SpudFiles

ok, for a school project, im making a coilgun. This is going to be a high power one, a company donated a 1200v 200A SCR(5200A surge), and i want to use it to its full potential. Im not asking about electrical aspects, i know all that, but i need help on the mechanical portion.

As some of you know, when that much power goes through a coil, it contracts will trmendous force. I was wondering what material to use as the barrel that can withstand this force, that easy, but it also needs thin walls for maximum efficency. PVC is way too thick. What AWG wire should i use? im thinking 10AWG.

thanks.

You have a few options, really. You could use something like polycarb, wrap a weaker material in carbon fiber, or use a non-ferromagnetic metal with a lengthwise cut along it to reduce eddy currents.

You should also consider injecting some type of epoxy into your coil after it's wound to increase strength.

No idea about the wire thickness.

carbon fiber O.o , i think ile use brass/copper with slots. How do i make slots in the barrel? i only have access to a machine shop until may 25th (school) do you use a mill drill? I heard epoxy is bad, its brittle, so i shatters. I need a flexible yet strong glue. if i used 1kv capacitors, then how much capacitance could i go for with 5000A at short circuit? only thing i can find is barry's RLC sim, but that needs inductance/

You can use a dremel to make the slots.
Are you sure you know all about the electrical aspects?
You need to know the inductance of the coil to calculate your capacitance, if you don't have any inductance you can get 5Ka with a 10uf capacitor if you want.
What projectile?
What capacitors do you have?
If they are highvoltage pulse capacitors you maybe will have better luck with a induction coilgun instead of a reluctance coilgun.
Edit:Typo

I think i know what i need to know, until i find out something
I havent got the caps yet. I think i think i will go with inductance, as for the caps i was thinking electrolyctic (there cheap). I found some from vishay that are about 5$ for 500J at 450V. that is unless you have some leftover from your 28.8KJ bank =)
As for the projectile, im not sure how big i should make it to avoid saturation. What if i wanted to do a electrothermal gun? that doesnt use inductive loads.

I think you best bet for the barrel is small ID plastic pipe, perhaps just a BIC pen body. Heck, see if you can imbed a fairly large gauge nail well into a piece of wood. The BIC pen would be a cheap barrel.

The gauge of the wire is a critical design parameter for a coil gun. It is a much more complex relationship than just "use a small G (large diameter) wire to minimize resistance".

The wire gauge affects the numbers of turns per unit volume (hence the inductance) and the resistance of the coil. It is not readily obvious, but for most coil guns you do not want to minimize the resistance of the coil. You want to balance the resistance of the coil (and the inductance of the coil and capacitance of your cap) with the amount of time needed to accelerate your projectile through the force zone. A large diameter wire will have lower resistance but the RLC relationship will give a power pulse that is too short for the ammo to respond to.

You should definitely fiddle with a good RLC simulator. I like <a href="http://www.coilgun.info/mark2/rlcsim.htm">Barry's</a>. (Wander his site, it is excellent.)

You should also remember that since the power pulse is pretty short you can grossly over-power the wire without it melting. IIRC, you can put 10 to 100 times more power through the wire than it's nominal rating. For example, 14G copper wire is normally rated for 15A at 120VAC (1800 watts, ~2HP). In a coil gun, 14G wire will easily handle at least ten times that much power (150A at 120VAC, 18KW). I believe Barry's site has a good discussion of the power handling capabilities of wire in a coilgun.

What are you going to use for your high voltage power supply?

yes, ive seen barry's site, its excellent. I played with his RLC a while back, as well as the air cored inductor sim. The SCR has a rating of ~5000A for 10ms, so i need to stay within that paramater. I also need to keep the reverse voltage low, as then ide need a grossly expensive diode to protect the SCR (some how i doubt there gonna give me another). With this ile put a diode anitparralel with the coil, on a coulple sites ive seen just a diode, but others use resistors to desipate the power, thus shorter pulse and less suck back. I dont want to use a BIC tube as the barrell, any projectile that will fit in there will saturate, ive already expierenced saturation past 40 joules with a bic tube.

To find a good wire thickness, i think ile discharge the power through a short lenght and see how much it heats.

About how long does it take ammo to respond to a strong magnetic field?
Also, barry coil sim, when used with RLC sim, does not give accurate readings, it doesnt factor in the projectile.

For the charging, i think ile use a MOT.

Correct me if I'm worng, but isn't an SCR a type of diode?

i think its a transistor switch that once it turns on it doesn't turn off till you disconnect the power.

Oh, I know that, but it's short for Silicone Controlled Rectifier, and thus only conducts in one direction....

TurboSuper wrote:Oh, I know that, but it's short for Silicone Controlled Rectifier, and thus only conducts in one direction....

Yes DC use only. A Triac would be the AC version of an SCR.

rp181 wrote:About how long does it take ammo to respond to a strong magnetic field?

That is a pretty difficult thing to calculate. I don't recall ever seeing a reliable simulation of the ammo actually moving through the coil. The way I approach the problem is kind of like this;

1. How fast do you want it too shoot?
2. Estimate the starting distance from the center of the ammo to the center of the coil. Usually people figure the center of the ammo starts at about the edge of the coil. (Of course, this is a parameter that you'll optimize once the gun is built.)
3. Assume average velocity during acceleration is half the muzzle velocity. (An incorrect assumption but close enogh.)
4. Average velocity X starting distance gives you a time.
5. You want your RLC circuit to fully (or nearly fully) discharge in that time. If it discharges faster than that time then your wasting energy. If it discharges slower than that time then you'll get suck-back. (Assuming you have enough total E to accelerate the projectile.)
6. Calculate the KE of your projectile at the target velocity. Pick an expected efficience for the gun. For a coil gun figure what, 1~5% efficiency? Using E=0.5(capacitance in farads)(voltage)² and the efficience gives you an estimate of the needed capacitance and voltage for the cap.

So, you need an RLC that discharges in the time period defined in step 4 and the total power in the cap as determined in step 6. Here is where you might decide to go with a smaller diameter wire. Smaller diameter wire has more resistance and increases the RLC of the circuit. You don't want the RLC (or just the RC) time to be too short, the ammo can't react fast enough. You need to match the RLC (or RC) to the target time from step 4.

You might want to look at <a href="http://www.inpharmix.com/jps/Jim%27s%20 ... g.html">my coil gun page</a>. The gun itself sucks, and the coils listed really suck, but there are some nice simple electronics for measuring the muzzle velocity and discharge curve of the RLC circuit. If you can measure those two things it makes it a lot easier to optimize the coil to your cap and ammo. Here is a typical data trace --

The first peak is the voltage across the coil (~300V at the negative peak), easy enough to get the discharge time and the RLC time constant. The next two peaks are the ammo (a BB) passing through a pair of photogates that are 1" apart. This partiular BB is moving at a blazing 45FPS. The detector electronics are pretty cheap, certainly less than $20.

starman wrote:

TurboSuper wrote:Oh, I know that, but it's short for Silicone Controlled Rectifier, and thus only conducts in one direction....

Yes DC use only. A Triac would be the AC version of an SCR.

Actually, SCRs are frequently used for AC circuits. As drex said, SCRs have a problem, there is now way to turn'm off except removing main power. This kind of defeats the purpose of the SCR in many circuits since the SCR is a high power switch. It takes another high power switch to turn it off.

To get around this problem, SCRs are frequently used in AC circuits. When the gate voltage is removed the SCR will turn itself off at the next zero crossing. Of course, that means that the SCR only conducts ~50% of the time when the gate is held on for long time periods.

SCRs are also used in capacitive discharge circuits, such as in a coilgun or photoflash. When the cap is fully (or nearly) disharged the SCR turns itself off. A capacitive circuit is not really a DC circuit, it is, in many ways, more similar to an AC circuit than a DC one. Though, as Starman said, the current flow is still in only one direction.

IIRC, a triac is just a pair of SCRs wired in anti-parallel.

If you wanted the switch to switch of at a certain time, you wouldnt use a SCR, you would use a IGBT. SCR's are commonly used in tri-phase power service.

jimmy101:

Thanks for the info and link. Part of the project is playing with factors like projectile starting point. Ile play with some differnt numbers and see what i get.

Turbo:

It is a diode, you just choose when you want it to start conducting, but wont conduct the otherway.

Anyone know how to calculate the bleeder resistor for a given capacitor?

Emailing some companies now to see if i can get some caps donated.

rp181 wrote:Anyone know how to calculate the bleeder resistor for a given capacitor?

By "bleeder resistor" you mean a resistor to discharge the big-ass cap through to make it safe?

As a rough estimate you can just use ohms law. The peak (initial) current out of a cap charged to 300V discharging through a 1K ohm resistor is 0.3A. 0.3A at 300V is 100 watts. So a 1Kohm, 1/4watt resistor will probably detonate.

If all you had was 1/4watt resistors then the total intial power is less than 0.25W (that would be a very conservative design). At 300V that would be 0.8mA. 300V at 0.8mA is a 375Kohm resistor. If your cap is 1,000uF then it'll discharge ~67% through a 375Kohm in ~6 minutes. The resistor will get hot but it shouldn't have any problems.

The best (cheapest, most readily available...) bleed resistor is just a couple of 120VAC light bulbs. A 100W light bulb is about a 100 Ohm resistor rated to 100W. That would be a pretty expensive resistor if you tried to buy an actual power resistor of that size. If your cap is charged to 300VDC then you probably want to use two 120VAC bulbs in series since initially you are running them at more than twice their rated voltage.

Assuming a 1,000uF cap charged to 300v then disharged through a pair of 100W bulbs in series;
RC = (200ohms)(0.001F)=0.2 seconds to disharge the cap to ~33% of it's initial voltage. Basically, the light bulbs would just blink.

Initial peak current through the bulbs would be 300V/200ohm=1.5A. Inital peak power would be (1.5A)(300V)=450W. Two bulbs, so 225W each. That's twice their power rating but the pulse only last for a fraction of a second.

My personal favorite bleed resistors are three or four of the small 25W 120VAC bulbs used in nightlights wired in series. (The kind that used to be used as Christmas tree lights.) Roughly 400ohms each, several in series will discharge a big-ass cap in a few seconds.