Particle rebound characteristics of turbomachinery cascade leading edge geometry

The objective of this research work is to investigate and understand the complex phenomena associated with the mechanism of particle impacts on turbomachinery cascade leading edge geometry. At present, there is a need for experimental work in basic and applied research to find out the parameters that are relevant to particle rebound characteristics on turbomachinery blades. In the present work, experiments were conducted with air velocity at 15 m/s (∼50 ft/sec) and at 30 m/s (∼100 ft/sec) using high-speed photography and Laser Doppler Velocimetry (LDV). Silica sand particles of 1000–1500 micron size were used for this study. In the present investigation, particle rebound data was obtained for cylindrical targets with radius of curvature representative of leading edge geometry (cylinder diameter = 4.5mm & 6.5 mm) using LDV. The numerical simulations, which are based on non-linear dynamic analysis, were also performed using the finite element code DYNA3-D. Several different material models viz elastic-elastic, elastic-plastic, elastic-plastic with friction & isotropic-elastic-plastic with dynamic friction and particle rotation were used in the DYNA3-D numerical analysis. The computational results include a time history of the displacement, stress and strain profiles through the particle collision. Numerical results are presented for the rebound conditions of spherical silica sand particle for different pre-collision velocities. The computed particle restitution coefficients, after they reach steady rebound conditions, are compared with experimental results obtained from LDV. A probabilistic model was developed to incorporate the uncertainties in the impact velocity in the numerical model. Histograms and Cumulative Distribution Functions (CDFs) for impact velocity were obtained from experimental LDV data. Ten randomly selected probabilities for each impact angle were used to calculate the impact velocity from cumulative distribution function. This randomly selected impact velocity was used as input to the DYNA3-D numerical simulation instead of a deterministic impact velocity. From these random simulations, mean values, for rebound velocity and velocity restitution ratio, were obtained. These results are compared with the results obtained from deterministic model.; Von-Mises effective stress and shear stress for multi-particle impacts are studied in detail. A DYNA3-D impact model in which several silica sand particles impact the cylindrical target was developed. The impacts were simulated in such way that the sand particles impact the cylindrical target in phase, out of phase, in the same longitudinal plane & different longitudinal plane. Maxima of effective stress (v-m) & shear stress with respect to time are plotted.; Particle rebound characteristics on leading edge geometry were obtained using two-dimensional LDV measurements and high-speed photography. A Wollensax Fastax camera was used to film the particulate flow. In the current investigation a camera speed of 2500 frames/sec was used. The developed film from high-speed photography was digitized. The data were processed using x-v image processing software. The restitution ratios of the particle collisions on pressure and suction surface of the blade leading edge are obtained for different impact angles at the two different impact velocities (15 m/s & 30 m/s).; A simple mechanical model was also developed to study inter-particle collisions. Results obtained from the model are compared with the results obtained from high-speed photography and DYNA3-D.; Blade surface erosion rate was calculated using an empirical model and is compared against the values obtained from cold tunnel experiments.