[max.kozak]

Hello all, I am attempting to run simulations in Fluent for a mesh that I generated in ICEM. The geometry is a pipe inlet into a vertical cylinder, which fills up and then exits out of another pipe. I have imported the initial coarse mesh and I have run a simulation. I now would like to perform a mesh independent test and I am having a few questions.

1.) I would like to do a global comparison test for my different mesh refinements where I look at multiple velocity points throughout the volume with each refinement. How do I view velocity at certain points in the volume, and continue to do so with each refinement?

2.) I would like to compute the boundary layer thickness. Is the best way to do this to make a slice plane and plot the velocity contour, determining at what position approx. 99% of the free stream velocity is reached?

3.) When I run my simulation, the residuals are high. The velocity residuals only decrease to about 1e-2, whereas the continuity residual stay close to 1. Could this be because the solution is initialized close to the actual solution value? It still seems to me the velocity residuals should further decrease, so I'm not sure how to determine convergence

I've attached my geometry in case that would be helpful. Thanks for any feedback on the questions I'm having.

[2phase]

Hi Max,

I hope I don't tell you something wrong, but from my point of view:

3) I guess you are wrong. if you were close, residuals should be very small. Have a closer look at what residuals are and try to understand. If you do not achieve convergence in steady flow you probably have to switch to transient.

1) as far as I know you can define these sample points in CFD-Post. Create a new location and look for the right settings (it's called sample data if you create lines or faces, I never actually tried with points)

Hope I could help.

Good luck, Martin

[MayTheFlowBeWithYou]

Are you looking at scaled or unscaled residuals