I am a 28-year old M.Sc. (tech), currently chasing my doctoral degree on electrical engineering in the Aalto University School of Electrical Engineering, Finland. My research mainly deals with resistive losses in the windings of random-wound electrical machines, but I’m very skilled in machine analysis in general. Below, you can find a brief description of my main areas of expertise.

Efficient Finite Element Analysis

I have a strong experience in writing computationally efficient finite element (FEM, also FEA) programs for electromagnetic analysis of electrical machines. I have especially focused on analysing eddy-current and circulating current losses in stranded windings, and I’m currently working on efficient time-domain analysis. Click here for more information.

Flux density plot of a 15 kW induction motor. Simulated in Matlab, with a library of my own making.

Reluctance Networks

I have programmed a library for solving nonlinear reluctance networks of arbitrary topology. Both permanent magnet and current sources can be modelled in both 2 and 3 dimensions. Time-stepping analysis is supported and encouraged, and changes in the network due to e.g. motion are handled in an efficient fashion.

Furthermore, I have developed a fully parametriced reluctance network model for a novel axial flux permanent magnet machine. Click here for more information. I am certain I could make a similar model for a more conventional machine topology quite fast indeed.

Reluctance networks can be quite accurate.
The phase flux linkages of an axial flux permanent magnet machine, calculated with my reluctance network and a commercial FEM software.

Uncertainty Quantification

Thanks to my work on random-wound machines, I am familiar with the basic stochastic finite element (sFEM, sFEA) approaches. This includes basic concepts such as polynomial and wavelet chaos, and surrogate models. Additionally, I have also tested some more advanced approaches such as the Proper Generalized Decomposition (PDG), its MiniMax extension, and some tensor-based approaches for high-dimensional problems.

I have also developed a sampling algorithm to model the uncertain winding process of stranded windings.