Bundle representation using Porous media
Hi,
I have tried to simulate a bundle of tubes by using the porous media option. First I did a simple 2d case of a periodic part of the bundle as represented in the handbook (tutorial 3). I did several simulation for a range of inlet velocities between low velocity and maximum velocity (compared with the real geometry). From the data I am able to plot the relation between the free inlet velocity and the pressure drop. Using the Ergun equation it is possible to determine c2 and alpha (by using a parabolic trendline). The order of magnitude seems correct for both parameters but after simulation in the final geometry c2 is under extimated. Do you have any previous experience in this field and could you advice me on the procedure to follow? Thank you for your cooperation. Hooky 
Re: Bundle representation using Porous media
The Ergun equation may not be a good/the best equation to use to represent a bundle of tubes  since the equation is developed assuming a packed bed of spheres. Maybe consider another equation for representing the solidfluid drag intereaction.
Greg 
Re: Bundle representation using Porous media
Wow that sounds pretty convoluted. Couldn't you just define an source term for the momentum equation which has different resistance depending upon the coordinate direction. You can do this quite easily in CFX5.

Re: Bundle representation using Porous media
Its easy in Fluent too  I suppose the point is what form this source term should take!

Re: Bundle representation using Porous media
There are empirical relations to determine the pressure drop in a bundle of tubes but it is not possible to separate the inertial resistance from the viscous resistance. It is relatively easy to calculate the inertial resistance for a bundle of tubes and in most of the case it is pretty accurate. However I would like to have an idea of the viscous resistance effect on the flow. Now, how could this coefficient be evaluated?
Thanks 
All times are GMT 4. The time now is 20:35. 