Hi I'm calculating some kind of mixer flow in 2 variants. 1. 1/4 model with cyclic BC and rotating coordinate sys 2. full model with sliding mesh

In both cases give roughly the same moment around the axis of rotation. In both cases if I monitor the moment over timestep or iteration the moment oscillates with very timestep or iteration chaoticly around a mean value.

Has someone ideas what could cause this behaviour ?

Regards,

Mark

What discretisation practises did you use in space and time? UD, Mars, Implicit, Crank-Nicholson?

Even if I use 1st order upwind in space and 1st order implicit Euler in time for the sliding grid case I get these high frequency oscillations, which seem to randomly distributed around the mean value.

regards,

Mark

Is your time step small enough not to cause your sliding motion to exceed one cell width for each sliding time step? I am not sure if this is necessary but maybe... How is your grid quality?