ThrustMaximizationProp

ThrustMaximizationProp is an automated pipeline for optimizing propeller geometry to maximize thrust under specific cruise conditions. It leverages Differential Evolution (DE) algorithms and Bezier curve parameterization to generate smooth, efficient blade shapes, validated using the XROTOR aerodynamic solver.

🚀 Features

• Advanced Optimization: Uses the pymoo library's Differential Evolution (DE) algorithm to explore the design space and avoid local optima.
• Smooth Geometry: Parameterizes chord and twist distributions using Bezier curves (via Control Points) to ensure manufacturable and aerodynamically smooth blade shapes.
• Aerodynamic Analysis: Integrates directly with XROTOR to evaluate propeller performance (Thrust, Power, Efficiency).
• Multi-Start Strategy: Supports multiple optimization runs with random seeds to ensure robustness and find the global optimum.
• Configurable: Fully driven by a JSON configuration file for easy adjustment of operating conditions, geometry bounds, and optimization settings.

🛠️ Prerequisites

• Python 3.8+
• XROTOR: The xrotor executable must be installed and accessible.
• Python dependencies:

  pip install numpy scipy pymoo

📂 Installation

1. Clone the repository:

   git clone https://github.com/shubhamtamboli/ThrustMaximizationProp.git
   cd ThrustMaximizationProp

2. Ensure xrotor is correctly set up in your environment or the path is configured in the interface scripts.

🏃 Usage

Run the optimization pipeline using main.py and provide a configuration file:

python main.py --file thrust-cruise.json --generations 100

Arguments

• -f, --file: (Required) Filename of the JSON configuration located in processing/inputs/.
• --generations: (Optional) Number of generations for the optimizer (default: 10000).

⚙️ Configuration

The optimization is controlled by a JSON file (e.g., processing/inputs/thrust-cruise.json).

Structure

{
    "optimization": {
        "n_gen": 100,          // Number of generations
        "pop_size": 50,        // Population size
        "n_restarts": 5        // Number of independent runs
    },
    "operatingConditions": {
        "V": 20.0,             // Cruise velocity (m/s)
        "rho": 1.225,          // Air density (kg/m^3)
        "rpm": 12000,          // Propeller RPM
        "maxPower": 50000.0,   // Max power constraint (W)
        "requiredThrust": 0.0  // Min thrust constraint (N)
    },
    "geometryGlobal": {
        "diameter": 0.30,      // Propeller diameter (m)
        "bladeCount": 8,       // Number of blades
        "rStations": [...]     // Radial stations for evaluation
    },
    "geometryBounds": {
        "chordMin": 0.05,      // Min chord (fraction of Diameter)
        "chordMax": 0.15,      // Max chord (fraction of Diameter)
        "pitchDegMin": 0.0,    // Min pitch (degrees)
        "pitchDegMax": 40.0    // Max pitch (degrees)
    }
}

📊 Output

• Console: Real-time progress of the optimization (Generations, Best Thrust, Constraint Violations).
• Results: The best design is saved to processing/outputs/ (or configured output path), typically including:
  • Optimized Control Points
  • Performance Metrics (Thrust, Efficiency)
  • Geometry Data (Chord/Pitch distributions)

🤝 Contributing

Contributions are welcome! Please open an issue or submit a pull request for any improvements or bug fixes.

📄 License

MIT License