Clay fabrication objectives

I’m doing some stuff in the ceramics lab, with the objective being to construct a self-replicating micromachine. I have four key objectives to achieve this:

1. Full recursion: do the entire fabrication process, from separation to firing and assembly, using entirely components that can be fabricated with that process. Alternatively, if this objective is misguided (perhaps using metals, glasses, plaster, or plastics would be more reasonable), I want to discover that.

2. Material properties characterization, simulation, and optimization: I want to have quantitative measurements of the main physical properties of the material in both green and fired states. In particular:

   1. Tensile strength and modulus.
   2. Shear strength and modulus.
   3. Density.
   4. Thermal expansion.
   5. Separately, fracture resistance, even though that can in theory be predicted from the other properties.
   6. Abrasion resistance.
   7. Dependences of these on relevant process parameters.
   8. Heat resistance (e.g. maximum service temperature, softening point if any, sintering point, melting point, perhaps variation of other properties with temperature below the softening point).
   9. Creep, plastic deformation, and related complications; although I expect these will be very low for fired clay ware, plastic deformation is an enormously important process for shaping wet clay, which can have almost zero elastic deformation and creep but enormous plastic elongation.

   Given an adequate numerical characterization of the materials’ properties, simulation of its behavior under different circumstances should become possible; by running a series of such simulations with different designs and performing error and differentiation analyses on the simulations, a substantial degree of automation in design should become possible. If the simulation performs adequately, it should be possible to run such optimization processes during the fabrication process in order to automatically improvise responses to newly available information.

   With the first two of these items, I can do FEA analyses and optimizations for static loading; with the third, I can do dynamic loading as well. The others are useful for particular cases.

3. Generation-time reduction by process intensification: I predict that the primary figure of quality for self-replicating machinery will be its mass growth rate, which needs to exceed the IRR of available investments in order to be economic, and which is an exponential function of the generation time.

4. Miniaturization: somewhat secondary to the above three considerations, smaller is better. This is largely a means to an end: thinner walls will dry faster and fire faster (up to a point), linearly smaller machinery will have a proportionally shorter generation time, and the demands its mass places on its mechanical properties will be proportionally smaller, both permitting more geometrical freedom and simplifying calculations. Miniaturization also helps with safety and cost of materials and energy. However, miniaturization has limits: it requires more precise manipulation and measurement, surface effects like stiction and sliding wear become more serious concerns, and in particular thermal processes like firing pottery become more difficult to complete at smaller scales.