Constitutive model for rock fractures: Revisiting Barton's empirical model

Barton's original model is modified in order to address the following limitations: the peak shear displacement is independent of normal stress and is zero for sawed fractures; for post-peak shear strength, Barton suggested a zero mobilized Joint Roughness Coefficient (JRC) after 100 times the peak shear displacement; Barton assumed zero dilation displacement at up to one-third of peak shear displacement. In this study, a database of results of direct shear tests available in the literature was collected and analyzed. An empirical equation is introduced to predict peak shear displacement, which considers the effect of normal stress and works for all types of rock discontinuities, even sawed fractures. The post-peak mobilized JRC can be obtained using a power-base equation, instead of employing Barton's table. This new empirical equation for post-peak mobilized JRC works for all ranges of displacements. The modified model can predict negative compressive dilatancy at small shear displacements.