Geometry and Meshing for Explicit Dynamics
Preparing the geometry and mesh for an efficient and successful explicit dynamics simulation often requires more effort compared to when using implicit solution methods. To a large part this comes from the conditional stability of the time stepping method. Also, the fact that it is preferred to use linear shape functions for the displacements adds requirements on the mesh to obtain an accurate solution.
The Solid Mechanics, Explicit Dynamics and Truss, Explicit Dynamics interfaces only support linear shape functions.
Geometry
The requirements on the geometry is a consequence of the subsequent requirements on the mesh to control the stable time step. In order to facilitate a good quality mesh for explicit dynamics consider the following:
Make use the Geometry Cleanup tool, Virtual Operations and other functionality available in the geometry sequence to simplify the geometry.
Meshing
The accuracy and cost of an explicit dynamics simulation is ultimately determined by the quality of the mesh. For example a few small or poor quality mesh elements will dictate the time step taken by the solver; thereby significantly increasing the solution time with little to no gain in accuracy.
Adhering to the following guidelines is good practice that will increase the quality of the simulation results:
-
-
-
-
-
-
The above recommendations on mesh element types follow from the use of linear shape functions for the displacements. It is well-known that linear simplex elements have poor accuracy and are prone locking, and therefore should be used sparingly. Also, hexahedra and quadrilaterals exhibit both shear and volumetric locking. However, their accuracy is much improved by using reduced integration. While better then simplex elements, the use prism elements in 3D should also be minimized. Even with reduced integration they exhibit locking and can thus lead to an overly stiff model response.
See Using Reduced Integration for more information.