Simulation and adjoint-based identification of a room acoustical benchmark case

Acoustic simulations are widely used in room acoustics. Ray-based methods are often preferred due to their low computational cost, although they have limitations in accuracy. Wave-based methods, such as finite differences in the time domain (FDTD), offer better accuracy but face difficulties in modeling complex acoustic boundaries. Since the acoustic boundary conditions are a crucial part of numerical simulations, it is a widely studied topic and several models are available. A promising solution to these challenges is volume penalization, which integrates the parameters flow resistivity and porosity into the governing equations. However, these parameters are typically obtained from tables with inherent uncertainty ranges, limiting the ability of the simulation to produce fully reliable results. In this study, the volume penalization method is applied to simulate the first case from the Benchmark for Room Acoustical Simulations (BRAS) with a Rockfon absorber. The boundary parameters are optimized during the simulation using an adjoint approach and the simulation is aligned with reference data from high-fidelity simuations and experimental measurements to achieve a more reliable acoustic simulation.