【作者】 李宝玉；

【导师】 戴光； 李伟；

【作者基本信息】 大庆石油学院 ， 安全技术及工程， 2005， 硕士

【摘要】 金属腐蚀所造成的损失很大,其中又以点蚀的危害最严重。由于点蚀的外观隐蔽,很难被探测到,但其破坏性极大,若任其发展,会导致金属穿孔直到整个结构被破坏。本文旨在研究金属点蚀的声发射监测,并试图利用小波分析和独立分量分析来识别点蚀过程中的声发射信号,这一研究的最终目的是为利用声发射技术监测点蚀萌生和扩展服务。这将在降低检测、维修费用、减少环境污染、提高设备的安全性等方面具有重要意义,将会带来巨大的经济效益和社会效益。通过对金属点蚀形成及发展过程的分析表明,点蚀是由(钝化)膜破裂引起的,而伴随膜破裂过程必然会产生应力波一声发射。另外,在点蚀发生时,微小氢气泡的破裂也会产生声发射,其原理与膜破裂相同。通过建立金属表面钝化膜破裂/气泡破裂产生声发射源机制的模型,估算出单个膜破裂所产生的薄板表面位移量级,从而计算出点蚀声发射信号的幅度值,该值远大于前置放大器的输入噪声水平。从理论上证明利用声发射技术可以监测到点蚀产生的声信号,并确定了点蚀声发射信号的参数特征和波形特征,这些对于利用声发射技术监测点蚀损伤都具有指导意义。通过建立有效的实验平台,提取金属点腐蚀的声学信号样本,深入研究点蚀过程中声发射信号的变化规律。试验表明,声发射能比超生和涡流等常规无损检测方法更早地发现材料的腐蚀损伤。通过研究点蚀过程中腐蚀损伤程度及腐蚀声发射信号随腐蚀时间的变化规律,获得了腐蚀损伤与声发射参数之间的变化关系。说明声发射技术可用于探测早期腐蚀、研究腐蚀发展规律及监测和评估腐蚀损伤等方面。结合点蚀声发射信号的特点及工程中对声发射源识别的需要,确定出适合于点腐蚀声发射信号小波分析的小波基;利用小波的多分辨分析对信号进行消噪,详细给出了点蚀声发射信号的降噪算法,通过仿真实验验证了算法的可靠性,实验结果证明它对声发射信号具有良好的去噪效果。提出了较适合于声发射混合信号的分离方法,即小波消噪与独立分量分析相结合的方法,对信号进行消噪与分离;利用Matlab 编制了独立分量分析软件,使独立分量分析更加直观、可靠。更多还原

【Abstract】 Metal corrosion is the most dangerous phenomenon attacking different construction materials, especially pitting corrosion. Pitting corrosion is a type of local corrosion attack as the result of which pits are formed, while the remaining metal surface practically remains untouched. This paper aims to research AE inspecting of metal pitting corrosion, and tries to identify AE signal of pitting corrosion by wavelets and ICA. The ultima intention of this research is serving for inspecting metal pitting corrosion by AE technology, which is important meaning in reducing the expense of checking and maintaining, in lessen environmental pollution, in increasing the security of equipment. By analysis for form and growing of metal pitting corrosion, film breaking arouses pitting corrosion with stress producing. Otherwise, small hydrogen air bubble will bring AE, which is the same theory with film breaking. Mechanism model of AE source was founded, and surface displacement quantity grade was estimated, so that amplitude was calculated. This amplitude is much more than value of noise. It is proved that pitting corrosion AE signal is inspected by AE technology. At that time, parameter characteristic and wave characteristic has been identified. This is instructed to inspect metal corrosion by AE technology. By establishing effective experiment plane, acoustic signal sample of pitting corrosion was distilled. Experiment indicates that AE can find material corrosion damage earlier than ordinary nondestructive testing. By studying change disciplinarian of corrosion degree and AE signal with corrosion time, change relation between corrosion degree and AE parameter has been achieved. This explains that AE technology can be used in detecting forepart corrosion, researching corrosion growing orderliness, inspecting and evaluating corrosion damage. Based on corrosion AE signal characteristic and the need of identifying AE sources, wavelet function which is fit for corrosion AE signal has been confirmed. By using wavelet Multi-Resolution Analysis, signal has been de-noised. De-noised arithmetic was given in detail, and the reliability of arithmetic was validated by imitated experiment. The result proved that de-noise effect was excellent. The way of separating AE mixed signal has been brought forward. This way is that ICA is combined with wavelet. Separating effect is excellent. ICA software makes ICA much more credibility.更多还原