Postdoc in micromagnetic modeling of new magnetic composites

In this large research project we will investigate if we can realize a new heavy-rare-earth-free permanent magnet through a nanocomposite of magnetic materials. To model these nanocomposites, we are looking for a postdoc to further develop high performance computing numerical methods in our state-of-the-art open source micromagnetic model, MagTense. MagTense is based on a core implemented in the Fortran programming language, and it relies on the platform CUDA for parallelization of the computation over several GPUs’ cores, and has interfaces with Matlab and Python for ease of use. However, powerful as it is, MagTense is at present limited in its capability to model complicated magnetic systems for two reasons. First, all magnetic sources in a simulation interact, leading to computational resources scaling with the number of sources squared – which is inherently problematic numerically. Secondly, the differential equation used to evolve a magnetic system forward in time might be stiff, thus requiring special numerical techniques. You will research and realize numerical techniques that can be used to solve or alleviate the above problems and use your developed methods to model and determine ideal nano-composites for permanent magnets. In collaboration with Aarhus University, you will get to validate your modelling results against experiments. You will also get to supervise a PhD-student hired on a similar topic. Your techniques will be published in the open source MagTense framework, and will be used in powering future beyond-state-of-the-art simulations within magnetism.