Improving Direct-Tension Testing Reliability and Material Characterization of Non-Proprietary Ultra-High-Performance Concrete

Abstract

Ultra-high-performance concrete (UHPC) is increasingly used in modern infrastructure because of its outstanding strength and durability, but accurately measuring its tensile behavior remains a significant challenge. Direct-tension testing using AASHTO T 397, particularly with a 220 kip wedge grip machine, has been documented to yield success rates as low as 36%. This paper addresses direct-tension testing challenges by evaluating non-proprietary UHPC mixtures with steel fiber dosages from 0.0% to 2.0% and silica fume contents of 9%, 12%, and 15% of total binder content, all tested at 28 days. Through a two-phase experimental program, this study covers improvements to specimen preparation, machine alignment, and grip pressure. These refinements increased the test success rate to 75% with the same testing equipment. The results show that the amount of fiber is the primary factor influencing tensile strength and post-cracking performance of UHPC, while moderate changes in silica fume content have only a minor effect. This study provides a practical and repeatable test protocol for direct-tension evaluation and demonstrates that optimizing fiber dosage is essential to achieve the desirable tensile behavior of UHPC. The proposed approach enables confident material selection and promotes wider use of UHPC.