Calculation in cylindrical coordinates

Franz Wingelhofer

I want to simulate the flow through an annular turbine cascade. Because of the periodicity in circumferential direction I must use cylindrical coordinates with x as the axis of rotation, y in radial and z in circumferential direction. Using unmatched grids on both sides in circumferential direction the calculation was stopped because some nodes could not get connected to nodes on the opposite boundary. After transformation of the cartesian coordinates into cylindrical coordinates the calculation started and reached a kind of convergency.

My questions are as follows: Is that I have done correct? Did the solver really calculate in cylindrical coordinates? Or did it calculate in cartesian coordinates anyway, but using a changed geometry?