Here’s a quick thought on a way to use EELVs for launching Orion that I was thinking about tonight. One of the reasons why the CEV is so big that it’s hard to launch on an existing EELV is because of the amount of service module propellant. Basically, the CEV is sized to provide nearly 1800m/s of delta-V. Now, you don’t actually need most of that for an ISS mission (except they use it as a contingency to provide enough delta-V during an abort over the North Atlantic in winter to make sure you can put the CEV down in a safer location), but you do for a lunar mission.
I don’t have the latest numbers, but some numbers I’ve seen for the CEV (from a presentation Pete Worden gave a few months back about using Orion for an NEO mission) put it at about 24.7klb dry and 20.5klb propellant. In other words almost half of the CEV launch mass is NTO/MMH for the service module. Once you factor in the much bigger tanks and engines for such a service module, the CEV isn’t really all that much heavier than a Dragon capsule for instance. But once you include all that lunar return propellant, you now have a capsule that’s so big that it’s hard to loft on all but the biggest existing launchers. In fact, due probably to the switch back to hypergols from the LOX/CH4 suggested in ESAS, right now in order to fit its mass targets, Orion is having to shed a lot of the redundancy and functionality it needs in order to perform its mission safely.
Right now, as I understand it, the current concept of operations is that the Ares-I puts the orion into a suborbital trajectory with most of the velocity needed for orbit. The Orion then provides the circularization burn to put itself in orbit.
The question I had was, if you are already using the Orion somewhat like a third stage, what if you actually did use it as a third stage? While you couldn’t launch a fully-fueled Orion into LEO on anything other than one of the EELV Heavies, transfer of hypergolic propellants is now a demonstrated capability, even in the US! So, I was curious how small of an EELV you could use and still get Orion into orbit (with enough propellant left for rendezvous and docking maneuvers), if you assumed that for lunar missions you could tank-up the CEV on orbit.
I happened to have some mass numbers from previous conversations about the Atlas V Phase 1 and 2 concepts that ULA did, so I put together a spreadsheet. First, I took the payload numbers from ULA’s site, and the Centaur and CCB mass numbers I had sitting around and estimated the total delta-V to LEO for the stack. Then, I took the rough numbers for Orion I had above, and treated it like a third stage. The Orion payload adapter was added as a dry weight to the Centaur stage, and the LAS was added as a dry weight to the CCB stage (since it would likely be tossed immediately after stage separation). As you can see, a 1.5x Phase 1 Atlas V could do the job, leaving a bit of performance for margin and maneuvering delta-V. Of course, if someone were serious about doing this, you would likely oversize the Phase 1 upper stage a bit to provide extra margin, to relax the constraints on Orion. Call it a 1.6x or 1.65x (referring to 1.6 or 1.65 times the propellant load of an existing Centaur).
For ISS missions, you wouldn’t even need to top the propellant off, as you’d have enough leftover performance after arrival for rendezvous, docking, and deorbiting. For a lunar mission, you would need most of a second launch worth of propellant, but that propellant would be a very low cost cargo (and could also reasonably be launched by other lower-cost commercial suppliers). By using the service module as an upper stage, you would get to test out the engine thoroughly on the way to orbit, which might reduce your risks of unforeseen problems cropping up on your way home from the Moon. Also, if for some reason the upper stage fails (though it has engine-out capability now unlike the Ares-1 US), you still have the extra propellant for contingency maneuvers to avoid landing in the North Atlantic.
The single stick Atlas-Vs meet almost all of the old NASA human rating requirements (and NASA had to lower their standards enough for Ares-I to pass that most of the few human rating requirements that Atlas-V didn’t already meet are no longer there), and most of the remaining requirements are improvements that ULA wanted to do anyway for their Bigelow collaboration. The nice thing about a Phase-1 Atlas V is that the CCB is unchanged from the existing Atlas-V, and even the upper stage changes all have prior design heritage. The Titan Centaurs used the wider body diameter that the Phase 1 design would use, most earlier Centaurs were dual-engine, and between Lockheed and Boeing’s contributions to ULA, the friction stir welding techniques that would be needed are mostly developed and qualified already. And the engine that would be used, the RL-10 is an existing one with excellent heritage, and very benign operating characteristics.
Anyway, I just thought this analysis was kind of interesting. Such a system would still be able to loft a standard Orion, it wouldn’t require a ton of new development work (a little stage work, but a close derivative of an existing stage done by a team that has a proven track record from doing several such stages within the past decade or two), and since it would be a single-stick system, it would likely compare very favorably (LOM/LOC-wise) with the Ares-I.
Just a random thought.