Fluids

Fluids

Variable time stepping method

    • ecrod
      Subscriber

      I am running a transient multiphase flow. If I select variable time stepping method in run calculation, I will have the following settings to adjust.

    • Rob
      Ansys Employee
      The various Courant Numbers will be defined in the documentation, so I'll leave you to find that. nIf you want a smoother transition, think about what those numbers represent: 0.8 and 1.2 Ie 20% up or down in size, if you set the value to be 1 what do you think might happen? nLastly, variable time stepping is great in theory. In practice, if your model has multiple and/or varying timescales how do you think the stepping will deal with that? n
    • subhamdas
      Subscriber
      I had a question. For explicit formulation(conditionally stable), discretising the governing equations and subsequently applying the stability criterion imposes a condition on the time step size(CFL condition). However, since the implicit formulation is unconditionally stable, hence no such condition is imposed on the time step.nHence, in case one uses the implicit formulation available in fluent, the CFL condition is ruled out, however, there should be a certain limit to the time step in order to acheive a physical solution. Is there an upper limit to the Courant Number for implicit formulation? I know the answer is yes but I needed an approximate value and the conditions governing the same.n
    • Karthik R
      Administrator
      Hello,nYou should be able to go with a higher value of time step with an Implicit formulation. I'd, however, not go higher than a Courant number of 5, especially for multiphase problems.nThank you.nKarthikn
    • subhamdas
      Subscriber
      Thanks that helped. But is the upper limit based on intuition or is there any explanation for the same. Although I have limited Courant number to 2 for most of the cases, I would like to understand the factors that govern the upper limit. Would it be different based upon the problem being simulated?n
    • Karthik R
      Administrator
      I generally tend to always be conservative (1 to 2.5) irrespective of the problem / application to obtain solutions. It's intuition and perhaps experience. nKarthiknn
    • subhamdas
      Subscriber
      Got it!n
Viewing 6 reply threads
  • You must be logged in to reply to this topic.