[clac]

I have a simple experiment for which I am getting an unexpected result. Perhaps someone here could help.

I have invisid, incompressible, swirling flow through an axisymmetric pipe. Axial and swirl velocity are specified at the inlet. The swirl velocity distribution is such that the fluid is rotating as a solid body. Static pressure is specified at the outlet with the 'radial equilibrium' option checked.

I know what the difference in pressure between the wall and the centerline should be by integrating the centrifugal force over the radius.

Using the Density based solver and a stationary reference frame I get the expected pressure gradient. When I specify a reference frame rotating about the pipe axis I get a different pressure gradient. The magnitude of the pressure gradient is dependent on the speed of the reference frame motion. How can this be as I am solving the same physical problem just in a different frame of reference (inlet velocities are specified in the absolute frame)?

Running this same experiment using the pressure based solver yields expected results. That is the pressure gradient from the center line to the pipe wall is not dependent on the reference frame speed.

Thanks for any help