Reduction Of Near-surface Dead Zone And Quantitative Detection Of Defects For Pipeline With TOFD

Pipeline is widely used in the industrial.To ensure its safe and stable operation,it is necessary to conduct regular nondestructive testing on the pipeline to discover the defects early.Ultrasonic time-of-flight diffraction(TOFD)is a nondestructive method to detect the defects by using the time difference between tip-diffracted waves.It is suitable for detecting the defects which may occur in pipeline welds and heat-affected zones.TOFD technology takes the travel time of the lateral wave as reference to locate the defect.Since the ultrasonic pulse has a limited bandwidth,the tip-diffracted wave of the near-surface defect is coupled with the lateral wave during detection,inducing the dead zone and resulting in the missing detection of the defect.However,when TOFD circumferential scan is implemented along the pipeline(i.e.,scan direction is along the circumferential direction of the outer surface of the pipeline),the propagation path of direct longitudinal wave(DLW)is not parallel to the curved surface of the pipeline,and the areas above and below DLW are both to detect.According to the propagation characteristics of the ultrasonic beam,taking DLW as the boundary,the travel times of tip-diffracted waves for the defects located above or below DLW are both longer than that of DLW and the defects are hard to detect quantitatively due to the superposition between DLW and tip-diffracted waves.In other words,the layered dead zone is generated at the near-surface of pipeline.In this paper,the formation mechanism of the layered dead zone in pipeline was analyzed theoretically and the main contents including:(1)The spectrum analysis method is used to establish the relationship between flaw depth and harmonic frequency interval and the distance from flaw tip to DLW are calculated.Meanwhile,the relative position between flaw tip and DLW is determined combining with the characteristics of the tip-diffracted waves in circumferential scan image,realizing the quantitative detection of the defects located in different dead zones.Simulated and experimental results show that the range of dead zone in pipeline is reduced by 40% with spectrum analysis method,and the relative quantitative errors of flaw depths are within 7.25%.(2)Considering that the spectrum analysis method only uses the defection information within the effective frequency band and ignores the relevant detection information within the low and high frequency ranges.On this basis,the autoregressive spectral extrapolation method is employed to expand the frequency band of the overlapped signal and compress the pulse width in time-domain to realize the separation of the overlapped signal in the dead zone.Simulated and experimental results show that the range of dead zone in pipeline is reduced by50% with autoregressive spectral extrapolation method,and the relative quantitative errors of flaw depths are within 11.8%.(3)The reference signal is difficult to determine when the deconvolution method is used to separate the overlapped signals since TOFD signals are dispersive in space–frequency distribution.First,the TOFD signals is decomposed into sub-band signals by using wavelet transform.Then,sub-band DLW signals are adaptively selected as the reference signal.Finally,the TOFD signals can be reconstructed with established Wiener filtering and spectral extrapolation methods to achieve the separation of the overlapped signal.Simulated and experimental results show that the range of dead zone in pipeline is reduced by 60% with adaptive deconvolution method,and the relative quantitative errors of flaw depths are within7.4%.