Innovative accelerated meshless computational fluid dynamics solution engine

Abstract:

In this project, we propose to develop innovative mathematical models to solve computational fluid dynamics problems. The proposed formulations employ the penalty formulation within the context of the indirect integral equations in their meshless form (the method of fundamental solutions). In doing so, a multi-domain formulation is proposed to ensure the stability of the system of equations for huge problems. Moreover, it allows iterative solvers. In order to have a truly meshless method for the non-linear Navier-Stokes equations, techniques based on using particular solutions via the dual reciprocity method (DRM) are proposed. All relevant mathematical solutions will be derived in closed form. The solution engine will be developed first for 2D applications, then it will be extended to 3D ones. In order to employ such innovative mathematical techniques for practical climate problems, GPU and/or Multi-Core acceleration of the developed codes are proposed. Both the PI and the CO-PI (as an integrated team) have the theoretical background and previous experience in dealing with such problems.
This project will build a multi-disciplinary and cross-institutional research group as two teams are involved; Team 1 is from the Faculty of Engineering, and Team 2 is from the Faculty of Science. Both teams belong to Cairo University. The project aims also to manifest the data center (which was previously established by the PI via three previously funded projects by the STDF/STIFA) as a unit of scientific excellence.
The project budget is 1,996,820 EGP over 2 years.