【作者】 胡昌洋；

【导师】 黄振峰；

【作者基本信息】 广西大学 ， 机械电子工程， 2008， 硕士

【摘要】 疲劳裂纹是构件断裂事故的主要根源之一,混流式水轮机转轮叶片经常出现大量贯穿性疲劳裂纹,严重危及电站的安全运行。混流式水轮机转轮叶片出现裂纹或断裂的事故仍屡见不鲜,多在停机检修时,用超声、磁粉探伤等方法检测转轮裂纹,而对裂纹在线监测,动态的预测裂纹扩展趋势仍是一项技术难题。目前所遵循的水轮机转轮定期检查的规定已越显弊端,对转轮叶片的裂纹状态进行监测已经越来越迫切。本文以国家自然科学基金项目(50465002)——“混流式水轮机转轮叶片裂纹监测的理论和方法研究”为背景,研究疲劳裂纹扩展趋势预测;利用声发射技术对疲劳裂纹进行动态监测,根据裂纹的声发射信号对裂纹的状态进行分析;对转轮安全评估方法研究,为实现水轮机转轮状态监测奠定理论基础。论文首先介绍了疲劳裂纹扩展机理、疲劳裂纹扩展预测方法,利用灰色模型和神经网络对疲劳裂纹扩展进行预估。针对水轮机转轮叶片产生的疲劳裂纹,通过三点弯曲疲劳试验,采用声发射技术对金属疲劳裂纹的产生、发展、断裂过程进行监测,与传统的裂纹检测技术相比,声发射不仅能够实时的捕捉到疲劳裂纹的产生,而且能够得到与理论上疲劳裂纹扩展速率曲线相似的变化规律,建立了裂纹声发射信号特征参量与应力强度因子之间的数学关系式,并利用人工神经网络对裂纹的扩展阶段声发射源进行了模式识别,判断裂纹的危害程度。介绍水轮机转轮结构、裂纹产生的位置,分析裂纹产生的原因和提出预防裂纹产生的措施。根据混流式水轮机转轮裂纹的实际情况,研究基于声发射和振动技术对转轮叶片裂纹在线监测的理论方法,并对转轮进行疲劳寿命计算和安全评估基础方法研究,为实现对转轮在线监测和故障诊断奠定理论基础。更多还原

【Abstract】 Fatigue crack is one of the main sources of component fault accident, Francis turbine runner blades often run a lot of fatigue cracks; seriously endanger the safe operation of power plants. The accident of Francis turbine runner blade cracks or faults is still commonplace, but the measures of detecting runner cracks, such as the ultrasound, magnetic detection, the Francis turbine need to be shut down at maintenance time, the crack monitoring on line and dynamic forecast crack growth trend is still a technical problem. At present, the shortcomings of turbine runner inspection by regular have been more and more significant, and the state of the runner blades cracks monitoring has become increasingly urgent.In this paper, the project ’Research of Theory and Method on Monitoring Crack of Francis Turbine Runner Blade’ supported by national fund of natural science (50465002) as the background, research on fatigue crack propagation forecasting approach; using acoustic emission technology dynamic monitoring fatigue crack, according to analysis by the crack of the AE signal to analysis the crack state; application acoustic emission and vibration technology to runner safety assessment, which established the basis of theory in realizing on-line monitoring the status of the turbine runner.Firstly, the fatigue crack growth mechanism and the fatigue crack growth prediction method were discussed. Also, the grey model and neural networks were used for the fatigue crack growth forecast, and gives the turbine crack forecast method, and made the adaptation Analysis.Faced to Francis turbine runner blades the fatigue crack on the fatigue crack growth mechanism, using acoustic emission technology detect metal fatigue crack on the rise, development, fracture process, through the stretch and three-point bending fatigue test to prove that with the traditional Compared to detect cracks, AE can not only real-time capture of a fatigue crack, and can be theoretically and fatigue crack growth rate curve similar to the changes in the law. The introduction of the AE parameters and crack growth rate between the formula and the establishment of a crack AE signal characteristics and parameters of stress intensity factor between the mathematical relationships. Neural networks were used to the crack expansion phase of the pattern recognition.After analyzing the turbine runner structure, the crack of the causes and prevention measures to crack the traditional overhaul of the deficiencies, and the turbine runner crack the online monitoring and fault diagnosis of necessity were discussed, too. According to Francis turbine runner crack the overhaul of the actual situation, to launch technology based on AE and vibration technology to conduct a security assessment runner, and the runner-line monitoring and fault diagnosis.更多还原