[Boone]

Hi all,

I am investigating an isothermal jet in cross flow created with 2 boxes linked by a hole with imposed and constant velocity profiles at the inlet of each room and a pressure difference between the two rooms. I am using RANS k-w SST model. I obtain oscillations in my results despite I reduced the timestep to dt=1e-7s which leads to a maximum CFL of 0.2. In addition, the profiles of velocity and k, among others, are modified when reducing the timestep. Time integration is performed using implicit Eulerian method. Does this mean that the flow is unsteady and needs URANS ?

If yes, are flows over a backstep also unsteady (because the reattachment distance is fluctuating over time)? What about the flow around a cylinder in case of vortex shedding?

May someone clarify this question of steadiness?

Thank you in advance for your help,
Boone

[FMDenaro]

Originally Posted by Boone

Hi all,

I am investigating an isothermal jet in cross flow created with 2 boxes linked by a hole with imposed and constant velocity profiles at the inlet of each room and a pressure difference between the two rooms. I am using RANS k-w SST model. I obtain oscillations in my results despite I reduced the timestep to dt=1e-7s which leads to a maximum CFL of 0.2. In addition, the profiles of velocity and k, among others, are modified when reducing the timestep. Time integration is performed using implicit Eulerian method. Does this mean that the flow is unsteady and needs URANS ?

If yes, are flows over a backstep also unsteady (because the reattachment distance is fluctuating over time)? What about the flow around a cylinder in case of vortex shedding?

May someone clarify this question of steadiness?

Thank you in advance for your help,
Boone

This topic has been already discussed in some previous posts.

[Boone]

I effectively found answers in this post: Question on steady and unsteady simulation.



If I understood well, I am probably in the case of an unsteady flow with a not reachable steady state due to its unstability to perturbations (approximate BC, truncation errors, imperfect level of convergence of the iterative solver).