Open AccessThreshold value and applicable range of nonlinear behavior detection method using second derivative of acceleration
Author(s) -
Iyama Jun,
Wakui Masaki
Publication year - 2019
Publication title -
japan architectural review
Language(s) - English
Resource type - Journals
ISSN - 2475-8876
DOI - 10.1002/2475-8876.12077
Subject(s) - acceleration , nonlinear system , vibration , range (aeronautics) , control theory (sociology) , sensitivity (control systems) , threshold limit value , noise (video) , stiffness , derivative (finance) , a priori and a posteriori , mathematics , computer science , algorithm , physics , acoustics , engineering , structural engineering , artificial intelligence , electronic engineering , financial economics , classical mechanics , quantum mechanics , economics , philosophy , image (mathematics) , aerospace engineering , environmental health , control (management) , epistemology , medicine
Abstract A previously proposed nonlinearity detection method using the second derivative “snap” of the recorded absolute acceleration requires the determination of a threshold value based on the yield strength of the target vibration system, which may not be known. Therefore, this study aims to extend this detection method by determining the mathematical relation between the snap and stiffness change and velocity of the vibration system; the results indicate that the threshold value required to detect nonlinearities can be explicitly expressed by mathematical equations. Although the accuracy of this detection method is affected by the intensity of noise and the time intervals of the acceleration records, the introduced mathematical model can both explain these effects and allow the user to decide a priori whether this method can be used to detect nonlinearities. Furthermore, the proposed mathematical model for nonlinearity detection was verified by dynamic response analysis with varying natural periods, showing that the detectable range estimated by the model agreed with the range where the accuracy of nonlinearity detection by snap increases.