Find answers, ask questions, and connect with our <br>community around the world.

Home Forums ANSYS Fluent Beginners to Advanced level 15. Steady simulation with different solvers

  • 15. Steady simulation with different solvers

    Posted by Ayoub on February 11, 2024 at 8:46 pm

    Hello everyone,

    Can you share with me your methodology for steady simulation, for instance how you chose the right under-relaxation factor in controls, when default ones don’t do the job i.e. diverged solution.
    I performed a comparison between: SIMPLE, SIMPLEC, PISO, Coupled, Implicit (density). For the test case of cylinder flow at Re=2.5 10^5 with inlet velocity V=200 m/s.
    Convergence was achieved for Coupled and Implicit only. These two require a CFL number by the user, I used CFL=10 for both.

    Tables below summarize the properties of and results for the different solvers.

    So, what is in your opinion the reason for diverged solution otherwise ?

    Thanks.

    Ayoub replied 1 month, 3 weeks ago 2 Members · 3 Replies
  • 3 Replies
  • Ayoub

    Member
    February 17, 2024 at 2:55 pm

    Up.

  • MIDHUN

    Member
    February 20, 2024 at 11:05 am

    Hi Ayoub,

    Usage of under relaxation factors are generally not needed for simple problems but when we work on some complicated problems in order to improve the convergence it can be used .We cannot solely depend on dropping of residuals alone when it comes to comparing different schemes ,we also need to monitor some significant parameter and check its variations and if the fluctuations are reduced then even though the residuals are like 1 e-3 or 1 e-2 we can say that the solution is converged .It can be pressure at outlet ,average mass flow , lift coefficient etc depends on our problem , so we need to analyze the residuals along with specific monitor parameters to get full idea on the comparison.

    Regards

    Flowthermolab SupportTeam

  • Ayoub

    Member
    February 25, 2024 at 8:49 am

    Thanks for your reply.
    Indeed, I monitored the value of lift and drag coefficients. They become stable only for : Coupled and Implicit (density-based) as I mentioned above.
    As you can see in the attached figures.

    What would you suggest to obtain a converged solution with these schemes ?

    Sincerely,

Log in to reply.

error: Content is protected !!