Centaur Based Earth Departure Stage

A couple of days back, when I was doing some of my preliminary thinking about the TeamVision architecture, I did some back-of-the-envelope calculations trying to see what kind of lunar transportation architecture could be done with off-the-shelf (or nearly off-the-shelf) hardware. One of the single biggest problems with the current NASA architecture is that almost everything is being done with custom developed, NASA-specific hardware. Almost all of the money budgetted for Exploration over the next decade is going to developing launchers and transfer stages, and in retaining the standing armies for those systems. Between development, workforce retention, infrastructure upgrades, and all other related expenses, you’re probably talking over $50B out of the $67B that NASA will spend on exploration over the next dozen years. Only a tiny pittance will actually be spent on lunar-specific hardware like the LSAM, the unmanned exploration hardware, and developing stuff to actually do once we get to the moon. Every year that the architecture is delayed will end up eating up over half of the budget just keeping all the people on the payroll!

An architecture that relied a lot more heavily on existing launch vehicles and upper stages would allow for much more actual exploration to be done.

Digression: Lunar Surface Rendezvous
So, I started poking around, looking at various combinations of stock, or nearly stock boosters and other hardware. The first conclusion I came to was that if you went to a two-man architecture and relied heavily on Lunar Surface Rendezvous type options you could greatly reduce the required IMLEO of the mission, and thus greatly reduce the number of launches needed to support it. I’ve already gone into LSR a little bit, and will definitely go into it in more detail in the future, but here’s the basic gist of Lunar Surface Rendezvous architectures. Basically, if you have to launch everything for a given lunar mission on a single lunar transfer stack, you will run into limits very quickly on how much you can deliver to the lunar surface.

One of the lessons not learned from the Shuttle was that if you try to make your “space truck” into a space Winnebago, complete with housing facilities for a bunch of people for several weeks, and insist on being able to carry a lot of cargo to boot, you end up with a rather bloated system. LSAM is headed back down the same road. In the end you end up getting the worst of both worlds–an overly expensive, overly bloated transportation system, and extremely cramped and inefficient living conditions. NASA goes on and on about how “roomy” the CEV and LSAM are, but the reality is that for all the weight being thrown into them, they don’t even have 1/4 of the space of a single Bigelow Sundancer module, in spite of probably expending more of their mass budget towards being roomy than the weight of such a module. It would be much better and cheaper, if you want to explore a site for a while, to preland a Sundancer (or Sundancer derivative) module, then send the crew. Unlike the LSAM which supposedly doesn’t even have an airlock, and can only support a crew of 4 for about 7 days (28 man-days in other words), a Sundancer and a bare-bones 4-person lander could support the same crew for probably closer to several months. And you don’t have to carry anywhere near as big of a lander, which allow for a much smaller IMLEO, and a much simpler architecture.

Even NASA intends to eventually do Lunar Surface Rendezvous, but only after they’ve designed their bloatware LSAM lunar winnebago. In other words, like with the shuttle, they’re bound and determined to always find a way to get the worst of both worlds. If they went with LSR from the start, and actually took advantage of the options that opens up, they could make their lives a lot easier.

Anyhow, getting back towards my original point, I started looking around at what existing or near-term upper stages might be available for a lunar project. The two main ones I was considering early on were Centaur V-1 and the Delta IVH upper stage. The Delta-IVH upper stage uses the heavier, higher Isp version of the RL-10 engine, and has a lower mass ratio due to not using a common bulkhead, and a few other things. But its mass ratio was still pretty good, and more importantly, it had more propellant. So I spent some time looking at it, and it really looks like you could do a lunar mission using two D-IVH launches, or a D-IVH launch and two Atlas V 401 launches. It wasn’t perfect, and didn’t include a lot of the features you’d really like in such a system, but it was off-the-shelf, and not crazy. However, the margins were a lot lower than I wanted, even with a rather minimalist two-person architecture, so I kind of back-burnered the idea for the time being.

Wide Body Centaur
Then I re-stumbled upon some work that Lockheed Martin is doing as an upgrade to their existing Atlas V fleet–the Wide Body Centaur (aka ICES, the Integrated Cryogenic Evolved Stage). You can find some details on the Atlas-Yesterday, Today, and Tomorrow page Lockheed put up a few months back. Apparently, the Centaur V-1 and V-2 flying with Atlas V today are very, very similar to the same ones that first flew. They’ve upgraded the avionics over the years, gone to better manufacturing processes as they came out, upgraded the RL-10s over time, but never really changed the size of the basic Centaur stage very much over the past 30+ years. However, newer manufacturing techniques like Friction Stir Welding of thin aluminum tanks have come out, as well as the need for longer duration missions, so a couple years back Lockheed started designing the next iterative improvement on their venerable design.

The three biggest changes are the size, the structural layout, and the tank material. Unlike previous Centaurs which IIRC were all in the 3-4m range, the newer WBC will use 5m diameter tanks. Since area of a circle goes with the square of the linear dimension, that means an equal length WBC Centaur stage will hold over 1.5 times as much propellants. Additionally, they are going to invert the direction of the common bulkhead. This requires the LH2 tank to be run at slightly higher pressures, but will greatly reduce the heat transfer into the LH2 on-orbit, as well as simplifying the sump. The third big change is going from the previously used ulta-thin wall stainless construction to a friction stir welded thin-wall aluminum construction. They’ve done a lot of manufacturing development work over the past two or three years (some on their own dime, and some in conjunction with NASA), and now feel confident in welding aluminum tank sections as thin as .04in or thinner. The upshot of this last point is that they can improve their already impressive 91% propellant fraction to nearly 95% in some configurations (probably depends on number of engines, and the duration of the mission–longer duration missions require some additional goodies like MLI and solar panels). The other upshot is that this makes it very easy for them to make different sized tanks. All you do is friction stir weld-in more or less barrel sections in the LOX and LH2 tanks.

These stages range in size from about 1.5x the size of the current Centaur, all the way up to 6x as big. They have a common thrust structure which can accomodate 1-6 RL-10s, depending on size and thrust requirements. They can be configured for short durations, or very long durations (greater than a year) depending on which options you add in (things like star-trackers, solar panels, etc). The best thing is that if they develop the WBC, it will be used on every single Atlas V flight after this point, which means they will get a lot of experience with the system soon. It also means that the development cost gets spread out over many more launches, and the incremental cost of a stage will also be tolerably low. There’s also the benefit that the WBC does not require any new launch pad infrastructure.

One more thing that I’ve mentioned previously, is that Lockheed now has a lot of experience with “settled” propellant technologies. Basically, by using the vented-off hydrogen run through a nozzle, they’re able to generate enough continuous acceleration to always keep the propellants settled at the bottom of their tanks. This takes most of the difficulties out of on-orbit propellant handling. Basically, “settled” propellant transfer ends up being very similar to terrestrial propellant transfer, instead of being the complicated mess that zero-g transfer tends to be. Lockheed still needs to actually finish reducing this concept to practice, but they have a very good point about how near-term feasible that is.

Anyhow, now that I’ve mentioned some of the benefits of the WBC design, I’ll just touch on two potential applications.

Wide Body Centaur LTV
The first interesting application was the one I started discussing at the start of this article, using the WBC as a lunar transfer vehicle. The big problem I mentioned with the existing Delta-IVH upper stage, and the Centaur-V1/V2 was that they were just a little too-small for use with a manned lunar architecture. You could do a 2-man architecture, but it would require either really minimalist landers and spartan crew capsules, or using two stages to get from LEO to LUNO, or require using the crew capsule or lunar lander to perform one or more of the burns. Using a WBC as a lunar transfer vehicle however would be a lot easier, due to their much larger propellant capacity.

Depending on if they also ever get around to fielding the Atlas V Heavy, you might even be able to do a “1.5 Launch” architecture. An Atlas V Heavy would launch putting just the upper stage with a bunch of excess propellant into orbit. Then, an Atlas-V 401 would launch the crew capsule, the lunar lander, and its own upper stage into orbit. The upper stage would be kitted out with slightly oversized tanks (1.8-2.0x current Centaur-V sized tanks instead of just 1.5x), the long-duration avionics suite, and some propellant transfer hardware. The manned lunar stack would rendezvous with the tanker, fill up it’s own WBC stage, and then do some final on-orbit checkout (waiting for the correct timing for a lunar departure). The lunar transfer WBC would perform both the Trans Lunar Injection burn as well as the Lunar Orbit Insertion burn placing the stack into Lunar orbit. The two-man lander and capsule would then descend to the surface leaving a partially fueled WBC in orbit for the duration of their mission. When they’re done, the lander would drop the crew capsule off on orbit, and the WBC would bring the capsule back to LEO–using 100% propulsive braking.

So, at the end of the mission, you’d have a reusable lander in Lunar orbit waiting for refueling and reuse, and you’d also have a lunar WBC sitting in LEO waiting to be reused. Based on the test results P&W claim, you could probably get 20-25 flights out of a given WBC before the engines wear out.

The same hardware used in a one-way cargo mode could put a Bigelow Sundancer module on the surface using only two launches.

Heck, maybe you could even steal a play from CSI’s “Soyuz Around the Moon” project, and buy a used WBC on-orbit. Most satellites end up weighing less than the maximum payload capacity of the launcher they fly on, so a lot of the times the launchers will launch ballast along with the satellite to keep the mass properties correct. What if the “ballast” for a launch were a lunar-mission kit for the WBC (and enough propellant to put it back into an LEO parking orbit if the satellite was being dropped off in GEO)?

The coolest thing about an architecture like this is how flexible it is. While you can start out with a 1.5 Launch 2-man architecture, you can eventually work your way up to much, much bigger missions, since the maximum size for the Wide Body Centaur contains about 290+ klb of propellants.

Wide Body Centaur as EDS
Which brings me to my second idea. The Wide Body Centaur would be an ideal “Earth Departure Stage”. It has more than enough performance, it has a lot of “long duration” features, and most importantly would be commercial-off-the-shelf hardware. Why blow extra billions developing a NASA-specific EDS, when a stock WBC with a lunar mission kit (solar cells, star trackers, and extra MLI) could do the job? Even if NASA insists on building the Shaft and the Ares V, using a WBC for their Earth Departure Stage would have several benefits:

  • Save $5B+ on development costs
  • Cut a year or two off of the current manned lunar return schedule
  • Use hardware that would have a flight heritage by the point lunar missions start
  • Save several hundred million a year on fixed costs associated with a NASA-specific EDS
  • Have a design that is capable of accomodating even bigger missions in the future (the largest of the WBC configurations is about 50% bigger than the EDS)
  • Have an EDS that uses flight-proven off-the-shelf propulsion systems, in an engine-out capable design
  • Have an EDS that has a much lower marginal cost since it will be using hardware in common with every single Atlas flight
  • Have an EDS that can store propellants for very long durations without requiring a complicated and expensive Cryocooler
  • Be able to field test EDS functionality even before Ares V is in operations

All in all, the WBC could meet or beat every spec that NASA needs out of the EDS. Congress should encourage Lockheed to develop this commercially, on their own dime. They could do this without spending any money, or commiting future congresses by stating in law that if Lockheed fields its WBC design (or if a competitor fields an equivalent or superior design) before a certain date, that NASA, will not be allowed to develop or operate its own EDS. Congress shouldn’t have to do this. They’ve already passed authorization legislation that requires NASA to purchase commercial hardware instead of developing their own stuff whenever possible, but NASA always seems to find an excuse.

The Wide Body Centaur technologies that Lockheed has been developing over the past few years have a lot of potential. Even if NASA insists on doing Ares I and V their way, everyone would benefit from encouraging Lockheed to field the WBC sooner rather than later.

One of the single biggest benefits is that if the WBC gets fielded, then we have a backup option for lunar exploration even if Ares I and V run into issues. Backup options, especially off-the-shelf options that don’t cost much if any NASA money to develop, would allow NASA to take the risks it needs to without having to worry that all will be lost if anything goes wrong with Plan A. Not to mention that by having a backup option, it will keep other contractors more competitive, because they know that their gravy train is not secure unless they really deliver a good, robust, and cost effective solution.


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 the founder and CEO of Altius Space Machines, a space robotics startup that he sold to Voyager Space in 2019. Jonathan is currently the Product Strategy Lead for the space station startup Gravitics. 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 the founder and CEO of Altius Space Machines, a space robotics startup that he sold to Voyager Space in 2019. Jonathan is currently the Product Strategy Lead for the space station startup Gravitics. 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.

11 Responses to Centaur Based Earth Departure Stage

  1. Anonymous says:

    How do we find if the reusable engine is worn if we do not return it to a shop like SSME? The question is largely prompted by looking at pictures of burn-throughs at Armadillo’s media site, but still…

  2. Brad says:

    “…and the WBC would bring the capsule back to LEO–using 100% propulsive braking.”

    Halt! Stop! Just wait one minute. You are going to have to walk me through the numbers for that to be credible. Full propulsive braking after TLI and after LOC of the entire lunar stack? Just how small a two-crew payload and how big a WBC are you talking about here?

  3. Ed says:

    I am all for lots of options. Going with only the Ares 1 and 5 means the use of NASA-only launchpads, but being able to use Atlas, Delta, Falcon etc means that a great many launch pads could be used.

    It seems to me that NASA is still ignoring the most important law of engineering: Murphy’s.

  4. Anonymous says:

    I certainly like the sound of a RLV disguised as an ELV. But how are you going to check out the systems and do maintainence on your lander, which is going to be exposed to lunar dust and the like.

  5. Jon Goff says:

    Checking out the lander would probably be best done on the lunar surface. I wrote up some ideas a couple of blog posts ago, have a few more that I haven’t written up, and some that get into MSS-proprietary ideas.

    But basically the whole idea that you have to tear an engine completely apart after each use in order to make sure it’s still ok is…archaic IMO.


  6. John Carmack says:

    >The question is largely prompted by
    >looking at pictures of burn-throughs
    >at Armadillo’s media site, but

    Do note that none of the failures we have seen are really statistical failures, they are just design failures when experiencing new regimes. We haven’t really had a design that has worked flawlessly at a particular performance point fail aparently randomly under the same circumstances.

    We are constantly changing our feed pressures and chamber pressures as we experiment, but XCOR’s experience with constant conditions seems to bode well for the reliability of rocket engines that aren’t at the very bleeding edge.

    John Carmack

  7. Anonymous says:

    Reusing the WBC transfer stage isn’t important. In serial production they wouldn’t cost more than about $50m per unit.

    The most expensive and heaviest component of the system is the lander, it would make a great deal more sense to make it reusable and leave it in low lunar orbit. Reusable hypergol tankers could be cycled ahead of every mission by solar electric tug.


  8. Anonymous says:

    There is nothing keeping the LSAM, or more specifically the LSAM descent stage, from being reused. The RL-10 has a huge runtime and is well overdesigned for its intended purpose. The RL-10 can also be used with LOX/Methane (All of this work was done in the 1960’s) as well.

    Look up the history of the RL-10. Its original development cost ~$3 billion dollars in 1960’s money.

    We ran some numbers and if you refuel the descent stage on the surface it can put ~25 metric tons into orbit and return to the surface.

  9. Anonymous says:

    As one of the originators/leaders of the WBC design it is a pleasure to see someone follow in almost precisely the path of thinking that we have been following for nearly four years. We revealed some of the concepts in the Lunar Lander paper presented in September- see the LM website. What is next?- you can easily replicate the math: Think Lunar orbital rendezvous combined with surface “rendezvous” or caching. Do not assume the LSAM has to go with the crew, and consider the ease of cryostorage at the moon vs LEO. And remember: “empty” tanks are REALLY valuable in the right places- not junk. Assume ascender reuse, LLO propellant storage, commercial propellant delivery FOB LLO, LEO supplied by MANY vendor teams. Build confidence in autonomous R&D in arbitrary orbital geometries including L1 and elliptical. This is mandatory for Mars. Lunar bases can be commenced at least 5 years earlier than NASA plans with a $20B savings in launcher development costs.

    Wanna go to Mars? Think 3.5X WBC working in axially conjoined pairs. Same propulsion core, several variants: ISRU/Power, Transit Habitat, Surface Habitat, Orbital Propellant Storage/Synthesis, Crew LMO Ascender/Descender, Propellant/Water LMO Ascender/Descender. Propellant FOB L1 cached and vehicles checked out in full-up flight configuration for months prior to depature. Think L1-initiated close-formation departure with powered gravity assist at perigee. Mars Insertion: power+aerobrake, circularization:aerobrake, landing: 90% powered using the DTAL lunar-proven architecture. The extensible WBC supplies all the tools and confidence for exploration worthy of the name (32 crew initial visit). Watch for upcoming papers. Real exploration that will make Americans proud of their space program again is within our grasp- if we dare to “think different”.

    ps- Imagine how handy ISS would be if it WASN”T in LEO. Now picture what it takes to move it and tell me what that looks like.

    specific imp

  10. Jon Goff says:

    specific imp,
    I was trying to get in contact with someone on the WBC team. Could you ping me on my gmail account (jongoff at gmail dot com)? Thanks,

  11. Martijn Meijering says:

    If you haven’t read it already, Taming Liquid Hydrogen (http://history.nasa.gov/SP-4230.pdf) is a very readable account of the history of the Centaur. Based on an idea for a lunar flyby using J-130+DHCSS+Orion proposed by the the DIRECT people, I had become interested in Centaur/DHCSS as makeshift earth departure stages. But reading about the history of the Centaur I realised that it wouldn’t really be a makeshift EDS, that’s what it was originally designed for! The Surveyors, Voyager and lots of probes in the seventies were actaully launched with Centaurs. It’s a bit small, but with L1 rendez-vous and refueling, you can make it work, even with Orion.

Leave a Reply

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