[paul1117]

Hi, I recently ran into a problem while doing simulations with fluent. The physical background is that the hot water in the tube is sprayed out in the form of jet and then diluted by the surrounding ambient cold water. In the small model simulation (x=40m y=2m z=13.3cm), the simulation results are OK and the symmetry is perfect, but in the large model (x=4000m y=2000m z=10m), the simulation results are not good.
The number of grids in the large model is about 7 million, the reliazable k-e turbulence model is used, and the turbulence parameter Settings of the inlet are estimated by CFD calculator. in addition, fluent always has the warning "turbulent viscosity limited to viscosity ratio of 1.00000e+5 in XXX cells" during simulation, but small models will not have it. It is not known whether the final asymmetric result has anything to do with this warning.

As shown in the first figure, hot water is sprayed out evenly from all sides of the circle, and then diluted by the flowing cold water. The direction of the cold water flow is from left to right. As can be seen, the figure does not achieve a symmetrical effect, and the contours are not symmetrical.
As shown in the second figure,The hot water inlet is at the bottom of the figure, and it is sprayed vertically upward. The black part in the middle has a top cover, which allows the hot water to spread around instead of shooting vertically upward. The flaws in the grid shown in the figure are the display problems of tecplot, which can be ignored.
From the third figure, you can see the process of hot water spraying upward, blocked by the top cover, and finally spreading around and floating up.


To summarize a few questions:
1. Is the warning of "turbulent viscosity limited to viscosity ratio of 1.00000e+5 in XXX cells" in fluent during simulation related to the asymmetry of the final simulation result?
2. The simulation results of small models are symmetrical, while large models are not. Is it related to the size of the grid? The largest grid of a small model is probably at the centimeter level, while the largest grid of a large model is generally at the meter level (the largest is about ten meters).
In fact, the core appeal is to solve the problem of symmetry, it is best not to use symmetric boundary conditions, but to use the whole model.
I hope someone can help me answer my questions, I would appreciate it

[LoGaL]

If anything, it is the other way around: that turbulence warning is a sign that your simulation is not converged, and that is why it is not symmetric. Only the converged solution satisfies symmetry.

So you should fix your problems with the simulation not converging.

[paul1117]

Thank you, Mr. LoGal. So how can I get my results to converge? Do I have to keep counting? But in my experience, it doesn't always work.

[LoGaL]

Hard to say.

Maybe you have a bad mash, maybe you have to keep counting, maybe the symmetric case stabilizes and the full case does not. Start by keeping counting though.