quad wound coil & SCR scheme / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

quad wound coil & SCR scheme

previous 1 2

Move Thread

LAN_403

Turkey9

Thu Jun 04 2009, 03:22AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Klugesmith-

p.s. don't forget the voltage reversal problem with your electrolytic capacitors. I bet real PITs (designed for efficiency) have an underdamped, oscillatory impulse. If you add a clamp diode then di/dt and plasma motor action are truncated after first 1/4-cycle, wasting the remaining energy in the circuit.

Actually I'm not using electrolytics.... I just got my caps today. DC link ones. Metallized polypropylene rated at 320uf at 1200V. I have four of them and they have really good esr and L ratings. Here they are. CDE was nice enough to sample them for my science fair.. YAY! I'm also going to use some kind of circuit to save the left over current from the under-damped oscillation.

So more plasma ionization (?), currrent, acceleration power, and specific impulse during the shorter discharge time, we hope.

The faster the rise time, the more acceleration of the ions. I'm not really worried about the ionization in my design, but fast rising electromagnetic field is crucial. One of the main points I was trying to achieve was a shorter rise time so I'm very excited about what you found. Did you use FEMM4.2 to do that? I've just started using that software... Is there anyway you could tell me a little about how you did it so I could replicate it? I don't need step by step but a hint or two would be nice. Thanks!

klugesmith

Sat Jun 06 2009, 06:22PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1715

Turkey9--

I envy your capacitors.

The conclusions in my last post (nothing but ratios) were all made by inspection, and I tried to explain the reasoning. Spiral illustration was drawn with MS Excel.

FEMM is a great tool (God bless Dr. David Meeker and Foster-Miller!), but I haven't played with it since 3 years ago. That was for solenoids, but it would work just as well with a pancake coil. Just a different orientation of narrow rectangular wound region, or row of discrete wire cross-sections, in the axial section view. If you understand the basic principles of low frequency EM fields, you can get coil inductance and a quantitative field map of the working space, and glitzy pictures for a science fair board

, without having to go analytical on the problem. More tediously, you could use Excel (I found an add-in that does elliptic integrals, necessary for field of a current ring).

At the risk of being too pedantic: consider a few more walking steps before running.

As I understand it, you have made initial choices of C, V, and coil size. You have one degree of freedom, the number of turns, to match the coil to the power source. It's like rewinding an electric motor for your own choice of voltage-current and speed-torque.
So if you haven't done so already, pick a starting point (N value) and figure the coil inductance and resistance by formula. Then figure the LCR damped oscillator parameters. Learn how to write the lowercase zeta character.

Now it gets more interesting. Changing magnetic flux induces an electric field, hence current and Lorentz force in conductive projectile. A useful step might be to model (and even test) your coil as a ring-launcher before making it a plasma thruster. The metal ring has an exactly known extent and conductivity, and can be tested at low energies, where the geometry is practically constant during the current impulse.

If this is a substantially efficient motor, then the primary circuit is strongly affected by currents in the secondary. To the LCR model we add a coupled L in a circuit loop with the the secondary R. Referred to the primary side, it looks like a reduction in L and increase in R and damping. Now the more damping, the better -- it represents energy delivered to the secondary (though to be useful work, the secondary has to move). With FEMM you could determine the secondary L and coupling coefficient.

Have fun!
-RIch

Turkey9

Wed Jun 10 2009, 04:44AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Well it seems to me that your reasoning is very sound and it sure has helped me a lot in visualizing the circuit. I've done a good amount of figuring for the LCR circuit already, but am no where near done. I'm trying to match it to my equipment so it's not going to be perfectly optimal (keep current below 3.5kA per coil section).

I'll make sure and model all this in real life, also. Thanks a lot for bringing to my attention the affect of the secondary side, now the physics fit a lot better with what I would expect to happen. I don't know the electric conductivity of an ionized substance, but I'd assume it's close to that of a metal, am I right? I used aluminum in FEMM a little bit already but I didn't check things such as inductance. So as the object being accelerated gets farther away from the coil, is the inductance going to change?

I'm not sure if this is really the right section for this thread but I guess plasma can still be considered a projectile!

Turkey9

Wed Jun 17 2009, 12:37AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

OK I modeled the coil in FEMM and got some interesting results. The inductance of a single coil section quadrupled compared to the same coil section geometry alone. The inductance of that single section is also the exact same as the inductance of the coil with only one wire. But the resistance seen is one fourth so that'll speed up the rise time! So I will in fact have one fourth the pulse time but the same strength magnetic field!

KLH

Thu Jun 18 2009, 11:15PM

Registered Member #1819 Joined: Thu Nov 20 2008, 04:05PM
Location:
Posts: 137

Turkey9 wrote ...

The inductance of that single section is also the exact same as the inductance of the coil with only one wire. But the resistance seen is one fourth so that'll speed up the rise time! So I will in fact have one fourth the pulse time but the same strength magnetic field!

The reason the inductance values of the coil section and the one-wire coil are the same is because, as discussed before, they are well coupled to each other, since they are wound tightly together. As you said, the resistance is also reduced by a factor of four. However, Barry's RLC simulator shows that in most practical coilgun circuits, the LC pulse time is not as dependent on the circuit resistance as it is on the actual reactance values. Reducing the resistance by a factor of 4 will reduce power dissipation and shorten the pulse time a little bit, but the pulse time will not be anywhere near one-fourth its original value.

In addition, the magnetic field strength is dependent on the total amp-turns (turns of coil * coil current). As long as the resistance of the LC circuit is kept negligible for a given system voltage, the same number of amp-turns (for example, whether the number of turns is 10000 and the amperage is 1 or the number if turns is 10 and the amperage is 1000) will give the same magnetic field strength.

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob