Hi, as you can imagine this is someone who is struggling to model flow through a porous medium. There are two things that are really annoying about Star CD and this type of flow. The first thing is the permeability coefficients; why is it that while in all the literature in fluid mechanics the permeability is given in [L^2], whereas in Star CD, you have to: a) define it in terms of a pressure drop, and then b) find a linear relationship with two coefficients given in alfa [M L^-4] and beta [M L^-3 T]!. The other thing is the porosity factor, I dont know if you have notice it but, I measured the concentration of the species [moles/L^3] in a void space and then in a porous medium and they are both the same, even though there suppose to be a solid fraction. I would be greatful if someone out there could tell me how to convert from normal permeability units in to 'alfa and beta' terms, without having to consider a pressure drop? and also why is the porosity not working properly? Kind regards CM

Yeah you are absolutly right. I have asked myself several times why CD uses this variant for porosity definition. In Fluent for example it is very simple. You can define your permeability in [L^2]and in addition to that you can define a pressure jump. Thats it. So I would recommend to use FLUENT if you have it.

And I agree for some alpha beta values you can't reach a converged solution!!! e.g for alpha = 3 and beta = 80 and a flowspeed of 20-30 m/s convergence is impossible.

I work with STAR-CD for six and FLUENT for 2 years. And from day to day I have to ascertain that FLUENT is superior in so much cases especially the new 6.1 version.

It is an absolut mystery why CD make simple things so complicated. (not only the porosity!!)

I hope some CD developers have a look at this thread so that they realize that there solution is not ideal.

CMB, Wow. It seems you are trying to make a square peg fit in a round hole. Instead of worrying about the 'alfa and beta', have you tried specifing the porous diffusivity and schmidt numbers in the Additional Scalar Properties panel? (If you have very little pressure drop, then don't try to 'convert from normal permeability units'). Good luck, -R

10-Sep-2004 Dear Sir

Sub: Technical Support

Could you elaborate on how Star-cd solves porous media flow using Navier-Stokes Equation for solving x,y,z momentum for a 3D case as Darcy's law is the governing equation for fluid flow in porous media.

I guess the pressure gradient term in Darcy Law acts as a momentum sink term (negative momentum source term) in the Navier-Stokes equation.

In a nutshell, connect with an example

(Pressure Gradient) Navier-Stokes and (Pressure Gradient)Darcy Law

Please explain for

Case 1 : I have only porous media flow with an example Case 2 : I have a channel flow with porous media fitted at one end of the tube with an example

Please Reply

Thanks

Bharath

10-Sep-2004 Dear Sir

Sub: Technical Support

Could you elaborate on how Star-cd solves porous media flow using Navier-Stokes Equation for solving x,y,z momentum for a 3D case as Darcy's law is the governing equation for fluid flow in porous media.

I guess the pressure gradient term in Darcy Law acts as a momentum sink term (negative momentum source term) in the Navier-Stokes equation.

In a nutshell, connect with an example

(Pressure Gradient) Navier-Stokes and (Pressure Gradient)Darcy Law

Please explain for

Case 1 : I have only porous media flow with an example Case 2 : I have a channel flow with porous media fitted at one end of the tube with an example

Please Reply

Thanks

you are right the flow through the porous medium in Star is modelled with Darcy's law, just as it says in the manual it is -Ki ui = grad p, where K is the permeability, u the velocity and p the pressure, this would be your reduced momentum equation in the porous medium. Then you can modify it for polar or cartesian coordinates or do whatever you want with it. The problem is that, it is not very straightforward the way you define the permeability in Star, it is still a mystery for most of us.

Can someone resolve the 'mystery' ?