Hi Vishal, I guess you have not changed any parameter across the two simulations? A possible workaround to your problem would be that of not restarting the Reynolds Stresses filter in the new simulation but combining the resulting RS fields at the end, using the CombineAvg FieldConvert module. I imagine this approach may require longer convergence times, as the current implementation of the Reynolds Stresses filter accumulates the mean employed for calculating RS during runtime and the mean would ideally require to converge in each simulation. Cheers, Andrea On 21 Mar 2019, at 15:00, Vishal Saini <vishal.saini.nitj@gmail.com<mailto:vishal.saini.nitj@gmail.com>> wrote: Hi all, I'm running an SVV LES simulation using the IncNavierStokesSolver. Ran the initialisation bit and finally started collecting the statistics using the filter below: <FILTER TYPE="ReynoldsStresses"> <PARAM NAME="OutputFile">AverageField_0pt1sOnwards</PARAM> <PARAM NAME="OutputFrequency">1000</PARAM> <PARAM NAME="SampleFrequency">1</PARAM> </FILTER> Checked the <uu> component on a sliced plane and it looked fine (top figure in the attachment). However, when I resume the simulation using the last written Average Field and the chk file as a starting point, the <uu> (and the other components) looks a bit corrupted (bottom figure in the attachment). <FILTER TYPE="ReynoldsStresses"> <PARAM NAME="OutputFile">AverageField_0pt5sOnwards</PARAM> <PARAM NAME="RestartFile">AverageField_0pt1sOnwards_stress.fld</PARAM> <PARAM NAME="OutputFrequency">1000</PARAM> <PARAM NAME="SampleFrequency">1</PARAM> </FILTER> May be I'm missing something! Has anyone encountered this before? If yes, is there a workaround? PS. Average velocities <U> look similar in the two cases. BR, Vishal <continuosVersusNonContinuousNektarSimulations_uu.png>_______________________________________________ Nektar-users mailing list Nektar-users@imperial.ac.uk<mailto:Nektar-users@imperial.ac.uk> https://mailman.ic.ac.uk/mailman/listinfo/nektar-users