Though the excavated structures like the atrium habitat and the found structures of lava tunnels would be the easiest and simplest ways to exploit the indigenous materials of lunar and planetary bodies and thus represent the most likely approaches to initial habitat development, they have the critical limitation of not being entirely controlled in location. They depend on the providence of pre-existing geography/geology and would very often not present ideal locations for some facilities. Thus, long-term, means of producing built-up structures would be necessary to allow for building habitats exactly where they are needed or wanted. Of necessity, these would need to similarly make the most from the simplest materials the natural environment has to offer, would need to integrate with the excavated structures as part of their expansion, and would employ the same structural components for their outfitting. Thus they would be based on largely similar architecture.
The most likely approach to this would employ the use of regolith as the basis of a concrete material –regolete– with which to build structures that, from the inside, are essentially identical to the vaulted spaces employed in the excavated habitats and feature the very same system of socket grids for outfitting. The first of such structures are likely to be simple pre-cast reinforced regolete vault modules producing structures similar to the grid vault structures of the telerobotic excavated settlement. These would be produced as unit boxes about 12m square that link and stack together and are sealed with a regolete mortar. Their thick walls –a composite of solid and foamed regolete in varying densities– would allow them to be used like simple surface bunkers but their more common use would be in trenches where they would be assembled in series, buried, and used to create long pressurized tunnels under loose granular surface strata that will not support conventional tunnel boring. Additional precast bunker forms such as domes would be developed to combine with these basic modules, allowing for more elaborate complexes. Again, these might be placed right on surface or within excavations. Areas undergoing strip mining of surface regolith would be likely locations, complexes of these modules installed within them before the tailings from mining activity are used to bury them and arrays of photovoltaics, radiators, and heliostats installed on top of them.
Of course, one can only employ precast structures of this sort at relatively small sizes and within some proximity of their production facilities because their high mass would make them difficult to transport long distances. A more flexible approach suited to much larger scale structures would be the simple ‘mound formed’ bunkers first employed in WWII by German engineers for the construction of large bomb-resistant structures. This construction method would employ simple carefully sculpted mounds of loose regolith as forms for the construction of domes by the simple in-situ pouring of regolete over the mounds and a reinforcement frame (integrated to the socket grid used for interior retrofit). Once the regolete is cured, the mound filling the now hardened vault shell is removed by excavation and piled up on the exterior for additional radiation shielding and impact protection and an optional slab foundation is poured.
This approach could also be employed in pre-existing surface craters, their rims excavated and terraced in the manner of the atrium habitats when the dome is finished. Being surface structures, they could employ the same light transmitting approach as the EvoHab hull, the domes made ‘virtually transparent’ by the use of image-corrected heliostat arrays and internal light emitters linked by thin light pipes that make the dome seem transparent despite being many meters thick. Thus Marshal Savage’s proposed transparent crater domes can be realized with much tougher and simpler materials.
One limitation of the mound forming technique is that very large amounts of regolith material would need to be moved around to make these form mounds. As experience with regolete improves automated extrusion techniques akin to the currently explored Contour Crafting and large scale architectural ‘fabbers’ would likely be employed as a more efficient method of automated surface construction. Eventually the same composite hull system as used with EvoHab orbital habitats would come into local mass production and use, allowing for progressively larger spans and easier adaptation.
A key characteristic of these surface dome habitats would be their ability to conjoin or merge domes into clusters to cover progressively larger surface areas once the maximum practical spans are achieved. The result would be a kind of ‘hyperbolic pendentive’ grid of vast area producing a sort of flowing or undulating ‘sky’ supported at the ‘groins’ of merged domes by flowing towers used to host terraced arcology-like residential structures. Such complexes could come to cover vast areas like a second skin under which the natural surface of the lunar or planetary body could be terraformed whole. In this way large areas of the Moon or Mars could be terraformed long before any more sophisticated wholesale methods of geoengineering are developed.