Recent studies have investigated the seismic response of controlled rocking masonry walls (CRMWs) that rely on gravity loads for self-centering and on supplemental energy dissipation (ED) devices to control the response (ED-CRMWs). However, such studies reported that limitations still exist due to ED devices being installed inside the wall, making repairs difficult or impossible following the yielding or fracturing of such devices. For these reasons, the current study develops a new system, namely, controlled rocking masonry walls with energy dissipation accessible in a steel base (EASt-CRMWs). In this system, walls are constructed upon a steel rocking base to allow for the installation of ED devices within the footprint of the wall. In addition, these ED devices are in the form of externally mounted cantilevered steel flexural yielding arms that can be easily replaced following a seismic event. To investigate the seismic response of the new system, the study reports the experimental results of six EASt-CRMWs tested under displacement-controlled quasistatic cyclic fully reversed loading. Various design parameters are investigated herein, including axial load, aspect ratio, vertical reinforcement, confinement technique, and size of the flexural arms. The experimental results are presented in terms of the force-displacement responses, residual drift ratios, and damage patterns, including the use of digital image correlation techniques. Using such results, iterative and simplified procedures are developed and validated to predict the monotonic force-displacement responses of the test walls. The experimental results show that the use of a steel rocking base considerably enhanced the seismic response of the walls by protecting their rocking toes at large drift ratios. Specifically, extremely low damage was observed at the end of the tests, where five walls reached drift ratios up to 4.0% with no strength degradation and with residual drifts of less than 0.1%. The vertical reinforcement and confinement properties of the rocking toes had almost no influence on the cyclic response of the test walls; however, the axial load, aspect ratio, and amount of ED altered their responses. The two prediction procedures were also able to simulate the monotonic responses of the walls at all displacement ranges. With these results, the current study demonstrates that the proposed EASt-CRMWs are expected to be a resilient system within the masonry construction practice by achieving low damage and rapid recovery following seismic events.
