Error when restart simulation
 

Good morning OpenFoamers!

I'm running a channel flow simulation with OF 1.5.x, I modified a little bit "channelOodles" to make a DNS of a channel and of course, I increase the mesh resolution to have a cell dimension like Moser, or Marusic show in their paper.

It's take long time to converge and sometime it stopped because there's not enough free space in hard disk. So I set, in controlDict:

starFrom startTime;
startTime 3000; (that for me it is the last field saved)

then I run again the simulation and what's happen? It happen that statistic, like "uprime", or "uvprime" are totaly wrong! I mean if "uprime_3000" is quite good, but not perfect, "uprime_3020" (next field saved) it's like "uprime_1000" or perhaps "uprime_100" that is the first field!

I'll post some pictures to be more clear:
Attachment 1315
Attachment 1316

/********************/

Then I try to change: instead to put:
startFrom starTime
startTime 3000

I wrote:
startFrom latestTime

and I comment the row "starTime ..."

But the result it's the same nothings change!!
Here there are pictures from second option:
Attachment 1317
Attachment 1318

/*******************/

What is it wrong?

Thank you.

MT