How mwasifanwar/aeroflow works
Project type and purpose
• This is a Python-based scientific computing “full-stack” pipeline (no user-facing GUI) that integrates:
• Computational Fluid Dynamics (CFD) simulation
• Geometry generation
• Physics-informed and surrogate neural networks
• Optimization algorithms (Bayesian, genetic, gradient)
• A minimal API layer for orchestration
• Visualization of results
• Purpose: iterate on aircraft/vehicle design by automating geometry → mesh → CFD solve → surrogate training → optimization → visualization.
File structure
• Top-level:
• main.py: CLI entry point
• config.yaml: global parameters
• src/api/server.py: REST API entrypoint (likely Flask/FastAPI)
• src/cfd: solver, BCs, post-processor
• src/geometry: airfoil & mesh generation, parameterization
• src/neural_networks: CNNs, physics-informed, surrogate models
• src/optimization: Bayesian, GA, gradient-based optimizers
• src/utils: config loader, helper routines
• src/visualization: flow visualization, plotting
• tests/: unit tests for CFD and NN
• No database; data flows through in-memory objects or file I/O.
Extract additional insights from README
• The pipeline blends CFD with deep learning surrogates for rapid design exploration.