Modeling and parametric identification of nonlinear hysteretic behavior

This study involves finding a method for modeling nonlinear hysteretic behavior that is physically meaningful, is efficient in terms of capturing the response and retaining the memory, and involves few parameters. Offline and online parametric identification approaches are also developed for this model. A distributed element model (DEM) with time-invariant parameters capable of capturing deterioration is studied with a particular focus on the way memory is stored in the model. It is observed that keeping track of the response at a few of the past extremes of input displacement, called the Sequence of Dominant Alternating Extremes, is enough for representing the effect of history. A generalized Masing model is proposed that captures the deteriorating response of certain models including the deteriorating DEM with any distribution of element yield displacements to arbitrary loading.; Offline parametric identification of this model is studied using nonlinear optimization techniques. It is shown that the identified parameters accurately determine the physical model and are robust when using the correct distribution function and also in the over-parameterized case. In the underparameterized case, relatively accurate response agreement is obtained without accurate parameter convergence. Such a result is not robust in the validation phase with arbitrary excitation. The effects of presence of noise on optimum parameters are also investigated.; Online parametric identification of this model as well as a model with viscous damping is performed by applying an online nonlinear optimization concept to a changing objective function, defined over a shifting window of recent data. A variation of the steepest descent method is used with significant modifications. To achieve the best performance for any given problem, a set of a priori numeric tests is suggested to design the identification scheme. The suggested tests are a contribution towards having more effective identification schemes using minimal information about the model and input. The design identification scheme exhibits very good performance in adaptive identification of the true values of the parameters and is rather robust in dealing with noise. The proposed approach has applications to online identification of much wider types of nonlinear rate-dependent hysteretic behavior.