1

I would like to know how to compute the "best" cell center for pressure calculation in case of concave polyhedral cells, if it makes any sense. My face centers are defined as faces barycenters and I'm just wondering how to correctly determine the pressure-point.

Any hint?

nicoguaro
  • 8,500
  • 6
  • 23
  • 49
Tom
  • 465
  • 4
  • 14
  • Would there be any reason not to subdivide the concave polyhedra into multiple convex polyhedra? It's hard to say without seeing the details of your method, but you might be degrading the accuracy unnecessarily by using these oddly-shaped cells. – Tyler Olsen Apr 19 '16 at 22:39
  • In fact, those concave cells result from the deformation of hexahedral cells. More precisely, the hexahedral cells a initialy cut by some discontinuities which split those cells in many pieces, not necessarily concave or convex. In fact the question is more about what should be the criteria in that case to define the "best" (with respect to some measure) pressure point. – Tom Apr 20 '16 at 07:44
  • Can't you remesh your geometry after this "deformation" process? – nicoguaro Apr 21 '16 at 15:50
  • Unfortunately no, this geometry is the actual geometry of the cells, needed to honor some geometric constraints on the model (discontinuities mainly). – Tom Apr 26 '16 at 08:23
  • That does not answer why can or cannot be subdivided. Constraints on the model are given for a differential equation, not for a discrete representation of it. BTW, please use the name of the person to answer, e.g., @nicoguaro. So we notice your replies. – nicoguaro Apr 27 '16 at 21:42
  • @nicoguaro Well I'm not sure to undertand, but your model boundaries are actual constraints on the geometry of your cells, and those discontinuities define some internal interface on your computational domain. So it really constrain the splitting indeed. – Tom Apr 28 '16 at 13:34

0 Answers0