I simulate the air flow in circular tube. The centerline of the tube is along z-axis. In the tube, there is the momentum source which is function of tube radius. To do this, I use 'Expression' of CFX-Pre.

I have checked the results in CFX-Post and found that using (x^2 + y^2)^0.5 and using 'r', as the variable for tube radius in 'Expression', provide the same results. However the output velocity profiles are not the function of tube radius. Then I suspect that 'r' may be not the proper variable for coordinate in radial direction for the tube. Do anyone have any idea for this?

Thank you very much for your kind consideration.

Post the expression you are using.

I'd like to provide the information related before answering your question.

The momentum source is the turbine in the tube. I model the turbine as the very thin disc which has the same axis as the tube and have diameter equal to the tube's diameter. This is the actuator disc concept. The tube is 100 m long and has 4m radius. The disc is at 50m after the tube entrance. The disc thickness is 0.1m.

To simulate axis-symmetry, a 5 degree section of the tube from the entire periphery is solved. (the cross sectional area of the tube likes a pie shape)

I specify 'Inlet' and 'Outlet' boundary condition to the inlet and exit of the tube, respectively. The 'symmetry plane' boundary conditions are applied at the two sides of the tube sector, and 'rotation' periodic domain interface are specified at the two sides of the disc sectors.

I specify the momentum source as the 'Theta Component' of 'General Momentum Source'. The momentum source is -(density * w * constant / r). I use 'w' because this is the axial velocity in the tube.

I hope this is clear to you. Thank you for your kind consideration.

I have checked in CFX-Post that 'density' and 'w' are constant in the momentum source (except near the center of the disc which the velocities are also changed).

I am so sorry I forgot to tell you one boundary condition. I specify the tube's wall and the wall of the disc as the 'free slip wall'.

I think there must be some error setting in my problem. I have tried one thing. I enter the axial momentum source at the disc, and no other source at the disc. The expression is -(constant * r).

The disc is at 50m after the disc entrance, and 0.1m in thickness. At 0.1m before entering disc, the axial velocities vary slightly, 13.94-14.01m/s. At the disc, however, the axial velocities are kind of constant when illustrated by contour plot. At 0.4 meter after the disc, axial velocity are vary slightly again, 13.94-13.49m/s. I do not know what is wrong???

Mixing symmetry and periodic BCs on one contiguous common plane isnt OK. Model each complete plane as a 1-1 periodic pair. If you are going to induce swirl in a tube use complete periodic BCs.

As regards the radial dependance of your momemtum source not working ... post your command file for inspection.

OK I'll try using only periodic.

According to the momentum, I use 'Expresion' in CFX-Pre. The expression is -(density * w * constant / r).