I have to simulate flow (incompressible with heat trasfer) in packed bed moderator. Literature suggests to use porosity modeling to model the tube bundles. However, I am not sure how accurate and physically meaningful these modeling are.

Is there any scientific way to calculate the constants in these porosity models (either Darcy or Brinkman Model).

Is there any other approach, as grids inside the tube bundles (approx. 400 tubes) are ruled out, to arrive at some physicallly meaningful modeling of the flow.

I am not familiar with the kinds of flows you are dealing with, but I have a suggestion. If you simulate the flow with periodic boundary conditions, maybe you don't need 400 tubes in your model.

Thanx Vinay..

Problem Discription (sorry I should have done it earlier)

I have to simulate the flow inside a cylinderical bad where two jet are coming from the horizontal centerline circumferencially after heating with each other at top they fall down inside the tube array and an outlet is at the bottom. I have already simulated the flow without the tube bundles using a three dimension code.

I really don't think periodic boundary conditions will help in this problem. If you think it can be used, I would certainly like to give another thought to this direction.

In my opinion, if the flow is along the channels formed by tubes, you can use subchannel method which is widely used in thermal-hydraulic analysis of the coolant flow in the reactor core. Subchannel method is quasi-one dimensional, but it also consider the mixing between the channels.