论文部分内容阅读
摘要: 在Vanmarcke修正首次超越破坏理论基础上,提出了利用Gausslegendre积分公式与Gausslegendre积分的点估计求解随机结构在随机地震激励下的动力可靠度方法。仿真分析结果表明,两种方法求得的可靠度均值与变异系数均随结构参数变异系数增大而收敛。尽管Gausslegendre积分法具有很高的计算精度,但维数超过3维时其计算效率降低明显;而Gausslegendre积分的点估计法则可明星改进高维系统的计算效率,同时还具有计算精度高的优点。计算结果表明:Gausslegendre积分法适合于多变量实际工程结构的可靠度分析。关键词: 随机结构; 地震激励;首次超越破坏;Gausslegendre积分; 可靠度
中图分类号:TU311.3; TU973.2+12文献标志码: A文章编号: 10044523(2015)02021106
DOI:10.16385/j.cnki.issn.10044523.2015.02.006
引言
确定性工程结构受随机地震激励的动力可靠度分析已较为成熟,一般可用首次穿越破坏理论求解[1]。然而工程结构在实际建造与使用过程中难免会出现诸如结构尺寸、刚度、阻尼等变化,即使变化较小,有时作为随机因素所产生的结果也是不容忽视的[2],因此将结构的随机性纳入到可靠度分析是十分必要的。
随机结构在随机激励下的可靠度分析一直是个难点,大部分的研究人员采用近似方法求解,若想得到较为精确的结果一般采用Monte Carlo方法,但相比其带来的计算量是得不偿失的。为此,陈颖等[3]基于首次穿越破坏的Possion极值破坏理论,建立了随机结构动力可靠度功能函数,并采用与响应面法结合的JC法对结构的失效概率进行了计算,但是每次计算需要进行迭代求解验算点。文献[4]基于随机有限元摄动法计算了结构在非平稳地震激励下的结构动力可靠度。张义民等[5]也采用摄动法结合Edgeworth级数技术分析了任意分布参数的梁结构刚度可靠性灵敏度问题。乔红威等[6,7]通过Hermite多项式逼近[8]与降维法[9]将多维动力可靠度问题转换成一维问题,具有较高的精度与效率。李杰与陈建兵[10]基于概率密度演化方法求解结构复合随机振动的可靠度,所提方法不需期望穿越阈率等假定,具有较高精度。本文将从另外一个角度去求解结构动力可靠度。即基于Vanmarcke的修正首次超越破坏理论[1],参照文献[4,7]计算无条件可靠度方法,本文提出了采用Gausslegendre积分公式[11]与Gausslegendre积分点估计[12]两种方法将无条件动力可靠度的复杂积分问题转化为点权积的形式求解,提高了计算效率,同时也保证了计算精度。
1确定性结构的随机地震响应分析
6结论
本文在Vanmarcke修正的首次超越破坏理论的基础上,提出了利用Gausslegendre积分公式与Gausslegendre积分的点估计求解随机结构随机地震激励下动力可靠度的两种方法。计算结果表明:
(1)GLM具有很高的计算精度,然而维数高于3维时其计算效率较低。
(2)GLM与GLPEM求得的可靠度均值与可靠度变异系数均随结构参数变异系数增大而收敛,且GLM的收敛性略好于GLPEM。
(3)GLPEM的计算精度略低于GLM,但计算效率改善明显,精度也较高,适合于多变量实际工程结构的可靠度分析。
总体而言,结构参数变量为低维时(不高于3)且结构相对简单,建议采用GLM,以获得较高精度与收敛性;而结构复杂,维数较多时则采用GLPEM以获得较高计算效率。值得注意的是,目前GLM与GLPEM适用于线弹性结构,而对非线性结构的适用情况还有待研究。
参考文献:
[1]朱位秋.随机振动[M].北京:人民交通出版社,1990.
Zhu Weiqiu. Random Vibration[M]. Beijing: China Communication Press, 1990.
[2]吴再光.具有随机参数的动力系统随机地震响应分析[J].应用力学学报,1995,12(1):90—93.
Wu Zaiguang. Stochastic seismic response analysis of dynamic system with random structural parameters[J]. Chinese Journal of Applied Mechanics, 1995,12(1):90—93.
[3]陈颖,王东升,朱长春.随机结构在随机荷载下的动力可靠度分析[J].工程力学,2006,23(10):82—85.
Chen Ying, Wang Dongsheng, Zhu Changchun. Dynamic reliability analysis of stochastic structures subjected to random loads[J]. Engineering Mechanics, 2006,23(10):82—85.
[4]Abhijit Chaudhuri, Subrata Chakrabort. Reliability of linear structure with parameter uncertainty under nonstationary earthquake[J]. Structure Safety, 2005,28:231—246.
[5]张义民,贺向东,刘巧伶,等.任意分布参数的梁结构刚度可靠性灵敏度分析[J].计算力学学报,2007,24(6):785—789. Zhang Yiming, He Xiangdong, Liu Qiaoling, et al. Stiffness reliabilitybased sensitivity analysis of beam structure with arbitrary distribution parameters[J]. Chinese Journal of Computational Mechanics, 2007,24(6):785—789.
[6]乔红威,吕震宇,关爱锐,等.平稳随机激励下结构动力可靠度分析的多项式逼近法[J].工程力学,2009,26(2):60—64.
Qiao Hongwei, Lü Zhenzhou, Guan Xinrui, et al. Dynamic reliability analysis of stochastic structures under stationary excitation using hermite polynomials approximation[J]. Engineering Mechanics, 2009,26(2):60—64.
[7]乔红威,吕震宇.平稳随机激励下随机结构动力可靠性分析[A].2007年第九届全国振动理论及应用学术会议[C].北京:2007:57—62.
Qiao Hongwei, Lü Zhenzhou. Dynamic reliability analysis of stochastic structures under stationary excitation[A]. The Ninth National Conference on Vibration Theory and Application[C]. Beijing: 2007:57—62.
[8]Baroth J, Bode L, Bressolette P, et al. SFE method using Hermite polynomials: An approach for solving nonlinear mechanical problems with uncertain parameters[J]. Computer Methods in Applied Mechanics and Engineering, 2006,195(44):6 479—6 501.
[9]Rahman S, Xu H. A univariate dimensionreduction method for multidimensional integration in stochastic mechanics[J]. Probabilistic Engineering Mechanics, 2004,19(4):393—408.
[10]陈建兵,李杰.复合随机振动系统的动力可靠度分析[J].工程力学,2005,22(3):52—57.
Chen Jianbing, Li Jie. Dynamic reliability assessment of double random vibration systems[J]. Engineering Mechanics, 2005,22(3):52—57.
[11]李庆杨,王能超,易大义.数值分析[M].北京:清华大学出版社,2008.
Li Qingyang, Wang Nengchao, YI dayi. Numerical Analysis[M]. Beijing: Tsinghua University Press, 2008.
[12]Rosenblueth E. Point estimates for probability moments[J]. Proceedings of the National Academy of Sciences, 1975,72(10):3 812—3 814.
[13]林家浩,张亚辉.随机振动的虚拟激励法[M].北京:科学出版社,2004.
Lin Jiahao, Zhang Yahui. Pseudoexcitation Method of Random Vibration[M]. Beijing: Sciences Press, 2004.
[14]Rackwitz R, Flessler B. Structural reliability under combined random load sequences[J]. Computers & Structures, 1978,9(5):489—494.
[15]Rosenblueth E. Twopoint estimates in probabilities[J]. Applied Mathematical Modelling, 1981,5(2):329—335.
[16]范文亮,李正良,王承启.多变量函数统计矩点估计法的性能比较[J].工程力学,2012,29(11):1—11.
Fan Wenliang, Li Zhengliang,Wang Chengqi. Comparison of point estimate methods for probability moments of multivariate fuction[J]. Engineering Mechanics, 2012,29(11):1—11.
[17]Christian J T, Baecher G B. Pointestimate method as numerical quadrature[J]. Journal of Geotechnical and Geoenvironmental Engineering, 1999,125(9):779—786. [18]Tsai C W, Franceschini S. Evaluation of probabilistic point estimate methods in uncertainty analysis for environmental engineering applications[J]. Journal of Environmental Engineering, 2005,131(3):387—395.
[19]张治勇,孙柏涛,宋天舒.新抗震规范地震动功率谱模型参数的研究[J].世界地震工程,2000,16(3):33—38.
Abstract: In this paper, Gausslegendre integral formulation and Gausslegendre integral points estimation method are proposed to solve the reliability of stochastic structure under seismic excitation, based on the firstpassage method modified by Vanmarcke. Numerical simulation shows that Gausslegendre integral method is only suitable for system whose dimension is less than 4, offering great precision and higher computational efficiency. By contrast, Gausslegendre integral points estimate method possesses excellent precision and much higher computational efficiency. Both the mean and variation coefficient of structural reliability solved by those two methods are converged with increase in variability coefficient of structure parameters.
Key words: stochastic structure; seismic excitation; firstpassage failure; Gausslegendre integral; reliability
中图分类号:TU311.3; TU973.2+12文献标志码: A文章编号: 10044523(2015)02021106
DOI:10.16385/j.cnki.issn.10044523.2015.02.006
引言
确定性工程结构受随机地震激励的动力可靠度分析已较为成熟,一般可用首次穿越破坏理论求解[1]。然而工程结构在实际建造与使用过程中难免会出现诸如结构尺寸、刚度、阻尼等变化,即使变化较小,有时作为随机因素所产生的结果也是不容忽视的[2],因此将结构的随机性纳入到可靠度分析是十分必要的。
随机结构在随机激励下的可靠度分析一直是个难点,大部分的研究人员采用近似方法求解,若想得到较为精确的结果一般采用Monte Carlo方法,但相比其带来的计算量是得不偿失的。为此,陈颖等[3]基于首次穿越破坏的Possion极值破坏理论,建立了随机结构动力可靠度功能函数,并采用与响应面法结合的JC法对结构的失效概率进行了计算,但是每次计算需要进行迭代求解验算点。文献[4]基于随机有限元摄动法计算了结构在非平稳地震激励下的结构动力可靠度。张义民等[5]也采用摄动法结合Edgeworth级数技术分析了任意分布参数的梁结构刚度可靠性灵敏度问题。乔红威等[6,7]通过Hermite多项式逼近[8]与降维法[9]将多维动力可靠度问题转换成一维问题,具有较高的精度与效率。李杰与陈建兵[10]基于概率密度演化方法求解结构复合随机振动的可靠度,所提方法不需期望穿越阈率等假定,具有较高精度。本文将从另外一个角度去求解结构动力可靠度。即基于Vanmarcke的修正首次超越破坏理论[1],参照文献[4,7]计算无条件可靠度方法,本文提出了采用Gausslegendre积分公式[11]与Gausslegendre积分点估计[12]两种方法将无条件动力可靠度的复杂积分问题转化为点权积的形式求解,提高了计算效率,同时也保证了计算精度。
1确定性结构的随机地震响应分析
6结论
本文在Vanmarcke修正的首次超越破坏理论的基础上,提出了利用Gausslegendre积分公式与Gausslegendre积分的点估计求解随机结构随机地震激励下动力可靠度的两种方法。计算结果表明:
(1)GLM具有很高的计算精度,然而维数高于3维时其计算效率较低。
(2)GLM与GLPEM求得的可靠度均值与可靠度变异系数均随结构参数变异系数增大而收敛,且GLM的收敛性略好于GLPEM。
(3)GLPEM的计算精度略低于GLM,但计算效率改善明显,精度也较高,适合于多变量实际工程结构的可靠度分析。
总体而言,结构参数变量为低维时(不高于3)且结构相对简单,建议采用GLM,以获得较高精度与收敛性;而结构复杂,维数较多时则采用GLPEM以获得较高计算效率。值得注意的是,目前GLM与GLPEM适用于线弹性结构,而对非线性结构的适用情况还有待研究。
参考文献:
[1]朱位秋.随机振动[M].北京:人民交通出版社,1990.
Zhu Weiqiu. Random Vibration[M]. Beijing: China Communication Press, 1990.
[2]吴再光.具有随机参数的动力系统随机地震响应分析[J].应用力学学报,1995,12(1):90—93.
Wu Zaiguang. Stochastic seismic response analysis of dynamic system with random structural parameters[J]. Chinese Journal of Applied Mechanics, 1995,12(1):90—93.
[3]陈颖,王东升,朱长春.随机结构在随机荷载下的动力可靠度分析[J].工程力学,2006,23(10):82—85.
Chen Ying, Wang Dongsheng, Zhu Changchun. Dynamic reliability analysis of stochastic structures subjected to random loads[J]. Engineering Mechanics, 2006,23(10):82—85.
[4]Abhijit Chaudhuri, Subrata Chakrabort. Reliability of linear structure with parameter uncertainty under nonstationary earthquake[J]. Structure Safety, 2005,28:231—246.
[5]张义民,贺向东,刘巧伶,等.任意分布参数的梁结构刚度可靠性灵敏度分析[J].计算力学学报,2007,24(6):785—789. Zhang Yiming, He Xiangdong, Liu Qiaoling, et al. Stiffness reliabilitybased sensitivity analysis of beam structure with arbitrary distribution parameters[J]. Chinese Journal of Computational Mechanics, 2007,24(6):785—789.
[6]乔红威,吕震宇,关爱锐,等.平稳随机激励下结构动力可靠度分析的多项式逼近法[J].工程力学,2009,26(2):60—64.
Qiao Hongwei, Lü Zhenzhou, Guan Xinrui, et al. Dynamic reliability analysis of stochastic structures under stationary excitation using hermite polynomials approximation[J]. Engineering Mechanics, 2009,26(2):60—64.
[7]乔红威,吕震宇.平稳随机激励下随机结构动力可靠性分析[A].2007年第九届全国振动理论及应用学术会议[C].北京:2007:57—62.
Qiao Hongwei, Lü Zhenzhou. Dynamic reliability analysis of stochastic structures under stationary excitation[A]. The Ninth National Conference on Vibration Theory and Application[C]. Beijing: 2007:57—62.
[8]Baroth J, Bode L, Bressolette P, et al. SFE method using Hermite polynomials: An approach for solving nonlinear mechanical problems with uncertain parameters[J]. Computer Methods in Applied Mechanics and Engineering, 2006,195(44):6 479—6 501.
[9]Rahman S, Xu H. A univariate dimensionreduction method for multidimensional integration in stochastic mechanics[J]. Probabilistic Engineering Mechanics, 2004,19(4):393—408.
[10]陈建兵,李杰.复合随机振动系统的动力可靠度分析[J].工程力学,2005,22(3):52—57.
Chen Jianbing, Li Jie. Dynamic reliability assessment of double random vibration systems[J]. Engineering Mechanics, 2005,22(3):52—57.
[11]李庆杨,王能超,易大义.数值分析[M].北京:清华大学出版社,2008.
Li Qingyang, Wang Nengchao, YI dayi. Numerical Analysis[M]. Beijing: Tsinghua University Press, 2008.
[12]Rosenblueth E. Point estimates for probability moments[J]. Proceedings of the National Academy of Sciences, 1975,72(10):3 812—3 814.
[13]林家浩,张亚辉.随机振动的虚拟激励法[M].北京:科学出版社,2004.
Lin Jiahao, Zhang Yahui. Pseudoexcitation Method of Random Vibration[M]. Beijing: Sciences Press, 2004.
[14]Rackwitz R, Flessler B. Structural reliability under combined random load sequences[J]. Computers & Structures, 1978,9(5):489—494.
[15]Rosenblueth E. Twopoint estimates in probabilities[J]. Applied Mathematical Modelling, 1981,5(2):329—335.
[16]范文亮,李正良,王承启.多变量函数统计矩点估计法的性能比较[J].工程力学,2012,29(11):1—11.
Fan Wenliang, Li Zhengliang,Wang Chengqi. Comparison of point estimate methods for probability moments of multivariate fuction[J]. Engineering Mechanics, 2012,29(11):1—11.
[17]Christian J T, Baecher G B. Pointestimate method as numerical quadrature[J]. Journal of Geotechnical and Geoenvironmental Engineering, 1999,125(9):779—786. [18]Tsai C W, Franceschini S. Evaluation of probabilistic point estimate methods in uncertainty analysis for environmental engineering applications[J]. Journal of Environmental Engineering, 2005,131(3):387—395.
[19]张治勇,孙柏涛,宋天舒.新抗震规范地震动功率谱模型参数的研究[J].世界地震工程,2000,16(3):33—38.
Abstract: In this paper, Gausslegendre integral formulation and Gausslegendre integral points estimation method are proposed to solve the reliability of stochastic structure under seismic excitation, based on the firstpassage method modified by Vanmarcke. Numerical simulation shows that Gausslegendre integral method is only suitable for system whose dimension is less than 4, offering great precision and higher computational efficiency. By contrast, Gausslegendre integral points estimate method possesses excellent precision and much higher computational efficiency. Both the mean and variation coefficient of structural reliability solved by those two methods are converged with increase in variability coefficient of structure parameters.
Key words: stochastic structure; seismic excitation; firstpassage failure; Gausslegendre integral; reliability