论文部分内容阅读
首先对响应面法在有限元模型修正中应用的主要问题,包括实验设计、参数筛选、响应面拟合及参数优化等进行了讨论,进而基于现场动力试验数据对一刚构-连续组合梁桥—靖远黄河大桥的有限元模型进行了修正。选取箱梁和桥墩的弹性模量以及连续墩顶支座及边跨联端支座的横向和竖向刚度为修正参数,采用中心复合设计法进行实验设计、计算样本值,分别构造两个响应面模型,其中一个考虑交互项的影响、一个不考虑交互项的影响,然后以这两个响应面模型代替初始有限元模型分别进行修正。修正结果表明,考虑和不考虑交互项的影响对修正后部分参数的取值有一定的影响,而对频率的修正精度影响不大。两种模型修正后的频率取值均与实测频率较吻合,修正后的模型能准确地反映桥梁的动力特性,可作为后续损伤识别及状态评估的基准模型。
Firstly, the main problems of response surface method, such as experimental design, parameter screening, response surface fitting and parameter optimization, are discussed. Then based on field dynamic test data, a rigid-continuous composite beam bridge - The finite element model of Jingyuan Yellow River Bridge was amended. Elastic modulus of box girder and bridge piers and transverse and vertical stiffness of continuous pier top support and side span support joint are selected as correction parameters. The central composite design method is adopted to design the experiment. The sample values are calculated and two responses are constructed respectively One model, considering the effect of interaction terms, does not consider the influence of interaction term, and then corrects these two response surface models instead of the initial finite element model respectively. The results of the correction show that considering the influence of the interaction term has some influence on the value of the modified part of the parameters, and has little effect on the accuracy of the frequency correction. The modified frequencies of the two models are in good agreement with the measured frequencies. The modified model can accurately reflect the dynamic characteristics of the bridge, and can be used as a reference model for subsequent damage identification and status assessment.