A Simplified Approach to Joint Shear Behavior Prediction of RC Beam-Column Connections

Abstract

An extensive experimental database of reinforced concrete (RC) beam-column connections subjected to cyclic lateral loading has been constructed. All cases within the database experienced joint shear failure, either in conjunction with or without yielding of longitudinal beam reinforcement, representing damage within a joint panel that was the main contributor to total lateral deformation. (Cases having damage within a joint panel caused by other premature failure modes (e.g., anchorage failure) are not included in the database.) Using the experimental database, envelope curves of joint shear stress vs. strain behavior were developed by connecting key points such as cracking, yielding, and peak loading. Joint shear stress and strain models at peak response have been developed by a Bayesian parameter estimation method based on the experimental database. At other key points, important influence parameters are also identified by constructing joint shear stress and strain models in conjunction with the Bayesian parameter estimation method. Then, a complete RC joint shear stress vs. strain model (including post-peak behavior) is suggested using simplified joint shear stress and strain models at peak response; effects of key parameters on the suggested behavior models are evaluated. Finally, the ASCE/SEI 41 joint shear behavior model has been examined using the constructed database—specific joint shear strength factors and plastic joint shear deformation values are recommended for use when following that approach.

Statistical models for shear strength of RC beam‐column joints using machine‐learning techniques

“…These parameters include material properties, in-plane and out-of-plane geometry, joint eccentricity, column axial load, and design parameters such as design joint stress demand and bond resistance. Finally, the comparison of the proposed joint shear strength model with that of simple multiple linear regression (MLR) as well as existing estimation methods proposed by Kim and LaFave [28,29], Tsonos [42], and Park and Mosalam [20] demonstrates the superior performance of the proposed model.…”

“…Beam-column joints can be classified as non-ductile and ductile joints in three different ways based on the existence of joint transverse reinforcement [28,29], the spacing of joint transverse reinforcement [21], and the deformation ductility of subassemblages [42]. As mentioned in Section 1, Kim and LaFave [28,29] developed a unified, simplified strength model for different types of joints, which includes three variables related to joint transverse reinforcement. For non-ductile joints (no transverse reinforcement), these variables should not be zero because the formulation was developed in a log-transformed space; thus, they used fictitious values for these variables.…”

“…The joint shear strength in the seismic codes [21,[45][46][47] is a function of the square root of f c . Additionally, through statistical observations, Kim and LaFave [28,29] showed that f c is the most influential parameter that affects the joint shear strength.…”

“…The joint configuration can be changed in accordance with out-of-plane geometry, namely the existence of transverse beams (TB). When a joint panel initiates expansion toward the out-of-plane directions (if the joint is subjected to higher joint shear stress demand than the joint capacity), passive confinement can be activated by longitudinal reinforcement of transverse beams preventing the dilatation of the joint panel [28,29]. Therefore, the out-of-plane geometry can affect joint shear strength due to this passive action.…”

“…The proposed equation reflects the most significant parameters that influence the joint behavior, and accounts for the compression-softening phenomenon associated with cracked RC. Kim and LaFave [28,29] developed an RC joint shear strength model for a set of subassemblages with proper confinement (no out-of-plane members, and no joint eccentricity) via a Bayesian parameter estimation method, and then proposed unified joint strength model for various types of subassemblages through a stepwise removal process. The entire database consists of 212 interior, 101 exterior, and 18 knee joints, all of which exhibited joint shear failure irrespective of beam yielding, and 15 of which have joints without transverse reinforcement.…”