3D Thermal Network Modeling for Axial-Flux Permanent Magnet Machines with Experimental Validation

A thermal resistance network model is presented for an axial flux permanent magnet (AFPM) motor. The model leverages similarity, typical to AFPMs, allowing use of a half slot, half tooth model, suitable for both single rotor, single stator (SRSS) and single rotor, dual stator (SRDS) AFPM machines. The presented approach leverages 3-dimensional (3D) thermal resistance networking, incorporating anisotropic material properties as well as analytical air gap and cold plate convection heat transfer coefficients for prediction of steady state thermal performance. This method offers significantly reduced computation time, when compared with conventional 3D finite element (FE) methods and eliminates the need for timely computation fluid dynamic (CFD) analysis. The proposed thermal network is configured such that the end user is able to ascertain thermal gradients and hot spots within the motor without computationally demanding 3D FEA and CFD analysis. Bulk thermal and temperature gradient predictions are first validated with a SRSS AFPM motor 3D FE model and lastly correlated with SRDS AFPM motor thermal test data, validating the presented modeling technique.