Rotation rates for providing pseudogravity decrease from a maximum 3 rpm for a small carefully selected crew to a rotation rate <1 rpm for a more general population. Similarly, radiation worker dosages of 5 rem/yr are constraints for the 102 and 104 habitats, while general population dosage limits of 0.5 rem/yr are used for the largest habitat.
A range of three atmospheric pressures has been selected for comparative purposes. Full and approximately half sea-level pressures are the maximum and intermediate selections for all three populations. The minimum pressure is a 100-percent oxygen physiological equivalent to the partial pressure of oxygen at 2400 in (8000 ft), 1200 in (4000 ft), and sea level on Earth for populations of 102, 104, and 106, respectively.
All habitats are limited to a maximum g level of 1.0; minimum g levels ranging from 0 to 0.7 are used as design constraints . These minimum g levels, combined with volumetric requirements, serve to define the habitat geometry for each of the four configurations investigated. The appearance of these volumetrically equivalent geometries varies as a function of how close the spherical configuration approaches the gmax limit at the maximum allowable rotation rate, as shown in figure 6. The example on the left depicts a sphere where the radius required to obtain the specified habitable volume is not close to the gmax limit. For this case an equivalent cylinder does not exist, the equivalent torus has a circular cross section, and the crystal palace has multiple layers. Habitable volume for the torus and crystal palace configuration is equal to their respective total volumes.
The example on the right in figure 6 shows a sphere for which the radius required to meet volumetric requirements would result in g levels exceeding gmax if the sphere were to rotate at the maximum allowable rate. To satisfy all design constraints, the radius is increased and the rotation rate decreased until the habitable volume just fits within the gmax/gmin limits. For this case, an equivalent cylinder rotating at the maximum rate does exist, as well as toroidal design with cylindrical walls. The equivalent crystal palace has fewer layers (only one layer is depicted) to remain within the g limits.
The appearance of an "equivalent" torus with cylindrical inner and outer walls imposes one additional design constraint. To prevent the inside cylindrical wall from buckling because of atmospheric-pressure-imposed compressive stresses, the "gravity load" on this surface can be made equal to or greater than the pressure load. This can be accomplished by suspending the entire internal mass from the interior wall and/or adding attached radiation shielding. An alternative approach provides compressive load-carrying capability by structurally stiffening the interior wall.
Radiation shielding is required on all habitats to protect the population from cosmic rays and solar flares. A choice exists, however, as to how this shielding is integrated with the structural shell. Direct attachment is undesirable because centrifugal loading added by the rotating shielding has a major influence on the structural shell thickness (and mass) requirements. The alternative is to mount the shielding outside the structural shell so that it either remains stationary, or counter rotates at a reduced angular velocity to obtain a net zero angular momentum for the entire habitat. This requires that alignment/positioning devices (i.e., mechanical bearings) be used to maintain a positive separation distance between the shell and shielding. To permit structural comparison of these two alternatives, both attached and detached shielding options were investigated.
SENSITIVITY ANALYSES
Both mass and cost sensitivity analyses were conducted for selected habitat elements. The structural mass sensitivities are presented first, followed by cost sensitivities for structural configurations (pressure shell and radiation shielding), atmospheric composition, and illumination implementation. All cost estimates were based on common ground rules which assumed use of lunar materials, unless the required elements were unavailable in lunar feedstock, in which case these materials were obtained from Earth. Asteroidal resources were not considered. Mass estimates were obtained by performing structural analyses of habitat pressure shell and radiation shielding configurations as a function of population, atmospheric pressure, minimum g level, and shielding integration techniques.
Structural Analyses
As the initial step of the habitat structural design trade, geometries were determined for the sphere, cylinder with hemispherical ends (when applicable), torus, and crystal palace on an equivalent habitable volume basis. The geometric relationships used are summarized in figure 7. Dimensions and important geometric parameters for the 65 shapes included in this analysis are listed in table.4. These characteristics were obtained by applying the design constraints identified in table 3.
Each of these habitat shapes was investigated for structural effects caused by the full range of established design
constraints. Membrane stress-analysis techniques were used for the spherical, cylindrical, and toroidal monolithic
geometries. For the modular crystal palace, element techniques based on O'Neill's work were used (ref. 20). The
membrane analysis was accomplished by resolving all shell loads (atmospheric pressure, skin inertial, internal
furnishings inertial, and attached shielding inertial) into combined normal and tangential distributed loads. This
enabled the required structural material thickness to be determined by the relationships:
where
Derivation of these equations is illustrated by the example for the spherical shell skin inertial loading due to centrifugal force depicted in figure 8.
The inertial force of the skin is
Therefore,
Components for internal furnishings inertia and attached radiation shielding inertia are similarly derived. These components are grouped and added together, substituted into the appropriate formula, and the resulting expression extensively manipulated to obtain the final equations listed in figure 9.
The crystal palace equations, also shown in figure 9, were obtained from O'Neill's work (ref. 20) by setting h = r and
incorporating the effects of skin inertial loading. Cable mass was calculated by neglecting the additional cable cross
section needed to support its own inertia (this assumption results in underestimated cable mass for larger habitats).
Attached radiation shielding was not considered for the crystal palace configuration. Employment of these crystal
palace formulas presupposes that the minor radius r is much smaller than the major radius R. To remain within this
constraint, a maximum minor radius of 10 m was selected. The crystal palace "r" radii shown in table 4 are obtained
by selecting suitable integer values for the number of modules
and solving for the required habitable volume.
Results of these structural shell analyses are presented in figures 10 through 17. Figures 10 through 13 show habitat structural shell mass as a function of minimum g level for fixed populations and atmospheric pressure.
The intermediate 51.7 kPa (half Earth normal) pressure was selected since representative structural sensitivity to the full pressure range is contained in figures 14 through 17.
The following general trends were obtained from data generated:
- At lower populations, the structural mass of spherical habitats is more sensitive to gmin than are other configurations (fig. 10).Attached shielding results in a substantial structural shell weight penalty.
- The structural penalty per person for attached shielding increases with population (habitat size).
- For larger habitats (figs. 12 and 13), the curves for toroids with attached and unattached radiation shielding bracket the "spherical habitat with unattached shielding" curve. This is important since the "torus with separate shielding" structural mass has been derived by neglecting the compressive loading condition of the inner cylindrical wall. Correction of this condition will shift this curve (increase the mass) closer to the spherical habitat curve.
Referring now to figures 14, 15, 16 and 17 :
- Habitat structural mass increases linearly in proportion to internal pressures.
- With a broad habitable g range (0.25 to 1.0), habitat pressure sensitivity is independent of configuration for larger
populations.
- For larger habitats with a constrained g range (0.70 to 1.0), pressure sensitivity is configuration-related, with the crystal palace showing the least sensitivity due to "fixed" cable mass.
Shielding masses for each of the habitats were obtained via the equations in figure 18 for the following alternatives:
- Sphere or cylinder - Shielding can either be attached or detached for all cases, and surrounds the structural shell.
- Torus - Circular cross-section toroids can have either attached or detached shielding. Torii with cylindrical walls also have an attached/detached option, but must either have sufficient shielding on the inner wall to equalize the internal pressure, or additional shell structure must be incorporated to sustain these compressive loads. All shielding configurations investigated were toroidal and surrounded the structural shell.
- Crystal palace - Only detached shielding was evaluated for the crystal palace. Since the crystal palace employs cables to react centrifugal loads, cylindrical rather than toroidal shielding was used.
All attached shielding was assumed to be immediately adjacent to the structural shell. Detached shielding was separated by 1.5 m. Owing to the increased area of separated shielding, its mass ranges from about 0.2 to 4 percent more than that for attached shielding. The following general shielding mass trends were obtained from separated shielding data shown in figures 19, 20, 21 and 22.
- Shielding mass sensitivity increases as the habitable g range is constrained (gmin =>1)
- Spheres exhibit greater sensitivity to shielding mass at lower populations than do other configurations.
- The shielding mass is two to three orders of magnitude greater than habitat structural mass for small habitats (102 population), but is only one order of magnitude greater for large habitats (106 population).
Cost Comparison Ground Rules and Assumptions
The assessment of alternative habitat design concepts can best be accomplished by cost comparison. Our analysis shows considerable cost differences owing to overall habitat material requirements, and between terrestrial supplies and the various products manufactured in space from lunar material. Habitat cost comparison has been conducted by assuming that all habitat atmospheric oxygen, pressure shell structure, and radiation shielding are derived from lunar resources. It was also assumed that other items such as lighting fixtures are required in sufficient quantity so that they can be economically manufactured from lunar materials. The only materials or products imported from Earth are those which are either unavailable in lunar resources, or which because of complicated manufacturing operations requiring expensive facilities coupled with relatively small quantity requirements can be more economically obtained from Earth. Since the sensitivity studies in this analysis were conducted at a major subsystem rather than a component level, no complicated Earth products of this sort were included. The special ground rules and assumptions used for costing of lunar and Earth habitat construction materials are as follows:
- Space processing and fabrication costs are obtained from data generated for the 11-year space manufacturing buildup period for the construction of solar power satellites from lunar resources (ref. 5). This information was used in lieu of equilibrium data because it was obtained by more rigorous methods and is better documented. This approach is conservative for comparison with Earth imports because space manufacturing equilibrium costs should be somewhat lower.
- Space processing and fabrication costs are based on the conversion of 1737 ktons of lunar material into useful products. A total of 5745 man-years and $6.45 billion are expended to perform this conversion, that is, personnel cost (including transportation and all support), equals $1.123 million/man-year.
- Material processing costs and fabrication costs constitute 25 and 75 percent of this total expenditure, respectively (R.Williams, personal communication, 1977).
- Costs for each major space-manufactured product obtained from lunar material are shown in table 5. The constituents of lunar material shown reflect some lunar beneficiation prior to delivery to the space manufacturing facility (SMF). The processing and fabrication factors reflect the relative difficulty of manufacturing that particular product (R. Williams, personal communication, 1977).
- Transportation costs from Earth to the SMF were derived for equilibrium conditions using third-generation launch vehicles. This approach is purposefully optimistic so that cost comparisons with space-manufactured products will be more conservative.
- A single-stage to orbit vehicle delivers payload in low Earth orbit (LEO) at $47.06/kg, and a heavy mass-driver reaction engine provides LEO to SMF transfer at $10.16/kg, for a total transportation cost of $57.22/kg Q. Engel, Dept. of Systems Engineering, Univ. of Illinois at Chicago, personal communication). For most items, this transportation cost will be an order of magnitude above the Earth purchase price, which has therefore been neglected.
| Product | Glass | Silicon | Oxygen | Shielding | Titanium | Iron | Aluminum |
|---|---|---|---|---|---|---|---|
| Percent of lunar material |
2.25 | 12.67 | 19.80 | 46.63 | 4.43 | 5.12 | 9.10 |
| Processing factor | 2 | 2 | 0.5 | 0.1 | 1 | 1 | 1 |
| Processing cost, $/kg |
2.95 | 2.95 | 0.74 | 0.15 | 1.47 | 1.47 | 1.47 |
| Fabrication factor | 2 | 2 | 0.05 | 0.01 | 1 | 1 | 1 |
| Fabrication cost, $/kg |
11.16 | 11.16 | 0.28 | 0.06 | 5.58 | 5.58 | 5.58 |
| Total mfg.cost, $/kg |
14.11 | 14.11 | 1.02 | 0.21 | 7.05 | 7.05 | 7.05 |
Habitat structural elements included in this comparison are the pressure shell and radiation shielding. Windows for
illumination are evaluated later. in this paper. Internal furnishings were not included because their allocated mass per
person was assumed to be equal for all populations within each of the two habitat classes. The smaller habitat class,
10
To provide better cost sensitivities for the larger-class habitats, the radiation shielding thickness for the 104 population configurations with agriculture was increased from 1.08 to 2.04 m. In addition, spherical habitats with 105 and 107 populations and shielding 2.04 m thick were also included in an attempt to establish the habitat size for minimum cost per person. A geometric definition of these added habitats is included in table 4.
The cost comparison performed assumes that all materials needed for the structural shell and radiation shield are obtained from processed lunar soil. The shell is assumed to be aluminum with a working stress of 218,000 kPa at a cost of $7.05/kg, and the shielding, molded industrial slag at $0.21/kg from table 5. If a significant percentage of shell alloying elements must be imported from Earth to meet the assumed working stress, then the shell costs must be increased accordingly.
All cost comparisons shown in figures 23, 24, 25, 26, 27 and 28, are for habitats with unattached shielding and an intermediate atmospheric pressure of 51.7 kPa. Unattached shielding and intermediate atmospheric pressure were both previously shown to result in reduced structural shell material requirements. Figure 23 shows the cost per person of the shell and shielding structural components for spherical habitats as a function of habitat size (population) and minimum g level. The shaded bands show the cost range from gmin = 0 to gmin= 0.7. It is interesting to note that the shielding cost exceeds shell cost up to a population of about 105 for larger populations the reverse is true. Similar results were obtained for toroidal and crystal palace habitats.
Figures 24, 25, 26 and 27, show cost as a function of habitat size (population) for various habitat configurations at fixed minimum g values. The smaller-class habitats (radiation shielding thickness = 1.08 m) are characterized by a very wide configuration cost spread at 102 population, narrowing to $20,000/person difference at 104 population for all values Of gmin. For the larger-class habitats (radiation shielding thickness = 2.04 m), the minimum cost per person always occurs between populations of 105 and 106 except for the special case of zero-g spheres, where the cost per person continues to decrease as population increases. The data for zero-g spherical habitats can be obtained by summing the left hand curves of Figure 23, . For most values of minimum g, toroidal habitats show a significant cost advantage ($10,000 to $20,000/person) over spherical and crystal palace habitats. It must be noted, however, that both toroidal and crystal palace habitat configurations require additional hub structures, which have not been included in this analysis. These hub features include cargo vehicle docking provisions, personnel access structure, and shielding alignment support structure, all of which are inherently provided by a spherical geometry. Also, as noted previously, the geometric constraints employed for this comparison result in toroids with cylindrical sections. With unattached shielding, these inner toroidal habitat cylindrical walls are compressively loaded. The additional structure required to prevent buckling was not included in the "detached shielding" structural weights used.
Large class habitat costs as a function of minimum g level are shown in figure 28. All geometries and populations exhibit lower cost per person for reduced minimum g.
Habitat Atmosphere Cost Comparison
The combination of habitat configuration, atmospheric pressure, and atmospheric composition have a significant effect on the total habitat atmospheric mass and its acquisition cost per person. Toroidal and crystal palace habitats are efficient from an atmospheric volume standpoint, and spherical and cylindrical habitats are inefficient since their total pressurized volumes exceed the habitable volumetric requirements of their populations.
Possible habitat atmospheric compositions of oxygen and diluent were previously established as a function of total pressure and population (see figs. 4 and 5 and table 3). The oxygen for the habitat atmosphere will be recovered during lunar material processing at the space manufacturing facility. Since it is produced in large quantities at the habitat construction site, the cost of oxygen is a reasonable $1 .02/kg
Possible inert atmospheric constituents include nitrogen or helium. During early space-settlement efforts, these diluents will probably be imported from Earth, with a delivery cost of $57.22/kg. Since the cost of atmospheric inerts is 56 times that of an equal quantity of oxygen, their use for large habitats will either be limited or a nonterrestrial source will be developed. Some scientists have suggested that volatiles may exist in frozen form at the lunar poles, but a more probable source is asteroidal material. It has been estimated that up to 0.3 percent of the mass of carbonaceous chondrite asteroids could be nitrogen (ref. 21). If nitrogen can be recovered from a convenient extraterrestrial source, its cost should be approximately equal to that of oxygen.
Figure 29 shows atmospheric cost sensitivity assuming that either nitrogen N2 or helium (He) diluents are imported from Earth. The cost of an O2/He atmosphere is significantly lower than an equivalent O2/N2 atmosphere since, at equal pressure, nitrogen is 7 times more dense, and is therefore equivalently more expensive to transport. All the curves in figure 29 are for habitats with total volume equal to habitable volume. Spheres and cylinders usually have total volumes greater than the habitable volume as shown in the right-hand column of table 4. For example, a large spherical habitat (106 population), with a 51.7-kPa atmosphere of O2N2 and minimum pseudogravity one-half Earth-normal, will have a volume 1.5 times greater than shown, or a cost revision from $36,600 (shown in fig. 29 ) to $56,300/person.
The primary conclusion reached from this cost comparison is that if a significant percentage (>20 percent) of a settlement atmosphere is imported from Earth, then the cost of these diluents may be a major contributor to total habitat cost.
Habitat Illumination Cost Comparison
The interior of a space habitat can be illuminated by either artificial fights powered by photovoltaic cells, or by natural sunlight admitted through windows. This study evaluated the costs of these alternative lighting methods for one representative habitat configuration. Since the natural illumination choice is probably reasonable only for larger habitats, the 10,000-person spherical habitat was selected for this evaluation. This sensitivity analysis was performed only, for photovoltaic versus natural illumination independent of habitat configuration and population variations. Further work should consider possible configuration and population effects.
Artificial lighting- Recent solar photovoltaic array estimates use a specific power generation factor of 138 W/kg (ref. 22). To account for power transmission, conditioning, and control equipment, an efficiency of 90 percent was assumed for an effective array output of 124 W/kg. Photovoltaic array cost was based on space manufactured components of 30 percent silicon and glass and 70 percent aluminum and other metals. The cost for light bulbs and fixtures was estimated using a specific weight of 0.03 kg/W of rated fixture output. Space-manufactured components were selected over Earth imports owing to the estimated acquisition cost ratio of 1/6. Table 6 displays the weights and costs for the various illumination system components. These calculations were based on the power requirements obtained from table 2. Weighted values were used for array sizing and maximum values for distribution fixtures. The use of maximum power requirements for fixtures reflects the difficulty in moving them as lighting needs vary. As indicated, the cost of illuminating the agricultural area is an order of magnitude higher than that for the living area. Of particular interest is the overwhelming influence of distribution costs, which account for 88 percent of the total artificial lighting cost. If Earth-imported fixtures had been assumed, the distribution cost would have exceeded $200 M, instead of the $33.73 M used.
Natural illumination - The corresponding analysis for determining weight and cost of illuminating with natural sunlight
is more difficult because of the many implementation options available. To reduce weight and the possibility of
meteorite damage, it is desirable to minimize the window area. This can be done by concentrating the sunlight before
it enters the habitat, but the concentration is limited by the 800-K softening temperature of window glass. It was
assumed that the maximum window temperature should be less than 600 K (337o C). To maintain this temperature,
the absorptivity of the glass must be balanced by its conductivity. The energy admitted to the habitat is also a
function of window absorptivity; specifically
(refs. 23, 24):
Only sunlight in the visible waveband is useful for habitat illumination. This constitutes 38 percent of the total solar
radiation (fig. 30). Admitting energy outside this band into the habitat results in many penalties, including the
requirement for larger windows and increased internal thermal control capacity. It is therefore desirable to use
concentration mirrors that selectively reflect sunlight in the visible spectrum into the habitat for illumination. The
degree of allowable concentration is influenced by glass thermal limits and absorptivity, which are functions of glass
thickness. The thickness of the window is determined by structural requirements. Based on terrestrial manufacturing
experience, the practical window weight limit is about 1 ton. A weight of 0.5 ton per window was selected or volume = Ad = 0.2 m3, where the glass density is 2450 gm3 . This equation is solved simultaneously with the structural equation for a pressure-loaded square plate with simply supported edges:
The solution resulted in a window 5.65 cm thick with an area of 3.5 m2.The transmissibility of 5.65-cm-thick glass is 0.84 in the visible wavelengths (ref. 25). Since the sum of reflectivity, absorptivity, and transmissibility must equal 1,
an absorptivity of 0.16 may be conservatively assumed.
The total energy transmitted through the windows by selective mirrors is therefore 36.5 kW/m2. This is equivalent to a solar visible energy concentration factor of 70. Window areas, weights, and costs are delineated in table 6.
| System and component data | Area illuminated | ||
|---|---|---|---|
| Living | Agricultural | Total | |
| Artificial light | |||
| Requirement, kW/person | 0.56/0.93 | 5.69/11.21 | 6.25/12.14 |
| Photo-voltaic Array Area m2 |
27,800 | 284,000 | 312,400 |
| Photo-voltaic Array Mass,kg |
44,900 | 459,000 | 503,900 |
| Photo-voltaic Array Cost,$M |
0.41 | 4.21 | 4.62 |
| Distribution Mass,kg |
281,700 | 3,397,000 | 3,678,700 |
| Distribution Cost,$M |
2.58 | 31.15 | 33.73 |
| Total Cost, $M | 3.0 | 35.4 | 38.4 |
| Natural sunlight | |||
| Requirement, kW/person | 1.86 | 5.69 | 7.55 |
| Glass window Area m2 |
610 | 1860 | 2470 |
| Glass window Mass, kg |
84,000 | 257,000 | 341,000 |
| Glass window Cost, $M |
1.19 | 3.63 | 4.82 |
| Mirrors/other equipment Area m2 |
54,200 | 165,300 | 219,500 |
| Mirrors/other equipment Mass, kg |
10,800 | 33,100 | 43,900 |
| Mirrors/other equipment Cost, $M |
0.08 | 0.23 | 0.31 |
| Additional Shield Area m2 |
610 | 1860 | 2470 |
| Additional Shield Mass, kg |
3,220,000 | 9,850,000 | 13,070,000 |
| Additional Shield Cost,$M |
0.68 | 2.07 | 2.75 |
| Total Cost, $M | 1.95 | 5.93 | 7.88 |
Illumination requirements were again obtained
from table 2 with the following assumptions. Since distribution in living areas may be cumbersome, twice the
maximum energy/person allocation was used. Agricultural areas are relatively open and a simple swiveling mirror can
control light, so the weighted power requirement was used.
The mirror estimates shown in table 6 were derived assuming 90 percent efficiency, a total concentration plus distribution factor of 80, and aluminum mirrors and support structure weighing 0.20 kg/m2.
There are several ways to provide radiation protection for habitat windows: one alternative employs very thick glass windows that supply all the shielding 'necessary; this is undesirable, however, since the thickness must be 1.02 or 1.92 in to meet the 5.0 or 0.5 rem/yr dose rate requirements, respectively. This increases the total absorptivity, and these windows weigh and cost more than 20 times those that are only 5.65 cm thick. Another approach is to use shielding constructed from lunar slag in a chevron configuration to protect thin windows. Comparable thicknesses are still needed but material costs are 67 times lower and manufacturing requirements are greatly simplified. Chevron construction requires about twice the amount of normal shielding, but as shown in table 6, the cost remains below that for thin-glass windows.
The results of the preceding discussion and a comparison of total costs in table 6 lead to several conclusions:
- If natural illumination is used, it should be concentrated with selective mirrors before it is transmitted into the habitat.
- Thick glass windows that supply radiation protection are expensive and impractical, table 6. Natural illumination should be used for agricultural areas since its cost is 1/6 that of artificial lighting.
- Natural illumination should also be considered for living/working areas unless distribution becomes awkward or expensive.
CONCLUSIONS
The introduction to this paper posed three questions to be addressed and, it is hoped, answered by these habitat structural/cost sensitivity analyses. Sufficient habitat options have been evaluated to obtain a quantitative answer to the initial question: How much do "Earth-normal" physiological conditions cost when applied to a space habitat structural design? Obviously, many "less than Earth-normal" design options are possible, but a representative example should be illustrative.
| Environmental Parameter | Earth normal conservative designs | Representative reduced cost habitat designs | ||||
|---|---|---|---|---|---|---|
| Sphere | Torus | Crystal palace |
Sphere | Torus | Crystal palace | |
| Atmospheric pressure,kPa | - | 101.3 | - | - | 51.7 | - |
| Atmospheric composition | - | O2/N2 | - | - | O2/He | - |
| Minimum g level | - | 0.7 | - | - | 0.35 | - |
| Habitat structure Mass, kT |
15,300 | 9,300 | 9,200 | 5,500 | 3,200 | 5,200 |
| Habitat structure Cost, $B |
107.6 | 65.3 | 65.1 | 38.6 | 22.7 | 36.3 |
| Separate radiation shielding Mass, kT |
80,200 | 100,700 | 81,600 | 45,500 | 55,200 | 58,700 |
| Separate radiation shielding Cost,$B |
16.8 | 21.1 | 17.1 | 10.0 | 11.6 | 12.3 |
| Atmosphere volume ratio | 2.78 | 1.00 | 1.00 | 1.25 | 1.00 | 1.00 |
| Atmosphere volume ratio Cost, $B |
293.0 | 105.5 | 105.5 | 7.2 | 5.8 | 5.8 |
| Total Cost,$B | 417.4 | 191.9 | 187.7 | 55.8 | 40.1 | 54.4 |
Notes: 106 population;V H =2,000m3/person: IM=53,000 kg/person; trad = 2.04m (dosage limit of 0.5 rem/yr); atmosphere volume ratio = total volume/habitable volume.
The data in table 7 compare 106 population habitats having identical habitable volumes (2000 m3/person), internal furnishings (53,000 kg/person), and radiation protection (0.5 rem/yr). The variables noted in the table are limited to habitat configuration, atmospheric pressure, atmospheric composition, and the range of pseudogravity in the habitable volume.
The total costs obtained offer some interesting comparisons between conservative and reduced cost designs, as well as for alternative configurations. As shown, when only habitat structure, shielding, and atmosphere are compared, the conservative to reduced cost design ratio is 7.5, 4.8, and 3.5 for spherical, toroidal, and crystal palace habitats, respectively. The answer to the first question is that design costs associated with Earth-normal conservatism are at least significant and may, in fact, be excessive. Also, table 7 cost data shows that the cost dispersion for conservative designs varies by more than a factor of 2.2 due to configuration differences, but the dispersion for reduced cost designs varies by only 1.4. This reduction in configuration sensitivity demonstrated by the lower cost habitat designs is due primarily to the reduced effect of their atmospheric costs.
The second question asks which "less than Earth-normal" physiological conditions offer the greatest habitat cost savings. Minimum g-level constraints and atmospheric pressure have strongly synergistic effects on structural shell design requirements and costs. The selected habitable g range significantly affects the overall structural shell geometry as demonstrated in table 4, while pressure affects the structural shell thickness. When evaluated independently, these influences are of the same magnitude. The structural mass difference for a 106 population spherical habitat is 4.7X106 tons for a gmin range from 0.7 to 0.25. This difference is due to geometry effects; it is not caused by a change in rotation rate, which varies only between 0.9 and 1.0 rpm. Similarly, the mass difference over the 76 kPa atmospheric pressure range is 8.IX106 tons and 3.7XI06 tons for gmin values of 0.7 and 0.25, respectively. Similar results are obtained for toroidal and crystal palace habitat geometries. Shielding mass and its resultant cost at a fixed population and radiation requirement are influenced solely by habitat geometry, which is a function of minimum g-level. The relative importance of structural shell and radiation shielding costs is also a function of habitat population. Below 105 inhabitants, shielding costs are higher, but above 106 inhabitants, structural shell are dominant.
The quantity of atmosphere diluent imported from Earth has the largest potential influence on habitat cost. The mass which must be imported depends on excess habitat volume, total atmospheric pressure, and the inert gas selected. These factors combined produce a cost factor of 40 for the spherical habitat example given in table 7.
Illumination alternatives also exhibit substantial habitat cost effects. The use of unselected/unconcentrated natural solar lighting results in window plus extra shielding costs similar in magnitude to those for the basic habitat structure. The costs for this technique are two orders of magnitude higher than for a selected/concentrated natural illumination system.
The third question, regarding identification of a recommended habitat configuration, has not been fully resolved by the limited analyses conducted during this investigation. Each habitat shape offers benefits for certain applications, and configuration selection may well depend on design considerations outside the scope of this study. These considerations include implementation of life-support functions, power supply, waste heat dissipation, shielding alignment, docking functions, and illumination for each of the habitat configurations.
Although an overall best habitat configuration has not been recommended, some interesting configuration
comparisons can be made:
- Torus - Offers the lowest habitat basic structural cost (pressure shell and detached radiation shielding) for almost
all combinations of population and gmin. It is also volumetrically efficient and contains only as much atmosphere as required. Unfortunately, several parts of the torus structure were not included in this analysis: spokes, hub, shielding alignment, and additional stiffening needed in compressively loaded cylindrical sections. Also the distribution of
illumination in a torus appears cumbersome. These factors will combine to increase the basic torus structural cost so
that its cost per person approaches that for spherical habitats.
- Crystal Palace - Has attributes and disadvantages similar to the torus. Its basic structural cost always exceeds
that for the torus and is less than that for spheres when gmin exceeds 0.5. It is atmospherically efficient but has
illumination distribution difficulties and requires unestimated spoke and hub structures. The crystal palace cylindrical
radiation shielding, which contributes to its high basic structural cost, also offers several potential advantages. Its flat
ends provide structure for spin axis alignment, and also radiation protection for the unpressurized volume within the
crystal palace's inner structural wall. If additional facilities for space manufacturing were placed within this volume, they would be provided free radiation protection.
- Sphere - Has a basic structural cost which exceeds that of the torus and crystal palace over most of the population and gmin ranges. It also exhibits an inefficient volume ratio, which results in high atmosphere costs. Its advantages include straightforward illumination distribution and no need for separate hub or shielding alignment structure. The sphere's excess pressurized and shielded volume could become advantageous by locating space manufacturing facilities in this region.
One important discovery about habitat configurations was made during this work: based on habitat basic structural cost, the cost per person has a minimum for habitats in the 105 to 106 population range. Even more interesting is the fact that this minimum is independent of habitat configuration; spheres, toroids, and crystal palaces all had minimum costs in this population range. This minimum is obviously dependent on the ground rules and scope of this investigation, but an equivalent minimum should also exist for revised ground rules and for more extensive habitat subsystem considerations.
RECOMMENDATIONS
Physiological research should be concentrated in those areas where less than Earth-normal conditions offer the greatest savings in habitat cost. Based on the work conducted during these sensitivity analyses, recommended research falls into two categories: (1) that necessary to validate our assumptions and (2) that highlighted by our results. Our assumptions included allowable habitat rotation rates for pseudogravity, and the use of a broad range of habitable g levels. Since higher angular velocities and broad habitable g ranges result in smaller, less massive, and lower cost habitats, it is very important that we fully understand the long-term physiological effects of these conditions. Specifically, the following research is needed:
- Rotation rates: The angular velocities used for this study were based on work performed in a 1g (Earth surface) environment. Initial human adaptation, long-duration habitability effects, and transition effects between rotating and nonrotating enclosures must be obtained in a reduced g environment.
- g-level variations: As inhabitants move radially within a settlement with a broad habitable-g band (0.25 -> 1 g), they will be subjected to continuous or stepped g-level variations. This effect, with special emphasis on human adaptation and transition, must be investigated.
- Low-g tolerance: More knowledge is required on long-duration physiological effects of low- and zero-g habitation. Experimental data are limited to those obtained during the 84-day Skylab mission.
The results of our sensitivity analyses indicate that the combined influence of atmospheric pressure and composition
have potentially the most significant effect on habitat cost. Therefore the following research is recommended:
- Low atmospheric pressure: Perform extended duration and adaptability testing for O2/N2 mixtures at less than sea-level pressure. Evaluate speech transmission effects in addition to respiratory effects.
- Low-density diluent: Conduct tests similar to those above for O2/He atmospheres at less than sea-level pressure.
As repeatedly pointed out in the preceding text, this habitat sensitivity investigation was limited to the basic habitat
structure (pressure shell and radiation shielding) plus some considerations of atmospheric composition and habitat
illumination. It is recommended that additional habitat sensitivity studies be performed to expand this work. The
following sensitivity investigations should be accomplished to pin a better understanding of certain habitat design
conditions and subsystem options on overall habitat mass and cost.
- Influence of secondary structures, such as spokes, hub, shielding alignment, and additional stiffening in compressively loaded sections, on overall habitat structural mass and cost comparisons.
- A detailed look at habitat illumination sensitivity for various configurations. Illumination implementation appears to be fairly important and should be treated in much greater depth than was possible in this paper.
- The effect of fixed personnel area requirements on habit structural weight and cost. (The work described in this paper assumed a fixed volumetric requirement.)
- Habitat heat rejection sensitivity as a function of habitat shape and radiation shielding design.
- Evaluation of life support function integration in various habitat configurations. These functions include air circulation and conditioning, power, water and sewer services.