Application of Sweep to Low Pressure Turbine Cascade Blade for Tip Flow Containment

Application of Sweep to Low Pressure Turbine Cascade Blade for Tip Flow Containment Rachel F. Trehan 1 and Bhaskar Roy 2 Indian Institute of Technology-Bombay, Mumbai, 400076, India A numerical investigation on a low-speed linear cascade has been done to study the impact of sweep applied at the tip of a turbine rotor blade on tip leakage flow. Two forward and two backward swept blade tip modifications have been applied to the T106 profile to create new blade cascade configurations. The aim of applying sweep to the blade s tip is an attempt to reduce tip losses. This can be achieved by reducing the leakage mass flow rate or by altering the leakage flow to reduce the leakage vortex strength, which is the main contributor to tip losses. A detailed comparison of the tip region including blade loading, mass flow, turbulent kinetic energy, pressure gradients, and velocity vectors, has been conducted to gain insight into the flow structures within the tip gap. A similar detailed comparison has been conducted for the main passage to determine leakage vortex formation, location, size, turbulent kinetic energy and interaction with secondary flow. Forward sweep was observed to reduce mass flow rate, however, an increase to the tip gap vortex size and strength was also observed; which raised the turbulent kinetic energy introduced into the leakage flow, thereby increasing the size and strength of the leakage vortex and increasing pressure loss. Backward sweep, contrarily, increased mass flow but reduced the tip gap vortex, thus decreasing the turbulent kinetic energy introduced. Therefore the leakage vortex size and strength was reduced, ensuring a reduction in pressure loss. Nomenclature c = chord C p = pressure coefficient C z = axial chord, C axial h = blade height or span LE = leading edge LP = low-pressure P, P 0 = static pressure, stagnation pressure P ref = reference pressure PS = pressure surface s = pitch, spacing ratio SS = suction surface TE = trailing edge TKE = turbulent kinetic energy X,Y,Z = tangential, spanwise, axial direction τ = blade tip gap clearance ζ = pressure loss coefficient 1 Graduate Student, Aerospace Engineering, Propulsion Lab, Main Aerospace Engineering Building, and AIAA Student Member. 2 Professor, Aerospace Engineering, Office 205B, Main Aerospace Engineering Building, and AIAA Associate Fellow. 1

I. Introduction ip leakage flow and its associated vortex is one of the major causes for loss of aerodynamic Tefficiency in turbine rotor blades. Leakage flow is primarily induced by the pressure difference that exists between the suction and the pressure side. The cross flow that leaks across the tip gap disturbs chordwise flow near the tip, reducing work done by the tip area and incurring mixing losses in the main passage flow. This mixing generates a tip leakage vortex, which normally forms near the suction side of the blade. The leakage vortex blocks the passage flow and causes unsteadiness downstream. Viscous effects in the tip gap may also result in cross flow separation on the blade tip, adding further to the tip losses. The size and intensity of the leakage vortex is directly influenced by the tip leakage flow. Reduction of the leakage mass flow improves the aerodynamic efficiency of the turbine. Tip clearance losses account for roughly one-third of the total stage loss experienced by axial turbines. 1,2 These losses are much higher than seen by the compressors due to the high curvature of the turbine blades. Recent turbomachinery publications have focused on tip geometry modification as a way to reduce the tip leakage flow and therefore reduce tip losses. Several investigations on tip squealers and winglets, such as those conducted by Li 3 and Krishnababu 4, have shown that certain blade tip modifications reduce the mass flow rate through the tip gap and therefore decrease the size of the leakage vortex and its penetration into the main passage flow. Depending on the applied tip modification, reduced mass flow rate is achieved by either creating a blockage in the tip gap, such as a recirculation zone, or by enlarging the tip separation bubble that exists on the pressure side of the blade tip. Other investigations have shown that redirecting the leakage flow streamlines can also result in a successful reduction of the leakage flow vortex size and strength. Kusterer 5 altered the blade tip to redirect the radial flow along the pressure side away from the gap and toward the casing. This reduced the pressure difference across the tip gap, which reduced the mass flow rate and decreased the leakage vortex. Tallman 6,7, however, was successful in reducing the leakage vortex by using tip modification to turn the leakage flow toward the camber direction inside the gap. The leakage flow therefore exited the gap at a direction more aligned with the main passage flow and is less conducive to forming a leakage vortex, thus reducing the size and strength of the leakage vortex. 7 Sweep and lean are other possible methods that may be employed to alter the tip flow and reduce the leakage vortex. The application of sweep for improved performance of compressor blades has been a point of interest for the last few years in 3D blade design techniques. There are many published articles available in open literature regarding the benefits and disadvantages of applying forward and backward sweep to compressor blades. Investigations for application of sweep to turbine blades, however, have been few and most available articles, such as those published by Pullan 8,9 and Spataro 10, focus on blades with no tip clearance to determine impact on blade profile losses and on secondary flow. 2

The present study focuses on a low-speed computational investigation into the application of sweep to a blade cascade for potential low-pressure turbine (LPT) application in an attempt to contain and reduce tip losses. A three-dimensional linear cascade for an aft-loaded LP turbine rotor Figure 1: T106 Blade Profiles blade, T106, with chord length of 100 mm and a tip clearance of 5 mm, τ/h = 3.45%, has been created and investigated using the commercial flow solver Star- CCM+. The blade design inlet and outlet flow angles are -37.7 and 63.2 with a stagger of 30.7. The axial chord length is 86 mm, with a blade aspect ratio of 1.45 and a pitch to chord ratio of 0.714. The C p distribution at various spanwise locations and blade surface pressure contours were reviewed and it was Figure 2: Cascade Blade Mesh Profiles determined that the leakage flow impacted the blade loading most significantly from 90% span height to the blade tip (100% span height). Further tip flow analysis revealed a tip gap vortex, as well as a leakage vortex, due to the large tip clearance. New blade profiles were created using sweep in an attempt to restore tip blade loading and reduce tip leakage flow and the leakage vortex. Sweep, defined here as a deviation of the stacking axis such that it is no longer orthogonal to the mainstream flow direction, was applied to the end 10% of the blade. From 0% (hub) to 90% blade height, the blade s stacking axis remained straight and orthogonal to the mainstream flow. From 90% span height to the tip, the airfoil sections were shifted along the axis of the design flow inlet angle, 37.7 from the axial chordline, forward (into mainstream flow) and backward (away from mainstream flow) as shown in Figure 1. Four new blades were created by sweeping the blade tip region, as described above, 15 and 30 in the forward and in the backward directions. The unswept blade, the 30 backward and the 30 forward swept blade mesh profiles are shown in Figure 2. The purpose of this study is to determine the direct impact of applied sweep on a typical turbine rotor blade tip flow through cascade studies in an attempt to reduce tip losses; therefore the 2D baseline airfoil shape has not been modified. However, such modification may be borne out of this investigation at a later date. II. Numerical Methods The CFD simulations presented in this paper have been performed with Star-CCM+, an engineering simulation commercial software, that includes various physics and turbulence models, 3D-CAD modeling, CAD embedding, CAE Integration, an automatic meshing tool, and 3

a post-processor. Five three-dimensional linear cascades were created using multi-block elliptic o-grid meshes with approximately 1.22 million hexahedral cells (+/- 10,000). Simulations were run using a coupled flow and energy solver, with a k-ω turbulence model at a constant density (M 1 < 0.1). The inlet boundary was defined as a velocity inlet with a low-speed free stream velocity of 15.355 m/s at 37.7 from the axial direction, parallel to the applied sweep. The meridional planes were set as periodic boundaries and the outflow was defined so that mass flow was conserved. The computational domain was swept in the tangential direction to maintain the same distance between the blade s suction and pressure surfaces and the periodic boundaries, preserving the blade s location in the middle of the tangential space. III. Numerical Results This study demonstrates the impact of forward and backward sweep on the tip flow field for a large tip gap clearance. The sectional airfoil C p distributions at various spanwise locations are analyzed for the unswept blade and the four swept blades. Figures 3 and 4 compare the blade C p distributions for the 30 swept and the unswept cases at 94% and 98% span (measured from the hub) to the C p found at the midspan. It is observed from the C p plots that forward sweep causes a significant deviation from the desired blade loading near the tip, when reviewed compared to the unswept blade. In Figure 3, at 94% span, the forward swept blades have two obvious regions of minimum pressure peaks in the aft region, whereas the backward swept and unswept cases only have one. These peaks indicate the chordwise Figure 3: C p at 94% Span Compared to Midspan C p location where the tip leakage vortex interacts with the flow on the blade s suction surface. At 98% span, the baseline and backward swept C p distribution curves also have two pressure peaks (Figure 4). The location of the first suggests the leakage vortex forms around 30-40% axial chord. Backward sweep, as demonstrated by Figure 3 and 4, helps to recover the C p distribution on the blade s suction surface closer to that observed at the midspan. It reduces tip leakage vortex interaction with the suction surface flow and minimizes the pressure loss in the aft region due to the large tip clearance. The 15 forward and backward swept cases demonstrate similar C p distributions as that of the 30 forward and backward swept cases. 4

Figure 4: C p at 98% Span Compared to Midspan C p The C p figures also illustrate how application of forward sweep leads to a recovery of C p distribution along the blade s pressure surface towards that observed at the midspan location. The rise of pressure along the pressure side combined with the reduction of pressure along most of the suction surface results in a larger pressure difference across the tip gap due to forward sweep. Backward sweep, however, reduces the C p along the pressure surface to a greater extent than that observed for the unswept blade, resulting in a reduced pressure difference across the gap. Both effects can be attributed to the modified geometry of the blade tip. Backward sweep facilitates more flow along the blade pressure surface to be drawn into the tip gap due to an obtuse flow-turning angle around the blade tip compared to that of the unswept blade (90 turning). The forward sweep, however, discourages tip flow due to an acute turning angle. The calculated cross mass flow rate across the open tip illustrated in Figure 5 confirms this. Figure 6 gives a detailed look at the flow field in the tip gap for the unswept blade, the 30 backward swept blade and the 30 forward swept blade. The figure depicts the tangential-spanwise plane at 40% axial chord. The static pressure contours and the velocity vectors provide insight into the impact of the tip gap flow entry angle on the cross flow. Due to the large size of the tip gap, a tip gap vortex is observed on the blade tip corner for all blade profiles. This is evidenced by the Figure 5: Mass Flow Rate thru Tip Gap Calculated Along Camber Line regions of low static pressure as well as the reverse flow velocity vectors near the pressure side of the blade tip. The leakage vortex can also be clearly observed as the leakage flow exits the gap and interacts with the mainstream flow. The obtuse flow angle, provided by backward sweep, reduces the vortex inside the tip gap and allows more cross flow (increased mass flow rate in Figure 5) to enter the gap at a slower acceleration than observed for the unswept blade. The reduction in tip gap vortex size and in cross flow acceleration minimizes the leakage turbulence as verified by the mass-averaged turbulent kinetic energy (TKE) plotted in Figure 7. The reduced cross flow TKE also impacts the leakage flow interaction with secondary flow, which will be discussed later. 5

Figure 6: Velocity Vectors plotted against the Static Pressure Contours at 0.4C axial (P ref = 100 kpa) Figure 7: Mass-Averaged Turbulent Kinetic Energy thru Tip Gap The sharper blade edge provided by the forward sweep has the opposite effect to the backward sweep. The acute turning angle increases the tip gap vortex size, thus decreasing the gap area for the leakage flow to pass through and thus decreasing the cross flow (reduced mass flow rate in Figure 5). The larger tip gap vortex in conjuncture with the increase in cross flow acceleration produces a more intense leakage flow, with an increase in mass-averaged TKE (Figure 7). Figures 8, 9, and 10 depict scalar contours in the tangential-spanwise (X-Y) planes at 10% C axial, 30% C axial, 50% C axial, 70% C axial, and 90% C axial measured at the blades midspan. The absolute total pressure and TKE contours provide an insight into the chaotic and turbulent nature of the tip flow and compare how the different blade tip modifications impact the leakage flow structures and the secondary flow. The unswept blade contours in Figure 8 portray three vortices present near the casing: the tip gap vortex, the leakage vortex, and a vortex due to secondary flow. Backward sweep, as mentioned previously, increases gap mass flow and reduces leakage flow turbulent kinetic energy. This produces a smaller, less turbulent tip gap vortex as well as a smaller and less turbulent leakage vortex, seen in Figure 9. The backward swept tip also causes the leakage vortex to form closer to the secondary flow vortex, encouraging them to merge near the blade s trailing edge. This reduces the TKE observed in the passage as compared to the unswept blade and reduces the impact of tip leakage on the passage main flow. Forward sweep, contrarily, reduces tip gap mass flow and increases leakage flow TKE as illustrated by Figure 10. The larger and more turbulent tip gap vortex impacts the leakage vortex as well as the secondary flow vortex. The leakage vortex and secondary flow vortex have grown in size so that they cover the entire passage near the blade s trailing edge. The 15 forward and backward swept cases demonstrate the same flow behavior as the 30 cases, although to a smaller extent. 6

Figure 8: Unswept Blade Static Pressure and TKE Scalar Contours in X-Y Planes Figure 9: 30 Backward Swept Blade Static Pressure and TKE Scalar Contours in X-Y Planes Figure 10: 30 Forward Swept Blade Static Pressure and TKE Scalar Contours in X-Y Planes 7

As the leakage vortex increases in size and blocks the passage, the capability for work extracted from the main flow reduces and downstream unsteadiness increases. Minimizing the leakage vortex size and intensity reduces losses and improves aerodynamic efficiency. The pressure loss coefficient is evaluated using: Figure 11: Passage Mass-Averaged Loss Coefficient at 1.1C axial Figure 12: Passage Mass-Averaged Loss Coefficient thru the Passage 8 to quantitatively define the impact of forward and backward sweep on total pressure losses. The pressure loss coefficients for all five blade tip configurations at 110% axial chord are plotted in Figure 11. Forward sweep, as supported by the C p distribution plots, increases the total pressure losses even though the cross flow mass rate is reduced. This increase in total pressure losses downstream from the blade can be attributed to the formation and decay of a larger, more turbulent leakage vortex. The pressure loss coefficient at 110% C axial increases by 4.1% for 15 forward sweep and 14.6% for 30 forward sweep. Correspondingly, backward sweep decreases total pressure losses due to the formation and decay of a smaller, weaker leakage vortex. The pressure loss coefficient at 110% C axial decreases by 4.3% for 15 backward sweep and 12.3% for 30 backward sweep. These trends of increased pressure loss for forward sweep and decreased pressure loss for backward sweep are evident in Figure 12 from formation of the leakage vortex around 30% C axial to 125% C axial. IV. Conclusion In this study, four blade tip modifications have been investigated computationally to determine impact of sweep on the tip flow field. A low-speed linear turbine cascade with a tip gap clearance of 3.45% has been simulated to determine the impact of forward and backward sweep on the tip gap vortex, the leakage vortex, and tip losses. The following major conclusions have been drawn regarding sweep. 1. Forward sweep modifies the blade tip geometry such that the flow entry angle to enter the tip gap becomes more acute, raising pressure along the pressure surface post midchord, and discouraging fluid from entering. The acute angle increases the separation on the

blade tip pressure side, increasing the size and strength of the tip gap vortex. The stronger, larger tip gap vortex increases the turbulent kinetic energy of the leakage flow that exits the gap. This increases the size and strength of the leakage vortex as well as the secondary vortex. 2. Backward sweep modifies the blade geometry such that flow entry angle becomes more obtuse, encouraging fluid to enter. The obtuse angle reduces the separation on the blade tip pressure side, decreasing the size and strength of the tip gap vortex. The weaker, smaller tip gap vortex reduces the turbulent kinetic energy of the leakage flow that exits the gap. This diminishes the size and strength of the leakage vortex, thus reducing the pressure losses. 3. As the sweep angle increases, the flow field effects become more pronounced as does the impact on the pressure loss coefficient for forward and backward sweep. 4. In future investigations, other sweep configurations, such as pure axial or pure tangential sweep, may be simulated with a smaller tip gap clearance to determine impact on leakage flow and two or three optimum designs may be tested in a low-speed cascade rig and compared to the computational results. References 1 Bunker, R.S., Advances in Turbomachinery Aero-Thermo-Mechanical Design Analysis, von Karman Institute for Fluid Dynamics Lecture Series 2007, GE Global Research, NY, USA. 2 Lakshminarayana, B., Fluid Dynamics and Heat Transfer of Turbomachinery, John Wiley & Sons, Inc., New York, 1996, Chaps. 2, 4, 6. 3 Krishnababu, S.K., Newton, P.J., Dawes, W.N., Lock, G.D., Hodson, H.P., Hannis J., and Whitney, C., Aerothermal Investigations of Tip Leakage Flow in Axial Flow Turbine-Part 1:Effect of Tip Geometry and Tip Clearance Gap, Journal of Turbomachinery, Vol. 131, 2009, pp 011006-1. 4 Li, W., Qiao, W., XU, K., and Luo H., Numerical Simulation of Tip Clearance Passive Control in Axial Turbine, Journal of Thermal Science, Vol. 17, No. 2, 2008, pp 147-155. 5 Kusterer, K., Moritz, N., Bohn, D., Sugimoto, T., and Tanaka, R., Reduction of Tip Clearance Losses in an Axial Turbine by Shaped Design of the Blade Tip Region, Proceedings of ASME Turbo Expo 2007: Power for Land, Sea and Air, GT-2007-27303. 6 Tallman, J., and Lakshminarayana, B., Numerical Simulation of Tip Leakage Flows in Axial Flow Turbines, With Emphasis on Flow Physics: Part I Effect of Tip Clearance Height, Journal of Turbomachinery, Vol. 123, 2001, pp 314 323. 7 Tallman, J., A Computational Study of Tip Desensitization in Axial Flow Turbines, Ph.D. Dissertation, Mechanical Engineering Dept., Pennsylvania State Univ., University Park, PA, 2002. 8 Pullan, G., and Harvey, N.W., The Influence of Sweep on Axial Flow Turbine Aerodynamics at Midspan, Journal of Turbomachinery, Vol. 129, 2007, pp 591-598. 9 Pullan, G., and Harvey, N.W., The Influence of Sweep on Axial Flow Turbine Aerodynamics in the Endwall Region, Journal of Turbomachinery, Vol. 130, 2008, pp 041011-1. 10 Spataro, R., D Ippolito, G., and Dossena, V., The Influence of Blade Sweep Technique in Linear Cascade Configuration, Proceedings of ASME Turbo Expo 2011, GT2011-45728. 11 Xiao, X., McCarter, A.A., and Laskhminarayana, B., Tip Clearance Effects in a Turbine Rotor: Part 1 Pressure Field and Loss, Journal of Turbomachinery, Vol. 123, 2001, pp 296-304. 9