ABSTRACT

The ultrasonic measurement model (UMM) has been theoretically studied over the past few decades, but there is little published work on its practical application for guiding defect detection. In this work, we present a model-based ultrasonic method to detect flaws in curved components using immersion of spherically focused transducers. The UMM of focused transducers is further investigated theoretically and verified experimentally, and some key issues in the model with focused transducers are addressed, including the necessity of correctly using effective geometrical parameters and the corresponding transfer function. The UMM is employed to guide the inspection of a magnesium alloy shell component which consists of several curved surfaces. The alignment between the transducer and the shell is conducted by comparing the measured and theoretically predicted interface-echo signals, the water path is selected by simulating the effective detection region, and the system gain is set by quantitatively analysing the amplitudes of flaw signals and noises. When the inspection conditions are determined, a variable threshold based on the predicted flaw signals is introduced to distinguish noise and flaw signals. The ultrasonic C-scan images of the shell are compared with the digital radiography (DR) images, and the comparison shows that the flaws of interest can be detected in the C-scan images, which verifies the effectiveness of the UMM-based detection method.

摘要

超声波测量模型（UMM）在过去的几十年里一直在理论上进行研究，但关于其在指导缺陷检测的实际应用方面的发表工作却很少。在这项工作中，我们提出了一种基于模型的超声波方法，使用浸入球面聚焦换能器来检测弯曲部件中的缺陷。对聚焦换能器的UMM进行了进一步的理论研究和实验验证，并解决了聚焦换能器模型中的一些关键问题，包括正确使用有效几何参数和相应传递函数的必要性。UMM 用于指导由多个曲面组成的镁合金外壳部件的检查。换能器与壳体的对位通过测量和理论预测的界面回波信号进行比较，通过模拟有效探测区域选择水路，通过对缺陷信号和噪声幅值的定量分析确定系统增益. 当检测条件确定后，引入基于预测缺陷信号的可变阈值来区分噪声和缺陷信号。将壳的超声C扫描图像与数字射线照相（DR）图像进行比较，比较表明可以在C扫描图像中检测到感兴趣的缺陷，验证了基于UMM的检测方法的有效性. 系统增益通过对缺陷信号和噪声幅度的定量分析来设定。当检测条件确定后，引入基于预测缺陷信号的可变阈值来区分噪声和缺陷信号。将壳的超声C扫描图像与数字射线照相（DR）图像进行比较，比较表明可以在C扫描图像中检测到感兴趣的缺陷，验证了基于UMM的检测方法的有效性. 系统增益通过对缺陷信号和噪声幅度的定量分析来设定。当检测条件确定后，引入基于预测缺陷信号的可变阈值来区分噪声和缺陷信号。将壳的超声C扫描图像与数字射线照相（DR）图像进行比较，比较表明可以在C扫描图像中检测到感兴趣的缺陷，验证了基于UMM的检测方法的有效性.