Skip to content

About

No description, website, or topics provided.

Resources

Code of conduct

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

Computational Chemistry with Python

This repository is where I "dabble" with the invisible forces that hold the universe together—specifically, the chemical bonds in rocket propellants. I'm currently self-learning how to use Python to simulate molecular energy so I can understand why some fuels (like Hydrazine) are so much more "exciting" (and energetic) than others.

🧪 What’s happening here?

I’m moving away from just reading about chemistry in textbooks and trying to actually compute it. I self-read about bond energy in chemistry and I am still learning the theory behind it. Most people use Kaggle for Machine Learning, but I’ve been "repurposing" Kaggle’s free GPUs as a cloud-based supercomputer to run physics simulations. It is fun to get free GPU resources and simulation environment in Kaggle.

Current Focus:

  • Potential Energy Surfaces (PES): Using the Hessian Matrix (from my Multivariable Calculus self-studies) to find local minima and stable molecular geometries.
  • Quantum ESPRESSO: Someone said this is good to simulate. I am still learning what is Density Functional Theory (DFT) calculation to "see" electron density.

Because the computation of full DFT for screening large molecules is very high, I found a pipeline using xtb. By automating the conversion from Moldraw-generated SMILES to GFN2-xTB optimized geometries via Python, I found a quicker workflow to analyze bond vibration modes / energies

If we want to go to Mars or build better aircraft, we need to understand the chemistry of propulsion at an atomic level.

About this repo

This repo was originally made by me to just tweak and play around with other DFT codes (especially GPU4PySCF), but now it is home to my Quantum ESPRESSO Kaggle notebook as well as some interesting outputs I am accumulating over time alongside RocketCEA and Cantera Gas and Propellant analyses. My code is costantly in development, however my pw.x build so far works flawlessly and is calibrated/tested across a variety of systems. Clone it and change paths to local, an AMD fork is currently under development and will be added here once done and when it can be tested. This repo is specifically for people with workflow similar to mine:

  • Extract .xyz file from any method (eg using Avogadro/xTB)
  • Run Quantum ESPRESSO with minimal setup friction using fast setup methods for easy tweaking and checking

About

No description, website, or topics provided.

Resources

Code of conduct

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Used by

Contributors

Languages