I have a 4 x 4 array of the geometry, put it in a "wind tunnel" in flow with ideal walls and run a bunch of points over the velocity range of interest. Take the data and plot a flow rate vs pressure drop curve for the geometry. You then put in the small sample geometry (where it asks for it) into flow as a pressure drop vs flow rate and then it will scale for whatever the actual geometry in the model is. For your app a tip would be to make the conical section thicker to conserve the mesh as it has such angles which will drive mesh count. Then adjust the medium thickness to the same value. If the pressure drop is expected to be high you will need more than a couple cells across the medium to capture high pressure drops. For fabric type mediums or sintered types your only option is data from manufacturer or do a real test. this works pretty well in my experience.
i have data from manufacturer regading the pressure loss see atatched,
based on that i calculated the alpha and beta values in porous equation.
i f i take isotropic material, i have a pressure drop which is same as pressure drop with out porous media,
if i choose unidirectional option, i am getting pressure drop values much greater than practical case,
i have doubt on which reference should be taken while selecting the direction
ploss2.gif 26.7 KB
Poruos media will not give you a good representation of your issue. From your model. the pressure drop will be very small. You could model this with a screen instead of a perforated body with very close results. You might have to modify the screen in the material data base for flow to be user defined to more closely model your perforations. In reality, the pressure drop through a strainer such as you designed is relatively insignificant comapred to the pressure drop when it is loaded with material.
I have used perforated plate on gas flows prior to entrance into a fabric mesh (porous media) on gas liquid separators. The perforated plate did not seem to create a computational botleneck, although it was only a 12 diameter domed plate. I would recommend that you just run the simulation on the perforated material that you have designed. The results you get in Flow will be more accurate than if you try to simulate your perforated pattern with a porous media, which it really is not.