[amin144]

I want simulate flow between two plate, transient flow which starts to flow from rest according to constant pressure gradient.

I use Finite-Diffrence-Method and fractional step scheme with a staggered mesh.

which boundary condition I should use for inlet velocity?

[amin144]

Anybody know about "the Womersley boundary conditions" ?

[amin144]

I understood
Womersley is BC which use in blood simulating

Anybody can help me what should I do for a starting simulation when I know pressure gradient but does not know velocity inlet?

[truffaldino]

I would use pressure inlet at inlet and pressure outlet at outlet and you will get a flow driven by difference of pressures at the beginning and end of your "pipe" consisting of top and bottom plates. That what people usually do when considering flows in long pipes.

Could you say more about your problem: is it just a simulation of flow between two plates, or they are a part of a bigger system?

[amin144]

Thanks Dear truffaldino
my problem is that in fractional step on staggered mesh with Finite Difference Method, it need velocity for inlet!

Is there any method that need pressure for inlet?
Or am I doing any big mistake about problem?

[truffaldino]

answer removed

[truffaldino]

Hello again Amin,

It seems that in order to use pressure BC on inlet/outlet on staggered grids one has to add extra momentum equations on the inlet/outlet cells for velocity components tangential to the inlet/outlet planes. The normal velocities will be fixed by continuity equation.

[amin144]

Thanks very much
Could you introduce me some references to study more about boundary conditions and transient simulatation?

[truffaldino]

Originally Posted by amin144

Thanks very much
Could you introduce me some references to study more about boundary conditions and transient simulatation?

You can take a look at the Implementation of Boundary Conditions section of a good CFD textbook: for instance book by Ferziger and Petric "Computational Methods for fluid Dynamics". They discuss, for instance, implementation one-sided derivatives at boundaries that you need to define laplacians in extra momentum equations at the inlet/outlet boundary cells of your staggered grid (i was talking about in my previous post).

I am not specialist in this field, but It seems that finite-difference method on staggered grid is not very popular for the kind of the transient channel flow problem you are solving, but anyway it can be used.

[amin144]

thanks very much dear Truffaldino
I am thinking of changing my method too