Assessment the effect of various meshing in finite volume precision in order to numerical solution of seepage diffusion equation beneath the Roller Compacted Concrete dam

Presenting a suitable numerical solution for modeling seepage phenomenon as one of the destruction factors that have significant affect on dam’s stability, is necessary. Between several kinds of numerical methods, finite volume method has so many advantages for simulating different fluid’s flow problems with complicated conditions. Different kinds of meshes such as structured and unstructured mesh can be used in finite volume method that implementing suitable mesh has considerable affect on the results of finite volume method by modeling the complicated topographies actually and increasing the computations accuracy. Therefore appropriate choice of meshing is one of the most important factors in numerical solution. In this article, first finite volume method with different kinds of meshes such as rectangular, orthogonal triangular, non orthogonal triangular and voronoi is utilized for discretization seepage differential equation.Second,powerful programming software, MATLAB, is implemented for writing flexible program which is able to execute mathematical-numerical model. In order to evaluate the effects of various meshes in final result’s accuracy, seepage problem is modeled beneath several RCC dams by applying presented MATLAB code. Riemann boundary condition is used in this code in order to model the boundary conditions precisely. These cases are also simulated in PHASE2 7.0(2010) software for validation the model results. Then the results of the model and reference are compared by executing statistical software, SPSS17 and their compatibility proportion is evaluated. The results of comparison indicates that type of the meshing is affected P-VALUE and is varied this index more than 0.1.The results of model with finite volume and reference with finite element method had no statistical difference (P-VALUE>0.05).

Keywords:

seepage , finite volume method , various meshing , increasing accuracy , Riemann boundary condition