Multi-Scale Methods and Applications of Ultrashort Laser Processing
This dissertation presents a study of the multiscale simulation methods and their applications in ultrashort laser shock peening (LSP) of metals. This study is motivated by the fact that microstructural evolution of near surface response of metals is very sensitive in ultrashort laser processes. More specifically, near surface deformation behavior is concentrated on dislocation movement and interaction among the metal nanoparticles. Six parts of work are presented in this study. In the first chapter, we present a modeling framework of femtosecond LSP. We implement atomistic-continuum based model in single crystal copper grain boundary structure. We replaced conventional Two Temperature Model (TTM) with Two Temperature Model-Molecular Dynamic (TTM-MD) simulations. For femtosecond LSP, initially electrons are heated by finite difference method in TTM part. Later on, electrons exchange heat to the lattice by electron-phonon coupling method. In the second chapter, we present a multiscale simulation method to study the microstructural responses of near-surface grain boundary structures of copper subjected to ultrashort femtosecond LSP. By integrating TTM with MD, we highlight the effects of laser process parameters on the near-surface response and corresponding phase change, formation of voids and their growth, and mechanism of dislocation nucleating and propagating from grain boundary. In the third chapter, we present a multiscale computational model of femtosecond LSP to calculate dislocation-gliding rate under laser peened surface. We use pressure profile from previous chapter as an input in this study. We implement crystal plasticity material model of single crystal copper in femtosecond LSP to understand dislocation deformation mechanism at the macroscale. Our investigation shows that generated dislocation is concentrated not only below laser peened surface but also propagated through the material. In the fourth chapter, we describe a multiscale computational framework for calculating residual stress during femtosecond LSP. This work is motivated by the advantages of femtosecond LSP without sacrificial overlay. At first, we use TTM-MD model to capture pressure profile. Calculated pressure profile is used as a surface load in the FEM framework. We implement Johnson-Cook material and damage model in the FEM scheme. A symmetry cell method is implemented to predict residual stress for a large scale model. In the fifth chapter, we model a framework of plasmonic gold nanoparticles under ultrafast pulsed laser heating. We couple Discrete Dipole Approximation (DDA) with atomistic simulation and details of the coupling mechanism are described in this chapter. By introducing the DDA, we calculate electric field ratio, which is then converted to laser heat source as an input to atomistic simulation framework.