Design And Optimization Of Pulsed Eddy Current Testing Sensor Of Multi-layer Riveted Structure In Aircraft

Riveted multi-layer structure is one of important components in aircraft frame, and detection of its deeply embedded defects a hot spot problem owing to its complexity in structure. Pulsed eddy current technique can detect such defects effectively. As a new NDT method with square wave as excitement, PEC has the advantage of wide spectrum, high precision, high efficiency, etc.The detection of cracks in the airframe, cracks around rivet, interlayer corrosion in the multi-layer structure is studied in this paper using PEC method. The efforts are focused on PEC probe design and related parameter optimization.Based on the design of traditional cylindrical probe, a new kind of soft magnetic material called amorphous alloy is used and compared with ferrite. The results show that the detection sensitivity of the probe using amorphous alloy is higher than using ferrite.For the detection of cracks around rivets, a rotary probe is designed. Experimental results show the cracks can be detected effectively by this kind of probe.A new type of probe based on two exciting coil has been designed in this paper for the detection of cracks around rivets. Two exciting coil are reversely-connected to produce a focused and intense magnetic field without the need of large current in the coil. A lot of experiments have been done for the optimization of PEC sensor parameters to improve the detection sensitivity of the probe. The experimental results show that the detection sensitivity reaches maximum when the exciting coil is winded 180 turns, the spacing of two exciting coils is 20-30 millimeter, and the angle between coil axis and horizontal direction is 60° ~ 90°.Array probe with eight-channels is designed to scan the defects in riveted multilayer metal structure. The eight TMR sensors are arranged in two rows. A series of experiments are done to study the positioning of TMR sensors. C scan imaging is implemented by extracting differential signal amplitude of each channel as a feature matrix.Finally, some suggestions for future work are presented.