Multi-layer magnet with separate coils

[Pharrahnox]

I am trying to make a very strong magnet that will only last for a short pulse, just a fraction of a second, with a high peak current.

I am going to use 1mm (18awg) wire, so the resistance should be pretty low, but with even just 25m, it would be expected to have 0.5Ω. I would like to have many turns of wire to make the magnet stronger, and I will be using 60V, so resistance could drastically affect the peak current.

Because I don't want an overly long coil, between 5-10cm, and a fairly small internal radius < 2cm, I will need multiple layers of wire. As long as I wind them in the same direction (clockwise or anti-clockwise) they shouldn't negate other layers' magnetic fields.

So what would happen if I were to have the same amount of layers, say 10 layers, and I connected half of them to a power source, and the other half to another identical power source? Would this then half the resistance, while still having the coils creating magnetic fields that would add together?

Would it create a magnetic field of twice the strength, due to a doubling of current, because of a halving of resistance (assuming the power source has negligible resistance)?

Also, how would it affect the inductance of the circuit(s)?

 

[UltrafastPED]

Perhaps you are reinventing Litz wire: http://en.wikipedia.org/wiki/Litz_wire
The links at the end of the article contain useful information on calculations for engineers.

Here are some other calculations/calculators:
http://physics.princeton.edu/romalis/magnetometer/coildesign/
http://www.midnightscience.com/formulas-calculators.html#formulas4

And some random papers:
http://cds.ismrm.org/ismrm-1999/PDF3/740.PDFA short pulse is equivalent to multiple frequencies of AC - you can do a Fourier decomposition based on your actual pulse shape to discover the frequency spectrum.

As the article points out, for higher DC currents the cross sectional area of the wire should be larger; for AC currents the "skin effects" become increasingly important as the frequency increases.

The Litz wire will split the current along multiple paths; all of the paths should have the same length, and should begin and end on common busses. This ensures that the magnetic fields generated by your pulse will be synchronized - and so by the superposition principle will result the maximum intensity.

 

[Pharrahnox]

Thankyou for the links.

I am using a DC current from a capacitor bank, and I have found a formula for the frequency:
f = \frac{1}{2∏\sqrt{LC}}
For my coil, it would be something like, C = 0.15F, L = 680μH, which would give a frequency of about 16Hz.

Since the frequency is so low, I won't need to worry about the skin effect, or the proximity effect, so the separation of the various coils would be solely for resistance purposes, to decrease the pulse duration.

But would it work as I am hoping? I am going to have 10 capacitors, and I am going to connect them in parallel, to increase the capacitance. But because the capacitance and resistance are higher than I want, the capacitors will take too long to discharge mostly.

So I thought if I had two separate banks of 5 capacitors, still in parallel, connected to two different coils of close enough to identical length, then the overall resistance would be halved, while still maintaining the same voltage source. This would result in double the current, but what would happen with the total inductance of the solenoid?

And if this is the case, then it would be even better to have 5 banks of 2 capacitors each, althought the capacitance would be lowered significantly.

Am I right in my assumptions, or is there something I'm missing/forgetting?