Computational Materials Science

Upadesh Subedi

Modeling microstructural evolution & failure of materials in extreme environments — from laser-driven multicomponent alloys to next-generation nuclear reactor materials.

Phase Field Method Materials Informatics Tensor Computation Machine Learning Nuclear Materials
Upadesh Subedi
US
01

About

I am a Doctoral Researcher at the Faculty of Mechanical Engineering, Silesian University of Technology, Gliwice, Poland, where I work on computational materials science with a focus on phase-field modeling, materials informatics, and tensor-based thermodynamic frameworks.

My PhD thesis, "Towards Digital Twins for Quantifying Laser-Microstructure Interface in Multicomponent Alloys Using Thermodynamic Tensor Model," developed computational frameworks coupling phase-field methods, thermodynamic (CALPHAD-based) modeling, and machine learning to predict microstructural evolution during laser-based processing of multicomponent and high-entropy alloys.

Current

Doctoral Researcher
Silesian University of Technology
Gliwice, Poland

02

PhD Research

My doctoral work centers on building digital twin frameworks for laser-material interactions in multicomponent alloys — combining phase-field simulation, CALPHAD thermodynamics, and deep learning to predict, accelerate, and interpret microstructural evolution.

🧊

Phase Field Modeling

Simulating microstructural evolution — grain growth, intermetallic formation, phase decomposition — under thermal and electric driving forces relevant to laser additive manufacturing and electromigration.

⚛️

Thermodynamic Tensor Models

Coupling CALPHAD-based thermodynamic descriptions with tensor-formulated free energy models to capture multicomponent alloy behavior efficiently within phase-field frameworks.

🤖

Machine Learning & Informatics

Conv-LSTM and U-Net frameworks for multi-generational microstructure prediction, accelerating phase-field simulations and enabling data-driven materials design (pyMPEALab, IMCATHEA).

🔬

Laser–Microstructure Interface

Quantifying meltpool dynamics, intermetallic growth (e.g. Ti2Cu, Cu6Sn5), and resulting microstructure in laser additive manufacturing of Ti6Al4V and other multicomponent systems.

03

Future Research Direction

Building on my PhD foundation, my future research aims to develop advanced computational materials models to predict failure in structural materials exposed to extreme environments — enabling the design of safer, more resilient materials for next-generation nuclear reactors and aerospace applications.

Plasma-Facing Materials for Nuclear Fusion

Modeling, simulating, and predicting the microstructural evolution, degradation, and fracture of plasma-facing wall materials exposed to extreme electric, magnetic, and thermal fields in nuclear fusion reactors — including void swelling, grain growth, radiation-induced phase separation, and fracture modeling, coupled with CALPHAD thermodynamics.

Phase FieldRadiation DamageFracture ModelingCALPHAD

Advanced Nuclear Fission Fuels (TRISO & SMR Fuels)

Computational studies of next-generation fission fuels such as TRISO particles and advanced fuels for small modular reactors (SMRs) — investigating evolution and degradation, diffusion phenomena, microstructural changes after fission reactions, and the impact of high thermal fields on fuel performance.

TRISO FuelDiffusion ModelingSMRMicrostructure Degradation

Additive Manufacturing of Multi-Material Alloys

Computational modeling for additive manufacturing of multi-material alloys for extreme environments — modeling multiphase microstructure evolution during additive manufacturing of high-entropy and multicomponent alloys built for nuclear and aerospace applications.

Additive ManufacturingHigh-Entropy AlloysMultiphase MicrostructureAerospace
04

Selected Publications

A non-isothermal multi-phase field approach to model the meltpool and IMC grains interaction in Ti-Au material

U. Subedi, N. Moelans, T. Tański, A. Kunwar — Computational Materials Science, 2025

Foretelling microstructural interface evolution with multi-generational convolutional-LSTM framework

U. Subedi, N. Moelans, T. Tański, A. Kunwar — Engineering with Computers, 2025

pyMPEALab Toolkit for Accelerating Phase Design in Multi-principal Element Alloys

U. Subedi, A. Kunwar, Y. A. Coutinho, K. Gyanwali — Metals and Materials International, 2022

View full list on Google Scholar ↗

Media & Documents

Simulation videos, thesis, presentations, and posters.

Phase-Field Simulation | Phase Growth

PF Modeling of laser processing in Ti-Au binary alloy system: Result showing IMC Grain Growth along with other evolving phases.

05

Software & Tools

🧪

pyMPEALab

A Python-based Multi-Principal Element Alloy Laboratory for predicting phases of MPEAs using neural network algorithms.

View on GitHub →
🖥️

IMCATHEA

A Python-based GUI application for detecting intermetallic phases in High Entropy Alloy / Multi-Principal Element Alloy systems using neural networks.

View on GitHub →
06

Get in Touch

I'm always happy to connect on topics around phase-field modeling, materials informatics, nuclear materials, and additive manufacturing — feel free to reach out.