Blade-Row Interaction in a High-Pressure Turbine

This paper identifies and analyzes the propagation of aerodynamic deterministic stresses through a two-spool counter-rotating transonic facility representative of modern and future turbine aeroengine sections. The test setup consists of a high-pressure stage, a diffusing turning midturbine frame with turning struts, and a counter-rotating low-pressure rotor. The flowfield downstream of the high-pressure stage is strongly influenced by the stator-rotor interaction. Such a mechanism interacts again with the downstream turning midturbine frame leading to a vanerotor-vane interaction, which affects the behavior of the low-pressure stage. The results presented were obtained using a fast-response aerodynamic pressure probe for unsteady measurements as well as three-dimensional unsteady Reynolds-averaged Navier-Stokes calculations. The work is presented in two parts. This first part focuses on the explanation of the flow physics that governs the convection of unsteady three-dimensional flow through the midturbine duct. Viscous and inviscid mechanisms are discussed as main drivers for the convection of wakes, secondary vortices, and shocks. The flowfield in the duct is characterized by three superimposed effects: 1) duct diffusion and radial pressure gradient together with turning strut potential field, 2) rotor unsteady work source, and 3) vane/blade interaction phenomena. The understanding of these mechanisms will eventually help to control the unsteadiness content in future architectures where reduced engine component length will enhance the interaction effects.

Section: Cfd Resultsmentioning

confidence: 99%

Unsteady Flow Evolution Through a Turning Midturbine Frame Part 1: Time-Resolved Flow

Lengani

Spataro

Paradiso

et al. 2015

“…5), the center positions of the tip-leakage vortex and the tip-passage vortex can be derived by the local maximum of the turbulence level. The relation between the turbulent-loss core of a vortex and its velocity field was investigated by Chaluvadi et al [37]. The exact positions are marked in the plots with dashed hairlines.…”

Section: Flow Sensitivity On Tip-gap Height and Injectionmentioning

confidence: 99%

Desensitization of the Flowfield from Rotor Tip-Gap Height by Casing-Air Injection

Behr

Kalfas

Abhari

2008

This paper presents an investigation that aims to desensitize the flowfield of an unshrouded high-pressure gasturbine stage from the radial clearance between the rotor tip and the casing. A novel method of cooling-air injection from the rotor casing into the rotor tip region is applied. The injection opposes the tip-leakage flow and affects the development of the rotor secondary flow. In a previous investigation, an increase of stage efficiency of 0.55 percentage points was achieved with an injection of 0.7% of the core mass flow. The present study investigates the effect of the injection on the flowfield for two different rotor tip clearances. Measurements conducted with a two-sensor fastresponse pressure probe provide data describing the time-resolved behavior of flow angles and pressures as well as turbulence intensity. The experimental investigation is done on the tip-gap heights of 0.65 and 1.00% span and injection mass flow rates representing 0, 0.7, and 1% of the turbine core mass flow. The results show that an increasing injection rate reduces the difference of rotor exit flow angle and flow unsteadiness between the two different tip-gap heights. The stage efficiency increases with injection for both tip-gap cases, whereas identical values are obtained at the injection rate of 1% of the turbine mass flow. Nomenclature BR = blowing ratio ( c c c = m c m ) Cp = pressure coefficient p p s;3 =p t;0 p s;3 c = absolute flow velocity, m=s c p = specific heat capacity at constant pressure, J=kg K DR = density ratio ( c = m ) _ H = enthalpy flux, W h = enthalpy, J=kg IR = impulse ratio (BR 2 =DR) M = torque, N m _ m = mass flow, kg=s p = pressure, Pa r = radius, m T = temperature, K Tu = turbulence intensity, % TKE = turbulent kinetic energy, J=kg u = rotor velocity, m=s v = rotor-relative velocity, m=s = efficiency = isentropic coefficient (c p =c v ) = density, kg=m 3 = standard deviation = loading coefficient (h=u 2 ) = flow coefficient (c x =u) Subscripts c = coolant jet I = rotor injection flow is = isentropic M = first-stage main inlet flow m = local main flow r = radial coordinate rel = rotor-relative s = static sw = streamwise coordinate T = first-stage total exit flow t = total w = wall x = axial coordinate = circumferential coordinate 0 = turbine inlet 3 = turbine exit Superscripts = mass average 0

“…A S INDIVIDUAL blade rows are driven to higher loadings and moved in closer proximity to one another, blade row interaction phenomena, both potential and vortical effects, are magnified and have become the focus of several research investigations [1][2][3][4]. Potential disturbances from both the upstream and the downstream blade rows decay exponentially and thus typically only affect adjacent blade rows.…”

Section: Introductionmentioning

confidence: 98%

Vane Clocking in a Three-Stage Compressor: Frequency Domain Data Analysis

Key

Lawless

Fleeter

2009

Blade row interactions affect compressor performance and durability. As design systems expand to account for these interactions, a better understanding of the underlying physics is necessary. In this paper, results from a vane clocking experiment in a three-stage compressor are discussed. Efforts are focused on the second stage, specifically the change in stator 2 wake profiles with respect to the placement of the stator 1 wake. The two clocking conditions presented position the wake from stator 1 at the leading edge of stator 2 and in the middle of the stator 2 passage. The time-accurate data are Fourier decomposed to determine the relative magnitudes of the frequencies in the spectrum. With data acquired at 50 circumferential locations spanning one vane passage for each clocking configuration, an enormous amount of data is collected, and a useful method for synthesizing this information is presented. Results show that, by placing the stator 1 wake at the leading edge of stator 2, the stator 2 boundary-layer response to the large incidence variations associated with the rotor 2 wakes is dampened, resulting in a thinner and more shallow stator 2 wake. Nomenclature CL LE = clocking configuration that positions stator 1 wake at leading edge of stator 2 CL MP = clocking configuration that positions stator 1 wake in the middle of the stator 2 passage f = frequency m c = corrected mass flow rate P o = total pressure R = rotor S = stator U = wheel speed V = flow velocity in the absolute reference frame = absolute flow angle = adiabatic efficiency