Hi I have noticed that some of my simulations are ending early.
example: I set 1000 as an end time, time step 1. and 1000 write interval but the simulation
ends at time 300. I still get the results ok.
Is this because it has converged and does not need to go any further?
Hi @gcanfield,
exactly - you can set prior to the simulation run when you consider the simulation to be “converged” via the residual control settings under Numerics (see screenshot). When all residuals have dropped below the specified threshold, the simulation is stopped and the last results written.
Best,
David
Excellent Thank you 
So if I see during a simulation that it has converged as much as its going to can I stop the simulation for the results as there’s no point in it continuing?
Hi @gcanfield,
so if you stop the run, it will also write out the last time step if that’s what you mean.
Best,
David
Hi David
Thanks I have tried this but I do not think it writes the last time step as when I try to post process it in paraview
it will not go to the last step. still not a problem though I will just not cancel.
Thanks again
Hello @gcanfield,
I guess you referring to e.g. your project SimScale where your first Run was cancelled at ~ timestep 900, but only timestep 0 has been written. I reported an issue and will keep you updated.
Best Alex
Hi Alex,
Yes that’s right but I have never been able to post process any canceled projects in the past.
It would be pretty handy if this was possible.
Thanks for your help
Ged
Hello @gcanfield,
sorry for my late reply. Recovering cancelled time steps worked in the past as well, but not in a robust way as you experienced. There were multiple stability fixes on that mechanism in the last weeks. I restarted your job in a copied project multiple times now and it always worked so I am confident that it works now also in a robust ways.
If you experience problems in the future with it, please tell me immediately. This feature is important and has high priority.
Best Alex
sorry I was logged in with a demo user 
Best Alex