I am doing two-way FSI simulation (Fluent + transient structural) of flow through simple tube. I have kept Reynolds number of 1000 and I have tried with the time step of 10-4 and 10-5. In the fluent, residuals are diverging (increasing after each coupling iteration) and in the end error comes with large deformation in solid elements. Simulation is working fine for lower time step (10-2). Does anyone have idea about how to resolve this?
Do you consider time-varying flow and pressure in your calculations, or constant average values, e.g. constant outlet pressure?
I have given inlet velocity of 0.2 m/s (Re =1000 ). At outlet simply constant zero gauge pressure is provided.
The easiest way is to calculate the case without deformation first, and then turn on the structural analysis. For such boundary conditions I recommend the steady state solver. The calculations should converge very quickly.
Do you mean that, at first I should run simulation without deformation which will be initial condition for two-way simulation?
Actually I need to capture initial transient wall deformation due to the wave propagating through the tube within the fraction of seconds. Will steady assumption of flow work here?
Thank you Krystian Jędrzejczak for your response.
No, in this case you have to perform transient calculations. In terms of initial conditions, it depends on what conditions you want to reproduce. I assume you want to recreate a hydraulic shock, then the initial conditions mean no fluid movement. Have you estimated the wave propagation speed in relation to mesh size and time step to smoothly represent both flow and deformation?
Yes, I have initialized with zero fluid moment. I have calculated time step of 10-4 based on the refined mesh and keeping courant number below 0.5. What do you mean by wave propagation speed? In my case of incompressible flow will it be same as the order of inlet velocity?
It is worth checking the calculations in terms of residuals fo simulation without deformations and checking the applied values of time steps in literature sources for similar velocities and pipe material with a similar Young's modulus.
- Badges
- Science topic
- Similar topics
- Correction