TL;DR
This paper introduces a novel hybrid sensor network combining piezoelectric (PZT) and fiber optic (FBG) sensors for comprehensive, real-time monitoring of composite structures throughout their entire lifecycle, from manufacturing to in-service health. They also developed an improved damage localization algorithm that significantly enhances accuracy compared to existing methods.
What Problem Does This Paper Address?
Carbon Fiber Reinforced Plastics (CFRPs) are increasingly used in critical applications like aerospace and automotive due to their excellent properties. However, their manufacturing process, especially using vacuum bag molding (VBM), can be complex and prone to defects (voids, dry spots) that compromise structural integrity. Furthermore, monitoring the health of these structures during their operational life is crucial for safety and longevity. Existing methods for monitoring either the manufacturing (curing) phase or the in-service phase are often limited in scope, real-time capability, or accuracy. Specifically, monitoring curing parameters like resin modulus and reaction progress can be challenging, and accurately locating damage in service using sensor networks requires sophisticated algorithms. This paper aims to bridge this gap by providing a unified, real-time, life-cycle monitoring solution for CFRPs.
Methods
The core of this work is a piezoelectric-fiber hybrid sensor network.
- During Manufacturing (Cure Monitoring):
- PZT Sensors: Embedded within the composite during manufacturing, these sensors monitor the resin's storage modulus and the progress of the curing reaction (POR) using guided wave propagation. This allows for real-time tracking of critical curing parameters.
- FBG Sensors: Also embedded, these sensors measure strain and identify the gelation point of the resin. This provides a complementary measurement to validate the PZT sensor data and precisely pinpoint a key stage in the curing process.
- Data Analysis: The amplitude of guided wave signals, processed via Hilbert transform, is used to derive the POR. An Energy Variation Index (EVI) is also proposed to track the loss factor of the resin.
- During Service (Damage Identification):
- PZT Network: The embedded PZT sensors are repurposed for in-service structural health monitoring (SHM). They are used to generate and receive guided waves to detect and locate damage.
- Modified Probability Diagnostic Imaging (PDI) Algorithm: The paper proposes an enhanced PDI algorithm to improve damage localization accuracy. This involves:
- Averaged Shape Factor: Instead of a single fixed shape factor, the algorithm considers multiple shape factors and averages the results.
- Multi-frequency Fusion: Utilizing multiple excitation frequencies (50, 60, 70 kHz) to capture damage-sensitive responses.
- Dynamically Adaptive Shape Factor: This is the most significant algorithmic improvement. It dynamically adjusts the shape factor for each sensor path in real-time based on the damage index (DI) of that path. This allows for more precise focusing on potential damage areas.
Datasets
The paper describes experiments conducted on fabricated CFRP specimens. No specific publicly available datasets were used for training or evaluation. The "datasets" are the experimental measurements obtained from the PZT and FBG sensor networks during the curing process and from simulated damage experiments on the cured composite plates.
Key Insights and Ideas
- Synergistic Hybrid Sensing: Combining PZT and FBG sensors offers a powerful dual capability: PZT for broad cure monitoring and FBG for precise gelation point identification, validating each other.
- Guided Waves for Cure Monitoring: Using guided waves to track resin modulus and reaction progress during curing is a novel application that provides real-time insights into the manufacturing process.
- Life-Cycle Monitoring is Achievable: The sensor network, once embedded, can seamlessly transition from monitoring manufacturing to monitoring in-service health, offering a truly integrated solution.
- PDI Algorithm Enhancement: The development of a dynamically adaptive shape factor is a crucial step in improving the accuracy of guided wave-based damage localization, moving beyond static or averaged parameters.
- Damage Index as a Guide: The insight that the damage index (DI) of a sensing path is directly related to its proximity to damage is key to developing the adaptive shape factor.
Key Results
- Cure Monitoring:
- The POR and EVI curves derived from guided wave data showed a strong correlation with the storage modulus and loss factor measured by DMA, respectively.
- The gelation point identified by FBG sensors was consistent with DMA and POR results, validating the accuracy of the monitoring.
- The hybrid sensor network successfully monitored the resin's curing process, including gelation and vitrification points.
- Damage Identification:
- The modified PDI algorithm with a dynamically adaptive shape factor significantly improved damage localization accuracy.
- The average Relative Distance Error (RDE) of the modified PDI method was 68.48% and 46.97% lower than the conventional PDI and fused PDI methods, respectively.
- The average Absolute Error (AE) was also substantially reduced, showing 64.89% and 43.10% lower than the other methods.
- The modified PDI method demonstrated superior performance in identifying simulated damage locations.
Limitations and Future Work
- Limitations:
- The paper focuses on specific types of CFRP and damage (simulated). The performance with different composite layups, damage types (e.g., internal delamination), or environmental conditions might vary.
- The "averaged shape factor" and "fused PDI" methods, while improvements, are still less effective than the dynamically adaptive approach.
- The paper primarily focuses on damage localization accuracy, with less emphasis on the real-time processing speed of the modified PDI algorithm.
- Future Work:
- The authors suggest that the embedded PZT sensor network can continuously perform real-time structural health monitoring, implying further development and validation for long-term in-service applications.
- Further research could explore the application of this hybrid sensor network and the improved PDI algorithm to more complex composite structures and various damage scenarios.
- Investigating the real-time computational efficiency of the dynamically adaptive PDI algorithm would be beneficial for practical deployment.
Summary updated