Hello, I'm simulating a rotating disk in the air but my simulation doesn't correspond to the experimental measurements. I've already meshed the disk and the control volume, and the mesh quality is about 0.3 and there aren't bad cells. Within this velocity, the temperature around the disk is about 100°C by experiment but during my simulation the temperature is only at 50°C. The aim of this simulation is to determine the moment at the axis. Now, I try to change the fluid (air) into ideal gas and change the options for the density and viscosity.

Could someone give me some advice about really important parameter ? What is really the consequence of the turbulent viscosity option in fluent?

Disk specification: Radius 95 mm; thickness 2 mm; roughness 5 micrometer, frequency 0-500 Hz

Model: k-epsilon RNG; coupled; energy

Best regards

anthony

Check the y+ value near the wall. If you are using the wall function approach, the y+ has to be above ~30. If, however you are using the enhanced wall treatement, then a y+ ~1 is desirable. Check the fluent user guide for recommendations on the grid for near wall computations.

Regarding the correspondence with experimental data, is the temperature measured in the experiment, the surface temperature of the disk, or the temperature of the air? If it the surface temperature of the disk, then you will have to predict the heat flux at the surface of the disk from your CFD simulation and then perform a heat conduction study in order the determine the surface temperature of the disk.

In any case, proper resolution of the mesh near the wall is important.

Hope that helps, Raj Kiran

Thank you,

Now we have remesh properly and we try to simulate but we still have some questions.

Is k-omega standard with shear flow correction, coupled, density (ideal gaz) and viscosity (kinetik theory) the best option?

I use the ploter under solve/monitor/moment and I choose the rotating wall to obtain the moment at axis z. Is it the best way in order to obtain the total moment on axis z?

How can I decrease the residuals (I'm near ten).

Thank you for answer, the values don't fit the measurement so we are really stuck.