INVESTIGATIONS ON THE HEAT TRANSFER AND FLOW CHARACTERISTICS IN A TRAPEZOID DUCT FOR TURBINE BLADE LEADING EDGE

To evaluate the application of the impingement cooling in a trapezoidal duct, particularly the influence on internal cooling of the cross flow and swirl flow. Experimental and numerical studies have been performed. The experiment focuses on the heat transfer characteristics in the duct, when the numerical simulation focuses on the flow characteristics. Four Reynolds numbers (10000, 20000, 30000 and 40000), six cross flow mass flow ratios (0, 0.1, 0.2, 0.3, 0.4 and 0.5) and two impingement angle(35 and 45) are considered in both the experiment and the numerical simulation. The temperature on the target wall and the exit side wall is measured by the thermocouples, when the realizable k-epsilon turbulence model and enhanced wall treatment are performed using a commercial code Fluent. The results show that only part of the jets contribute in the heat transfer enhancement on the target wall, the other jets improve a large anticlockwise vortex occupied the upper part of the duct and drive strong swirl flow. The heat transfer on the exit side wall is enhanced by the swirl flow. The cross flow is induced in the duct by the outflow of the end exit hole. It deflects the jets and abates the impingement cooling on the target wall in the downstream region but has no evidently effect on the heat transfer on the exit side wall. Higher impingement angle helps to augment the impingement cooling on the target wall and improves the resistance ability of the jets against the effect of the cross flow. The heat transfer enhancement ability on the target wall and exit side wall in the present duct is compared to that of a smooth duct. The Nusselt number of the former is about 3 times higher than that of the latter. It indicates that the impingement and swirl play equally important roles in the heat transfer enhancement in the present duct. Empirical dimensionless correlations based on the present experiment data are presented in the paper.