Hello Karyan,
First off, welcome to SimScale and our forums!
Your problem setup for a pressure drop analysis very similar to what many SimScale users try to achieve.
I took a look at the v3 Project as it follows the best practice to define both a reference pressure at 1 side of a flow channel + a flow rate at another side of the flow channel. Obtaining a CFD is a lot more stable this way rather than defining the flow rate only on the inlet/outlet sides (as you’ve done with the v2 project).
Looking at the convergence residuals of your solution, it seems that the initial simulation with 1,000 iterations was not yet converged to a solution. I’ve attached the residual convergence plot from your results below:
Further, it is useful to see the convergence of a steady-state solution for the quantity of interest; in this case, the pressure difference between the inlets and the outlet. You can use Result Control objects to track the field variable averages on these inlet/outlet surfaces using the Surface Data > Area Averages object.
I’ve re-run your setup with this result control added, but you can take a look at this result control page to learn how to do it yourself. The pressure on the inlet faces are fixed as you, the user, have defined them within the boundary condition setup. However, you can see the lack of a convergence solution for the outlet face (green):
There are a few steps that can be taken to resolve this convergence issue:
- Continue the run with more iterations: Since the pressure on the outlet face is oscillating but progressing towards convergence, you can continue the run for more iterations to see if it progresses to a steady-state solution. Since the 2 inlet ducts converge to the larger diameter duct, this heavily turbulent region may simply just need more iterations to converge to a steady-state solution. Linked here is an article explaining how to continue a run for more iterations.
- Refine your mesh: The current mesh size can be further refined, especially on the boundary layers, to ensure you properly capture the viscous effects on the pipe surface where the pressure friction loss occurs. This will require more computational time, but can improve the validity of your results.
- Use the Subsonic Analysis Type (Pro Account only): The Subsonic analysis type uses a proprietary solver with a robust meshing algorithm that is well suited for internal flow problems, especially at predicted pressure loss. This is the analysis type that works very well “out-of-the-box” for these type of flow control problems and is often recommended to our Pro users.
Continuing your current V3 project setup run to 3,000 iterations provided much more stable results, predicting a pressure drop in line with what is expected using fundamental air-duct pressure drop calculators; of course, this calculator is only really valid for straight line pipe/duct flows, not for cases where there is more turbulence (therefore more pressure losses) due to the flow from 2 channels joining together into one channel.
I’ve presented the updated plots below:
Pressure plot:
Convergence Residual Plot:
Let me know if this helps!
Cheers,
Omar
