// USC Viterbi · Aerospace Engineering · M.S. 2026

Samay
Shah

CFD · Combustion · Physics ML

I build high-fidelity simulations and machine learning tools for aerospace propulsion — from swirl-stabilized combustors to adaptive kinetics solvers.

View Projects Get in touch →
Scroll
// 01 — About

Who I am

I'm a Master's student in Aerospace Engineering at USC Viterbi, graduating May 2026. My work sits at the intersection of high-fidelity simulation and machine learning — specifically applying ML to accelerate combustion chemistry in CFD workflows.

Before USC, I worked as a Stress Analyst in the aerospace industry, giving me hands-on experience with structural analysis in production environments. I bring both research depth and engineering practicality to the problems I work on.

I'm actively pursuing roles in aerospace simulation, propulsion, and Physics-informed ML — particularly at companies pushing the boundary between cloud computing and simulation-driven design.

SU2
CFD Solver
v5
Cantera
PyTorch
ML Framework
M.S.
USC Viterbi '26


// Skills
CFD
  • SU2
  • Gmsh
  • OpenFOAM
  • Mesh generation
  • Turbulence modeling
Combustion
  • Cantera
  • Chemical kinetics
  • Premixed flames
  • Counterflow flames
  • Stiff ODE solvers
ML / Python
  • PyTorch
  • PINNs
  • GNNs
  • Surrogate modeling
  • NumPy / Matplotlib
Structures
  • Stress analysis
  • FEA
  • NASTRAN
  • Fatigue & fracture
  • Load cases
// 02 — Projects

Things I've built

01  —    completed
Swirl-Stabilized H₂ Combustor CFD
High-fidelity 3D simulation of a hydrogen combustor using SU2. Modeled swirl-stabilized turbulent reacting flow with species transport, Gmsh mesh generation, and automated Python post-processing pipelines.
SU2 Gmsh Python Turbulence H₂ Combustion
GitHub →
02  —    completed
Adaptive Chemical Kinetics via ML
Physics-informed ML surrogate for combustion chemistry. Two-policy Jacobian-based adaptive solver in Cantera v5 — selects reduced mechanisms on-the-fly to cut stiff ODE solve times while preserving accuracy.
Cantera PyTorch Kinetics Adaptive ODE ML Surrogate
GitHub →
03  —    completed
Cantera Multi-Fuel Flame Suite
Systematic premixed and counterflow flame study across H₂, CH₄, and C₃H₈ at varying pressures. Pickle-based simulation cache, automated multi-subplot visualization, and laminar flame speed comparisons.
Cantera Combustion Matplotlib Python Flame Speed
GitHub →
04  —    in progress
Physics-Informed ML for Reacting Flows
Exploring GNN and PINN architectures for mesh-agnostic inference on combustion fields. Targeting scalable, cloud-native simulation workflows that reduce time-to-solution for complex reacting flow problems.
PINNs GNNs PyTorch Reacting Flow Cloud Sim
GitHub →
// 03 — Experience

Where I've worked

2024 — Present
Graduate Researcher — Combustion & ML
USC Viterbi School of Engineering
Developing ML-accelerated chemical kinetics solvers for combustion CFD. Research focuses on reducing computational cost of stiff ODE systems through adaptive mechanism selection and surrogate modeling.
Prior
Stress Analyst
Aerospace Industry
Performed structural stress analysis on aerospace components — fatigue, fracture, and static load cases. Applied FEA tools in a production engineering environment with flight-critical hardware.
2024 — 2026
M.S. Aerospace Engineering
University of Southern California
Specialization in propulsion and computational fluid dynamics. Coursework in turbulence, combustion, and advanced numerical methods.
// 04 — Contact

Let's talk

Open to full-time roles in aerospace simulation, propulsion, and Physics-informed ML.

Email samayshah5@gmail.com → LinkedIn linkedin.com/in/samayhs →