Study on the early warning mechanism for real-time monitored structural responses of a historical timber building

Highlights

• The early warning mechanism for historical timber building is presented.

• Early warning thresholds based on statistical and extreme value analysis of past monitored structural responses are proposed.

• Early warnings for substructures are realized by efficiency coefficient data fusion method.

Abstract

Structural Health Monitoring (SHM) is of great importance for the preventive protection of historical buildings. Real-time monitoring and early warning of the structural response have advantages of timeliness and early detection of abnormal structural conditions. Feiyun Wood Pavilion, a historical timber building in China, was taken as the object to study the early warning mechanism for real-time monitored structural responses. An early warning mechanism for historical timber building has been presented under different data accumulated time. The early warning thresholds of structural responses based on statistical and extreme value analysis of the accumulated monitoring data are proposed. Early warning for substructure of the building is realized by data fusion based on the efficiency coefficient method. The effectiveness of the proposed mechanism was verified by real monitoring data. The proposed mechanism gives an early warning solution of the SHM that can be references for existing and other historical buildings.

Introduction

Historical buildings have a long history. In the service process, due to the long-term effects of existing environmental loads and adverse factors such as fatigue, corrosion effects and material aging, the structure will inevitably experience damage accumulation and a reduction in resistance [1]. Most historical buildings undergo required maintenance, but most of the maintenance projects have been initiated to rescue the structure or building after obvious damage has occurred. In recent years, the “preventive protection” idea of “prevention before disaster is better than post disaster protection” has been gradually accepted in the field of cultural relic protection [2]. Preventive protection emphasizes the importance of determining the risk factors faced by historical buildings and obtaining their rules of development and change based on information collection and risk assessment. Through disaster prevention, daily maintenance, scientific management and other measures taken to reduce or eliminate the risks that threaten historical buildings, the condition of these buildings will be maintained in a good state for a long time, and their comprehensive protection can be realized. From the basic connotation of preventive protection, it can be seen that its foothold is 'prevention' rather than 'repair'.

Structural Health Monitoring (SHM) technology obtains information on the structure and relevant environmental factors by using sensing and data acquisition devices and then evaluates the service status of the structure through data analysis to provide an early warning of the safety status of the structure and surrounding environment and to offer reference and guidance for structural repair, maintenance and management decisions [3], [4], [5], [6], [7], [8]. In recent decades, this technology has become a popular research subject in the field of structural engineering and has rapidly developed. SHM technology has been widely applied to many important engineering structures [9], [10], [11], [12], [13], such as bridge, dam and high rise building. SHM has been performed on historical buildings with the gradual maturity of technology. An optical inclinometer and digital network camera were installed on the Portogruaro Civic Tower to monitor the building inclination [14]. The Aquila tower was monitored by wireless sensor technology, and its displacement, acceleration and temperature were measured [15]. Accelerometers, temperature sensors and humidity sensors were installed on the bell tower of the Benedictine Abbey of San Pietro in Perugia, Italy, to gain knowledge on the effects of changes in temperature and humidity on the natural frequencies of slender masonry buildings [16]. A continuous dynamic monitoring system has been installed on the historical masonry bell tower of Ficarolo in Italy to investigate the dynamic characteristics of the structure and check the possible evolution of the structural behavior [17]. A SHM system was installed on the Roman amphitheater (Arena) of Verona in 2011 through a sensor network installed in relevant positions on the monument to test the main cracks and dynamic parameters and control the reversibility of the seasonal displacements or deformation trends of the monument [18]. Seven Literary World has installed a structural health monitoring system based on fiber Bragg grating, which aims to monitor the structural deformation around the building [19]. The main hall of Potala Palace in Tibet is also equipped with strain transducers, tilt sensors and other sensors to monitor the long-term stress and deformation of the structural components [20], [21].The Yingxian Wooden Tower is equipped with strain transducers, tilt sensors and crackmeters to obtain the deformation information of the tower [22]. Due to the important cultural value of historical buildings, their repair projects should be carefully implemented. On the one hand, the implementation of SHM technology can catch the abnormal state of a structure earlier to ensure the safety of a historical building through maintenance; on the other hand, it can judge the change rate and stability of the structural state through the cumulative change in data to provide a scientific basis for whether to start a repair or not, which can also avoid destructive or unnecessary repair.

An early warning is one of the main functions of SHM. The purpose of early warnings for structural safety is to improve the prevention of and the ability to respond to structural emergencies and to control, reduce and eliminate the caused structural damages. The core content of an early warning mainly includes the selection of early warning indicators, the setting of early warning indicator thresholds, and the construction of early warning models. Recent early warning researches mainly focus on monitoring data abnormality warnings and structural performance warnings. Yi et al. [23] summarized the basic types of abnormal data and their manifestations so that the early warning of abnormal data can be realized. Hemandez-Garcia and Masri adopted a multivariate statistical process control method based on principal component analysis, independent component analysis, and improved independent component analysis to provide the early warning of bridge monitoring data [24]. Yi et al. [25] analyzed the early warning of abnormal data in bridge deformation monitoring by using the cumulative sum (CUSUM) chart. Rao et al. [26] constructed a Hankel matrix for the monitoring data and then established a principal component analysis model for the matrix to implement the early warning of abnormal data. Based on the vibration acceleration monitoring data, the modal flexibility [27] and modal curvature [28] derived from the modal parameters are extracted as early warning indicators, which can effectively identify structural damage. Magalhaes et al. [29] realized the early warning of damage to a long-span arch bridge based on the Markov distance of the frequency residual after excluding time-varying load effects. Huang et al. [30] proposed an early warning method for the performance of expansion joints based on the Shewhart mean control chart and used the kernel density estimation method to determine the control limit of the control chart.

In structural early warning, the setting of an early warning threshold is always a difficult problem in SHM technology. For engineering structures, the early warning threshold can be set according to the design allowable value, theoretical calculation value, numerical analysis value and monitoring data value. However most of the historical buildings have no design data, their initial state is unknown, and most of the existing structures have damage and deformation, which introduces great difficulty into the theoretical calculation and numerical analysis [31]. Although some simplified methods can be adopted for structural calculations, there is still a large gap between the realization of the calculations and the real situation of the structure. In contrast, the measured data and information obtained by SHM can comparatively reflect the performance state of the structure, so it is feasible to use the analysis and processing results of the monitoring data to provide a warning for the change in the structure state of historical buildings. A historical building structure has typical time-varying characteristics. With the continuous accumulation of monitoring data, the early warning threshold should be checked, supplemented, modified and optimized regularly to produce a more scientifically reasonable result.

Taking the structure and its SHM system of the Feiyun Wood Pavilion which is a typical representative of historical timber building in China as a research object, this paper study the early warning mechanism for real-time monitored structural responses. The design purpose and equipment composition of the SHM system on Feiyun Wood Pavilion is introduced. The conceptual early warning mechanism for historical timber buildings under conditions of past monitoring data with different accumulated time is presented. The early warning mechanism and thresholds of monitored structural responses based on statistical and extreme value analysis of the accumulated monitoring data are proposed. Early warning for different substructures can be realized by data fusion based on the efficiency coefficient method. The effectiveness of the proposed set of analysis is verified by real monitoring data from the SHM system of Feiyun Wood Pavilion.