Hello,
I have a heat transfer simulation in Ansys CFX and it is a thermodynamic cycle and thus have a periodic solution. I do not know how convergence works for this problem , because it does not converge to a solution and is periodic. How should I study the convergence for this problem ? I would appreciate your help.
Your question so far is somewhat vague, at least for me. Am I right in assuming that you have some kind of hydrodynamic and thermodynamic transport equations, with a periodic forcing term? If so, what are you seeing and what is your difficulty?
I agree with Leyvraz. But based on the information you gave I would say: perform mesh independency tests and time step independency tests, like in any other non-stationary simulation. Since it seems as if your simulation is demanding I would recommend to refine the simulation (mesh-wise and step-wise) until your variable of interest (e.g. Work being done by the cycle or Heat extraxcted) does not change. Then you can assume your answer as valid. Not an expert, but that is what I would do.
Try to monitor some quantities such as pressure in a specific position and see if its trend is periodic.
Franco
@Patricio Javier Valades-Pelayo : How can I use the convergence criteria for the specific variable , like Work being done or Heat extracted?
@Seyed Alireza, sorry to be off the topic for long time. Assesing convergence in a periodic simulation is done by taking some path dependent function (e.g. Work obtained, irreversible energy losses, efficiency, etc.) and increase the simulation "resolution" (i.e. mesh size and timestep) until that quantity is independent of meshsize; independent in this case means that the trajectory it follows (oscillations amplitude, frequency, etc) is the same for different meshes...or if some variable such as work (area below the curve of a P,V diagram that) is the same for different meshes.
Usually, each simulation will need to be run until the state variables, in different cycles, follow the exact same trajectory (meaning the the dynamic system is periodic, the is oscilations are not growing or shirnking in any cycle).
- Badges
- Science topic