Lasting

Introduction

How can you make things so that they still work after a long time? More specifically, unless we find another Earth, survival off this planet will involve artificial environments in artificial structures. If people are going to live off Earth, those structures are going to have to be dependable indefinitely. How is that to be achieved?

The two main approaches seem to be:

a.- Passive - make the structure intrinsically stable and robust.

b.- Active - make the structure in such a way that it can be continually repaired or renewed, and have systems in place to carry out such activities.

An example of a. would be the great pyramids in Egypt. They've lost their surface facing and been robbed of their contents, but they've maintained their overall shape and internal spaces. Explanations for their longevity include the use of a building material, stone, that is stable in that environment, the environment itself with its low rainfall and lack of frosts and the use of an intrinsically stable geometry for their gravitational environment, vis, the top is much smaller than the bottom, there are no overhangs and the slope is not very steep.

A possible example of b. might be a bristlecone pine tree. Some of these may be of an age similar to that of the great pyramids. While the core of the trunk is no longer living, it is constantly added to by the essence of the tree, the outer, living, layers of cells that are constantly maintaining, renewing and replacing themselves.

OK, so what does this say about making structures off Earth that last? What approach should be taken to the latter and what aspects of design are likely to help, with either approach?

First, a look at perhaps the ultimate passive approach, living on other heavenly bodies.

Planets and moons

Living on a planetary surface (or that of a large moon) would have a variety of advantages for robustness. Gravity would be provided in the 'natural' way by mass, rather than requiring spin to provide pseudogravity. This immediately reduces complexity. The gravity would not, of course, be variable to suit the circumstances, as would be possible with spun habitats, and would be unlikely to match Earth gravity exactly. We don't yet know what the effect of prolonged existence in, for example, Mars gravity (~1/3 of Earth's) would be. However, it is likely to be less deleterious than no gravity at all and compared with spun pseudogravity, there would not be the problems of rotation induced vestibular disturbances with associated nausea.

A planetary surface would provide cosmic ray shielding over at least one hemisphere (under foot) and if there were atmosphere or magnetic fields it could contribute towards shielding over the other. Even in the absence of the latter, the ready availability of material for shielding, either by burrowing or piling on, would make shielding a very straightforward matter.

In addition to passive gravity and passive shielding, a planetary or lunar surface would provide a source of raw materials readily to hand, without the effort of lifting them out of a gravity well. This could be either volatiles (moons of outer planets) or metals (Earth's moon) or both (Mars).

Aspects of the problems of volatiles which are essential to life, when you are not actually living on a source of them as above, are discussed next.

Loss of volatiles

In a space based structure (and some planetary or lunar ones) it would be essential to preserve the available volatiles (e.g. water, atmospheric gases, components of life cycles) which, by definition, would be prone to leakage away to the vacuum of space. An 'active' approach to coping with such leakage would be to seek out fresh sources of volatiles (comets, outer moons) and use these to replace those lost. Passively, having an excess of such materials over that needed for current life processes would be a likely strategy. The other passive approach would be to design a structure to limit losses of volatile materials.

Design features to limit loss are fairly obvious:

a. Have thick walls.

b. Don't have volatiles near an outer surface that may be damaged (e.g. punctured by micrometeorites).

c. Keep surface area (through which leaks may happen) small in relation to volume.

d. Don't have many openings to the outside that may be prone to leakage around seals (e.g. airlocks).

e. Don't have structures that require rotating seals.

f. Keep internal pressure as low as is practical.

g. Keep internal temperature as low as is practical.

In a little more detail:

a. and b. are different aspects of the same point. A thin membrane will allow gases to leak across it. A thin metal wall, as on a radiator, could be easily punctured. Both suggest that to limit loss of volatiles one should have thick solid walls, relying perhaps on passive conduction to take heat to an outer surface for radiative dispersal. Clearly this might limit internal activity compared with a more active method a heat dumping.

c. relates to a. also. Losses would be less with less wall area as well as with greater wall thickness. Reducing surface area for a given habitable surface would count against, for example, extended toroidal colonies compared with simple cylinders (see Geometries), although this would depend on the relative area of end caps and thus on the length to diameter ratio and the end cap geometry. For example, a simple cylinder with flat ends would have a surface area of 2prh+2pr². For an extended torus, if one simplifies by assuming that the end areas are negligible and that the inner surface is negligibly different from the outer surface, the surface area would be described by 4prh. Clearly the 'break even' point for this would be when 4prh = 2prh+2pr². Simplifying, this becomes 2prh = 2pr² , prh = pr², rh = r² and h = r. Thus, when h (the colony height or length) equals the colony radius (i.e. 1/2 the colony diameter) the two approaches have the same surface to habitable surface ratio. Any colony with a greater length to diameter ratio than 0.5 (typically this can be as high a 5.0) would have less surface to habitable surface ratio as a simple cylinder than it would as an extended torus. Obviously with hemispherical end caps the break even point would be at a higher ratio (1.0), but the general point still holds for most colony designs. Any colony design that involved more than one pressure vessel would also have the effect of increasing the relative surface area for leakage, although from another point of view it might be argued that it would prevent loss of all volatiles in one catastrophic event (as has been pointed out for Radiators).

d. is obvious, but in this case a design for passive endurance may have the opposite requirement from one for active endurance (and a lot of to-ing and fro-ing, e.g. see What rings are for).

e. would tend to rule out designs with non-rotating radiators (see Radiators ).

f. there may be a little leeway here, since life can exist at lower than one atmosphere pressure (e.g. fairly readily up to 10,000 feet altitude on Earth).

g. the main application of this may be to adopt a stop-go policy with periods of relative hibernation for the whole colony. Replication of some Earth like seasons would be one approach, or it could be more drastic/longer. Obviously this is less of a passive design aspect and more of an active adoption of a strategy of periodic passivity (if you see what I mean).

Next, a look at simplicity and complexity.

Other aspects of designing to last

In designing to last, the need is to consider space habitats as more architecture than aerospace. Extending the analogy, it would probably be better if it were akin to old architecture that has stood the test of time, rather than to the modern light and airy stuff designed to minimum cost and to the limits of the materials involved. Designs probably need to be simple rather than complex and clever.

Walls have been suggested as having a non-rotating outer shell for shielding and a rotating inner shell for atmosphere retention and artificial gravity. Such a design would require bearings and careful control of clearance between the two parts. It may be that a simple single shell outer wall would be less efficient in some respects, but it would probably be more likely to still be 'working' after a thousand years than would a complex double shell of differing rotational characteristics.

Considering radiation shielding again, simple passive shielding by mass would clearly be more reliable and robust than active shielding by electrical and or magnetic fields, although it may be in other ways less efficient (if artificial active fields can actually be shown to work). For example, if you want the habitat being shielded also to be mobile and, therefore, of as low a mass as possible, field shielding may be a preferred approach (and could possibly also be applied for braking or manoeuvring of starship habitats). Field shielding can, of course, also be passive, as discussed above for planets.

Thick metal is often considered as a wall material, but there have also been suggestions for using pre-stressed reinforced concrete (Sheppard, D. J. 'Concrete Space Colonies'. Spaceflight. 21: pp. 3-8.) for space habitats, an argument being that the material is, in some ways, more resistant to damage and corrosion. However, this approach still needs use of metal, in the form of reinforcing wires, which tends to undermine some of the arguments in its favour (it has been suggested that the wires may be checkable and replaceable, which is moving away from passive robustness towards the active approach). Concrete may also have other disadvantages such as its capacity for taking up gases in contact with it. This has already been implicated in causing problems in Biosphere 2, for example. G. K. O'Neill, in his famous book 'The High Frontier', pointed out that windows would be the most vulnerable part of a space colony, given the relative fragility of the transparent materials available. This may be an argument for avoiding windows altogether and lighting by artificial means (see below), thus allowing a simple, uninterrupted, strong wall.

Clearly there would be some cross benefits in designing space habitats to last. Thick, single shell, (probably) metal walls would be simple rather than complex for greater reliability, but would also reduce loss of volatiles, provide intrinsic radiation shielding and would be of considerable benefit in resistance to (small) impacting objects.

The lack of overhangs was mentioned in the introduction as a favourable design aspect. Obviously a rotating space habitat has, at the very least, the outer surface as a continuous 'overhang' of sorts. However, while this overhang is unavoidable, there are other possible ones that could be avoided. Going back to the comparison of cylindrical and, more complex, extended toroidal colonies mentioned above, the latter have a ceiling as well as a floor for their habitations and this has the potential for shedding parts of itself onto the inhabitants. The same would apply to any rain pipes or lights incorporated into such a ceiling (or near the centre of a cylindrical design). This problem could be got around by having 'rain' sprayed from floor level (or 'natural' rain cloud formation) and lights mounted on the floor to illuminate the floor on the opposite side of the cylinder.

Other active approaches

Some active approaches to endurance of space habitats have been touched on above, including intermittent hibernation behaviour to reduce loss of volatiles, seeking out replacements for lost volatiles, shielding by artificially generated fields and designing composite walls that would be built to be accessible for maintenance.

There are other examples of active approaches to endurance. In the Daedalus probe design, the BIS included 'wardens', automated repair and maintenance robots (see Artists' Impressions). This active approach was for a mission that was meant to last decades. On a mission lasting hundreds of years it probably would have been necessary for the wardens to repair themselves. No such provision was made, or needed, for the study in question.

O'Neill, in 'The High Frontier' discussed maintenance teams replacing damaged window panels on large colonies. There would be an obvious need to do this quickly to limit loss of volatiles, which would require constant vigilance.

In 'What rings are for' the possibilities of using external or internal tori for active maintenance of habitats is discussed. The need for access points with this method, with consequent increased leakage risk, has already been mentioned above.

Conclusions

Aside from living on other planets or moons, the best approach to designing a robust habitat that would endure without too much intervention would seem, from the arguments above, to end up with a (relatively) simple windowless cylinder with thick metal walls, minimal access points, artificial lighting in the floor and nothing up off the floor to fall on peoples heads. If this sounds like Arthur Clark's 'Rama' (see Artists' Impressions), that is probably because I'm a fan. However, having said all of that, space habitats aren't planets and any design will inevitably require some active maintenance. As an example, if you rule out windows, you'll have to maintain the artificial lighting and its power supply (which some people may consider to represent more of a risk than windows). Clark's design was for an interstellar craft where artificial lighting would be necessary anyway.

The level of active maintenance considered acceptable if large space habitats are ever constructed will probably vary with the anticipated use and with the wishes and psychology of the population. As a lazy pessimist, I'd tend to favour a passive robust approach. However, it may be that a colony type with a more active bias would also be somewhat lighter (for travel), more efficient and able to reproduce itself more rapidly, and could thus end up being the longer term winner.

Return to front page - Starships and Space Colonies