Hi,

I have a simple shear flow with complicated interface conditions and am struggling to formulate far field boundary conditions for a staggered type grid. Applying du/dx=dv/dx=0 on the boundary I still have an unknown pressure term outside my domain. Any ideas?

Also at the boundaries should I apply the boundary conditions above to the equations (Stokes) or simply disretise them and use them instead of the normal equations?

Thanks

James

This is for incompressible flow"

For velocities just set u(N) = u(N-1) {along X) v(N) = v(N-1) (along X)

For pressure set pressure such that pressure at boundary to be zero P(N+1) = - P(N-1) Where N+1 is pressure point just ouside of domain and N-1 is point just inside domain (for staggered grid)

Thanks, thats great.

One final thing, would you substitute u(N)=u(N-1) into the discretised Stokes equations (so that dp/dx=d2u/dy2 etc) or would you simply use u(N)=u(N-1) instead of the bulk equations?

Also do you know of any good books/sites on FDM and in particular staggered grid arrangements as on the whole I have found the literature average.

Thanks again

James

At outflow boundary the zero-gradient condition for (u,v,p) may work well. But you need fix something of the inflow.