摘要
研究了采用声发射技术检测大型氧化剂储罐腐蚀损伤状态的可行性。根据液体火箭氧化剂储罐主要的结构材料5A03铝合金在实际使用中的腐蚀机理,选取不同浓度的硝酸作为腐蚀介质,建立5A03铝合金腐蚀的试验方案,利用声发射技术对腐蚀过程进行监测,获得了各浓度水平下的声发射信号。试验结果表明,声发射信号撞击数的多少能够反映合金不同的腐蚀损伤程度,不同浓度硝酸中5A03铝合金腐蚀声发射信号的上升时间、持续时间、振铃计数、能量等特征参数的分布具有较大差异,可通过90%的分布区间加以区分。利用所建立的BP神经网络能够以很高的正确率对5A03铝合金储罐腐蚀损伤程度进行模式识别。
The study is made to prove the feasibility of using the acoustic emission technology to detect tbe corrosion destruction condition of the liquid rocket oxidant vessels, whose main structure material is 5A03 aluminum alloy. The simulative test project using different concentrations of nitric acid to corrode 5A03 alloy is made according to the corrosion mechanism of the oxidant vessels in use. The corrosion process of 5A03 aluminium alloy is monitored by the acoustic emission technology for the first time in this article. The results show that the hits number can reflect the corrosion destruction condition of the alloy, and the distribution of the characteristic parameters, such as rise time, duration, ring-down counts and energy, is different obviously, and can be differed by 90% distribution interval. At last, the BP artificial neural network is build up to judge the corrosion damage degree with high accuracy, and the acoustic emission technology is proved feasible and predominant to inspect the corrosion condition of the oxidant vessel.
出处
《无损检测》
2011年第11期24-28,共5页
Nondestructive Testing
关键词
铝合金
声发射检测
腐蚀
特征分析
神经网络
Aluminum alloy
Acoustic emission testing
Corrosion
Characteristic analysis
Artificial neural network
