【作者】 李力；

【导师】 汪小凯；

【作者基本信息】 武汉理工大学 ， 汽车运用工程， 2020， 硕士

【摘要】 随着我国工业经济迈向高质量发展的新阶段,《中国制造2025》提出将产品质量作为建设制造强国的生命系。大型环件已经广泛应用于能源石化、航空航天、国防军工等重要工程领域。大型环件成型缺陷若不能被及时发现,将会严重威胁相关装备的服役安全。目前,国内大型环件企业基本采取人工携带便携式探伤设备的方式对环件进行缺陷检测。然而该方法受检测员综合素质影响较大,易造成缺陷的错判漏判,检测效率低,可靠性差。因此需要开发一种针对大型环件自动化超声检测系统。本文针对超声无损检测技术高效率、高精度识别大型环件内部缺陷的问题,开展了以下几个方面研究:(1)为提高大型环件自动化检测效率,首先确定了阵列超声无损检测方法;然后基于多元高斯声束模型,研究不同频率阵列超声探头在不同水层、多介质中的声场分布规律,确定高低频扫查探头个数和扫查深度,并进行合理的传感器阵列和结构设计,提出了阵列探头的扫查方式和运动轨迹,绘制自动化检测时识别缺陷的DAC曲线。(2)为提高大型环件自动化检测精度,在COMSOL软件的声学仿真模块中建立水/钢双介质的简化模型,研究超声波从被发射至被接收的传播历程,分析对检测信号能产生干扰的因素,提出调整脉冲发射时序的方法消除干扰,降低自动化检测的缺陷误判率,提高检测质量。(3)由于超声脉冲宽度及尾部震荡使脉冲反射法存在检测盲区,造成无法准确识别大型环件近表面缺陷的问题,提出一种基于傅里叶变换相位谱差值的频率周期识别盲区缺陷的方法。首先由理论公式推导出相位谱差值的周期频率与盲区缺陷位置关系,通过数值仿真模拟多个缺陷和噪声干扰时的盲区信号,结合脉冲频带宽度提取相位谱差值有效信息,借助傅里叶变换的性质求出有效信息的频谱周期,最后结合频谱周期与盲区缺陷位置关系,实现盲区内的缺陷检测。(4)根据环锻件提出的要求和系统的性能指标设计了总体方案,搭建了大型环件的自动化检测系统,设计系统运行的基本流程,并基于VC++开发了多通道阵列超声自动化检测软件系统。该软件采用了多线程编程技术、双缓存绘图技术以及波形的动态跟踪技术,有效提高了软件的数据读取速度、图形绘制速度和缺陷识别的精度。通过对某大型矩形环件的实际检测,能够快速、有效检测出深度2mm-200mmΦ2mm当量缺陷。更多还原

【Abstract】 As China’s industrial economy moves towards a new stage of high-quality development,"Made in China 2025" proposes to use product quality as the lifeline of building a manufacturing power.Large ring parts have been widely used in important engineering fields such as energy and petrochemical,aerospace,and defense and military industries.If large-scale ring forming defects cannot be found in time,it will seriously threaten the service safety of related equipment.At present,most of the large ring parts in China use hand-carried portable flaw detection equipment for defect detection.This method is greatly affected by the comprehensive quality of the inspector,and it is easy to cause the wrong judgment and omission of the defect.The detection efficiency is low and the reliability is poor.Therefore,it is necessary to develop an automatic ultrasonic detection system for large rings.This paper aims at the problem of high-efficiency and high-precision identification of internal defects in large rings by ultrasonic non-destructive testing technology.(1)In order to improve the automatic detection efficiency of large-scale rings,the array ultrasonic non-destructive testing method was first determined;then based on the multivariate Gaussian beam model,the sound field distribution laws of different frequency array ultrasonic probes in different water layers and multiple media were studied for reasonable reasons.Sensor array and structural design;eliminate the interference between different ultrasound probes by adjusting the pulse emission timing.(2)In order to improve the accuracy of large-scale ring automation detection,a simplified model of water / steel dual media is established in the acoustic simulation module of COMSOL software.The propagation process of ultrasonic waves from being transmitted to being received is studied.The factors that can interfere with the detection signal are analyzed.The method of adjusting the pulse emission timing eliminates interference,reduces the false error rate of automated detection,and improves the quality of detection.(3)Aiming at the detection dead zone caused by the pulse reflection method,it is difficult to identify near-surface defects of large rings and cause missed detection.A method of identifying blind zone defects based on the frequency cycle of the phase spectrum difference of the FFT transform is proposed.First,the relationship between the periodic frequency of the phase spectrum difference and the position of the blind spot defect is derived from a theoretical formula.The blind zone signal when multiple defects and noise interference are simulated by numerical simulation,and the effective information of the phase spectrum difference is extracted by combining the pulse band width.The spectrum period of valid information is obtained by nature,and finally the combination of the spectrum period and the position of the blind spot defect is used to realize the defect detection in the blind spot.(4)Based on the process planning of large-scale ring ultrasonic testing,a multichannel array ultrasonic automatic detection system for large ring based on VC ++was developed.The software uses multi-threaded programming technology,doublebuffer drawing technology,and waveform dynamic tracking technology,which effectively improves the software’s data reading speed,graphic drawing speed,and accuracy of defect identification.Through the actual inspection of a large rectangular ring,it can quickly and effectively detect the equivalent depth of 2mm-200mmΦ2mm defects.更多还原