Condition monitoring of structures using limited noisy data modal slope and curvature of mode shapes

摘要

Purpose - Condition monitoring (CM) has become significantly important, particularly in the context of ensuring safety, reliability and future usefulness of civil infrastructural systems. Most of the age old structures require immediate attention. Nondestructive tests and/or load tests along with routine maintenance inspections are common practice. However, most of the NDT techniques are location-dependent and are conducted in a piecewise manner. The paper aims to discuss these issues. Design/methodology/approach - Numerical methods incorporating inverse techniques are a global approach to identify structural parameters using dynamic responses. However, measurement at all degrees of freedom does not seem to be feasible, due to practical constraints. Parameter identification of structures based on limited dynamic responses like modal slope and curvature mode shapes at the element level in a finite element platform is proposed in the present paper. The structural property for each element is derived adopting a two-phase analysis process, consisting modal extraction and structural parameter identification. It is important to study the accuracy of the predicted parameters with the number of measured modes. The structural property is identified using measured responses at those selected MDOF. Findings - The proposed method is demonstrated in detail with a numerical example. The method seems to be an attractive proposition as the results obtained are very accurate even with noise-contaminated data. Research limitations/implications - However, for practical problems, the experimental validation is significantly important prior to its application in real-life problems. Practical implications - The developed model seems to be feasible for practical applications after experimental validation, as it is able to identify the structural parameters from limited noisy dynamic responses in frequency domains measured for few modes. Social implications - Structural CM is the need of the hour, particularly for infrastructural systems including buildings and bridges, etc. System identification with a global dynamic response at few measurement locations may address the issue of health assessment of structures, which will have great social implications with respect to safety. Originality/value - The proposed numerical model is originally developed by the authors with judicial modifications and combination of earlier research contributions to achieve greater accuracy. Limited measurement and the effect of random noise with numerical example are considered for the successful validation.