An Integrated Computational Environment for CFD

The aim of this project is to develop an integrated, software package for applications in the field of Computational Fluid Dynamics (CFD). The package, called CFDice (CFD Integrated Computational Environment), will integrate three stand-alone packages, all developed at the department : a mesh generator (IGG), 2 three-dimensional (3D) Navier-Stokes solver (EURANUS) and a visualisation package (CFView). The integration will be based on object-oriented programming techniques and on the use of a database, shared by the three applications. The resulting CFDice package will be one of the few if not the only package able to 1) support complete and complex CFD projects (from the geometry definition, over the simulation, up to the analysis of results) and 2) based on state-of-the-art numerical schemes and programming techniques. CFDice will significantly decrease the cycle time for the simulation of flow problems and, through the database, will facilitate storage and retrieval of data and results.This project concerns the prolongation of an existing project, aiming at the development of an integrated, software package for applications in the field of Computational Fluid Dynamics (CFD). In this extension 1) the unstructured approach with adaptation, developed in the original project, is further refined and optimazed and 2) the code is generalized for applications in the field of fluid-structure interactions, by extending it for time-accurate simulations on moving meshes.The time-accurate simulations will be absed on an implicit formulation, solved in pseudo-time with explicit schemes that are accelerated with multigrid, residual smoothing and local time stepping. A novel and efficient approach is proposed to generate the sequence of coarse grids for the multigrid and to transfer information between grids. Mesh movements will be accounted for by rebuilding meshes from very coarse, a new strategy, which is now already used as alternative for derefinement. This will ensure a good quality mesh, even for important movements. These complex applications require efficient sensors that ensure an adequate capturing of all important flow phenomena but avoid excessive refinement. Part of the research therefore focuses on the development of such sensors.