Question about convergence criteria

[Tu-Shiqin]

Hi Dear all,

I have some questions about the termination of calculation in Anura3D.
As shown in Section 6.5 of ScientifciManual, the convergence criterion for each time step is adopted for quasi-static problems. I want to observe the dynamic process of the system which will be static eventually. The end of simulation is determined by the number of load step and time per loadstep. Incomplete dynamic process will be obtained if fewer time steps are choosed, while excessive time steps leads to low computing efficiency since the model is already static after a certain time. For example, the sand is quasi-static after 50 timesteps in the "column collapse" simulation (section 6 of Tutorial manual).

Questions:

1. How can convergence criteria be applied to the whole process of dynamic simulation instead of every load step?
2. Is there any convenient operation that can obtain the kinetic energy of each time step (or dimensionless energy ratio E)?

Thank you all for any helps.
Tu

[ayerrocolom]

Dear Tu,

Anura3D always performs dynamic simulations. If the convergence criterion is activated, the only thing is that the code will stop when the out-of-balance force and the kinetic energy ratios of the system both are below a certain tolerance. In the CPS, you can also define a maximum number of time steps, so the calculation will stop even if a quasi-static equilibrium is not reached. The user should check if the code has finished because the calculation truly achieved the equilibrium or if you reached the maximum number of time steps. You can check in the ENG file, where the information regarding out-of-balance force and the kinetic energy ratios is recorded for each time step.

I hope this helps.
Alba

[Tu-Shiqin]

Thank you for your answers!