the state of the art in control and health monitoring of civil engineering structures[1].Salawu (1997)discussed and reviewed the use of natural frequency as a diagnostic parameter in structural assessment procedures using vibration monitoring.Doebling, Farrar, et al. (1998)presented a through review of the damage detection methods by examining changes in dynamic properties.Zou, TongandSteven (2000)summarized the methods of vibration-based damage and health monitoring for composite structures, especially in delamination modeling techniques and delamination detection.
4 Sensors and optimum placement
One of the problems facing structural health monitoring is that very little is known about the actual stress and strains in a structure under external excitations. For example, the standard earthquake recordings are made ofmotions of the floors of the structure and no recordings are made of the actual stresses and strains in structural members. There is a need for special sensors to determine the actual performance of structural members. Structural health monitoring requires integrated sensor functionality to measure changes in external environmental conditions, signal processing functionality to acquire, process, and combine multi-sensor and multi-measured information. Individual sensors and instrumented sensor systems are then required to provide such multiplexed information.
FuandMoosa (2000)proposed probabilistic advancing cross-diagnosis method to diagnosis-decision making for structural health monitoring. It was experimented in the laboratory respectively using a coherent laser radar system and a CCD high-resolution camera. Results showed that this method was promising for field application. Another new idea is thatneural networktechniques are used to place sensors. For example,WordenandBurrows (2001)used the neural network and methods of combinatorial optimization to locate and classify faults.
The static and dynamic data are collected from all kinds of sensorswhich are installed on the measured structures.And these datawill be processed and usable informationwill be extracted. So the sensitivity, accuracy, and locations,etc. of sensors are very important for the damage detections. The more information are obtained, the damage identification will be conducted more easily, but the price should be considered. That’s why the sensors are determined in an optimal ornearoptimal distribution. In aword, the theory and validation ofoptimumsensor locationswill still being developed.
5 Examples of health monitoring implementation
In order for the technology to advance sufficiently to become an operational system for the maintenance and safety of civil structures, it is of paramount importance that new analytical developments are ultimately verified with appropriate data obtained frommonitoring systems, which have been implemented on civil structures, such as bridges.
Mufti (2001)summarized the applications of SHM of Canadian bridge engineering, including fibre-reinforced polymers sensors, remote monitoring, intelligent processing, practical applications in bridge engineering, and technology utilization. Further study and applications are still being conducted now.FujinoandAbe(2001)introduced the research and development of SHMsystems at the Bridge and Structural Lab of the University of Tokyo. They also presented the ambient vibration based approaches forLaser DopplerVibrometer (LDV) and the applications in the long-span suspension bridges.
The extraction of the measured data is very hard work because it is hard to separate changes in vibration signature duo to damage form changes, normal usage, changes in boundary conditions, or the release of the connection joints.
Newbridges offer opportunities for developing complete structural health monitoring systems for bridge inspection and condition evaluation from“cradle to grave”of the bridges. Existing bridges provide challenges for applying state-of-the-art in structural health monitoring technologies to determine the current conditions of the structural element,connections and systems, to formulate model for estimating the rate of degradation, and to predict the existing and the future capacities of the structural components and systems. Advanced health monitoring systems may lead to better understanding of structural behavior and significant improvements of design, as well as the reduction of the structural inspection requirements. Great benefits due to the introduction of SHM are being accepted by owners, managers, bridge engineers, etc..
6 Research and development needs
Most damage detection theories and practices are formulated based on the following assumption: that failure or deterioration would primarily affect the stiffness and therefore affect the modal characteristics of the dynamic response of the structure. This is seldom true in practice, because①Traditional modal parameters (natural frequency, damping ratio and mode shapes, etc.) are not sensitive enough to identify
and locate damage. The estimation methods usually assume that structures are linear and proportional damping systems.②Most currently used damage indices depend on the severity of the damage, which is impractical in the field. Most civil engineering structures, such as highway bridges, have redundancy in design and large in size with low
natural
frequencies.
Any
damage
index
should
consider
these
factors.③Scaledmodelingtechniques are used in currentbridge damage detection. Asingle beam/girder models cannot simulate the true behavior of a real bridge. Similitude
laws
for
dynamic
simulation
and
testing
should
be
considered.④Manymethods usually use the undamaged structural modal parameters as the baseline comparedwith the damaged information. This will result in the need of a large data storage capacity for complex structures. But in practice,there are majority of existing structures for which baseline modal responses are not available. Only one developed method(StubbsandKim (1996)), which tried to quantify damagewithout using a baseline, may be a solution to this difficulty. There is a lot of researchwork to do in this direction.⑤Seldommethods have the ability to distinguish the type of damages on bridge structures. To establish the direct relationship between the various damage patterns and the changes of vibrational signatures is not a simple work.
Health monitoring requires clearly defined performance criteria, a set of corresponding condition indicators and global and local damage and deterioration indices, which should help diagnose reasons for changes in condition indicators. It is implausible to expect that damage can be reliably detected or tracked by using a single damage index. We note that many additional localized damage indiceswhich relate to highly localized properties ofmaterials or the circumstances may indicate a susceptibility of deterioration such as the presence of corrosive environments around reinforcing steel in concrete, should be also integrated into the health monitoring systems.
There is now a considerable research and development effort in academia, industry, and management department regarding global healthmonitoring for civil engineering structures. Several commercial structural monitoring systems currently exist, but further development is needed in commercialization of the technology. We must realize that damage detection and health monitoring for bridge structures by means of vibration signature analysis is a very difficult task. It
contains several necessary steps, including defining indicators on variations of structural physical condition, dynamic testing to extract such indication parameters,
defining the type of damages and remaining capacity or life of the structure, relating the parameters to the defined damage/aging. Unfortunately, to date, no one has accomplished the above steps. There is a lot of work to do in future.
桥梁健康监测应用与研究现状
摘要
桥梁损伤诊断与健康监测是近年来国际上的研究热点,在实践方面,土木工程和航空航天工程、机械工程有明显的差别,比如桥梁结构以及其他大多数土木结构,尺寸大、质量重,具有较低的自然频率和振动水平,桥梁结构的动力响应极容易受到不可预见的环境状态、非结构构件等的影响,这些变化往往被误解为结构的损伤,这使得桥梁这类复杂结构的损伤评估具有极大的挑战性.本文首先给出了结构健康监测系统的定义和基本构成,然后集中回顾和分析了如下几个方面的问题:①损伤评估的室内实验和现场测试;②损伤检测方法的发展,包括:(a)动力指纹分析和模式识别方法, (b)模型修正和系统识别方法, (c)神经网络方法;③传感器及其优化布置等,并比较和分析了各自方法的优点和不足.文中还总结了健康监测和损伤识别在桥梁工程中的应用,指出桥梁健康监测的关键问题在于损伤的自动检测和诊断,这也是困难的问题;最后展望了桥梁健康监测系统的研究和发展方向.
关键词:健康监测系统;损伤检测;状态评估;模型修正;系统识别;传感器优化布
置;神经网络方法;桥梁结构
1概述
由于不可预见的各种条件和情况下，设计和建造一个结构将永远不可能或无实践操作性，它有一个失败的概率百分之零。结构老化，环境条件和回用情况的例子，可能会影响可靠性和结构的使用寿命。有定期检查需要检测恶化，正常运行和环境造成的攻击或检查下列极端事件，如强震地震或飓风。为了量化这些系统性能的措施，当其还在使用中，需要一些手段来监测和评估土木工程的完整性。由于波音737阿罗哈意外，04月28日1988年发生，这些利息也助长了近年来在结构健康监测和非破坏性的损伤检测领域的研究。
根据Housner等编著（1997年），结构健康监控被定义为“使用奥芬原位，无损检测和结构特点，包括结构检测变化的响应可能表明损坏或退化的分析”。这个定义还确定了弱点。虽然研究人员已经尝试了NDE与健康监控一体化，重点是收集数据，而不是评价。我们需要的是一种有效的方法来从使用中的结构中收集数据和处理数据以评估关键性能的措施，如可用性，可靠性和耐用性。因此，由Housner定义（1997）应被修改，并且结构健康监测可以定义为“使用奥芬原位，无损检测及结构特点的分析，包括结构反应分析，以识别的目的，损害发生后，确定损害的部位，评估结构的安全性并评估结构上被毁坏的结果.”一般来