4. Conclusion

Modeling and experimentation are inseparable when it comes to validating any thermal model, whatever the numerical method used. The experimental phase provides experimental temperatures, boundary conditions and internal heat sources. Simulation allows us to obtain very specific conductivities (windings, fluids, contact thermal resistances) and, ultimately, points that cannot be reached experimentally. It is also preferable to carry out these phases at different rotation speeds (different temperatures) in order to initially set these parameters and then validate them formally.

It goes without saying that such a model is valid within a restricted operating range. If it is possible to interpolate the results (with a degree of caution), extrapolations become more random (major changes in fluid thermophysical parameters).

The future of numerical...