If one goes to my project https://www.simscale.com/workbench/?pid=3253427884872203341 to simulate gas flowing through a 2d domain, they can see that in simuations 2d domain, Simulation runs named 0.03 ends breaks time limit at 67% finish. It runs a transient simulation for 0.03s.
After this run, I tried another simulation with the same config but only for 0.02s. It is named 0.01-0.02, however 0.01-0.02 breaks time limit at only 50% finish. It is supposed to be a faster simulation, but somehow it takes more time than the 0.03 ends simulation
Hi yapabox,
I spent some time looking at the Solver logs and comparing your residuals for the two different 0.002 & 0.003 simulation runs to understand why the earlier fail for shorter time period, because like you it didn’t seem to gel with me.
I saw that even at the very start/beginning of the Solver logs there were differences between both, in the residuals, which if they were identical in every respect should provide an identical residual result.
I went through your setup and checked a few items and everything seemed to be identical from a quick scan & comparison. I then had a look at the residuals plots for both and compared. The plots were very similar in shape but with a different horizontal axis, but the peaks and troughs are occurring at the same horizontal time (sec) value. So the results are the same.
So my novice Simscale guesstimate is the that this ‘Adjustable time step’ should be set to False if you wish to get a similar time failure response for the 0.002 & 0.003 simulation runs. Regarding the settings for the ‘Adjustable time step’ I tracked down this documentation and there are others which may also assist in the ‘Help / Documentation’ SimScale Documentation | Online Simulation Software | SimScale
Whilst my review hasn’t provided you with a definitive answer that a more experienced Simscale person may be able to provide, you can possibly switch the ‘Adjustable time step’ to False and run both again and compare. However, there is an underlying problem in terms of both simulations are failing, but I think this is a matter for an experienced Simscale person.
I hope my second set of eyes considering your simulations has assisted a little and I wish you all the best on your future simulation modelling.
