AEROSP 568 - Computational Methods for Engineered Systems Design

Course Description: Examines computational and statistical methods for the design of complex engineered systems. Topics include design of experiments, surrogate modeling and surrogate-based optimization, sensitivity analysis, and mixed-variable design space exploration. Aerospace examples highlight advanced aircraft and propulsion system concepts, including methods for co-optimization of aircraft, propulsion, and energy/power management. (3 credits)

Click here to download the course syllabus.

Note: This course was previously taught as AEROSP740 - Complex Systems Design & Integration in Winter 2024 and Winter 2025.

Who Should Take This Course:

• PhD and Master’s students in Aerospace, Mechanical, IOE, and related fields who use or plan to use computational models in design or research and want a more systematic way to design studies, build models, and interpret results.
• Aerospace seniors aiming for graduate school or design / analysis roles in industry

Prerequisites: Programming skills required. A background in statistics, aircraft and/or propulsion system design, optimization is advantageous but not required.

Topics covered:

–Foundations of Complex Systems

• Introduction to complex systems and systems thinking
• Nonlinearity and emergent behaviors
• Systems design processes, the digital Vee framework
• Validation and verification
• Introduction to quality engineering; Taguchi method
• Computational design methods

–Future Flight Concepts - a brief overview

• Current challenges and opportunities in aviation
• Novel propulsion technologies (electrification, batteries, hydrogen)
• Economic and life-cycle considerations

–Physics-based modeling for hybrid-electric systems

• Electrified propulsion system architecture and component modeling
• Graph-based architecture representation and generation
• Energy-based flight mission performance analysis
• Operational considerations: energy and power management strategies
• Aircraft and propulsion system sizing and synthesis
• Co-design of system, sub-systems, and hybrid operations

–Statistical and Data-driven Methods

• Introduction to probability and statistics for engineering design
• Analysis of main and interaction effects (Screening, Morris method, Pareto charts)
• Sampling strategies, Design of Experiments (factorials, space-filling designs including Latin Hypercubes), Monte Carlo sampling
• Linear and non-linear regressions and surrogate modeling techniques (Response Surface Methodology, Radial Basis Functions, Kriging, Artificial Neural Nets)
• Model adequacy checks, residual analysis, validation, testing
• Multi-dimensional trade studies using prediction profiles
• Global and local sensitivity analysis (variance-based and moment-independent methods)
• Mixed-variable design space exploration and surrogate-based optimization