Current to fuse copper wire?

I have a project involving loads of current circulating in a tuned
circuit at 2-3 khz, and yesterday I suceeded in fusing a length of 35
sq mm copper multistrand welding cable - Any ideas what sort of
current would do that? Googling I did find one reference to current
density to fuse, and had the 35 sq mm been a solid bar it would fuse
at 5,800 amps.
This was part of the feeder circuit to an induction furnace coil.
Originally piped in 12mm copper pipe with a cross sectional area of 51
sq mm I had replaced it with 15mm pipe with a cross section of only
31.5 sq mm, so had paralled up a piece of 35 mm welding cable to give
a total of 31.5+35 = 66.5 sq mm and thought it'd be easily enough. But
as the pipes are water cooled and the wire wasn't the result was a
fused wire ! (I'm now re-doing it in 3/4" pipe with a 68mm cross
section so >>should
Reply to
Andrew Mawson
Loading thread data ...
Did you account for the skin effect at 3kHz? The combination of the current being carried in the outer strands, and the poor heat conduction between the strands carrying the bulk of the current and the less heavily loaded strands, may be more responsible for the failure than the total current.
Was the cable fused all the way thru?
Re the tubing, I've run around 5000A at 60 Hz thru short lengths of 3/8 sch80 copper pipe (140 mm^2 cross section) with no problem. I'm afraid I don't recall how much water was flowing thru the tubes; no more than a couple GPM, though.
Ned Simmons
Reply to
Ned Simmons
Ned had got the answer Andrew... skin depth for copper at 3kHz is 1.19mm, so the 35mm^2 welding wire will have an effective cross section of a bit more than 20mm^2 depending on the packing density of the strands in the cable.
Damn, I wish I could afford to be your next door neighbour :-)
Mark Rand RTFM
Reply to
Mark Rand
You also need to take into account eddy current heating. I assume this isn't Litz wire, so the eddy currents behave somewhat as if it was solid copper. If you didn't have a load in the furnace, then all the heating effect goes right into the coil.
You might be able to stuff a cable inside the water pipe, to get more area and still get cooling. You'd need somne pretty fantastic resonating components to handle 5,800 A continuous, or anything near that level.
Jon
Reply to
Jon Elson
current
conduction
afraid I
Good point re: skin effect : I'm advised it's about 1.2mm at 3 KHz but have been unable to source the calculations.
AWEM
Reply to
Andrew Mawson
tuned
current
enough. But
conduction
Well Mark if I carry on at this rate there should be a house price fall locally so maybe you will be lucky
AWEM
Reply to
Andrew Mawson
current
Jon,
Yes the 'cable up the pipe' method is used for the four 3/4" rubber water / current hoses from the electronics to the furnace body, but up 15mm pipe the water flow would have been too restricted - never mind the 3/4" pipe should sort things and will answer the skin effect issues as well.
As for the resonating components, the proprietory coil is square section copper pipe and the capacitors are vast and both are water cooled - at a peak I'm pushing 100Kw at this from a diesel generator driving a CFEI electronics unit which is microprocessor controlled. It is indicating currents of the order of 300 / 400 amps from the drive unit, but the circulating currents in the tuned circuit are of course far higher.
I have an ancient (1950's) clamp ammeter of the sort that have the clamp arms as the transformer laminations, the target cable as the primary, and a simple moving coil meter and shunts across the secondary. The furnace has two flow and two return pipe/cables. Measuring one cable when running at 25Kw it was pushing 800-900 amps indicated. But the meter is calibrated at 50 Hz and at the time I was resonating at 2100Hz and I have no idea how the higher frequency will affect the readings. From the fused copper wire it suggests that they may be about right !!!
AWEM
Reply to
Andrew Mawson

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.