排水固结三维非线性有限元及基于遗传算法的材料参数反演技术应用研究.doc

排水固结三维非线性有限元及基于遗传算法的材料参数反演技术研究 摘 要 随着我国基本建设的发展，越来越多的道路不得不建在软土地基上。由于软土地基的压缩性高，渗透性低，固结变形持续时间长，所以改进现有的沉降计算方法及准确的预估地基沉降量就有了重要的工程意义和理论价值。特别是土体勘测密度不够，土体的多个参数的选择对计算结果影响较为显著，如何确定实际工程中土体参数对计算结果的影响，实现对设计进行校核已逐步受到岩土工程界的重视。 本文根据Biot固结理论与有限元原理，用FORTRAN语言编制了三维地基土体Biot固结的非线性有限元程序ZBP。该程序中土体采用了Duncan-Chang双曲线本构模型，可以考虑土体的应力－应变的非线性特征，可以计算软土路基任意时刻的沉降，并能够考虑施工分级加载对沉降的影响。 为了减少砂井地基三维有限元模拟和分析的工作量，引入一种新型单元来考虑塑料排水板。该单元类型在不增加有限元结点的情况下把排水板模拟成线单元。其主要考虑排水板的排水性能。同时，推导了排水板单元的力学劲度矩阵和渗透矩阵，利用FORTRAN语言编制了相应的程序，并把它结合到三维Biot固结程序ZBP中。 本文以深圳市宝安区宝安大道南段为研究背景，以K6+475断面为研究对象，将遗传算法与Biot固结理论相结合，建立了基于遗传算法的软土地基多土层的多参数优化反演分析模型。利用FORTAN语言编制了相应的反演程序ZGA。对Duncan-Chang模型的部分参数进行了反演分析，根据反演的土体参数值，再利用ZBP，预测了下阶段土体变形发展情况，实例中计算值与实测值的拟合情况良好，预测结果符合工程发展规律。通过工程实例分析，很好的说明了遗传算法适用于软土地基多土层的土体参数优化反演分析，其反演结果可用于位移预测。 关键词：Biot固结，塑料排水板，遗传算法，参数反演，沉降预测 3-D nonlinear FEA of drainage consolidation and research of parameter inversion based on genetic algorithm Abstract Along with the development of countrys capital construction, more and more highways have to be built on soft soil foundation. Because of the high compressibility, low permeating, long time of consolidation, it is possessed of great engineering meaning and theoretical value to improve the existing settlement calculation methods and estimate the settlement of the soft soil foundation. Especially when the surveys are not enough, the choice of soil body parameters can make very obvious influence to the calculated result. It is paid attention progressively, concerning about how to confirm the influence by multi-parameters in practical project and how to carry on the check of designing. According to the Biot Consolidation and FEM theories, the program ZBP in FORTRAN language has been worked out. Considering the nonlinear relation between the stress and strain, Duncan-Chang model has been applied in the program. This program can calculate the settlement of the soft soil subgrade at every moment. It could also consider the influence of settlement by the construction grade. In order