I have written a CFD code which uses staggered grid and QUICK discretization scheme to solve an incompressible turbulent flow problem. But I have problem with far upwind nodes at boundaries. I want to implement second order boundary conditions. Can anyone introduce me some references or make some advices on this topic?
See pages 125-132 of book "An introduction to CFD, The finite volume method" by K.H. Versteeg.
Implementing second order boundary conditions needs two mirror nodes and depends on boundary condition type (i.e. Von neumann or Dirichlelt) and grid spacing (uniform or non-uniform).
In some problems, using second order boundary condition can lead to numerical instability (See book by HIRSCH) that may be conflicted with bad boundary condition implementation.
See pages 125-132 of book "An introduction to CFD, The finite volume method" by K.H. Versteeg.
Implementing second order boundary conditions needs two mirror nodes and depends on boundary condition type (i.e. Von neumann or Dirichlelt) and grid spacing (uniform or non-uniform).
In some problems, using second order boundary condition can lead to numerical instability (See book by HIRSCH) that may be conflicted with bad boundary condition implementation.
- Badges
- Science topic