Reliability of Extracted Modal Parameters by Random Decrement of Acceleration Signatures

Abstract

Modal parameters provide important information on dynamic properties of structures. In operating condition, since it is difficult to measure input loadings, methods should be applied where don’t require measuring inputs. Such methods which identify modal parameters of structures by measuring their responses are called Operational- or Output Only- Modal Analysis (OMA) techniques. There are many time and frequency domain operational modal analysis techniques. Generally a form of impulse or free vibration response is required to use most of these techniques. However, in practice structures are usually subjected to some immeasurable or unknown random inputs. In these situations Random Decrement (RD) transformation can reduce these responses to equivalent free decay or correlation functions. Therefore, RD technique coupled with those methods, which require a form of impulse or free vibration response offer a valuable tool for identifying the dynamic characteristics of structures from operational or ambient responses. Unfortunately, in the literature there are some constrains on using random decrement signatures. For example by complicated mathematical relations it is shown that random decrement technique is applicable only if the inputs are uncorrelated zero mean Gaussian white noises. In addition, it is proved that only random decrement of displacement and velocity is equivalent to the corresponding free decay responses or correlation functions the random decrement of acceleration response is never equivalent to the corresponding free decay responses or correlation functions. However, there are many papers which have used random decrement of acceleration responses and extracted modal parameters accurately! In this paper it is tried to show simply and clearly whether it is possible to obtain modal parameters from random decrement acceleration signatures or not. To do that, a numerical simulation of a discrete dynamic system with viscous damping is carried out and the results of numerical methods are compared with those come from analytical solution. Numerical simulation is used since it is completely controllable. Finally, it is tried to identify power and the Applicability cases of random decrement method.