Electromagnet design - help me sort these core materials?

[cvanc]

OK I've got a DC electromagnet design that is physically large which is driving me to alternative materials for the core.  The largest dimension is several feet and the winding is #8 copper.

Here are the materials I think I can realistically build this from, but I need to know how these compare magnetically. I want as much efficiency boost as I can get from the core but something like soft iron is right out the window.

Can you help me get this list in the right order for efficiency?  (remember, DC magnet)

Copper
Aluminum
Steel
Garolite
G12
Conductive carbon fiber
Anything else I should consider?

All input much appreciated, thank you.

[uer166]

All materials listed except steel have permeability of ~=1, which is same as air, and will behave as such. Steel* will also be a pretty lousy choice here as well..

*edit: meant stainless specifically.

[Andy Chee]

The core material is not just a form in which to hold and shape the wire.

The core material forms an integral part of the magnetic circuit.

[T3sl4co1l]

Uh, for what parts?  There's at least three important ones I can think of: core, bobbin, winding.

I think it's pretty clear the core needs to be made from some iron alloy (electrical steel), and the winding needs to be made from copper (minimum size) or aluminum (minimum weight/cost).

As far as efficiency (magnetic flux density and volume per power input), there is only one way to save massively, and that's by putting in a hard magnet.  Or, I guess, supercooling everything, preferably using superconducting wire -- but, that's a rather more expensive proposition.

You don't say what parameters are open to design at this stage, nor what budget, or material selection (other than a non-exhaustive list of possibilities), or even what specifications are (beyond generally being big?), so, there's not a whole lot to go on here.  For example, we can't even determine if mild steel might in fact be sufficient; Bsat tends to be lower than proper alloys (not to mention magnetic hardness), but if say 0.6-0.9T is enough, why not, right?

Tim

[johansen]

The cost savings of aluminum will pay for the more expensive grain aligned steel which can get you to 1.9T and slightly higher
.but only if you have room to work with.

If not then copper costs you well over 20 times as much (3.5 times the density and 4 times the cost per pound, plus markup.)

But it has a higher electrical conductivity which means lower losses... Which may never pay for themselves.

Motors and transformers prefer copper not just because of its lowe loses but its greater thermal conductivity which can transport the heat out of the core better.
The smaller the core the less the core size and weight.

Many distribution transformers have a thick aluminum sheet secondary of just a handful of turns but a copper coil of round wire with paper insulation between each layer.

So the copper is less than 75% fill factor but the aluminum is near 95%.
Its done this way because the copper primary has to be able to handle 100Kv impulse, but not the secondary.


Last time i looked up copper sheet metal the cheapest i could find is 16$ a pound. Aluminum was 1.

[Kleinstein]

The thermal conductivity of aluminum is not that much worse than copper. A difficulty with aluminium is however making the contacts.
If there is space aluminium windings may be a real option.

With windings that have am outer diameter significant larger than the inner diameter it can be worth using thicker wire at the outside. Ideally the wire cross section scales like the turns length.

The core may experiance significant mechanical forces too. This may requite some more structural steel and not just something selected for the magnetic properties. A plus for the usual transformer laminations is that they can be cut to shape relatively easy. For a single unit the pure material costs are only one part. Manufacturing and logistics (getting the matirial) are also important.

[johansen]

The thermal conductivity of aluminum is not that much worse than copper. A difficulty with aluminium is however making the contacts.

Sounds like AI

Cu is 400mwk al is 200 to 237