Shear strength of reinforced and prestressed concrete beams using shear friction

Abstract

A new approach for the prediction of shear failure of reinforced and prestressed concrete beams, based on shear friction, has been under development at the University of Calgary for over a decade. In this thesis, the shear friction model has been investigated and refined in order to account for the existence of axial forces, the effect of the longitudinal reinforcement ratio, the effect of the beam size, and the effect of the concrete compressive strength. The proposed shear friction model is applicable to both disturbed and non-disturbed regions, and to beams with or without stirrups. For calibration purposes, and to compare with available CSA A23.3 shear design methods, a large database of 559 prestressed and reinforced concrete beams with and without stirrups was collected. The reinforced beams were without and with axial tensile or compressive loading.

The experimental part of the study consisted of two test series. Five large beams without stirrups with low amount of longitudinal reinforcement were tested. In addition, six tests were performed on three beams with stirrups. The experiments were designed to investigate different elements of the shear friction model.

The results show the predicted shear strengths for beams without stirrups, using the new shear friction model, to be much superior to those of the CSA A23.3 simplified method, the CSA A23.3-94 general method, and the CSA A23.3-draft 52003. This is illustrated by the fact that these shear friction predictions covered a wider range of aid ratios (from 1.0 to 7.0), yet were characterized by the smallest coefficient of variation. In addition, the shear predictions using the shear friction model showed no bias over wide ranges of the most influential parameters: concrete compressive strength (f~), member height (h), and longitudinal reinforcement ratio (Ptw)·

For beams with stirrups, however, the shear predictions showed a high coefficient of variation, although remained consistent over the whole ranges of the most influential parameters. In its proposed form, the shear friction model is not simple enough to be accepted as a design code method. Recommendations for future research were made, and more work is still needed to simplify the model without compromising its precision.