考量挫曲束制斜撐圍束單元撓曲之整體面外穩定性研究

Abstract

摘要 由於挫曲束制斜撐(BRB) 能大幅提升結構之勁度、強度與韌性，含BRB之構架(BRBF)經濟性與良好耐震消能行為已獲肯定。近年來已被廣泛應用於鋼建築結構系統之中。然而，由於其斷面變化及圍束不連續的特徵，於先前研究中BRB整體面外挫曲的情況時有所聞。過去對於BRB整體面外穩定性雖已有相當之研究，但現行習見之評估方法運用三項檢核，分別檢討圍束鋼管、連接段與接合板穩定性；唯此方法未考量彼此之間的互制效果，以及施工誤差及面外變位等影響穩定性的參數；此方法亦未對於BRB的有效長度作嚴謹的定義。有鑑於此，本研究之宗旨在於根據力學模型與理論，提出可靠之BRB整體面外穩定性評估模型。 日本 Takeuchi教授等人根據BRB及其接合板於挫曲時的塑性變形行為，推導出一系列之挫曲強度預測模型。基於Takeuchi教授等人提出之模型，並考量槽接式挫曲束制斜撐(WES-BRB)圍束單元較長之特徵，本研究提出考量圍束單元撓曲效應及接合板旋轉效應的BRB挫曲強度預測模型。由模型發現，圍束單元的尺寸與圍束單元內充水泥砂漿或混凝土之貢獻是影響整體穩定性的關鍵參數。本研究亦使用有限元素模型分析接合板的面外旋轉勁度與強度，及初始端部面外錯位、初始圍束單元面外變形對整體穩定性之影響。並提出計算接合板力學參數及因初始面外錯位、變形造成額外端部彎矩需求之建議公式。 為驗證模型之可靠度，本研究設計了四組圍束單元尺寸不一、接合板厚度不一以及有無接合板加勁板，總長5.8公尺、100噸級標稱降伏強度之WES-BRB試體。於國家地震工程研究中心的多軸向試驗系統(MATS)進行反覆載重測試。本研究針對其中一組試體施加1%試體總長之端部面外錯位，藉比對有無面外錯位之實際挫曲強度，來探討其對整體穩定性之影響。 實驗結果證明，預測強度與實際挫曲強度僅6%之誤差。若於接合板長邊配置加勁板，則實際挫曲強度上升13%。1%試體總長之端部面外錯位造成9%之挫曲強度損失。若圍束單元管徑增加24%，預測之挫曲強度將大幅上升80%。由光學量測結果顯示，隨著軸力上升，初始面外端部錯位會導致較嚴重的面外變形，而大幅增加挫曲趨勢。實驗中亦發現BRB試體圍束單元的初始面外變形比預期更嚴重。這應是在試體養護、運送或安裝過程中，不當的吊裝所造成。本研究所提出之BRB整體面外穩定性評估模型，能應用於實際BRBF設計，防止BRB面外挫曲。

Abstract Buckling-restrained braces (BRBs) have been widely used as cost-effective energy dissipaters for seismic steel buildings. However, several cases of out-of-plane (OOP) instability have been observed from past researches. In a prior research, a BRB specimen installed on first floor buckled with severe flexural deformation along the restrainer and plastic hinges forming at gussets during a full scale two-story RC frame test. Simplified procedures commonly applied could predict the high possibility of the buckling. These procedures are based on three independent stability assessments for the steel casing, the connection and the gusset respectively. Nevertheless, these three separate stability checks do not consider their coupling effects on the overall stability. Furthermore, the procedures appear to be over-simplified, using unreasonable assumptions on the end conditions. Takeuchi et al. proposed a set of advanced procedures to measure the global OOP stability. However, their buckling models cannot predict the aforementioned failure mode. This study adopts Takeuchi’s procedures and proposes a buckling model considering flexural restrainer and gusset rotations. In addition, evaluation methods are developed for FEM analysis in order to compute the gussets’ rotational stiffness and strength, as well as the additional moment demand at gussets caused by OOP end drift and initial imperfection along BRB restrainer. In order to verify the effectiveness of the proposed procedures, four full-scale BRB specimens each of 5.8m long with a 988-kN nominal yielding strength, varying restrainer stiffness, gusset thicknesses, with/without edge stiffeners or OOP drift demands were tested. The proposed model satisfactorily predicts specimens’ failure modes and buckling strengths with errors less than 6%. Test results show a 9% drop in buckling strength due to a 57-mm OOP end drift. Edge stiffeners detailed on the gussets’ long sides improved the buckling strength by 9%. The proposed model exhibits an improvement of over 80% in the buckling strength with a 24% enlargement in the restrainer diameter, indicating the critical effects of the restrainer’s flexural stiffness. The deformed shapes of specimens throughout the loading indicate that OOP drift tends to trigger more severe flexural deformation as axial loading increases, leading to a higher buckling potential. In addition, the initial imperfection along BRB restrainer was larger than expected due to improper supporting condition during mortar curing, shipping and handling. The research results can be adopted to improve the practice of BRB frame design.