Implementation of Lagrangian Surface Tracking for High Performance Computing

In almost all technically relevant combustion applications, flames occur in turbulent flows. The interaction of turbulent flows with flames is still not fully understood due to the large range of time and length scales which govern combustion processes. One method of studying this interaction is by tracking thermo-physical trajectories of material points on flame surfaces. These trajectories give insight into the local flame dynamics and help to understand the influence of turbulence on flame properties. In this work, a Lagrangian tracking algorithm is presented which performs the tracking of material points on iso-surfaces. Because this tracking method is used in large-scale direct numerical simulations of combustion processes, the focus of the implementation lies on performance. By tracking the position of the Lagrangian particles in barycentric coordinates, efficient algorithms for spatial interpolation and the intersection of particle trajectories with iso-surfaces can be utilized. The code is written in a general way and not restricted to reacting flows but can be used to track any iso-surface. Additionally, the algorithm works by decomposing the computational cells into tetrahedra. This allows the tracking method to work on unstructured meshes with arbitrary cell shapes. The tracking method is implemented in OpenFOAM and applied to the direct numerical simulation of a 3D turbulent flame. The simulations are conducted with a custom solver which makes use of automatically generated, highly optimized code for the computation of chemical reaction rates, which performs up to 20 times faster than OpenFOAM’s implementation. For the turbulent flame, applying the tracking method increases simulation times by less than 5 %, so that the current implementation is well suited to be used during large scale simulations.