so what would its weight really be? Just look at Ares and Orion weight increases to be skeptical of current estimates.
    – Centaur’s propellant conditioning is already flight proven. How does one condition the multi-tank propellants on Altair???
    – Centaur’s tank heating is a known. Altair’s multi-tank design will make tank heating huge.
    – Propellant residuals are already proven. What will be the residuals and differential pull through on Altair.
    – Mixture ratio control and propellant mass gauging are already proven. Once again, Altair’s multi tank design makes this a challenge.
    – Centaur’s feed system including chilldown are already known. The manifolding and long feed line routing of Altair will result in large unknowns and likely much larger masses.
    – RL10 conditioning is already proven for Centaur. The imbedded RL10 on Altair will result in a cold engine that without a heating system agravates RL10 start.

    This list can go on and on, but starting with Centaur hugely reduces development risk, system weight, residual mass and uncertainties providing a great efficient foundation for the lander propulsion system.

    On the other hand, ULA knows squat about landers, terminal guidance, etc. I think that it would make a lot of sense for ULA to simply provide the basic propulsion system while NASA or Lockheed turn it into a lander.

  34. Brad says:

    Hi Jon,

    Re: reusable lander ascent propellant

    I have just carefully re-read the LM documents, and it seems to support my conclusions about ascent propulsion. Storable propellants are a better choice for ascent propulsion.

    First off the penalty of storable propellants for short missions isn’t very great. Figure 1 of the “Lunar Lander Configurations…” document shows little mass penalty for storable propellant ascent propulsion. For a two stage system (presumably of no more than 14 day mission duration) the penalty is about 12% (45 tons compared to 40 tons). And that penalty gains the reliability and finer flight control that storable propellants provides, important qualities, even more so for a manned lander.

    Secondly the penalty for long-term storage of cryogenic propellants is spelled out pretty convincingly. Even if all the proposed boil-off mitigation systems work as advertised the propellant loss over six months is 18% (0.1% per day). In the same document the description of the all-cryogenic propellant “Concept 3: reusable single-stage lander” states such a lander is likely limited to shorter mission durations.

    It seems to me the problem with an all-cryogenic propellant reusable manned lander is it falls into a gap between needs and capabilities. The commitment to a lunar depot infrastructure that the reusable lander needs implies lunar operations that include a permanent lunar base. And an all-cryo lander is most capable of sortie missions instead of long duration lunar base crew rotation missions.

  35. Chuck2200 says:

    Jon;
    Have you considered the possible use of drop tanks on the lander? You are correct that the numbers support the performance of the single stage concept, but that could be bettered if a significant percentage of the descent propellant was housed in tanks that the crew detached and left behind on the surface. This eliminates the mass of unused tankage during the ascent.
    Chuck

  36. Jonathan Goff Jonathan Goff says:

    Chuck,
    It isn’t crazy, but the benefit might not be as much as it seems. The Centaur tanks themselves weigh less than 1 ton, though the lander tanks might be heavier. The big problem though is that as soon as you do that, you now have to outfit extra tanks in zero-G, and you have to be shipping a decent amount of hardware along for each mission.

    What I could see making sense though is adding drop tanks for bigger missions. Like say you want to land a really big payload, you add some extra drop tanks that you leave behind on the surface. Not sure though. Would have to think about it.

    ~Jon

Leave a Reply

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