Hello,
I am a beginner in CFD simulations as well as using Simscale. I am trying to simulate an incompressible flow (water) through a channel with both converging and diverging cross-sections. I have defined the boundary conditions as
- Inlet - Defined Pressure (Fixed) = 1bar gauge
- No slip at walls
- Outlet - Pressure (average) = 0 bar (gauge)
The z-direction is along the length of the channel.
Post processing the result, I find that some of the quantities don’t converge, i.e. Ux and Uy. Additionally, I see that from statistics of the inlet cross-section and outlet cross-section, there is a mismatch of the volumetric flow rates,
i.e. Inlet Flow rate > Outlet flow rate.
If someone has any pointers to things which I have missed or overlooked, that would be extremely helpful.
Link - SimScale
Thank you
Hello!
I had a look at the project and am posting some comments below:
- When using statistics, it’s a good idea to use cell data instead of node data.
- In simulations in general, since it’s an iterative process, you will never have a 100% mass/momentum/energy conservation in the post-processor. There will always be some residual (i.e. the residuals plot). The closer to convergence you are, the better conservation you will get
- Having a look at statistics on the inlet/outlet faces, you are getting 8.4e-6 m3/s at the outlet and 8.41e-6 m3/s at the inlet, which is pretty close (0.1%). Monitoring convergence with result controls helps.
Hi Richard,
Thanks for the inputs. Yes, you are right, the inlet and outlet flow rates match very well. I had probably made a mistake during selection of respective faces. I clicked on face 2 while face 1 was also selected, so the average volumetric flow rate somehow skewed up. Had to de-select statistics icon again before proceeding to select face 2.
But I am still wondering if the there is any issue with the convergence of Ux and Uy which seem to oscillate in value over the many simulation runs. Even their residuals seem high.
