Experimental and computational investigation of inclined jets in a crossflow

Matthew J. Findlay
1998
The flow field characteristics of four different geometries of a row of square jets in a crossflow at velocity ratios relevant to gas turbine film cooling applications have been examined using experimental and computational methods. The geometries considered were: long and short entry length streamwise-inclined jets, spanwise-inclined jets, and compound-angle jets. Mean velocity and turbulence measurements were made using a three-component LDV system. Jet penetration, spreading, and film
more » ... effectiveness were measured using a flame ionization detector. Numerical simulations were performed using three different turbulence models: the standard κ — ε model, Menter's baseline blended κ — ε/κ — ω model, and Menter's shear stress transport model. The flow field at the jet exit is strongly influenced by the crossflow, as well as by the inlet conditions at the entrance to the jet orifice. At low velocity ratios the jets do not penetrate beyond the upstream boundary layer thickness. As the velocity ratio increases, the jet penetrates beyond the boundary layer resulting in stronger interaction with the crossflow. Considerable anisotropy of the turbulent flow field is observed. The film cooling effectiveness is best at the lowest velocity ratio as the jet is deflected strongly towards the floor of the wind tunnel, although the improvement is more significant for the streamwise injection case. At the highest velocity ratio the spanwise jets provide the best film cooling effectiveness but provide increased blockage to the crossflow. The results of the preliminary computational analysis indicate that the flow field produced by each of the geometries provides a serious challenge for numerical modelling. Mean velocity gradients and turbulence kinetic energy levels are typically underpredicted by the computations due to the assumption of equilibrium turbulence inherent in the models. The use of an isotropic eddy viscosity model must be reconsidered in light of the measured turbulence anisotropy and gen [...]
doi:10.14288/1.0080962 fatcat:3q5nuei5nral3nykvvvf5yrjwq