Electrical shake-down test for stators

[Beaufort]

Is it possible to pass current through the 3 phases of an axial flux stator in order to  approximate a temperature rise due to I^2R losses in the windings?  The idea is to check thermal transfer characteristics at higher RPM's without hooking up a large driving motor to the alternator.  I'm thinking of 15-25 amps at 12V...does it matter for thermal reasons to go with AC or DC?  It seems that keeping it DC would keep the rotor from running away during the test?

[Flux]

You can pass current through the stator windings to create heat easily enough. It may not match the working conditions with no forced cooling.

If the magnet rotors are not present you can use ac or dc. If the magnets are present you had better stick to dc to prevent the dreadful vibration you will get with ac.

In normal operation the current flows through 2 phases for 2/3 the time so you will need less current with this testing method to simulate the working conditions.

Normally when running the line current is about 76% of the dc current into the battery so you should feed that into two phases in series. With no forced cooling perhaps 50% is a more reasonable figure but I don't think much is known about actual cooling when running.

Flux

[Beaufort]

Thank you...sounds like I'm on the right track.  The only major thing I'm not going to get is the force pulling the coils toward the magnets as they pass.  It seems like many of the stator failures here are a combination of too much current (high heat) and the coils push the softened stator covering into the magnets (which rubs away and gets weaker and weaker).  It would be good to include this somehow in order to get a more accurate power/time rating, but I don't know how.  The end goal here is to have a good feeling about sustained and peak power ratings on a given alternator, and trying to learn about heat flow and cooling in the process.

I'll publish something if I get useful data.

[Flux]

The magnetic forces on the coils are not that high except during a sudden short ( braking).

As long as your resin doesn't go soft you should be ok. Perhaps some measurement of hardness as you heat it would be a guide. All resins soften with heat, normal epoxy is virtually useless above 100C. Polyester is not really rated for high temperatures but short term it seems to hold reasonably well to about 140C. Vinyl is supposed to be better. Special epoxies can survive 250C in VPI motor windings but I don't know how soft they are ( core slots hold coils).

Bear in mind that the failures you see here have burnt copper and that is with wire rated for 200C service so you can guess the temperature they have reached.

I also have reservations about sticking coils together with super glue, I have no idea of temperature rating of cyanoacrolate bur I suspect it is just a messy contaminant at 200C and will be pushing the coil apart. These are the sort of things you could find out in your life tests.

In the end you can't run electrical machines at 40% efficiency and expect huge ratings unless you do something drastically different from normal construction. Best keep to a sensible rating or adopt mppt and get efficiency up to 70% plus then this burn out worry mainly goes away.

Flux