| Engineers constantly seek advancements in the performance of aircraft and power generation engines, including, lower costs and emissions, and improved fuel efficiency. Nickel-base superalloys are the material of choice for turbine discs, which experience some of the highest temperatures and stresses in the engine. Engine performance is proportional to operating temperatures. Consequently, the high-temperature capabilities of disc materials limit the performance of gas-turbine engines. Therefore, any improvements to engine performance necessitate improved alloy performance.; In order to take advantage of improvements in high-temperature capabilities through tailoring of alloy microstructure, the overall objectives of this work were to establish relationships between alloy processing and microstructure, and between microstructure and mechanical properties. In addition, the projected aimed to demonstrate the applicability of neural network modeling to the field of Ni-base disc alloy development and behavior.; The first phase of this work addressed the issue of how microstructure varies with heat treatment and by what mechanisms these structures are formed. Further it considered how superalloy composition could account for microstructural variations from the same heat treatment. To study this, four next-generation Ni-base disc alloys were subjected to various controlled heat-treatments and the resulting microstructures were then quantified. These quantitative results were correlated to chemistry and processing, including solution temperature, cooling rate, and intermediate hold temperature.; A complex interaction of processing steps and chemistry was found to contribute to all features measured; grain size, precipitate distribution, grain boundary serrations. Solution temperature, above a certain threshold, and cooling rate controlled grain size, while cooling rate and intermediate hold temperature controlled precipitate formation and grain boundary serrations. Diffusion, both intergranular and grain boundary, was identified as the most pertinent mechanism. Variations in chemistry between alloys created different amounts of gamma/gamma' misfit strain, which affected precipitate size and morphology.; Next the question of how a disc alloy with differing microstructures would respond to constant or cyclic stresses as a function of time was addressed. To this end, mechanical testing at elevated temperatures was conducted, including tensile, hardness, creep deformation, creep crack growth and fatigue crack growth. Overall, mechanical properties were primarily related to the cooling rate during processing with hold temperatures being secondary. Whether the impact was positive or negative depended on the behavior under consideration. Fast cooling rates improved yield strength and creep resistance, but were detrimental to creep crack growth rates. The ability of precipitate particles to impede dislocation motion was the most frequently cited mechanism behind structure-property interaction.; Neural network models were successfully generated for processing-structure predictions, as well as for structure-property predictions. Training data was limited, none-the-less models were able to predict outputs with minimal relative errors. This was achieved through careful balance between the number of inputs and amount of training data. Despite the demonstrated correlation between microstructure and yield strength, microstructural quantities did not need to be directly inputted. Neural networks were sufficiently sensitive as to infer these effects from processing and chemistry inputs. This result improves the efficiency of this technique, while also demonstrating the capability of neural network techniques.; A full program of heat-treatment, microstructure quantification, mechanical testing, and neural network modeling was successfully applied to next generation Ni-base disc alloys. From this work the mechanisms of processing-structure and structure... |
