Discussion Forum

Hello I am doing the 3D simulation of the flow of a gas (air) flowing through a pipe. The Reynolds number turns out to be 3600 depending on the inlet geometry and flow conditions. This implies a turbulent regime. In order to ensure the resolution of the wall, I tested several models of turbulence, in a long pipe, with less geometric details. From this I got the conclusion that the models which describe the wall better are the STT and k-epsilon low Reynolds model. I chose k-epsilon Low-Re as my turbulence model.

The total geometry is larger but I already want to be sure of the good resolution of this area. The problems appears especially when I try to ensure a circulation of the fluid inside the cylindrical geometry. Looking at the red line, you can get an idea of where I want the air to pass through. I want the air to pass within the cylinders, but all the domain is defined as air (except walls).

Elements are free tetrahedral.I am having problems creating boundary layers. Specifically the error is described by: Couldn’t shift boundary node at all. What I understand is that the mesh layers next to the internal walls could not be inserted. I already tried to refine the mesh as much as possible and even changing the turbulent model, but I get the same problem every time.

What recommendation could you give me so that I can bypass this problem? Do you thinks is possible to generate this kind of mesh??

Thanks in advance, Jhoan