[the_end90]

Hi all,
when simulating my steady state flow bench I have Problems in convergence: at the first test case, with Maximum valve lift, the convergence whas not so good but acceptable, but if I Change the mesh to simulate a reduced valve lift the calculation diverges to Overflow.
I am using a Coupled flow with high y+ Standard k-Epsilon model stedy.
Can I Change something to improve?

[lcarasik]

1) Are you doing flow at high mach numbers?
2) Why are you using k-epsilon?
3) What are more details of your step up? You didn't provide a lot of details to go off of.
4) What does diverges to overflow mean?

[the_end90]

Originally Posted by lcarasik

1) Are you doing flow at high mach numbers?
2) Why are you using k-epsilon?
3) What are more details of your step up? You didn't provide a lot of details to go off of.
4) What does diverges to overflow mean?

Hello Icarsik,
my best apologies for giving too few Infos.
No the Mach number is pretty low to 0.5 Maximum, only some cells are at Mach 1 but they are barely visible
I am doing the Thesis and so the idea is to start with a Basic phisic model and then move to something better, for you what is the most useful physics?
TO be more detailed I use a classic polyhedral mesher with Extrusion layer on the inlet and outlet that are a Stagnation inlet and pressure outlet, I also use symmetry planes on the simmetry sections, I imported everything and before meshing I took care to take to Zero the Errors in the surface repair section
Now I struggled to make it able to not go to Overflow, but I do not have a convergence even after 10000 iterations. I have big oscillation in the residuals.
If I ignore residuals and build up a new case, that is if I move up the valve of 2 mm from the before Quote and remesh and restart without Clearing the solution, the mass flow starts to drop until it goes to stack and says "invalid Operation" on the solver. the Problem is physically that mass flow drop that is not correct at all because the movement of the valve should not make the mass flow rate to drop much more than some percent in real testing

[lcarasik]

Is your flow compressible? Why are you using the coupled solver?
I agree with your approach to start with the basic physics first but I want to make sure you didn't accidentally select the wrong options. Star-CCM+ is really bad about explaining what each choice means.
Are you sure you need to be using a stagnation inlet?
What does overflow mean?
Can you post a picture of your residuals?

Can you explain what you are actually trying to simulate?

[the_end90]

Actually after some trials I switched to Segregated flow solver but the situation did not change a bit as You can see from images I have the residuals that moves up and down without stability. The stagnation inlet is always used when simulating a steady flow bench as the boundary conditions are given in terms of absolute pressure.
I have uploaded a photo of the mesh and their values and a doc with the mesh values used, also I attach a photo of the situation when it goes to Overflow when solving.
Practically what I do is the following:
In repair parts I offset the valve up of 2mm and I remesh everything, but when I do make the simulation restart it actually starts with a very high bump of the residuals from 1e-3 up to 1e2 then it continues to have higher and higher residuals making always more and more temperature corrections, from 3 cells in the first iteration than up to 30000cells

[lcarasik]

1. I strongly recommend you do not use the coupled solver unless you are using highly compressible flows, compressible flows experiencing mach numbers higher than 1.3, and natural circulation flows.
2. I am not entirely convinced the stagnation flow inlet is correct. What was the information given related to your boundary conditions?
3. What is overflow?
4. Your mesh seems very coarse. If you can do it I'd recommend using around 250k cells.
5. How are you defining convergence in your study? Do you have any additional variable you are measuring throughout the simulation to see when they reach an asymptote?
6. Your behavior might be unsteady you might not be able to run a steady state solution to get your values.

[the_end90]

Sorry to answer only now,
1. If you see from the Images above and the Infos given in the last message I am now using Segregated solver but the solution had not improved
2. I have Information on total pressure at the inlet and at the outlet, but I have already a full chamber Simulation running, without using simmetry planes and it does not give any Problem.
3. Overflow is when the residuals goes up to 1e8 and more and the calculation stops
4. I tried to increase the number of cells but the Situation did not improve, I will try again
5. I am measuring the mass flow at the inlet and at the outlet and I consider the convergence when those values does not variate more than 0.5% and their difference is negligible
6. the same, as I had a full chamber Simulation running and this with only half chamber is not, I do not know why
PS: i tried to improve the mesh reducing the Surface Growth Rate from 1.3 to 1.2, it went in Overflow, with this message
A floating point exception has occurred: floating point exception [Invalid operation]. The specific cause cannot be identified. Please refer to the troubleshooting section of the User's Guide and the Residual did what You can see in the Image linked