Abstract
Film cooling has been widely applied to cool hot section components in modern aero-engines, typically turbine airfoils. Blowing cooling air via discrete holes from inside the airfoil realizes coolant film which protects the exposed surface against harsh gas stream. Film cooling holes are characterized in two types, namely, round and shaped. Shaped holes are designed to provide higher film effectiveness with marginal heat transfer enhancement compared to round holes, typically by diffusing and spreading the coolant jet with expanded exits.
A number of factors affect the film cooling performance: hole shape, flow rate parameter (e.g. blowing ratio), density ratio, freestream turbulence, etc. Among these, definitely the hole shape has dominant effects on the performance. Therefore, it has been one of the primary interests in the film cooling research to design/select/optimize the hole geometry, where various ideas, some relatively simple and others even more complicated, were proposed.
The primary aim and contribution of the present paper is to challenge the shaped hole multi-objective optimization on a realistic setup. A medium-loaded modern HPT cascade under transonic condition was employed, where the effects of surface curvature, pressure gradient, Mach number, etc. on the thermal and aerodynamic behavior of film cooling can be reproduced faithfully as in actual operating turbines. Heat transfer enhancement and cascade aerodynamic loss as well as film effectiveness were considered as objective functions. Film effectiveness and heat transfer coefficient were evaluated across almost the entire surface downstream because a shaped hole working well only near the hole often shows very limited overall thermal advantage, based on the authors’ design experience. Any hole design assuming the use of emerging new techniques as AM and/or other breakthrough in manufacturing capabilities in the future is completely out of the present scope. Restrictions in design parameters were carefully set for the resultant hole shape to be well manufacturable with conventional process.
Another important consideration in the present work is to generalize the observations as much as possible from the optimization results. The fundamental risk in optimization work is that whole the logical context led to the optimal solution is often very difficult to follow for designers. As the process being a black box, any minor redesign task needs another full set of optimization, which is not the right course to take, obviously. Even though the present work focused on one specific setup, a factorial effect analysis was carried out in order to make the observations generalized.
