节点文献

基于随机骨料模型的混凝土细观损伤与宏观强度研究

Studies on Meso-Level Damage and Macroscopical Strength of Concrete by Random Particle Model

【作者】 刘强

【导师】 彭一江;

【作者基本信息】 北京工业大学 , 水工结构工程, 2005, 硕士

【摘要】 混凝土是工程中广泛使用的一种建筑材料。其力学特性的研究对充分发挥材料的强度,提高设计水平和降低工程造价具有十分重要的意义。科研及技术人员采用不同的试验设备和试验手段对混凝土的力学特性进行了大量的试验。近年来,随着高速度、大容量计算机的出现以及有限元法等计算技术的迅速发展,传统的试验方法及设计分析方法面临着挑战。研究人员基于不同的试验模型,采用不同的数值计算方法进行了数值实验。数值实验可以克服普通实验室试验的一些不足,有利于解释实验中发生的各种现象,并为改进试验设计提供数值分析基础。此项研究工作是目前混凝土材料研究领域的重要发展趋势之一,细观力学的研究已经在混凝土的微观结构和宏观力学特性之间架起了一座桥梁。本文采用随机骨料模型和双折线损伤模型相结合,在细观层次上对混凝土在单轴荷载作用下的力学特性和破坏机理进行了数值研究。并对混凝土的强度尺寸效应和材料参数的影响进行了数值计算与分析研究。论文从混凝土的内部结构入手,认为它是由粗骨料、水泥砂浆基体和界面粘结带组成的非均匀质的复合材料。由Walraven 公式确定各级骨料颗粒,并用蒙特卡罗法,在平面上随机确定骨料的位置和数量,生成随机骨料结构,然后再将其剖分为有限元网格。利用电算程序将不同的材料特性分配给相应的单元,即形成了数值试验所用的“试件”。最后采用本人自编的有限元程序计算在逐级加载条件下混凝土内部单元的应力与应变,采用双折线损伤模型判别单元的损伤及破坏情况并行逐次迭代计算,直至计算出混凝土的宏观强度和裂纹扩展结果。本论文提出将计算所得双折线损伤模型中损伤单元、残余强度单元和最终破坏单元的个数的稳定性作为迭代过程中的位移收敛条件,程序计算中对不同性质的单元采用不同颜色的方法,从而可以在计算过程中同步显示试件中单元的损伤程度和裂纹扩展过程。论文将数值实验结果与试验成果进行了对比分析。本论文主要做了以下几个方面的工作: (1) 阐述了混凝土的一般特性及进行混凝土细观研究的必要性。(2) 分析研究了混凝土的细观损伤破坏理论,以及混凝土的变形和破坏的规律。(3) 对混凝土骨料分布的随机特性及Fuller曲线的适用性进行了验证和研究。

【Abstract】 The concrete si a kind of building material which is widely used in engineering. Studies for the mechanical properties of concrete are very important to increase the strength of concrete material, improve design level and reduce the construction cost. Some tests of concrete used to be done by researchers using some different equipments and test methods. In recent years, the tradition test methods and design approaches have been confronted with the challenges of the appearance of high speed and large capacity computers and the development of finite element method. Scientific researchers carried on a lot of numerical experiments by different test models and different calculating methods. Numerical experiments are easy to overcome the limitation in laboratory tests and explain the various phenomena happened in the experiments, and they can provide some very useful numerical analysis data to improve experiments. The research work is one of the important development trends in the areas of meso-level mechanics research currently, the study in the meso-level mechanics has already built a bridge between the meso-level structure and the macroscopic mechanics properties. In this paper, the random particle model and double lines damage model are used, and uniaxial mechanical properties of concrete and the damage mechanism are integrated on meso-level. The influences of size effect and parameter of materials are researched. The inner structure of concrete specimens is analyzed in meso-level and it is taken as composite structure of aggregates, matrix and the interfacial transition zone(the zone between aggregate and matrix). Using Walraven function and Monte-Carlo method ,the positions and number of aggregates are identified in plane, hence the random aggregate structures are generated. Then they are separated into the finite elements mesh. After the different material properties are assigned to respective elements according to the locations of elements in individual phases, a two-dimensional numerical model for the composite structure of concrete is formed. The numerical experiment can be carried out using a finite element method program. The program has the ability to carry on the automatic distribution to the strength parameter of materials by projection method. The classification loading principle is first time used by counting the number of damaged elements, residual strength elements and the latest broken elements using double lines damage model in the research. The program also can fill the different elements with deferent color in deferent step, and show these damaged elements and broken elements on screen. The results of the numerical test and laboratory test are compared. The following works have been finished in this paper. (1) The general characteristic of the concrete and the necessity on meso-level study of the concrete were reviewed. (2) The meso-level damage, broken theory and the regularities of deformation and destruction for concrete were studied.

  • 【分类号】TU528
  • 【被引频次】25
  • 【下载频次】1376
节点文献中: 

本文链接的文献网络图示:

本文的引文网络