Research Overview
Our research focuses on understanding and controlling the relationships among manufacturing processes, solidification, microstructure, defects, and performance in metallic materials. We integrate advanced experiments, in-situ sensing, and multiscale computational modeling to develop predictive understanding and new strategies for materials processing, with particular emphasis on metal additive manufacturing.
Our work is organized around five research themes:
1. Multiscale Modeling of Solidification and Microstructure
We develop physics-based computational models to predict solidification and microstructure evolution across the multiple length and time scales encountered in solidification based processes, in particular, casting and metal additive manufacturing. Our research connects processing conditions and thermal histories to nucleation, dendritic and cellular growth, competitive grain growth, segregation, and resulting grain structures.
A major focus is muMatScale, our open-source and highly parallel solidification modeling framework. muMatScale enables grain and subgrain structures to be simulated over engineering-relevant domains while retaining the spatial resolution needed to capture rapid-solidification phenomena. The framework provides a bridge between process-scale thermal simulations and local microstructure evolution and is designed for efficient execution on multicore workstations and HPC systems.
2. Process–Structure–Defect Relationships in Metal Additive Manufacturing
We investigate how processing parameters and melt-pool behavior govern surface morphology, porosity, microstructure, and defect formation in metal additive manufacturing. By combining controlled LPBF experiments, materials characterization, and physics-based modeling, we seek to identify measurable signatures of process instability and establish predictive relationships between processing conditions and build quality.
Current research includes melt-pool stability, lack-of-fusion and keyhole-related defects, surface roughness, porosity, solidification cracks, and the use of surface and process signatures to assess subsurface material quality.
3. Microstructure Control and Hard-to-Print Alloys
Additive manufacturing creates opportunities to control solidification at spatial and temporal scales that are difficult to access using conventional manufacturing. We explore new processing strategies to tailor grain and subgrain structures while mitigating cracking and other defects in alloys that are challenging to process by fusion-based additive manufacturing.
Our research includes pulsed-laser processing, remelting and rescanning strategies, grain refinement, columnar-to-equiaxed transition, and physics-based approaches for understanding and suppressing solidification-related cracking.
4. Solidification Science and Advanced Manufacturing Processes
Many of the microstructures and defects observed in additive manufacturing originate from fundamental solidification phenomena that are shared across welding, casting, remelting, and other fusion-based manufacturing processes. We study these mechanisms experimentally and computationally to establish transferable understanding of nucleation, grain growth, solute transport, segregation, fluid flow, and defect formation.
This research provides fundamental knowledge that supports applications ranging from laser-based additive manufacturing to directed energy deposition, casting, and vacuum arc remelting of advanced alloys.
5. In-Situ Monitoring, AI, and Digital Twins
Collaborating with ARTS-Lab, we develop sensing and computational approaches for observing manufacturing processes in real time and transforming process data into actionable information about material and component quality. Our research integrates optical, thermal, event-based, and mechanical sensing with physics-based models and data-driven methods to identify process anomalies, characterize defect signatures, and advance digital-twin-enabled manufacturing.
A long-term objective is to move beyond process monitoring toward real-time prediction and validation, where experimental observations continuously inform computational models and support decisions during manufacturing.