Research · May 2026 – Present

Plasma, simulated

I write and run simulations of the plasma a laser drives off a target, in C on UAB’s Cheaha supercomputer, for a physics lab that grows thin films with lasers and machine learning.

Undergraduate Researcher, Computational Plasma Physics · Computational plasma physics research group, UAB Department of Physics

  • 1 of 3students on the simulation team
  • 12students in the lab
  • CheahaUAB’s supercomputer
  • C + Basiliskthe simulation stack
In motion

Laser meets target

One pulse

A stylized pulse strikes an aluminum target and a plume swells from the spot, then fades.

The work

What I do

Project: Laser-generated plasmas in length scales relevant for thin film growth and processing: simulation and experiment.

  1. Simulations in C

    Write and run coupled laser-ablation and fluid-dynamics simulations of laser-generated plasmas in C with the Basilisk PDE framework.

  2. Supercomputer runs

    Run workloads on UAB’s Cheaha supercomputer; learned Linux and Bash from scratch to manage HPC jobs, environments and data.

  3. Ground truth for ML

    Work with synthetic plasma datasets in which plasma temperature (T) and ionization fraction (Xe) serve as ground truth for testing ML models.

  4. Analysis in Python

    Write Python scripts that process, visualize and analyze simulation output into the plots and comparisons the team uses to evaluate models.

  5. ML-guided growth

    Support research on ML-guided pulsed laser deposition (PLD), where Gaussian Process Bayesian Optimization builds surrogate models of plasma conditions from a small number of experiments.

AVS 72

On the record

PublicationCo-author

“Machine Learning-Assisted Pulsed Laser Deposition Based on Plasma Optical Emission Signatures”

D. Carpenter, Z. Nasiri, S. Gupta, R. Mayatskiy, S. B. Harris (Center for Nanophase Materials Sciences, Oak Ridge National Laboratory), R. P. Camata

Accepted for an oral presentation · AVS 72nd International Symposium · Pittsburgh, Nov 8–13, 2026

Across 1,000 independent trials, the model found optimal plasma conditions in as few as three adaptive iterations: a step toward autonomous, real-time control of thin-film growth.

Tools

  • C
  • Basilisk
  • Python
  • Bash
  • Linux
  • HPC (Cheaha)
  • Gaussian Process Bayesian Optimization
  • Scientific visualization
See every role