From constitutive models to shear capacity prediction: A comparative review of international design provisions for ultra high-performance fiber-reinforced concrete beams

Abstract

Ultra-High-Performance Fiber-Reinforced Concrete (UHPFRC) has gained increasing global adoption due to its superior mechanical performance, durability, and strain-hardening capacity. However, international design provisions remain inconsistent in their treatment of constitutive behavior and shear resistance, leading to significant differences in predicted beam capacity. This study conducts a comprehensive review and large-scale benchmarking of eight major UHPC/UHPFRC design provisions including FHWA-HRT-23–077 (2023, USA), ePCI (2021, USA), AFGC-SETRA (2013, France), CSA A23.1 Annex U (2019, Canada), JSCE (2008, Japan), KICT (2012, Korea), as well as French standards NF P 18–470 (2016) and NF P 18–710 (2017) to evaluate their constitutive modeling assumptions and predictive reliability. The review synthesizes their underlying compressive and tensile constitutive assumptions, as well as their flexural and shear design philosophies, while the quantitative benchmarking is focused specifically on shear capacity prediction in UHPFRC beams. For this purpose, a database of 492 shear-dominated specimens was established, covering prestressed and non-prestressed members with varied geometries, reinforcement ratios, fiber contents, material strengths, and shear span-to-depth ratios. Because FHWA-HRT-23–077 shear design relies on crack-localization parameters seldom reported in the literature, a supplementary predictive model for ft,loc was developed to enable consistent implementation across the database. Statistical comparisons indicate that FHWA-HRT-23–077 provides the most consistent agreement with experimental results, mainly due to its localization-based tensile parameter, strain-compatible shear formulation, prestressing contribution, and diagonal-compression upper-bound limit. Other provisions showed greater conservatism or scatter due to differences in tensile modeling, fiber-contribution assumptions, reduction factors, calibration background, and required material inputs. Overall, the results emphasize the need for harmonized UHPC/UHPFRC shear design provisions supported by standardized material characterization and failure-mode reporting. © 2026 Published by Elsevier Ltd.