How lon-fee/aviary works
Project’s Major Domains and Layers
• Core Engine: At the heart is the aviary package. The core modules (found under aviary/core, aviary/api.py, and aviary/mission) perform the main modeling, mission-analysis (with separate FLOPS-based and GASP-based options), and solution of aircraft performance equations.
• Subsystems: A huge part of the project is divided into functional “subsystems” (e.g., aerodynamics, energy, geometry, mass, propulsion) located under aviary/subsystems. Each subsystem is further split into versions (flops_based and gasp_based) and encapsulates domain-specific calculations.
• Models and External Data: The aviary/models directory holds various aircraft model definitions and engine decks. This also includes definitions for external subsystems and decks used during simulations.
• Interface and Visualization: The aviary/interface module provides user interaction points (CLI entry points, graphical input, report generation), while aviary/visualization handles aircraft 3D models and dashboards.
• Testing, Validation, and Documentation: There are extensive tests (under aviary/interface/test, aviary/utils/test, aviary/mission/test, and aviary/validation_cases). Documentation and developer guides in aviary/docs further describe the workflows and design rationale.
How Components Interact
• Data Flow:
– Inputs (user provided CSV decks, configuration files, engine/aero decks) get processed by the respective subsystem builders.
– The mission module collects these outputs and passes them through either the 2DOF or energy-height mission analysis pipelines.
– Results are combined with weight and sizing computations from the core modules.
– The interface layer then provides report generation and visualizations.
Architectural Patterns
• Separation of Concerns: The project is compartmentalized into clearly defined modules (e.g., mission, subsystems, interface, visualization) – each being responsible for distinct aspects of the simulation.
• Layered Architecture: Separate data input (models, decks), computation (subsystems, core mission solvers), and output (reports, visualization) layers. Indicate where OpenMDAO’s capabilities streamline optimization and performance calculations.