基于DSF模式的在役承压设备早期损伤识别与评价研究

项目名称： 基于DSF模式的在役承压设备早期损伤识别与评价研究

项目编号： No.51205133

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 机械工程学科

项目作者： 胡明慧

作者单位： 华东理工大学

项目金额： 25万元

中文摘要： 结构损伤是在役承压设备（如反应器、塔器、储罐、压力管道、锅炉等）常见失效模式，早期疲劳损伤在复杂环境下很可能快速发展为大损伤,对承压设备结构安全具有很大的潜在危害。面向在役承压设备安全可靠运行的需求，针对其在结构损伤识别与评价中的基础科学问题展开研究。主要内容包括：(1)考察控制模式（应力、应变、刚度等参数）对损伤评定的影响，形成早期疲劳损伤宏细观规律评定的一般方法。（2）引入DSF（Damage Sensitive Feature）损伤敏感特征模式，基于连续损伤力学理论，探索从微观到宏观的统一损伤特征量模型，建立判别损伤位置和损伤程度的DI（Damage Index）损伤因子模型。（3）揭示高维特征空间中损伤表征量之间的交互作用及其对早期损伤的影响规律，建立基于损伤特征提取的在役承压设备早期损伤识别与评价理论。本项目的成果对在役承压设备安全评定、早期损伤识别及准确的工程方法奠定基础。

中文关键词： 承压设备；损伤识别与评价；DSF 模式；动态参数；特征提取

英文摘要： Structural damage is familiar failure mode for inservice pressure equipments such as reactors, tower vessels, storage tanks, pressure pipings, pressure boilers. In complex environment, the early fatigue damage detection may be quickly extended into large damage. Large potential danger exists in structural safety of pressure equipments. Based on the safety reliability of inservice pressure equipments, we research the basic scientific questions of damage detection and assessment. The main research as: (1)We investigate the structural damage evaluation correlation with the control patterns such as stress, strain, stiffness mode. Based on macroscopic and microscopic constitutive, the early fatigue damage evaluation is defined by using the physical and mechanical parameters. (2)Based on damage sensitive feature (DSF) pattern and continuum damage mechanics theory, a computationally attractive damage index (DI) method is proposed for structural damage detection. The DI model can successfully locate damage and quantify the damage level. The unified damage characteristic model is found from microstructure to macrostructure. (3)It provides an insight to the structural damage correlation with the characteristic parameter in high dimensional feature space. Based on parameter feature extraction, the early damage detection an

英文关键词： Pressure equipments；Damage detection and assessment；DSF pattern；Dynamic parameter；Feature extraction