Estimation of backbone model parameters for simulation of exposed column base plates

An approach is presented for the estimation of the parameters required to simulate the nonlinear monotonic (i.e., backbone) rotational response of Exposed-Column-Base-Plate (ECBP) connections subjected to moment and axial compression. A trilinear backbone curve is selected to represent the rotationa...

Full description

Autores:
Villar-Salinas, Sergio
Kanvinde, Amit
Lopez-Almansa, Francisco
Tipo de recurso:
Fecha de publicación:
2024
Institución:
Universidad Tecnológica de Bolívar
Repositorio:
Repositorio Institucional UTB
Idioma:
eng
OAI Identifier:
oai:repositorio.utb.edu.co:20.500.12585/12740
Acceso en línea:
https://hdl.handle.net/20.500.12585/12740
https://doi.org/10.1016/j.jcsr.2024.109034
Palabra clave:
Exposed-column-baseplates
Moment-rotation curves
Axial compression ratio
Regression models
Performance-based assessment
LEMB
Rights
openAccess
License
http://creativecommons.org/licenses/by-nc-nd/4.0/
Description
Summary:An approach is presented for the estimation of the parameters required to simulate the nonlinear monotonic (i.e., backbone) rotational response of Exposed-Column-Base-Plate (ECBP) connections subjected to moment and axial compression. A trilinear backbone curve is selected to represent the rotational response, defined by three deformation and two strength parameters; these properly represent the stiffness, strength, and ductility of the connections. This approach is accompanied by a tool to facilitate convenient estimation of the parameters. The approach is based on a combination of behavioral insights and physics-based models (for some parameters) as well as regression for other parameters, which are estimated from a dataset of eighty-four experiments on ECBP connections conducted over the last forty years in the United States, Europe, and Asia. Predictive equations are provided to estimate the various parameters defining the nonlinear response, and their efficacy is examined by comparing them with the test data; in addition, well-established techniques are implemented to avoid collinearity and the overfitting of regression models. The results show that the models presented in this work provide robust and accurate predictions for in-sample and out-of-sample data. Limitations are outlined.