Mechanics and Mechanisms of Creep and Ductile Fracture

The main aim of this dissertation is to relate measurable and hopefully controllable features of a material's microstructure to its observed failure modes to provide a basis for designing better materials. The understanding of creep in materials used at high temperatures is of prime engineering importance. Single crystal Ni-based superalloys used in turbine aerofoils of jet engines are exposed to long dwell times at very high temperatures. In contrast to current theories, creep tests on Ni-based superalloy specimens have shown size dependent creep response termed as the thickness debit effect. To investigate the mechanism of the thickness debit effect, isothermal creep tests were performed on uncoated Ni-based single crystal superalloy sheet specimens with two thicknesses and under two test conditions: a low temperature high stress condition and a high temperature low stress condition. At the high temperature, surface oxidation induced microstructural changes near the free surface forming a layered microstructure. Finite element calculations showed that this layered microstructure gave rise to local changes in the stress state. The specimens also contained nonuniform distribution of initial voids formed during the solidification and homogenization processes. The experiments showed that porosity evolution could play a significant role in the thickness debit effect. This motivated a basic mechanics study of porosity evolution in single crystals subjected to creep for a range of stress states. The study was performed using three-dimensional finite deformation finite element analysis of unit cells containing a single initially spherical void in a single crystal matrix. The materials are characterized by a rate-dependent crystal plasticity constitutive relation accounting for both primary and secondary creep. The effect of initial void spacing and creep exponent was also explored. Based on the experimental observations and results of finite element calculations a quantitative mechanistic model is proposed that can account for both bulk and surface damage effects and assess their relative roles in the observed thickness debit effect. Another set of calculations aim at relating the crack growth resistance and fracture surface morphology to material microstructure for ductile structural metals. The process that governs the ductile fracture of structural materials at room temperature is one of nucleation, growth and coalescence of micron scale voids, and involves large plastic deformations. Experimental studies have shown that fracture surfaces in a wide variety of materials and under a wide variety of loading conditions have remarkable scaling properties. For thirty years, the hope to relate the statistical characterization of fracture surfaces to a measure of a material's crack growth resistance has remained unfulfilled. Only recently has the capability been developed to calculate sufficient amounts of three dimensional ductile crack growth in heterogeneous microstructures to obtain a statistical characterization of the predicted fracture surfaces. This development has enabled the exploration of the relation of both fracture toughness and fracture surface statistics to material properties and microstructure when the fracture mechanism is one of void nucleation, growth and coalescence. The relation of both toughness and the statistical properties of fracture surfaces in calculations of heterogeneous microstructures to various microstructural features is discussed and a remarkable correlation between fracture surface roughness and fracture toughness is shown for the first time.