DETECTION OF DISPERSED DAMAGES OF 2D STEEL FRAMES USING AN ENHANCED MODEL UPDATING METHOD

During many decades,maintenance and behavior evaluation of civil structures are drastically become of more importance problem of urban planning. Many researchers have tried to present various efficient and effective methods to evaluate more reliable and much precisely structural damages especially in the beginning of the damage formation. Among different proposed methods, model updating method is truly known as an accurate and reliable method for detecting location and severity of structural damages. Model updating method is a properly substitution for the traditional methods. In this paper, a methodology for model updating of complex andor irregular two degrees of freedom moment-resisting steel structures and braced steel structures is presented. In order to increase accuracy and performance of this methodology, model updating is designed to carry out in two stages. Damages may be much dispersed and their severities may be so small or much intense. This methodology is powered by an iterative sensitivity analysis using nonlinear constraint optimization, which is worked based on the Newton trust region algorithm. In order to shown capability of the proposed methodology, different damaged steel frames are examined. Model updating parameters are defined as stiffness reduction coefficients of members or a member is divided into some smaller parts. Updating is obtained during reducing differences between measured and analytical modal parameters of the structure. These parameters may be the values of frequencies and modal shapes of the real damaged structure and its related analytical model. Extensive results obtained from analysis of different models with various dispersed damages which were so small or very influenced in severity, shown that the proposed methodology has able to detect three levels of the structural health monitoring stages, accurately and more precisely. These three levels are as follows: (1) detection that the considered structure is healthy or damaged, (2) detection of damage locations (along whole or part of the members of the considered structure), (3) detection of the severity of the damaged member, which may be so small or so intense.