Condition monitoring of composite structure with Fibre Optic Sensors

Abstract

The usage of composites in aircraft increases the fatigue resistance and the strength to weight ratio of the structure, but their anisotropic mechanical properties (low strength along the thickness direction) make the composite structures sensitive to the impact damage. This has resulted in a conservative damage tolerance design and the potential interest in utilising a sensor network integrated with the structure to realize real-time structural health monitoring (SHM) for the composite structures. Different types of sensors can be used to form an SHM sensing system. Amongst them, fibre optic (FO) sensors are regarded as one of the most promising sensors for integration with composite aircraft. They can be used for point sensing, such as Fibre Bragg Grating (FBG) sensors, or continuous sensing, such as Rayleigh backscattering sensors. However, the application of FO sensors in SHM has its own challenges; e.g., they are passive sensors and can only measure the changes in strain, temperature, and/or load in the host structure, which is challenging for damage detection under the operational conditions of the structure, such as an aircraft since it is hard to distinguish whether the strain changes are caused by the external load or internal damage. One solution to overcome this limitation is to form a hybrid system by combining the FO sensors with an active system such as piezoelectric (PZT) transducers, which excite the structure with a given waveform and use the FO sensors to record the structure’s response. This thesis addresses the challenge of FO sensors’ application in different stages of usage monitoring (damage detection, temperature effect decoupling, real-time load monitoring, and shape sensing). In detail, the major subjects of this research are 1) to address FO sensors’ poor sensitivity, temperature influence, and low robustness under non-uniform external loading caused by the FO sensors’ spectrum distortion in their application in measuring guided wave signals 2) To improve the spatial resolution and thus the accuracy of FO sensor-based shape sensing systems by using distributed FO sensors and their associated demodulation algorithm; and 3) To realize real-time 2D load distribution monitoring in composite plates with the distributed FO sensing network, which goes beyond the current load monitoring system's state of the art from slender structures to more complex structures such as composite plates. First, a novel FO sensor namely the FBG-based FP sensor as well as its measurement system is developed for the hybrid damage detection system (a system combining both piezoelectric actuators and FO sensors). The research reported in this thesis demonstrated that the application of this new type of FO sensors in the hybrid damage detection system has improved the current hybrid system by increasing the FO sensors’ strain sensitivity, compared to the conventional hybrid system which consists of PZT and FBG sensors. The sensor development process is presented from principle to numerical simulation and finally experimental validation of its improved performance in measuring guided wave signals on composite plates under different frequencies. Second, the temperature influence on the new FO hybrid system is also addressed by applying amplitude calibration for both guided wave and strain sensitivity of FO sensors. The guided wave temperature calibration method requires the measurement of local temperature on the structure. Thus, the newly developed FO sensor’s capability in measuring both Lamb wave signal and temperature is also presented. The new hybrid system’s temperature calibration method and the temperature demodulation and decoupling method are described from its principal to experimental validation on composite plates. The new sensor’s resistance to spectrum distortion is also shown and validated experimentally. Finally, this thesis also focuses on improving the FO sensors network’s performance in shape sensing and load monitoring in terms of spatial resolution and a more realistic loading assumption (non-uniform or 2D external load distribution). The spatial resolution is improved by replacing the point sensing FBG with distributed FO sensors: Rayleigh backscattering sensor (RBS) and deriving its demodulation algorithm. The 2D external load distribution is realized by deriving the demodulation algorithm based on plate theory instead of beam theory. In composite aircraft’s load monitoring, strain sensing along different directions is required which is addressed in this work by optimizing the FO sensor network’s configuration.