ALBERT

Description: A comprehensive assessment is made of the predictive capability of the average passage flow model as applied to multi-stage axial flow compressors. The average passage flow model describes the time average flow field within a typical passage of a blade row embedded in a multi-stage configuration. In this work data taken within a four and one-half stage large low speed compressor will be used to assess the weakness and strengths of the predictive capabilities of the average passage flow model. The low speed compressor blading is of modern design and employs stators with end-bends. Measurements were made with slow and high response instrumentation. The high response measurements revealed the velocity components of both the rotor and stator wakes. Based on the measured wake profiles it will be argued that blade boundary layer transition is playing an important role in setting compressor performance. A model which mimics the effects of blade boundary layer transition within the frame work of the average passage model will be presented. Simulations which incorporated this model showed a dramatic improvement in agreement with data.

Description: Experiments were performed on a low-speed multistage axial-flow compressor to assess the effects of shrouded stator cavity flows on aerodynamic performance. Five configurations, which involved changes in seal-tooth leakage rates and/or elimination of the shrouded stator cavities, were tested. Data collected enabled differences in overall individual stage and the third stage blade element performance parameters to be compared. The results show conclusively that seal-tooth leakage ran have a large impact on compressor aerodynamic performance while the presence of the shrouded stator cavities alone seemed to have little influence. Overall performance data revealed that for every 1% increase in the seal-tooth clearance to blade-height ratio the pressure rise dropped up to 3% while efficiency was reduced by 1 to 1.5 points. These observed efficiency penalty slopes are comparable to those commonly reported for rotor and cantilevered stator tip clearance variations. Therefore, it appears that in order to correctly predict overall performance it is equally important to account for the effects of seal-tooth leakage as it is to include the influence of tip clearance flows. Third stage blade element performance data suggested that the performance degradation observed when leakage was increased was brought about in two distinct ways. First, increasing seal-tooth leakage directly spoiled the near hub performance of the stator row in which leakage occurred. Second, the altered stator exit now conditions caused by increased leakage impaired the performance of the next downstream stage by decreasing the work input of the downstream rotor and increasing total pressure loss of the downstream stator. These trends caused downstream stages to progressively perform worse. Other measurements were acquired to determine spatial and temporal flow field variations within the up-and-downstream shrouded stator cavities. Flow within the cavities involved low momentum fluid traveling primarily in the circumferential direction at about 40% of the hub wheel speed. Measurements indicated that the flow within both cavities was much more complex than first envisioned. A vortical flow structure in the meridional plane, similar to a driven cavity, existed within the upstream cavity Furthermore, other spatial and temporal variations in Row properties existed. the most prominent being caused by the upstream potential influence of the downstream blade. This influence caused the fluid within cavities near the leading edges of either stator blades in space or rotor blades in time to be driven radially inward relative to fluid near blade mid-pitch. This influence also produced large unsteady velocity fluctuations in the downstream cavity because of the passing of the downstream rotor blade.

Description: At the NASA Lewis Research Center, a comprehensive assessment was made of the predictive capability of the average passage flow model as applied to multistage axial-flow compressors. This model, which describes the time-averaged flow field within a typical passage of a blade row embedded in a multistage configuration, is being widely used throughout U.S. aircraft industry as an integral part of their design systems. Rotor flow-angle deviation. In this work, detailed data taken within a four and one-half stage large low-speed compressor were used to assess the weaknesses and strengths of the predictive capabilities of the average passage flow model. The low-speed compressor blading is of modern design and employs stator end-bends. Measurements were made with slow- and high response instrumentation. The high-response measurements revealed the velocity components of both the rotor and stator wakes. From the measured wake profiles, we found that the flow exiting the rotors deviated from the rotor exit metal angle to a lesser degree than was predicted by the average passage flow model. This was found to be due to blade boundary layer transition, which recently has been shown to exist on multistage axial compressor rotor and stator blades, but was not accounted for in the average passage model. Consequently, a model that mimics the effects of blade boundary layer transition, Shih k-epsilon model, was incorporated into the average passage model. Simulations that incorporated this transition model showed a dramatic improvement in agreement with data. The altered model thus improved predictive capability for multistage axial-flow compressors, and this was verified by detailed experimental measurement.

Description: Benchmark aerodynamic data are presented for compressible flow through a representative S-duct configuration. A numerical prediction of the S-duct flow field, obtained from a subsonic parabolized Navier-Stokes algorithm, is also shown. The experimental and numerical results are compared. Measurements of the three-dimensional velocity field, total pressures, and static pressures were obtained at five cross-sectional planes. Aerodynamic data were gathered with calibrated pneumatic probes. Surface static pressure and surface flow visualization data were also acquired. All reported tests were conducted with an inlet centerline Mach number of 0.6. The Reynolds number, based on the inlet centerline velocity and duct inlet diameter, was 2.6 x 10(exp 6). Thin inlet turbulent boundary layers existed. The collected data should be beneficial to aircraft inlet designers and the measurements are suitable for the validation of computational codes. The results show that a region of streamwise flow separation occurred within the duct. Details about the separated flow region, including mechanisms which drive this complicated flow phenomenon, are discussed. Results also indicate that the duct curvature induces strong pressure driven secondary flows. The cross flows evolve into counter-rotating vortices. These vortices convect low momentum fluid of the boundary layer toward the center of the duct, degrading both the uniformity and magnitude of the total pressure profile.

Description: Compressible, subsonic flow through a diffusing S-duct has been experimentally investigated. Benchmark aerodynamic data are presented for flow through a representative S-duct configuration. The collected data would be beneficial to aircraft inlet designers and is suitable for the validation of computational codes. Measurements of the 3D velocity field and total and static pressures were obtained at five cross-sectional planes. Surface static pressures and flow visualization also helped to reveal flow field characteristics. All reported tests were conducted with an inlet centerline Mach number of 0.6 and a Reynolds number, based on the inlet centerline velocity and duct inlet diameter, of 2.6 x 10(exp 6). The results show that a larger region of streamwise flow separation occurred within the duct. Details about the separated flow region, including mechanisms which drive this complicated flow phenomenon, are discussed. Transverse velocity components indicate that the duct curvature induces strong pressure driven secondary flows, which evolve into a large pair of counter-rotating vortices. These vortices convect the low momentum fluid of the boundary layer towards the center of the duct, degrading both the uniformity and magnitude of the total pressure profile.

Description: Full three-dimensional Navier-Stokes computational results are compared with new experimental measurements for the flowfield within a round diffusing S-duct. The present study extends previous computational and experimental results for a similar smaller scale S-duct. Predicted results are compared with the experimental static and total pressure fields, and velocity vectors. Additionally, wall pressures, velocity profiles in wall coordinates, and skin friction values are presented. The CFD results employ algebraic and k-epsilon turbulence models. The CFD computed and experimentally determined separated flowfield is carefully examined.

Description: Full three-dimensional Navier-Stokes computational results are compared with new experimental measurements for the flowfield within a round diffusing S-duct. The present study extends previous computational and experimental results for a similar smaller scale S-duct. Predicted results are compared with the experimental static and total pressure fields, and velocity vectors. Additionally, wall pressures, velocity profiles in wall coordinates, and skin friction values are presented. The CFD results employ algebraic and k-epsilon turbulence models. The CFD computed and experimentally determined separated flowfield is carefully examined.