20-07-2012, 03:27 PM
PREDETERMINATION OF THE ELECTROMAGNETIC LOSSES IN HIGH SPEED ELECTRICAL MACHINES
PREDETERMINATION OF THE ELECTROMAGNETIC.pdf (Size: 801.21 KB / Downloads: 28)
INTRODUCTION
Several methods for predicting electromagnetic losses
have been proposed and tested by different teams. This
predetermination is particularly important for the
machines with a high rotating speed bemuse under
certain conditions the electromagnetic losses can be
relatively important and thus constitute a nonnegligible
design factor.
In this paper a method for predicting electromagnetic
losses, is proposed wich associates simulation results
with experimental losses measurements on the sample
of iron sheets under various flux density waveforms.
First of all, we present a synthesis of the state of the art
based on different methods for predicting the losses.
Equivalent magnetic circuit method
This method is based on the exact analytical
determination of the flux waveforms in the different
regions of the machine. Firstly, the equivalent magnetic
circuit needs to be accurately established for the relative
stator-rotor position (2). In order to determine the
permeance values and the flux waveforms in various
parts of the magnetic circuit, an iterative method is
generally used in following the procedure:
The iterative process starts with given initial excitation
mmf and magnetic induction B in the airgap values
chosen arbitrairy. Afterwards, using the non linear B-H
characteristic of the magnetic material, the permeance
values in the different parts of the machine are
determined. Finally, the mmf computed value is
compared to the initial one. In the case where the
waveforms of the voltage supply are known, the flux
waveforms can firstly be deduced in the stator parts and
then in the rotor parts by neglecting the stator winding
resistance per phase. Thus, the supply voltage and the
flux linkage are related by the expression:
U = dbldt .
CONCLUSION
In this paper we present a method which allows us to
predict the electromagnetic losses in a rotating
machine. The computed results obtained show that the
division of the machine in a limited number of zones
leads to a 10 to 30% error in the experimentally
measured losses; this error is usually accepted for this
type of measurement.
These limits are due to the actually known data which
do not allow us to account precisely for'the effects of
the residual stresses and for the machine cutting of the
iron sheets on the losses. Another reason which affects
the measured losses is the heterogeneity of the
magnetic field existing in the machine. In other words,
is it reasonable to simply add the electromagnetic losses
occuring in two neighbouring regions which are
subjected to two differents fields, or is there a coupling
effect which must be accounted for in measuring the
total electromagnetic losses? This is what we will work
on next.