The first step I needed to complete to make the new fluorite facility was cutting those 12-inch fluorite crystals into a plate that could be used to power a stirling engine. So, I cut a 2 inch thick plate from the center. I decided to leave the triangular edges on the plate, since they aren't really in the way. Then, I cut a grid into the crystal only a half of an inch deep. I didn't want to harm the heat production capabilities too much, but I still wanted to have the better thermal conduction out of the crystal.

Ultimately, that meant that I had a plate that still comprised about 40% of the initial crystal's volume. That also means that this singular plate has about 25% more volume than the 8-inch crystal I grew before. Meaning I needed to prepare accordingly for the stirling engine, since it would produce a decent amount of heat.

A quick test showed me that, indeed, it did produce a large amount of heat, and I hadn't even traveled that close to the surface before I had to turn around because I started needing to use mana on thermal hands to keep myself from being burned while carrying the plate. That made me realize that I needed to assemble some of the stirling engine down in the low mana area, and haul it up to the surface for final production. At the very least, the future crystal plates will need to be attached to their metal housing underground before being brought to the surface.

Between the potential for the pure iron fluorite to produce even more heat per unit volume, and the much larger fluorite crystal grower, I'm actually somewhat concerned about the final temperatures involved. On the other hand, I can cut those plates thinner while maintaining a high heat production, distributed over a larger surface area, so it'll probably be quite ideal for a stirling engine. There are also questions about mass production that come to mind. Namely, each octahedron can only have one large central plate cut out, or two slightly smaller ones, then all further plates will be ever decreasing in size.

I'll have to wait until we have our first extra large crystal to really evaluate the proper method moving forward, but I suspect that I'll want the largest possible central plate for powerful machines, but then the remaining plates we'll actually choose an intermediate size, and cut them all to that size, so that we can have a standard plate for construction.

In any case, I ended up making a replica of the crystal plate out of stone so that I could design the whole stirling engine on the surface. Next, I brought the attaching point down to the crystal, got it attached, and carried that whole chunk of metal and crystal to the surface by using wood as an insulator to carry the increasingly hot piece of metal.

After I got everything hooked in, the stirling engine functioned well for a little while before steadily losing torque. At this point, I'd already spent six days on this project, so I was a little frustrated. After taking the whole thing apart, it was clear that the fine lead mesh had partially melted and formed something of a clog in the regenerator section. I hadn't really thought much of it, since the other large and efficient stirling engines were being used at cryogenic temperatures, but here that wasn't the case.

So, after another day, I had cleaned out the regenerator and replaced it with steel wire. After that, the stirling engine ran better than before. The actual power output was quite impressive, and I was confident that this one engine could easily run both furnaces and one of the grinders. One downside to stirling engines is that it's very difficult to properly harness the energy being produced, and in fact, I can tell that It's losing some efficiency in the hot side piston from sight alone.

To really harness the energy well and not have a long travel distance, the piston is actually quite a bit larger than the plate, meaning that the center of the piston casing plate is quite a bit hotter than the outside. While I've done my best to insulate the piston, it's still experiencing losses as a result. It's as if I'm using a blowtorch on the center of a metal dining table, and trying to cook on the other side of it. The good news is that I'm able to learn from this experience in designing the next engine.

Rather than making such a thick plate, for the next engine, I'm going to use two crystals to make 4 plates an inch thick, and put all four on the piston, equidistant from the center and the edge. The next engine is intended to power the rest of the mechanical grinders, so it might be slightly overkill, as I expect it to be more powerful than the last one. However, given how close I want this facility to be to the lab, I'm sure there are other things we could hook in to the spare mechanical power, like a freezer.

----------------------------------------

I was initially quite confused when the second engine wasn't working as well as the first one was. I had already spent another five days building this second engine, and I was remembering the fact I'd had to change to steel wool, and was worried I had a problem similar to that. When I had opened it up, however, I didn't notice anything really out of the ordinary. However, after using thermal hands to protect myself, and get a rough gauge of the temperatures involved by looking at how much mana I lost to keep myself safe over a short period of time, I figured out the problem.

The narrative has been illicitly obtained; should you discover it on Amazon, report the violation.

While the net energy going into the stirling engine was higher for the second engine, and the heat was better distributed, the highest temperature it was achieving was lower by a decent amount. I'm only hypothesizing here, but due to the source of the heat being more concentrated before, the overall temperature was higher. That extra distance from the center of the piston plate acted as a quasi-insulator allowing the piston as a whole to hold on to that energy better, since the bottom there was insulated somewhat by a layer of rock. This design has far more of the total crystal volume exposed to air, however.

So, it was leading to a two-fold efficiency loss. First, the efficiency of a carnot cycle, and by extension a stirling cycle, is determined by the hot and cold end steady state temperatures. So while I gained a small amount of efficiency in getting the whole piston head up to temperature, I was reaching a lower operating temperature, reducing the efficiency drastically. Second, I was rejecting more of the heat I wanted to turn into work to the atmosphere, causing even more losses.

So, I came up with two fixes for this. While I'd love if I could make some rock wool, it'd be incredibly tedious to make it with stone shaping, so I did the next best thing, and made a few layers of fine honeycomb pattern made of rock around the piston to help insulate it more. Second, rather than having the crystal plate exposed to air directly, I made a series of air vents in line with the surrounding insulation leading to the plates. While that might limit the mana getting in by a small amount, it should allow the air to heat up and move less, reducing convective losses, and allowing the crystals to reach a higher temperature.

That brought the second engine quite a bit passed the first in terms of total power output. However, it did take me two additional days to implement those changes. After taking another two days to implement those changes back onto the first engine, and the improvement was at least measurable, though the second engine still outperformed the first by a decent margin. I'd expected that, however. While the second engine is made of smaller plates, their total volume is twice the first, and while there are a decent amount of losses thanks to the crystal volume scaling factor, it only makes up about half the difference in total power output in this case.

----------------------------------------

With the power sources for the fluorite processing facility done, the only thing left to do was actually build it. I placed an order for four mechanical grinders along side the various mechanical components that I need to get power from point a to point b. Rather than try to convince Zeb that I needed a construction team with words, I brought him to see the two stirling engines I'd already made, and discussed how these basically run all the time, and aren't dependent on wind or water meaning they can be built just about anywhere. They do slow down quite a bit during the eclipse, but that's far less limiting than the other two sources.

Plus, thanks to stoneshaping, they're actually lower maintenance than our large windmills and the hydroelectric facility. Though the waterwheels are still easier to maintain overall. Basically, getting this facility online, and potentially training some stoneshapers and blacksmiths on how to make the stirling engines opens up an entire field of magically powered labor. Though I didn't tell him that most of my plans for the initial mechanical labor are cryogenically condensing air to re-grow our lost mana crystals. After that though, there are tons of applications. If we could make a mechanical lathe, for instance, we'd speed our production of parts up drastically.

After all that, I think I may have oversold him on it, because three construction teams showed up. This upset Tiberius somewhat, because there was a near constant stream of workers going down to our intermediate crystal to charge for four days until the facility was completed.

The building is a fairly unremarkable rectangle measuring sixty feet by thirty feet, with a twenty foot ceiling. It's constructed similar to the lab next door, where the floor is recessed slightly, and part of the building is built into the hillside, meaning it has no basement. The facility seems quite large on its face, but in actuality, most of the work occurs in a relatively small amount of the floor space. The rest of the facility is storage for fluorite and fluorite glass. Multiple labelled areas for handling raw, purified, and specially doped glasses were scattered around the facility. Fans bring air in to help fuel the furnaces, and also take air out of the same area.

I also realized that I should probably use the spare energy from the stirling engines to install some large intake and outlet fans in the facility. Between the furnaces, stirling engines, and natural heat fluorite that will be stored, the building could get very hot. I specifically have the air flowing perpendicular in the building as compared to the wall where the cold side pistons on the stirling engines are facing the outside world, just to help maintain efficiency and temperature control.

I took two days doing checks to make sure everything worked as intended. Now that I'm fairly confident everything is good, I think I can bring in some goblins and train them on how to run the place. I want to start making some iron doped fluorite and hydrogen fluorite as soon as possible.