code_saturne User's Forum

Hello,

I'm using version 2.0 and I have a very basic question regarding the calculation of moments following a restart from a calculation in which moments were already being calculated.

My understanding is that, assuming I do not set IMOOLD = -1, the restart calculation would continue to add to the cumulated moments from the previous calculation and I would have the new time averages at the end of my new calculation when I divide by the overall cumulated duration.

In my restart listing, under calculation of temporal means, the index for IMOOLD is equal to the index for IMOM. I would really just like to confirm that if, for example, I calculated moments over 5000 time steps in one calculation, then restarted this for a further 5000 time steps, the total averaging time would be 10000 time steps.

I apologise for the basic question, but I look forward to a clarification.

Kindest regards,

Kristin

Hello,

I believe your understanding is correct, but as this is something even experienced users often find tricky, I would play it safe by testing this behavior on a few time steps, on a "well designed" test: for example define a very simple/small channel with a scalar initially at 0, with an inlet value of 1, and which "fills" the domain by convection in n time steps. Run n time steps, the restart. With the desired behavior, your mean value should be quite under 1, while a restart with imoold = -1 should give you a mean value around 1.

Cheers,

Yvan