桥梁混凝土结构耐久性数值模拟的细-宏观转译理论

Durability of the bridge concrete structures shows significant time-dependent and spatial variability under the influence of material, environmental effects, component details and stress conditions. It is very difficult to seek both accuracy and efficiency in the current monoscopic numerical simulation methods. In this research, it is intended to explore and establish the mesoscopic to macroscopic translation theory in durability numerical simulation of carbonation and chloride ingression. It is expected that this research will offer viable solutions to some fundamental questions in the durability performance simulation and design for bridge concrete structures. Firstly, verification indices and methods for the mesoscopic and macroscopic numerical simulation are proposed, based on which the current methods can be verified and promoted. Secondly, mesoscopic numerical simulation is adopted to study the scientific laws of aggregates variability influencing the ingression results and the relationship between the mesoscopic variability and macroscopic responses. According to the study, two mesoscopic to macroscopic translation modes are proposed, one of which is based on the effective diffusion coefficient, and the other is based on the non-constant environmental effect input. Then a new macroscopic numerical simulation method with consideration to the aforementioned translation is established, settling the contradiction between simulation accuracy and calculation efficiency. Thirdly, the translation theory is improved with considerations to the characteristics of the durability of the bridge concrete structures, 1) translation theory for the multi-factor coupled effects is studied to precisely simulate the actual engineering state, and 2) the statistic and probabilistic characteristics of the numerical simulation results are discussed and expressed in the durability design for engineering applications. The accomplishments in this research are expected to provide the parameters, instruments and methods of the durability design for bridge concrete structures with novel fundamental data and theory.

桥梁混凝土结构耐久性受到材料、环境、构造、受力状态等影响，具有显著时空变异性，现有单一尺度的数值模拟方法难以兼顾精度与效率。本课题拟针对碳化与氯离子侵蚀，探索建立耐久性数值模拟的细-宏观转译理论，解决桥梁混凝土结构耐久性能模拟与设计中的基础科学问题。首先，研究提出细观与宏观数值模拟校验指标与方法，校核与优化现有计算方法。然后基于细观数值模拟，研究骨料随机性对侵蚀结果的影响规律，揭示细观差异与宏观响应的对应关系；分别基于有效扩散系数和非定常环境作用输入提出细-宏观转译模式，并进一步建立宏观尺度上考虑转译的数值模拟方法，解决精度与效率的矛盾。在此基础上，面向桥梁混凝土结构耐久性特点对转译方法进行完善：研究多因素耦合作用的转译方法，准确模拟实际工程状态；研究数值模拟结果的统计与概率特性，及其设计表达形式与应用方法。项目成果将为桥梁混凝土结构耐久性设计的参数、工具、方法提供新的基础数据和理论。

桥梁混凝土结构的耐久性具有特殊性、复杂性，以及显著的时空变异性；目前理论与实践仍有差距的重要原因是对影响退化过程各种因素的变异性的描述不充分；基于单一尺度研究方法与工具解决这一问题有一定难度。.研究总结了主流混凝土细观建模方法的算法机理，对不同建模方法的算法效率和模拟效果进行了对比分析，结合不同的建模需求（包括大尺寸、高密度以及骨料堆积效应模拟）推荐了相应的细观建模方法；基于真实材料截面高精度图像识别结果，提取了骨料颗粒几何参数的概率特征，提出了骨料“三参数模型”以及对应的多重抽样建模方法，实现混凝土材料的精细化建模，为耐久性退化多尺度数值模拟提供支撑。.研究考虑骨料随机分布影响，开展混凝土氯盐侵蚀、碳化反应、水分传输和热量传导的精细化数值模拟，结合工程结构复杂服役环境开展多因素耦合的耐久性能数值模拟；基于骨料颗粒的空间随机分布特性，结合概率代表体积单元方法提出了细宏观跨尺度转译模型改进的介观尺度模型，并对转译单元尺寸进行优化，通过对比分析校验了转译模型在效率和精度上提升。针对耐久性能退化中经典的物质传输和热量传递机理，引入元胞自动机方法，将传统有限元方法中全局浓度矩阵求解转化为材料局部状态更新，大幅提升耐久性分析的求解效率。.研究针对桥梁混凝土结构服役环境特征，引入细宏观转译模型，开展多因素耦合作用的结构全断面快速仿真分析，为桥梁混凝土结构耐久性能演化预测提供支撑；结合钢筋锈蚀概率和保护层开裂损伤时机等关键指标，实现了开展长期服役结构耐久性能和服役性能的可靠度求解分析；研究了气候变化以及人类活动导致的除冰盐使用和环境作用变化影响，研究了不同管养维护方案对结构长期耐久性能的影响作用，为实际工程结构服役管养决策提供支撑。.研究成果实现了建立桥梁混凝土结构耐久性退化细宏观数值模拟转译理论的总体目标，为混凝土结构耐久性预测、评估和设计防护等提供了关键的基础模型和科学方法。