Out-of-plane behaviour of concrete block walls with unbonded reinforcement

The use of grout in conventional reinforced masonry construction increases the cost and time of construction but, when used in combination with reinforcing steel, allows walls subject to out-of-plane loads an enhanced ability to span between lateral support levels. Reinforced concrete block walls constructed in this manner can typically span at least two stories in constrast to the limited single storey capacity of unreinforced walls. However, the use of grout as needed for the construction of these walls increases their self-weight, and requires an additional trade on-site. A novel, potentially cost-efficient, approach to achieve reasonable load-carrying capacity in masonry walls was therefore investigated that involves the use of minimally stressed reinforcement anchored at the top and bottom of the wall. This allows for a grout-free structural system that relies upon arching to resist the flexural effects resulting from out-of-plane loads and so make more effective use of the compressive capacity of the masonry assembly. An experimental program was therefore conducted at the University of Saskatchewan to investigate the performance of concrete masonry block walls reinforced with non-prestressed, unbonded reinforcement. This study included a total of 21 walls that were built to identify potential alternatives to unreinforced and conventionally grouted and reinforced walls. The strength and serviceability of these walls was evaluated. All walls in this program were two and a half blocks wide and 14 courses tall and were built in running bond using standard 200 mm concrete blocks. Six replicates of both unreinforced and partially grouted, conventionally reinforced walls served as control specimens. An analysis of the data obtained during testing revealed that the walls with unbonded reinforcement were inherently stable with maximum loads approaching those of partially grouted, conventionally reinforced walls. Furthermore, an analytical approach is presented herein that is based on the assumption that the walls with unbonded reinforcement could be modeled using a three hinged mechanism. The analytical model was found to match with the experimentally obtained load versus mid-height deflection data reasonably well throughout the post-cracking range