| 摘要: | 台灣因位於菲律賓海板塊與歐亞大陸板塊碰撞之地體構造中，地震頻繁進而引致土壤液化災害。土壤液化係因地震力作用下引致砂土顆粒間的應力無法傳遞，致使上部砂土失去承載力，造成結構物失穩或差異沉陷等。台灣橋樑墩座或重要結構物大都採用樁基礎設計，在地震時土壤產生液化時，其對樁行為之影響值得探討，目前我國大多採用日本道路協會(1996)及Tokimatsu and Yoshimi(1983) 規範，但其規範是否適合台灣本地折減係數規範。
本研究利用振動台模擬一維及二維震動狀態下，土壤液化時樁-土互制行為，期能提供樁於土壤液化時折減係數訂定規範之參考。本研究採多層式剪力盒，以鋁製圓管空心樁鎖固剪力盒固定板，並於樁上方分別鎖固不同載重質量塊，以模擬長、短周期。砂土材料使用飽和越南石英砂，以霣降方式控制相對密度50%，振動台上輸入921集集地震(TCU052測站)及331花蓮外海地震(LIA050測站)事件，樁內設置加速度計及水壓計，剪力盒外側位移計，量測液化前後砂土層應力及應變關係、樁身彎距分布。
本研究試驗結果顯示，當液化時土壤剪力將強度弱化，且土壤液化行為係由淺往深層土壤漸進式發展。在土壤液化前，土層加速度反應之振幅會隨土層深度由深至淺逐步增加，其可能土層未達液化前仍可傳遞剪力，當土層液化時，土壤圍束力消失導致土層無法傳遞剪力；土壤主頻與不同結構物主頻在地盤受到主震前後並無明顯改變。由剪應力及剪應變關係圖分析土壤相對密度中顯示，液化土層剪力模數比，隨著土壤相對密度逐漸增加，液化土層剪力模數比也會逐漸放大。
;Taiwan is situated at the tectonic boundary where the Philippine Sea Plate collides with the Eurasian Continental Plate, resulting in frequent earthquakes that, in turn, induce soil liquefaction hazards. Soil liquefaction refers to the phenomenon where, under seismic loading, the stresses between sand particles can no longer be transmitted effectively, causing the overlying sand layer to lose its bearing capacity, thereby leading to structural instability or differential settlement.
In Taiwan, bridge piers and other critical infrastructures are predominantly designed with pile foundations. During seismic events, the effect of soil liquefaction on pile behavior warrants thorough investigation. At present, most domestic practices refer to the specifications of the Japanese Road Association (1996) and Tokimatsu and Yoshimi (1983). However, whether such specifications are suitable for use in Taiwan—particularly in terms of reduction factors—remains to be evaluated.
This study employs a shaking table to simulate pile-soil interaction behavior under both one-dimensional and two-dimensional seismic excitation during soil liquefaction, aiming to provide a reference basis for establishing reduction factor specifications for piles in liquefiable soils. A multi-layer shear box is adopted, wherein hollow aluminum circular piles are fixed to the base plate of the shear box, and different masses are mounted atop the piles to simulate long- and short-period structures. The soil material used is saturated quartz sand from Vietnam, prepared by pluviation to control a relative density of 50%.The shaking table is subjected to input motions from the 1999 Chi-Chi earthquake (TCU052 station) and the 2022 offshore Hualien earthquake (LIA050 station). Accelerometers and pore water pressure transducers are installed within the piles, while displacement meters are placed outside the shear box to measure the stress–strain response of the sand layer before and after liquefaction, as well as the bending moment distribution along the pile shaft.
The experimental results of this study indicate that soil shear strength degrades during liquefaction, and that the progression of soil liquefaction develops gradually from shallow to deeper layers. Prior to liquefaction, the amplitude of soil acceleration responses increases progressively from deep to shallow layers, suggesting that soil layers are still capable of transmitting shear forces before reaching a liquefied state. Once liquefaction occurs, the loss of confinement in the soil prevents the transmission of shear forces. Furthermore, the predominant frequency of the soil and of various structures shows no significant change before and after the mainshock. Analysis of the shear stress–strain relationship reveals that, in liquefied layers, the shear modulus ratio increases with increasing relative density of the soil. |
