基于时域残余力向量特征分解的结构损伤识别方法研究
Structural damage identification method based on the time domainresidual force vector
-
摘要: 提出基于时域残余力向量的结构损伤识别方法。从结构振动方程出发,推导强迫振动条件下由结构动力响应求解时域残余力向量的表达式。时域残余力向量包含与损伤过程相关的时间信息,可以用来识别损伤发生的时刻。通过结点时域残余力向量值的变异系数建立损伤位置指标,识别损伤位置。建立结构残余力矩阵特征分解方法,特征分解获得的特征向量在损伤前后不发生改变,而特征值与单元刚度相关。因此,可以从特征值中提取单元刚度参数构建损伤程度参数,直接识别损伤程度。将时域残余力向量方法应用于武汉军山大桥模型的损伤识别,结果表明:提出的方法在单点激励和多点激励下均可高精度地识别出损伤发生的时刻、位置和程度。提出的方法通过对时域残余力向量特征分解提取损伤时刻、位置和程度指标,计算过程直观且精度高,避免了繁琐的时频变换、优化迭代等计算过程。Abstract: It is presented a method for structural damage identification based on the time-domain residual force vector. Starting from the structural vibration equation, an expression for the time-domain residual force vector is derived from the dynamic response of the structure under forced vibration conditions. The time-domain residual force vector contains the time information related to the damage process and can identify the moment of damage. Afterwards, the damage localization indicator is established by the coefficient of variation of the time domain residual force vector of the nodes to determine the damage location. The structural residual force matrix is eigen-decomposed, whereby the extracted eigenvectors are invariant before and after the damage and the eigenvalues are a function of the damage level. Therefore, the damage level indicator is constructed by extracting the elemental stiffness parameters from the eigenvalues to directly identify the damage extent. This paper applies the time-domain residual force vector method to the damage identification of the Wuhan Junshan Yangtze River Bridge model, which can identify the time, location, and damage level of damage occurrence with high accuracy under both single and multiple point load inputs. The method extracts the damage time, location, and extent indicators by decomposing the time-domain residual force vectors. The calculation process is intuitive and highly accurate, avoiding the complex time-frequency transformation and optimization iterations.