Riassunto analitico
As the trends in the automotive sector are changing and electric cars are
gaining more and more popularity, the thermal management of their motors is becoming one of the most critical factors to achieve better performance and higher efficiency.
This work consists in the development, in the SymSpace environment,
of a high-order Lumped Parameter Thermal Network (LPTN) capable of
simulating the thermal behaviour of Wound-Field Synchronous Machines,
in both steady-state and transient conditions. In addition it includes the
implementation of a possible novel cooling technique, consisting in a form
of direct stator cooling by means of channels cut in the teeth.
The aforementioned LPTN is tested and validated by comparing its results with finite element simulations performed with data coming form a
commercial wound-field machine and then is used to simulate the temperature evolution in such machine during a WLTP Class 3 drive cycle.
Finally the geometry and position of the stator cooling channels are
optimized, with the Optimizer software available in the SymSpace package, taking into account the trade-off between the thermal performance
and the negative effect, produced by the presence of the channels, on the
electromagnetic performance.
Moreover, once the optimal parameters are found, the approximate
power of a water pump, needed to push the coolant through the channels,
cooling circuit and radiator, is calculated.
The results show that the LPTN is capable of predicting the most important machine temperatures with a maximum error of around 3% with
respect to the finite element simulation.
The machine equipped with optimized cooling channels is able to achieve
stator maximum temperature around 36% lower and rotor maximum temperature around 8% lower, with respect to the same machine cooled by an
housing water jacket.
The negative effect on electromagnetic performance is compensated by
the reduction of the copper losses due to lower coil temperatures.
|
