考虑粉粒含量影响的循环荷载下砂土边界面模型及其适用性研究

Experimental studies on artificially mixed sands with non-plastic fines have shown that these soils are much more liquefiable than clean sand. The non-plastic fines can affect the liquefaction strength of silty sands and consequently affect the deformation of engineering structure induced by soil liquefaction due to earthquake shaking. Thus, it is necessary to develop an elastoplastic cyclic constitutive model to characterize the effect of non-plastic fines. The main objective of this planning project is to develop a unified bounding surface model for the silty sand containing different non-plastic fines contents under cyclic loading and then to incorporate the unified bounding surface model in a fully coupled dynamic effective-stress finite element procedure (UWLC) for liquefaction analysis of the effect of non-plastic fines content on the ground deformation. Firstly, unified formula will be derived from a number of silty sands of different non-plastic fines contents from existing literature to characterize the dilatancy, critical state line, phase transformation line, bounding surface, critical state stress ratio, elastic modulus, and plastic modulus. Then, such formula will be used to develop a unified bounding surface model for the silty sand under cyclic loading following the basic framework of stress-controlled, critical state compatible, bounding surface model for sand presented by Dafalias and Manzari (2004) and Boulanger and Ziotopoulou (2013). Thirdly, a series of monotonic and cyclic triaxial tests will be conducted to validate the capacity of the proposed constitutive model by taking the non-plastic fines content, confining pressure, relative density, and cyclic stress ratio into consideration. Fourthly, the proposed constitutive model will be implemented in UWLC to establish a computation method for the liquefaction analysis of the site containing sand or silty-sand layer. And, finally, numerical analyses will be conducted on a quay wall that was damaged due to soil liquefaction during the 1995 Japan Kobe earthquake to investigate the effect of non-plastic fines content on the deformation of the quay. The corresponding results can be useful to improve the seismic performance of engineering structures normally located on the site containing sand or silty-sand layer in offshore engineering and ocean engineering and to develop the remedial measures against liquefaction.

试验研究表明，若忽略砂土中粉粒的影响，则可能低估砂土液化给临近工程结构所带来的危害，因此，需要充分估计粉粒含量对饱和砂土液化强度及液化变形的影响。本项目以含粉粒砂土为研究对象，依据国内外典型含粉粒砂土室内试验，建立粉粒含量对砂土微观孔隙结构影响的评价指标；考虑粉粒含量、围压、相对密度、循环应力比等因素，开展含粉粒砂土动静三轴系列试验，研究粉粒含量对砂土剪胀规律、边界面及相变线的影响；基于统一临界状态线的思想，建立考虑粉粒含量影响的循环荷载下砂土边界面模型，并利用三轴试验结果，验证所提模型的有效性；基于Biot动力固结理论，研发本构模型程序模块，并对1995年日本阪神地震中液化场地上沉箱式码头岸壁震害进行数值分析，建立粉粒含量对含粉粒砂土液化变形及其所引起岸壁位移的评价方法。研究成果对评估沿海（滨）地区含粉粒砂土层天然及人工回填地基、人工岛地基抗液化性能提供理论依据与有效方法。

试验研究表明，若忽略砂类土中细粒的影响，则可能低估或高估砂类土液化给临近工程结构所带来的危害，因此，需要充分估计细粒含量对饱和砂土液化强度及液化变形的影响。本项目采用福建标准砂与三种不同粒径的非塑性细粒进行混合，考虑细粒含量、围压、循环应力比等因素，开展了一系列三轴不排水压缩试验和循环三轴试验；提出了适用于单调加载和循环加载条件下砂类土状态相关弹塑性本构模型；在商业有限元程序ABAQUS和完全耦合动力分析框架中实现了该本构模型的二次开发，建立了定量评估细粒对场地及工程结构影响的动力分析方法；以桩基工程为实例，通过自主研发程序将所提的状态相关弹塑性理论与应变楔理论结合，提出了定量评估细粒对桩基水平响应影响的静力分析方法。结果表明：（1）中等密实状态的砂类土在高细粒含量的情况下可发生静态液化，且平均粒径比Rd越大静态液化越容易发生，因此，评估饱和砂类土强度需要考虑非塑性细粒粒径的不利影响；（2）等效骨架状态参数ψ*可以很好的表征饱和砂类土的液化强度，这与本项目开展的循环三轴试验结果的发现较为一致；（3）所提模型的计算结果与本项目开展的及已有文献中的砂类土三轴试验结果能够较好吻合，验证了所提模型的适用性；（4）进行了Wildlife实际液化场地地震反应分析，计算出的地表加速度时程和孔压时程与实测值能够合理吻合，进一步验证了模型二次开发的可靠性；（5）在相同的孔隙比条件下，桩基水平承载力随着细粒含量的增加而降低，细粒对桩身挠度影响比其对桩身弯矩影响更大。研究成果对评估沿海（滨）地区含砂类土层天然及人工回填地基、人工岛地基抗液化性能与设计提供理论依据与有效方法。