Measurements are performed in a stationary linear cascade of turbine blades for a range of clearances varying from 0.0 to 5.5 percent blade chord. The downstream flow field is measured with a seven-hole pressure probe at several axial locations. The results are analyzed both in terms of vorticity and total pressure loss. The former quantity provides insight into several aspects of the leakage flow, including size and strength of the leakage vortex in relation to the tip-clearance size and the bound circulation of the blade;​ the diffusion rate of the tip-leakage vortex and its interaction with the passage flow. In this context, a simple model of the vortex based on the diffusion of a line vortex is found to be quite accurate. Evaluation of the losses in the downstream measurement planes in relation to the losses measured inside the tip gap help clarify the dominant tip-leakage loss mechanisms. Based on experimental observations, full verification of existing turbomachinery tip-leakage loss prediction schemes is performed. An improved tip-leakage loss model is developed using one of the existing models as the starting point.

Following the stationary linear cascade tests, relative tipwall motion is simulated by a moving belt. Three clearance sizes of 1.5, 2.4 and 3.8 percent blade chord are examined. The tip-gap and downstream flow together with blade pressures are measured for several wall speeds. As the relative wall speed is increased, the mass flow rate through the gap is significantly reduced. The dominant mechanism behind this trend is identified. The relative wall motion alters the tip vortex and passage vortex structures substantially. The trends in the blade loading are found to be in agreement with the observed changes in the vortex structures. Based on the experimental results, the tip-vortex and tip-gap flow models are extended to the case of relative tipwall motion.

Succeeding the experimental and theoretical work on tip-leakage flow for plain tip geometry, blade tip treatments to reduce tip-leakage losses are tested with stationary and moving tipwall for clearance sizes of 2.4 and 3.7 percent blade chord. Three types of winglets consisting of pressure-side, suction-side and a combination of the two are considered in comparison to the regular blade tip. Losses extracted from measurements downstream of the blades suggest some reduction of tip-leakage losses with the application of winglets. The possible physical mechanisms by which winglets reduce the losses are discussed on the basis of downstream flow field data and tipwall static pressures.