【作者】 李昊；

【导师】 王社良；

【作者基本信息】 西安建筑科技大学 ， 建筑与土木工程， 2020， 硕士

【摘要】 常温奥氏体相形状记忆合金(Shape Memory Alloy,简称SMA)具有独特的超弹性效应,并表现出耐腐蚀、抗疲劳及与其他基体耦合难度低等优点。近年来,超弹性SMA基耗能装置在结构被动减震控制领域得到了广泛的应用,但在SMA的本构模型研究、受控结构的地震反应分析方法以及优化设计等方面仍存在一些不足。本文将预应力SMA拉索安装于一个三层两跨的钢框架缩尺模型中,并结合改进的SMA模型与成长记忆库-自适应免疫遗传算法,进行了有控结构与无控结构的地震反应分析。主要工作和结论为:(1)对超弹性SMA丝进行了多种工况下的单轴循环拉伸试验,考察了循环训练次数、应变幅值和应变速率对其滞回性能的影响,并分析相关力学参数的变化规律。基于超弹性SMA丝力学性能测试结果,在广义粘塑性框架下,对Graesser宏观唯象理论本构模型进行了拓展。新模型考虑了马氏体正/逆相变过程中特征参量的差异以及大应变幅值下的马氏体硬化特性,并通过引入的内变量演化方程,描述应变幅值和应变速率对超弹性SMA滞回性能的影响,仿真结果表明模型的误差在5%以内。(2)设计并制作了一个三层两跨的空间杆系结构缩尺模型,对其进行无控状态与被动控制(安装SMA拉索)下的振动台试验,并对试验的位移与加速度峰值进行评价,结果表明SMA拉索可以有效抑制结构在地震下的位移响应,各层的层间位移降幅最大可达32.3%。(3)为了减少免疫遗传算法中固定的变异率所引起的精度误差,克服动态记忆库在变异概率较大时容易丢失最优值的不足之处,本文基于自适应变异率准则和抗体记忆库的积累机制,设计了一种成长型记忆库-自适应免疫遗传算法。采用该算法对多组二维函数进行了寻优测试,结果表明改进后的算法种群多样性好、收敛速度高、算法精度高、全局搜索能力强。(4)基于Matlab语言,编写了无控及SMA拉索被动控制空间杆系结构的时程分析程序。采用成长型记忆库-自适应免疫遗传算法SMA拉索被动控制系统进行优化设计。仿真分析结果表明:SMA拉索的数量和安装位置对结构地震响应控制效果有较大影响。当使用固定数量的SMA拉索时,拉索布置于两个平面优于一个平面,隔层布置优于集中布置。更多还原

【Abstract】 Austenite phaseshape memory alloy(SMA)in The room temperature has a unique superelastic effect and exhibits the advantages of corrosion resistance,fatigue resistance and low difficulty in coupling with other substrates.In recent years,superelastic SMA-based energy dissipation devices have been widely used in the field of passively seismic control of structures,but there are still some deficiencies in the study of SMA constitutive models,seismic response analysis methods for controlled structures,and optimal design.In this paper,the prestressed SMA cable is installed in a three-layer twospan steel frame scale model,and combined with the improved SMA model and the growth memory-adaptive immune genetic algorithm,earthquakes with controlled and uncontrolled structures are carried out Reaction analysis.The main work and conclusions are:(1)The uniaxial cyclic tensile test was carried out on superelastic SMA wire under various working conditions,the effects of cycle training times,strain amplitude and strain rate on its hysteretic performance were investigated,and the changes of related mechanical parameters were analyzed law.Based on the test results of the superelastic SMA wire mechanical properties,under the generalized viscoplastic framework,the constitutive model of Graesser’s macroscopic phenomenological theory is expanded.The new model takes into account the difference in characteristic parameters during the martensite forward / reverse phase transformation and the martensitic hardening characteristics at large strain amplitudes,and describes the strain amplitude and strain rate to superelasticity through the introduction of internal variable evolution equations Impact of SMA hysteresis performance.The superelastic SMA hysteretic curve is simulated by Matlab/Simulink module,and the predicted results are compared with the experimental results.Simulation results show that the model error is within 5%.(2)A scale model of a three-story,two-span bar system structure was designed and produced,and a shaking table test under uncontrolled and passive control(with SMA cables installed)was carried out,and anlys the displacement and acceleration peaks of the test.The evaluation shows that the SMA cables can effectively suppress the displacement response of the structure under earthquake,Interlayer displacement can be reduced by up to 32.3%.(3)In order to reduce the algorithm error caused by the fixed mutation rate in the immune genetic algorithm and overcome the shortcomings that the dynamic memory library easily loses the optimal value when the mutation probability is large.Then,Added adaptive mutation rate criterion and accumulation mechanism of antibody memory bank in immune genetic algorithm;Designed a growing memory bank-adaptive immune genetic algorithm.This algorithm was used to test multiple sets of two-dimensional functions.The results show that the improved immune genetic algorithm has better population diversity,improved convergence speed and accuracy,and enhanced global search capabilities.(4)Based on Matlab language,The time-history analysis program based of SMA cables’ passively control system is written.The growth memory-adaptive immune genetic algorithm is used to optimize the design of the SMA cable passively control system.The result of simulation analysis shows that the number and positions of SMA cables have a great influence on the the structure’s seismic response.When the number of cables is known,the cable arrangement in two flats is better than one flat,and distributed layout is better than centralized layout.更多还原