Orion On HLS -- An Odd Couple Minus The Third Wheel

SLS is meant for Orion and Orion for SLS...

...or so the conventional wisdom goes, but Orion was actually designed for a different mission architecture that included the Altair lander and the greater launch capabilities of the Ares V rocket. In that architecture, Altair would have handled the insertion into lunar orbit, and Orion would have only been responsible for the return.

I've written about this before in my post on single launch Apollo Redux architectures that could have been. When Altair and Ares V were cancelled with Constellation, Orion was left without a way to do a roundtrip to low lunar orbit, so for Artemis, NASA ended up settling for NRHO, which is closer to Earth in delta v terms.

This enables Orion to get in and out of lunar orbit on its own, but makes it substantially harder to actually land on the Moon, increasing the total lander delta v requirement from around 4km/s to around 5.5km/s, as well as dramatically reducing the frequency of rendezvous opportunities and typically increasing minimum mission duration.

NASA PR has tried to spin NRHO as a good staging point mainly because of its unbroken line of sight with Earth and nearly unbroken line of sight with lunar landing sites, but the communication problem is easily and better solved with a set of small satellites around the Moon. There has also been some reference to the advantages derived from the relative stability of the orbit, but there are highly stable "frozen" low lunar orbits, one of which is quite well suited for the polar regions Artemis is targeting.

If your goal is to land on the Moon, given a clean sheet design, LLO is much better than NRHO as a staging point for a lander. The choice of NRHO for Artemis is really just the product of a hardware limitation.

But what if we didn't have the hardware limitation?

Ironically, with the HLS variant of Starship as the lander, there is a very easy way around the hardware limitation that also comes with a massive side benefit.

I described it briefly in a twitter thread months ago and have shared more details about it in other conversations, but a blog post with all the relevant details in one place was long overdue.

What it comes down to is that HLS Starship is an overpowered enough lander to do both the job of Altair and the job of EDS (Earth Departure Stage) from the original Constellation program without breaking a sweat. Not only that, but it has also got plenty of room to launch Orion on its way up to LEO.

HLS is designed to launch as the second stage of a super heavy launch vehicle, refill in Earth orbit, fly out to lunar orbit, rendezvous with Orion, go down to the lunar surface and come back up to lunar orbit to rendezvous with Orion again.

A shockingly simple way to improve on this scheme is to have Orion ride HLS all the way to lunar orbit. With a somewhat fuller refilling, HLS could carry Orion to NRHO, but even better is a mission profile that goes all the way to LLO, which would actually reduce overall propellant requirements relative to the Artemis baseline. In other words, HLS could launch Orion without any extra launches and gain lunar cargo capacity or potentially need one fewer tanker flight in the process.

What would this require?

Basically, little more than a payload adapter. Orion would sit on top of HLS, tail to nose, pointing in the right direction for launch and with full abort capability as it has when riding on SLS.

Structurally what's needed is primarily a load-bearing barrel section with approximately the diameter of Orion's ESM. This would connect the two spacecraft while providing vertical space for Orion's main engine and the tip of HLS's nose. This barrel section could be slightly conical, tapering either down or up -- tapering down from Orion to HLS to minimize length or up from HLS to Orion to transfer force more directly and provide better aerodynamics, running tangent to HLS's skin. Or it could just be cylindrical for simplicity -- a lot of flexibility here.

For the tapered down and cylindrical options, a structural ring, welded either to HLS's nosecone (either on the inside or the outside) or onto the payload adaptor, would serve to prevent buckling. If welded to the outside of HLS, it could also double as the mating hardpoint. Since HLS Starship won't have TPS tiles or fins, it will be a lot easier to mount hardware to its nose.

Once in orbit, HLS will need to refill. This is a somewhat risky operation, so it is best to separate Orion from HLS, keeping it at a safe distance during refilling, and then bringing it back and docking it for the TLI burn. To retain abort capability during engine burns, we'll want the crew inside Orion and Orion docked tail to nose. Structurally, this is also the orientation Orion is better suited for.

This means retaining the payload adapter and having reconnect capability there -- essentially an unpressurized docking mechanism, either where the payload adapter joins ESM or where it joins HLS. My preference is for it being between the payload adapter and HLS, with the payload adapter being permanently fixed to the ESM. This reduces mass taken to the lunar surface and back, while keeping this as a 2-piece system (with the adapter fixed on one side), avoiding the complexity of a jettison in lunar orbit. It also offers better visibility when Orion and HLS are docked nose to nose, as well as lower risk of a damaging collision during docking and undocking. This adds a bit of extra mass to Orion (in the neighborhood of half a ton if built out of aluminum), but Orion has plenty of delta v margin and can easily do the TEI burn with it.

I've been asked if Starship's nose is strong enough to carry Orion through launch, and the answer is yes, easily, given the structural ring I mentioned earlier to prevent buckling (a few vertical stiffeners might also be warranted). A 4m diameter cylindrical section with 3mm thickness bearing 35t (including the LAS) at a conservatively high 5g (could easily keep it to 3g), would have compressive stress of around 45MPa. You'd have a bit more than that when factoring in cosine losses from the taper of the nose, and locally a bit more from vibrations and pitch/yaw maneuvers. If I recall correctly, Starship's structural steel is most closely related to 304L steel, which has a yield strength over 200MPa and UTS over 500MPa at room temperature. So, roughly an order of magnitude safety factor before the material gives out.

Some have expressed concern about how long astronauts might have to wait in orbit while HLS is refilled, but there's no need to worry there, since SpaceX is already planning to use a depot system for Artemis. For this mission architecture, all the propellant would be ready and waiting in LEO before HLS and Orion are launched. Propellant would then be transferred from the depot to HLS in one shot -- in likely on the order of an hour. Including rendezvouses, check outs, and waiting for alignment for the TLI burn, Orion would probably spend several hours in LEO.

Once the TLI burn is complete, Orion can undock, turn around, and dock nose to nose so the astronauts have all of the space and equipment in HLS available to them for most of the outbound transit. For the LOI (lunar orbit insertion) burn, the astronauts can go back into Orion, undock, turn around and dock tail to nose. Then flip again and transfer crew into HLS for descent (Orion would be left in LLO). After lunar stay and ascent, HLS would rendezvous again with Orion, and crew would return to Earth in Orion, leaving HLS in lunar orbit.

This mission profile allows for similar abort modes to what would be available with SLS -- using the LAS until it is jettisoned, and Orion's main engine afterward. Abort with Orion main engine during launch, TLI, and LOI would require a shutdown of HLS's engines, but similarly with SLS it requires a shutdown or at least throttle down of core or upper stage engines.

Another common concern is about the structural loads on Orion. Orion deploys its solar panels in LEO and because of that has much lower g limits for the TLI burn and beyond. Based on an ESA document on Artemis I, it seems Orion is rated for 1g with solar panels deployed ("Electrical Propulsion Subsystem" section). That section is a bit confusing, though, since it seems to have two conflicting claims for the g's that are reached when propelled by ICPS, neither of which would be correct. One way to reconcile this is to see the 0.5g figure as an approximation (rounded up), and the 1g figure as a design limit and/or derived from g's sustained when propelled by EUS on Block 1B (for Artemis III and on). 1g is ease to stay under with this HLS+Orion mission profile, and even if the limit is actually 0.5g or lower, it is manageable.

The key is the propellant HLS has to carry for lunar descent and ascent after the final burn with Orion (LOI). Assuming 2km/s for both ascent and descent, 100t ship mass at final engine cutoff (dry + cargo + residuals), and 375s isp, we're looking at nearly 300t after Orion separation and a minimum mass of over 320t with Orion. If the Raptors can throttle down to ~100tf (roughly 40% throttle), we can stay under 1g with 3RVacs. With 1 RVac and 1 SL engine, it would be possible to keep the g's even lower (or stay under 1g with less throttling), and with a single engine (either a gimbaling center SL in the current 6-engine configuration or a gimbaling center RVac in a new configuration), it would be possible to keep it well under 0.5g.

The mass, isp, delta v, and thrust figures above are approximate (especially the terminal mass figure), but if the stack ends up being substantially lighter at separation than I've suggested, it would always be possible to increase the mass by carrying extra propellant and/or cargo.

Another question is the fate of HLS. One of the goals of Artemis is to develop reusable lander technology, and Starship is designed to be reusable. Artemis I will leave HLS behind in NRHO, but it seems undesirable to commit to using it as an expendable vehicle in future missions. There are a few different reuse options that are possible (shoutout to @SpaceHaxx for bringing this up):

The lowest effort option is to leave it in lunar orbit to serve as a space station. Going a little further, with a second refilling in lunar orbit, it would be possible to bring it back to Earth orbit or to land it on the Moon to serve as a hab. This is also possible without sending tankers to LLO with a permanent depot ship in LLO. If a terminal mass (inert + payload upmass) of around 100t or lower is achieved, an HLS that is fully filled in VLEO (with 1200t propellant capacity) can deposit roughly 100t of excess propellant in the LLO depot before heading down to the surface. Picking it back up after ascent and crew transfer would give it more than enough delta v to either make a one-way trip to the surface or move to an elliptical Earth orbit.

If it was considered important to reuse HLS in the same capacity, we could go further still. Once in elliptical Earth orbit, a bit more refueling could enable it to come down to LEO, and a fill up in LEO could technically allow a fully powered descent to Earth without necessitating a thermal protection system capable of handling reentry heating. With over >9km/s of delta v budget, an unladen and fully fueled HLS could comfortably perform a powered descent as if it were doing an SSTO launch in reverse.

If it were up to me, I would probably opt for leaving the first 1-3 HLS Starships in LLO and taking down the following several to the lunar surface. Eventually, a standard Crew Starship with Earth atmospheric reentry capability would take over at least the Earth-lunar orbit leg.

Conclusion

Launching Orion on HLS Starship could be done with two simple pieces of new hardware: a ~4m diameter steel docking collar (doubling as a structural ring) welded around the nose of Starship, and a payload adapter (aluminum cylinder with stiffeners -- orthogrid is a good option) welded or bolted onto Orion's ESM.

In addition, some aerodynamic studies would need to be performed, SuperHeavy would need to be certified for crew launch, and an Orion-suitable crew access arm added to the Starship launch tower at KSC.

For this relatively small amount of work, NRHO could be bypassed (though Gateway could still be accessed), a higher lunar mission cadence (not limited by SLS) could be sustained, and the $2.5B/yr spent on SLS could be redirected to other worthwhile endeavors, like perhaps building a base on the Moon or working towards crewed Mars missions.