Conclusions

This project was not successful in constructing a composite spin-cast mirror for two primary reasons:

• The materials used for the composite - Ultracal industrial cement as the matrix and bi-directional woven carbon fiber as the fiber - did not form a suitable composite.
• The cement surface was unpolishable and I was unable to find another polishable material that could be attached to the cement surface.

However, the project produced encouraging results in the direction of constructing such a mirror. Some highlights are:

• It is possible to spin-cast an even, smooth paraboloid from cement or epoxy - both of which are more viscous than water - using machinery with the proper tolerances.
• The focal length of the mirror can be determined by the rate of rotation, although not in the way I predicted. This needs further investigation and may involve the viscosity of the material being spun in the cast.
• The materials used so far are cheap and readily available, the most expensive being the carbon fiber. Most of the grant funding was spent on equipment, materials that ended being unusable for the project, and supplies related to the construction and use of the equipment. The hope is that other cheap and available materials with better thermal properties and surface hardness can be found and tested.

I plan to continue work on this project after researching what other materials could be used. Some ideas for new materials are:

slip clay

which is a clay suspension that may behave similarly to the cement while casting, but could be fired like a ceramic after casting. This has the advantage of better surface hardness and structural properties than cement, but the disadvantage of requiring a kiln.

balsa wood

or another lightweight wood used as a core and surrounded by a polishable, spin-castable material. If the direciton of the grain was chosen correctly this could provide as light a solution as a carbon fiber composite. However, if the thermal expansion of the wood is considerably different from the surrounding material, there may be problems with disfiguring at the mirror surface. It may also be difficult to obtain a lightweight wood in the size and shape required.

Another direction for further research is the use of a more intricate support system for the mirror. In this case, a non-structural, thin casting could be polished and then supported at hundreds of points by thin supports that attach to a structural underlayer.