Graduate Research Assistant , University of Utah ,United States
Donovan Birky is a PhD candidate in Mechanical Engineering at the University of Utah, with a passion for applying computational techniques and machine learning to solve complex material modeling problems. His expertise spans finite element simulation, machine learning for constitutive model development, uncertainty quantification, and high-performance computing. He has contributed to innovative research in areas like plasticity modeling, microstructure-sensitive damage models, and uncertainty analysis. Birky has collaborated with prominent institutions like Sandia National Labs, where he worked on material damage models and neural networks for material science applications. His work has resulted in multiple impactful publications in renowned journals and conferences. Outside of research, Birky has excelled academically, earning recognition such as the Rhode Baker Most Outstanding Student award and multiple placements on the Dean’s List. His interdisciplinary work bridges material science, engineering, and machine learning, with a focus on real-world applications in industry.
Profile
Education
Donovan Birky is currently pursuing a PhD in Mechanical Engineering at the University of Utah, where he has maintained a GPA of 3.93/4.00 since 2020. In Fall 2022, he received his Master of Science degree in Mechanical Engineering from the same institution. As part of his doctoral research, Birky is a member of the Materials Prognosis from Integrated Modeling and Experiment (M’) Lab, contributing to the development of advanced material damage models using machine learning. Prior to his graduate studies, Birky earned a Bachelor of Science in General Engineering from Fort Lewis College, Durango, CO, graduating with a GPA of 3.92/4.00. During his time at Fort Lewis, he was awarded the Rhode Baker Most Outstanding Student award for Physics and Engineering and was consistently on the Dean’s List. Birky also contributed as a STEM tutor for the TRIO Student Success Center, further demonstrating his dedication to academic excellence.
Experience
Donovan Birky has significant research experience as a Graduate Research Assistant at the University of Utah since 2020, working in the Materials Prognosis from Integrated Modeling and Experiment (M’) Lab. His research focuses on the integration of machine learning with material modeling to develop damage models for advanced materials. His previous internships at Sandia National Laboratories have further refined his skills in computational mechanics and material science. In 2021 and 2022, Birky worked on projects developing grain growth and pore growth datasets, using Sierra finite element codes for machine learning applications. He also contributed to the development of a material model calibration tool. Birky’s research at Fort Lewis College, funded by Sandia National Labs, centered on applying AI and machine learning to solve structural dynamics problems. His work at both academic and research institutions showcases his ability to tackle complex engineering challenges using cutting-edge computational techniques and high-performance computing.
Research Focus
Donovan Birky’s research is centered around computational material modeling and machine learning, with a particular focus on improving the accuracy and interpretability of constitutive models used in material science. His work integrates advanced techniques such as finite element simulation, genetic programming-based symbolic regression (GPSR), and uncertainty quantification to develop more robust models for material damage and plasticity. Birky is also interested in creating microstructure-sensitive deformation models that can better predict material behavior under varying conditions. A key aspect of his research is the application of high-performance computing for running large-scale simulations and training machine learning models, enhancing model reliability and efficiency. His work includes developing yield surface models for porous metals and improving existing damage models, like the Gurson model, through symbolic regression. Birky’s research aims to bridge the gap between fundamental material science and industry, offering tools for product design and certification in real-world engineering applications.
Publications
- Predicting the dynamic response of a structure using an artificial neural network 📊💻 (2022)
- Complementing a continuum thermodynamic approach to constitutive modeling with symbolic regression 📐🔬 (2023)
- Generalizing the Gurson model using symbolic regression and transfer learning to relax inherent assumptions 🔄⚙️ (2023)
- Physics-Informed Machine Learning for the Development of Microstructure-Sensitive Deformation and Damage Models 🧬🔍 (2021)
- Methods for Generating Interpretable Yield Surface Models With UQ Based on Data With Multiple Sources of Uncertainty 📈📚 (2024)
- Interpretability and Generalizability of Constitutive Models using Symbolic Regression 🤖📏 (2024)
- Uncertainty Quantification for Interpretable Constitutive Models using Genetic Programming-based Symbolic Regression 🎲💡 (2023)
- Interpretable machine learning for uncertain, microstructure-dependent constitutive models 🔍🧪 (2023)
- Physics-informed Machine Learning for Development of Interpretable, Improved Material Damage Models 🔬🔧 (2022)
- A Data-driven Approach for Improving the Existing Gurson Material Damage Model Using Genetic Programming for Symbolic Regression 📊📉 (2022)
- Improved Plasticity and Damage Models By Symbolic Regression of Microscale Finite Element Simulations 🧠📐 (2021)
- Efficient Clustering of the Dynamic Response of a Structure Subject to Impulse Loading ⚡📊 (2021)
- Learning Implicit Yield Surface Models with Uncertainty Quantification for Noisy Datasets 🌐🔍 (2021)
- Applying genetic programming symbolic regression to solid mechanics 🔧🧠