以非線性歷時分析驗證鋼筋混凝土建築之耐震性能

Abstract

非線性歷時分析是一種更高階且精確的結構分析方法，它不僅可以充分反映結構物的動態行為，亦能精確地洞察結構桿件的非線性行為，是最貼近真實的結構分析方法。傳統的等效側力法係將結構於地震中之動態反應，以施加於結構側向之靜態力取代，並以折減地震力之方式考慮構件之非線性行為。然而，若構件進入非線性行為，此時結構物的耐震性能，其實是取決於構件的非線性變形能力，而非構件的強度，因為構件進入非線性行為後，強度增長有限，整體結構主要係透過構件的非線性消能行為抵抗地震能量，因此結構在真實地震中的非線性位移，才是攸關結構耐震能力之關鍵，此即基於位移的性能設計法之核心概念。 本研究以一棟位於臺北一區的16層樓鋼筋混凝土建築物為研究對象，該建築業已依照建築物耐震設計規範及建築物混凝土結構設計規範設計完成。本研究輸入多筆最大考量地震之歷時及構件非線性行為參數於結構分析程式ETABS中，並依照國內外相關規範，檢核本建築是否滿足規範對於整體結構及局部構件之接受準則。 此外，本研究亦將非結構牆納入分析模型中，探討非結構牆於最大考量地震中，其破壞情況是否如同實務界常假設的全部破壞，亦或是僅有部分結構牆產生破壞，以及這些非結構牆是否會對結構物有不良的影響如短梁效應。分析結果顯示，在最大考量地震下，並非所有的非結構牆均完全失去強度而破壞，而且非結構牆破壞的順序及位置，亦未具統一的規則及形式。 非線性歷時分析之前後處理資料量極大，本次研究共開發了17個非線性歷時分析的工具程式，不僅可以大幅節省處理時間，亦可對分析結果有更深入的洞察。 在執行多次非線性歷時分析後，本研究發現結構物的反應，在真實地震下，具有高度變異性，即使是同一個結構物的同一個桿件，在不同的地震歷時下，常常會有截然不同的反應及破壞模式。透過非線性歷時分析，結構設計者可以更全面地掌握建築物於地震下的真實反應與性能，並加以改善或優化，以增進建築物之安全性與可靠度。

Nonlinear time history analysis (NLTHA) is an advanced and precise method for evaluating seismic performance, capable of capturing both the global dynamic response of structures and the localized inelastic behavior of structural components. In contrast to the widely adopted Equivalent Lateral Force (ELF) method—which substitutes dynamic effects with simplified static loads and considers inelasticity through strength-reduction factors—NLTHA directly reflects how structures dissipate seismic energy through nonlinear deformations. Once components yield, strength increases are limited, and seismic performance becomes largely governed by deformation capacity, making displacement demand a key indicator in performance-based seismic design. This study investigates a 16-story reinforced concrete building located in Taipei, originally designed according to Taiwan’s Seismic Design Specifications and Concrete Structural Design Code. A series of Maximum Considered Earthquake (MCE) ground motion records were applied in ETABS, incorporating nonlinear behavioral models of structural components. Structural responses were then evaluated against the acceptance criteria outlined in several relevant codes, covering both global structural performance and individual member demands. Additionally, non-structural walls were modeled to assess their failure patterns under MCE conditions. The study questions the common assumption of total failure and explores whether partial damage occurs and if such walls induce negative effects like the Short Beam Phenomenon. Results indicate that non-structural wall failure is neither uniform nor fully predictable, and complete loss of strength is not always observed. Given the data-intensive nature of NLTHA, the study also developed 17 customized tools to automate preprocessing and postprocessing tasks, enabling significant time savings and deeper insight into structural behavior. Multiple analysis runs revealed a high degree of variability in structural responses—even for the same component under different ground motions—underscoring NLTHA’s essential role in realistic seismic performance assessment and in guiding the optimization of structural safety and reliability.