Date of Award

Winter 2024

Project Type

Dissertation

Program or Major

Mechanical Engineering

Degree Name

Doctor of Philosophy

First Advisor

Marko M Knezevic

Second Advisor

Brad B Kinsey

Third Advisor

Igor I Tsukrov

Abstract

With the market expansions and increasing number of scientific explorations, the demands for manufacturing metallic parts to satisfy specific requirements continue to increase. A slight improvement in the manufacturing process can result in great cost savings in time and energy. The research and development time to identify these areas of improvements can be greatly reduced using numerical models. One such model is the mean-field elastoplastic self-consistent (EPSC) framework studied extensively and advanced in this dissertation. In the past two decades, many material behavior formulations were implemented in EPSC, and the model has shown versatility and accuracy in predicting a wide range of experimental observations. EPSC is computationally efficient, allowing it to be coupled with finite element analysis to predict larger scale deformations. However, the advantage of EPSC is also its shortcoming. EPSC approximates material properties as averages and cannot predict the effects of spatial fluctuations, limiting EPSC’s ability to describe the behaviors on a more physical level. In this dissertation, EPSC is first advanced to describe the martensitic phase transformation with physics-based dislocation mechanics model. Then, EPSC is extended to incorporate field fluctuations (FF) formulations. The resulting model is named FF-EPSC and calculates the second moments of stress, strain, lattice spin, and misorientations to give an approximation of the intragranular spread of material properties. The second moments enable both a grain fragmentation and a static recrystallization model. FF-EPSC is first applied to AA5182 and IF-steel to predict the grain fragmentation and recrystallization textures. Then, FF-EPSC is applied to simulate a continuous process of rolling, recrystallization, and cup drawing of AA6022. Finally, based on the second moment formulations, EPSC is further advanced with a procedure to introduce spatial relationship to rotation spreads. The spatial distribution enables EPSC to calculate geometrically necessary dislocations (GND) and the final model is therefore named strain gradient (SG)-EPSC. SG-EPSC is first applied to model mechanical responses and texture evolution of α-Ti and then applied to compare with measured GND evolution of AA6016 under three strain paths.

Download

Share

COinS