Nodal Based Mappings Between Finite Element Meshes
Due to the demands on mathematical models that simulate physical properties and behaviors of automotive components, different finite element meshes often partition the same component for each disparate simulation. Commercial packages that transfer mesh dependent results from one finite element analysis to initialize another typically use a variation of a specific nodal based mapping. The simplicity and ease of implementation make this mapping very attractive. In its simplest form, the mapping first assigns a node in the analyzed model the average value of the elements that share the node. A node in the unanalyzed mesh receives its value from the closest node in the analyzed mesh. Nodal based interpolants on the new element initialize elemental values. When an element in the unanalyzed mesh requires a constant value such as its thickness, the element receives the average value at the nodes that define the element. This brief paper documents some undesirable characteristics of this nodal based transfer of material properties from one mesh to another.