About Umean
 

dear all,
I put fieldaverage function into controlDict, and I run the simulation for, say 100s, then I stop here. Since at the beginning of the simulation non-physical phenomenon might occur, this would make the average filed, like UMean and pMean inaccurate.
To eliminate the effect from non physical phenomenon, I resume the simulation at t=100s. From then on, are UMean and pMean calculated from t=100? or are the values before t=100s still be counted into the average field?

Dear Kai,
Usually the averaging must be done after the full convergence i.e. in your case you must turn on averaging after t=100sec (Not from t=0sec) then continue your run for another 100sec i.e. you should sample your data at t=200sec.