3 ms·
Nice article! I like Gmsh a lot. I'm a big FE user, but don't have any experience with structural problems as it's not relevant to my field. I note that you say
by physicsguy 6y ago
Nice article! I like Gmsh a lot. I'm a big FE user, but don't have any experience with structural problems as it's not relevant to my field. I note that you say that tetrahedral elements aren't useful for structural problems, but why is this?
In my field, for e.g., generally people use Lagrange basis functions and it's cheaper (and sufficiently accurate) to just subdivide the mesh and use more 4 node tetrahedra than swap over to 10 node tetrahedra or a different element type. Is changing the basis function family not acceptable in structural mechanics?
- alibabaSX 6y agoThanks! The linear tetrahedron is known to lock in structural finite element. When performing mesh refinement, it turns out that the "convergence slope" is much worse with tet elements than with wedge or hex elements. To my understanding, these locking phenomenons are quite "field-dependent". And this is one of the main reason why new elements are often proposed is structural fe: they try to alleviate locking. About changing the basis function family, I guess it has been tried but it wasn't successful. In the end it always boils down at showcasing that your element passes mathematical tests (such as the patch test), and demonstrate that it is physically correct, even when the element is distorted. These are the subjects discussed by Bathe in his book "Finite element procedures".
- physicsguy 6y agoInteresting! Thanks for the info :)