[zese]

Hi everyone

would you please help me about my outlet boundary condition at cfx software? I want to allocate zero gradient boundary condition at the outlet but I dont know how.
thanks in advance

[-Maxim-]

Hi,
you should find that in the manual/help easily.
Define Boundary Type, then go to the Boundary Details tab and select as Turbulence Option "Zero Gradient" in the drop-down menu.

[zese]

thanks Maxim but I am modeling a laminar swirl flow. I am not sure whether zero gradient condition is possible or not? second, for zero gradient condition should I consider my outlet as an opening condition?

[Opaque]

As the first response missed your expectations, it would be best if more information is provided

So far we know your flow is laminar; however, it is not clear if it is compressible, or incompressible, nor if you are modeling heat transfer or not. Single material, or multicomponents material, etc.

Just a reminder on boundary conditions: They are not independent of each other.

For example, flow in a pipe with mass flow inlet BC cannot be combined with Zero Gradient for velocity at the outlet. It is an ill-posed problem.

Hope the above helps,

[zese]

Originally Posted by Opaque

As the first response missed your expectations, it would be best if more information is provided

So far we know your flow is laminar; however, it is not clear if it is compressible, or incompressible, nor if you are modeling heat transfer or not. Single material, or multicomponents material, etc.

Just a reminder on boundary conditions: They are not independent of each other.

For example, flow in a pipe with mass flow inlet BC cannot be combined with Zero Gradient for velocity at the outlet. It is an ill-posed problem.

Hope the above helps,

Dear Friend
I am going to simulate a swirl laminar flow in a pipe. my flow is incompressible and at the inlet I have 16 blades to make swirl an my inlet boundary condition is based on velocity inlet. it is single phase without solving heat transfer equations.at the outlet, I want to consider zero gradient boundary condition but I dont know how.

thanks in advance

[Gert-Jan]

Could you please explain why you want zero gradient at the outlet? Because that is an uncommon boundary condition.

In fact you want flow through your outlet in strict normal direction. I remember from my first CFD classes in 1996 that the ancient software called Sepran allowed me to do this. However the results were very wierd. The pressure field required to obtain this prescribed normal outlflow was completely unrealistic. I can't imagine that you want something like this.

So, please explain in detail why you want this. It looks like a X-Y-problem to me.

[zese]

Originally Posted by Gert-Jan

Could you please explain why you want zero gradient at the outlet? Because that is an uncommon boundary condition.

In fact you want flow through your outlet in strict normal direction. I remember from my first CFD classes in 1996 that the ancient software called Sepran allowed me to do this. However the results were very wierd. The pressure field required to obtain this prescribed normal outlflow was completely unrealistic. I can't imagine that you want something like this.

So, please explain in detail why you want this. It looks like a X-Y-problem to me.

Dear Gert Jan

I have two reason to do that
First, I want to validate my study with an article written in 2001 that the author has set zero gradient condition at the outlet. In addition, Based on the experimental study the outlet boundary condition has not been represented in the paper. For example I dont have any idea about the pressure in the outlet. Once I tried with mass flow rate boundary condition with this idea that mass at the outlet and inlet is similar based on continuity equation but it didnt give good results. I wanted to try convective boundary condition but it was very difficult and studying manual didnt help me.

[Gert-Jan]

As far as I know it it not possible to enforce zero gradient at the outlet. You can set normal speed (=uniform). Maybe better: specify the Cartesian components. In this way you might get pretty far. Then you are really prescribing what is happening at the outlet.