Abstract
The mechanical performance of composite materials under cryogenic environments presents significant challenges to structural reliability. Current limitations in in-situ characterization techniques hinder the comprehensive understanding of damage evolution mechanisms under cryogenic bending loads. To address this, flexural damage behavior of carbon fiber reinforced polymer laminates at temperatures as low as 123 K was systematically investigated using in-situ fiber-optic acoustic emission (AE) testing. A refined damage mode identification method, integrating mode decomposition analysis and a novel deep learning algorithm, was adopted to elucidate the cryogenic damage mechanisms. Results reveal that cryogenic environments significantly reduce the damage initiation strain threshold and compress the temporal intervals between damage modes, thereby promoting homogenization of damage development and the dissipation of mechanical energy. Although cryogenic temperatures strengthen the resin matrix and the bonding at the matrix-fiber interface, matrix embrittlement at 123 K markedly decreases the delamination resistance, serving as the key contributing factor to strength degradation. Notably, the refined damage identification methodology achieves over 99 % classification accuracy in identifying four critical damage modes across different temperature conditions while effectively recovering hidden information related to fiber/matrix debonding and fiber breakage. This study advances the understanding of cryogenic damage mechanisms in composite materials and establishes a robust framework for real-time damage assessment in cryogenic engineering applications.
| Original language | English | 
|---|---|
| Article number | 112994 | 
| Journal | Composites Part B: Engineering | 
| Volume | 308 | 
| DOIs | |
| Publication status | Published - 1 Jan 2026 | 
Keywords
- Acoustic emission
 - Cryogenic damage behaviors
 - Damage identification
 - Deep learning
 - Mode decomposition
 - Polymer composites