Analysis and calibration of default steel strip pullout models used in Japan

The calibration of default pullout capacity models for smooth and ribbed steel strip reinforcement, used in reinforced soil walls in Japan, was carried out more than two decades ago and was based on a small number of physical tests available at that time. The writers have collected and organized a much larger database of more than 600 laboratory pullout box and in situ pullout tests from among the Japanese literature. The new database is a useful reference for design engineers to match project-specific soils to previous pullout tests and to check the accuracy of the current lower-bound design curves proposed in the late 70s and 80s. Today, only the ribbed-type steel reinforcement strips are used. The new data show that a three-parameter exponential function better captures the trend in pullout data for ribbed steel reinforcement than the current bi-linear models adopted from European practice. The formulations also have the advantage of being smoothly continuous with depth. Parameter values are determined for default pullout models that can be used in load and resistance factor design (LRFD) and for the current lower-bound (factor of safety) allowable stress design (ASD). The current PWRC model and a newly proposed model for ribbed steel strip reinforcement, that include the soil coefficient of uniformity (Uc) in their formulations, are shown to be no more accurate than the simpler default models without this term.