search

abarret / clotting_model

An IBAMR implementation of Fogelson and Guy, 2008
2 stars 0 forks source link

Use second order elements for FSI #26

Closed abarret closed 2 years ago

abarret commented 2 years ago

This replaces the first order elements with second order elements for FSI. This prevents locking, in which correct deformations can not be represented by the finite element basis. Note that use of this now requires a smaller time step to tether the structure in place.

jmcragun commented 2 years ago

Would you be willing to explain the context/concepts of what's going on here? My experience with FEM is only tangential. I suppose we could also discuss it in the meeting tomorrow.

abarret commented 2 years ago

This is the LBB (or inf-sup) condition. The description can be pretty technical, but it has to do with the incompressibility condition. We're using a finite element basis to describe the deformations of the structure, but the structure is incompressible (which means the divergence of the structure velocity is zero). We need to make sure there are enough degrees of freedom to account for both the deformations and incompressibility.