Dimensional vs non-dimensional simulation

[nkul...@ncsu.edu]

Hello,

I have been trying to set up a dimensional solution and a non-dimensional and trying to compare their velocity profiles. At t = 3s, I get a decent boundary layer profile for the non-dimensional solution at Re = 10,000 (figure below).

But for the dimensional simulation I don't get the proper profile even at t = 40s. The BL region is very small even for Re = 1,000. I calculated the property of water at 25 C (this is an isothermal simulation) density = 997 kg/m3 and dynamic viscosity = 0.00089 kg/m-s. I followed the Tutorial/Fully Developed Flow example and checked that the same units are being used. Could I get some help on this please?

Thank you

Nilay

[stek...@gmail.com]

Hi, its hard to tell from a distance whats going on, I'd set up two cases, dimensional and non-dimensional, that should yield the "same" solution and re-check everything until it works. I put "same" in quotes because you have to be aware that nondimensionalization will include rescaling domain geometry and rescaling velocities etc., so the actual numerical values of all quantities will of course be different. For example, did you maybe forget to rescale the velocity/mass flow at the inlet boundary? For a closer look you'd have to send your case files.