Temforge

This pipeline generates large volumes of simulated electron diffraction patterns by first constructing random Cu/Au alloy supercells, then relaxing them with a molecular dynamics engine, extracting a central region of interest, and finally simulating diffraction using a multislice electron microscopy code. Key features include automated supercell generation with adjustable species fractions, variant creation that preserves a central ROI, energy minimization via a KIM EAM potential, ROI extraction, and tilt‑series diffraction simulation. The workflow is orchestrated through a series of Python scripts and a SLURM batch script, allowing users to submit thousands of jobs on a high‑performance computing cluster. It targets researchers in materials science and computational physics who need realistic diffraction data for machine‑learning models or to benchmark experimental techniques. By automating the entire process, it eliminates manual setup, reduces errors, and accelerates data production.