Turbine blades experience significant surface degradation with service. Previous studies indicate that an order-of-magnitude or greater increase in roughness height is typical, and these elevated levels of surface roughness significantly influence turbine efficiency and heat transfer. This paper presents measurement and a mean-line analysis of turbine efficiency reduction due to blade surface roughness. Performance tests have been conducted in a low-speed, single-stage, axial flow turbine with roughened blades. Sheets of sandpaper with equivalent sandgrain roughnesses of 106 and 400 μm have been used to roughen the blades. The roughness heights correspond to foreign deposits on real turbine blades measured by Bons et al. [1]. In the transitionally rough regime (106 μm), normalized efficiency decreases by approximately 4% with either roughened stator or roughened rotor and by 8% with roughness on both the stator and rotor blades. In the fully rough regime (400 μm), normalized efficiency decreases by 2% with roughness on the pressure side and by 6% with roughness on the suction side. Also, the normalized efficiency decreases by 11% with roughness only on stator vanes, 8% with roughness only on rotor blades, and 19% with roughness on both the stator and rotor blades.

Turbine blades experience significant surface degradation with service. Previous studies indicate that an order of magnitude or greater increase in roughness height is typical, and these elevated levels of surface roughness significantly influence turbine efficiency and heat transfer. This paper presents measurement and a mean line analysis of turbine efficiency reduction due to blade surface roughness. Performance tests have been conducted in a low speed, single-stage, axial flow turbine with roughened blades. Sheets of sandpaper with equivalent sandgrain roughnesses of 106 and 400 μm have been used to roughen the blades. The roughness heights correspond to foreign deposits on real turbine blades measured by Bons et al. [1]. In the transitionally rough regime (106 μm), normalized efficiency decreases by approximately 4 percent with either roughened stator or roughened rotor and 8 percent with roughness on both the stator and rotor blades. In the fully rough regime (400 μm), normalized efficiency decreases by 2 percent with roughness on the pressure side and by 6 percent with roughness on the suction side. Also, the normalized efficiency decreases by 11 percent with roughness only on stator vanes; 8 percent with roughness only on rotor blades; and 19 percent with roughness on both the stator and rotor blades.

This paper presents an experimental study of the behavior of leakage flow across shrouded turbine blades. Stereoscopic particle image velocimetry and fast response aerodynamic probe measurements have been conducted in a low-speed two-stage axial turbine with a partial shroud. The dominant flow feature within the exit cavity is the radially outward motion of the main flow into the shroud cavity. The radial migration of the main flow is induced by flow separation at the trailing edge of the shroud due to a sudden area expansion. The radially outward motion is the strongest at mid-pitch as a result of interactions between vortices formed within the cavity. The main flow entering the exit cavity divides into two streams. One stream moves upstream towards the adjacent seal knife and reenters the main flow stream. The other stream moves downstream due to the interaction with the thin seal leakage flow layer. Closer to the casing wall, the flow interacts with the underturned seal leakage flow and gains swirl. Eventually, axial vorticity is generated due to these complex flow interactions. This vorticity is generated by a vortex tilting mechanism and gives rise to additional secondary flow. Due to these fluid motions combined with a contoured casing wall, three layers (the seal leakage layer, cavity flow layer, and main flow) are formed downstream of the shroud cavity. This result is different from the two-layer structure which is found downstream of conventional shroud cavities. The seal leakage jet formed through the seal clearance still exists at 25.6 percent axial chord downstream of the second rotor. This delay of complete dissipation of the seal leakage jet and its mixing with the cavity flow layer is due to the contoured casing wall. Time-averaged flow downstream of the shroud cavity shows the upstream stator’s influence on the cavity flow. The time-averaged main flow can be viewed as a wake flow induced by the upstream stator whose separation at the shroud trailing edge induces pitchwise non-uniformity of the cavity flow.

A detailed flow analysis has been carried out in a two-stage shrouded axial turbine by means of intrusive and non-intrusive measurement techniques. Multi-sensor Fast Response Aerodynamic Probe (FRAP) and 3D-PIV system were applied at two locations downstream of the first and second rotors. Several radial planes were measured focusing on the blade tip region in order to obtain a unique set of steady and unsteady velocity data. The investigation deals with the aerodynamics and kinematics of flow structures downstream of the first and second rotors and their interaction with the main flow in a partially shrouded turbine typical of industrial application. The first part of this work is focused on the flow field downstream of the first rotor while the second part studies the leakage flow in the cavity of the second rotor and its interaction with the main stream. The interstage region is characterized by interactions between the tip passage vortex and a vortex caused by the recessed shroud platform design. Flow coming from the blade passage suddenly expands and migrates radially in the cavity region causing a localized total pressure drop. The time evolution of these vortical structures and the associated downstream unsteady loss generation are analyzed. The partial shroud design adopted in this geometry is beneficial in terms of blade stress and thermal load; however flow field downstream of the first rotor is highly three dimensional due to the intense interaction between cavity and main streams. A flow interpretation is provided and suggestions for improved design are finally addressed based on the steady and unsteady flow analysis.

This paper presents an experimental study of the behavior of leakage flow across shrouded turbine blades. Stereoscopic particle image velocimetry and fast response aerodynamic probe measurements have been conducted in a low-speed two-stage axial turbine with a partial shroud. The dominant flow feature within the exit cavity is the radially outward motion of the main flow into the shroud cavity. The radial migration of the main flow is induced by flow separation at the trailing edge of the shroud due to a sudden area expansion. The radially outward motion is the strongest at midpitch as a result of interactions between vortices formed within the cavity. The main flow entering the exit cavity divides into two streams. One stream moves upstream toward the adjacent seal knife and reenters the main flow stream. The other stream moves downstream due to the interaction with the thin seal leakage flow layer. Closer to the casing wall, the flow interacts with the underturned seal leakage flow and gains swirl. Eventually, axial vorticity is generated due to these complex flow interactions. This vorticity is generated by a vortex tilting mechanism and gives rise to additional secondary flow. Because of these fluid motions combined with a contoured casing wall, three layers (the seal leakage layer, cavity flow layer, and main flow) are formed downstream of the shroud cavity. This result is different from the two-layer structure, which is found downstream of conventional shroud cavities. The seal leakage jet formed through the seal clearance still exists at 25.6% axial chord downstream of the second rotor. This delay of complete dissipation of the seal leakage jet and its mixing with the cavity flow layer is due to the contoured casing wall. Time-averaged flow downstream of the shroud cavity shows the upstream stator’s influence on the cavity flow. The time-averaged main flow can be viewed as a wake flow induced by the upstream stator whose separation at the shroud trailing edge induces pitchwise non-uniformity of the cavity flow.