为研究单掺钢纤维及混掺钢-聚乙烯醇(PVA)纤维对超高性能混凝土(UHPC)断裂性能的影响,设计11组梁试件,包括1组未掺入纤维的对照组、5组单掺钢纤维试件和5组混掺钢-PVA纤维试件,进行断裂试验。基于荷载(P)-裂缝开口位移(CMOD)曲线,获取了弯曲强度、双K断裂韧度及断裂能等。研究表明:钢纤维掺量显著改善UHPC的韧性和延性,使其由脆性破坏转变为延性破坏,并且P-CMOD曲线出现“二次强化”现象,尤其当钢纤维掺量>1.5%时,起裂荷载和峰值荷载有较为明显的提升;掺入PVA后,钢-PVA纤维UHPC的起裂韧度、失稳韧度及断裂能均有不同程度的提高;当纤维总体积掺量为2.5%时,以20%的PVA纤维取代钢纤维,实现了断裂性能与成本的综合优化。基于扩展有限元法所提出的数值模拟和实测的P-CMOD曲线与破坏形态较为吻合,可为混掺纤维增强UHPC的设计提供参考。

[Objective] This study investigates the fracture performance of ultra-high performance concrete (UHPC) reinforced with steel fibers and steel-polyvinyl alcohol (PVA) hybrid fibers through combined experimental tests and extended finite element method (XFEM) simulations. The objective is to determine an optimal hybridization strategy that enhances fracture resistance and cost efficiency, thereby providing theoretical support and practical guidance for engineering applications. [Methods] Notched beam specimens were tested using the three-point bending method. The program included one control group, five groups with varying steel fiber dosages (0.5-2.5% by volume), and five groups reinforced with hybrid steel-PVA fibers, maintaining a total fiber volume of 2.5% while adjusting PVA replacement ratios from 0 to 1.0. P-CMOD (load-crack mouth opening displacement) curves were used to evaluate flexural strength, initiation toughness, unstable toughness, and fracture energy. Parallel XFEM simulations were developed in ABAQUS, where fracture initiation was governed by maximum principal stress criterion and crack growth was modeled with energy-based softening laws. Experimental and numerical outcomes were compared to assess the predictive accuracy of XFEM. [Results] 1)The addition of fibers transformed the fracture behavior of UHPC from brittle through-crack failure to ductile non-penetrating fracture. Three distinct modes were identified: brittle single-crack, ductile single-crack, and ductile multi-crack. Steel fibers mainly provided bridging and anchorage, delaying unstable crack growth and enhancing energy dissipation, whereas PVA fibers controlled micro-crack initiation and dispersed stresses effectively, often rupturing instead of pulling out. This complementary mechanism revealed a clear division of roles, highlighting a “synergistic hybrid effect” that improved toughness and post-cracking performance. 2)Quantitatively, increasing steel fiber dosage yielded significant improvements. At 2.5% steel fibers, the initiation load, peak load, initiation toughness, unstable toughness, and fracture energy increased by 146.55%, 60.94%, 145.13%, 56.28%, and 45.58%, respectively, compared with specimens containing 1.0% steel fiber. Hybrid reinforcement further optimized performance. At a total fiber content of 2.5%, replacing 20% of steel fibers with PVA (γ=0.2) increased initiation toughness by 6%, while unstable toughness decreased by only 2%, representing the most favorable balance between toughness and economy. In contrast, higher PVA replacement ratios (γ>0.2) reduced flexural strength and fracture energy due to fiber agglomeration and uneven dispersion within the UHPC matrix. 3)Cost analysis further emphasized the advantages of hybridization. Copper-coated steel fibers cost approximately 6.5 RMB/kg, whereas PVA fibers were about twice as expensive. By replacing 20% of steel fibers with PVA at 2.5% total content, material costs were reduced by 11.6% compared with 2.5% steel fiber UHPC, without compromising fracture resistance. This finding underscored the engineering value of hybrid design, particularly for large-scale applications requiring both high durability and economic efficiency. 4)XFEM simulations closely reproduced experimental outcomes. Simulated P-CMOD curves were generally enveloped within the experimental results, and predicted crack paths matched observed failure modes. Average relative errors were 4.21% for peak load, 3.88% for unstable toughness, and 13.62% for initiation toughness, which were within acceptable limits. Moreover, XFEM captured the delayed crack penetration behavior in hybrid fiber specimens, showing how fiber synergy effectively slowed crack growth. This predictive capability demonstrated the suitability of XFEM for analyzing complex hybrid fiber systems, reducing experimental workload while offering mechanistic insights into crack evolution. [Conclusion] Steel-PVA hybridization significantly enhances UHPC fracture behavior and reduces cost, confirming the benefits of a synergistic reinforcement approach. The main conclusions are as follows: 1) Fibers convert UHPC failure from brittle through-crack rupture to ductile failure characterized by irregular, non-penetrating cracks, improving structural integrity and durability. 2) Increasing steel fiber dosage enhances toughness and ductility, with contents above 1.5% yielding substantial improvements in fracture parameters and shifting the load-bearing response from brittle to ductile. 3) A replacement ratio of γ=0.2 is optimal, strengthening crack initiation resistance and sustaining fracture toughness while reducing material costs by 11.6%. Excessive replacement (γ>0.2) negatively affects strength and fracture energy, highlighting the need for balance in hybrid design. 4) XFEM effectively simulates crack initiation, propagation,and post-cracking responses, achieving strong agreement with experiments.The method offers a reliable tool for predicting fracture performance in hybrid UHPC and can support performance-based design with reduced reliance on extensive laboratory testing.