Health monitoring of cylindrical structures using MFC transducers
Date of Issue2016-12-08
School of Civil and Environmental Engineering
Wave propagation techniques are widely used in structural health monitoring (SHM) because of their easily recognizable and controllable characteristics. Using wave propagation in SHM, controllable ultrasonic stress waves activated in the structures that be distorted if there exist discontinuities like cracks, delaminations, and corrosions. The received signals are analyzed, and the discontinuities can be identified. In cylindrical structures such as pipelines, cracks are more likely to occur along the longitudinal (axial) direction, and they can be fatal to the serviceability of the structures. Unfortunately, the conventional ultrasonic crack detection methods which use longitudinal waves are not very sensitive to this type of cracks. The purpose of this research work is to find an appropriate SHM method for cylindrical structures by using surface attached piezoelectric macro-fiber composite (MFC) to generate guided wave in cylindrical structures. MFC transducers oriented at 45˚ against the neutral axis of the specimen are used as both actuator and sensor to generate longitudinal and torsional waves and to pick up the signals, respectively. Firstly, MFC generated torsional wave pack is used for the axially oriented crack growth monitoring of cylindrical structures. Numerical simulations are performed using ANSYS and nodal release method is used to model the progress of crack growth. Experimental studies are conducted to verify the simulation results. Root mean square deviation (RMSD) method is proposed to capture the slight amplitude changes between the signals collected from the specimen with different crack sizes. Both the numerical results and the experimental data suggest that the axial-direction crack propagation in cylindrical structures can be well monitored using this wave propagation approach. The proposed SHM system then extended with an additional piece of MFC transducer. The new system is not only able to pick up the axial crack growth but also able to identify the axial crack position in the cylindrical structure. The crack position is determined by the time of flight of the wave pack, while the crack propagation is monitored by measuring the variation in the crack induced disturbances, namely, the RMSD crack index. Both numerical simulations and experimental tests on aluminum pipes have been carried out for verification. The results demonstrated that the crack position can be identified, and its growth can be well monitored with the proposed approach. Based on the same principle and experiment setup, the detection of crack size and orientation in the cylindrical structure are studied. First, a crack of finite size is induced in a laboratory specimen. Later, the size is gradually increased along various orientations. The effects of the crack size and transmitted waves, captured by the sensor, are correlated with the RMSD values of the torsional wave packs and the longitudinal wave packs. The results show that both size and orientation of the crack can be evaluated based on the proposed method. The system developed in this thesis is easy to setup, cost efficient and able to achieve automatic continuous online monitoring with good results.