Fix JMAG analyzer's copper loss calculation method

[noguchi-takahiro]

The following line in jmag_2d.py is a way of calculating copper losses from JMAG:

eMach/mach_eval/analyzers/electromagnetic/bspm/jmag_2d.py

Lines 175 to 182 in a86c775

	@property
	def copper_loss(self):
	copper_loss_per_phase = (
	((self.current_trms / 2) ** 2 + self.current_srms ** 2) * self.R_wdg
	)
	copper_loss = self.m * copper_loss_per_phase
	return copper_loss

This code calculate the copper loss per phase and then multiplies it by number of phase to obtain the total copper loss. Here I believe m represents the number of phase (which should be 6 in the case of the parallel winding). However, if we see the definition of m:

eMach/mach_eval/analyzers/electromagnetic/bspm/jmag_2d.py

Lines 159 to 162 in a86c775

	@property
	def m(self):
	m = len(self.machine_variant.coil_groups)
	return m

this uses len() function to get the length of the coil group, which for example is defined as:

eMach/examples/mach_eval_examples/bspm_eval/ecce_2020_bspm.py

Line 83 in a86c775

"coil_groups": ['b', 'a', 'b', 'a', 'b', 'a'],

In this case, m should be return as 6 which is same as the number of the phase. However, we sometimes define this dictionary as follows:

This is example is coming from Sandia machine with 24 slot machine. In this case, it returns 24 (number of slot) when we calculate m.

I think this results in overestimate the copper loss by 24/6 = 4 times?