Differences according to turbulent model


I leave the link for a CFD simulation that I have been doing. The pipe is similar to a tutorial, 1000 long, 10 mm diameter. I made two simulations of laminar flow (0.1 m/s), one with tetrahedral mesh with layer inflation and another with automatic hexahedral mesh. The pressure difference between outlet and inlet resulted 62.2 Pa in the first case and 37.5 Pa in the second. The difference by traditional method (factor f = 64 / Re) a is 29.8 Pa. Why did it differ so much just by changing the type of mesh and why the differences with traditional method ?. I also made four simulations of turbulent flow (0.4 m/s). Two simulations with hexahedral mesh already used and two with the tatrahedral mesh. For both meshes I used a simulation with k-epsilon model and another with k-omega SST model. The results for the pressure difference are: a) Tetrahedral mesh, k-epsilon: 688 Pa b) Hexaedral mesh, k-epsilon: 670 Pa c) Tetrahedral mesh, k-omega SST: 314 Pa d) Hexaedral mesh, k-omega SST: 386 Pa. My calculations with traditional method indicate 314 Pa. What is the reason for so much difference between the results? I do not know which model to trust.

Link to simulations:


Hi @cnordio!

The @PowerUsers_CFD and me will have a look at this issue and get back to you as soon as we know more.




Hi @cnordio!

Totally forgot to mention that you can have a look at a similar setup here: Turbulent Pipe Flow - Validation. Make sure that your parameters are adapted and set the initial conditions for the velocity as well as the gauge pressure to default (zero).

Let me know how things go!