Accuracy of efficiency maps for permanent magnet synchronous machines depends on loss and torque calculations. Losses that occur in the stator or rotor core material as well as copper loss that occurs in the winding at high speeds are non-linear, and can be predicted by Finite Element Analysis (FEA) precisely. Though prediction is precise with FEA, the technique introduces substantial computational overhead. Replacing FEA with analytical estimates of the losses expedites loss modeling, but reduces precision of loss estimates, making efficiency calculations less precise as well. The prediction method proposed in this paper potentially offers accuracy similar to that attained using FEA while reducing the computational overhead typically required. The prediction method requires a small number of finite element solutions of the torque and the losses at different operation points which can then be utilized to accurately predict the value of the torque and losses in the machine’s operating range of interest. In this paper, the prediction method is compared to full FEA for estimating the efficiency of a permanent magnet synchronous machine. The prediction method is shown to closely predict the efficiency while requiring one third of the computational resources required by full FEA.
