七層樓鋼造二元構架系統複合模擬: 中等韌性箱型鋼柱之一層樓實尺寸子結構試驗

Abstract

在本研究中將會針對複合模擬(hybrid simulation)此實驗方法進行改進。於有限元素分析軟體PISA3D中建立七層樓鋼造二元構架系統模型，在此平面構架中，其中一跨為特殊抗彎構架(SMRF)，另一跨為挫屈束制斜撐構架(BRBF)。所使用的試體為一組全尺寸一層樓鋼造子構架，於實驗過程中，水平自由度使用位移控制，垂直自由度使用力量控制，而根據PISA3D事前分析的結果，旋轉自由度則不用控制也能自動滿足。在進行複合模擬時，一組額外的虛擬力施加在構架中的一層樓柱頂，此虛擬力的主要目的為滿足試體與分析模型中彼此間的位移諧和性。一般來說，PISA3D無法模擬因柱挫屈造成的縮短量，故需要透過額外的虛擬力將模型中的柱縮短至指定位置。中間柱的縮短量透過試體量測而得，而外柱的縮短量則是透過ABAQUS事前模擬得知。實驗結果證明此研究中所提出之新的複合模擬方法可行，相較於傳統的複合模擬，其結果更能反應構架真實的行為。 為了研究整體構架的行為受到中間柱有不同程度的縮短量、地震力方向、二階效應及柱底旋轉效應的影響，進行一系列的虛擬複合模擬(virtual hybrid simulation)，其結果顯示當同時考慮二階效應及柱底旋轉效應且中間柱有顯著縮短量時，一樓水平位移會有顯著的提升，並且當中間柱與其中一根外柱同時有明顯挫屈時，其對一樓水平位移的影響會比只有中間柱有明顯挫屈時來的顯著。 此研究結果顯示進行構架動力歷時分析時，若所有桿件都是由數值模型模擬，與考慮柱真實回復力及柱挫屈造成垂直位移的結果相比，在一樓水平位移上有低估的情形。說明PISA3D在模擬建築物受到地震的情況下，其模擬方法仍有改進的空間。

This research introduces some advancement in the implementation of hybrid simulation (HS). A structure under the consideration is a 7-story dual system with one bay is a special moment resisting frame (SMRF) and the other bay is a buckling-restrained braced frame (BRBF). The finite element analysis program “Platform of Inelastic Structural Analysis for 3D Systems” (PISA3D) is used as the analysis kernel for the HS. The specimen is a full-scale cruciform beam-column connection subassemblage. The displacement- and force-control are adopted to achieve the control target when the first-story served as the physical substructure (PS). Preliminary PISA3D simulations indicates that without control the rotation DOF of column top is acceptable. The fictitious force method is used to address one of the principal challenges in this series of HS: to incorporate the column shortening due to local buckling into the HS, while the PISA3D does not support the modelling. In addition, the HS considers shortening of exterior first-story columns by incorporating preliminary ABAQUS simulation results in each integration step in the HS. Test results confirmed that the proposed modelling and control methods could successfully integrate the information available only in the laboratory and ABAQUS simulation into the HS, and provide more realistic structural response that otherwise cannot be obtained by traditional numerical simulation or testing methods. Moreover, to investigate the influence of the different level of shortening in interior column, direction of the excitation, P-Delta, and column base rotation on the first-story horizontal displacement, a well calibrated specimen model in the ABAQUS is adopted to generate the interior column restoring force in a series of the virtual hybrid simulation (VHS). The analysis result indicates that the influence of P-Delta and the column base rotation on the first-story horizontal displacement can be significant with the severe column shortening in the interior column, and when the significant column shortening due to local buckling occurred in the interior and one of the exterior column at the same time, the influence on the first-story horizontal displacement was more major than that when only interior column performed significant shortening. This research indicates the results of first-story horizontal displacement from the PISA3D analytical analysis are underestimate compared to the results from the HS which considered the specimen actual restoring force and vertical deformation due to column local buckling which shows that the PISA3D still has room for improvement in its simulation method when simulating buildings subjected to earthquakes.