This representation is more than academic. It explicitly shows that threshing millet is a multistep process, and each step may require different methods or apparatus. In the first process above, stripping the florets off the stalk is relatively easy, as is separating florets (mostly with still attached grain) from rachi. So this process must be coupled to the second process, separating floret from seed, followed by winnowing. That proved harder. The third process (abstractly) suggests a direct way to remove the seed from the panicle, without removing the floret from the rachi. This would be a simpler path to isolating grain -- if we can figure out how to implement it. The challenge is the millet construction: the seed is tightly held in the floret, and the floret is not so tightly held to the rachi.


Later we made this diagram showing 3 different paths we could take to isolate clean grain. We initially thought the middle path (2) would be most achievable.


Breaking up grain from floret can be thought of as "endothermic" (or more technically correct, "endergonic")-- it doesn't happen by itself (within a year), but requires energy.

Separating or winnowing seed from floret results in a more ordered state, so this step requires the extraction of entropy. According to the Second Law of Thermodynamics, decreasing entropy in the millet can only happen if we increase the entropy by even more elsewhere. I think the concept of "Gibbs Free Energy" : G = U + PV - TS, where U, P, V, T, S are energy, pressure, volume, temperature, and entropy respectively, helps quantify this. During winnowing the entropy term, TS, decreases, requiring additional U + PV to make the "reaction" proceed. In other words, winnowing REQUIRES adding energy and/or increasing the volume of the system (e.g., blowing debris all over). Gibbs shows there are trade-offs. Making a "clean" winnowing system that doesn't scatter the chaff everywhere will FUNDAMENTALLY take more energy input than a messy system.

Since separating incurs a fundamental thermodynamic energy cost, we should minimize separation steps, and focus on removing the seed from the floret, keeping the floret on the stalk (path 1 on the second figure above). Key to achieving this is understanding the tensile forces holding the millet components together, and how they respond to stresses. For example, the millet panicle has high strength to axial forces, but very little torsional strength, as can be realized by pulling and twisting a millet stalk in your hands, or as we discovered after more extensive experimentation with the Corn sheller concept.

In conclusion, both threshing and winnowing millet requires energy input (which is of limited supply for our goal of a human-powered system). Applying this abstract thermodynamic analysis has constrained the space within which we can hope to find solutions. Q.E.D.



Home Previous Next Contact Us