In recent years, the combination of construction 3D printing technology and ultra-high-performance concrete (3DP-UHPC) leverages the rapid construction and controllable shaping advantages of 3D printing along with the ultra-high strength and high toughness advantages of UHPC. Existing research on 3DP-UHPC, constrained by current printing technologies, has primarily focused on shorter steel fibres (6 mm and 10 mm), whereas the steel fibre length commonly used in engineering UHPC is 13 mm. Additionally, research on the micro- and meso-structural characteristics of 3DP-UHPC and the resulting anisotropic mechanical properties and damage-fracture mechanisms is relatively scarce. Therefore, revealing the multiscale effects of anisotropic compressive mechanical behaviour in 3DP-UHPC is of vital importance. This study employs 13 mm steel fibres to investigate the influence of different fibre volume fractions (0%, 0.5%, 1.0%, 2.0%) on the compressive performance of 3DP-UHPC. By combining digital image correlation technology (DIC) and X-ray computed tomography (CT) methods, we aim to clarify the coupling mechanism between fibre orientation and interlayer/interstrip interfacial defects, providing critical insights for material optimisation and engineering deployment of 3DP-UHPC.

In the experiment, 3DP-UHPC specimens with steel fibre volume fractions of 0%, 0.5%, 1.0%, and 2.0% were prepared. First, a particle packing-based mix design was carried out. By testing the particle size distribution of the raw materials and applying the modified Andreasen-Andersen model, the optimal packing proportions of the mortar matrix components were determined. Subsequently, in accordance with the GB/T 2419—2005 standard, a flow table test was conducted using an electric jolting apparatus to evaluate the printability of mixtures with varying fibre volume fractions. Furthermore, high-resolution X-CT was employed to acquire the internal structure of the specimens, from which microstructural parameters such as pore morphology and steel fibre orientation were extracted. Finally, uniaxial compression tests were performed using DIC to measure the compressive strength and to capture, in real time, the processes of crack initiation, propagation, and coalescence, thereby revealing the evolution characteristics of the failure modes. In this process, the synergistic effects between interlayer interfacial defects and fibre orientation were further analysed, and their coupling mechanisms during damage evolution were investigated, thereby providing deeper insights into the microstructural mechanisms governing the anisotropic mechanical behaviour of 3DP-UHPC.

The test results indicate that in the 3DP-UHPC specimens, a total of 37 irregularly shaped large pores were observed at interlayer and interstrip interfaces, with equivalent diameters ranging from 2000 μm to 3200 μm. These accounted for only 0.174% of the total pore volume, while the remaining 92.195% consisted of nearly spherical or ellipsoidal pores, resembling the typical pore morphology of cast UHPC. Notably, the maximum (3200 μm) and average (264 μm) equivalent pore diameters in the printed specimens were both smaller than those reported for cast UHPC (4500 μm and 700 μm), suggesting that the extrusion process during printing effectively compacts the material and facilitates air removal. Therefore, the total porosity of 3DP-UHPC (2.24%) was lower than that of cast UHPC (3.45%-5.90%).

Steel fibres within 3DP-UHPC exhibit a clear tendency to align along the printing direction (x-axis), with more pronounced orientation in the x-z plane than in the x-y plane. Specifically, 65.2% of fibres were aligned at θ = 75°-90° in the x-z plane, compared to 32.8% at φ = 75°-90° in the x-y plane. This anisotropic distribution stems from the filament geometry, where the height (17 mm) is smaller than the width (36 mm), creating spatial constraints that favour alignment in the vertical cross-section. In total, 3,324 fibres (52%) were oriented within the 60°-90° range, reflecting a strong directional preference induced by the layer-by-layer deposition process.

The compressive strength of 3DP-UHPC exhibits directional anisotropy, influenced by layer interfaces, fibre volume fractions, and fibre orientation. At fibre volume fractions of 0.5%, 1.0%, and 2.0%, compressive strength consistently followed the order x < y <z, with the weakest-to-strongest directional differences being 2.95%, 7.82%, and 19.50%, respectively. These values are all below 20% and notably lower than those reported in earlier studies. The achieved reduction in anisotropy is attributed to the optimisation of printing parameters in this study, particularly adjustments to printing speed and path, which improved inter-filament contact, enhanced pore filling, and mitigated the adverse effects of interfacial defects.

3DP-UHPC outperforms cast UHPC in compressive strength across all principal directions (x, y, and z) at fibre volume fractions of 0%, 0.5%, and 1.0%. This advantage is due in part to the absence of fibre sedimentation, water bleeding, and entrapped air defects that may occur during vibration-based casting. The extrusion-based fabrication in 3D printing mitigates these issues, resulting in improved uniformity and mechanical properties. At a fibre volume fraction of 2.0%, the printed specimens exceeded cast specimens in compressive strength along the z-axis by 10%, while showing slightly lower values in the x- and y-directions.

Increasing the steel fibre volume fraction enhances the compressive strength of both 3DP and cast UHPC. For the cast specimens, an increase in fibre content from 0% to 2.0% led to a 49.9% improvement in compressive strength. In the 3DP-UHPC specimens, the same increase led to strength enhancements of 16.8%, 11.0%, and 37.7% in the x-, y-, and z-directions, respectively, indicating that the reinforcing effect of fibres is directionally dependent. Even at a modest fibre volume fraction of 0.5%, compressive strength stabilised around 133 MPa across all directions, indicating a beneficial role in mechanical consistency. In contrast, fibre-free specimens showed brittle failure dominated by interfacial defects, resulting in greater variability in strength measurements.

The 3D printing process significantly influences the microstructural characteristics and mechanical performance of 3DP-UHPC. Although certain interlayer and interstrip voids are inevitably introduced during the 3D printing process, the extrusion-induced "degassing" effect helps lower the overall porosity, resulting in a denser microstructure in 3DP-UHPC compared to conventionally cast UHPC. Under the combined effects of nozzle extrusion and printing path guidance, steel fibres tend to align along the printing direction, in contrast to the random distribution observed in cast specimens. This directional alignment improves the crack-bridging efficiency of fibres. The compressive failure behaviour exhibits pronounced anisotropy: under loading in the x-direction, cracks preferentially propagate along interlayer interfaces, while in the z-direction, fibre alignment enhances bridging capacity and delays crack growth. The coupled effects of fibre orientation and interfacial quality are primary contributors to this anisotropy. Enhancing the extrusion-interlocking mechanism between adjacent filaments can improve interfacial densification, thereby mitigating mechanical anisotropy. The steel fibre volume fraction also plays a critical role in both strength and stability: increasing the fibre content significantly enhances compressive strength, particularly under loading in the z-direction; in contrast, lower fibre volume fractions, although less effective in improving strength, contribute to greater performance uniformity and structural consistency. This study further elucidates the multiscale mechanisms through which 13 mm steel fibres influence compressive anisotropy in 3DP-UHPC. It identifies key microstructural features, such as pore distribution and fibre orientation, and preliminarily establishes their correlation with macroscopic mechanical behaviour, providing valuable insights for optimising the structural performance of 3DP-UHPC.