【摘要】 通过对地震作用下不同形式的网架结构进行大规模数值计算,以最大节点位移超过结构跨度的1/50为失效准则,分析了其动力失效模式及特征,将决定网架结构动力失效模式的因素提炼为四个方面:结构8P杆件比例、全时程的最大位移、平均单杆塑性应变能以及临失效前平均位移与最大位移的比值。由于网架结构各失效模式之间存在模糊性,利用模糊C-均值方法分析以上四个因素,进而将网架动力失效模式归为三类:失稳型局部失效、强度型整体失效和强度型局部失效。结果表明,模糊C-均值方法可以有效地划分网架结构的失效模式,并将自然语言描述的失效模式转变为数字语言的表述,得到各失效模式的典型数字特征,可用于识别其他网架结构,为结构的性能化设计打下基础。更多还原

【Abstract】 Based on large-scale numerical calculations for different forms of grid structures subjected to earthquake,dynamic failure modes and characteristics were analyzed,assuming the failure criterion is that the maximum displacement exceeds 1/50 of the span.Dynamic failure modes of grid structures are determined by four factors: the proportion of 8P bar,the maximum displacement during full-time history,the average plastic strain energy,and the ratio of the average to maximum displacement before expiration.As the failure mode is fuzzy,the four factors were analyzed using fuzzy C-means method,and the failure modes were classified into three categories: instability-based progressive collapse,strength-based overall collapse and strength-based progressive collapse,which are described by natural and numerical language respectively.The results show that fuzzy C-means method can effectively classify the failure modes of grid structures,and transform the description from natural to numerical.The numerical prototypes of failure modes were obtained,which can be used to recognize other unknown types of grids,and lay the foundation for performance-based design.更多还原