Part of what I’m building for work includes making things watertight, which is always a little hellish. Since we love to makerbot everything we can, last week I experimented with making PLA 3D-printed parts watertight and waterproof.
I print on the lowest quality (for the fastest print) and at a pressure head of about 2cm the untreated parts didn’t leak overnight. Anything above that and you get leaking within a couple hours. Slow, but steady, and this also means you getting water inside the walls of your part which could become a serious issue later on as it probably won’t dry up. I was recommended tetrahydrofuran (THF), which is meant to be like acetone is for ABS — a solvent for PLA.
It works really well — I used some low-lint applicator q-tips to apply it to the print and afterwards it survived overnight with about 5cm of pressure head without any leaks. This was all that is necessary for my application, though you might find that at higher pressures you require more coats or a different tactic.
This particular bottle of THF looks pretty scary, but the chemical has about the same toxicity as acetone. However, like any good engineer you should look up the MSDS yourself and learn about how it is a mutagen if ingested, etc. etc. I actually left the piece to be ventilated for about an hour after I applied the THF because it I could still smell it.
THF is also used for vapor-smoothing PLA parts, though the toxicity issues get more serious there and I recommend this post that is detailed about how to be safe.
Another fun water-engineering fact I have run into is that you can run brushless DC motors underwater with no problems. Brushed motors would be a different animal because you could get sparks through the water when the commutator is swapping coils. Hypertextbook puts the resistivity of water at about 2e4 Ohm-m, whereas air is in the 1e16 Ohm-m range (so says Wikipedia), so you can imagine sparking in ways the designers had not anticipated. Brushless motors really don’t have this problem, and the resistivity of water is still far above the resistivity of copper (i.e. the wire in the windings) plus windings are generally insulated well. I might be a little wary of running a brushless DC motor in seawater, which Wikipedia says has a resistivity of 2e-1 Ohm-m. This is still far and above copper (10e-8 Ohm-m) but I wouldn’t want to take my chances, especially with a low-quality motor.
I did hour long tests at low (0C) and high (80C) water temperatures and saw no failures, though of course that doesn’t mean I’m not reducing the life of the motor. I imagine the biggest effect the water has is on the bearings; it’s hard to find good manufacturer’s information about cheap CPU fans, but they do list that that it has one ball bearing. Ball bearings tend to have shorter lifespans at higher temperatures, but the water might actually help to lubricate them so it’s a bit of a tradeoff.
Remember that the viscosity of the water slows down any fan meant to run in air, though in my case this was a benefit.