Heat transfer between spherical particles and plane walls is of relevance for many industrial applications. Research done by TU Berlin has focused on the heat transport in a laminar regime. In these cases, heat conduction could be identified as the main heat transport mechanism (Brösigke, Herter, Rädle, & Repke, 2017). The question that then arises is to what extent turbulence influences the quantitative contributions of the different heat transport mechanisms. To clarify that question, further research has to be conducted on heat transfer between a spherical particle and a plane wall in turbulent regimes. For these cases, the applied method of direct numerical simulation (DNS) is computationally expensive. One possibility to reduce the computational cost is the use of Reynolds-averaged Navier-Stokes (RANS) equations. These equations contain the Reynolds stress tensor, which can be described by different turbulence models. ^Within the scope of this research, the k-epsilon turbulence model was chosen for its robustness. DNS and k-epsilon turbulence models produce qualitatively similar results. Nevertheless, quantitative differences of up to 50% can be reported. By optimizing the model parameters, a quantitative representation of DNS data could be attained. Therefore, the aim of this project is to develop an automatically operating interface, which compares the DNS and k-epsilon turbulence model results and numerically optimizes the model parameters in respect of the velocity field U, the turbulent kinetic energy k and the dissipation rate epsilon. In order to evaluate the physical legitimacy of parameter sets - as well as their application boundaries - the basic, theoretical principles of the k-epsilon model were compiled and discussed. The findings of this research were compared with sensitivity analyses. ^These were carried out to gain a practical understanding of the k-epsilon model with regard to the examined flow. By comparing theoretical and practical correlations, physically justified constraints for the parameter optimization were developed. Additionally, an appropriate optimization procedure was selected. Several modified optimizations have been executed to increase the efficiency of future optimizations. Finally, the improvement of the fields was checked by applying the optimized parameters on a similar flow. Besides model parameter optimization, the interface will be able to facilitate process parameter optimization for future applications.