Magnetic Module - angular periodicity

[Lavanya-Bala]

Hello,

I previously simulated the IPMSM_A and the magnetic module worked fine. I am now trying to run a similar IPM machine but I face this particular error "In Magnetic model, Na_tot=2048 is not divisible by the machine angular periodicity (3). Angular periodicity removed". I don't see what I am missing here as I am running the same script used in https://www.pyleecan.org/tuto_Simulation_FEMM.html for both. Any idea why this is happening?

[BonneelP]

Hello,

When defining your simulation.input you need to set "Na_tot" which is the angular discretization. Then when defining your magnetic module you have set "is_periodicity_a" to use the FEMM periodicity. Unfortunately for now the way we handle periodicity is a bit strict: if we can't divide Na_tot by the machine periodicity we disable it. The Prius periodicity is 8 so 2048 is Ok but for your machine you can use Na_tot=2049 for instance and it should solve the problem ;)

Best regards,
Pierre

[Lavanya-Bala]

Thank you. That makes sense but doesn't work either. Changing Na_tot given the same error and removes the periodicity.

[SebGue]

Hello,

I haven't checked but maybe you could try 2046. I.e. if your machines has 3 pole pairs you should set a number Na_tot that is a multiple of 6 to also allow anti-periodicity.
Best regards, Sebastian

[Lavanya-Bala]

Hello,
While running the magnetic and electrical module in the version 1.3.4 I get the following error (For now, reverting to 1.3.2 helps):

out_femm1 = simu_femm1.run()
File "\site-packages\pyleecan\Methods\Simulation\Simulation\run.py", line 52, in run
self.run_single()
File "\site-packages\pyleecan\Methods\Simulation\Simu1\run_single.py", line 31, in run_single
self.mag.run()
File "\site-packages\pyleecan\Methods\Simulation\Magnetics\run.py", line 24, in run
out_dict = self.comp_flux_airgap(output, axes_dict)
File "\site-packages\pyleecan\Methods\Simulation\MagFEMM\comp_flux_airgap.py", line 171, in comp_flux_airgap
filename=self.import_file,
File "\site-packages\pyleecan\Methods\Simulation\MagFEMM\solve_FEMM.py", line 156, in solve_FEMM
femm.mi_analyze()
File "\site-packages\pyleecan\Classes_FEMMHandler.py", line 1919, in mi_analyze
self.callfemm("mi_analyze" + self.doargs(*arg))
File "\site-packages\pyleecan\Classes_FEMMHandler.py", line 70, in callfemm
raise Exception(ErrorMsg)
Exception: error: (anti)periodic BCs applied to dissimilar segments

Could you please let me know If I have missed something here?

[BonneelP]

Hello,
this error means that something went wrong in the definition of your machine in FEMM. Can you share a screenshot of the fem file ?
I just release v1.3.5 that includes some modifications on FEMM, maybe updating can solve your issue.
Best regards,
Pierre

[Lavanya-Bala]

Thank you, this update solved it.
But running an optimisation code brings the following error:
TypeError: uniform() takes 3 positional arguments but 9 were given
Traceback (most recent call last):
xout4 = solver.solve()
File "\pyleecan\Methods\Optimization\OptiGenAlgNsga2Deap\solve.py", line 286, in solve
raise err
File "\pyleecan\Methods\Optimization\OptiGenAlgNsga2Deap\solve.py", line 99, in solve
pop = self.toolbox.population(self.size_pop)
File "\AppData\Roaming\Python\Python37\site-packages\deap\tools\init.py", line 25, in initRepeat
return container(func() for _ in range(n))
File "\AppData\Roaming\Python\Python37\site-packages\deap\tools\init.py", line 25, in
return container(func() for _ in range(n))
File "\AppData\Roaming\Python\Python37\site-packages\pyleecan\Methods\Optimization\OptiGenAlgNsga2Deap\create_toolbox.py", line 76, in create_indiv
value = design_var.get_value(design_var.space)
File "", line 1, in
TypeError: uniform() takes 3 positional arguments but 9 were given

I can't seem to figure out what the issue is here.

[BonneelP]

Hello,

It seems that the error is within the design_var. Can you share the definition of your design variables?

Best regards,
Pierre