Rebuilding Science in a Magic World

I tested different configurations one at a time for a day each, with a day of downtime to adjust the build in between each test.  By the third test, I decided to make a second stirling cryocooler for testing, so I wouldn't have to take a day of downtime to make modifications.  By the fifth test's completion, I had the second cryocooler to work with.  I then decided to make a third one, and run two tests in parallel at a time, since I still had quite a bit of free time while the tests were running.

After fifteen days, I had three cryocooler tests running at a time, with one down for changes at a time.  With three cryocoolers, I was also finding myself having to spend some time each day or night producing the dry hydrogen to recharge the stirling system with, since each time I changed a component, I'd lose the gas during maintenance.  Unfortunately by this point, I still hadn't achieved any success with liquifying air across the ten different test runs, and without a thermometer that could reach that low, I really didn't know how effective I was being.  I did get some test results that at least led me to some preliminary conclusions on the matter.  

My testing began with three different sizes of steel wool packed into three different sizes of regenerators.  None of which yielded any liquid air.  I moved on to the next prepared metal wool, copper, which again across the nine tests, didn't produce any liquid air.  I got the same results for zinc and lead as well after a total of 36 tests over.  However, the copper tests were unable to freeze the thermometer in all but one test, and zinc shared a similar fate in three tests.

In the copper wool tests, the only one that did still freeze the thermometer was the larger regenerator with the smallest wool size.  In the zinc wool tests, the smallest regenerator failed with the largest and medium sized wool, and the middle size regenerator failed with the largest wool.

That information, while not perfect, indicates that higher thermal conductivity in the regenerator material is a problem, and that the regenerator might still be undersized.  It also seems to indicate that I want higher surface area.  As much as I'd like to just use metal shavings to drastically increase surface area, I don't have a good way to keep them in place in the regenerator, so I'm stuck with metal wools instead.

Thankfully, lead seems to work well, which means it is in competition with steel wool.  Lead is quite soft, which should make it much easier to make a finer grain wool to use as well.  So, I've decided that the next stages of testing will only involve steel and lead wools alongside new regenerator sizes.

The new tests were actually quite interesting.  The new tests included the previous finest metal wools, and two new finer grains, along with larger regenerators.  With our dies, this is probably the finest steel wool we can make, but I think we can still make even finer wool from lead.  Ultimately, this round of testing was 18 trials long, with three regenerators and three wool sizes for the two metals.  The largest of the new regenerators packed with the finest wools of both steel and lead also failed to freeze ethanol, however, the larger grained wools still worked in that regenerator.

My hypothesis is that as the wool gets finer grained, it has a higher resistance to airflow, causing losses.  Despite the fact the largest regenerator with the smaller wools failed, the smallest of the new set of regenerators with the finest grained lead wool produced a miniscule amount of liquid air in the 24 hour period.  Or at least, I assume it was liquid air based on it's behavior.  That meant to me that I was potentially on the right track.

With that result under my belt, I decided to focus in on improving the lead wool and regenerator design in the next set of trials.  I'll try to push the limit of how fine we can make the lead wool to see how effective we can make things.

After a week of testing, we were able to make an extremely fine grained lead wool by briefly heating the wire just before pulling it through the die and then catching the shaved piece on a plate to prevent the lead's weight from tearing the shaving.  It was very labor intensive to make this ultra-fine grained wool, but it turned out to be worth it.

Since I believed we had somewhat dialed in on the correct property of the regenerator material, I only used two sizes of lead wools in the next tests, though I did use eight different regenerator sizes.  The third smallest of these regenerators loaded with the finest of the wools produced the most liquid air, followed by the second smallest regenerator for total volume produced.  So, I made an intermediate size between those two, and found that this intermediate was the new highest producer.  I repeated that process of making intermediate sizes two more times before I settled on what I believed to be a fairly optimal size regenerator.

As for the amount of liquid air produced in a 24 hour period, we were still only looking at about a fluid ounce.  There was a secondary problem that likely reduced the amount produced, however.  That problem was buildup of ice and solid CO2.  So I took a few days modifying the building where the crycoolers were housed.  The room now has desiccants inside as well as in the newly installed air vents, to help reduce the total volume of water in the room.  I watched the new production over it's 24 hour period, and periodically scraped off ice and CO2 that built up, and ultimately, this cryocooler design produced about 3 ounces of liquefied air.

When we move to larger production, we'll likely need to have workers manually scrape material from the cold condenser piston occasionally as well.  Though, since I have these smaller cryocoolers, I do sort of want to put them to good use in the final facility by installing them in the air intake lines to the facility after the desiccant, to potentially condense out some of the CO2 before hand while also pre-cooling the air into the facility.  It won't be much cooling, but I'd hope that some is better than none.

With that, I'm fairly confident that I should move on to the larger cryocooler designing stage now.  It will also need a significant amount of testing done, and winter is only a month and a half away.  I want to make sure that by spring, I'm ready to do testing.

The first issue I had with the larger design was that I needed a complete redesign of the hot side heat exchanger.  Because the whole design was much larger than before, I needed a proper heat exchanger with small tubes to effectively transfer heat from the internal gas to the copper piping I was using.  Since I was using small copper pipe, that also increased resistance, meaning I needed more powerful pistons to drive the exchanger.  Due to the size of the whole device, I then needed to add a pump to a water reservoir that could run water over the heat exchanger, and the whole design kept growing and growing in size.  It grew to the point that the only way to properly drive this cryocooler would be to use the dam, as it would be the only thing providing enough sheer horsepower to turn the whole thing.

I'm thankful that I decided on a batch process, so that downtime can happen each year to repair the engine as needed.  The copper piping will probably need to be replaced at least once a year, and new grease will need to be applied regularly as well.  Hydrogen will need to be recharged frequently, and the regenerator's internal material will need to be replaced on occasion to keep efficiencies high.

I ended up spending quite a bit of time building the housing facility for the cryocooler.  It's a facility built onto the base of the dam, similar to the hydroelectric facility.  The main spillway is situated between this facility and the hydroelectric one, meaning I really don't have any more space for facilities powered by this dam if I need one in a future.  The hydroelectric facility does at least have extra space for more generators and machines though.

This facility could potentially house a second large stirling cryocooler in the future, and has space for the eventual distillation column.  The facility has thick, triple layer walls with an air gap between each wall.  The way in and out is through two sets of doors, where ideally only one set of doors is opened at a time to reduce air leakage. 

There are four moderately sized vents to provide airflow into and out of the building, each of which has desiccant trays, that can easily be removed and regenerated as needed.  On the inside, they each also house a box with a stirling cryocooler cold side condenser inside to pre-cool the air being drawn into the facility.  A main crankshaft from the dam will then power those four smaller coolers, while the larger stirling cryocooler will have it's own power source.

Ultimately, all the construction and design took until the last month of the year to complete.  Even with all that, I still need to do testing and resizing of components on the large cryocooler to try to optimize it's production of liquefied air once the spring rains refill the reservoir.