Size effect on quasi-brittle fracture of concrete-like materials is commonly investigated by testing geometrically-similar specimens with artificial notches. It can also be studied by testing plain specimens without any artificial notches to show the statistical influence of pre-existing defects/cracks using…
Size effect on quasi-brittle fracture of concrete-like materials is commonly investigated by testing geometrically-similar specimens with artificial notches. It can also be studied by testing plain specimens without any artificial notches to show the statistical influence of pre-existing defects/cracks using Weibull strength distribution. Those un-notched specimens can be further studied non-statistically, using the Fictitious Crack Model (FCM), which emphasizes the influence of crack-bridging within the fracture process zone (FPZ) on quasi-brittle fracture of concrete by adopting a single bridging-stress and crack-opening relation. This study discusses the merits and limitations of common size effect models, including Weibull strength and FCM, in dealing with quasi-brittle fracture of concrete specimens with and without notches, and shows it is necessary to adopt a tri-linear local fracture energy distribution, which considers the boundary influence from both the front and back specimen boundaries. The front boundary influence is particularly important to quasi-brittle fracture of un-notched specimens, as the limited stable crack growth before unstable fracture is well within the front boundary zone. Therefore, size effect models, based on constant fracture energy or constant local fracture energy distribution (suitable only to a long crack away from boundaries), cannot be used to reliably predict quasi-brittle fracture of un-notched concrete specimens.