PSI - Issue 64

Shaofeng Qin et al. / Procedia Structural Integrity 64 (2024) 168–174 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

172

5

(3)

ℎ = ( 0 −1)×100%

where, i 0 is the initial current without loading while i is the current as the displacement changes.

3. Results 3.1. Flexural tests and electrical measurement

Fig. 2a displays the fiber orientation angle of the LCF group relative to the printing direction, with a high R-squared value (R 2 =0.9682) indicating close alignment with the printing direction. The permissible angle range between fiber orientation and printing direction is ideally within ±18° due to the injector head size limitation when comparing the length of CF (Fig. 1a). The cumulative probability within ±18° exceeded 70%, demonstrating effective control over fiber alignment parallel to the printing direction in extrusion-based 3D printing. In Figure 2b, the fiber orientation angles for the three groups are summarized. The MCF group exhibited slightly lower cumulative probability within ±18° compared to the LCF group but notably higher than the RCF group. Moreover, the lower R-squared value of the RCF group suggests random fiber alignment. Interestingly, the resistivity of the LCF group was lower than that of the RCF group (Fig. 2c), being different from the previous study in that the percolation network was easily formed via random fiber distribution [William et al. (2023)]. The variation may be attributed to the smaller region for voltage measurement (20 mm, see Fig. 1c), and directionally arranged fibers make it easier to form channels that connect electrodes (e.g., 12 mm + 12mm > 20 mm). Under random conditions, CFs tend to connect more easily, but a larger number of fibers are needed to reach the measured electrode. The MCF group shows similar results in both average and error, possibly due to uncertain and unstable connections between the random and longitudinal regions. In terms of mechanical properties, the stress vs. strain plots of various groups including the reference pasts without CFs are depicted in Fig.2d. Compared to the Ref. samples without fiber enhancement, all the fiber-cement composites (LCF, RCF, and MCF group) presented significantly higher flexural strength, indicating the strengthening effect due to the addition of the CFs. Interestingly, the stress vs . strain behaviors of fiber-cement composites show notable deflection softening after cracking, unlike the brittle fracture seen in Ref. group, suggesting fiber bridging in the tension zone. Fig.2e reveals that the average flexural strength of groups with specified fiber alignment (LCF and MCF) is slightly higher than the RCF group; however, LCF and MCF groups exhibit larger variations. This suggests that the random fiber arrangement may provide a stable distribution on the failure section, while the controllable extrusion based printing may result in uneven fiber distribution (higher shear stress near the injector head wall than at the center of extrusion) and potential weak interface defects of the printed composite strip. In the previous research, the high strength cement-based matrix incorporated with more than 1.0% volume fraction of shorter CFs (3 mm length) presented significant deflection hardening, and besides, the CFs aligned with the printing direction provided higher flexural strength than the random one [Hambach et al. (2016)]. The comparison implies the difference in failure situations (single or multiple cracking modes), which can be achieved by the material tailoring in aspect ratio and length of CFs, the strength and toughness of cement matrix, fiber-cement bonding as well as the additional volume fraction and alignment of CFs. The porosity of the midspan of the groups was characterized via mercury intrusion porosimetry (MIP) testing, as shown in Fig. 2f, and the fiber-cement composite presented a similar distribution of the harmful pores (larger than 4.5 nm). The similarity in porosity and strength results, as well as a larger gap in the resistivity between Ref. group and fiber-cement composites, imply that the electrical and strengthening mechanical properties of c-FRCC are heavily determined by the fibers rather than the defects and pores.