Wet joints are the cast-in-place connections that tie precast concrete segments together in prefabricated bridge construction, an approach that has become an inevitable trend in modern civil engineering because it shortens construction periods, reduces overall costs, and gives engineers tighter quality control than traditional cast-in-place methods. Because every load crossing from one precast segment to the next passes through these connections, the material properties and geometric configuration of a wet joint directly dictate the structural performance, safety, and service life of the entire bridge system. Earlier work has underlined how much is at stake: one study found that a 200-millimeter-wide ultra-high performance concrete joint achieved 8.8 percent higher tensile capacity than a 500-millimeter-wide normal concrete joint, while another reported that a Type II ultra-high performance concrete joint raised load capacity by 35.7 percent under shear and 143.2 percent under flexure compared with a Type I joint. A dovetail joint paired with steel wire mesh interface treatment has been shown to lift cracking load by 42.9 percent.

Traditionally, these joints have been filled with normal concrete, but that choice carries penalties. Normal concrete wet joints typically demand large structural volumes, require complex reinforcement layouts, and involve time-consuming, labor-intensive construction that can delay project timelines. Ultra-high performance concrete offers an alternative with exceptional durability, superior compressive and bending strength, high stiffness, and outstanding crack resistance, and its excellent anchorage strength to steel reinforcement allows shorter development lengths that simplify bar-lapping and speed construction. The catch is cost and carbon: conventional ultra-high performance concrete is expensive and cement-hungry, running counter to the construction industry’s push toward low-carbon development. Researchers have therefore explored industrial solid wastes such as fly ash, slag, and coal gangue as partial cement replacements, and more recently have developed coarse aggregate ultra-high performance concrete, in which added coarse aggregates maintain or improve compressive strength and elastic modulus while substantially cutting binder volume, lowering cost, and inhibiting early-stage hydration shrinkage cracking.

The new study targeted a gap in that literature. Most research on metakaolin in cement-based materials has focused on conventional or non-high-strength concrete, and work on metakaolin-modified ultra-high performance concrete has concentrated on hydration chemistry and standard rectangular specimens; explicit investigation of metakaolin within coarse aggregate ultra-high performance concrete wet joints has remained highly limited. To probe how material formulation and joint geometry interact under flexural loading, the team cast five large-scale test beams, each 1400 by 180 by 120 millimeters with a calculated span of 1200 millimeters, and loaded them in four-point bending so that a 400-millimeter pure-bending region at midspan contained the joint, the interfaces, and some normal concrete. Both ends of every beam were normal concrete. Four beams carried rectangular joints filled respectively with plain ultra-high performance concrete, metakaolin-modified ultra-high performance concrete with 10 percent metakaolin, coarse aggregate ultra-high performance concrete with 16 to 20 millimeter basalt gravel, and a combination of both. The fifth beam served as the geometric control, pairing a rhomboidal joint with the combined metakaolin and basalt material. All joints were 150 millimeters wide with a 110-millimeter steel bar lap length.

Material characterization at 28 days revealed the trade-offs the beam tests would later expose. Plain ultra-high performance concrete reached a cube compressive strength of 125.56 megapascals, an axial compressive strength of 84.62 megapascals, an elastic modulus of 42.20 gigapascals, a flexural strength of 29.67 megapascals, and a tensile strength of 7.1 megapascals. Adding 10 percent metakaolin raised cube compressive strength to 135.24 megapascals and elastic modulus to 45.29 gigapascals, but flexural strength fell to 19.54 megapascals. The basalt gravel mixture showed the opposite pattern: cube compressive strength dropped to 109.40 megapascals while elastic modulus climbed to 55.28 gigapascals, yet flexural strength fell to 15.09 megapascals. The combined metakaolin and basalt formulation recovered a cube compressive strength of 126.31 megapascals and achieved the highest elastic modulus of the group at 59.72 gigapascals, with a tensile strength of 7.3 megapascals. The reinforcing bars were HRB400 steel with a measured yield strength of 455 megapascals and tensile strength of 596 megapascals, and the normal concrete reached a cube compressive strength of 54.00 megapascals.

The bending results tracked those material trends with striking consistency. The plain ultra-high performance concrete joint beam cracked at 38 kilonewtons and failed at 126 kilonewtons; the metakaolin joint beam cracked at 39 kilonewtons and failed at 131 kilonewtons, an ultimate load 3.97 percent higher than the conventional joint. The basalt gravel joint beam was the weakest of the group, cracking at 34 kilonewtons and failing at 115 kilonewtons, an 8.73 percent reduction in ultimate load. Adding metakaolin to the basalt mixture restored most of that loss, with the combined-material rectangular joint beam reaching 37 kilonewtons at cracking and 123 kilonewtons at failure. The rhomboidal joint beam led on every index, cracking at 41 kilonewtons and failing at 133 kilonewtons. The authors note that the 3.97 percent ultimate-load gain from metakaolin may fall within typical experimental variability, though it was accompanied by a positive contribution to cracking resistance. They also flag a caveat: preloading to 30 kilonewtons during test setup, applied to verify equipment calibration and specimen seating, approached the elastic limit for the lowest-cracking specimen and may have induced micro-damage that slightly lowered its measured cracking load.

Crack development followed a consistent sequence across all five beams. Before cracking, the beams behaved elastically; at the cracking load, the first tiny cracks appeared in the normal concrete near the joint rather than in the joint itself. By 80 kilonewtons, cracks had multiplied and extended upward with a tendency to penetrate the beam, and at ultimate load their number and width peaked, with some cracks running through the whole section. Ductility, measured as the difference between ultimate deflection and cracking deflection, told a parallel story. The metakaolin joint beams showed gaps of 10.3 and 9.6 millimeters, exceeding the 9.9 and 8.7 millimeters of their plain and basalt-only counterparts, indicating improved ductility, while the basalt gravel beams showed the poorest deformation resistance and more interface cracks. The researchers attribute the metakaolin benefit to a potential late-stage pozzolanic reaction, with incompletely reacted metakaolin at the interface continuing to react and generate hydrated calcium silicate and other compounds that strengthen the bond. The basalt penalty is ascribed to a weakened interfacial transition zone around the coarse aggregate, which disrupts steel fiber distribution and weakens fiber-matrix bonding, reducing flexural tensile strength.

Geometry changed the mechanics in a different way. The rhomboidal joint enlarges the contact area between the ultra-high performance concrete and the normal concrete and introduces a mechanical interlocking effect: whereas a rectangular joint bears the full applied load across its interface, the inclined surfaces of a diamond-shaped joint resolve that load into components along and normal to the interface, reducing the force the interface must carry. In the rhomboidal beam, cracks formed mainly in the normal concrete area, the compressive and tensile strains measured in the joint were significantly smaller than in the rectangular counterpart, and the beam retained high deformation resistance after cracking. Strain gauges on the longitudinal reinforcement, placed 30 millimeters on either side of the interface and 130 millimeters away at midspan, showed that the strongest joint formulations shared more bending shear stress with the joint, reducing the strain carried by the steel. Across all beams, the tensile and compressive strains of the normal concrete ran roughly twice those of the ultra-high performance concrete under the same load, reflecting the lower elastic modulus of the normal concrete.

The load-deflection curves passed through the three classic stages of elastic response, elastoplastic behavior as cracks grew and the steel took over tensile forces, and stiffness degradation as the reinforcement yielded, with the rhomboidal beam showing the steepest curve after steel yielding, indicating the greatest post-yield stiffness. To check the reliability of the results, the team compared measured capacities against two established flexural strength calculation methods. The first, which assumes uniform tensile stress on the tension side, predicted ultimate moments of 29.8, 30.9, 27.2, 29.1, and 31.5 kilonewton-meters for the five beams, while the second, a linear stress variation distribution method, predicted 23.2, 24.1, 21.1, 22.6, and 24.5 kilonewton-meters, about 22 percent lower and consistent with prior findings. The authors conclude that the linear stress distribution method provides the more accurate prediction for these geometrically optimized joints, and that combining metakaolin chemistry with rhomboidal geometry offers an economical, sustainable strategy for high-performance wet joints in prefabricated bridge construction, though the gains in ultimate load are modest and the findings rest on a small laboratory campaign of five beams under static four-point loading.

Subject of Research: Flexural performance of metakaolin-modified coarse-aggregate ultra-high performance concrete wet joints with optimized geometry in prefabricated composite beams

Article Title: Flexural behavior of CA-UHPC wet-joint beams: Effects of metakaolin, basalt gravel, and joint shape optimization

Article References: Ma, Z., Nishiwaki, T., Ahmad, M. A., Yin, Y., Yi, F., & Tao, H. (2026). Flexural behavior of CA-UHPC wet-joint beams: Effects of metakaolin, basalt gravel, and joint shape optimization. Results in Engineering, 32, Article 113609. https://doi.org/10.1016/j.rineng.2026.113609

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113609

Keywords: ultra-high performance concrete, wet joints, prefabricated bridges, metakaolin, basalt gravel, coarse aggregate UHPC, four-point bending test, flexural strength, rhomboidal joint geometry, interfacial bonding, ductility, sustainable construction