Inverse Material Determination for Small Porous Absorber Samples
In this contribution, we report progress in inverse material determination based on iteratively optimized FE simulations. Due to the consistency of simulation and measurement setup, this hybrid approach is in principle very well able to compensate for influences of in situ environments or edge effects caused by finite material samples. However, when using fitting algorithms which can lead to local minima, the choice or determination of initial estimates for the material parameters is of crucial importance to obtain satisfactory results. After presenting the complete workflow of the proposed method, focus is placed on the measurement of particularly small absorber samples. Starting from a strategy that utilizes an analytical sound field model to determine the initial estimate, the improvement of the results using a more sophisticated machine learning based approach is investigated. The design of the machine learning models trained specifically for this purpose is presented by Herrmann et al. in a separate contribution to this conference.