Olga Kartuzova

Separation Control on a Very High Lift Low Pressure Turbine Airfoil Using Pulsed Vortex Generator Jets

Boundary layer separation control has been studied using vortex generator jets (VGJs) on a very high lift, low-pressure turbine airfoil. Experiments were done under high (4%) freestream turbulence conditions on a linear cascade in a low speed wind tunnel. 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. 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. Jet pulsing frequency, duty cycle, and blowing ratio were all varied. Computational results from a large eddy simulation of one case showed reattachment in agreement with the experiment. In cases without flow control, the boundary layer separated and did not reattach. With the VGJs, separation control was possible even at the lowest Reynolds number. Pulsed VGJs were more effective than steady jets. At sufficiently high pulsing frequencies, separation control was possible even with low jet velocities and low duty cycles. At lower frequencies, higher jet velocity was required, particularly at low Reynolds numbers. Effective separation control resulted in an increase in lift and a reduction in total pressure losses. Phase averaged velocity profiles and wavelet spectra of the velocity show the VGJ disturbance causes the boundary layer to reattach, but that it can reseparate between disturbances. When the disturbances occur at high enough frequency, the time available for separation is reduced, and the separation bubble remains closed at all times.

Experimental and Computational Investigations of Separation and Transition on a Highly Loaded Low-Pressure Turbine Airfoil: Part 1 — Low Freestream Turbulence Intensity

Boundary layer separation, transition and reattachment have been studied on a very high lift, low-pressure turbine airfoil. Experiments were done under high (4%) freestream turbulence conditions on a linear cascade in a low speed wind tunnel. Pressure surveys on the airfoil surface and downstream total pressure loss surveys were documented. Velocity profiles were acquired in the suction side boundary layer at several streamwise locations using hot-wire anemometry. Cases were considered at Reynolds numbers (based on the suction surface length and the nominal exit velocity from the cascade) ranging from 25,000 to 300,000. At the lowest Reynolds number the boundary layer separated and did not reattach, in spite of transition in the separated shear layer. At higher Reynolds numbers the boundary layer did reattach, and the separation bubble became smaller as Re increased. High freestream turbulence increased the thickness of the separated shear layer, resulting in a thinner separation bubble. This effect resulted in reattachment at intermediate Reynolds numbers, which was not observed at the same Re under low freestream turbulence conditions. Numerical simulations were performed using an unsteady Reynolds averaged Navier-Stokes (URANS) code with both a shear stress transport k-ω model and a 4 equation shear stress transport Transition model. Both models correctly predicted separation and reattachment (if it occurred) at all Reynolds numbers. The Transition model generally provided better quantitative results, correctly predicting velocities, pressure, and separation and transition locations. The model also correctly predicted the difference between high and low freestream turbulence cases.Copyright

Experimental and Computational Investigations of Low-Pressure Turbine Separation Control Using Vortex Generator Jets

Boundary layer separation control has been studied using vortex generator jets (VGJs) on a very high lift, low-pressure turbine airfoil. Experiments were done under low freestream turbulence conditions on a linear cascade in a low speed wind tunnel. Pressure surveys on the airfoil surface and downstream total pressure loss surveys were documented. Cases were considered at Reynolds numbers (based on the suction surface length and the nominal exit velocity from the cascade) of 25,000, 50,000 and 100,000. Jet pulsing frequency, duty cycle, and blowing ratio were all varied. In all cases without flow control, the boundary layer separated and did not reattach. With the VGJs, separation control was possible even at the lowest Reynolds number. Pulsed VGJs were more effective than steady jets. At sufficiently high pulsing frequencies, separation control was possible even with low jet velocities and low duty cycles. At lower frequencies, higher jet velocity was required, particularly at low Reynolds numbers. Effective separation control resulted in an increase in lift of up to 20% and a reduction in total pressure losses of up to 70%. Simulations of the flow using an unsteady RANS code with the four equation Transition-sst model produced good agreement with experiments in cases without flow control, correctly predicting separation, transition and reattachment. In cases with VGJs, however, the CFD did not predict the reattachment observed in the experiments.Copyright

LES Flow Control Simulations for Highly Loaded Low Pressure Turbine Airfoil (L1A) Using Pulsed Vortex Generator Jets

Seven different cases were examined experimentally and computationally to study LPT flow control using pulsed VGJs for the L1A airfoil. These cases represent a combination of variation in Reynolds number, Re (25,000, 50,000 and 100,000), based on the suction surface length and the nominal exit velocity from the cascade, blowing ratio, B (from 0.25 to 1), dimensionless frequency, F (from 0.035 to 0.56) and duty cycle, DC (10% and 50%). The data was obtained for the pressure distribution along the airfoil and downstream in the wake as well as for velocity profiles at six different stations downstream of the suction peak. The CFD was done with LES utilizing version 6.3.26 of the finite-volume code ANSYS Fluent. The CFD did provide further insight to better understand the physics of flow control. The comparison between CFD and experimental results for Cp, velocity profiles and Ψint is reasonable for all cases examined. Two of the cases examined did indicate that the higher DC could compensate for the lower F value. However, the effect of increasing the frequency appears to be stronger than increasing the DC value. The results from CFD using the Q-Criterion clearly illustrate how a separation bubble will persist at the lower frequency case (Case (2)) and the disturbances created from the jet flow do not have enough energy or time to travel further downstream to cause reattachment. On the other hand, the higher frequency case (Case (6)) did exhibit a penetration of the disturbance created by the jet into the separated region and caused reattachment, especially at the trailing edge. It appears that the jet was capable of breaking the large bubble into smaller ones with reattachment in between.Copyright

CFD Modeling of the Multipurpose Hydrogen Test Bed (MHTB) Self-Pressurization and Spray Bar Mixing Experiments in Normal Gravity: Effect of the Accommodation Coefficient on the Tank Pressure

A CFD model for simulating the self-pressurization of a large scale liquid hydrogen storage tank is utilized in this paper to model the MHTB self-pressurization experiment. The kinetics-based Schrage equation is used to account for the evaporative and condensi ng interfacial mass flows in this model. The effect of the accommodation coefficient for calculating the interfacial mass transfer rate on the tank pressure during tank selfpressurization is studied. The values of the accommodation coefficient which were considered in this study vary from 1.0e-3 to 1.0e-1 for the explicit VOF model and from 1.0e-4 to 1.0e-3 for the implicit VOF model. The ullage pressure evolutions are compared against experimental data. A CFD model for controlling pressure in cryogenic storage tanks by spraying cold liquid into the ullage is also presented. The Euler-Lagrange approach is utilized for tracking the spray droplets and for modeling the interaction between the droplets and the continuous phase (ullage). The spray model is coupled with the VOF model by performing particle tracking in the ullage, removing particles from the ullage when they reach the interface, and then adding their contributions to the liquid. Droplet-ullage heat and mass transfer are modeled. The flow, temperature, and interfacial mass flux, as well as droplets trajectories, size distribution and temperatures predicted by the model are presented. The ul lage pressure and vapor temperature evolutions are compared with experimental data obtained from the MHTB spray bar mixing experiment. The effect of the accommodation coefficient for calculating the interfacial and droplet mass transfer rates on the tank pressure during mixing of the vapor using spray is studied. The values used for the accommodation coefficient at the interface vary from 1.0e-5 to 1.0e-2. The droplet accommodation coefficient values vary from 2.0e-6 to 1.0e-4.

Modeling Ullage Dynamics of Tank Pressure Control Experiment during Jet Mixing in Microgravity

A CFD model for simulating the fluid dynamics of the jet induced mixing process is utilized in this paper to model the pressure control portion of the Tank Pressure Control Experiment (TPCE) in microgravity1. The Volume of Fluid (VOF) method is used for modeling the dynamics of the interface during mixing. The simulations were performed at a range of jet Weber numbers from non-penetrating to fully penetrating. Two different initial ullage positions were considered. The computational results for the jet-ullage interaction are compared with still images from the video of the experiment. A qualitative comparison shows that the CFD model was able to capture the main features of the interfacial dynamics, as well as the jet penetration of the ullage.

CFD Simulations of Unsteady Wakes on a Highly Loaded Low Pressure Turbine Airfoil (L1A)

A study of a very high lift, low-pressure turbine airfoil in the presence of unsteady wakes was performed computationally and compared against experimental results. The experiments were conducted in a low speed wind tunnel under high (4.9%) and then low (0.6%) freestream turbulence intensity conditions with a flow coefficient (ζ) of 0.7. The experiments were done on a linear cascade with wakes that were produced from moving rods upstream of the cascade with the rod to blade spacing varied from 1 to 1.6 to 2. In the present study two different Reynolds numbers (25,000 and 50,000, based on the suction surface length and the nominal exit velocity from the cascade) were considered.The experimental and computational data have shown that in cases without wakes, the boundary layer separated and did not reattach. The CFD was performed with Large Eddy Simulation (LES) and Unsteady Reynolds-Averaged Navier-Stokes (URANS), Transition-SST, utilizing the finite-volume code ANSYS FLUENT under the same freestream turbulence and Reynolds number conditions as the experiment but only at a rod to blade spacing of 1.With wakes, separation was largely suppressed, particularly if the wake passing frequency was sufficiently high. Similar effect was predicted by 3D CFD simulations. Computational results for the pressure coefficients and velocity profiles were in a reasonable agreement with experimental ones for all cases examined. The 2D CFD efforts failed to capture the three dimensionality effects of the wake and thus were less consistent with the experimental data.As a further computational study, cases were run to simulate higher wake passing frequencies which were not run experimentally. The results of these computational cases showed that an initial 25% increase from the experimental dimensionless wake passing frequency of F = 0.45 greatly reduced the size of the separation bubble, nearly completely suppressing it, however an additional 33% increase on top of this did not prove to have much of an effect.Copyright

Computational Simulation of Cylindrical Film Hole with Jet Pulsation on Flat Plates

Film cooling of flat plates with pulsation were simulated using FLUENT ™ commercial code with realizable k-e turbulence model. The simulations were done for nominal blowing ratios 0.5 and 1.5, duty cycle = 50%, and Strouhal number ranging from 0.0119 to 1.0. Pulsation helps to lower the amount of cool air from the compressor, which is desirable for film-cooling applications. Pulsed jets performance significantly depends on geometry and blowing ratio. From the cases studied and for steady flow with attached jets, pulsation considerably decreases the film-cooling effectiveness. On the other hand, for steady flow cases where jet liftoff occurs (e.g., higher blowing ratios), pulsation helps to increase the film-cooling effectiveness.

CFD Simulation of Jet Pulsation Effects on Film Cooling of Flat Plates

In this paper fourteen different CFD cases for CFH (cylindrical film hole) and LDIFF (film hole with laterally diffused exit) geometries were conducted to study the film cooling of flat plates. Those cases included different blowing ratios: 0.5, 1.25 and 1.5 and both steady flow and pulsed jets. In the jet pulsation cases the Duty Cycle was taken 50% and the Strouhal number ranged from 0.0119 to 1.0 for the CFH geometry and from 0.0119 to 0.38 for the LDIFF geometry. Fluent commercial code with realizable k–e turbulence model was used in this study to investigate how the pulsed jet performance was affected by varying: 1) pulsation frequency, 2) blowing ratio and 3) jet geometry. For the CFH geometry (B = 0.5) the pulsed jet showed lower film cooling effectiveness than the steady state for all cases examined. However, the frequency effects varied according to the downstream location from the jet exit. Immediately near the jet trailing edge the effectiveness increased as the frequency increased. Downstream (x/D above 3) the effectiveness for both St = 0.0119 and 0.38 almost agreed while lower effectiveness were noted for St = 0.19. For St = 1.0, the effectiveness was above the other frequencies (for all x/D values) but still below the steady state ones. As for the LDIFF geometry (B = 1.25) the effect of frequency was negligible and the pulsed jet showed lower film cooling effectiveness than the steady state. Two different blowing ratios (0.5 and 1.5) were examined for the CFH geometry. The pulsation had different effects in the two cases. At B = 0.5, lower effectiveness performance everywhere was obtained for pulsed cases compared to steady ones. For B = 1.5 pulsation results were highly dependent on the frequency. For low frequency (St = 0.0119) the effectiveness was below the steady state one for all x/D values. For higher frequency (St = 0.38) the effectiveness was higher than the steady state one for all x/D values. As for St = 0.19 and 1.0 the results were in between the above two frequencies. A spatially averaged effectiveness was developed to enable comparing “overall” performance of all cases examined. This was done by choosing an area downstream of the jet that covered from the jet “trailing edge”, x/D = 0 to x/D = 10 and also covered a 1/2 pitch on both sides of the jet in the spanwise direction. Using the defined spatially averaged effectiveness with 50% of the coolant (Duty Cycle) an overall reduction in the film cooling effectiveness was found to be: 52.73% for the LDIFF (B = 1.25), 38.12% for the CFH (B = 0.5) and an overall enhancement of 14.77% for the CFH (B = 1.5). Knowing that low effectiveness in case of B = 1.5, CFH geometry was caused by the jet “lift-off”, results above clearly indicate that jet pulsation will be more effective for cases with detached jet under steady state conditions. Although pulsation didn’t bring overall benefit to film cooling, there were cases where pulsed jets helped to increase effectiveness over the steady state conditions. Therefore, present results might be useful for evaluation of the effect of pulse frequency on film cooling effectiveness in real life applications, where jets pulse naturally due to the pressure fluctuations in the engine.Copyright