[yaseen wsu]

dear all
first question
is there any difference if multiphase volume fraction coupled (there is five equation in cfx solver), but if this option is discleared (four equation in cfx solver, p with velocity in u v w), what is the idea behind them, because in Naiver stokes eq. there are velocity (u,v,w) and pressure with (sources)
second question
in most of the posts in the forum, the responser of the questions talked about sensitivity analysis for mesh, time step, please Iam a Msc student my time is limited is there is a pdf or link, on sensitivity analysis, till now I dont know S.A. for mesh and convergence.
thanks for your help
http://www.cfd-online.com/Forums/att...1&d=1459107553

[ghorrocks]

The volume fraction coupling option does not change the Navier Stokes equations. It just determines whether the volume fraction equations are solved in the same linear solver matrix as the momentum and pressure equations (making an even bigger matrix), or if they are solved separately (which does not require such large matrixes, but does make coupling the equations a bit harder in some cases).

[yaseen wsu]

thanks for your reply, can you answer the above second question about sensitivity analysis

[ghorrocks]

A basic sensitivity analysis is you adjust an input parameter (convergence tolerance, time step size, mesh size etc) and see its effect on an output which is important to you. That output could be pressure loss or drag or whatever is important in your case. You refine the input parameter until the output parameter has converged such that the changes in output is small enough that it is acceptable for your requirements.

But this is only a simple exercise. There are much more rigorous and effect (and faster) ways of doing it if you are prepared to do some research. The book "Computational Fluid Dynamics" by Roache is the seminal textbook on CFD accuracy. This link also has some more detailed discussion: http://journaltool.asme.org/Template...umAccuracy.pdf