economics of modelling electromagnetics vs structural space products

[Phillip Huggan]

Modelling ceramics for space products is easier than is modelling metals and metal-ceramic interfaces. Ceramics can be built from cheaper chemical components, while metal items need metal ore or pure metal. For reducing Monte Carlo run time, a ceramic wave function is easy enough to find. From a certain distance away from the node (atom), the lattice spacing doesn't matter. Merely staggering the wavefront and throwing the functions in a reverse projector gives a coarse simplification. Also, Boehmite is cheaper.
To make electromagnetically active objects, it is necessary to find not just one decently minimal energy wavefront, you need to find all the wavefronts electrons peak at given an excitation, then you put them in a reverse projector. At some point the fixed node ceramic atom locations need to be combined respecting fermion math, with the new electrons locations data.
I suspect near-field generating nanotech will be a jobs creator, but not as profitable as just the ceramic and iron bulk products. Without ice moon mines, I'd think only NASA would easily bother scaling up EM products too much. There is a continuum of product costs: sapphire with many inclusions might be moderately expensive, and rare metals like gold are even more expensive than are RF coil metals, but might be discovered in the solar system. I'm not sure whether antennae are cheap or expensive, giant radar arrays being there this half century or next half century, in the balance.

 

[Catastrophe]

Phillip, can you please go back a couple of steps and tell us what you are talking about?

Cat

 

[Phillip Huggan]

Sure. I'm learning modelling now. Figured half of machine learning in 2006 but never learned models until now. Before I kind of knew hard manufacturing hurdles ever since "Mining the Sky", the prerequisite intangibles of how to quickly and cheaply do experiments weren't clear. Making things out of BeO, Iron Oxide, and Sapphire using ball milling looks like cheap ball milling improvements. Metals is harder as there is no cheap precursor feedstock. I've just now learned the software is much harder for metal products. Monte Carlo and DFT update each atom. That isn't necessary. You can reverse clone and reverse projector the updates before doing much electronic math, to work with larger functions at little loss of accuracy. When you do this for above 3 substances, it is easier math and modelling (crack propagation and twinning defects still somehow need to be specifically modelled); extensions of Oak Ridge, UK and California papers. MC tells you the location of atoms using electron state modeling. To model a little ion engine, you need the atom locations as well as the, at least likely, electron wave excitation energies. This requires the atom locations as well as many electron locations for a given volume of substance. And these all need to be combined. The extra math will make metals harder to make reliably in space. It is beyond 2022 models. And things like antennae and solar cells are in between, not sure if 1.5x or 4x as hard to model as are the above three substances.
I used to want to go to U of T's Advanced Materials Lab to make nanotech or metallurgy physically, but now I'm sure RF coils for health and space can be R+D-ed better using software 1st. I suppose I'm just happy three locations have had good universities.