yugendheran.dev
Freiberg, DE
ykt@portfolio:~$

Yugendheran Kumaran Thirumurugan

M.Sc. Computational Material Science · TU Bergakademie Freiberg
Mechanical Engineer ✕ Simulation ✕ AI & ML ✕ Scientific Computing

Python FEA / FEM Deep Learning MSC Adams Signal Processing AWS EC2 MATLAB CATIA V5 LLMs / RAG
// about me
Who I Am

I'm a Master's student in Computational Material Science at TU Bergakademie Freiberg with a strong foundation in Mechanical Engineering. I sit at the intersection of physics-based simulation and data-driven modelling — and I'm deliberately building toward that intersection.

What drives me is automating the complex — turning raw measurement data into reliable insights through clean, reproducible pipelines. I've built nonlinear FEA solvers from scratch in Python, deployed autonomous pipelines on AWS EC2, and integrated LLMs as reasoning engines in production workflows.

I believe the most consequential engineering advances emerge where physical simulation meets artificial intelligence — and I'm actively building that skill set through academic projects, research work, and independent exploration of LLMs, RAG, and generative AI.

// identity
LocationFreiberg, Germany
StatusM.Sc. Student (2024–2026)
BackgroundB.E. Mechanical Engineering
LanguagesTamil · English C1/C2 · German B2
// education
TU Bergakademie FreibergM.Sc. 2024–2026
Chennai Inst. of TechnologyB.E. 2019–2023
// focus areas
SimulationFEA, MBS, CFD
AI / MLANNs, CNNs, Transfer Learning
AutomationPython Pipelines, AWS
LLM / GenAIClaude API, RAG, Agents
Technical Skills
Programming
PythonMATLAB BashVBA C++NumPy PandasSciPy Matplotlib
AI & ML
PyTorchTensorFlow scikit-learnANNs CNNsTransfer Learning Claude APILLMs RAG
Simulation
MSC AdamsHyperMesh FEniCSFemap Nastran LS-DynaANSYS Fluent Abaqus
CAD
CATIA V5SolidWorks Fusion 360PTC Creo
Cloud & Tools
AWS EC2REST API Telegram APIGit JupyterExcel
Methods
Signal ProcessingNonlinear FEM Fatigue AnalysisDFMEA Six SigmaLean Mfg
Experience
Student Research Assistant
Technische Universität Chemnitz
Oct 2025 – Present · Germany
  • Built automated Python & MATLAB pipelines for end-to-end signal processing of acoustic emission data
  • Developed ANN regression models for fatigue life prediction from machining signals
  • Automated training/evaluation workflows enabling systematic parameter sweeps
Trainee Engineer
Magna Automotive India Pvt Ltd
Jun 2023 – May 2024 · India
  • Rigid Body Dynamics simulations (MSC Adams) for automotive latch & brake systems
  • FEA model setup in HyperMesh for safety-critical closure components
  • Python automation scripts reducing manual post-processing and accelerating design iterations
Intern
Minda Corporation Ltd
Mar 2021 – Apr 2021 · India
  • Starter motor performance parameter design and calculation
  • Technical drawing creation for structural evaluation
Publications & Certifications
Performance Analysis of Tidal Turbine Blades for Different Composite Materials
2023 · Published Research
Simcenter Femap Nastran
Aug 2022 · Official Certification
Autodesk Fusion 360
Oct 2019
Autonomous Vehicle (Beginner)
Nov 2019
// ./projects --list
Project Portfolio
🔬 Research Project · Active
Fatigue Life Prediction via Acoustic Emission Signals
Technische Universität Chemnitz
Building a fully automated measurement data pipeline in Python and MATLAB — from raw AE signal ingestion through feature extraction (time-domain, frequency-domain, spectral) to ANN regression model training and evaluation. The workflow runs systematic parameter sweeps across machining conditions and produces reproducible, structured documentation for the research team.
PythonMATLAB Signal ProcessingANN Fatigue AnalysisPyTorch
Oct 2025 – Present
🤖 Personal Project
Autonomous Crypto Futures Trading System
Python · Anthropic Claude API · AWS EC2
A fully autonomous algorithmic trading pipeline integrating the Anthropic Claude API as the core reasoning engine. Ingests live market data, computes RSI and custom indicators, and determines trade entry, stop-loss, and take-profit levels with zero manual intervention. Includes Telegram Bot for real-time position updates and backtesting via Yahoo Finance. Deployed and running continuously on AWS EC2.
Claude APIAWS EC2 REST APIAgentic Pipeline Telegram BotPrompt Engineering
2025 – Present
🧮 Academic Project
Nonlinear FEA Solver — Penalty Contact Method
Personal Project · Python / FEniCS / NumPy / SciPy
Developed a 2D nonlinear finite element solver from scratch implementing plane-stress triangular elements, penalty-based contact mechanics, sparse stiffness matrix assembly, and Newton–Raphson iteration. Computed von Mises stress fields, validated contact forces against analytical solutions, and conducted a systematic mesh-refinement convergence study — all the way to publication-quality post-processing figures.
FEniCSNumPy SciPyNonlinear FEM Newton-RaphsonContact Mechanics
Jul 2025
🧠 Academic Project
CNN-Based Elasticity Tensor Prediction
TU Bergakademie Freiberg
Built CNN regression models to predict all 9 independent elastic constants (C11–C66) directly from microstructure images — without explicit microstructure characterisation. Applied transfer learning with VGG16, ResNet50, and DenseNet121 using PyTorch Autograd for gradient-based optimisation. Benchmarked architectures by test MSE and documented full training curves and model summaries.
PyTorchCNN Transfer LearningVGG16 ResNet50Material Science
Jun–Jul 2025
🌊 Bachelor Final Year Project
Liquid-Cooled Battery Thermal Management — CFD Simulation
Chennai Institute of Technology · ANSYS Fluent / PTC Creo
Designed a liquid cooling system for an EV battery pack using non-conducting mineral oil. Created simulation-ready CAD geometry in PTC Creo, then performed full CFD simulation in ANSYS Fluent to model internal flow distribution, pressure drop, and thermal behaviour across the pack. Validated results against thermal performance targets and produced full technical documentation.
ANSYS FluentPTC Creo CFDThermal Management EV Battery
2022–2023
🏎️ Team Project · Captain
SAE e-Baja India — Electric Off-Road Buggy
Chennai Institute of Technology · Team Captain
Led an interdisciplinary student team through full vehicle design, manufacturing, and testing phases. Identified an overweight problem and engineered a solution by integrating a Maruti Suzuki Alto 800 brake booster — significantly improving braking force without custom parts. Validated the braking system against SAE Baja safety regulations through iterative testing and analysis.
Vehicle DesignSAE Baja Team LeadershipBrake Systems Manufacturing
2022–2023
// broadcast --channel tech
The Mech × AI Dispatch
What You'll Get
🧠 AI in Engineering
ML models applied to real mechanical problems — fatigue, vibration, thermal, materials
⚙️ Simulation Advances
FEA, CFD, MBS breakthroughs and how AI is accelerating simulation-driven design
🔧 Tools & Workflows
Python automation, LLM integration, and practical pipelines you can actually use
// connect --open
Get In Touch
I'm open to research collaborations, internship opportunities, project discussions, and conversations about AI in engineering. Drop me a message — I respond to everything.
Send a Message
🏠
Home
👤
About
🔬
Projects
📡
Newsletter
💬
Contact
👨‍💻
👨‍💻
Yugendheran
● Online · Freiberg, DE
Yugendheran
Hey! I'm Yugendheran — Engineer, Researcher, and AI enthusiast. Drop me a message and I'll get back to you soon!
status: available for collaborations & opportunities
response_time: ~24 hours
location: Freiberg, Germany
mode: M.Sc. student + researcher