Tensile behavior of strain-hardening cementitious composites in elastic, ultimate, and rupture stages, experimental investigation on the accuracy of testing methods

In the absence of a standard methodology for evaluating the tensile behavior of ductile construction materials, this research investigates the efficacy of existing methods in capturing the authentic behavior of a strain-hardening cementitious composite. Four widely used testing methodologies (unidirectional tensile test, three-point bending, four-point bending, and Brazilian disc) were scrutinized, each varying in geometry. Additionally, a newly developed testing method was employed. The results reveal that the Brazilian disc method fails to accurately represent the material's true behavior. Taking the direct tension (DT) test as the control method, this exhibits a straight mid-section crack with an elastic strength of 3.47 MPa, a strain-hardening peak of 2.68 MPa, and a failure strain of 0.101. In contrast, the three-point bending (3 PB) and four-point bending (4 PB) tests underestimated strain and fiber bridging capacity, with the 3 PB method exhibiting greater discrepancies (for instance, a -96% strain deviation at the elastic limit) due to shear stresses, whereas the 4 PB method displayed improved performance (specifically, a -6% stress deviation at peak). Moreover, the SBD test, particularly the thick-SBD variant, emerged as a promising alternative, closely aligning with the DT method within a ± 20% error margin for both stress and strain metrics. The geometry of the specimens significantly influenced the results, with smaller size-to-radius (S/R) ratios in SBD tests confining tensile zones, while higher thickness-to-diameter (t/D) ratios increased the risk of shear failure. These findings emphasize the necessity for optimized testing methodologies, such as the SBD, to effectively assess ductile cementitious composites.