hi,

i'm basically trying to simulate the flow around a blunt object wihtin a pipe, but my object is porous. i've calculated for an input velocity speed such that it is laminar through the pipe (w/o the blunt object).

I'm not sure what kind of viscous model i should use. when i used k-epsilon, i had problems converging but when i used a laminar model it is converges easily

maybe these models are all wrong also. maybe someone can advise me.

also, my pipe length before and after my blunt object is only one object length,, i know it should be 3-4 lengths, but the mesh is very dense it does not make it possible. any advise on this?

thanks os much in advance

newbie

Why are you trying to use any turbulence model if you have already concluded that the flow is laminar?

What kind of Reynolds numbers are we talking about here?

Thanks for the replies!

my reynolds number is 140, yes i calculated it to be laminar, but that is w/o the porous blunt inside the tube.

so not sure of how i should approach

thanks

I don't know how to decide whether you should really go in for a Turbulence model. However, what do the experiments say? Is the flow inside OR around the porus media turbulent?

Also you may want to try some low-Re variants available in fluent. These can be activated through the TUI.

Porosity might not necessarily induce turbulence in the flow field. Also the magnitude of the induced turbulence,if generated might be small enough to be negligible for the current model.Try to use a non uniform mesh with more grid points near the object boundary. Also try to use an higher order scheme as the first order schemes introduce artificial diffusion.

-H

porous objects should give rise to even low Re flows due to small dimensions. turbulence if any will be created more at the entrance. It may be ignored however in view of already low Re at entrance. you will have dramatic pressure loss between entrance and the porous object, further pressure drop in pores/channels of the object and then increase in pressure at the downstream end of object. Hence, your pressure fluctuations need to be captured well especially in and around the object. A laminar model will do. an appropriate mesh will go a long way in deciding accuracy. try higher order schemes for greater accuracy.

Swarup.