Search

Different types of rotating disk boundary layers were previously shown to be absolutely unstable. The present study checks if this is also true for the flow between narrowly spaced co-rotating disks with merged boundary layers and imposed throughflow. The described method is able to reproduce the absolute instability of the von Karman boundary layer and of radially outward flow between co-rotating disks with separated boundary layers. Nevertheless, no absolute instability is found for the case of narrow disk spacing with merged boundary layers for both radially inward and outward flow. The data does however show convective instability. Stability maps are provided for the analysed parameter space.

In order to identify transitional flow, the linear stability of inward flow in a narrow gap between co-rotating disks is studied. Two methods are developed. The local approach assumes periodic nature of the flow in radial and circumferential direction, whereas the computationally more expensive biglobal variant only requires periodicity in circumferential direction. Stability maps are provided for a single set of operating conditions spanning a range of Reynolds numbers at fixed rotor geometry and rotational speed. The theoretical findings are compared to a set of experimental velocity profiles that includes the transitional region according to profile shapes. Apart from some explainable inconsistencies, the determined stability limits agree with each other and with results from previous studies.

The inward flow between two parallel, co-rotating disks undergoes a thorough examination by analytical, experimental and numerical means. The analytical approach utilizes the asymptotical truncated series solution provided by Batista (2011) and extends it by a correction for an arbitrary mean tangential velocity at the rotor inlet. Taylor series expansions of the analytical results provide an estimate for the orders of magnitude of velocity components and the polynomial order of their profile shapes. The common assumption of parabolic velocity distributions is only appropriate in the radial direction. In parallel, a unique test rig provides the experimental counterpart of the velocity profiles inside a rotor gap, that is suitably narrow for turbomachinery applications. The optical flow measurements are based on a novel calibration technique and volumetric particle tracking evaluation. Both laminar and turbulent operating conditions are examined. Finally, numerical studies using commercial CFD software provide insight into the flow field inside the test rig rotor where experimental methods fall short and provide an additional means to investigate the effects of the approximations made in the derivation of the analytical results. The velocity distributions acquired by analytical, numerical and experimental means agree well, the asymptotical nature of the analytical solution by Batista (2011) can be observed. The comparison of experimental and numerical results of a turbulent case suggests that the Shear Stress Transport turbulence model reproduces turbulent flow inside the rotor gap appropriately.