An Assessment of Time and Frequency Domain Identification Methods Applied to Non-contact Experimental Data

The use of time or frequency domain in structural identification has advantages and disadvantages in both cases. For example, in structures with frequency-dependent mechanic properties, as viscoelastic materials, frequency domain is clearly more interesting than time domain. On the other hand, if non-linear systems are considered, time domain may present some advantages. Regarding these last considerations, it is clearly observed that both domains are very important in structural analysis. Besides that, the use of image processing based structural monitoring system has increased over the last years. Professional integrated monitoring systems are been developed and, in a short time period, it is expected an augmentation of the application of this non-contact instrumentation method. In the present work, an assessment of time and frequency domain identification methods applied image processing results is carried out. To this end, time domain methods like Ibrahim Method, Random Decrement Method, Exponential Decrement Method and frequency domain methods like Peak Picking Method, Frequency Decomposition Methods, are applied to experimental results obtained from laboratory tests performed with image processing techniques. As a conclusion of this work, some recommendations for the use of each tested method applied for the class of experimental data (image processing based data) are expressed, regarding theirs efficiencies and accuracies.