基于浆-骨体系对收缩应力响应的混凝土收缩变形机理及混沌预测方法研究

In order to achieve excellent stress performance, the water-cement ratio of concrete used in modern concrete structures is decreasing, and the strength of concrete restrained is increasing, which has brought more and more serious shrinkage and cracking problems to the structure. Accurately grasp the shrinkage characteristics of concrete is the key to solve this problem. Project from a new perspective, based on the two basic points which restrict the shrinkage characteristics of concrete, the theoretical and experimental research is carried out under the coupling environment of stress field, temperature field and humidity field. On the basis of revealing the relationship among pore structure, moisture and shrinkage stress of concrete, and clarifying the changing law of shrinkage stress of concrete, the shrinkage stress response test is used. The response characteristics of mortar, aggregate and mortar-aggregate system to shrinkage stress are revealed, and the method of determining the elastic modulus of concrete to the shrinkage stress is established. Combined with the theory of micromechanics of concrete, the theoretical model of concrete mortar-aggregate coupling deformation with general mechanical significance is established, and the mechanism of shrinkage and deformation of concrete is further clarified. Based on the artificial intelligence method of fuzzy theory and the construction of concrete shrinkage multivariable chaotic system, a concrete shrinkage characteristic prediction model based on multivariable chaos theory and ultimate learning machine is proposed. The concrete shrinkage characteristics can be accurately described and predicted. The research results play an important role in ensuring the safety, applicability and durability of concrete structures.

为追求卓越的受力性能，现代混凝土结构所用混凝土水灰比不断降低，混凝土所受约束强度不断增强，给结构带来了日益严重的收缩开裂问题。准确掌握混凝土收缩特性是解决这一问题的关键。项目以新角度，从制约混凝土收缩特性的两个基本点出发，通过在应力、温度以及湿度场等多场耦合环境下的理论及试验研究：在揭示混凝土孔结构-水分-收缩应力三者关系，阐明混凝土收缩应力变化规律的基础上，利用收缩应力响应试验，揭示砂浆、骨料及其形成的浆—骨体系对收缩应力响应特性，建立混凝土相对于收缩应力的弹性模量确定方法；结合混凝土细观力学理论，建立具有普遍力学意义的混凝土浆—骨耦合变形模型，进一步阐明混凝土收缩变形机理。基于模糊理论的人工智能方法，构建混凝土收缩多变量混沌系统，提出一种基于多变量混沌理论和极限学习机的混凝土收缩特性预测模型，实现对混凝土收缩特性的准确描述与预测。研究成果对于保障混凝土结构安全、适用、耐久具有重要作用。

为准确掌握混凝土收缩特性，预测混凝土收缩，从而解决混凝土收缩开裂问题。项目通过理论分析、数值模拟和试验研究，阐明了混凝土孔结构-水分-收缩应力三者关系， 揭示了混凝土收缩应力及砂浆、骨料及其形成的浆—骨体系对收缩应力响应特性；建立了混凝土相对于收缩应力的弹性模量确定方法；结合混凝土细观力学理论，建立了混凝土浆—骨耦合变形模型，形成了数值计算方法；基于人工智能方法，利用树模型和混沌模型构建了基于多特征影响的多场耦合环境下混凝土收缩预测模型，实现对混凝土收缩特性的准确描述与预测。研究成果成功应用于国内多个重大工程，有效防止了混凝土结构施工期开裂，取得了良好的经济、社会效益。