[aykay]

Hi CFD community,

Consider me a Fluent beginner who is exploring CFD analysis on their own.

I am trying to calculate RPM of a 3D fan that rotates in response to the normal incoming airflow. I want to do this with 1DOF dynamic mesh motion using overset grid.

The fan is enclosed in 1DOF overset gird volume inside a bigger farfield volume (as shown in image taken from reference paper).

First, I verified setting up dynamic mesh motion with overset grid in 2D, The body completed several rotation cycles (seen from animation). I could not quantify the accuracy of results because I didnt have any 2D benchmark to compare my results to.

However, in 3D case, the overset grid keeps oscillating to about 1 degree roll angle. After convergence, I plotted the mesh at t=0.4sec. The mesh wasn't skewed/distorted, fan seemed to have moved slightly, residuals converged upto e-4. However, according to reference papers, the fan should have completed about 4 full cycles, and converged to a constant roll rate by this flow-time. I am calculating rotation rate using roll angle from "motion history" option in dynamic mesh setup.

Please evaluate my case and give your valuable suggestion on how to obtain correct RPM using 1DOF 3D overset grid.

Mach no = 2.16
Fan MOI= 0.001 kgm^2
time step = 1.4e-3 (also tried 9.8e-4)
mesh = 7.5M cells, unstructured 3D domain
no. of time steps = 2990
flow time = 0.4s

I have also tried changing mesh size, mesh y+, body MOI, time-step. Nothing has given me correct rotation rate (or rolling moment).