2. External Design

DaedalusaL4 is both a residential and commercial settlement. For it to survive and profit, it must attract residents to live there. Therefore, the aim of this project is to design a living environment in space, which is as comfortable and as desirable to the residents as any Earth environment. It must also be cost effective if it is to be seen as a real alternative to living on Earth, not just a short-lived gimmick, a playground for the rich and famous or an ‘isolation resort’ for the planet’s elite.

 There are many parts to realising this objective. First, it must provide the basic functions of life support - food, water, oxygen etc. It must preserve the health and well being of the residents, posing no greater risks than are experienced here on Earth, or are balanced by the nature of the activity involved.  (Certain jobs may be hazardous.)  In short, it must seek not only to preserve life, but quality of life also. This covers a wide area, ranging from radiation shielding, to fluoridation of the residents’ water. As the list is long, they are only mentioned specifically when appropriate, but have been taken into consideration at all points in the design process.

Ideally, the settlement will receive finance from those organisations who will make profit from what its situation has to offer (see 20.1 Financing the Settlement) - those involved in the production of SPSs, the mining and refining of asteroids and the production of goods that can only be produced in a low gravity environment (high quality plastics, synthesised proteins etc.). Other industries that can take advantage of the vacuum and the low temperature ‘beyond the station wall’ may also decide to join in, as costs decrease.  It is also an ideal location for many other activities, e.g. research and space observation.  Further industry will be required to support the residents on board and make DL4 as self-sufficient and financially realistic as possible.  Some suggested industrial activities are covered in the sections 11, Industry, and 12, Research.

These industries will require one or more of the following features:

·         low gravity

·         low temperature

·         low pressure/vacuum

·         access to non-shielded areas (radiation research)

·         cheap power

·         heat

·         cheaply transported and refined materials

·         extra-station construction/assembly

·         docking facilities.

On the basis of these requirements we set about choosing acceptable limits for gravity, radiation levels, and area required.  We then chose a suitable shape, shielding, lighting, power source and method of creating spin, taking these limits into consideration.

2.1        Requirements


The idea of living in a weightless environment may seem appealing at first thought, however, when practicalities are considered, there is little argument that a residential space station requires gravity. The decalcification that residents would experience would prevent them from returning to live normally on Earth, and that is just the start. Muscles lose proteins, size and strength, and undergo changes in metabolism, neuromuscular control and individual muscle fibre characteristics. Prolonged hypercalciuria (excess calcium in the bloodstream due to decreasing bone mass) may lead to the formation of renal stones (ref. 82).

As hard as we looked, neither could we find any practical way of avoiding these problems, or simulating gravity aside from the traditional centrifugal method. [Inserting titanium rods in residents’ bones seemed a little impractical, not to mention painful to say the least (ref. 81).] We can therefore assume that the station must spin to produce its ‘gravity’.

The speed of this rotation affects the residents by disrupting the fluid balance within the ear. Most people can adjust to 3 revolutions per minute (rpm), but do suffer a type of motion sickness when adapting. This adaptation would however be severely hindered by the changes in rotation rate caused by residents constantly moving between the effectively non-rotating zero-g area and the rotating 1g area.  Therefore the lower the rotation rate the more desirable it would be. A rotation rate of approximately one rpm was considered best from a medical standpoint. This should serve to limit the effect the Coriolis force has on the inhabitants.  This force can render simple movements complex and cause the eyes to play tricks while turning the head can make stationary objects appear to gyrate and continue to move once the head has stopped moving.  All people (bar the extreme few) can adapt to 1 rpm (ref. 18, 26). A low rotation rate makes things like sport a lot easier - the ball will land closer to where you expect it to at 1 rpm than 3 rpm. For the benefit of all concerned, we set 1 rpm or lower as our target figure.

The gravity that you experience also depends on your distance from the axis about which you are spinning (i.e. - your radius). As radius increases, so does the gravity according to the formula , where simulated gravity, angular velocity and R = radius of settlement (see Appendix A).  The problem of balancing radius with rotation rate is considered further on.

As well as areas at ‘normal’ gravity levels, the station must also have areas with different g-levels (zero-g, micro-g).  After all, DL4 won’t be built to be exactly the same as Earth but also to exploit the attractions of space.


The environmental radiation that DaedalusaL4’s residents will have to face consists of a Galactic and Cosmic component, from distant stars and even more distant galaxies, and a Solar component, from the Sun (ref. 8, 18).  Both these must be shielded against for the lifetime of the settlement as even minor radiation exposure (0.5 - 1 Sv, Sv = Sievert = 1 Joule/kg of exposed material), depending on radiation type, can lead to radiation sickness.  Luckily trapped particles that contribute to radiation in orbits between 1 and 2, and 4 and 6 Earth radii (the Van Allen belts) can be ignored due to DaedalusaL4’s location.  However, the settlement also loses through this, as being at L4 removes it from the protective enclosure of the Earth’s magnetic field.  Thus high energy particles have to be stopped along with wave form radiation.

Galactic Cosmic Radiation (GCR) consists mainly of high energy (> 0.1 GeV) protons (85%), a-particles (14%), and heavy nuclei (1%).  The latter heavy nuclei tend to come from all atoms with atomic numbers greater than 2 and less than 29.  GCR, on a short term basis is not life-threatening instead being career limiting, however, in the long run it poses the greatest threat to the residents due to its deeply penetrating capabilities.

More life threatening in the short term, due to its variable nature, is solar radiation.  Once or twice every 11 years, during periods of solar maximum in the solar cycle, the sun produces protons, electrons, and heavy nuclei accelerated to energies between 107 and 109 eV.  These dangerous releases can result in a thousand-fold increase in the normal proton and electron density of about 5/cm3.  The normal plasma wind, being burnt off the Sun, poses little threat as the flow is released mainly from longitudinal positions that are unfavourable for the direct transfer of particles to L4 along the interplanetary magnetic field lines.  Any radiation that does reach DL4 from these ‘background’ releases is of a reasonably low energy and can thus be blocked quite easily.

All space flights outside the Van Allen belts have, so far, been lucky enough to avoid solar flares. A long term space settlement would not be in a position to avoid these. The inhabitants of the station would also be exposed to the radiation for a considerably longer length of time than any space flight so far.

Therefore an acceptable level of radiation exposure must be decided on. The United States Federal Government has set a standard of less than 0.5 rem/yr (» 0.005Sv) for the general population. While levels of up to 50 rem/yr are not proven to cause any damage but the consequences of long term exposure at these levels are not fully known or understood. As such the more conservative 0.5 rem/yr seems the better option.


One important point to remember is that the station must be relatively spacious. On Earth, the size of your house/apartment is often taken as a sign of your wealth/social status. Forcing people to live in the bare minimum of space gives rise to psychological problems, which can easily be avoided. The comfort and well being of the residents is one of the most important aspects of design.

The well reasoned arguments for the amount of space required by each section have been laid out in the specific sections (see Sections 8 Residential, 9 Agriculture, 10 Recreation etc.). This resulted in the following list of areas required.


At 1g (m2)

Other g levels (m2)

Total (m2)

















Industry & Research








table 2.1a - Area required

Additionally, when choosing the dimensions of the settlement to fit the area required, the length of the lines of sight must be taken into account. Long lines of sight are desirable for several reasons. Not only will they offset any sense of overcrowding and claustrophobia but also inhibit the development of several potentially harmful psychological conditions e.g. Solipsism syndrome (see 18.3 Psychological Concerns). Long lines of sight would also help to alleviate the sense of boredom and monotony that could develop in an environment such as this.

Despite allowing plenty of room for these aspects of the settlement we must also ensure that materials and space are used efficiently.  This does not contradict the previous point, everything must have enough space, but nothing should be wasted.

When you start with nothing, as you do in space, everything comes at a price. The cost may be lower from the Moon, or from asteroids, than from the Earth, but it is a price none the less. It’s all very well saying that DL4 should be a nice place to live, but it has to be a place first. Costs must be kept down where possible to ensure:

a)      Not to bankrupt the world economy (!!!)

b)      Not to incur a negative public image as money that could be ‘well spent alleviating poverty/building schools/curing cancer etc.’ is thought to be ‘wasted’

c)      Not to make the costs prohibitive to potential investors

d)      That return to these investors is realised in as short a space of time as possible,
and most importantly,

e)      To encourage people to view the station, not as novelty destination for those with too much time or money on their hands, but as an accessible settlement, where normal people live everyday lives, with a different view out the windows.

These limits meant that while not compromising the health and well being of the residents (to a significant extent), we had to design an orbital settlement that was economical with space, materials and resources.  While allowing room for expansion to a certain degree, we sought to make sure that there is not a huge amount of space that will never/cannot be used (e.g. - large areas at unsuitable gravity levels).

2.2     Design

 (For 3D views of the chosen design, see our Image Gallery and Movies Page)


To start with, all shapes with a square edge around the axis of rotation were ruled out. Structurally, they would not be sound, and they are impractical, as to maintain a 1g gravity level a circular floor is required.

After a brainstorming session produced masses of very weird designs (potatoes or swiss-rolls anyone?) we came back to considering the now ‘standard’ manipulations of the curve. We had the following options: cylinder, sphere, dumbbell, torus and composite shapes (ref. 37). By calculating the radius, line of sight and spin rate, we started to examine the pros and cons of each design. Finally, a torus, with a central sphere and four smaller moveable spheres, was chosen. Other options were ruled out for the following reasons.

A cylinder was too large. In order for it to maintain structural integrity (to ensure that the square edges do not come under too much stress) the length must be many times (circa 10 times, ref. 37) greater than the radius. This means that a suitable shape, in terms of area required at 1g and length to radius ratio, would have a very short radius. This puts it beyond our target rotation rate.  Even when structural integrity was compromised to a point where radius was equal to or greater than length, there remained many levels which would not be used. A proposed solution was that these be removed to create a torus-type dwelling with square edges. Using the most widely available materials, this would not have been structurally sound when rotated and also posed problems for construction. A large settlement would allow for expansion, but the initial cost must also be taken into account. For reasons outlined in Appendix B – Population, we set out to design for a population of 10,000, so a cylinder was far too large.

There is also a similar argument against a sphere. This was ruled out, as its useable space to volume ratio was too low. The areas at uniform gravity levels are arranged in cylindrically around the axis of rotation. At one end of the scale, where the habitation area lies close to the axis of rotation, the small radius pushes rpm over our established limit. At the other, where the area of habitation is a small strip around the equator of the sphere (parallel to the axis of rotation), the volume of the sphere becomes huge, and the surface area unmanageable. In either case, there are huge areas within it that have gravity levels that are not suitable for ‘normal’ use, and so would not be required. The attractions of a sphere improve with increasing size of population; however, with 10,000 people it was not seen as feasible.

The dumbbell seemed promising, as it would be ideal for the production of solar power satellites because of its large radius. There were however several design problems that arose, as we looked closer. The two ‘ends’ seemed to be quite isolated. Also, to have a large enough area at normal g levels the spheres would need to be too large. The tube between the spheres also posed a problem: if it was large enough to contain ‘levels’ for research or a large lift for the transport of industrial goods and people, then it would become uneconomical at the lengths that make it ‘ideal’ for the construction of Solar Power Satellites (SPSs). The zero-g area that is available is also too small. The structural stresses of having a very large radius between the ends to support the construction of SPSs also narrowed the choice of material that we could have used

A torus has the advantage that you can make it as long or as short as you want, by altering the major radius, and then altering the minor radius, so you have the area that you require.  It is also relatively simple to divide, if this is required, e.g. - in case of emergency.  On its own, the primary disadvantage is that there are relatively few variations in gravity, and there is no zero-g or micro-g area.

Most composite shapes that have been designed were for larger populations than we are considering.  To modify them would mean making them smaller, thus increasing the rotation rate, and making them unsuitable when residents’ comfort is taken into account. The over-division of a beaded torus or multiple dumbbells seemed cumbersome, incorporating too many transport and communication problems. 

This left us to design our own ‘composite shape’ using the best features of the other shapes.

A torus is the ideal shape for the 1g area, because it provides a large amount of floor space at a uniform gravity level, and does not require large amounts of atmosphere for gravity levels that would not be used. All parts are easily accessible by residents, yet some of the habitat remains hidden from view at all times.

The central area must either be a sphere or a cylinder.  A cylinder was ruled out because its required structural length would not be conducive to simple construction.  Therefore a sphere was chosen, by a process of elimination (see 3 Construction).

The central sphere will be connected to the torus by means of four tubes. If these were big enough to accommodate research all the way ‘down’, they would be too large, and all the space would not be used. If permanent extensions to the tubes were built at one point, it would be quite limiting if the industry or research group that needed them changed their requirements, viz. gravity levels. For that reason, it was decided to include spheres that could be moved along the external length of the tubes as required.

Cross Section of DeadalusaL4

Torus - minor radius

To get the right amount of surface area at habitable gravity levels, we considered two primary designs of tori. The first was a short torus, with a large minor radius, and many different floors within acceptable gravity variations. The other had a single habitable floor, across the diameter of the torus and parallel to the axis of rotation. Both ideas had pros and cons attached.

The idea of many different floors means that the tubes that connect the torus to the central sphere are considerably shorter. However, from the resident’s point of view, we felt that having many floors could be confusing, and would not do anything to help them overcome the idea of living in a ‘tin can in space’ (see 18 Health Care). A torus with a larger minor radius would require a smaller major radius, otherwise there would be more space than is required. A smaller major radius meant that the rotation rate increased beyond levels we thought acceptable. Also, as you move closer to the centre, the floor becomes more noticeably curved, and the line of sight decreases. 

A torus with a single floor would be more comfortable for the residents. There would be a greater sense of openness to the station and it allows for a greater ceiling height. With a smaller, longer torus, people are able to take a relatively long walk around the torus (in ours 5.95125 km), before they get back to where they started from. The larger major radius means that the rotation rate can be slowed, so decreasing the Coriolis effect. On the other hand, the lifts that connect the torus and the central sphere have to be longer. As they are longer, the tubes require more sturdy materials for their construction. This increases the cost. It was decided that, on this point, comfort of the residents was of more concern than economies of construction and the latter option was chosen.

The maximum minor radius that could be used was determined using the figure for required surface area at normal gravity, and taking into account the major radius, and the rotation rate that would be required to create normal gravity. Following that various radii were compared, and the minor radius of DaedalusaL4 was fixed at 125m. This was deemed to strike a balance between space across the floor, ceiling height and the line of sight along the curve of the torus. At a minor radius of 100m, the increase in that line of sight was so small (» 6m) that it was not considered worth increasing the major radius by that much. It also seemed that it was slightly too small ‘across’ and ‘up’.  An increase in the minor radius to 150m would make the ceiling height above the residents too large, and would mean that there would be a lot of space that would not be used. It would also decrease the line of sight and make the major radius smaller, thus the rpm required to get normal gravity would increase (albeit only slightly).

A radius of 125m also allows for single buildings with more than one floor, so the allowance for floor-space can be maximised. (This applies particularly to agriculture, where the 434,600m2 of actual floor space will be double stacked). It also allows for sufficient storage of water and goods, as well as services, such as electricity and transport, to run underneath the floors. The habitat floors can also be thick enough to support the growth of tall trees.

Radius (minor)




Major radius required




rpm required to give 1g




Line of sight (given average height of 1.7m)




table 2.2a - Comparison of minor radii considered

Torus - major radius

The major radius was then adjusted so that the floor across the diameter of the torus at 90° to the major radius was at 1g, and fell within the guideline figure we set ourselves at the start of the design process.

Minor radius (r)


Floor area required at 1g


Radius to torus floor level


Major radius (R)


Using the formula , where acceleration due to gravity and angular velocity, we calculated the rpm required for normal gravity.  At 0.975 rpm, this is well within the 3 rpm that people can adapt to and the 1 rpm that we set as an upper limit (ref. 18, 37).

Torus – shielding (radiation)

Three basic types of radiation shielding were considered: passive bulk shielding, electromagnetic shielding and chemical radioprotection.  Electromagnetic shielding was ruled out, as lower exposures can be achieved with passive bulk shielding than can be acquired with the same mass needed for the electromagnetic shield configuration.  Also, electromagnetic generators are suspected to cause a health risk, which, although the effects are not yet fully understood, might be more harmful than those particles with an energy below the shield cut-off point.  Even if such a shield was efficient, those particles stopped by the shield could produce Bremsstrahlung radiation (electromagnetic radiation produced as high energy particles decelerate, ref. 21).  Chemical radioprotectors, such as APAETF (aminopropyl-aminoethyl thiophosphoric acid), are impractical on such a large-scale population.  Therefore, passive bulk shielding was selected by elimination.  The question then remaining was the configuration of this shielding.

Bulk shielding produces certain problems in that thin to moderate shields are effective at screening most low energy radiation, however as shield thickness increases, shield effectiveness decreases.  This is caused by the high-energy particles in GCR being decelerated by heavy elements in the shield and in turn producing large quantities of photons in the form of X-rays.  To overcome this problem on DaedalusaL4 a balance had to be found between the shield’s protective capabilities and its own addition to harmful radiation.

DL4’s radiation shielding will consist of a series of layers, starting with a crushed lunar regolith matrix.  In some areas (see fig. 2.2d) this matrix will lie outside the hull in a non-spinning ‘shell’.  (This is to avoid the need to increase the mass of the aluminium hull so that it has sufficient structural strength to support an internal shield at >1g levels of gravity.)  In others, initial layers of lunar soil will be applied to the inside of the outer hull using a polymer-based resin to seal in the next soil layer.  Such a resin, like polyamide binders, could be made from in-situ materials on the Moon, Asteroids or even Mars (ref. 9).  The next layer will consist of crushed regolith to a depth of about 3m.  This will be loosely packed and provide a matrix for water, with which the soil will be flooded.  This layer will be sealed on the inner side with another layer of polyamide bound regolith.

Lunar soil consists largely of oxidised silicon, aluminium, magnesium, calcium and iron.  These range in atomic mass from 24 (magnesium) to 56 (iron), however the principal components are silicon (28) and aluminium (27), which are considered to be heavy enough to shatter ionising particles but light enough not to cause indirect deceleration and therefore Bremsstrahlung radiation.  This soil alone will cause a dramatic reduction in radiation, however the water (solidified to ice), that the regolith provides a matrix for, will also act as an effective barrier, as it is composed of two parts hydrogen and will be able to shatter invading ions.

Using ice within the soil will also provide the additional advantage that incoming ions will simply diffuse rather than being amalgamated into the crystal structure of the next (metal) layer.  Such an amalgamation would disrupt the normal crystal structure and lead to a ‘hardening’ of the metal lowering its ability to stop radiation and increasing the risk of it shattering if it ever impacted with micro-meteorites.  This protection will lower the need to constantly regenerate the shield through heating and means that only the external meteorite shield will periodically (every decade) require replacing.

The next section of radiation shielding will consist of a layer of Zerodur or a similar lunar ceramic (see 4 Materials) if replacement layers are ever needed.  This, while not adding to the shielding properties of the hull, will provide a suitable base for the solid metal shield.  This shield will consist of a sandwich of copper-tungsten and aluminium, providing the final protection against the extremely high-energy particles of solar flares and GCR.  The copper tungsten portion will consist of an alloy composed of 90% weight density tungsten (W, Ar = 184) and 10% copper (Cu, Ar = 64).  The inclusion of this alloy in the traditional aluminium shield provides an approximate improvement in shield effectiveness of 2 : 3 (Al shield : Cu-W/Al shield) in the proposed L4 environment during solar maximum and an estimated 1 : 2 (Al shield : Cu-W/Al shield) during solar minimum (ref. 7).  This final ‘sandwich’ layer of shield would only require 0.5g/cm2 to provide final protection for DaedalusaL4 as well as providing a reflective surface for sunlight (on the ceiling – dyed blue).


Such a thin metal shield would be best applied during construction, while the interior of the settlement is still in a state of vacuum.  This would provide the required conditions for the application of the aluminium and copper-tungsten layers through a xenon ion electrolytic spray method.  This involves the use of aluminium or Cu-W (depending on the layer being applied) as the cathode of an electrolysis cell in which xenon acts as the electrolyte.  A charge is passed through the xenon, turning it into a high-energy plasma that travels at high velocity to the cathode.  Upon impact with the metal of the cathode, the metal is disintegrated, forming an even spray of aluminium or Cu-W over the Zerodur ceramic base.  This method distributes the impurities in the metal (as the Al will have been refined from lunar soil).  A similar distribution of irregularities is provided for in the lunar soil segment simply by crushing the regolith.

Within the settlement, electromagnetic interference (EMI), caused by the systems onboard or by external radiation disruptions, can be prevented by using EMI shielding.  However, the traditional aluminium ‘cage’ will prove inefficient for use onboard DL4 and as the ‘cages’ are required to constantly shield the equipment, even during transit, any items brought from Earth must bring its own shielding with it – making transit more expensive.  Thus, it was decided that graphite fibres, intercalculated (diffused) with bromine, atoms should be used.  These have mechanical and thermal properties identical to pristine graphite fibres but their resistivity is lower by a factor of 5, even lower than stainless steel.  They will also provide weight savings of up to 80% and even if only applied to the power systems onboard they would still save up to 15% in mass (ref. 29) compared with current EMI shield configurations.


Torus – shielding (meteorite)

For the purposes of meteorite shielding external to DL4, where the outer hull of the settlement is exposed to space (i.e. – where radiation shielding is not external to the settlement and where windows are not in place), the configuration of the Mars Trans Hab Module MOD Shield (see fig. 2.2b) was chosen (ref. 22).  The shield offers the advantage of initially being transported compressed to a thickness of about 5cm.  Once in place, outside the settlement, the foam inside the shield is inflated to its full thickness of 30cm.  This 30cm is composed of a front layer of (see fig. 2.2c) 2 ply 25.4 micron Mylar followed by 3 Nextel® AF-10 ceramic fabric layers, each separated by 10.2cm.  This gap is supported using Sonex Polyurethane low-weight open-cell foam.  The foam support is hollowed out with 7.6cm diameter cores spread out to reduce the shield mass by 45%.  The final wall consists of 5 layers of Kevlar® fabric (style 710) and a final 3 plies 25.4 microns Mylar.  This configuration, held together using RTV 3145 adhesive, is capable of resisting impacts of a velocity of 6.82kms-1 and incident objects about 6.35mm in diameter.  This ability could prove necessary in the region of L4 due to possible, although unlikely (ref. 3), accumulation of Kordylewski Clouds at the stable libration point.

Torus - lighting

One of the elements that design must provide for is adequate light throughout the station as well as day and night periods to allow human and plant circadian rhythms.  Much of the station could be lit by electrical light, as power will be in abundant supply (see 2.2 Power Generation Systems). The easier option would be to light the entire station in this manner, but this is not an option because of psychological considerations (ref. 37), such as Solipsism syndrome, and the aesthetic appeal of natural light.

For this reason, a design must be put in place so that a reasonable amount of natural light is available to the people and plants, within the residential area (located within the torus). There were various options that were considered for this purpose.

All of these ideas centred on the reflection and re-distribution of light using different arrangements of mirrors. The more complex required precision, which would be hard to obtain, and would have to be altered to account for the irregular orbit of the station about L4, leading to difficulties as moving parts are liable to break or suffer wear.

Ruling more complex ideas out, we eventually chose a mirror angled at 45º to the plane on which the settlement will lie (see fig. 2.2f).  This will rotate about its own axis, centred above the axis of rotation of the torus, once every 24 hours thus giving day/night cycles.  However as natural light need only be supplied to the torus and the central sphere, a full mirror, covering the entire settlement and measuring 3.014km in diameter, would be redundant. For this reason, there will be reflective areas, made from sodium foils, around the circumference, and at the centre of a structural shell. The first will extend 424m towards the centre from the rim reflecting light into the torus and then a second deformable ring extending from 257m to 156.97m from the centre of the mirror (see fig. 2.2e) will light the central sphere.  The rest of the mirror shell will be devoted to the production of energy, being covered with solar cells (see 2.2 Power Generation Systems).

The window section of the torus must allow light in, to illuminate the residential areas, diffuse this light to ensure an even distribution and also screen the residents from all harmful solar and galactic radiation.  While a thick layer of glass will shield from UV rays the light must also have other elements removed from it that the Earth’s atmosphere normally screens.  To achieve this, laminated layers of lead glass will be installed.  At a density of 3.27gcm-3 (± 0.07) this glass contains 30% lead (ref. 14).  Such a percentage of lead is not the highest possible (values reaching 70%), but instead a trade off between radiation protection and increasing brittleness as lead content is increased in the ‘anhydrous’ lunar glass (which itself has excellent mechanical properties).  In space, this brittleness must be prevented from causing any structural weaknesses within the settlement.  Hence the lead glass is laminated and will be covered with an outer layer of glass (in effect double-glazing).  The outer layer of glass, synthesised entirely from lunar soils, will be frosted white, using lunar ceramics, to diffuse light and prevent exposure of the residents’ eyes to directly reflected sunlight.  [Unfortunately this prevents external views throughout DL4 however an observation window has been provided (see 7.1 Internal Design).]  Such diffusion will develop the effect of a ‘sky’ within the torus, as the blue ‘roof’ will reflect the light softly to the residential areas.  The synthesis of the outer glass from lunar materials allows for replacement glass to be easily produced in the event of particle impact damage on the window.  In the event of solar flares/storms the residents will have the capability of closing Al/Cu-W ‘shutters’, mounted on the structural ribs of the windows, to increase their radiation protection.


Central Sphere - dimensions

The central sphere provides an ideal starting point for construction due to its self-contained nature.  During this period the sphere will be spun at higher rpm rates to allow better working conditions at 1g for the workers, therefore enabling them to tolerate longer stays.  However once complete the sphere will be used to house the zero-g recreation area, zero-g research area, docking facility, rectifying antenna and most of DL4’s construction industry.

Therefore, the central sphere must be large enough to accommodate the living area for the construction workers, docking facilities, and room for the production of the aluminium stressed skin, which will make up the torus (see 3 Construction).  We estimated that a radius of 300m is a reasonable extrapolation of current technology, and that this would suffice to meet the needs of the station (see 7.1 Internal Design, Appendix A).

Central Sphere – shielding

During construction the central sphere will be rigidised with an ice shell in its substructure.  While under normal usage this ice cavity will provide extensive radiation shielding and therefore further radiation shielding will only be required to a small depth inside the sphere.  This shielding will take the form previously outlined for the torus, with the exception that the lunar-soil based water/ice matrix will only measure 1.5m in depth.  All other elements will remain the same.

The sphere will additionally be externally coated with Trans-Hab Meteorite Shielding of the same configuration as outlined for the torus.



Central Sphere – lighting

Within the central sphere lighting will be as natural as possible, with light being reflected from the central ring of deformable sodium foils (see fig. 2.2e) on the mirror.  This light will enter into the sphere through a window encircling the sphere in a band starting 111.7m from the sphere’s centre and carrying on until the level of the thermal radiators and rectifying antenna (rectenna) hub (see fig. 2.2g).  Light will be able to travel through the transparent layers of the rectenna thus illuminating the sphere.  The window will be composed of sections of glass constructed in the same manner as the glass used for torus construction.

Moveable Spheres

The moveable spheres will house research labs, and will be moved so they lie at the required gravity level. The area needed at micro/varying-g was estimated at 9,000m2. When this is divided by four, and area is allowed for the passage of the lifts through the centre of the spheres, a diameter of 50m was chosen (see Appendix A).  The movement of these small spheres will be facilitated thanks to magnetic induction linear motors that will use the aluminium connecting tubes, running through 10.5m diameter sleeves in the centre of the spheres, as tracks on which to travel.  The nature of this motor allows frictionless motion and hence lowers the risk of damage to the station from external moving parts.  Radiation and meteorite shielding will be provided for the moveable spheres in the same way as the central sphere.

Connecting Tubes (Spokes)

The connecting tubes must house the lifts, and facilitate the supply of services to and from the moveable spheres and the central sphere.  There will be two lift tracks in each tube while additional space must also be incorporated to service both the lifts and the tube, should any faults occur. A diameter of 10m was chosen to be capable of housing all the required services and structurally strong enough to connect the central sphere to the torus.  Radiation shielding in the spokes will be provided internally in a similar manner to the torus.  However meteorite shielding will be installed externally using a reduced foam configuration of the Trans-Hab shield.  This thinner set-up will enable the linear motors of the moveable spheres to function through the shield layers.  Movement of people and equipment between the spheres and the spokes will be catered for by 3m high doors, placed every 18m (from the base of one access port to the base of the next) along the spokes’ 515.9m (from the outside of the central sphere to the outer wall of the torus) length.  These doors will be placed on each of the sphere’s floors (see 7 Internal Design) in order to connect with the lifts running through the spokes.

Power Generation Systems

When choosing a power system for the space settlement, we considered three main options:

1.       Nuclear power

2.       Solar cell (PV) power

3.       Solar thermodynamics

Nuclear power has the advantages that a large amount of energy is easily created, and in relation to other options, it takes up a lot less space. However, nuclear power is generally not in favour with the public at the moment.  If anything goes wrong, it can be potentially dangerous, so the system needs constant monitoring.  It may be more difficult to get corporate or government sponsorship for the project if nuclear power is involved.  Also if a fission reactor were to be used the necessary heavy fuel elements would have to be shipped from Earth.

Photovoltaic power is the most widely used power source in space; most materials are readily available from the Moon or asteroids. It is relatively cheap, and the facilities required will have been developed for the production of solar power satellites.

The final option, solar thermodynamics, works on a similar principal, turning the sun’s energy into power. Initial suggestions are that the efficiency is higher than that provided by photovoltaics, but the disadvantage is that there are moving parts, that will inevitably break over their lifetime. This means that if they were to be used as the power source of DL4, they would need to be easily accessible for repairs. Therefore, they would not enjoy the best possible location and could not function to their greatest efficiency.

Taking this into consideration the best option, based on current technology, was to use photovoltaic cells. There are however further factors that can increase the efficiency of photovoltaics depending on the materials used and the location.

Power Generation Systems – materials

Different materials can be used for the PV cells. Options include cadmium sulphide, copper sulphide, copper indium diselenide, crystalline silicon, and amorphous silicon.

The cheapest way of producing these cells is using resources from the Moon or asteroids. Low escape velocity means lower costs (fig. 3a). The production of solar power satellites (SPSs) will be one of the main economic activities of DL4.  It therefore makes sense to use the permanent facilities for such activities to produce PV cells to power the settlement itself. It would also be convenient if the production of the cells could begin as close as possible to the start of construction.  The ideal source for these materials therefore is the Moon.

This rules out many of the previous options, as they are not abundant in lunar soil.  (Future discovery of significant mineral deposits may make them viable options.)


Maximum Percentage Present



ppm Present








































Selenium, Cadmium


table 2.2b - Major Constituents of Lunar Soil (ref. 4)

It does however, leave the possibility of both crystalline and amorphous silicon cells.

Many methods have been suggested for the processing of lunar soil to produce silicon, however, the most convincing argument seems to be for the use of fluorine, brought to the Moon in the form of potassium fluoride, to react with lunar silicon and form fluorosilane (then reduced to pure silicon).  The fluorine can then be recovered from the MgF and CaF2 by reaction with KO2 (ref. 4 ).  This process has the advantage that is less sensitive than other methods proposed (e.g. – aluminium reduction) to the composition of the material used, so the rock used would not have to be significantly tested before use.

An additional advantage of this process is that it produces a high yield of oxygen from the lunar regolith, and that it produces metals such as aluminium, titanium and iron, and other by-products like magnesium and calcium oxide, which are either already purified or are relatively easy to separate in pure form.  The disadvantage of this process is that it requires an initial amount of potassium fluoride to be brought from Earth. As this is recovered, and as production of solar cells will continue after those for the settlement itself have been produced, it is not seen to be a significant barrier for its use.

However, if this does pose a problem, aluminium reduction could be used.  This process has the advantage that the aluminium is available on the Moon. The aluminium acts simultaneously as a reducing agent for the silicon dioxide, and as a solvent for the silicon that is formed.  The silicon can then be crystallised out in nearly pure form, which also saves on energy required for purification.  The reduction reaction is exothermic, so it only requires heating to melt the slag, and start the reaction (ref. 4).

If amorphous silicon is chosen, then silane gas must be manufactured. All the present processes for the manufacture of silane requires materials that are not available on the Moon. Depending on the methods chosen, these can be recycled. This taken into account with the fact that the actual mass of material required is quite small, and that the production of PV cells for SPSs will continue after the cells for powering DL4 are made, means that there are no significant obstacles to the use of amorphous silicon cells.

Normal crystalline silicon cells have an efficiency of about 15%, but some cells have been made with an efficiency of 19%. The efficiency of amorphous silicon is comparatively lower at about 10%. The advantage of amorphous silicon cells is that very little refined silicon is needed. (The active thickness may be as low as 1-2 microns). Amorphous silicon cells are also more radiation tolerant than single crystalline silicon, and the production process of the solar cells themselves can be made quite simple. If we were to use current technology, some materials would have to be sourced from Earth, or asteroids.  However many of these materials only make up the structure of the satellite as opposed to the cells themselves.  Therefore as total mass is not a major obstacle when items are manufactured in space, the cells could be fabricated on glass or aluminium, instead of terrestrial polymer films that are currently used.

There will be two other elements required for PV cell production.  Firstly, aluminium, which is the most likely candidate for use as a substrate, as well as for the structure and support of the cell.  If amorphous silicon is chosen, the required hydrogen, used for the production of silane gas (SiH4), can be refined from water sourced on the Moon.

Proponents of both types of solar cells have predicted future increase in efficiency, radiation tolerance etc., therefore it was decided to neither recommend one nor the other. However, it should be noted that as the mirror is located quite close to the station, it can be repaired or its cells replaced quite easily, therefore efficiency may be of higher priority than the frequency with which they have to be replaced.  (The same criteria would not apply to the choice of cell for Solar Power Satellites, due to their location.)

Power Generation Systems – location

The other factor which is a design consideration regarding power is the location of the solar cells. There are three options:

1.       On the outer hull of the settlement

2.       As a totally separate SPS which is permanently sun-facing

3.       Part of the mirror structure, above the station

The largest and heaviest part of a solar cell arrangement is the supporting structure.  Therefore, to minimise materials required and hence costs, it would seem sensible to use a structure that is already there for another purpose – the settlement itself, or the mirror shell.

Placing the solar cells in the approximately 3.2km2 free space on the mirror frame means that for almost twelve hours a day, they will be sun facing. If they are placed on both sides of the mirror, they will receive almost twenty-four hour exposure. For the times when the mirror is aligned parallel to the sun’s rays, a power storage system will suffice (see 15 Power Distribution).

Placing the PV cells on the outer hull of the settlement itself was ruled out as any side of the torus that is sun-facing at a particular point in time will always be encircled by the sheath of radiation shielding.

Creating and Maintaining Spin

The only feasible way to simulate gravity in space is through centrifugal force - i.e. spin. However, the method by which spin should be created and maintained is not clear-cut. Spin will have to be initiated when the station is constructed. Also, internal and external traffic may cause some degradation of spin over long periods. Therefore spin will have to be ‘topped up’ periodically.

The easiest and simplest way to create and maintain spin seems to be to attach small thrusters in balanced pairs around the rim of the station. These thrusters would have to be:

This seems to rule out traditional chemical rockets, as they are not particularly efficient and pose a risk of explosion.  Some form of electric propulsion  (e.g. an ion drive similar to that used on the NASA DS1 mission, or a Russian plasma thruster type drive) seems more appropriate.

The main advantages of electric propulsion are:

The main disadvantage of electric propulsion has always been that it can only provide low thrust propulsion. However in this situation, this is not a problem. The length of spin-up time is not particularly critical, and a low acceleration would allow the workers to become accustomed to the increasing rpm.  Additionally low-thrust acceleration is ideal for maintaining spin.

Finding a propellant for an ion drive that is available cheaply to DL4 is a problem - heavy inert elements like xenon that have been used to date are not available.  However, sodium, present in reasonable quantities (about 0.5% average of samples taken by the Apollo missions), is a possible alternative. It is extractable without much effort, relatively easy to store and would work well with an ion drive – it is easily ionised, not damaging to the engine materials, and a relatively heavy element. Current designs would need some modification in order to use sodium, however it still quite feasible.