The role of absolute pressure in deposition testing is reviewed from first principles. Relevant dimensionless parameters for deposition testing are developed and dynamic similarity conditions are assessed in detail. Criteria for establishing appropriate conditions for laboratory studies of deposition are established pursuant to the similarity variables. The role of pressure is particularly singled out for consideration relative to other variables such as temperature, particle size, and test article geometry/scaling. A case study is presented for deposition in a generic array of impinging jets. A fixed quantity (2g) of 0–10micron Arizona Road Dust (ARD) is delivered to the impingement array at three different temperatures (290, 500, and 725K) and at fixed pressure ratio. Deposition results are presented for operating pressures from 1 to 15atm. Surface scans show that the height of deposit cones at the impingement sites decreases with increasing pressure at constant temperature and pressure ratio. This reduction is explained by the lower “effective” Stokes number that occurs at high particle Reynolds numbers, yielding fewer particle impacts at high pressure. A companion CFD study identifies the additional role of Reynolds number in both the impingement hole losses as well as the shear layer thickness.

A new turbine cascade has been constructed that is designed to investigate the performance of actual nozzle guide vane hardware at temperatures representative of modern gas turbine engines. The facility is designed to investigate internal and external deposition, analyze the effectiveness of new cooling techniques, characterize material systems such as metal substrates or coatings, and assess the aerodynamic performance of a vane. The results presented here are the first results obtained in this new facility. External deposition on cooled CFM56 nozzle guide vanes has been explored at inlet temperatures of 1090° C, 1265° C, and 1350° C. Results at 1090° C have been compared to similar results in a previous facility. External deposition tests at temperatures greater than 1100° C on actual turbine hardware have not been reported publicly prior to this paper. These results show that deposition is concentrated at the stagnation line at all three inlet conditions. The amount of deposition on the vane pressure surface increased with increasing inlet temperatures.

The role of absolute pressure in deposition testing is reviewed from first principles. Relevant dimensionless parameters for deposition testing are developed and dynamic similarity conditions are assessed in detail. Criteria for establishing appropriate conditions for laboratory studies of deposition are established pursuant to the similarity variables. The role of pressure is particularly singled out for consideration relative to other variables such as temperature, particle size, and test article geometry/scaling. A case study is presented for deposition in a generic array of impinging jets. A fixed quantity (2 g) of 0–10 μ Arizona road dust (ARD) is delivered to the impingement array at three different temperatures (290, 500, and 725 K) and at fixed pressure ratio. Deposition results are presented for operating pressures from 1 to 15 atm. Surface scans show that the height of deposit cones at the impingement sites decreases with increasing pressure at constant temperature and pressure ratio. This reduction is explained by the lower “effective” Stokes number that occurs at high particle Reynolds numbers, yielding fewer particle impacts at high pressure. A companion computational fluid dynamics (CFD) study identifies the additional role of Reynolds number in both the impingement hole losses and the shear layer thickness.

An experimental and numerical investigation is conducted to assess the fluid dynamic mechanisms of control by vortex-generator jets for shock-induced separation in a highly loaded low pressure turbine (LPT) blade. Two- and three-dimensional steady RANS computations are performed to evaluate their ability to reproduce the main features of such a complex flow. The test blade is part of a compressible LPT cascade that exhibits shock-induced separation at an exit Mach number of 0.8. Active flow control is implemented through a spanwise row of discrete vortex-generator jets (VGJs) located on the suction surface. The control performance of VGJs in these transonic conditions has an optimum blowing ratio beyond which losses increase. Three-dimensionalities in the flow field are established by discrete VGJ-boundary layer interaction as suggested by Particle-Image Velocimetry (PIV) acquisitions at different spanwise locations. Blade pressure distributions and wake total pressure losses are acquired to evaluate the control performance and compared with calculations. Two-dimensional numerical investigations by RANS simulations suggest that the effect of increased expansion over the passage is a product of massflow injection only. Three-dimensional RANS results are interrogated to give a more detailed representation of the flow features around the jets, such as the jet vortex dynamics and spanwise modulation of the potential field. The analysis of this experimental and numerical information identifies the mechanisms contributing to the performance of skewed jets for control of shock induced separation in a highly loaded LPT blade. The results provide indications on the accuracy of RANS simulations, identifying the challenges of using RANS (2D or 3D) to solve such complex flows.

No abstract