Structures are prone to damage during their service lives, caused by factors such as corrosion, fatigue, impact and overloads. Structural damage causes deviations of geometric or material properties from nominal or baseline values. Thus, techniques for nondestructive damage detection in aerospace, civil and mechanical engineering applications are essential to ensure safety, reliability and operational life. Experimental modal analysis (EMA) has become an increasingly accepted method for determining the overall health of a structure. It uses differences of measured modal properties from baseline or nominal values to identify and assess the location and severity of damage.; Damage-induced changes in natural frequencies, damping ratios and mode shapes can be detected using experimental modal analysis, which has recently enjoyed intense research for application in structural health monitoring. A main theme of the research has been to formulate a parameter, calculated using the measured changes in eigenparameters, which exhibits a significant “jump” at the damage location. This damage-sensitive parameter thus serves to localize the damage and assess its severity. It should likewise minimize the likelihood of false indications or missing damage. In this research we use finite element analysis of a damaged cantilevered beam to study a previously proposed pragmatic parameter called the Strain Energy Damage Index (SED) which is reported by several authors to exhibit the desired jump. A nominally similar parameter, which we call SED12, is likewise studied, based on the hypothesis that the modal strain energy differences between damaged and undamaged modes are may be neglected. However, even though they localize the damage, the actual value of the SEDI and SEDI2 parameters are found to be insensitive to the underlying damage severity, measured as a normalized decrease in the modulus of elasticity. In this research a new parameter, called the Modal Moment Index (MMI), is introduced based on the assumption that the ‘modal moment’ is unaffected by damage. This parameter likewise ‘jumps’ at the damage location. In addition, it is linearly proportional to the relative decrease in the modulus of elasticity. Consequently, MMI appears to be a suitable damage-sensitive parameter for experimental modal analysis. The MMI is successfully applied to beam and plate structures.; Two-stage damage detection techniques are introduced. In the first-stage damage in the structure can be located and to some extent quantified. In the second-stage the damage can be located and quantified with higher resolution.; Noise embedded in the vibration signals is an obstacle to damage detection using experimental modal analysis, especially when the damage is not severe. It alters the measured data of the undamaged and damaged structures. Compared to conventional averaging methods denoising the measured signals of the undamaged and the damaged structure, using wavelet analysis, is shown to yield improved information in detecting and assessing the severity of damage.
展开▼
