UC San Diego

Reinforced concrete shear walls are commonly used to provide lateral strength and stiffness to concrete buildings in seismic regions. Typically installed in the wall face, mechanical anchors are responsible for connecting various nonstructural systems to the main structure. During an earthquake, anchors in reinforced concrete structural elements need to retain their strength and stiffness, despite the inevitable presence of cracks and damage in the concrete, developed as a consequence of the lateral cyclic loading. Anticipating damage to the concrete, which will naturally influence anchor response, current guidelines to qualify anchors for seismic applications require adequate performance in cracked concrete to assure minimal anchor load loss. However, these guidelines are based on anchor performance in pure flexural cracks, as this is the typical damage condition occurring in reinforced concrete frame elements, which has been studied for decades. The response of anchors to a mix of flexure and shear cracks, i.e., the complex situation realized in shear-flexure structural components such as shear walls, however, has largely not been studied.

To address the paucity of data regarding anchor behavior in cracked concrete, the behavior of anchors installed horizontally in three full-scale reinforced concrete shear walls with different aspect ratios (wall height/length) is studied in this dissertation. Notably, two types of post-installed anchors were investigated in these tests, namely: i) expansion anchors and ii) bonded anchors. One slender and two identical low-aspect ratio walls were designed according to current U.S. design codes. Simulated seismic loading was imposed at the top of the wall using an equivalent cyclic displacement history, while uniform compression was applied on the slender and one of the two identical low-aspect ratio shear walls. One of the low aspect ratio walls was tested without axial compression to investigate its effect on the anchor response. Anchors were continuously loaded to their design tension while the walls were cycled. The slender full-scale wall failed in a predominantly flexural mode, precipitated by buckling and fracture of the boundary reinforcement. The two identical full-scale low-aspect ratio walls failed in a mixed flexure-shear response, with severe web concrete crushing and buckling and rupture of the boundary reinforcement. Anchor axial load and displacement data, continually measured during the wall cyclic tests, confirmed the sensitivity of the performance of anchors amidst the presence of a variety of cracked concrete conditions, especially in walls prone to develop large shear stress and shear induced damage when subjected to lateral cyclic loads. Following the wall cyclic tests, tension failure tests performed on the anchors indicated that their residual tension load capacity was significantly compromised by concrete damage. Such damage was concentrated in specific wall regions, such as the boundary elements and the plastic hinge region in slender walls, or along the diagonal struts, the boundary elements and near the base of low-aspect ratio walls. Of the two types of anchors tested, expansion anchors observed the most significant load loss (and consequentially axial displacement) in the presence of both the wall cyclic loading and the residual tests on the anchors themselves.

Following the experimental program, a multiple vertical line finite element model was used to predict the response of each of the tested full-scale shear walls. Numerical analyses cross-comparison with test results demonstrated a high level of accuracy of the selected modeling approach. As such, an expanded parametric study was conducted to understand the extent of severe concrete strains on the crack distribution and width, using a smeared crack approach. Wall models designed for the parametric study were intended to explore different geometry, reinforcement and axial compression to study the damage distribution within the wall elevation. Crack pattern distribution plots developed using the parametric study results were used to identify regions where anchors would be vulnerable to load loss upon achievement of service, design and severe seismic damage. Ultimately, the findings from this dissertation shed light on the vulnerability of anchors placed in reinforced concrete shear walls, where damage in the form of mixed mode cracking and spalling can be expected. Future design guidelines would benefit from precluding crack sensitive anchors in the most highly damaged regions of these essential lateral force resisting components of the structural system.