Risk Assessment and Hazard Prediction of Seismic Liquefactio

发布时间：2024-07-05 19:45

　　饱和砂土液化是最为典型的地震灾害之一。因此,准确地对液化危害进行评估是岩土工程领域的重要任务之一。地震液化评估中,需要对三个主要的方面进行考虑:1)液化敏感性;2)触发液化的动态载荷的评价;3)液化的影响。前两个在分析描述液化能否发生的液化势方面已经得到了检验。此外,在评估液化破坏效应时,主要是指横向位移,这是进行液化危害评估的最后一个方面。在所有的方法中,基于历史数据或室内试验结果的经验和半经验模型是最为普遍且易于工程师和研究人员所使用的一种分析手段。现阶段,智能算法已被运用于检测参数与危险因素之间的相关性。然而,这种方法依旧缺乏防止模型被过度训练的验证阶段。虽然细粒含量(FC)对孔隙水压力的产生具有复杂影响这一结论已经得到了验证,但是在现有模型中并没有考虑该影响参数。所有的模型都包含了此参数,但对其取值并没有加以任何限制。此外,已经证明,标准化累积绝对速度(CAV5)为孔压的增长和液化的发生提供了最充分和有效的推动作用。然而,在现有模型中,动量大小(Mw)和水平峰值地面加速度(PGA)通常与液化危害评估相关。此外,由于天然易变性以及对土体性质缺乏了解,几何条件、地震荷载,岩土工程问...

【文章页数】：176 页

【学位级别】：博士

【文章目录】：
摘要
Abstract
1 Introduction
    1.1 Approaches for the assessment of potential of liquefaction triggering
        1.1.1 Stress-based approaches
        1.1.2 Strain-based approaches
        1.1.3 energy-based approaches
            1.1.3.1 Approaches based on earthquake case histories
            1.1.3.2 Approaches based on the Arias Intensity
            1.1.3.3 Approaches based on laboratory test results
    1.2 Approaches for prediction of lateral displacement due to the liquefaction
        1.2.1 Review of Empirical and Semi-Empirical Models to Predict Lateral Displacement Dueto Liquefaction
            1.2.1.1 Newmark sliding block analysis
            1.2.1.2 Nonlinear analyses
    1.3 Cumulative Absolute Velocity
        1.3.1 Unpublished results
    1.4 Fine Content Critical Value
    1.5 Artificial Neural Network
    1.6 Response Surface Methodology
        1.6.1 Select of regression model function
        1.6.2 Design of experiment
            1.6.2.1 The concepts used in the design of experiments
                1.6.2.1.1 Variables
                1.6.2.1.2 Factor
                1.6.2.1.3 Levels
                1.6.2.1.4 Response
                1.6.2.1.5 Effect
                1.6.2.1.6 Interaction
                1.6.2.1.7 Optimization process
        1.6.3 Experimental design
            1.6.3.1 Central composite design
                1.6.3.1.1 Types of central composite design
                    1.6.3.1.1.1 Circumscribed designs
                    1.6.3.1.1.2 Face centred
            1.6.3.2 Box- Behnken design (BBD)
            1.6.3.3 Doehlert design
        1.6.4 RSM advantages
        1.6.5 RSM disadvantages
        1.6.6 Coding of the input variables
        1.6.7 Hypothesis Test
    1.7 Monte Carlo simulation and uncertainties
    1.8 Monte Carlo simulation using artificial neural network for parametric sensitivity analysisproposed in this study
    1.9 Organization of the Thesis
2.A New Equation to Evaluate Liquefaction Triggering Using the Response Surface Methodand Parametric Sensitivity Analysis
    2.1 Introduction
    2.2 Case History Database
        2.2.1 Earthquake magnitude and peak accelerations
        2.2.2 The selection and measurement of qc1 Ncs values
        2.2.3 Moss landing state beach
        2.2.4 Wildlife liquefaction array
        2.2.5 Miller and Farris farms
        2.2.6 Malden Street
        2.2.7 The classification of site performance
    2.3 Proposed Model and Equation to Evaluate Liquefaction Triggering
    2.4. Results
        2.4.1 RSM Equation to Evaluate Liquefaction Triggering
        2.4.2 Sensitivity Analysis with the Monte Carlo Simulation Method
    2.5 Summary and Conclusions
3. Energy Evaluation of Triggering Soil Liquefaction Based on the Response Surface Methodand Parametric Sensitivity Analyse
    3.1 Introduction
    3.2 The mechanisms of energy dissipation in sand
        3.2.1 Hysteresis loops
        3.2.2 Equal linearization and damping ratios
        3.2.3 The use of dissipated energy to quantify capacity
    3.3 Databases and artificial neural network models
        3.3.1 First artificial neural network mode
        3.3.2 Second artificial neural network mode
    3.4 The RSM Equations
    3.5 Comparison of the predicted capacity energy liquefaction by the RSM equations andexisting model
    3.6 Comparison of the predicted capacity energy liquefaction by the ANN models andexisting models
    3.7 Sensitivity analysis
    3.8 Summary and Conclusion
4. New Equations to Evaluate Lateral Displacement Caused by Liquefaction Using theResponse Surface Method and Parametric Sensitivity Analysis
    4.1 Introduction
    4.2 Patterns of Lateral displacement deformation
    4.3 Models for lateral displacement measurement
    4.4 Dataset
    4.5 Artificial Neural Network Models
        4.5.1 Comparison of ANN models with Extra Model
    4.6 The RSM Equations for Predicting DH
        4.6.1 Comparison of RSM Equations with Extra Models
    4.7 Sensitivity Analysis
    4.8 Results and Discussion
    4.9 Summary and Conclusions
5. Conclusions and Prospects
    5.1 Conclusions
    5.2 Innovation Points
    5.3 Outlook
References
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Research Projects and Publications During PhD Period
Acknowledgement
Curriculum Vitae



本文编号：4001437