help high initial residuals

Aly

Hi

Hope any mathematician can help me.

I am not using a cfd code but using a propriety code (not much inofrmation is avaiable about it) based on two fluid model with staggered mesh scheme. I have used steady state flowrates values as input BC with outlet BC constant at atmospheric pressure, I got results that were far from experimental values.

next I tired putting experimental values (flow rates) as input BC keeping outletBC constant at atmpospheric pr. I also tried the reverse keeping experimental measured time series at outlet and keeping inlet BC (steady state values) at inlet BC. what I want to know is that in both cases there are few scattered high residual peaks in initial simulation till 400s (t=0 to 3600s) and then the residuals drops down to 10-8 and below. I was wondering whether the trends (of varaibles like P, V etc) obtained in first 500s are the artfacts of these initail high residuals peaks or physics behind the phenomena?

looking forward to hear any suggestion.

Aly