[ashtonJ]

Dear all,

Attached image shows a simple one-dimensional representation of a vascular tree.

I calculated a resistance (R = pressure drop / flow) for each outlet (E1 to E5) using Poiseuille pressure-flow relationship:

R=(8*mu*L)/(pi*r^4)

Now I want to apply these resistances (R1 to R5) as the outlet boundary conditions in CFX-Pre.

Does anybody know how I can do this.

Thanks,
AshtonJ

[ghorrocks]

Have you calculated the resistance from the inlet to each of the outlets? If so then haven't you defined the flow and you don't need to do a CFD simulation?

[ashtonJ]

Originally Posted by ghorrocks

Have you calculated the resistance from the inlet to each of the outlets? If so then haven't you defined the flow and you don't need to do a CFD simulation?

Thanks Glenn.

Yes, I have calculated the resistance from inlet to each branch. I don't have the pressure drop for each branch, so how I am supposed to find the flow?

What I am thinking is to do CFD using inlet pressure boundary condition and the calculated resistances as the outlet boundary conditions in order to determine the total pressure drop for each branch. Is this a correct approach?

[ghorrocks]

For CFD you do not calculate the resistance inside the domain. The solver does that for you.

But before you do anything you need to think what you are trying to understand. Do you want to know the way the flow splits up between the outlets? As you appear to know the resistance of your network then this can be calculated analytically. There is no need for CFD.

[ashtonJ]

Let me explain my problem a bit in detail:

There is an invasive technique, called fractional flow reserve, to determine the pressure drop at the outlets of a vascular tree.
Now I want to develop a CFD-based method to non-invasively find the pressure drop. For the CFD modelling, I only have the pressure at the inlet of the above vascular tree and don't have the pressure/velocity data at the outlets to use as the boundary condition. Therefore I need to find appropriate outlet boundary conditions. I was thinking to determine the resistance of each branch based on the length and the diameter of each branch and then use it as the outlet boundary condition. However, you say that this is not correct, so what would be your recommendation?

[ghorrocks]

You are misunderstanding the concept of a boundary condition. The outlet boundary condition applies the condition imposed by whatever is outside the domain to the domain. In other words, there is some fluid system beyond the boundary condition and you have cut this off in the simulation. So you replace it with a boundary condition such that this boundary condition behaves like the system you are not modelling.

So the conditions at the outlet boundary conditions are set by the flow conditions OUTSIDE the domain.

You have correctly done this by replacing the inlet with a pressure boundary condition. Now you have to determine what the outlets are.

[ashtonJ]

Thanks so much Glenn. fully understood.
For my problem, I create the vascular tree model from angiography images, normally due to the resolution of the images I am able to include only a limited number of branches in my model and have to cut off the rest.
I have no flow related or pressure information at the outlets and have no idea what outlet boundary condition I should use to model the whole system as realistic as possible!

[ghorrocks]

You have to know what happens beyond the outlet boundary or you cannot proceed.

Here is an example:

Imagine branch E1 is open to atmosphere just beyond the outlet boundary location (maybe a haemorrhage ), and branches E2 to E5 are all clamped closed just outside the outlet boundary. In this case obviously no flow goes out E2 to E5 and all the flow goes out E1. You would model this with a pressure boundary at atmospheric pressure at E1 and a no flow boundary (maybe a wall) at E2-E5.

This shows how the flow conditions outside the domain affect the flow inside the domain, and also how the flow outside the domain affects the choice of boundary condition.

[philip0513]

Originally Posted by ashtonJ

Thanks so much Glenn. fully understood.
For my problem, I create the vascular tree model from angiography images, normally due to the resolution of the images I am able to include only a limited number of branches in my model and have to cut off the rest.
I have no flow related or pressure information at the outlets and have no idea what outlet boundary condition I should use to model the whole system as realistic as possible!

Hi Ashkan,

Did you figure out your question?

Thanks,
Philip

[ghorrocks]

I quite clearly stated that Ashkan's model is not fully defined until he works out sensible exit boundary conditions. It is not possible for Ashkan to succesffully complete his model based on the information he provided so far.

If you would like us to help you please explain what you are trying to do.

[philip0513]

Originally Posted by ghorrocks

I quite clearly stated that Ashkan's model is not fully defined until he works out sensible exit boundary conditions. It is not possible for Ashkan to succesffully complete his model based on the information he provided so far.

If you would like us to help you please explain what you are trying to do.

Hi Ghorrocks,

I am facing a similar issue with single inlet and multiple outlets. I know the inlet and outlet pressures and there are downstream resistances (constant values). My current simulation is in steady state setup in Fluent; however, I am finding trouble to figure a way to apply those resistances to their corresponding outlets. I would really appreciate if you can provide some guidance.

Sincerely,

[ghorrocks]

This is the CFX forum. For help with Fluent try the fluent forum. I do not know how to apply resistance to an outlet in fluent.