A possibly irrelevant nostalgic detour, but I'm reminded of a thought experiment I contributed to once [1].. Called Mars 9 Tons at a Time. Basically use Zubri

A possibly irrelevant nostalgic detour, but I'm reminded of a thought experiment I contributed to once [1].. Called Mars 9 Tons at a Time. Basically use Zubrin's Mars Direct as the baseline plan, but do it without the Ares 5 he described or another HLV. Which has the effect that we would launch everything direct to Mars, no rendezvous or docking prior to the Martian surface. Primarily on the hypothesis that docking the parts together would add the cost of an assembly facility, and complicating the components enough to be dock-able together or heavily suboptimal to support simple assembly. But to avoid the costs of the Ares V, we wanted to do the entire thing using nothing but existing off-the-shelf available LVs to do the entire thing. If we'd gone the route of even a small custom LV, we would have the constant overhead of maintaining multiple launch pads that sat idle for 2 years at a time, just to launch something each launch window to Mars. Instead, our plan began with the Delta IV-H lobbing payloads trans-Mars. It has a payload of just over ~9T doing that. It could also fly three or four payloads per launch window, Vandenburg and the Cape, both at the start and the end of the window for a two week turnaround on each launch pad. Of course, suboptimal timing and suboptimal inclinations will cut into that 9T slightly. Further, we could make use of things like the Mitsubishi H-IIA [2], Atlas V 551 or Heavy (would need modest new development, or the 551 which is a bit weaker), Ariane V and maybe Proton (tho Proton is quite weak on high energy flights). Falcon 9 Heavy could eventually be added to that list (I think we were already talking about F9H in 2007...). That gets you maybe 6-9 payloads per launch window to Mars, of 6-9 tons depending on the launcher. That was our working budget for each payload. We managed to account for just about everything to do with the surface base using nothing but those small payloads... Planned out a standard design for a small descent vehicle in I think 1500kg to 2500kg which could land most of the rest of the mass as payload. We could do Earthworks (or Marsworks... dirtmoving / digging) using using some small remote-controlled construction equipment of 2-3T each. [3] Food, water and other supplies could pretty easily fit in smallish pieces like that. Used solar panels in place of Mars Direct's nuke for political feasibility / scalability to small payload deliveries. Also talked a bit about assembling a shelter using either the landed capsules for rooms, or assembling rooms from parts (I don't remember which). A part I was a bit uneasy about that some others thought was trivial enough, was building a new launch vehicle from parts on the surface... We'd be using small modular segments to build stages, scaffolding to assemble everything together and connect the stages, and all of this would be done outdoors. I was kinda worried about getting sand between the staging components or in the engines. The part we didn't manage to fit into the initial constraints of 9 ton payloads, all direct-to-Mars, was the actual crew transport there (we were fine-ish for the return). We stretched kinda far, pulled compromises nobody would really take, and got a 1-crew capsule to weigh in at 11T-12T. And that's with basically no redundancy! The entire thing would be a small can with a solar wing for power, life support for 6 months, and no way for the pilot to survive a launch failure from Earth. The capsule would fly straight to Mars, pilot would don a suit, depressurize the capsule and get in a MOOSE-alike [4] module and travel to the Martian surface in that. They would also be dependent on the entry trajectory for their life, as they'd have to walk to base on landing. From there our plans started to look a bit more "traditional" with either an HLV snuck in there just for the crew (and not the base) or orbital assembly of an orbital ship. Say, something with plenty of life support of its own, just does Earth-Mars-Earth and gets the crew to and from the base. The ship itself would be a rather substantial project to build, but at least it could ignore landing much payload. And the ascent vehicle from Mars would only need to worry about orbit, not escape velocity. [1]https://web.archive.org/web/20070810053610/http://uplink.space.com/postlist.php?Cat=&Board=businesstech&page=0&view=collapsed&sb=5&o=0 https://web.archive.org/web/20070810053610/http://uplink.spa... Ok, so the thread Mars 9 Tons at a Time wasn't archived at all. But there it was, August 2007. [2]https://en.wikipedia.org/wiki/H-IIA https://en.wikipedia.org/wiki/H-IIA [3] http://www.cat.com/en_US/products/new/equipment/excavators/mini-excavators/18172626.html http://www.cat.com/en_US/products/new/equipment/excavators/m... For example, is a smallish piece of equipment which can take many different parts on the end of the arm, work very slowly as a bulldozer, and fit inside our payload budget. We'd need something mechanically different, but probably that size / style of machine, to work remotely in very low temperatures. Shortly after our that thread, one of the construction equipment companies produced concept art describing something like that on The Moon with actuators (I think) in place of the hydraulics. Some folks complained that they were just advertising their stuff on The Moon to look cool. [4] http://www.astronautix.com/craft/moose.htm http://www.astronautix.com/craft/moose.htm Not terribly confidence-inspiring.