High-throughput Experimental and Computational Investigations of Metallic Glass Structure and Glass Forming Ability

Despite intense interest, identifying the structural origin of glass forming ability in metallic alloys remains a challenge due to the difficulty of describing the evolution of the long-range disordered structure from the liquid. In this thesis, we integrate high-throughput experimental methods with computational simulations to study glass formation and the resulting mechanical properties, with a primary focus on the Al-Ni-Zr system. Based on our investigation of the structural and cluster evolution using molecular dynamics simulations, we report the variance of the fraction of different types of atomic clusters in the liquid as a potential parameter to predict glass formation. The predictive power of the variance in the liquid state was verified by comparison with alloy libraries synthesized by a highly efficient laser deposition technique. Experimentally, glass formation was found over a wide compositional range centered on Al21.4Ni23.9Zr54.7, which is in excellent agreement with the simulations. Because the variance of cluster fractions at temperatures above the crystallization temperature is independent of quench rate as well as any particular cluster type, we believe this method could be extended to any alloy system, including those of higher complexity. Building upon this work, we examine the fundamental factors that determine the distribution and volume fraction of the crystal nucleation in simulated Al20Ni60Zr20 metallic glass/crystalline composites. The results show that the initial distribution of the atoms does not contribute to the final faction of atoms that form BCC-coordinated crystals in the composite. However, one major factor that affects the crystalline fraction is the temperature at which the stable nuclei form. The stability of Al-centered \u3c0, 3, 6, 4\u3e clusters also plays an important role in the final percentage of the ordered atoms. Finally, nanoindentation was performed to identify trends in hardness and indentation modulus with composition. The relationship between cluster structure and the observed mechanical behavior was evaluated by molecular dynamic simulation in Al-Ni-Zr system. By addressing the local mechanical property-cluster structure-glass forming ability relationship in this system, this study expands the understanding of the relationship of atomic structure, macroscopic mechanical behavior and glass forming ability