Strain Measurements Under Damage and Environmental Effects from a Real-Scale Masonry Testbed via Smart Brick Sensors

Structural health monitoring (SHM) of masonry structures suffers from a critical scarcity of real-world strain data, particularly measurements capturing progressive damage under controlled yet realistic conditions. This dataset addresses that gap by providing strain time-series acquired through smart bricks: piezoresistive brick-like sensors, embedded directly within the fabric of the structure. Their seamless integration into the masonry allows for continuous, minimally invasive monitoring. The measurements were collected from a full-scale masonry prototype, a two-story building archetype constructed outdoors and left exposed to natural environmental conditions over multiple monitoring campaigns. As a result, the recorded strain signals carry the simultaneous imprint of two distinct sources of variation: structural damage and environmental effects (thermal cycles, humidity, seasonal drifts). This dual nature makes the dataset particularly suited for developing and benchmarking compensation strategies, novelty detection methods, and damage identification algorithms under realistic operating conditions. Three damage scenarios of increasing severity were induced on the structure. The first involved the sudden release of two central tie-rods (out of four originally installed). The second consisted of incremental static overloading applied to the roof slab. The third, and most severe, was a progressively induced differential foundation settlement. Each of the three accompanying CSV files corresponds to one damage scenario. All files share the same structure: the first two columns encode the timestamp (date and time), followed by thirteen measurement channels (SB1 to SB13), each representing the strain output of one smart brick. Each file contains approximately 720 time steps. For sensor placement within the structure and further details on sensor layout and damage scenarios, readers are referred to the linked publication, where a damage detection strategy based on linear cointegration is demonstrated using this dataset. Additional publications exploiting this dataset will be linked as they become available.砖石结构的结构健康监测（Structural Health Monitoring, SHM）领域长期面临真实世界应变数据严重匮乏的问题，尤其是在可控且贴近实际的工况下捕捉渐进式损伤的测量数据。本数据集通过智能砖（smart bricks）——即直接嵌入结构本体的压阻式仿砖传感器——采集得到应变时间序列数据，填补了这一空白。这类传感器可无缝融入砖石结构，实现连续且微创的监测。 本次测量数据采集自一座足尺砖石结构原型——一座搭建于户外的两层建筑原型，在多轮监测任务中持续暴露于自然环境中。因此，记录得到的应变信号同时包含两类不同变异来源的特征：结构损伤与环境影响（包括温度循环、湿度变化与季节性漂移）。这种双重特性使得本数据集特别适用于在真实工况下开发、验证补偿策略、异常检测方法以及损伤识别算法。 该原型结构共设置了三级严重程度逐步升级的损伤工况：第一级为突然松开原安装的四根中心拉杆中的两根；第二级为对屋面板施加渐进式静态超载；第三级（亦是最严重的一级）为逐步诱发基础不均匀沉降。 随数据集附带的三个CSV文件分别对应一种损伤工况，所有文件的结构一致：前两列为时间戳（日期与时刻），其后为13个测量通道（SB1至SB13），每个通道对应一块智能砖的应变输出。每个文件约包含720个时间步。关于传感器在结构中的布设位置、传感器布局以及损伤工况的更多细节，请参阅关联的学术论文，该论文基于本数据集验证了一种基于线性协整的损伤检测策略。后续基于本数据集的相关研究成果将陆续更新关联链接。