Wall Interference in the Discharge Flow in a Linear Cascade Wind Tunnel

Linear nozzle cascade flows in a free jet tunnel may exhibit a significant wall-boundary interference at the edge of the test section. Supersonic discharge tests at the University of Leicester have shown that spurious expansions are generated at the free jet boundaries that over-expand the discharge flow. Numerical modelling at the University of Rome has detailed the generation and propagation of these disturbances from the jet boundaries to the inner test section and has been used to guide the design of a perforated wall, to rectify the flow. Tests have been conducted in a linear cascade of Rolls-Royce T2 profile blades at a discharge isentropic Mach number of 1.27. Three tunnel end wall geometries have been tested: (i) a perforated plate with evenly distributed round holes, (ii) a plate with wide slots in the streamwise direction and (iii) a plate with narrow slots in the streamwise direction. All geometries have about the same 50% geometric porosity. Spark schlieren flow visualisation and simultaneous side wall pressure measurements were used to document the discharge and to compare it against the baseline flow of a test with open boundaries (free jet). Experimental and numerical results highlight significant end wall interference effects in test cases (i) to (iii). These cause a lack of pitchwise periodicity in the flow. The discharge flow from the slotted board is reasonably well modelled by the numerical method in which a thick porous wall model is used. The predicted wall pressure distribution is in good quantitative agreement with the measurements. This has given confidence in using the numerical method as a design tool to improve the slotted board geometry. Specifically, numerical simulations have shown that a 14.3% porosity tailboard is able to minimise the wall interference, reinstating a sufficiently periodic flow inside the cascade.