August 16, 2021, 06:34
|
Final result changes if simulation is interrupted and restarted
|
#1
|
New Member
Raoul
Join Date: Apr 2021
Posts: 16
Rep Power: 5
|
Hi everyone, I am dealing with a nonNewtonian laminar flow that I solve through a customized Pimple solver (as more equation are required). I have recently noticed that if I stop the simulation and then I restart from the latest time step the results will considerably change with respect to the ones obtained not stopping the simulation.
I am stopping the simulation in a forced way (I know it is not optimal, but I always make sure through the log that the "latest" folder has been written completely). After that, when I restart from "latestTime" everything starts changing, reaching after a couple of seconds a completely different state.
I am using the adaptive time step, but after I restart, the time step value is the same of the unarrested one.
Do you have any suggestion about this? Did it ever occur to you?
|
|
|