Our main research interests include Integrated Computational Materials Engineering (ICME), multi-scale computational modeling and optimization, uncertainty quantification, reduced order modeling and machine learning to study structural behavior in length-scales ranging from microstructure to component-level in moderate to extreme environments. A snapshot of our research topics is illustrated below.
Embedded Files
Our research interests are including (but not limited to) the following topics:
Multi-Scale Modeling of Metallic Microstructures
We develop computational models to bridge processing-microstructure-properties for metallic microstructures. These models investigate how thermo-mechanical processing variables affect the evolution of microstructures, as well as the development of material properties throughout components.
Multi-Scale Computational Design of Mechanical Metamaterials
We build computational tools to design mechanical metamaterials. We focus on the characterization of non-periodically repeating metamaterial architectures, the development of numerical homogenization methods, building mechanical property spaces, and multi-scale design of metamaterial architectures.
Materials Design under Uncertainty
The deterministic computational models are not powerful for precisely capturing the structural response of engineering components due to the variations in structural properties (aleatoric uncertainty), or modeling/mathematical inaccuracies in computational simulations (epistemic uncertainty). Such variability can propagate over the computational models and significantly impact the expected performance. Our goal is to build stochastic multi-scale computational environments to model and optimize engineering materials (metals, composites, polymers, metamaterials) by considering these aleatoric and epistemic uncertainties.
Computational Modeling of Ferromagnetic-Paramagnetic Phase Transition of Magnetic Materials
The existing knowledge on the determination of the magnetic phase transition zone is based on simplified analyses as the comprehensive effects of the long-range interactions among the magnetic spins, external parameters, and uncertainties have been neglected. We propose to determine the ferromagnetic-paramagnetic phase transition onset and quantify the likelihood of the transition by examining: i) high-order interactions between the magnetic spins and external fields; ii) long-range effects including the grain boundaries and dislocations; iii) propagation of the uncertainties in external fields and temperature.
Machine Learning and Data-Driven Modeling and Design of Materials
The multi-scale computational techniques have not been utilized for engineering-scale structural optimization problems due to the excessive computational cost of the multi-scale simulations involving billions of degrees of freedom. Machine learning-reinforced and data-driven models are found to provide efficient mathematical surrogates that are perfectly suited to overcome this computational burden. Our goal is to use this mathematically-rigorous approach to represent the multi-scale and complex interplay between processing, materials, and properties with their data-driven surrogates.
Physics-Based and Data-Driven Modeling for Thermo-Mechanical Processing of Additively Manufactured Metals
We develop physics-based and data-driven models for additively manufactured metals to investigate the formation and evolution of microstructural features, including crystallographic texture and grain topology. Applications include the design of Ti-Al and Nickel-based alloys during electron beam melting and MELD processes to be used for aerospace systems.
FUNDING
FUNDING
Report abuse