基于声发射能量模型的胶合木结构蠕变机理与演变规律研究

Stability and durability of modern timber buildings mainly depend on the creep behavior of their structural glulams. In theory, creep property is usually described by constitutive relation models on the deformation. Nevertheless, all the existing creep models are built with linear and continuous functions. Obviously, these models disagree with the real nonlinear creep process. In fact, deformation is only the macro-scene of the creep behavior. Conversely, the creep process can be rebuilt by strain energy, and then the strain energy will be released by acoustic emission (AE) signals. For this reason, a series of constructive relation models based on AE characteristics will be built to describe the creep mechanism and evolvement in this project. At first, a multi channel AE acquisition system will be constructed with AE sensors and high-speed data acquisition equipments, and then, the approaches are designed to extract real AE signals from the original noisy ones and reconfigure AE waves based on wavelet theory. The propagation of AE in different directions in the glulams also will be studied through experiments with artificial AE sources. Secondly, regarding the AE frequency as a parameter, a strain-frequency-time constructive relation model will be built to identify the stages of creep behavior of the glulams. In order to reveal the creep mechanism and evolution, the cumulative energy of AE waves during creep process will be used as a parameter and then a strain-energy-time constructive relation model will be constructed. At last, with the help of location approaches of AE sources, a series of evaluation criterions will be established to dynamically monitor and predict the damage degree of creep behaviors in glulams. The research results can provide a new effective method for analyzing and forecasting of creep behavior in structural glulams that is helpful to guarantee the security of the glulams. Furthermore, the results can be directly used to real time dynamically monitor and evaluate the creep damage of modern timber buildings.

胶合木结构蠕变行为是决定木结构建筑稳定和耐久的关键因素，蠕变模型是研究蠕变特性的根本方法，现有蠕变模型都是基于形变的线性连续数学模型，与实际非线性蠕变过程不符。形变只是蠕变行为的宏观表现，能量才是反映蠕变行为的特征变量，而蠕变过程又是通过声发射方式释放应变能，为此，本项目拟基于蠕变过程声发射特征构建胶合木结构蠕变本构关系模型，从能量角度揭示蠕变机理及演变规律。首先，基于小波理论研究蠕变过程声发射信号在胶合木中的传播规律；其次，分别以声发射信号频率和能量为特征变量，采用数值分析方法构建反映蠕变成因和状态的应变-频率-时间本构关系模型，以及反映蠕变机理和演变规律的应变-能量-时间本构关系模型；最后，在蠕变模型的基础上，依据声发射源定位理论，研究胶合木结构蠕变损伤动态评价与预测模型。不仅为胶合木结构蠕变机理与规律研究提供基于能量的理论模型，同时提出了一种有效的木结构建筑动态安全监测方法。

胶合木结构蠕变行为是决定木结构建筑稳定和耐久的关键因素，形变只是蠕变行为的宏观表现，能量才是反映蠕变行为的特征变量，而蠕变过程又是通过声发射（AE）方式释放应变能，为此，本项目拟基于蠕变过程AE特征构建胶合木结构蠕变本构关系模型，从能量角度揭示蠕变机理及演变规律。研究内容主要包括：1）搭建多通道木材AE信号采集系统，针对AE信号降噪问题，设计相应的信号处理与波形重构算法；2）研究对应不同木材损伤类型的AE信号时频域特征；3）研究AE信号在胶合木及木材中传播的速度与能量衰减规律；4）设计木材AE源定位算法，研究胶合木及木材损伤动态监测与评价方法；5）依据弹性波理论，研究AE信号在木材中以横波与纵波形式传播的速度与能量衰减规律。研究取得的主要成果为：1）自主搭建了8个通道、最大采样频率为2MHz的木材AE信号采集系统，提出了基于小波及EMD的木材AE信号波形重构算法；2）采样信号相关性分析与谱白化方法，建立AE信号在胶合木及木材中的各向异性速度计算模型以及相应的能量衰减模型。3）针对胶合木及木材不同损伤类型，提出了基于频率的AE事件辨识方法，并且依据AE信号随机特性，提出了基于AE信号信息熵和分形维数的木材损伤过程应变能集中释放判别方法，从而建立基于AE信号的胶合木及木材损伤动态监测与评价方法。4）依据弹性波理论，建立了AE信号分别在针叶和阔叶树材中表面横波与内部纵波的速度与能量传播模型。项目搭建的多通道AE信号采集系统为木材AE检测技术的理论与应用研究提供了基本的技术保障，项目研究所得到了AE信号在胶合木及木材中的速度及能量模型是木材AE无损动态监测基本理论的补充与完善，项目研究提出的AE源定位算法和基于频率的AE事件辨识方法，为胶合木及木材损伤动态监测与评价提供了一种有效途径。