The magnetic flux leakage(MFL) type non-destructive testing(NDT) system is widely used to detect metal loss of the underground gas pipelines. In the system, sensor module is consisted of permanent magnet, magnetic yoke and hall sensors to detect corrosion defect or any other damages of the gas pipeline.
To increase the magnitude of the sensing signals, it is necessary to increase the change of the magnetic leakage flux in the region of defect. The optimal design method of the magnetic system with permanent magnet and yokes is described. In case the operating point on the magnetic saturation curves of the object is too low, the object will not be magnetically saturated in the defect region, so the defect signals become weak. In case it is too high, the change of the magnetic flux in the defect region will be small, so the amplitude of the sensor signal becomes weak. The operating point of the magnetic system is optimized so as to maximize the change of the magnetic flux in the region of the defect.
During the measurement, average speed of the PIG module is 4～5 m/sec. But, in most cases, the speed of the PIG module is not constant and varying inevitably from 0 to over 10 m/s because of the irregular geometry of the underground pipeline such as curvature, joint, and wrinkle structures. So, it is necessary to compensate the velocity induced distortion signals as to obtain the pure defect signals from measured signals. The method to deduce the speed of the PIG module from the sensing signals are described and the compensation scheme to eliminate the velocity induced signal distortions are developed.
In each leg, a magnetizing yoke and magnet were equipped with 3 sets hall sensors to detect the MFL signals. For the measurement, we made a gas pipe of 30 inches diameter with several types of artificial defect. Artificial rhombic defect could be successfully identified from the defect signals. The computed MFL signal obtained by a nonlinear finite element method is verified by actual measurements.