High-Fidelity Computational Analysis of the Aerothermal Performance of In-serviced Jet Engine Blades

In today’s civil aviation, the struggle for higher jet engine efficiencies has pushed the manufactures into the continuous challenge of developing new and better design and optimization strategies. In the information age, it is only natural that a great deal of this effort is going to be carried out by means of computational analysis. That is to say, these design and optimization strategies rely heavily on the use of computational models, and thus the search for a better design is hinged upon the search for a better model. A notable product of this search is the “robust design” philosophy, which aims to consider the variability in geometry and operating conditions that every component will invariably experience in real-world conditions. In general, the key element in this evolution process is for the model to be capable of accounting for more and more aspects of the reality of the problem at hand, while still being affordable in terms of computational costs. In this case, the problem is represented by the aero-thermal behavior of the jet engine’s most characteristic components: the blades. As mentioned above, to increase the fidelity of the model, key aspects that characterize the real operation of these components can be included in it, beginning with the geometry. While most of the computational performance analysis is conducted on nominal designs, it is important to consider that, during most of their service life, the turbine blades are going to operate with a geometry that is increasingly affected by deviation from nominal. This is due to both manufacturing variation and in-service damage. These geometric deviations can be measured on the blades after an engine overhaul, providing highly useful information on the damage modes of the engine. By digitalizing these geometries, engineers can quantify and parametrize the geometric deviation. Furthermore, by creating computational grids around these geometries, a high-fidelity CFD study revolving around the performance of the real blades can be carried out, shedding light on the correlation between the geometric deviation parameters and aerodynamic performance loss. Naturally, this geometric deviation also has a significant impact on the thermal behavior of the blades, affecting the distribution of the Heat Transfer Coefficient (HTC) over the blades’ surfaces. Even when modelling the nominal case, it is often common practice to use a simplified version of the geometry, where the internal cooling system is replaced with source terms. Although this reduces the costs of the CFD simulations, it obviously subtracts from the model’s accuracy. Furthermore, it is particularly important to model the fluid-solid thermal exchange, and the rotor-stator unsteady interaction. All these fidelity-related aspects that can impact the model’s accuracy are investigated in the present work