# Sizing cables in pulsed power applications

I am looking for rules for sizing cables for short pulses. For example, if I were to run 2000A DC current for 0.5 seconds in a cable, what size would I need? Sizing as if it were continuous current seems like overkill. I could calculate the temperature rise for each wire size and then pick one, but how do I determine an acceptable change in temperature?

• What's the duty cycle? And it might be helpful to think in terms of watts, rather than current. – Nate Strickland Apr 15 at 23:55
• What is your expected average transmission loss? [W/meter' x thermal resistance ['C/W] * thermal time constant [seconds] It takes seconds to heat up and a long time to cool down. – Sunnyskyguy EE75 Apr 16 at 0:52

If you plan on 2kA for 0.5 s, you may want to limit the temp rise to the temperature rise of 80'C

Consider:
4/0 copper cable D = 11.7 mm Area = 107 mm² 0.1608 mΩ/m

• Fusing current in 10 seconds = 3200 A

But if were computing the short circuit capacity of wire rated before fusing open, this was done by the Southern California Edison Company published in June 1928 .

• Good graph, but the numbers are unreadable. – EinarA Apr 16 at 1:46
• I can see if you click on image to zoom 2kA @ 0.5s = 100kcir-mils ( on AWG table it shows 100kcmil = (1/0)cable = 8.25mmD. ) meaning it should not melt (but might smoke) so 4/0 cable might be best for repeat pulses. – Sunnyskyguy EE75 Apr 16 at 2:12
• Zoom didn't work for me, but I only have a phone connection. Merely curious. – EinarA Apr 16 at 2:42

You know the temperature rise just like every other wire...Max Insulation Temp - Maximum Ambient Temp.

Calculating temperature rise due to pulses would require things like thermal impedance if you go with the physical models and those are basically simulations. 2000A is a lot of current so will produce a lot of resistive voltage drop in your wire so try this first instead...it's much easier:

1. Determine your allowable voltage drop.
2. Determine the resistance required to achieve this.
3. Find an AWG with this resistance.
4. Check a standard table to see if that AWG can't already support 2000A continuously. If it can, you're done (although it still might be tough to find a table with AWGs that can go up to 2000A).