The quality of test data is an important consideration in conducting field experiments on civil infrastructure. In addition to possible errors due to the experimental setup, the uncertainties due to incomplete knowledge of a structure's behavior and its interactions with the natural environment greatly affect the reliability of the system identification results. This paper discusses the uncertainties related to ambient vibration testing of a long-span steel arch bridge and possible ways to mitigate them. The consistency of the identified parameters is examined through statistical analyses.
Condition evaluation of large constructed structures has been widely researched over the last decade. In numerous studies dynamic testing has been used as the primary experimentation tool for measuring the dynamic characteristics and extracting various proposed indicators of structural condition. Despite these efforts, dynamic testing of constructed systems has not yet evolved to a point that it can be standardized as a tool for condition evaluation. Writers believe that two major sources of uncertainty in dynamic test based condition evaluation are the reasons for the gap between concepts and meaningful real-life applications. Most dynamic test methods are built on principles of observability, linearity and stationarity. The first major source of uncertainty is due to constructed systems and their loading environments’ inherent complexity leading to limitations in the application of the above tenets. The second category of uncertainty is related to the experiment, i.e. sensing, data acquisition, processing and analysis for different dynamic test methods. Ambient vibration (i.e. operational modal analysis) and impact testing are two different tools aiming to identify same parameters. Comparative evaluation of different test methods at the presence of different levels of uncertainty will enable us to assess the reliability of dynamic testing tools for condition evaluation. Writers designed a set of experiments on a laboratory physical model to investigate the effects of these two groups of uncertainties on modal parameter identification. Results will be discussed along with mitigation measures of uncertainties in dynamic testing of constructed systems.
Health monitoring of infrastructure systems for their proactive management to make the best use of limited resources for their optimum life-cycle performance, protection and preservation is a promising paradigm. Structural health monitoring (SHM) technologies have sufficiently evolved to accomplish real-time post-hazard evaluation of highway bridges. Solution of the pressing and critical problem of post-hazard bridge condition evaluation is possible by problem-focused, coordinated cross-disciplinary research. A newly initiated research program, which has been formulated for creating knowledge to construct an intelligent bridge SHM system, is presented in this paper. An overview of the state-of-the-art research in SHM is outlined.
In this paper a structural identification (St-Id) case study on a laboratory physical model is presented. The main objective is to understand the reliability of ambient monitoring as the principal global experimentation tool for St-Id and to address the issues related to the projection of the laboratory study to a bridge structure. Linear deterministic FE modeling, controlled static load tests, impact tests and ambient vibration tests are utilized as St-Id tools to supplement each other in the study. The results showed that even under controlled laboratory conditions, uncertainties in the St-Id process cannot be completely avoided and govern the reliability of an St-Id study. Issues regarding the successful application of ambient vibration testing on a real-life bridge structure based on laboratory model results are addressed.