The design of turbine cooling systems remains one of the most challenging processes in engine development. Modern turbine cooling systems indeed invariably combine internal convection cooling with external film cooling in complex flow systems. The heat transfer and cooling processes are at the limit of current understanding and engine designers heavily rely on empirical tools and engineering judgment to produce new designs. These designs are moreover developed in the context of continuously increasing Turbine Entry Temperature (TET) as the latter leads to improvement of Specific Fuel Consumption (SFC). The present contribution fits into the frame of the ongoing FP7 ER-ICKA project. It focuses on achieving a significant progress in understanding turbine blade passages internal cooling systems by gathering high quality experimental data and by developing cooling state-of-the-art design capabilities based upon computer codes calibrated through these experimental data.
In this context, the paper will describe the design optimisation and analysis work performed for two different internal cooling passages configurations, namely a static leading edge LP configuration passage (baseline experimentally tested at Stuttgart University) and a rotating mid-chord HP configuration passage (baseline experimentally tested at ONERA). The aim of the work was to develop a design methodology to optimise turbulence promoting ribs shape and arraying to improve the thermal behaviour of the internal cooling passages while avoiding excessive head loss. The optimisation was driven using decoupled rib design parameters for each ribbed wall to enhance flow interactions and maximise disturbances, to maximise potential increase in Heat Transfer Coefficients (HTCs). Any improvement in the thermal behaviour of the cooling system may indeed allow to either reduce the coolant mass flow rate requirements or increase the TET.
To drive these optimisations, the ultimate target was hence to reduce the maximum blade metal temperature. To this end, suitable cost functions (objectives and constraints) have been derived and implemented. They will first be presented and discussed along with the parameterisations, so as to define the complete optimisation specification. The computational chain setup, among which the challenging mesh regeneration choices set based on a mesh dependence study will then be detailed. Validation of the CFD evaluation against the experimental results will be described for the static baseline configuration at least (rotating test measurements are still ongoing) and the optimisation results, which have led to significant gains in HTCs, will finally be analysed, data mining techniques allow to identify key parameters, path taken in the conception space and major trends.