Re: Three phase induction motors

["David Meeker" <dcm3c@xxxxxxx>]

--- In femm@xxxx, Alberto Perez <albertoperarnau@xxxx> wrote:
> Hello
>  
> We are joining a group of people that are working with
> FEMM. We simulations on three phase induction motors.
> The first results are quite acceptable according to
> the analitical theory. But we have found some problems
> to define the current. Is the input value the average
> (RMS) or the peak value of the current?.

The input values are peak value of the current. More specifically, 
femm represents all currents, voltages, etc. as in harmonic problems 
as complex numbers where the amplitude of the complex number 
corresponds to the peak value (instead of RMS), and the argument (i.e. 
arg(a+j b)=arctan(b/a) ) corresponds to the phase angle. 

> If we define a closed rotor slot. It becomes saturated
> very soon in the interface between the slot and the
> airgap. Do we have to change the material properties
> in this area?

Currently, the program assumes linear material properties for harmonic 
problems, even if you define a nonlinear B-H curve. I've been 
thinking about doing an approximate nonlinear harmonic solver where 
the idea is to get the "right" amplitudes for the fundamental in the 
presence of saturation. (sort of like the approach described in Paoli, 
G.; Biro, O.; Buchgraber, G., "Complex representation in nonlinear 
time harmonic eddy current problems," IEEE Trans. Mag.34(5): 2625 
-2628) This is the sort of thing you are probably after here, since 
the narrow pieces of iron in your design saturate early on by design, 
and a lot of flux gets shorted through these sections if the 
saturation isn't modeled. However, I can't really say when I will 
have this solver ready--it's just something that I've been thinking 
about. Your idea about changing the matierial properties in this area 
(e.g. replacing part of it with air, perhaps?) would probably be a 
reasonable kludge. 

> For the magnetostatic calculation the stator current
> introduced is the magnetising current and not the load
> current which is three time higher.
>  
> Do you have any comments to our problem definition?
> They will be wellcome.
>  
> 
>  
> Albert Pérez albertoperarnau@xxxx
> Emili Valero emilivalero@xxxx
> Atanasi Jornet jornet@xxxx