摘要
针对索杆张力结构的施工张拉设计,建立了一种定量评价不同施工张拉方案对索长误差效应控制效果的分析方法。假定索长误差变量服从正态分布,分别针对原长控制法和索力控制法推导了索长误差和预张力偏差的解析关系式,进而给出了随机预张力偏差的均值和方差计算式。分析了主动索的合理选择能够减小索长误差效应的原因。以控制结构最大预张力偏差水平为目标,提出了一种基于遗传算法的主动索优选策略。以实际索杆张力结构为例,利用建议方法分析了该结构在不同张拉方案下预张力偏差的分布特性,并对张拉方案进行了优化分析。计算结果表明:索长误差对索杆张力结构初始态预张力造成的偏差不可忽视,合理选择张拉方案是控制索长误差效应的重要措施。
For the construction pretensioning design of cable-strut tensile structures, a method was developed to quantitatively evaluate the control effect of pretensioning scheme on the structural response caused by cable length errors. The variable of cable length error was assumed to approximately yield the Gaussian distribution. The relationships between cable length error and pretension deviation, respectively for length-control pretensioning method and force-control pretensioning method, were deduced in analytical form. Further, the expressions of mean value and variance of random deviation of structural pretensioning forces were put forward. The reason why the effect of cable length errors can be reduced by the effective choice of active cables is interpreted theoretically. Based on the Genetic Algorithm, an optimization method for choosing the active cables was further suggested in order to control the maximum pretension deviation. A practical cable-bar tensile structure is employed as an illustrative example, and its properties of pretension deviation distribution corresponding to different pretensioning schemes were analyzed by the method suggested in this paper. An optimization analysis of pretensioning scheme was also carried out. The results indicate that the pretension deviation caused by cable length errors is not negligible for cable-bar tensile structures. But this effect of cable length errors can be effectively controlled under a reasonable pretensioning scheme.
出处
《建筑结构学报》
EI
CAS
CSCD
北大核心
2012年第5期71-78,共8页
Journal of Building Structures
基金
国家自然科学基金项目(50978226)
关键词
索杆张力结构
随机索长误差
预张力偏差
遗传算法
张拉分析
cable-bar tensile structure
random length errorpretensioning analysisof cable
pretension deviation
genetic algorithm
