Realization of Semi-active Control by MR dampers in Multi-Story Building Structures

在第一年，设计并制造了最大阻力接近300 N的小型MR阻尼器。通过一系列正弦加载试验，验证了阻尼器力学性能的稳定性。同时也证实了阻尼器的分析模型可以用一个简单的阻尼器和一个可变库仑滑块的组合来表示，这就是所谓的宾汉模型。通过比较随机加载试验得到的荷载-挠度关系，验证了模型的准确性。在进行半主动控制的缩尺钢框架振动台试验之前，研究了框架单独和设置被动粘滞阻尼器后的阻尼特性。在此基础上，将磁流变阻尼器安装在车架上，验证了半主动控制的有效性。结果表明，该阻尼器的唯一缺点是阻尼器的阻尼比往往滞后于输入信号来控制阻尼器的阻尼能力。次年，又研制了一种小型MR阻尼器，用于基础隔震建筑模型的半主动控制。阻尼器的力学行为再次表示为宾汉模型。对于自振周期为2-3秒的隔震建筑模型，其电阻对输入控制信号的时滞几乎可以忽略。在这种情况下，两种类型的半主动控制算法的使用和比较。一种是瞬时最优控制算法，另一种是基于输入建筑物的半周地震能量和阻尼器吸收的滞回能量之间的平衡。

In the first year, a small MR damper with the maximum resistance of nearly 300N was designed and manufactured. Through a series of sinusoidal loading test, the stability of mechanical behavior of the damper was confirmed. It was also confirmed that the analytical model of the damper could be represented by a simple combination of a dashpot and a variable Coulomb slider, which is called a Bingham model. The exact ness of the model was confirmed by comparing the two load-deflection relations, experimental and simulated ones, obtained by a random loading test. Prior to the shake table test of a scaled steel frame subject to semi-active-control, damping property of the frame alone and the one with passive viscous dampers are studied. Then, the MR damper was installed into the frame and the effectiveness of semi-active control, simulating origin-restoring friction damper, was confirmed. It was found that the only negative point with the damper was that the resistance of the damper tends to be delayed to the input signal to control the capacity of the damper.In the following year, another small MR damper to perform semi-active control on base-isolated building model was manufactured. Mechanical behaviors of the damper were again represented by the Bingham model. In the case of base-isolated building model with the natural period of 2-3 seconds, the time lag of the resistance with regard to input control signal was found almost negligible. In this case, two types of semi-active control algorithms were made use of and compared. One is an instantaneous optimum control algorithm and the other is based on the balance between the half-cycle input seismic energy into the building and the hysteretic energy absorbed in the damper.