Generation of Unstructured 3D Streamline Pressure-Potential-Based K-Orthogonal Grids

Laurent Souche. ( 2004 )
in: Proc. 9th European Conference on the Mathematics of Oil Recovery (ECMOR)

Abstract

An original technique for building optimal streamline-pressure-potential (SPP) unstructured grids is presented. This method entails the generation of Voronoi Diagrams of n-dimensional objects for partitioning a given volume of interest (basically a reservoir in the geological sense) into a set of 3D streamtubes. Unlike other methods, this process is fully 3D and does not involve any extrusion. Moreover, no restrictions are imposed on the complexity of the flow pattern and all the generated grid cells are such that their faces are either locally perpendicular or collinear to the flow direction, which has been proven to be profitable when using Finite Volume discretization schemes [1]. The originality of this method lies in the fact that the polyhedral grid is represented by a 3D raster image. This raster image can either be used directly for feeding the flow smulator or converted first to a combinatorial representation through a vectorization process.

Download / Links

    BibTeX Reference

    @INPROCEEDINGS{,
        author = { Souche, Laurent },
         title = { Generation of Unstructured 3D Streamline Pressure-Potential-Based K-Orthogonal Grids },
     booktitle = { Proc. 9th European Conference on the Mathematics of Oil Recovery (ECMOR) },
          year = { 2004 },
      abstract = { An original technique for building optimal streamline-pressure-potential (SPP) unstructured grids is presented. This
    method entails the generation of Voronoi Diagrams of n-dimensional objects for partitioning a given volume of interest
    (basically a reservoir in the geological sense) into a set of 3D streamtubes. Unlike other methods, this process is fully 3D
    and does not involve any extrusion. Moreover, no restrictions are imposed on the complexity of the flow pattern and all
    the generated grid cells are such that their faces are either locally perpendicular or collinear to the flow direction, which
    has been proven to be profitable when using Finite Volume discretization schemes [1]. The originality of this method lies
    in the fact that the polyhedral grid is represented by a 3D raster image. This raster image can either be used directly for
    feeding the flow smulator or converted first to a combinatorial representation through a vectorization process. }
    }