Research Focus
We develop simulation workflows and AI models for molecular systems, soft matter, electrolytes, and reactive materials.
Molecular DynamicsReaxFFGNNCUDAMethods
Our work connects physics-based simulation, differentiable modeling, generative design, and high-throughput screening.
LAMMPSGROMACSPyGHPCApplications
Representative applications include viscosity prediction, polymer electrolyte design, reactive force-field validation, and molecular property optimization.
ElectrolytesPolymersLubricantsResearch Experience
- Built GPU-oriented molecular dynamics prototypes for force decomposition, neighbor-list construction, and component-level validation.
- Designed high-throughput simulation pipelines for molecular property prediction and benchmark evidence collection.
- Integrated surrogate models and graph neural networks with chemical feature engineering for candidate screening.
- Maintained reproducible HPC workflows with Slurm, containerized runtimes, and structured verification artifacts.
Selected Publications and Manuscripts
AI-assisted screening of molecular additives for transport-property optimization.
Component-wise validation of reactive force fields on GPU molecular dynamics kernels.
Graph-based surrogate modeling for multi-target molecular property prediction.
Workflow automation for reproducible molecular simulation benchmarks.
Contact
For collaboration, simulation workflows, or model deployment inquiries, contact the MDCore research team through the project-maintained communication channel.