均質化具空間變異性土體之有效滲透係數

Homogenization of Effective Soil Hydraulic Conductivity with Spatial Variability

陳致宇(Zhi-Yu Chen)

指導教授 ： 卿建業

國立臺灣大學/工學院/土木工程學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202101628

摘要

土壤的空間變異性為常見造成大地工程設計時產生不確定性的因素之一，然而工程師往往因為某些實際考量將土壤視為均質以便分析。為了選擇較具代表性的土壤材料性質作為設計參數，本研究針對土壤的滲透係數參數，提出一均質化模型，有效地代表具空間變異性土體之滲流行為。
本研究透過模擬穩態隨機場，並藉由有限元素軟體Abaqus 進行隨機有限元素分析獲得等值滲流行為的有效滲透係數。同時，以空間平均模型計算滲透係數隨機場的均質化滲透係數進行比較。初步結果顯示，土壤受到非均勻驅動情形會影響空間平均的權重分布。本研究藉由適當的方法獲得權重分布後，發現與均質有限元素模型輸出的水力坡降增量與滲流量增量有關，進而提出了適用於計算土壤元素權重的新pseudo incremental energy (PIE) 模型，搭配權重幾何平均模型便能獲得均質化滲透係數。
透過新PIE模型針對常見滲流案例，進行單層土及多層土在不同空間變異程度下進行分析。研究結果顯示，大部分案例採用新PIE模型的均質化模型所獲得之均質化滲透係數皆與隨機有限元素分析獲得之有效滲透係數呈現高度的相關性。值得注意的是，此方法僅需進行一次的均質有限元素分析即可得到合適、具代表性且同時考慮土壤空間變異性的均質化滲透係數。此外，本研究也透過現地抽水井案例驗證該方法的可行性與正確性。

關鍵字

空間變異性 ； 隨機場 ； 有限元素分析 ； 有效滲透係數 ； 新PIE模型 ； 均質化模型

並列摘要

Inherent spatial variability is one of the uncertainties in geotechnical engineering design, while engineers often do not consider it and treat soil properties as constant values for some practical reasons. In order to find a representative soil property as design parameter, this research focuses on soil hydraulic conductivity and proposes a homogenization model enable to characterize seepage behavior with spatial variability.
First, we simulate stationary random field and then conduct random finite element analysis (RFEA) from Abaqus software to obtain the effective hydraulic conductivity (keff), which means a homogeneous hydraulic conductivity actually felt by soil. Meanwhile, we compute spatial averages of hydraulic conductivity random field to obtain homogenized hydraulic conductivity and compares to keff. Preliminary results show that non-uniform mobilization of soil will affect different weights in soil elements. Therefore, we conduct a single run of homogenization finite element analysis (FEA) to find out which parameters are relevant to weights. Results show that weights have correlations with hydraulic gradient increment and flow rate increment. Subsequently, we establish a new pseudo incremental energy (PIE) model. With new PIE model adopting weighting geometric average model, we can obtain the homogenized hydraulic conductivity.
This new PIE model will be verify through some common seepage cases concluding single layer and multiple layers with different spatial variability settings. Results show that most cases have good approximation by adopting new model. It is note-worthy that the new model only needs a single run of FEA then it can approximately represent the effective hydraulic conductivity and also considers spatial variability. Moreover, this research verifies the new model through an in-situ case and the results are also quite reasonable.

並列關鍵字

Spatial variability ； Random field ； Finite element analysis ； Effective hydraulic conductivity ； New pseudo incremental energy model ； Spatial average model

參考文獻

Bevan, M. J., Endres, A. L., Rudolph, D. L., Parkin, G. (2003). The non-invasive characterization of pumping-induced dewatering using ground penetrating radar. Journal of Hydrology, 281(1-2), 55-69.

Bevan, M. J., Endres, A. L., Rudolph, D. L., Parkin, G. (2005). A field scale study of pumping-induced drainage and recovery in an unconfined aquifer. Journal of Hydrology, 315(1-4), 52-70.

Cami, B., Javankhoshdel, S., Phoon, K.-K., Ching, J. (2020). Scale of fluctuation for spatially varying soils: Estimation methods and values. ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, 6(4), 03120002.

Ching, J., Chen, Y.-C. (2007). Transitional Markov chain Monte Carlo method for Bayesian model updating, model class selection, and model averaging. Journal of Engineering Mechanics, 133(7), 816-832.

Ching, J., Hu, Y.-G. (2016). Effect of element size in random finite element analysis for effective Young’s modulus. Mathematical Problems in Engineering, 2016.