Current numerical computational methods used to perform warhead design and analysis are commonly based on the eulerian finite difference method. The eulerian method is used to allow the material to be advected through a mesh, and the finite difference method is used to provide a natural way to explicitly account for the material flux from a donor cell into neighboring receiver cells through the convenience of ordered mesh indices (i, j, k). The eulerian finite difference method is adequate, however, in its ability to treat complicated mesh geometries, irregular boundaries, and variations in both mesh size and cell shape. As a result, the eulerian finite difference method can not be economically used for solving general three-dimensional problems. The lagrangian finite element method, on the other hand, is very adaptive at handling complicated geometries while reducing the number of elements without compromising the numerical accuracy. The lagrangian finite element method is, however, incompetent in solving problems with severe material/mesh deformation. The rezoning and dezoning procedures can be effectively used to overcome the severe mesh deformation problem occurring in the lagrangian finite element method if the mesh discretization only contains a small number of regions with severe mesh deformation. The lagrangian finite element method with rezone and dezone capabilities appears to have the advantage over the eulerian finite difference method for solving impact dynamics provided the procedures to perform the rezone and dezone are conservative. In phase i we will develop a conservative rezoning procedure to replace the non-conservative rezoning procedure currently used in dyna2d.
Keywords: FINITE ELEMENT FINITE DIFFEREN REZONING CONSERVATIVE RE EULERIAN LAGRANGIAN WARFARE NUMERICAL COMPU
