缩尺模型混合模拟试验测量误差分析及试验验证

Measurement error analysis and experimental verification of hybrid simulation testing method for reduced scale model

基于缩尺模型子结构混合模拟试验方法,将正态分布随机数引入测量误差,利用MATLAB和Open SEES混合编程技术,分别对剪切型模型、弯剪型模型进行了考虑测量误差的子结构混合模拟数值模拟分析。数值模拟结果表明:在满足JGJ/T 101—2015《建筑抗震试验方法规程》的误差范围内,缩尺比为1/2、1/4缩尺模型的测量误差与足尺模型相应测量误差基本相当。在缩尺比例不小于1/4情况下,随着地震峰值加速度增大,总的累积相对差值呈增加趋势;在相同地震峰值加速度作用下,总的累积相对差值随梁柱线刚度比增加呈减少趋势。通过1/2钢框架子结构混合模拟试验,验证了所提出的缩尺模型子结构混合模拟试验方法的精准性和可行性。

Based on the presented hybrid simulation testing method for the reduced scale model, the numerical simulation of the measurement error, which is simulated by the random number of the normal distribution, on both the shear model and the bending-shear model have been finished by using the multi-language programming technology with MATLAB and OpenSEES. The measurement error used in the numerical simulation analysis, which meets the code of JGJ/T 101-2015 ＇Specification of testing methods for earthquake resistant building＇ , is obtained from the normal distribution random number by using the Matlab. The numerical simulation results demonstrate that the errors from the 1/2 and 1/4 reduced scale model are basically equivalent with the ones from the full-scale model in both the relative error of the peak point and the accumulated error. With the reduced-scale more than 1/4, the accumulated errors increase with the increase of the peak ground acceleration. Meanwhile, the accumulated errors decrease with the increase of beam column linear stiffness ratio under the same peak ground acceleration. The hybrid simulation test of the 1/2 reduced-scale steel frame structure was completed in the laboratory, and the accuracy and feasibility of the hybrid simulation testing method are verified by the test results.