ENZO-Inspired 1D HLL Shock Tube Solver

A hybrid C++ / Python simulation pipeline that implements a 1D HLL Approximate Riemann Solver for the Euler equations.

This project demonstrates a high-performance numerical core (C++17) wrapped in a robust automation layer (Python), capable of simulating classic fluid dynamics problems like the Sod Shock Tube.

Features

• Numerical Core: C++ implementation of the HLL (Harten-Lax-van Leer) flux scheme.
• Time Integration: Explicit time-stepping with dynamic CFL condition.
• Hybrid Workflow: Python handles compilation, input parsing, data aggregation, and visualization; C++ handles the heavy compute.
• Visualization: Automated generation of profile plots and evolution GIFs.
• Reproducibility: Metadata preservation (JSON) alongside raw binary data (NPY).

Directory Structure

.
├── cpp/                 # C++ Source (Numerical Engine)
│   ├── grid.h/cpp       # Mesh and State management
│   ├── solver.h/cpp     # HLL Flux & Time integration
│   ├── utils.h/cpp      # I/O and Helpers
│   └── main.cpp         # Entry point
├── python/              # Python Tools (Orchestration)
│   ├── parser.py        # Config file parsing
│   ├── run.py           # Build system & Simulation runner
│   └── visualize.py     # Plotting & Animation
├── data/                # Inputs & Outputs
│   ├── shock_tube.enzo  # Simulation parameters
│   ├── outputs/         # Raw CSV snapshots
│   └── plots/           # profile plots and animation
└── README.md

Quick Start

1. Prerequisites

• C++ Compiler: g++ (supporting C++17)
• Python 3.x
• Dependencies: numpy, pandas, matplotlib

  pip install numpy pandas matplotlib

2. Configure & Run

The parameters are defined in data/shock_tube.enzo (ENZO-style format).

#1. Run the simulation (Auto-compiles C++ code)
python3 python/run.py data/shock_tube.enzo --rebuild

#2. Visualize the results (Generates plots and GIF)
python3 python/visualize.py

3. View Results

Check the data/plots/ directory:

• simulation.gif: Animation of density, velocity, and pressure evolution.
• profile_*.png: Static profiles at different time steps for pressure, density and velocity.

Sample Output

Configuration

The solver is controlled by .enzo files. Example:

nx = 400
x0 = 0.0
x1 = 1.0
interface_position = 0.5
gamma = 1.4
cfl = 0.45
t_final = 0.2
output_dt = 0.01  # Snapshot cadence in simulation time
bc_type = outflow # Options: outflow, reflective

Limitations & Future Work

• Spatial Reconstruction: Currently First-Order Godunov (Piecewise Constant). Future upgrade: Piecewise Linear Method (PLM) for 2nd order accuracy.
• Geometry: Strictly 1D Cartesian.
• Performance: Uses CSV I/O for clarity. For large-scale runs, this should be replaced with HDF5 or unformatted binary I/O.