順向邊坡趾部基礎開挖數值分析之實例分析及問題探討

Case Study on the Numerical Analysis of the Building Foundation Excavation on the Dip-Slope

本研究以台灣東北部某校區順向坡趾部之圖資大樓基礎開挖工程為例，探討於順向邊坡坡趾進行基礎開挖施工時，所須之分析設計考量與因應措施及監測系統施測結果，並將數值分析結果與現地監測資料作印證與比對，藉以瞭解於順向邊坡建築開挖數值分析結果與監測資料之差異性。初步分析結果顯示，實際案例原設計以斜坡明挖方式施工，經由PLAXIS 與STABL 分析，於暴雨及地震情況皆有可能發生破壞，其滑動面有沿破碎帶下滑之趨勢，故斜坡明挖較不適用；其後變更設計為垂直擋土地錨背拉工法，於各情況分析比較發現基地以垂直擋土地錨工法開挖較明挖工法合宜，其中降雨，地震情況及動態分析滑動面沿破碎帶下滑之趨勢減少，且安全係數大都符合法規之規定值。本案例結果亦顯示彈塑性分析法所得安全係數多數偏不保守，較不適用於斜坡之基礎開挖情況。另於開挖分析結果顯示PLAXIS 程式分析壁體變位之趨勢，因受地錨預力造成壁體往上邊坡變位，但分析變位量與實測值略有不同。經由傾斜管、地錨荷重計、傾度盤…等監測儀器之實測值與數值分析程式所模擬之分析結果比較，其趨勢大致相同，惟部份實測量略高於分析之值，可能係假設條件與初始情況差異所致。

A case study involving slope foundation excavation located in northeast Taiwan is presented. Factors considered in the analyses and the corresponding measurements from a building foundation excavation on the toe of dip-slope is discussed. Extensive efforts were made to compare the numerical analysis results with the in-situ instrumentation measurements. Based on the analysis results from PLAXIS and STABLE, it demonstrates that the original proposed open-cut method is unsuitable for this dip-slope excavation condition. It was found that the sliding failure may occur under the rainfall and earthquake conditions. The sliding surface is extending along with the bedding slips. To minimize those safety problems, the alternative construction method using the pile retaining wall and the tie-back anchors are proposed. According to the numerical analysis results, both the rainfall and earthquake conditions for the alternative method are able to conform to the safety requirements outlined in the Taiwan Building Codes. It was concluded that the elasto-plastic analysis is not conservative for this slope foundation excavation. The analysis results from PLAXIS have the closest agreement with the field measurements. Furthermore, the deformation trends obtained from PLAXIS analyses match with those from field instrumentation measurements. However, the deformation values from the analyses are less than those from the field measurements. This could be attributed to the creep effects and assumed conditions.