How to make a good Capacitor?

From what I remember, Capacitor work by storing electric charge on an input plate that is barely seperated from an output plate by some non-conductive material. The charge builds up on the input plate until it leaps the gap of the non-conductive material and floods the output plate with all the accumulated energy at once. Increasing the size of the plates increases the ammount of energy the capacitor can store, and increasing the width of the non-conductive material does much the same. You can actually have a capacitor without the input/output plates touching anything; air or vacuum acts as the non-conductive spark gap.

How does Hafnia compare to water? That's what I'm currently using in my capacitors.

Mica, Polyethylene, or Hafnia are favorites. Hafnia can get expensive.

Not if you build them correctly. It is possible to get nearly perfect (50%) system efficiency with single capacitor designs. (And this doesn't force low performance with modest 300km range weapons being only slightly weaker than the 1.08 version equivalent).

This can make minimal rail designs break though due to higher initial accelerations, and in *some* cases using multiple smaller capacitors can give better total cost and performance because the rail can be slightly lighter.

Peak efficiency of multiple capacitor designs is lower than a single capacitor design, and maximum rate of fire will be calculated as much slower than for a single minimal capacitor for the same shot energy.

My annoyance right now is that the game assumes the capacitor is always fully discharged when firing. Best muzzle velocity is achieved with an over-sized capacitor and nearly constant acceleration. The constant acceleration indicates that the voltage on the capacitor is not dropping much during the shot. So where is the extra energy in the capacitor going after the projectile leaves?

Does anyone have a rail-gun design with acceleration that rises along the barrel? That would indicate that the accelerating current is limited by inductance and that system efficiency could exceed 50%.

Pulse forming networks can produce fairly rectangular pulses without much leftover energy. So that might be what's going on. The weight of that should be modeled though, ideally.

There's going to be an efficiency cap because as you're putting kinetic energy into the projectile, you're also storing energy in the railgun's magnetic field. When the projectile leaves the rails, that energy is going to get dissipated as arc flash.

I think you could recover the field energy, though, by lengthening the barrel enough that the cap-gun system rings all the way to the next current zero, with the cap charged in reverse. Then you flip the polarity of the cap and top up the charge for the next shot.