I’ve been trying to simulate water flow through a simple pipe system of varying height. From what I can tell based on the mesh parameters I’ve used, the mesh should be properly made, but it seems some of my convergence plots remain unstable, especially the domain and outlet plots. Is this still an issue with the mesh, or am I missing some other parameter?
2 of the simulations:
Hi, this is Fillia from SimScale Support. Have you calculated the Reynolds number of the case?
Would this be to calculate the entrance length? Since the simulation involved setting a pressure differential instead of manually setting the velocity, would using the velocity from the simulations I already have be accurate enough?
The simulations tended to give a maximum velocity of around 9.9e-5 m/s near the inlet. I used the method below for calculating reynolds number and then used the equation 0.06 x Reynolds number x hydraulic diameter for the entrance length. For the first simulation I had listed, I got 5.3 e-8 m for the entrance length, and the actual pipe I used has a length of 1e-3 m, so at the time I assumed flow to be fully developed.
Hi there, I see that you created more runs, with different sets of boundary conditions. In general, it is suggested to use a Velocity Inlet and a Pressure outlet, like you are currently doing 
Try using the ‘potential flow initialization’ activated too. You can find this listed on the Simulation Control tab.
I was able to improve the convergence to a point where most of the fluctuations seem to have stopped. However, some of the other values don’t seem to be converging, but are steadily increasing or decreasing. I’m not familiar with this type of deviation, is this still an issue with the meshing?
Hi, I hope you had a great weekend! Run 7 seems to work like a charm! Oscillations at such low residuals can be negligible. However, a quick calculation I previously did gave me a really low Reynolds number, so I assume the flow can be laminar. Have a look here: Reynolds number - Wikipedia for flow in a pipe (with a square cross-section). If that is not the case, make sure to add an inflate boundary layer on the walls. The standard mesher provides an automatic BL addition.
I’m trying to model fluid flow through a simple pipe, and it seems I was able to refine the mesh to the point where most of the variables converge, except for those in the residual, which seem to remain unstable. Is there a specific feature I should be refining in order to solve this problem?
Hi again
I see 20 runs in total inside the project, and that you managed to create many runs with this oscillation afterwards. What did you change?
Best regards,
Fillia
Things I tried changing include adding region refinements, surface refinements and feature refinements to the mesh, changing the inlet from a pressure inlet to a velocity inlet (having a pressure inlet seemed to mess with other convergences outside of residuals), and changing the velocity inlet between from fixed value to flow rate, and changing the specific value of the flow rate. With the volumetric flow rate, I can’t seem to get rid of the fluctuations in the residual convergence graphs.
I also tried changing the outlet from a fixed pressure to a mean pressure most recently, but that seemed to make the fluctuations worse.
Great! thanks for sharing.
