An Investigation of the Application of Roughness Elements to Enhance Heat Transfer in an Impingement Cooling System

The inclusion of roughness elements on the target surface of a turbine aerofoil impingement cooling system is an attractive means of heat transfer enhancement. In such a system, it is important to minimise additional pressure loss caused by the roughness elements and thus their shape, size and position need to be optimised. The research showed how heat transfer enhancement is normally achieved at the expense of extra pressure loss. A hexagonal roughness element designed by the authors showed up to 10% heat transfer enhancement with minimal extra pressure loss. The present work includes shear pattern visualisation on the target surface, pressure loss measurements and heat transfer coefficient measurements for an impingement cooling system with simply shaped roughness elements-specifically cylindrical & diamond pimples. Flow visualisation results and pressure loss measurements for the above configurations provided criteria for selecting the shape, size and position of the roughness elements. The detailed heat transfer measurements on the target surface and over the roughness elements were used to explain the heat transfer enhancement mechanisms. It was found that the largest contribution to heat transfer is the impingement stagnation point and the developing wall jet regions. However, the research showed that the low heat transfer coefficient region could be made to contribute more by using strategically located roughness elements. A hexagonal rim was designed to cover the complete low heat transfer coefficient region midway between neighbouring jets. The effect of the height, cross sectional shape and wall angle of the hexagonal rim were studied using a series of heat transfer and pressure loss experiments. The transient heat transfer tests were conducted using a triple thermochromic liquid crystal technique and the thermal transient was produced by a fine wire mesh heater. The heat transfer coefficient over the pimples was measured using a hybrid transient method that analysed the thermal transient of the copper pimple. The detailed heat transfer coefficient distributions over the complete area of the target surface provided comprehensive understanding of the performance of the hexagonal rim. Tests were conducted at three different mass flow rates for each configuration. The average and local jet Reynolds numbers varied between 21500 and 31500, and 17000 and 41000 respectively.