不依赖基准模型数据的钢筋混凝土梁桥多损伤阶段状态评价

Damage identification of bridge is a vital challenge around the world, and the vibration-based linear methods have been widely used. However, many of them require the modal data of intact structure as baseline and are generally applied to the linear situation. In practice, the damaged bridges are built decades ago, the baseline parameters are difficult to obtain, and the fatigue damages usually exhibit nonlinear behavior. Therefore, the research on damage characteristics, numerical simulation and corresponding identification methods are conducted for the damage detection of reinforced concrete bridge. The detailed contents are as follows: Firstly, the development law for cracks of reinforced concrete bridge under external loads is revealed based on laboratory tests for bridge model and the relationships are observed between cracks and mechanical responses. Secondly, the constitutive equations for concrete, steel bars and their bonding interface are proposed based on irreversible processes thermodynamics and continuum damage mechanics. Three dimensional simulation of reinforced concrete bridge is realized. Finally, the linear and nonlinear damage indicators are constructed through responses from impulsive and harmonic vibration tests, respectively. Different damage identification strategies are presented for various stages of cracks. Proof-of-concept validation is carried out to verify the effectiveness of the proposed methods. The implementation of this project can breakthrough in key techniques of damage detection under real damage situation without considering baseline parameters and provide novel methods for the damage assessment of reinforced concrete bridge.

桥梁结构损伤识别是世界各国面临的重大难题，基于振动的线性损伤识别方法得到广泛应用。但该方法一方面需要完好结构响应作为基准数据，另一方面主要针对结构线性损伤。然而，损伤桥梁大多服役时间较长，基准模型数据难以获取，且其疲劳损伤呈现非线性特征。因此，本项目以广泛使用的钢筋混凝土梁桥为研究对象，对其损伤特征、数值模拟和识别方法展开研究。具体内容包括：通过室内模型梁试验，揭示混凝土梁桥在外部荷载作用下裂缝发展规律及其与结构力学响应相关关系；基于不可逆热力学和连续损伤力学，提出适用于表征非线性损伤的混凝土、钢筋及其界面粘结本构模型，实现混凝土梁桥三维仿真；通过结构脉冲及谐波激励响应，构造线性和非线性损伤识别指标，提出适用于不同裂缝阶段的不依赖基准模型数据的识别策略，并进行概念性验证。本项目的顺利实施可以突破真实损伤状况下不依赖基准模型数据的损伤识别技术，为钢筋混凝土梁桥损伤识别提供新思路和新方法。

我国在役桥梁中钢筋混凝土梁桥的占比很大，针对该桥型的损伤识别对于保障结构安全运营意义深远。考虑到在役桥梁的基准模型数据获取困难，提出一种不依赖基准模型数据的损伤识别方法具有重要的现实意义。本项目通过钢筋混凝土模型梁四点弯曲试验和分级加载，获取模型梁的裂缝发展规律。基于声发射实时监测和导电膜技术，通过声发射参数及导电膜电阻变化，实现了裂缝演变规律表征。通过钢筋混凝土桥梁多裂缝阶段静动力响应分析，构建了裂缝与静动力响应间关系模型，为损伤识别方法提出奠定基础。构建钢筋混凝土梁三维有限元仿真模型，实现了试验加载过程的有效模拟，计算得到了钢筋混凝土梁的变形规律以及裂缝产生过程。基于双线性疲劳裂缝理论和非线性接触单元模型，实现了钢筋混凝土梁桥疲劳裂缝的非线性仿真。通过非线性疲劳损伤对谐波振动响应敏感性分析，阐明了加速度时程曲线图、相谱图、功率谱图等谐波振动响应与裂缝位置和深度的关系，明确了不同损伤位置与不同损伤深度下疲劳裂缝对结构谐波振动响应的影响。通过构建相谱图畸变值、谐能滋生值和总能滋生值损伤识别指标，实现了非线性接触疲劳裂缝损伤表征。依赖疲劳裂缝梁模态振型数据，构造了振型递变值和振型曲率两类损伤位置识别指标，实现了疲劳裂缝的损伤位置识别。以谐波响应参数为粒子群优化支持向量回归模型的输入参数，实现了疲劳裂缝的损伤程度识别。通过神经网络拟合及切比雪夫多项式拟合得到结构未损情况下的模态曲率值，进而计算得到结构损伤前后的模态曲率差值，实现开口裂缝的损伤位置识别。以归一化的模态曲率差值作为模糊推理系统的输入参数，实现开口裂缝的损伤程度识别。选取槽梁为室内试验模型，验证了本项目所提出方法的实际应用效果。采用粒子群优化模糊C均值聚类算法，形成损伤识别样本库，采用受噪声影响的振型数据和不完备的振型数据进行结构损伤位置及程度的损伤识别，增强了方法的鲁棒性。