Physics-Based Framework for Computational Simulation of Reinforced Concrete Structures

最近，美国土木工程师学会（American Society of Civil Engineers）就评估基础设施的状况所作的报告，将整体评级定为D+，这是相当不理想的。结构的长期老化和恶化对自然和人为灾害事件中的结构弹性构成重大威胁。钢筋混凝土结构由于钢筋与混凝土之间的相互作用而难以进行分析建模。因此，该项目的总体目标是提高工程界的能力，通过建立一个大规模的基于物理的基础设施的模拟计算框架来分析钢筋混凝土结构的计算。准确的计算模拟模型，它将有可能模拟恶化剂，如腐蚀，裂缝，化学侵蚀混凝土等，这项研究将促进老化混凝土结构的破坏机理的理解。该研究所涉及的严格的数学和计算发展将为研究生提供理想的培训平台，并将为土木工程本科课程的教育发展带来机会。准脆性材料和结构在长期劣化情况下的破坏行为的精确分析建模是一个复杂的问题，仍然没有解决。可靠的模拟需要准确描述各种中尺度现象，这是不考虑现有的唯象计算模型。细观尺度（混凝土中粗骨料颗粒的尺度）模型捕捉这些细观尺度的故障特征，但从计算的角度来看，它们是非常密集的。基于最近制定的离散模型称为格离散粒子模型（LDPM），有效的多尺度技术适用于放大离散系统将在这项研究中开发。该项目对知识库的主要贡献是制定和验证一个多尺度框架，用于模拟钢筋混凝土（1）最常见的长期退化机制，例如，冻融循环、碱-硅反应、收缩和徐变;（2）包括将混凝土模型与钢筋及其与钢筋腐蚀相关的膨胀效应耦合的粘结算法;（3）最后，（4）具有用于钢筋混凝土框架模拟的多尺度算法。

The recent American Society of Civil Engineers report on the assessment of infrastructure condition resulted in an overall grade of D+, which is rather unsatisfactory. Long-term aging and deterioration of structures pose a significant threat to structural resiliency in the event of natural and man-made hazards. Reinforced concrete structures are difficult to model analytically because of interaction between concrete and reinforcing bars. Hence, the overarching goal of this project is to improve the capability of the engineering community to analyze reinforced concrete structures computationally by establishing a large-scale physics-based computational framework for the simulation of infrastructure. With accurate computational simulation models it would be possible to simulate deteriorating agents such as corrosion, cracks, chemical attack on concrete etc. This research will advance the understanding of failure mechanisms of aged concrete structures. The rigorous mathematical and computational developments involved in the research will provide an ideal training platform for graduate students and will bring opportunities for educational advancement in the civil engineering undergraduate curriculum.Accurate analytical modeling of failure behavior of quasi-brittle materials and structures in presence of long-term deterioration is a complicated problem that is still unsolved. Reliable simulations require accurate descriptions of various meso-scale phenomena, which are not taken into account by available phenomenological computational models. Meso-scale (the scale of coarse aggregate particles in concrete) models capture these meso-scale features of failure but they are extremely intensive from a computational point of view. Based on a recently formulated discrete model called the Lattice Discrete Particle Model (LDPM), effective multi-scale techniques suitable for up-scaling discrete systems will be developed in this research. The main contribution of this project to the knowledge base will be the formulation and validation of a multi-scale framework for the simulation of reinforced concrete (1) accounting for the most common long-term deterioration mechanisms, e.g., freeze-thaw cycles, alkali-silica reaction, shrinkage and creep; (2) including a bond algorithm to couple the concrete model with steel reinforcing bars and their expansive effect associated with steel corrosion; and, finally, (3) featuring multiple scale algorithms for the simulation of reinforced concrete frames.