Unsteady Surface Pressures Due to Wake-Induced Transition in a Laminar Separation Bubble on a Low-Pressure Cascade

2004 ◽  
Vol 126 (4) ◽  
pp. 544-550 ◽  
Author(s):  
R. D. Stieger ◽  
David Hollis ◽  
H. P. Hodson

This paper presents unsteady surface pressures measured on the suction surface of a LP turbine cascade that was subject to wake passing from a moving bar wake generator. The surface pressures measured under the laminar boundary layer upstream of the steady flow separation point were found to respond to the wake passing as expected from the kinematics of wake convection. In the region where a separation bubble formed in steady flow, the arrival of the convecting wake produced high frequency, short wavelength, fluctuations in the ensemble-averaged blade surface pressure. The peak-to-peak magnitude was 30% of the exit dynamic head. The existence of fluctuations in the ensemble averaged pressure traces indicates that they are deterministic and that they are produced by coherent structures. The onset of the pressure fluctuations was found to lie beneath the convecting wake and the fluctuations were found to convect along the blade surface at half of the local freestream velocity. Measurements performed with the boundary layer tripped ahead of the separation point showed no oscillations in the ensemble average pressure traces indicating that a separating boundary layer is necessary for the generation of the pressure fluctuations. The coherent structures responsible for the large-amplitude pressure fluctuations were identified using PIV to be vortices embedded in the boundary layer. It is proposed that these vortices form in the boundary layer as the wake passes over the inflexional velocity profiles of the separating boundary layer and that the rollup of the separated shear layer occurs by an inviscid Kelvin-Helmholtz mechanism.

Author(s):  
Rory Stieger ◽  
David Hollis ◽  
Howard Hodson

This paper presents unsteady surface pressures measured on the suction surface of a LP turbine cascade that was subject to wake passing from a moving bar wake generator. The surface pressures measured under the laminar boundary layer upstream of the steady flow separation point were found to respond to the wake passing as expected from the kinematics of wake convection. In the region where a separation bubble formed in steady flow, the arrival of the convecting wake produced high frequency, short wavelength, fluctuations in the ensemble averaged blade surface pressure. The peak-to-peak magnitude was 30% of the exit dynamic head. The existence of fluctuations in the ensemble averaged pressure traces indicates that they are deterministic and that they are produced by coherent structures. The onset of the pressure fluctuations was found to lie beneath the convecting wake and the fluctuations were found to convect along the blade surface at half of the local freestream velocity. Measurements performed with the boundary layer tripped ahead of the separation point showed no oscillations in the ensemble average pressure traces indicating that a separating boundary layer is necessary for the generation of the pressure fluctuations. The coherent structures responsible for the large amplitude pressure fluctuations were identified using PIV to be vortices embedded in the boundary layer. It is proposed that these vortices form in the boundary layer as the wake passes over the inflexional velocity profiles of the separating boundary layer and that the rollup of the separated shear layer occurs by an inviscid Kelvin-Helmholtz mechanism.


2007 ◽  
Vol 111 (1118) ◽  
pp. 257-266 ◽  
Author(s):  
R. J. Howell ◽  
K. M. Roman

This paper describes how it is possible to reduce the profile losses on ultra high lift low pressure (LP) turbine blade profiles with the application of selected surface roughness and wake unsteadiness. Over the past several years, an understanding of wake interactions with the suction surface boundary layer on LP turbines has allowed the design of blades with ever increasing levels of lift. Under steady flow conditions, ultra high lift profiles would have large (and possibly open) separation bubbles present on the suction side which result from the very high diffusion levels. The separation bubble losses produced by it are reduced when unsteady wake flows are present. However, LP turbine blades have now reached a level of loading and diffusion where profile losses can no longer be controlled by wake unsteadiness alone. The ultra high lift profiles investigated here were created by attaching a flap to the trailing edge of another blade in a linear cascade — the so called flap-test technique. The experimental set-up used in this investigation allows for the simulation of upstream wakes by using a moving bar system. Hotwire and hotfilm measurements were used to obtain information about the boundary-layer state on the suction surface of the blade as it evolved in time. Measurements were taken at a Reynolds numbers ranging between 100,000 and 210,000. Two types of ultra high lift profile were investigated; ultra high lift and extended ultra high lift, where the latter has 25% greater back surface diffusion as well as a 12% increase in lift compared to the former. Results revealed that distributed roughness reduced the size of the separation bubble with steady flow. When wakes were present, the distributed roughness amplified disturbances in the boundary layer allowing for more rapid wake induced transition to take place, which tended to eliminate the separation bubble under the wake. The extended ultra high lift profile generated only slightly higher losses than the original ultra high lift profile, but more importantly it generated 12% greater lift.


Author(s):  
S. Sarkar

An attempt is made to describe the physical mechanism of transition of an inflexional boundary layer over the suction surface of a highly cambered low-pressure (LP) turbine blade influenced by the periodic passing wakes. Large-eddy simulations (LES) of wake passing over the T106 profile for a Reynolds number of 1.6×105 (based on the chord and exit velocity) are performed using wake data extracted from precursor simulations of cylinder replacing a moving bar in front of the cascade. The three-dimensional, time-dependent, incompressible Navier-Stokes equations in fully covariant form are solved using a symmetry-preserving finite difference scheme of second-order spatial and temporal accuracy. The present LES results are compared with experiments and DNS. The operating condition of a high-lift LP turbine blade leads to the formation of a separation bubble on the suction side. The interactions of incoming wake with this separation bubble complicate the transition process. Enhanced receptivity of inflexional boundary layer causes amplification of the perturbations produced by the passing wake leading to the formation of coherent vortices within the boundary layer. The transition mechanism during the wake-induced path is highly influenced by the convection and breakdown of these coherent vortices. Streamwise evolution of turbulent kinetic energy and production illustrates that these vortices play an important role in generation of turbulence and thus to decide the transitional length, which becomes time-dependent. LES results resolve a multimoded transition on the suction surface and the calmed region. The calmed region is nothing but an attached flow with low production as the boundary layer tends to relax after wake passing; the level of turbulent intensity suggests that the boundary layer is in a state of transition rather than laminarized.


Author(s):  
Hidekazu Kodama ◽  
Ken-ichi Funazaki

Abstract This paper describes the interpretation of a generation mechanism of profile loss of low pressure turbine (LPT) blades from a viewpoint of blade drag forces. On the analogy of profile drag of an isolated body, the profile loss of a cascade blade is subdivided into two components, the loss due to friction drag and the loss due to pressure drag. The friction drag is equal to the integral of all axial component of shearing stresses taken over the surface of the blade. The pressure drag, which does not exist in an inviscid flow, is due to the fact that the presence of the boundary later modifies the pressure distribution on the blade. The losses due to friction drag and pressure drag are evaluated for two kinds of blade profiles using the results of steady incompressible Reynolds Averaged Navier-Stokes (RANS) simulations at three different Reynolds numbers (Re), 57,000, 100,000 and 147,000. It is found that the trend of the total profile loss with Reynolds number is mainly determined by the trend of the loss due to pressure drag with Reynolds number. A rise in the total profile loss of the blade with a laminar separation bubble on the suction surface at low Reynolds number is mainly attributed to the increase in the pressure drag due to thickened suction surface boundary layer by the enlarged separation bubble. The friction drag and the pressure drag are also estimated for the measured data of low speed linear cascade tests with a moving-bar mechanism. In the estimation, the pressure drag is derived from the estimated total profile loss and the estimated friction drag by using boundary layer integral equations. It is found that the trend of total profile loss with incoming wake passing frequency is almost determined by the trend of the loss due to pressure drag with the wake passing frequency.


Author(s):  
Samuel C. T. Perkins ◽  
Alan D. Henderson

Studies on the effects of stator reduced frequency in low pressure turbines have shown that periodic wake-induced unsteadiness can increase steady flow circulation by as much as 15% and reduce losses compared to a steady flow datum. A large separation bubble downstream of peak suction that formed under steady flow conditions was periodically suppressed by wake passing events, resulting in significantly reduced losses within the boundary layer. This research extends this concept to a controlled diffusion compressor stator blade with a circular arc leading edge. The blade was placed inside a large scale, two-dimensional, cascade with a rotating bar mechanism used to simulate an upstream rotor blade row. The blade profile has been shown to experience leading edge separations and subsequent transition on both the pressure and suction surfaces due to a velocity overspeed caused by discontinuities in surface curvature. Testing was carried out at reduced frequencies of 0.47, 0.94 and 1.88 at the design inlet flow angle 45.5° and Reynolds number based on chord of 230,000. The freestream turbulence intensity was 4.0%. A range of experimental measurements were used to look at the blade’s performance: high resolution time-averaged blade surface static pressure measurements, inlet and exit 3-hole probe traverses and instantaneous, ensemble averaged and time average surface mounted hot-film measurements for the calculation of turbulent intermittency and quasi wall-shear stress. Results showed that increasing the stator reduced frequency from, 0–1.88, increased the overall blade pressure loss. The losses generated by the pressure surface and suction surface differed significantly and are affected very differently. The pressure surface demonstrated a clear reduction in loss with an increase in reduced frequency whereas the opposite trend was seen on the suction surface. Wake-induced turbulent strips suppressed the formation of leading edge separation bubbles that formed under steady flow conditions and in between wake passing events. Wake-induced turbulent strips reduced in width and level of turbulent intermittency through the favorable pressure gradients leading to peak suction and grew in the adverse pressure gradient of the velocity overspeed. The flow between wake-induced turbulent strips partially relaminarised through the favorable pressure gradient leading to peak suction.


2005 ◽  
Vol 128 (2) ◽  
pp. 221-231 ◽  
Author(s):  
S. Sarkar ◽  
Peter R. Voke

The unsteady pressure over the suction surface of a modern low-pressure (LP) turbine blade subjected to periodically passing wakes from a moving bar wake generator is described. The results presented are a part of detailed large-eddy simulation (LES) following earlier experiments over the T106 profile for a Reynolds number of 1.6×105 (based on the chord and exit velocity) and the cascade pitch to chord ratio of 0.8. The present LES uses coupled simulations of cylinder for wake, providing four-dimensional inflow conditions for successor simulations of wake interactions with the blade. The three-dimensional, time-dependent, incompressible Navier-Stokes equations in fully covariant form are solved with 2.4×106 grid points for the cascade and 3.05×106 grid points for the cylinder using a symmetry-preserving finite difference scheme of second-order spatial and temporal accuracy. A separation bubble on the suction surface of the blade was found to form under the steady state condition. Pressure fluctuations of large amplitude appear on the suction surface as the wake passes over the separation region. Enhanced receptivity of perturbations associated with the inflexional velocity profile is the cause of instability and coherent vortices appear over the rear half of the suction surface by the rollup of shear layer via Kelvin-Helmholtz (KH) mechanism. Once these vortices are formed, the steady-flow separation changes remarkably. These coherent structures embedded in the boundary layer amplify before breakdown while traveling downstream with a convective speed of about 37% of the local free-stream speed. The vortices play an important role in the generation of turbulence and thus to decide the transitional length, which becomes time dependent. The source of the pressure fluctuations on the rear part of the suction surface is also identified as the formation of these coherent structures. When compared with experiments, it reveals that LES is worth pursuing as an understanding of the eddy motions and interactions is of vital importance for the problem.


Author(s):  
Ralph J. Volino

Boundary layer separation control with pulsed vortex generator jets (VGJs) has been studied on a very high lift, low-pressure turbine airfoil in the presence of unsteady wakes. Experiments were done under low (0.6%) and high (4%) freestream turbulence conditions on a linear cascade in a low speed wind tunnel. Cases were considered at Reynolds numbers (based on the suction surface length and the nominal exit velocity from the cascade) of 25,000 and 50,000. Wakes were produced from moving rods upstream of the cascade with flow coefficient 1.13 and rod spacing equal 2 blade pitches, resulting in a dimensionless wake passing frequency F = fLj-te/Uave = 0.14, where f is the frequency, Lj-te is the length of the adverse pressure gradient region on the suction surface, and Uave is the average freestream velocity. The VGJs were injected at the beginning of the adverse pressure gradient region on the suction surface with maximum jet velocity in each pulse equal to the local freestream velocity and a jet duty cycle of 10%. Several different timings of the VGJs with respect to the wakes were considered. Pressure surveys on the airfoil surface and downstream total pressure loss surveys were documented. Instantaneous velocity profile measurements were acquired in the suction surface boundary layer and downstream of the cascade. In cases without VGJs, the boundary layer momentarily reattached in response to the wake passing, but separated between wakes. The VGJs also caused reattachment, and if the VGJ pulsing frequency was sufficiently high, separation was largely suppressed for the full wake passing cycle. The timing of the VGJs with respect to the wakes was not very important. The jet pulsing frequency needed for separation control was about the same as found previously in cases without wakes. The background freestream turbulence effect was negligible in the presence of the larger wake and VGJ disturbances.


Author(s):  
Syed Anjum Haider Rizvi ◽  
Joseph Mathew

At off-design conditions, when the blade Reynolds number is low, a significant part of the blade boundary layer can be transitional. Then, standard RANS models are unable to predict the flows correctly but explicit transition modeling provides some improvement. Since large eddy simulations (LES) are improvements on RANS, the performance of LES was examined by simulating a flow through a linear, compressor cascade for which experimental data are available — specifically at the Reynolds number of 210,000 based on blade chord when transition processes occur over a significant extent of the suction surface. The LES were performed with an explicit filtering approach, applying a low-pass filter to achieve sub-grid-scale modeling. Explicit 8th-order difference formulas were used to obtain high resolution spatial derivative terms. An O-grid was wrapped around the blade with suitable clustering for the boundary layer and regions of large changes along the blade. Turbulent in-flow was provided from a precursor simulation of homogeneous, isotropic turbulence. Two LES and a DNS were performed. The second LES refines the grid in the vicinity of the separation bubble on the suction surface, and along the span. Surface pressure distributions from all simulations agree closely with experiment, thus providing a much better prediction than even transition-sensitive RANS computations. Wall normal profiles of axial velocity and fluctuations also agree closely with experiment. Differences between LES and DNS are small, but the refined grid LES is closer to the DNS almost everywhere. This monotonic convergence, expected of the LES method used, demonstrates its reliability. The pressure surface undergoes transition almost immediately downstream of the leading edge. On the suction surface there are streaks as expected for freestream-turbulence-induced transition, but spots do not appear. Instead, a separating shear layer rolls up and breaks down to turbulence at re-attachment. Both LES capture this process. Skin friction distribution reveals the transition near the re-attachment to occur over an extended region, and subsequent relaxation is slower in the LES. The narrower transition zone in the DNS is indicative of the essential role of smaller scales during transition that should not be neglected in LES. Simulation data also reveal that an assumption of laminar kinetic energy transition models that Reynolds shear stress remains small in the pre-transitional region is supported. The remaining differences in the predictions of such models is thus likely to be the separation-induced transition which preempts the spot formation.


Author(s):  
Jeffrey P. Bons ◽  
Rolf Sondergaard ◽  
Richard B. Rivir

The effects of pulsed vortex generator jets on a naturally separating low pressure turbine boundary layer have been investigated experimentally. Blade Reynolds numbers in the linear turbine cascade match those for high altitude aircraft engines and industrial turbine engines with elevated turbine inlet temperatures. The vortex generator jets (30 degree pitch and 90 degree skew angle) are pulsed over a wide range of frequency at constant amplitude and selected duty cycles. The resulting wake loss coefficient vs. pulsing frequency data add to previously presented work by the authors documenting the loss dependency on amplitude and duty cycle. As in the previous studies, vortex generator jets are shown to be highly effective in controlling laminar boundary layer separation. This is found to be true at dimensionless forcing frequencies (F+) well below unity and with low (10%) duty cycles. This unexpected low frequency effectiveness is due to the relatively long relaxation time of the boundary layer as it resumes its separated state. Extensive phase-locked velocity measurements taken in the blade wake at an F+ of 0.01 with 50% duty cycle (a condition at which the flow is essentially quasi-steady) document the ejection of bound vorticity associated with a low momentum fluid packet at the beginning of each jet pulse. Once this initial fluid event has swept down the suction surface of the blade, a reduced wake signature indicates the presence of an attached boundary layer until just after the jet termination. The boundary layer subsequently relaxes back to its naturally separated state. This relaxation occurs on a timescale which is 5–6 times longer than the original attachment due to the starting vortex. Phase-locked boundary layer measurements taken at various stations along the blade chord illustrate this slow relaxation phenomenon. This behavior suggests that some economy of jet flow may be possible by optimizing the pulse duty cycle and frequency for a particular application. At higher pulsing frequencies, for which the flow is fully dynamic, the boundary layer is dominated by periodic shedding and separation bubble migration, never recovering its fully separated (uncontrolled) state.


Author(s):  
Debasish Biswas ◽  
Tomohiko Jimbo

Unsteady transonic flows in diffuser have become increasingly important, because of its application in new propulsion systems. In the development of supersonic inlet, air breathing propulsion systems of aircraft and missiles, detail investigations of these types of flow behavior are very much essential. In these propulsion systems, naturally present self-sustaining oscillations, believed to be equivalent to dynamically distorted flow fields in operational inlets, were found under all operating conditions. The investigations are also relevant to pressure oscillations known to occur in ramjet inlets in response to combustor instabilities. The unsteady aspects of these flows are important because the appearance of undesirable fluctuations generally impose limitation on the inlet performance. Test results of ramjet propulsion systems have shown undesirable high amplitude pressure fluctuations caused by the combustion instability. The pressure fluctuations originated from the combustor extend forward into the inlet and interact with the diffuser flow-field. Depending on different parameters such as the diffuser geometry, the inlet/exit pressure ratio, the flow Mach number, different complicated phenomena may occur. The most important characteristics are the occurrence of shock induced separation, the length of separation region downstream of the shock location, and the oscillation of shock location as well as the oscillation of the whole downstream flow. Sajben experimentally investigated in detail the time mean and unsteady flow characteristics of supercritical transonic diffuser as a function of flow Mach number upstream the shock location and diffuser length. The flows exhibited features similar to those in supersonic inlets of air-breathing propulsion systems of aircraft. A High-order LES turbulence model developed by the author is assessed with experimental data of Sajben on the self-excited shock oscillation phenomena. The whole diffuser model configuration including the suction slot located at certain axial location around the bottom and side walls to remove boundary layer, are included in the present computation model. The time-mean and unsteady flow characteristics in this transonic diffuser as a function of flow Mach number and diffuser length are investigated in detail. The results of study showed that in the case of shock-induced separation flow, the length and thickness of the reverse flow region of the separation-bubble change, as the shock passed through its cycle. The instabilities in the separated layer, the shock /boundary layer interaction, the dynamics of entrainment in the separation bubble, and the interaction of the travelling pressure wave with the pressure fluctuation region caused by the step-like structure of the suction slot play very important role in the shock-oscillation frequency.


Sign in / Sign up

Export Citation Format

Share Document