Using a turbulent model when the flow is entirely laminar.
I may have made a mistake. I have a gas coming into a cylinder (2inches diameter 35cm long), at one end, via an 1/8th inch diameter pipe and exits via an 1/8 inch diameter pipe at the other end. I calculated the turbulence intensity for the inlet boundary using the empirical correlation for pipe flows. This gave a very low number (0.06%, Reynolds: 2130). i.e. it was laminar, unsurprisingly due to the diameter size of the pipe and the velociy (4m/s) the gas was going at. I'm used to dealing with turbulent flows, so I automatically switch on k-epsilon without thinking to get me started.
Anyway, I've went through mesh dependancy study etc and suddenly realised that the flow in the volume is entirely laminar (max cell reynolds number ~100). Since I've used the k-epsilon model and the flow is laminar, is the result rubbish? Should I have used k-omega or should I have just used laminar flow? The result seems reasonable, so I never questioned it initially.
I am facing a similar problem but the difference is that my results from turbulent & laminar model is completely opposite unlike yours. I just posted a thread as I am confounded with the results.
For a single system, as long as the conditions remain the same, whether we use a turbulent or laminar model, the results should be similar ; might vary a little of course ; but more or less similar.
|All times are GMT -4. The time now is 18:34.|