Computer-based methods such as shape and topology optimization offer a high potential to serve increasing light-weight ambitions within a significantly decreased product development time. To this end, results must comply with a wide range of challenging technical requirements, for which optimization methods are constantly being improved and extended. LEOPARD is a comprehensive topology optimization tool for the development of structural components especially in the automotive industry. The focus is on handling nonlinear simulations, ensuring high manufacturability and providing results in the form of modifiable Computer Aided Design (CAD) models with smooth surfaces. For the optimization of some vehicle components, the challenge arises that the load position is uncertain in reality. For instance, for the optimization of crash-loaded components, any possible impact position must be taken into account to ensure the required robustness in real-world crash events. Using common optimization methods, all of these impact load cases would have to be simulated, which makes practical execution impossible due to the very high computing time. In this talk, a method is presented that saves computation time on such optimizations by significantly reducing the amount of impact load cases that have to be simulated. The approach consists of only simulating a few impact points and, based on the results, approximate the results of the other impact points. This is achieved by using a special form of spatial interpolation, called displacement interpolation, introduced in the mathematical field of optimal transport. Unlike linear interpolation, displacement interpolation takes into account the underlying geometry, thus allowing for a reasonable approximation of stress distributions for intermediate impact positions that were not explicitly simulated. The method is tested using two and three-dimensional examples with a comparably high number of linear static as well as non-linear transient load cases. The modified topology optimizations yield promising results, showing remarkable resemblance to the references, where all the possible load cases were simulated. This allows optimizations considering a large number of possible impact positions to be carried out in a practicable manner. In addition, further possibilities resulting from the introduced approach for displacement interpolation are presented as an outlook, such as the morphing of different optimization results.