۱- Introduction

۲- Physical response of connections

۳- Hysteretic model formulation

۴- Estimation of model parameters

۵- Unconstrained calibration

۶- Evaluation of the ECB hysteretic model against experimental data

۷- Summary, recommendations, and limitations

References

Abstract

Embedded Column Base (ECB) connections are commonly used in mid- and high-rise steel moment frames, to connect the steel column to the concrete footing. Although recent research has shown these connections to be highly ductile, they are typically designed to be stronger than the adjoining column, resulting in significant cost. To enable assessment of strong-column-weak-base systems that leverage the inherent ductility of these connections, an approach is presented to simulate their hysteretic and dissipative response. The proposed approach simulates ECB connections as an arrangement of two springs in parallel, to reflect moment contributions due to horizontal and vertical bearing stresses. This is informed by recent work that provides physical insight into the internal force transfer within these connections. The springs’ response is defined by the pinched Ibarra-Medina Krawinkler (IMK) hysteretic model, which is able to capture both in-cycle and cyclic degradation in strength and stiffness. The model is shown to reproduce the response of ECB connections with reasonable accuracy. Guidelines to calibrate model parameters are presented; these include physics-based estimation of selected parameters such as strength and stiffness, accompanied by empirical calibration of ancillary parameters associated with cyclic deterioration. Limitations are discussed.

Introduction

Column base connections in steel moment frames may be classified as of the exposed or embedded type. Exposed base plate connections (such as the one shown in Fig. 1a) are common in low-rise (1–۳ story) moment frame buildings, where the base moment, shear, and axial force demands are relatively modest. These are less preferable for midor high-rise moment frames, since the higher moment demands necessitate a large number of deeply embedded anchor rods and/or thick base plates. In these cases, Embedded Column Base (ECB) connections, such as the one shown in Fig. 1b are more preferable. These connections resist base moments and forces through a combination of bearing stresses on the column flanges and the embedded base plate. Besides, exposed base plate connections may be shallowly embedded under a slab-on-grade cast on top of the base plate. This shallow embedment (typically less than 300 mm) increases the strength and stiffness of the connection. Exposed base plate connections are well-researched, with validated models for strength (Drake, and Elkin [1]), stiffness (Kanvinde et al. [2], Trautner et al. [3]), component hysteretic response (Torres-Rodas et al. [4]), and methods for design (Fisher and Kloiber [5], Gomez et al. [6]). In contrast, ECB connections (constructed as per US practice) have attracted research attention only recently; this work includes some of the first experiments on deeply embedded column bases (Grilli et al. [7]), and shallowly embedded column bases (Barnwell [8]). These experiments have led to validated strength models and design methods (Grilli and Kanvinde [9] for deeply embedded, and Barnwell [8] for shallowly embedded), as well as stiffness characterization approaches (Torres-Rodas et al. [10] for deeply embedded, and Tyron [11] for shallowly embedded). A secondary finding of these studies is that ECB connections are ductile (rotation capacity in the range of 0.03–۰٫۰۸ rad) for the specimens tested by Grilli and Kanvinde [9], and Barnwell [8], even when not explicitly detailed for ductility. In seismic regions, where ECB connections are generally designed to remain elastic (AISC [12]), making this finding to be important. More specifically, ECB connections (and more generally, column base connections in seismic moment frames) are designed to resist a moment equal to 1.1R My p of the connected column (i.e., a “strong-base-weak-column design”).