[Bazinga]

Good day everyone,

I just started using LES for my first very basic 2D atmospheric flows over a cube and I am wondering what residual target i should use for every time step.

For RANS i usually took the residual target of <10^-6 as describet in the COST Action 732.

Are there any practical guidelines that help defining the convergence criteria for LES?

[FMDenaro]

Originally Posted by Bazinga

Good day everyone,

I just started using LES for my first very basic 2D atmospheric flows over a cube and I am wondering what residual target i should use for every time step.

For RANS i usually took the residual target of <10^-6 as describet in the COST Action 732.

Are there any practical guidelines that help defining the convergence criteria for LES?

I suggest at each time step to ensure a low normalized residual in mass, momentum, energy and any other equation you solve, for incompressible flows is relevant you solve accurately the mass equation

[Bazinga]

Dear Mr. Denaro,

thank you for your answer.

I planned to monitor all residuals for the equations of the Navier Stokes equations as you described.

Till now I only worked with residuals and not normalized residuals. I will read some more on how to normalize the residuals, but I have one follow up question: What do you consider a "low" normalized residual?

[FMDenaro]

Originally Posted by Bazinga

Dear Mr. Denaro,

thank you for your answer.

I planned to monitor all residuals for the equations of the Navier Stokes equations as you described.

Till now I only worked with residuals and not normalized residuals. I will read some more on how to normalize the residuals, but I have one follow up question: What do you consider a "low" normalized residual?

when correctly normalized, is somehow common to have no more than 4-5 magnitude order smaller than the initial residual order. However, I suggest to be carefull in the mass equation