杜永峰, 李慧, 苏磐石, 赵国藩. 非比例阻尼隔震结构地震响应的实振型分解法[J]. 工程力学, 2003, 20(4): 24-32.
引用本文: 杜永峰, 李慧, 苏磐石, 赵国藩. 非比例阻尼隔震结构地震响应的实振型分解法[J]. 工程力学, 2003, 20(4): 24-32.
DU Yong-feng, LI Hui, Billie F. SPENCER, Jr.. REAL MODE SUPERPOSITION METHOD FOR ANALYSIS OF SEISMIC RESPONSE OF NON-PROPORTIONALLY DAMPED ISOLATED STRUCTURES[J]. Engineering Mechanics, 2003, 20(4): 24-32.
Citation: DU Yong-feng, LI Hui, Billie F. SPENCER, Jr.. REAL MODE SUPERPOSITION METHOD FOR ANALYSIS OF SEISMIC RESPONSE OF NON-PROPORTIONALLY DAMPED ISOLATED STRUCTURES[J]. Engineering Mechanics, 2003, 20(4): 24-32.

非比例阻尼隔震结构地震响应的实振型分解法

REAL MODE SUPERPOSITION METHOD FOR ANALYSIS OF SEISMIC RESPONSE OF NON-PROPORTIONALLY DAMPED ISOLATED STRUCTURES

  • 摘要: 采用子结构瑞利阻尼模型表达隔震体系的非比例阻尼矩阵,将实振型分解法与拉普拉斯变换方法联合应用,建立了任意多自由度非比例阻尼隔震体系时域动力响应的工程算法。以本文作者承担设计的三个实际隔震工程作为算例,在两种不同输入地震波和三种不同隔震层阻尼水准下,用本文算法计算地震响应,并与复振型分解法、Wilson-θ法、及Matlab下的Simulink这三种方法的峰值响应进行了对比。结果表明,在设计阻尼水准(隔震层阻尼比为0.21)下,本文算法与Wilson-θ算法的精度大致相当,所求得的位移、速度及加速度峰值的最大相对误差分别不超过0.07%、0.19%和0.27%。即使在极大阻尼水准(隔震阻尼比为0.81)下,本文算法所求得的所有峰值响应的最大相对误差均不超过4%。这表明,当隔震阻尼比不是特别大时,本文提出的算法完全可以满足工程计算要求。

     

    Abstract: An approximate method for analyzing seismic response of non-proportionally damped base isolated structures in time domain is derived by combining real mode superposition and Laplace transform method. The non-proportional damping matrix is expressed in the form of sub structural Rayleigh damping model. Three of the isolated buildings which were designed by the authors have been used as numerical examples, inputting two different earthquake waves under 3 different isolation damping levels. A comparison of calculated results is made with complex modal superposition, the Wilson-θmethod coded in double precision Fortran, and the Simulink method in Matlab. It is shown that under the design damping level defined in this paper, i.e., the isolation damping ratio is 0.21, the result of the present method is basically as accurate as the Wilson-θmethod, and the maximum relative errors of the peak displacement, velocity and acceleration are less than 0.07%, 0.19% and 0.27%, respectively, of their maximum response. Even if under an extremely heavy damping level (isolation damping ratio 0.81), the maximum relative errors of all peak responses of interest are still less than 4%, indicating that the method presented in this paper is sufficiently accurate for use in engineering analysis.

     

/

返回文章
返回