projects
Research and engineering projects in computational mechanics, digital twin development, and AI/ML-enabled simulation tools.
All work listed below was conducted under industry research programs. Implementations remain proprietary; descriptions reflect generalized research contributions.
2023 – Present
-
Digital Twin Framework for Advanced Manufacturing
Developing end-to-end digital twin infrastructure integrating physics-based process simulation, real-time sensor data, and validation workflows for advanced manufacturing programs.
- Architecting sequential thermal–mechanical coupling framework with Python-based automation.
- Validating residual stress and distortion predictions against in-situ thermocouple measurements.
- Delivering parameterized HPC job-management pipeline for cluster-scale execution.
- Integrating boundary-condition identification and model calibration workflows for virtual process qualification.
-
AI/ML-Enabled Materials & Process Prediction Platform
Building GUI-based software tools combining transformer-based language models with physics-based simulation for materials property prediction, inverse design, and LLM-assisted FEA automation.
- Architecting ML-powered software combining transformer language models with Abaqus-based simulation.
- Benchmarking transformer architectures against high-fidelity RVE micromechanics outputs.
- Developing GUI front-end exposing FEA workflows to non-CAE users.
- Implementing inverse-design workflows for composite property targets.
2022 – Present
-
Thermomechanical Process Simulation — Solid-State Manufacturing
High-fidelity coupled thermal–mechanical FEA of solid-state additive manufacturing processes — covering heat generation, deposition sequencing, residual stress, and experimental validation.
- Developed spatially-distributed heat generation model as a Fortran user subroutine.
- Built path-synchronized element activation for multi-layer deposition sequencing.
- Validated residual stress prediction workflow against thermocouple measurements for multi-layer AFSD.
- Automated end-to-end simulation pipeline from CAD geometry through HPC execution and post-processing.
2021 – Present
-
Multi-Physics Modeling of Quenching Processes — Aluminum & Airframe Components
Developing multi-physics digital twin and planning tools for industrial quenching — coupling thermal, metallurgical, and mechanical phenomena to predict distortion, residual stress, and microstructure evolution in aluminum and airframe structural components.
- Coupled thermal–metallurgical–mechanical FEA for aluminum and metallic structural components.
- Optimization framework for tailoring quenching processes of representative airframe components.
- Digital twin architecture for large-scale metallic-structure quenching.
- Recognized with Vertical Flight Society Best Paper Awards in 2024 and 2025.
2020 – 2022
-
Defect-Aware Composite Structural Analysis — Aerospace Structures
Delivered global-to-local aircraft wing simulation frameworks incorporating manufacturing distortion, XFEM-based fatigue crack growth, CDM progressive damage, and CFD-driven thermal loading.
- Built global-to-local sub-modeling framework for aircraft wing structural assessment.
- Implemented XFEM-based fatigue crack growth with defect-informed initiation.
- Integrated CDM progressive damage in composite laminates.
- Coupled CFD-driven thermal loading with structural model; mapped manufacturing distortion onto as-built geometry.
2019 – Present
-
Multiscale RVE Micromechanics & Digital Material Generation
Developed digital material generation toolkits for composite laminates — predicting ply-level elastic, strength, and thermal properties using RVE-based micromechanics, with explicit treatment of manufacturing defects.
- Developed digital material generation toolkit predicting ply-level elastic, strength, and thermal properties.
- Implemented explicit RVE modeling of voids, fiber waviness, and ply misalignment.
- Linked micro-scale defect distributions to macro-scale structural performance.
- Integrated toolkit into broader simulation and digital twin pipelines.
2014 – 2017
-
Composite Fabrication & Experimental Characterization
Designed and executed experimental programs covering composite panel fabrication, mechanical and thermal testing, and microscopy-based defect characterization.
- Fabricated composite panels (layup, debulking, autoclave cure) under ASTM-compliant procedures.
- Performed mechanical testing: Mode I/II fracture, fatigue, short-beam shear, stress relaxation.
- Conducted thermal characterization (DSC, CTE, thermomechanical response).
- Microscopy-based defect characterization feeding directly into FE model validation.
-
Nonlinear FEA & Multiscale Modeling — Honeycomb Core Structures
Investigated forming and creep behavior of honeycomb sandwich structures using advanced nonlinear FEA and micro–macro multiscale modeling.
- Nonlinear FEA of honeycomb forming under spherical, cylindrical, and multi-radius bending (MSC Marc).
- Micro–macro multiscale modeling of woven composite honeycomb cores.
- Custom FORTRAN subroutines for anisotropic hyperelastic creep behavior.
- Foundation methodology for later defect-aware composite analysis and digital twin work.