Session: 01-05-01: Foundations of AI and Machine Learning for Aerospace Structural Analysis

Paper Number: 183376

183376 - Ai-Assisted Modeling, Simulation, and Design Optimization of Composite Slender Structures 

The analysis and design of slender composite structures, such as helicopter rotor blades and wind turbine blades, pose considerable challenges because the arrangement of materials can be finely adjusted in a wide range of length scales to meet different performance requirements. This demands the integration of cross-sectional analysis and system-level simulation and leads to steep learning curves when gluing different tools together. Motivated by the need to reduce the barrier to entry for high-fidelity composite structural design, we present to use CompositesAI, an AI-assisted platform that leverages large-language models (LLMs) and domain-specialized models to support users in conducting cross-sectional analysis and design optimization of composite slender structures.

Our approach couples natural-language and structured-query interfaces with a curated knowledge base and automated generation of input files for cross-­sectional analysis tools (e.g., VABS/PreVABS). Users describe the slender composite geometry and layup in simple terms; the system interprets the input, retrieves relevant material and layup data, suggests modelling assumptions (orthotropy, hybrid laminate, geometry discretization), checks consistency, and generates the required simulation configuration. The platform further integrates parametric optimization modules enabling exploration of layup, material, and geometry variations to meet required performance.

We demonstrate the utility of the system on representative slender structural cases, showing that non-specialist users can assemble a valid cross-sectional model, invoke simulation, and iterate design variations far more rapidly than in conventional workflows. This work underscores the synergy between domain-specific simulation tools and general-purpose language-based AI: by bridging natural-language user input and high-fidelity composite structural analysis, CompositesAI provides a path toward more interactive, accessible, and accelerated structural-design workflows for the next generation of composite slender structures.

Presenting Author: Su Tian AnalySwift

Presenting Author Biography: Dr. Tian received the degree in Aerospace Engineering from Purdue University in 2022. His expertise is in multiscale design optimization of advanced composite structures. He is currently a Research Scientist at AnalySwift, mainly responsible for the development of the design tools for composite structures.