The development and implementation of parallel finite element methods is a complex and error-prone task that can benefit significantly by simulating models of them first. However, such simulations are useful only if they can be confirmed to accurately predict the performance of the parallel system being modeled. The purpose of this contribution is to present a methodology to validate a Petri Nets-based modeling and simulation approach for parallel 3-D unstructured mesh refinement. To meet this goal, a parallel mesh refinement model is implemented based on a detailed software prototype and parallel system architecture parameters in order to simulate the functionality and runtime behavior of the algorithm. Estimates for key performance measures are derived from these simulations and are validated with benchmark problem computations obtained using the actual parallel system. The potential benefits of using this methodology for designing high performance parallel mesh refinement algorithms are illustrated.
展开▼