[JPBodner]

Hi all

I am working on a model of an internal flow with a low Re (~1400), but there are some geometric features where local turbulence could be created. This includes some twists and turns as well as an internal jet (or large-expansion). I am simulating water flow through the passages of a small device My main interest is predicting pressure drops through the system.

I presume that it's best to solve with turbulence even though the Re is below 2000, because there are some regions of swirling-like behavior. (If you disagree and would just go laminar, let me know) But I am not sure what type of turbulence model is best for this type of flow. Low-Re k-e? k-w?

Any suggestions given the information I've provided?

Thanks in advance.

[andy_]

Traditional turbulence models generally assume a high Reynolds number turbulent flow in their formulation which you do not have. The low Reynolds number versions of high Reynolds turbulence models tend to be more of a curve fitting exercises rather than introducing more physics based terms into the equations. They are normally an improvement for the types of flows they have been setup to address but not for general cases and transition.

For accuracy you should run an unsteady 3D laminar solution and resolve what is causing the possible transition to turbulence. Do you have the computing resources to do this?