GenSDF: An MPI-Fortran based signed-distance-field generator for computational fluid dynamics applications

Abstract

This paper presents a highly efficient signed-distance field (SDF) generator designed specifically for computational fluid dynamics (CFD) workflows. Our approach integrates the Message Passing Interface (MPI) for parallel computing with the performance benefits of modern Fortran, enabling efficient and scalable signed distance field (SDF) computations for complex geometries. The algorithm focuses on localized distance calculations to minimize computational overhead, ensuring efficiency across multiple processors. An adjustable stencil width allows users to balance computational cost with the desired level of accuracy in the distance approximation. Additionally, GenSDF supports the widely used Wavefront OBJ format, utilizing its encoded outward normal information to achieve accurate boundary definitions. Performance benchmarks demonstrate the tool’s ability to handle large-scale 3D models ($\sim O\left(10^7\right)$ triangulation faces) and computational grid points $\sim O\left(10^9\right)$ with high fidelity and reduced computational demands. This makes it a practical and effective solution for CFD applications that require fast, reliable distance field computations while accommodating diverse geometric complexities.