Concrete breakout strength prediction of anchors in Steel Fiber Reinforced Concrete based on material test datasets

Abstract The use of steel fiber‐reinforced concrete (SFRC) has seen a significant rise in the construction industry due to advancements in material science, improved quality, and cost efficiency. Similarly, the importance of fastening systems in construction has grown, necessitating robust research on the performance of fasteners in SFRC to ensure safe design practices. This paper addresses the need for a unified methodology to predict the performance of anchors in SFRC amidst the diversity of steel fiber and anchoring products in conjunction to the variety of concrete material characteristics. Hereby, material data are implemented, which involves a unification of 4‐point and 3‐point bending test results, supported by machine learning. This study presents both a predictive model and a design formula for anchors embedded in SFRC, leveraging machine learning to enhance the experimental dataset by generating artificial 3‐point bending results. A comprehensive dataset comprising 168 anchor tensile experiments and 606 associated material tests is analyzed. The findings demonstrate that a generalized methodology based on material testing can achieve high accuracy in the design of anchors in SFRC.