Learning Forum

[Hafiz]

Hello,

I am working on a 3D cylindrical shape reactor. Currently, I am doing a mesh independence study to assess the dependence of the simulation results. For that purpose,

I created a report definition of area-weighted avg of radial and tangential velocity in the XY plane and monitored them until they reach the steady-state along with continuity residuals which have reduced to 10-4.

So, my question is whether plotting that monitor point value of radial velocity vs the number of cells in the simulation is a correct way to study mesh independence?

[Rahul Kumar]

Hello, 

I do not see anything wrong in the method, however, it would depend on the type of flow that you are dealing with and also the end goal of your simulation. 

Regards, 

Rahul

[Hafiz]

Thank you for your comment. However, someone told me that instead of plotting the monitored radial velocity steady-state value of each simulation, I should plot a residual of radial velocity vs no. of cells. But I do not know how to obtain residual for radial velocity. The flow is Laminar and residuals for continuity with x, y, and z velocity are plotted only. Actually, radial velocity in my "closed" rotating system is nearly zero ranging from 10E-04 to 10E-06 and the end goal is to predict the flow pattern of radial velocity accurately. If I consider the tangential velocity, I get nearly constant steady-state value in all mesh cases which reveals that mesh resolution does not alter the tangential velocity. This is why I chose radial velocity to study mesh independence since it changes much with mesh resolution. 

What is your insight about this?

[DrAmine]

Method is okay. If you know which quantity is important for your case better to use it as residual velocity. But you are the only one to know that as I assume you did not start simulating without an aim.

I usually look into Nusselt numbers Heat transfer coefficients swirl number pressure drop wall shear, holdups, smd,and,....

[Hafiz]

Thank you Amine for your kind suggestion. As I mentioned, I monitored "Area weighted avg of radial velocity" over a plane by creating a plane monitor and used the final steady-state value when the solution is converged. I mean from this that

i) Monitored radial velocity reached the steady-state

ii) Residuals have reduced to 10-04

Then I picked the final value and plotted it vs no. of cells.

Now as you said it is better to use it as residual velocity. I don't understand your point completely. Would you please explain it more? 

How can I plot residual for radial velocity?

Best,