Intelligent Manufacturing Principles for Structural Health Monitoring in Advance Composite Structures: A Machine Learning- Based Driven Placement Approach

Abstract

This paper presents a fundamental investigation into the composition and placement of mechanoluminescence smart material ink to develop a predictive model framework for machine learning-based lead placement in structural health monitoring of nextgeneration fabrication methods. To address the need for multifunctionality and structural health monitoring capabilities in next-generation composite structures, this study proposes a closed-loop integration of material-driven design, multi-modal additive processing, and intelligent manufacturing principles. The proposed method utilizes mechanical simulations and in-situ sensing to establish a priori knowledge to optimize the placement, orientation, and manufacturing conditions for smart sensor embedment to maximize sensing efficiency and part performance. Experimental validation involved the hybrid machining of compact tension samples to produce a crack front in additively manufactured material. Microfluidic deposition facilitated mechanoluminescent sensor embodiment near the crack front, and the sample was then subjected to a fracture mechanic analysis, whereby the crack growth rate and direction was monitored. The resulting light emission from the sensor was captured using a mixture of high-speed video and optical sensors to determine the mechanical energy transferred during crack propagation. The resulting dataset informed an algorithmic approach to govern the placement and process parameters of the embedded mechanoluminescent sensors throughout the polymeric structure. This approach enriches the knowledge pool of digital design discovery, showcasing the capabilities of material-driven learning schema for multifunctional composite fabrication. Our investigation provides important insights into the design and development of next-generation materials with enhanced structural health monitoring capabilities. This approach provides a valuable tool for the real-time monitoring of crack propagation and material failure, which has important implications for the development of safe and reliable structures in various industrial applications.