A. StwoRac€ © AE-GE CEM Proceedings of ASME TURBO EXPO 2001 International Gas Turbine & Aeroengine Congress & Exhibition May 8-11, 2000, Munich, Germany 2000-GT-0565 Unsteady Flow and Whirl-Inducing Forces in Axial-Flow Compressors. Part |- Experiment A.F. Storace, D. C, Wisler, H.-W. Shin and B. F, Beacher, GE Aircraft Engines Cincinnati, Ohio F.F. Ehrich, Z. 8. Spakovszky, and M. Martinez-Sanchez Massachusetts Institute of Technology, ‘Cambridge, Massachusetts S. J. Song ‘Seoul National University Seoul, Korea ABSTRACT. ‘An experimental and theoretical investigation has been conducted fo evaluate the effects seen in axial-flow ‘compressors when the centerline of the rotor is displaced from the centerline of the static structure ofthe engine. This creates circumferentially non-uniform rotor-tip clearances, unsteady flow, and potentially inereased clearances if the rotating and stationary parts come in contact, The result not only adversely affects compressor stall margin, pressure rise capability and efficiency but also generates an unsteady, destabilizing, aerodynamic force, called the Thomas/Alford force, which contributes significantly to rotor whirl instabilities in turbomachinery, Determining both the direction and ‘magnitude of this force in compressors, relative to those in turbines, is especially important for the design of ‘mechanically-stable turbomachinery components. Part I of this ‘wvo-part paper addresses these issues experimentally and Part, II presents analyses from relevant computational models. ‘Our results clearly show that the Thomas/Alford force ‘ean promote significant backward rotor whiel over much of the operating range of modern compressors, although some regions of zero and forward whirl were found near the design point, This is the first time that definitive measurements, coupled with compelling analyses, have been reported in the literature to resolve the long-standing, disparity in findings concerning the direction and magnitude of whitl-inducing forces important in the design of modern axial-flow compressors 1.0 INTRODUCTION AND THE NATURE OF THE ISSUES 11 Introduetion, Incteases in clearances resulting from rubs between rotating and stationary turbomachinery ‘components operating at tight levels of clearance most frequently result. from forces induced by the following situations: rotor unbalance (associated with imperfections in fotor manufacture or assembly), lateral deceleration during a hard landing, lateral forces induced by high-g and high-rate-of twin maneuvers, thermal bowing and/or asymmetric ‘ovalization of the easing, especially for fans. However, a potentially much more destructive mechanism for inducing rubs is whirl instability. Any radial deflection of the rotor relative to the stator creates circumferentially-non-uniform clearances and unsteady aerodynamic forces on the rotor as each blade traverses the ‘varying clearance gap. These unsteady forces are orthogonal to the deflection and therefore are a significant driver of rotor ‘whirl instabilities. The forces inerease in magnitude as the deflection increases so that above the onset speed where destabilizing forces overwhelm the stabilizing damping forces, the deflections are ultimately limited only by damage to the interacting parts or by damping forees. Consequently an accurate determination of their magnitude and direction is of ‘major importance in the design of safe, stable turbomachinery ‘components. Examples of whirl are hysteretic whirl, whit! associated with fluids trapped within cylindrical rotor cavities and plain journal bearings, ete 1.2 The Nature of the Iss ‘The unsteady, destabilizing, terodynamic eross-axis stiffness force that promotes rotor whirl was frst postulated by Thomas (1958) 1 Copyright © 2000 by ASMEMin Clearance (Afoi Loadina Increases) Min Clearance {Airol! Loading Decreases) Min Clearance (Atoll Loading Increases) Max Clearance (Airfoil Loading Decreases) ) Accepted model _for turbine tives forward whirl, is postive, Max Clearance (Alfol Loading Increases) b) Alford’s model for compressors dives forward whirl, [hs positive Max Clearance (Airfoil Loading Decreases) ©) Ehvich’s model for compressors tives backward whi Bis negative. Fig. 1 Models of whir-inducing forces in turbines and compressors. Net force F,= Fm + Fy acts perpendicular to the axis of displacement and drives rotor whit and Alford (1965) to explain rotor whit instabilities seen in steam turbines and jet engines respectively. Therefore, this force is generally refered to a8 the Thomas/Alford fore. Whirl in Turbines. Fora defected turbine rotor, it has teen shown experimentally that the airfoils in the closure zone are more highly loaded by aerodynamic forces tan the airfoils in the open clearance zone because the former are operating note efficiently (Kofskey and Nusbaum, 1968), This station is shown schematically in Fig. Ia for turbine rotoe whose centetine has been displaced upward along the ordinate by an mount +Y. This gives minimum clearance atthe top of the {urbine and maximum clearance atthe bottom. The forces at these two lotions ae the vector sum of the mean blade fore, Fy and the unsteady blade force resulting from the centerline olet, FAs suggested by Thomas (1958), summing the forces perpendicular to the ais of displacement results in a net force, fy = y+ Fa, duc tothe diference in airol loading Since Fx acts normal to the axis of aisplacement, its called a eross-xis (cross-coupled stiffness fore, As seen in Fig a, te dretion Df Fy acts to drive the rotor in orbital (whirling) motion about the nonalsplaced centerline in the same diection as rotor rottion, ie, Fx promotes forward whirl fr turbines. ‘Thomas postulated the fllowing model to compute a cross-coupled {erodyamiestfiess coefficient in terms of the torque ecting and a B coefficient. eeeeen Ee a $Y" Dy Measurements of transverse destabilizing fores in unshrouded turbines give postive values in the range fiom 2 10 5, (ilies, 1975 and 1977: Wohlab, 1975; Martnez-Sanchez and Jaroux, 1991). Whirl in Compressors. Alford (1965) hypothesized the same phenomenon for compressors, whereby aerodynamic, cross-axis forees caused by asymmetsic tip clearance feed energy into the whirling motion of the rotor. Alford reasoned! that during rotor whirl the circumferential variation in radial tip clearance causes a circumferential variation in efficiency so thatthe blading with the smallest clearance would be the most efficient, Alford further hypothesized that the compressor would pump to a circumgerentially-uniform exit staie pressure and therefore the mote efficient blading at tight clearance would have @ lower loading than the blading with larger clearance 180 degrees away. This situation, illustrated in Fig, Ib, shows thatthe net force, Fx, tends to cause forward rotor whirl, Thus, Alford concluded that compressors have positive B's so that whitl-inducing forces for both compressors and turbines are in the same direction. Ehrich (1993) hypothesized differently from Alford He reasoned that compressor airfoils with the smaller clearance would sustain a higher static pressure differential across their tips and would therefore be more highly loaded than the airfoils with larger clearance, 180 degrees away. As shown in Fig. lo, this dictates that the net destabilizing force in compressors, Fy , tends ta produce rotor whit! counter to the direction of rotation. Thus Ehrich concluded that compressors tend to have negative fi coefficients so that the direction of \whisl-inducing forces for compressors would be opposite to those for turbines. ‘There has been a disparity in the findings in the literature conceming the direction of rotor whitt in compressors, Vance and Laudadio (1984) found experimentally that the Thomas/Alford force is rolor-speed dependent and 2 Copyright © 2000 by ASME.‘mostly positive, except for some special combinations of rotor speed and stage torque where the direction of the force was reversed. Colding-Jorgensen (1992) found the same generality for the shape and slope of the relationship of ® coefficient vs. flow coefficient as later reported by Ehrich (1993), but the Colding-Torgensen results suggested a more positive level of the parameter than the negative levels reported in Ehrich’s work. Ehrich (1993) further showed that the experimental data of Vance and Laudadio (1984) implied that, for certain values ff torque and speed in their low-speed blower tests, the destabilizing forces tend to drive backward whirl, Other evidence was also accumulating in theoretical and experimental results of Yan, etal (1995) to indicate negative coefficients for compressors. In engine field experience, aerodynamic cross-axis forces were cited by Akin, Fehr and Evans (1988) as the destabilizing mechanism in the high-pressure rotor instability of the TF30 P1L1+ engine when it went into production in mid 1986. Vibration reject rates were as high as 50 percent until the instability was eliminated by using a squeeze-film damper atthe high pressure turbine bearing In view of the importance of the (& coefficient in esigning. stable turbomachinery components, the disparity between Alford's and Ehrich's conjecture, the mixed findings of researchers on the issues, the need for designers to often use very conservative methods, and the absence of a decisive resolution of rotor whirl issues, we formulated the experimental and analytic program described in Parts 1 and UL NOMENCLATURE ‘Ae Ansara fo} Cup™ Chord Kengt at tp fem) = Mean blade diameter (m) Fr Fu Fa Tange aia and ean foros on oor lads (N] y= Net unbalanced fore perpendicular to rotor deflection [N] Fy~ Net unbalanced fore pall tortor defection (N} Fu~ Unsteady bade ores (NI H= Blade height fer) Ky = Crose coupled stifnesscoeicint [Nn] . Throttle coefficient Pompressor inlet pressure and temperature (Kpa, degrees C] ‘Stoic pressure ating on blade surface (Kpa] {9= V2pmiU2, Dynan pressure for normalizing pressures [Kp] fa) Jaen or = Retr wea speed fe) X= Coordinate orthogonal the ¥ 2.0 OBJECTIVES AND DEFINITIONS 2.1 Objectives. The overall goal of Pat I was to provide ‘definitive resolution of the long-standing disparity in findings concerning the direction and magnitude of rotor whirl in compressoss. ‘There were three major objectives of Past 1. The first was to quantify any changes in compressor performance and airfoil loading produced in compressors when the rotor centerline becomes displaced or offset from that of the stator ‘The second was {0 determine which of the two models best describes whitl in compressors, the Alford Mode! of Fig, 1b or the Ehrich model of Fig. le. The third was to determine the direction and magnitude of rotor whitl-inducing, aerodynamic forces in axial-flow compression systems used in moder turbomachinery, including their design implications. Comparing the results from analytical and ‘computational models relative to the experimental data will be the subject of Part IL 2.2 Definitions. The following definitions will be helpful Rotor Whir! Instability - Unstable rotor whitl is defined as the self-excited orbital motion of the rotor centerline about its nominal or undisplaced centerline induced by a destabilizing tangential force which overcomes the stabilizing ‘extemal damping forces, There are several potential sources of such destabilizing forces. This paper focuses on the ‘Thomas/Alford forees. Vertical coordinate inthe direction of he deflection, 4Y = Rotor deletion [em] Z> Axial coordinate B= Normalized eross-coupled sifhesscoeicient, (the “Thomas/Alford Coefficient”) Rotor tip eleatance fem) oud seal clearance [em] n= V9 x 100%, torque eiieney, points = 2, Row coeteent BAU, Y= TALBU}OR, A): Work sefice, alee) } Presse ect, (0 Blade azimuth angle, ra p= Density pin speed ofthe rotor (rads) Supers = Averaged valu Subsernts {= Rotor tp Design point Stull point 3 ‘Copyright © 2000 by ASME