In-Plane Shear Investigation on Masonry Walls

Analysis has been performed on approximately 200 shear walls of various masonry types, reinforcement, and grouting scenarios to determine the trends of various wall parameters to strength. The database is assembled from that belonging to researchers over the last 30 years.

Strut-and-tie methods from current concrete codes are employed and new methods proposed to calculate the in-plane shear strength of masonry walls. In addition, a simplified method based upon strut-and-tie methodology is presented to calculate the masonry contribution to the in-plane strength of these walls, with contributions from axial load and horizontal reinforcement based upon past data and efficiency of reinforcement, respectively.

The proposed strut-and-tie model was validated by the use of structural analysis and finite element software. Out of 35 fully grouted walls from the database and 13 finite element walls constructed with a macro-model approach, strength calculations produced an average [experimental shear strength/predicted shear strength] of 1.63 with a coefficient of variation [COV] of 0.26, while 23 partially grouted walls from the database along with 12 finite element analysis walls produced an average of 1.44 with COV of 0.25. Average for 108 clay and concrete walls utilizing the simplified strut-and-tie method is 1.25 with a COV of 0.24 for fully grouted walls and 1.17 with a COV of 0.31 for 87 partially grouted walls.

A shear friction equation is proposed in which the contribution to friction from applied normal force and vertical reinforcement are separated. The coefficient from applied normal force is obtained from previous unreinforced masonry tests and from inclined plane tests that examined various degrees of surface roughness. A coefficient of friction of 0.80 is proposed for this contribution. The contribution of vertical reinforcement perpendicular to the sliding surface was obtained based upon previously reported dowel tests. A value of 0.45 is proposed. The proposed model is compared to full-scale wall tests and was found to decrease the variability in prediction by almost a factor of two for concrete masonry over models that combine both the axial load and steel contribution terms and utilize only one coefficient.