Porting a parallel rotor wake simulation to GPGPU accelerators using OpenACC

Abstract

OpenACC allows offloading calculations to accelerators such as GPGPUs using short compiler directives. We will present our OpenACC port of the existing complex MPI-parallel simulation code Freewake, which allows hybrid parallelized calculations on multicore CPUS and GPGPU accelerators. Freewake is part of the rotor simulation code S4 of the Rotorcraft department of the Institute of Flight Systems (FT-HS). It simulates the 3d flow around a helicopter’s rotor under active control using a vortex lattice approach.