[New Paper] Bouc–Wen class models considering hysteresis mechanism of RC columns in nonlinear dynamic analysis

A journal paper, titled “Bouc–Wen class models considering hysteresis mechanism of RC columns in nonlinear dynamic analysis” was recently published in the International Journal of Non-Linear Mechanics. The paper was co-authored by Sebin Oh (a former M.S. student member of SSRG, now a Ph.D. student at the University of California, Berkeley), Dr. Taeyong Kim (a former Ph.D. student member of SSRG, now a postdoctoral researcher at the University of Toronto), and Prof. Junho Song.

The permanent link via the DOI number of the paper is https://doi.org/10.1016/j.ijnonlinmec.2022.104263. The full reference information is as follows.

Oh, S., T. Kim, and J. Song* (2023). Bouc–Wen class models considering hysteresis mechanism of RC columns in nonlinear dynamic analysis. International Journal of Non-Linear Mechanics. Vol. 148, 104263.

Aiming to improve the predictive performance of the Bouc–Wen class models for RC columns under general loading conditions, this study first identified two hysteretic characteristics that the existing Bouc–Wen class models could not describe effectively — acute deterioration and pinching relaxation. Based on a thorough examination of the hysteresis mechanisms, we introduced two parameters into the modified Bouc-Wen Baber-Noori (m-BWBN) model, proposed in Dr. Taeyong Kim’s Ph.D. dissertation in 2021. The bounds of the model parameters and the detailed identification procedure were provided for practical uses of the proposed model. A numerical algorithm for estimating the resisting force given displacement vector was also provided in the Appendix of the paper.

ABSTRACT: Bouc–Wen class models have been widely used to describe the hysteretic behaviors of structures in nonlinear dynamic and stochastic analyses. Since the existing Bouc–Wen class models cannot fully represent the hysteretic characteristics of reinforced concrete (RC) columns with structural degradation, the model fitted to the hysteresis loop from a specific input loading may not provide accurate response predictions under general loading conditions. To address this issue, we first identify two hysteresis mechanisms of RC columns, namely acute deterioration and pinching relaxation. Next, a new Bouc–Wen model is proposed to describe the hysteresis mechanisms based on the modified Bouc–Wen–Baber–Noori (m-BWBN) model. The proposed model considers the hysteresis mechanisms by two additional parameters, which are adequately bounded to facilitate the parameter identification process. Two different input loadings, i.e., a monotonically increasing one and the other with a sudden increase and decrease of the amplitude, are introduced to investigate the effects of the input loading patterns on the model calibration. The fitted models are applied to predict the responses under a quasi-static cyclic loading and a real earthquake ground motion. The effectiveness of the proposed model is demonstrated by comparison with nonlinear finite element analyses of RC columns in a database. While the m-BWBN and the proposed model show good agreements in the model calibration, only the proposed model calibrated by the input loading with sudden amplitude changes successfully predicts the responses of degrading RC columns. The proposed model will contribute to highly efficient and accurate predictions by nonlinear dynamic and stochastic analyses of a degrading structure modeled by an equivalent single-degree-of-freedom system.

A typical FE model used in this study for an RC column

Two cyclic loading histories used for model calibration

Calibration results of the m-BWBN and the proposed models using the first loading history (left) and the second loading history (right) compared to the FE analysis results

Two cyclic loading histories used for response prediction: a quasi-static cyclic load (left) and a dynamic load of the El Centro earthquake (right)

Load-displacement loops under the quasi-static cyclic predicted by the four models: m-BWBN model calibrated with the first load (top left), m-BWBN model calibrated with the second load (top right), proposed model calibrated with the first load (bottom left), and the proposed model calibrated with the second load (bottom right)

Time histories under the El Centro earthquake predicted by the four models: m-BWBN model calibrated with the first load (top left), m-BWBN model calibrated with the second load (top right), proposed model calibrated with the first load (bottom left), and the proposed model calibrated with the second load (bottom right)