This thesis deals with the data exchange on the interface between computational domains in the context of distributed computing where the grids on the interface are not matching. A framework for parallel non-matching grid mapping is developed and implemented as an appli- cation in the finite-element software Kratos Multi-Physics [1]. After an introduction to the topic, the motivation is stated and the relevant theory is reviewed. This includes the main algorithms that are developed and used, mainly spatial neighbor searching in distributed computing and mapping techniques for non-matching grids. Following is the establishment of design goals and requirements for the implementation, since mapping is a crucial part for a software dealing with multiphysics. Then an overview of the application is presented and the implementation of the introduced theory and algorithms is explained in detail. A special focus is given to a good integration in Kratos as well as an efficient implementation that works both for serial and parallel execution. Furthermore the application is designed to be suitable for a later extension in the form of other mapping techniques. This is achieved by decoupling the core of the application from the implementation of the mappers. Besides of the core of the application, a nearest neighbor mapper is implemented. The capabilities of the application are shown in the form of several testcases. These focus on different aspects such as the mapping technique itself or the performance in distributed computing. A conclusion of the thesis is then given as well as an outlook on further implementations and improvements of the presented application. The usage is explained in the appendix.     «

This thesis deals with the data exchange on the interface between computational domains in the context of distributed computing where the grids on the interface are not matching. A framework for parallel non-matching grid mapping is developed and implemented as an appli- cation in the finite-element software Kratos Multi-Physics [1]. After an introduction to the topic, the motivation is stated and the relevant theory is reviewed. This includes the main algorithms that are developed and used...     »