In the aviation industry, passenger comfort is one of the highest priorities. The motivation for this project is to reduce the noise the passengers perceive (e.g. from the turbine). As an acoustic quantity, the Sound Transmission Loss (STL) of polymer foams is investigated in this work. The approach is to predict the STL response of foams based on their micro-structure. To do so, a tool is developed which generates arbitrary geometries. Those geometries are considered to be the periodically arranged unit cells of a foam. Using the Multi-scale Asymptotic Method (MAM), the parameters for the acoustic Johnson-Champoux-Allard-Lafarge (JCAL) material model are computed. As a final step, the poro-elastic Biot-Johnson material model implemented in ABAQUS (since version 2019) is used to predict the acoustic response based on the cell structure. This poro-elastic material model does not only consider the path along which the air is flowing through the structure of the foam but also accounts for the interaction between the air and the polymer frame. Furthermore, this material model is a homogenized material model which allows to model the acoustic response of complex components efficiently without modelling the micro-structural details for the finite element simulation. As the goal is to provide an ideal micro-structure to improve the STL response of polymer foams, the effects of one step to the next one is investigated. Firstly, the effect of geometric features onto the JCAL parameters is studied. These parameters then define the STL response, so the knowledge of the effects of each parameter is of interest. Finding an improved micro-structure is then reversed engineering. Knowing the effect on the STL, a set of target JCAL parameters can be defined. This set of parameters is then to be reached by altering the micro-structure leading to an optimized micro-structure for a polymer foam in terms of the acoustic response.