Tracer: Parametric Geometry Generation for Fusion Device Designs

The development of integrated modeling frameworks enables self-consistent analysis of the fusion device performance. However, a parametric geometry representation is needed to unlock automated design optimization. We have developed a parametric geometry tool – Tracer – to parametrically represent device components at medium and high-fidelity levels. Tracer is a python package build on top of Cubit python API. Cubit is a software developed at Sandia National Laboratory to enable generation of two-dimensional and three-dimensional geometries. The software also allows generation of robust meshes. The Tracer tool is connected to multiphysics framework Fusion REactor Design and Assessment (FREDA) enabling fast changes to the design considering multiphysics objective function and multi-fidelity model hierarchy. In contrast to existing parametric tools, the focus of Tracer is on multiphysics analysis with physicsappropriate meshing. Moreover, the geometry representation in Tracer is not specific to any device topology. Currently, any tokamak geometry can be represented within the software. We are working on extending the tool to stellarator designs. The parametric geometry representation is coupled to the plasma physics codes output via definition of the first wall. Therefore, changes in plasma configuration can easily be reflected in engineering analysis. The software generates the CAD file that contains geometry information together with the physics appropriate mesh. TRACER utilizes automated meshing capabilities of Cubit to generate meshes for each physics. Additional files are generated to allow easy integration with multiphysics software. These include material files, surface definition files to apply boundary conditions and any other files needed for the input generation streamlining. Integration with the FREDA framework will allow automated design optimization considering both plasma physics and fusion engineering metrics.