I reviewed some literature amd found that second order tetra has good enough accuracy to compete with hex mesh. Is that correct?

KV

I have some complex geometry that I am trying to hexa mesh. But, I have very little time. Neither do I have any training or practice for gaining expertise in hex mesh. But, I read in some literature that second order tetra has good enough accuracy to compete with hex mesh. Is that correct?

Adriano_Koga

hex mesh provides good accuracy with smaller node count (hence smaller stiffness matrix), as long as the mesh if refined enough where it really needs to be refined. But Tet10 is accurate enough, and specially in complicated geometries would be prefered. The drawback is the computational cost, so in order to get a good compromise, you should work well with local refinements, to reduce the nodal count overall, as it could increase a lot memory consumption.

So, besides the element type itself, you should consider what is the available hardware, and areas of interest, and the geometry complexity. By the way, for Tetra dominant mesh, SimLab and Inspire could deliver a quite nice result.