桁架圍束式挫屈束制支撐構架耐震設計分析與試驗研究

Abstract

挫屈束制支撐 (buckling-restrained brace, BRB) 能經濟且有效地提升結構之勁度、強度、韌性與消能行為，BRBF 已廣泛運用於建築中做為主要耐震結構系統。桁架圍束式挫屈束制支撐 (truss-confined BRB, TC-BRB) 之圍束單元由中央圍束鋼管於其周圍配置任意數量、方向與尺寸之剛性桁架所構成之新型 BRB，藉由桁架構架幾何配置提供所需之撓曲剛度，因此得以減低含內灌砂漿之中央鋼管斷面撓曲剛度；進而減少材料用量與自重同時維持 BRB 容許發展之強度；此優勢特別利於具長跨與高軸力容量斜撐構件之應用。為推廣此新型 BRB，本研究承襲前研究者之理論模型與設計方法，於國家地震工程研究中心多軸向試驗系統執行第二階段新造兩組具不同桁架圍束系統型態、1/5縮尺總長 6.3 米、100噸級之 TC-BRB 試體 2CT 與 2VT 之反覆加載試驗。為更一步探討 TC-BRB 做為長跨巨型斜撐實際運用於高層結構之可行性，本研究提出一幢 23 層三維建築，每六層採之字形配置長跨 TC-BRB，由特殊抗彎架 SMRF 與 BRBF 組成之二元系統結構設計例，建立耐震設計流程，並各採 21 組三向地震進行非線性歷時分析，以探討在三種不同地震危害度下之耐震性能。 構件第二階段試驗結果顯示，2CT 與 2VT 其挫屈破壞強度 Plim 分別為 1633 與1647kN，而理論模型對於 Plim 之預測最大誤差為 17%，比對第一階段試驗結果證實殘餘應力的影響必須量化，預測模型亦須改良；但兩試體最大核心應變皆達 2.23% 弧度，其累積塑性應變 CPD 累積亦超過 200，滿足規範需求並展現較第一階段更優良之穩定性表現。本階段試驗比對有限元分析結果，顯示藉考量殘餘應力效應之有限元素模型能準確模擬之變斷面試體 2VT 之反應，誤差僅 1.5%。 23 層設計例前三振動週期分別為 2.12(45°)、2.10(135°) 及 1.15秒(旋轉)，顯示結構系統有較大的側向勁度；且前三模態之質量參與比例各約為 75%。21 組歷史性三向地震放大至台北二區DBE放大因子在 2.61 至 8.34 之間，動力分析結果顯示在三個地震危害度作用下，各樓層最大層間位移角SLE下大部分樓層小於 0.5% 弧度，而高樓層局部樓層則接近 1% 弧度；而 DBE、MCE 則不超過 2% 弧度；而跨六層之 BRBF 各層最大層間位移角為 0.400、0.958、1.192% 弧度，顯示結構體每六層內未受斜撐束制樓層有局部較大反應產生，但驗證 TC-BRB 構件試驗所採設計側位移角 1% 弧度為合理假設。而最大基底剪力分別為 2832、4652、5158 噸，分別為採 I = 1.25 計算 LRFD 設計基底剪力 2560 噸之 1.14、1.69、1.84倍，而 BRBF 於 DBE 下佔整體基底剪力 81%；BRBF 角柱軸力 DCR 於 DBE 與 MCE 下各層最大平均值分別達 0.82 與 0.86；轉換層邊梁水平拉或壓力設計考量可採用 0.5 倍上下層 BRB 最大設計拉力強度之水平合力。累積塑性變形 (CPD) 於 MCE 下單支 TC-BRB 最大累積至 50.1。非線性歷時分析結果驗證所提設計流程與分析模型之可靠性，與跨樓層斜撐配置之結構體其高效率與經濟性。有限元素模型分析實尺寸降伏強度為 1500 噸，40 米長跨 TC-BRB 結果顯示自重為 58 噸，前兩自然振動頻率為 1.91 與 2.53 赫茲，本研究根據兩階段試驗提出簡化設計流程提供使用者快速設計桁架圍束單元。

Buckling-restrained brace (BRB) can effectively improve the stiffness, strength, ductility and energy dissipation capability of the structures, and BRBF is widely used in structures as the main seismic structure system. Truss-confined BRB (TC-BRB) is a novel type of BRB whose restrainer is composed of several steel open-web truss frames outside a central steel casing. Properly configuring the truss frames, TC-BRB’s restrainer can effectively develop the overall restraining rigidity. Thus, the cross section of the central steel casing and the infilled mortar in the TC-BRB can be much lighter than those in a conventional BRB. The reduction of the overall material and self-weight is particularly advantageous in the cases of long-span and large axial capacity BRB designs. Based on the test results, a theoretical model and design procedures developed in the previous research, this study conducts the second phase of the cyclic loading test program. Two 1/5 scaled TC-BRB specimens, each of 6.3m long with 100 tonf nominal yield strength in the constant- and varying-depth truss designs were tested in NCREE. In order to illustrate the feasibility of using TC-BRBs as the long-span mega truss braces in high-rise buildings, a 23-story steel structure is exampled. In this three-dimensional SMRF and BRBF dual system, the TC-BRBs are zigzag-configured across five or six stories in the exterior elevations. Seismic design requirements of the SMRF, the setback of the floor beams, the lateral support requirements of the corner columns and the collector beams are investigated using the nonlinear response history analysis (NLRHA) and 21 sets of ground motion records. The second-phase component test results show that the buckling failure strengths (Plim) of 2CT and 2VT are 1633 and 1647kN, respectively. The prediction error from using the theoretical model on buckling failure strength is 17%. As also observed in predicting the first-phase VT test results, the effects of residual stress must be incorporated into the prediction model in order to improve its accuracy. Nonetheless, the maximum core strain of the two specimens reached 2.23% radians, and the cumulative plastic deformations (CPDs) of both 2CT and 2VT exceeded 200. Applying a linearly-reduced Young Modulus vs. strain relationship to simulate the residual stress effects in the chord members of the truss system, the ABAQUS finite element model (FEM) analysis satisfactorily predict the experimental failure strength of specimen 2VT with an error of only 1.5%. The first three modal periods of the example 23-story building are 2.12s (45°)、2.10s (135°) and 1.15s (rotation), and each of the first three modal mass participation ratios is about 75%. The scaling factors of the 21 ground motions scaled to DBE for Zone 2 of Taipei City are ranging from 2.61 to 8.34. The NLRHA results show that the maximum average lateral drifts computed from the floors at two ends of the TC-BRB are 0.400, 0.958 and 1.19% radians, respectively under the SLE, DBE and MCE. This suggests that the long-span TC-BRBs seem to have transformed the 23-story structure into a mega 4-story building, and helped to reduce overall lateral drifts and make their distribution more uniformly. It also validates the assumption of applying 1% radian as the design story drift adopted in both two phases of the component tests. However, it appears that there are some locally enlarged story drifts in the higher stories. Under the SLEs, the maximum average inter-story drifts of most stories are less than 0.5% radian, except in some higher stories are closed to 1% radian. Under the DBEs and MCEs, the maximum average inter-story drifts are all less than 2% radian. The maximum base shears under the three hazard level earthquakes are 2832, 4652, 5158 tonf, respectively. These are 1.14, 1.69, 1.84 time the LRFD design shear of 2560 tonf using an importance factor of I =1.25. The base shear of the BRBF accounts for 81% of the overall base shear under DBE. The maximum average DCRs of the BRBF corner column axial force under DBE and MCE are 0.82 and 0.86, respectively. The horizontal maximum tension or compression force in the edge beam at the transfer floor can be computed from 50% of the sum of horizontal components of maximum design tensile forces of the two adjacent TC-BRBs from above and below the transfer floor. The NLRHA results demonstrate the reliability of the design procedures, the high efficiency of the example structure system. The FEM analysis results of the realistic 40-meter-long-span, 1500 tonf yield capacity TC-BRB indicate that the self-weight is 58 tonf, the 1st and 2nd modal frequencies are 1.91 and 2.53 Hz, respectively.