以傾斜錐貫入試驗識別水平自相關參數

Abstract

土壤參數的空間變異性可藉由自相關函數(auto-correlation function, ACF)的參數來量化，具體而言是關聯性長度。過去的研究主要關注垂直自相關參數的辨識；只有極少數的研究關注水平自相關參數的辨識。本研究提出了一種新方法：使用傾斜錐貫入試驗(cone penetration test, CPT)作為輔助，以提升水平向自相關參數的辨識能力。本研究首先分析了各個場址的CPT數據，以展示僅從多個垂直CPT辨識水平自相關參數的弱勢。本研究亦通過數值案例探討了有效辨識水平自相關參數的CPT排佈方式。最後，本研究探討了通過多個垂直CPT，外加一個（模擬的）傾斜CPT數據，來辨識水平自相關參數的可行性。

The spatial variability of soil data is characterized by the auto-correlation function (ACF) parameters, specifically scale of fluctuation. In the literature, the study mainly focuses on identification of vertical auto-correlation parameters; the identification of horizontal auto-correlation parameters received limited attention. In this study, a novel approach to enhance the identification of horizontal auto-correlation parameters using an inclined cone penetration test (CPT) as auxiliary sounding is proposed. The CPT data from various real site locations are analyzed, to demonstrate the shortcomings of identifying horizontal auto-correlation parameters from multiple vertical CPTs. Numerical study is conducted to investigate the possible approach (CPTs layout) for identifying horizontal parameters. Subsequently, the potential of identifying horizontal auto-correlation parameters from a combination of vertical CPTs and synthetic data from an inclined auxiliary CPT is examined.