Electronics > Projects, Designs, and Technical Stuff
Everything wrong with amateur railguns and how to make them better
(1/8) > >>
Honda Rider 271:
First, I am here for guidance not to pontificate so please add to my post and correct any and all mistakes and misconceptions.

Awesome comprehensive reference reflecting state of the art in the late 80s: 300+ pages I haven't finished it yet: https://apps.dtic.mil/dtic/tr/fulltext/u2/a192102.pdf

OK so tons of individuals have built railguns and these railguns all have a few things in common:

1) Relatively small, high cost capacitor bank power supply
2) Terrible efficiency below 1%
3)  Because of 1 and 2, really weak projectiles

The most powerful individual railguns that appear to have been built are on the order or 30kj of input energy (edit: apparently sam barros had one with 100kJ). Even 10% of that in terms of kinetic energy would be pretty impressive, on the order of a high power hunting rifle. Unfortunately from firing demonstrations it's clear that the actual projectile energy these designs deliver is on the order of 100 joules or less: somewhere between a small-bore air rifle and a .22. Man I live in Texas, surely we can do better.

In order to make an AWESOME railgun one or more of the above issues needs to be solved. First the pulse power supply:

While it's possible to find relatively large high power capacitors for salvage this leaves you at the mercy of what's available. Usually salvage cap lots are significantly limited in number. So you may be able to assemble a modest bank for relatively little money, but you probably won't be able to build a huge one. Conversely, if you want to build a big bank, you can buy as many capacitors as you want from the production market but this is extremely expensive with even a 100Kj bank costing on the order of several 10s of thousands of dollars. Maybe it's possible to find a physics lab scrapping their trashcan sized caps but this strikes me as especially impractical.

Building your own capacitors for any reasonable bank size is not an option (correct me if I'm wrong).

Building your own high energy inductor IS a realistic option and it is probably possible to create an inductor storing on the order of 1MJ energy for a few thousand dollars. I'd settle for 100kj. The material of choice would probably be aluminum.

So the inductor should probably be the choice of energy storage for amateurs who want to build a badass railgun--NOT capacitors unless you are so lucky as to come across a reasonably priced, gigantic salvage lot.

Charging the inductor is an open question. Since the charge time is on the order of 100x longer than the discharge, maybe it's possible to use a relatively small bank of high C-rate lithium polymer.

Now the second part, increasing firing efficiency.

Why could amateur railguns be so inefficient? A few guesses:

--Improper dimensions and geometry of rails and projectile
--Not taking skin effect into account, leading to much more energy wasted thru resistive heating than expected
--Pulse width too short (cap bank left to its own devices can easily dump all the energy nearly 2 orders of magnitude faster than it takes the projectile to leave the rails even with short rails). This could be tuned with an inductor in series with the cap bank (but we don't want a cap bank, we want POWER)
--Friction. Even if you have a loose-fitting projectile and rely on plasma to bridge the gap your projectile could still be intermittently spot-welding or sticking to the rails. I don't know how much loss this would be responsible for but maybe a high pressure low temperature gas could help. Totally speculative.
--Reliance on single-stage rails. It's known that adding additional lateral rails in series can increase efficiency. I don't know how important this is or even what such a design would look like in practice. I think this one is mostly irrelevant for the time being since it should be possible to achieve efficiencies higher than <1% even with a single pair of rails.

Action plan:
--Obtain solid grasp on what makes an efficient railgun
--Figure out muzzle energy level desired (let's say 5kj, on the order of a 12G shotgun slug)
--Figure out what efficiency is reasonable with all realistic optimizations
--Design inductor capable of delivering (muzzle energy / efficiency), inductor supply
--Design switching for inductor charge and discharge
--Design rails and supporting structure, projectile and electrical interconnects
--Design loading / injection system

Anyway if you read all that maybe you have something to add. Let's make amateur railguns BADASS instead of pitiful exercises in wasting money and energy.
Honda Rider 271:
Update!

First I bought some capacitors. But before that some newly found low-hanging fruit:

1) Amateur projectiles (the armature and anything it drives) are too massive for the small electrical energy levels involved. Kinetic energy is .5MV^2 so other things equal reducing the projectile mass by a factor of 10 should increase kinetic energy by a factor of 10 even after accounting for new 90% smaller mass. Miniaturizing the projectile to a reasonable mass (<10g) might include conductivity and or thermal problems reducing the theoretical energy gain. The small scale high velocity laboratory railguns I've found in academic research all shoot extremely low mass projectiles.

2) Staging. Low power railguns will need such a small projectile to achieve high velocities that a plasma armature might be most practical. Staging is thought to work better in plasma armature designs and many research railguns appear to use staged input vs dumping the power in all in one location. I will try both cap bank + coil for pulse elongation and staged inputs to see which works better.

Capacitors:

I found 28 450v 4400uF caps locally for cheap. Cap energy is .5CV^2. This theoretically gives just under 12.5kJ maxed out but accounting for a 10% lower actual capacitance and 420V working voltage so nothing blows up 10kJ is a more realistic working energy. I paid around 30-40 joules / usd. If I can find large quantities of big caps at similar prices a 100kJ bank would be affordable. Even so, I think a large homemade inductor may become affordable somewhere before 1MJ. That's a $30K cap bank.

Pulse lengthening inductor:

Depending on the actual resistance of wiring, rails, armature and switching if installed the 28p cap bank can discharge itself anywhere on up from a few tens of microseconds. This is way too short for a reasonable rail length so I will try a relatively small inductor in series. Reasonable-size conductors may have too much resistance but the inductor can be bathed in an insulated box full of liquid nitrogen which is cheap and readily available. Staging is also a solution.

Injection system:

Pneumatic injection seems easiest. Ideally the charged cap bank will be switched to the open circuit of the rails and triggered by the projectile making contact. I'm not sure how much efficiency will be lost with armature switching vs using a big thyristor.
I wanted a rude username:


For maximum efficiency, I suspect you will need to build a pulse-forming network.

If you want to get really exotic, you could use a light-gas gun as your injector ...
T3sl4co1l:

--- Quote from: I wanted a rude username on December 27, 2019, 05:16:03 am ---If you want to get really exotic, you could use a light-gas gun as your injector ...

--- End quote ---

Better yet, instead of charging a capacitor, run a compressor to charge an air tank -- far greater efficiency, plus the lower peak power requirement means you can run it on a few LiPo packs instead of a fuckoff massive capacitor. :)

Absolutely serious -- this is all about transforming energy, right?  Why choose literally the absolute worst method when you could make a practical device instead? :P

Tim
I wanted a rude username:
Or one of those convenient, primer-actuated, deflagration-powered, remote lead delivery systems with a bore diameter not exceeding 12.7 mm. But the thing to remember is ... practical is inversely proportional to cool.  ;D
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod