樁基礎橋梁含功能性支承之振動台試驗研究

Abstract

本研究探討含功能性支承墊之單垮橋梁模型受沖刷導致基礎裸露後，耐震性能及樁土互制行為之變化。本研究採用大型剪力盒裝置砂土試體以及橋梁模型並以0倍、3倍、6倍基樁直徑作為裸露深度之分組以探討基礎裸露之影響，砂土試體採用相對密度50%之乾石英砂。本研究亦對於功能性支承進行探討，採用理想化的簡支橋梁以及滑動支承橋梁研究支承形式對於耐震性能的影響。為能夠對功能性支承橋梁於受震下，摩擦係數以及橡膠支承墊的滑動行為進行更進一步的探討，本研究先進行固定基礎試驗並將其作為樁基礎試驗之對照組，試驗輸入地震主要為El Centro 100gal至400gal以及TCU076 100gal至300gal。研究著重於上部結構受震反應、土層加速度以及橋柱與基樁之彎矩需求容量比。試驗結果顯示，當基礎裸露深度逐漸增加時，結構系統之週期、上部結構位移以及基樁之彎矩需求容量比會逐漸增加，而上部結構之加速度以及橋柱之彎矩需求容量比卻會隨之下降。隨著輸入地震增加超過300gal後，滑動支承橋梁之上部結構開始發生滑動現象導致上部結構傳遞至橋柱之慣性力受到限制，因此滑動支承橋梁之柱底彎矩需求容量比小於簡支橋梁鉸接端之柱底彎矩需求容量比。 為探討樁土互制行為，本研究以SAP2000建立數值分析模型並進行迭代運算，考慮土壤動力特性同時採用公路橋梁耐震設計規範之補充研究計算等效線性之土壤彈簧。數值模型中給定任意土壤彈簧初始值並以數值分析結果得到之基樁側向變形以計算各層土壤之剪應變並藉此對於土壤彈簧之側向勁度進行折減，不斷重複該流程至結果收斂。此外，上層土壤受到基樁之擾動較為明顯，故該區土壤較為鬆散而無法提供足夠之側向勁度，因此建議對其進行適度折減。數值分析結果與試驗結果之比較顯示，對於頂層土壤進行適當折減之數值分析結果同時於大地震以及較小地震皆可得到良好的上部結構反應，對於橡膠支承墊之滑動情形亦能準確預測。

This study investigated the seismic performance and soil-structure interaction of a simply-supported bridge model through shaking table test. Dual laminar sand boxes were utilized to accommodate a scale-down bridge model with specified exposed length of pile foundation, as a ratio of pile diameter D, from 0D, 3D to 6D. The soil property of dry fine silica sand is controller by relative density of 50%. Besides, the boundary condition of rubber bearing system were also changed to represent semi-rigid or idealized hinge and roller. A reference case of bridge model without piles and fixed on the ground was also conducted to identify the friction coefficient and sliding behavior of rubber bearing under seismic loading. The selected input ground motion is El Centro, and the peak ground acceleration (PGA) were in the range of 100, 200, 300, and 400 gals. Dynamic response of superstructure, soil layer, and Demand/Capacity Ratio (DCR) of moment for both column and pile were recorded. The experimental results showed that structural period, displacement demand of superstructure, and the DCR of pile were increased, since exposed length of pile foundation was increasing. However, the acceleration of superstructure and the DCR of column were decreased. Meanwhile, when PGA was larger than 300 gals, since sliding of superstructure occurred, the inertia force transferred from superstructure to column was limited. Therefore, the DCR of column obtained from moveable condition is smaller than the one from hinge bearing. In order to account for soil-structure interaction, an iteration procedure to determine the equivalent stiffness of soil springs in SAP2000 was proposed. The initial stiffness along the pile may be randomly selected, and then updated by the calculated lateral displacement of pile from elastic dynamic time-history analysis. It was until the reduction of shear modulus for non-cohesive soil as used in SHAKE 91 was converged to stop the iteration for soil spring. A further reduction of shear modulus was recommended to reflect the disturbance between the top of pile and surrounding sand. It is found the structural performance, including acceleration and displacement of superstructure, as well as the sliding of rubber bearing can be precisely predicted, in the case of none, minor, and major scouring of pile, respectively.