Adhesively bonded joints of hybrid flax/jute fibre reinforced composites: An experimental study

Materials

Unidirectional flax fabric, 2/2 twill flax fabric and bidirectional jute fabric were used as reinforcements. The unidirectional flax fabrics (300 g/m²) and 2/2 twill (200 g/m²) were supplied by BComp (Fribourg, Switzerland), and the jute fabric (350 g/m²) was supplied by SisalSul (São Paulo, Brazil). The basic parameters of the fabrics used are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Basic parameters of fabrics ^a .

Parameters	Material
	Flax		Jute
Weave style	Unidirectional	2/2 twill	Plain
Density (g/cm3)	1.50 ± 0.02	1.47 ± 0.02	1.30 ± 0.01
Weight (g/m2)	300 ± 5%	200 ± 5%	350 ± 5%
Thickness (mm)	0.58	0.39	0.75

For the manufacture of the composite adherends, a two-component epoxy resin system, HEX 135 SLOW, supplied by Barracuda Advanced Composites (Rio de Janeiro, Brazil), was used as matrix. A general-purpose epoxy bi-component adhesive, AR345, supplied by Barracuda Advanced Composites (Rio de Janeiro, Brazil) was used in the fabrication of the SLJs. The tensile data of the resin and adhesive used were obtained in previous works and can be seen in Table 2.

OPEN IN VIEWER

Table 2.

Tensile data of the resin and adhesives ³⁴ .

Polymer	Tensile strength (MPa)	Young’s modulus (GPa)	Tensile strain (%)
HEX SLOW 135	60.92 ± 1.42	3.25 ± 0.008	3.20 ± 0.20
AR345 a	53.09	3.09	4.3

Substrate specimen fabrication

The composite plates were fabricated using the compression molding technique with a steel mold and a heated hydraulic press (Solab SL-20, São Paulo, Brazil). The resin system used was HEX 135 SLOW, with a resin/hardener weight ratio of 100/33. The detailed process fabrication can be found in a previous study. ³⁵ A summary of the fabrication method is presented as follows (see Figure 1): Initially, the fibre fabrics were cut to dimensions of 220 × 150 mm. A drying process of the fabrics was carried out to remove moisture, the fabrics underwent heat treatment in a muffle furnace at 100°C for approximately 15 min until a stable weight was reached. After this, the lay-up process was then carried out manually by placing a fabric layer onto a metallic mold, followed by the uniform application of resin using a roller. This process was repeated until all fibre layers were impregnated with resin. To ensure thickness uniformity and minimize void formation, metal spacers, silicone, and polyurethane adhesive tape were used. The metallic mold, containing the laminated fibre layers, was then placed in a heated hydraulic press, where the curing process was performed at 75°C under 30 MPa pressure for 8 h, following the resin manufacturer’s recommendations. After curing, the composite plates were cooled under pressure before removal from the mold. The fabricated composite plates were then cut into specimens using a mini circular saw, with the substrates cut to dimensions of 25 mm × 107.5 mm.OPEN IN VIEWER

Seven different composite configurations were manufactured, including neat jute composites with 11 layers, as well as neat flax fibre composites via unidirectional and twill weave fabrics, both with 11 layers. Additionally, interlaminar hybrid composites were produced, consisting of a 5-layer jute fabric core with symmetrically placed flax fibre layers on the outer surfaces (two and three layers). Schematics and the nomenclature of the composite plates according to the material and reinforcement are shown in Table 3.

OPEN IN VIEWER

Table 3.

Nomenclature of the specimens as a function of material and reinforcement.

Composite	Description	Schematic
Neat Jute	Eleven layers of bidirectional jute fabric reinforced epoxy
Neat Flax UD	Eleven layers of unidirectional flax fabric reinforced epoxy
Neat Flax Twill	Eleven layers of twill flax fabric reinforced epoxy
J/F-UD2	Two layers of unidirectional flax fabric on each side of a five-layer bidirectional jute core reinforced with epoxy.
J/F-UD3	Three layers of unidirectional flax fabric on each side of a five-layer bidirectional jute core reinforced with epoxy.
J/F-Twill2	Two layers of twill flax fabric on each side of a five-layer bidirectional jute core reinforced with epoxy.
J/F-Twill3	Three layers of twill flax fabric on each side of a five-layer bidirectional jute core reinforced with epoxy.

The fibre weight content in the hybrid composites was maintained at approximately 50% for the samples reinforced with twill flax and 55% for the samples reinforced with unidirectional (UD) flax, while the resin + hardener system accounted for the remaining mass of the final composite. The mass ratio between jute and flax in the hybrid J/F-Twill2, J/F-Twill3, J/F-UD2, and J/F-UD3 composites were approximately 65/35, 60/40, 60/40, and 50/50%, respectively. For the neat composites reinforced with twill and UD flax fibres, the fibre content was maintained at approximately 50% and 70%, respectively. As for the neat jute sample, the fibre content was kept at approximately 50%. The plate thicknesses were approximately 4 mm for the hybrid flax/jute samples, 3 mm for the neat samples reinforced with twill flax or UD flax, and 5 mm for the neat jute samples. It is important to note that the laminate configurations investigated in this study involve concurrent variations in fabric architecture, thickness, and fibre content, arising from the intrinsic physical characteristics of the fabrics employed. Although these parameters are inherently interrelated in practical hybrid laminate designs, all joints were manufactured using symmetric adherends to ensure balanced stiffness and geometric consistency within each configuration. Therefore, the adopted experimental approach still provides a consistent basis for assessing the predominant influence of UD and twill flax architectures on joint mechanical performance and failure behavior. Figure 2 shows the representative tensile stress–strain curves of all the materials used as adherends, while a summary of their main properties is presented in Table 4. These composite properties were previously characterized in a previous study by the authors, ³⁵ which provides additional details supporting the values reported here.OPEN IN VIEWER

OPEN IN VIEWER

Table 4.

Main properties of composite materials used as adherends.

Composite	Tensile strength (MPa)	Young’s modulus (GPa)	Tensile strain (mm/mm)
Neat Jute	82.56 ± 3.58	6.12 ± 0.16	0.015 ± 0.001
Neat Flax Twill	156.42 ± 14.38	13.92 ± 0.96	0.013 ± 0.001
Neat Flax UD	378.43 ± 6.48	33.89 ± 0.62	0.017 ± 0.001
J/F-UD2	163.65 ± 3.45	10.69 ± 0.61	0.017 ± 0.001
J/F-UD3	182.48 ± 7.13	15.91 ± 0.15	0.016 ± 0.001
J/F-Twill2	107.13 ± 6.73	7.42 ± 0.18	0.015 ± 0.001
J/F-Twill3	117.98 ± 2.66	7.88 ± 0.20	0.016 ± 0.001

Single-lap joint specimen fabrication

The schematic geometry of the SLJ specimens is shown in Figure 3.OPEN IN VIEWER

The single lap joints (SLJ) were fabricated according to the ASTM D5868 ³⁶ standard. The adherends were cut to dimensions of 107.5 mm × 25 mm, and their bonding surfaces were mechanically treated with 100-grit sandpaper at a +45°/−45° angle relative to the longitudinal axis. This abrasion aimed to increase the surface roughness, improving the mechanical interlocking between the adhesive and the adherend. The sanding residues were removed using compressed air and 99% alcohol, followed by thorough cleaning with 99% acetone to degrease the surface and minimize adhesive failures. ¹⁸

An aluminum mold and steel spacers were used to ensure alignment during fabrication. The epoxy adhesive used was a two-component system mixed in a 100:45 weight ratio, following the manufacturer’s recommendations. After homogenization, the adhesive was manually applied to the bonding area, ensuring uniform distribution and minimizing void formation. The adhesive layer thickness (0.2 mm) and overlap length (12.5 mm) were controlled using steel spacers, preventing the formation of spew fillets (see Figure 4). After assembly, the mold was closed and fixed to apply the necessary pressure for curing. The joints were cured in a hydraulic press at room temperature for 24 h. Alignment pads were fixed to the adherends to minimize bending moments during tensile testing. Any excess adhesive expelled during curing was removed by post-curing abrasion to avoid interference with the mechanical performance of the joints. ³⁷ OPEN IN VIEWER

Measurements and characterization

The SLJs were tested in traction using an Instron^® 5966 universal testing machine (Norwood, Massachusetts, USA), equipped with a 10 kN load cell. The tests were conducted at a crosshead speed of 1 mm/min. Four specimens were tested for each group at room temperature. To ensure proper alignment of the bonded area in the grips, shims with dimensions of 25 mm × 25 mm were placed at the grip areas, as can be seen in Figure 5. After the tests, tensile data and load-displacement curves were extracted for analysis and discussion in the results.OPEN IN VIEWER

For the analysis of fracture surfaces, scanning electron microscopy (SEM) was employed using a TM4000Plus Hitachi microscope (Tokyo, Japan) operated at 15 kV. All analyses were performed under low-vacuum conditions, and the samples were not metal-coated.

Joint failure modes

The failure process of adhesively bonded composite joints is highly complex. According to the international standard ASTM D5573, ³⁸ the failure modes of these joints can be categorized into seven types: (1) adhesive failure, (2) cohesive failure, (3) thin-layer cohesive failure (TLC), (4) fibre-tear failure (FT), (5) light-fibre-tear failure (LFT), (6) stock-break failure (SB), and (7) mixed failure (i.e., when two or more failure modes are present). After examining all the fractured specimens, three main failure modes were identified: fibre-tear failure, light-fibre-tear failure and stock-break failure. Fibre-tear failure, also known as delamination, involves a more severe and widespread removal of fibres and bundle failures. LFT is characterized by a thick adhesive layer on one side, with no adhesive on the other, along with some surface resin and a few fibres removed from the interface. Stock break failure refers to total failure of the adherend material. Mixed failure modes are common for NFRCs as well as natural/synthetic hybrid composite joints, due to the complex stress state of the overlap zone.^21–26

Figure 6 presents the representative failure modes for the joints studied. From Figure 6(a), it can be seen that the failure mode of the Neat Jute SLJs was a stock break failure. This failure occurs at the exact end of the overlap of one of the joint ends and is because jute fibres and their composites are more brittle.^22,24 Consequently, the JFRP is less able to rotate its edges to accommodate the increasing peel stress caused by load eccentricity in an SLJ shear test. Figure 6(b) and (c) show the J/F-UD2 and J/F-UD3 SLJs failure modes, where a widespread light fibre-tear failure (LFT) can be seen. For the J/F-Twill2 and J/F-Twill3 SLJs (see Figure 6(d) and (e)), a widespread fibre tear (FT) or delamination failure can be observed. Finally, it is possible to observe in Figure 6(f) and (g), that the failure modes of the flax neat SLJs, (i.e., Neat Flax Twill and Neat Flax UD), presented a widespread FT and LFT failure mode, respectively. In general, the main difference in failure modes was due to the flax fabric architecture, with failures closer to the interface in the UD flax joints (neat and hybrids) and generalized delamination in the Twill flax joints (neat and hybrids).OPEN IN VIEWER

In order to investigate the fracture mechanisms and fracture surface morphology, a SEM analysis of the fractured bonded areas was performed on Neat Flax Twill and Neat Flax UD SLJs. Representative SEM micrographs of the adherend surfaces after failure are presented in Figure 7. Since the failure modes were similar between the pairs (i.e., LFT in the UD flax-reinforced samples and generalized delamination in the twill flax-reinforced samples), the failure modes of these SLJs will be used to further discuss the general failure modes. From Figure 7(a), the fracture of the Neat Flax Twill SLJ shows severe and widespread delamination, with multiple layers removed and failures in some fibre bundles. This was the predominant failure mode for all flax twill SLJs. Figure 7(b), on the other hand, illustrates the fracture of the Neat Flax UD SLJ, where the failure mode was dominated by light fibre-tear (LFT). Resin removal from the adherend, fabric and fibre imprints, along with the superficial delamination of a fibre in the thick adhesive layer on the adherend’s right side, can be observed.OPEN IN VIEWER

This study suggests that the architecture of flax fabrics used as reinforcement, namely UD and Twill, influences crack development and, consequently, the failure mode of the joints. For example, in SLJs with flax UD fibre-reinforced substrates (see Figure 7(b)), which exhibit a compact fabric of twisted yarns aligned in a single direction, the resin content between the yarns tends to be reduced, thereby increasing the effective contact area between the yarns and the matrix, as previously demonstrated in the literature. ³⁹ Additionally, fabrics oriented at 0° help prevent crack propagation into deeper regions of the substrate. As a result, the crack tends to follow the path of least energy dissipation, occurring in the resin or at the resin/fibre interface, as shown in the literature.^40,41 Thus, when the applied load on the joint reaches a critical point of plastic deformation of the resin, the crack initiates at one of the adhesive/resin interfaces of the adherend. Being unable to penetrate further, it develops superficially within the resin, removing it almost entirely along with some residual fibres from the first layer. This results in a generalized LFT failure in all SLJs with flax UD adherends. Additionally, a chevron-like pattern is observed in the adhesive layer of one of the adherends, resulting from the imprint of the fibres from the first layer of the other adherend. This pattern is due to the twisting of the unidirectional fabric fibres (see Figure 7(b)).

On the other hand, in SLJs with flax twill fibre-reinforced substrates (see Figure 7(a)), the bidirectional structure of the twill fabric directly influenced the failure behavior. In this configuration, the warp fibres, aligned with the principal direction of the external load, bore most of the applied tension, while the weft fibres, oriented at 90° to the load, were more susceptible to debonding due to rotation. This facilitated the pull-out of the weft fibres due to the complex stress field in the overlap, particularly the high peel stresses at the overlap edges. The pull-out of the weft fibres from the first layer created vulnerable zones that allowed the crack to penetrate between the composite layers. As the crack advanced, the interaction between the debonding stresses and the fabric structure intensified the propagation of delamination more broadly. The result was generalized delamination, as observed in all SLJs with twill flax fabric-reinforced adherends.

Some studies demonstrate how the orientation of the lamina adjacent to the adhesive layer influences the determination of the failure mode of single-lap joints (SLJs). It was shown in the literature that the orientation of the lamina in contact with the adhesive layer can significantly influence the failure mode in single-lap joints (SLJs).^39,42 A 0° interface ply resulted in cohesive fracture within the adhesive layer, which could occur near the contact region between the adhesive and the adherend resin. On the other hand, a 90° interface ply led to crack propagation into the composite adherend, following a predominantly intralaminar path. Additionally, it was observed that the crack tends to stop when reaching the first 0° ply, indicating that the presence of fibres oriented in this direction helps to restrict crack propagation within the composite. Sadeghi et al. ⁴³ investigated the failure modes in E-glass/epoxy composite adhesive joints, focusing on the effect of the fibre angle adjacent to the semi-rigid adhesive layer. Experimental results and numerical analysis indicated that increasing the fibre angle from 0° to 90° significantly altered the failure modes. For the 0° angle, failure occurred predominantly at the adhesive interface, with cohesive fracture of the adhesive layer. As the angle increased, failure shifted towards the composite interior, propagating intralaminarly. A similar relationship was reported by Oliveira et al., ²⁷ who showed that configurations promoting controlled, interfacial or near-interfacial fracture (TLC/LFT-type) tend to maintain crack propagation confined to the adhesive–adherend interface, while architectures prone to deeper fibre–resin damage or heterogeneous fracture favour unstable intralaminar propagation. Overall, flax architectures that constrain the crack path to the interface delay the onset of severe delamination and promote more stable failure progression.

Effect of woven architecture and hybridization on the bonded joint efficiency

Figure 8 shows the representative load-displacement curves of the SLJs, while Figure 9 presents the quantitative statistical data. In Figure 8, it can be observed that the overall behavior of the joints was slightly ductile, with a final brittle failure. The only exception was the Neat Jute samples, due to the fragile nature of the jute composite and the rupture failure of the stock (see Figure 6(a) – failure mode). From Figure 8, it can be noted that all samples initially exhibit linear elastic behavior up to approximately 0.2 mm of displacement, after which nonlinear behavior begins to be observed. This is possibly due to plasticity and damage initiation in the adhesive layer. It is also evident that the apparent stiffness of the joint varied depending on the type of composite material used as the adherend. For example, in Figure 8(a), it is noted that the Neat Jute joint showed slightly higher apparent stiffness than the joints with adherends reinforced with flax twill (neat and hybrid). In Figure 8(b), it can be seen that the Neat Jute joint, when compared to joints with adherends reinforced with unidirectional flax, showed slightly lower stiffness than all the samples. Among the joints with neat flax composite adherends, the SLJ Neat Flax UD exhibited higher apparent stiffness compared to SLJ Neat Flax Twill. The experimental results suggest that fabric architecture significantly influences the apparent stiffness of the joint. Although the neat flax substrates show distinct mechanical properties resulting from their respective fabric architectures (Table 4), the hybrid joints exhibit a response strongly governed by the fibre arrangement at the bonding interface. The unidirectional architecture, characterized by a greater fraction of fibres aligned with the loading direction, provided a more direct load-transfer path, resulting in higher apparent stiffness compared to the twill counterparts. In contrast, the interlaced geometry of the twill fabric promotes stress redistribution effects that reduce the efficiency of load transfer through the adherend. Notably, the hybrid joints exhibited stiffness similar to their neat counterparts (i.e., comparing J/F-UD2 and J/F-UD3 with Neat Flax UD, or J/F-Twill2 and J/F-Twill3 with Neat Flax Twill), indicating that hybridization can achieve performance levels comparable to neat composites with little influence from the number of external layers. Overall, this behavior reflects the coupled influence of architecture, thickness, and fibre content on the global mechanical response of the joints.OPEN IN VIEWER

OPEN IN VIEWER

Figure 9.

Quantitative data for the SLJ tests as a function of flax woven architecture: (a) Flax Twill and (b) Flax UD.

From Figure 9 and Table 5, the quantitative data of the joints can be observed. It is noted that the performance of the joints was strongly influenced by the architecture of the flax fabrics used in the hybridization of the jute/flax adherends, which is reflected in the observed failure loads. The Neat Jute joint, although presenting inferior mechanical properties in the adherend, achieved failure loads (3131.32 N) like those of the joints with twill flax adherends (neat and hybrid). This performance can be attributed to the greater thickness of the jute joint (5 mm), which contributed to a higher load limit. The stock-break failure mode (see Figure 6(a)) indicates that adhesion was efficient, and that the substrate sustained the load until rupture. Although jute fibre is more brittle, as previously reported in the literature, ²² the greater joint thickness allowed the adherend to withstand the load up to the failure point, unlike twill flax, whose delamination failure compromised its structure before the complete failure of the joint.

OPEN IN VIEWER

Table 5.

Quantitative data for all the joints tested as a function of flax woven architecture: Flax UD and Flax Twill.

Joint type	Failure load (N)	Displacement (mm)	Lap shear strength (MPa)
Neat Jute	3131.32 ± 288.43	1.16 ± 0.03	10.02 ± 0.92
J/F-Twill2	3080.27 ± 61.79	1.15 ± 0.28	9.86 ± 0.20
J/F-Twill3	3130.36 ± 136.48	1.15 ± 0.29	10.02 ± 0.44
Neat Flax Twill	3168.08 ± 133.78	1.16 ± 0.28	10.14 ± 0.43
J/F-UD2	4289.94 ± 149.51	1.13 ± 0.09	13.73 ± 0.48
J/F-UD3	4576.62 ± 304.35	1.10 ± 0.06	14.65 ± 0.97
Neat Flax UD	4684.54 ± 443.79	1.05 ± 0.09	14.99 ± 1.42

In Figure 9(a), when comparing the hybrid twill flax joints with the Neat Jute joint, a plateau behavior in the average failure loads was observed, where SLJ J/F-Twill2 and J/F-Twill3 reached 3080.27 N and 3130.36 N, respectively. On the other hand, in Figure 9(b), when comparing joints with hybrid adherends reinforced with UD flax, considerable improvement can be observed. For example, SLJs J/F-UD2 and J/F-UD3 showed improvement with increases of approximately 27% and 32%, respectively.

When compared with previously reported flax-based SLJs, the present hybrid joints exhibited competitive performance. Oliveira et al. ²⁷ reported failure loads of approximately 3284 N for Flax-Twill 250, 3892 N for Flax-Twill 400 and 2952 N for Flax-Biax configurations, whereas the present J/F-UD2 and J/F-UD3 joints reached average failure loads of 4289 N and 4577 N, respectively. While these average values are higher, the experimental scatter typical of natural fibre composites due to inherent fibre variability and processing sensitivities suggests that the results are within a similar performance range. The apparent superiority of the UD-hybrid configurations can be attributed to the higher properties of the hybrid substrates with UD flax—tensile strength and composite stiffness (see Table 4), which help minimize the rotation of the overlap edges during loading, as well as by the fabric architecture, which conditions the failure to a more controlled and less severe mode (light fibre tearing—LFT) and delays the onset of delamination failure, as observed in the previous section.

When comparing the 3-layer hybrid twill flax SLJ (J/F-Twill3) with the 2-layer one (J/F-Twill2), the difference in the average failure load was marginal (an increase of ∼1.6%), suggesting a limited benefit from the additional layer. For the 3-layer unidirectional flax hybrid SLJ (J/F-UD3) compared to the 2-layer one (J/F-UD2), a slightly higher increase of approximately 6.3% is noted, which is small when compared to the inherent variability (standard deviation) of the tests. However, the difference in failure loads becomes more evident when comparing hybrid joints with different flax fabric architectures (i.e., twill vs UD), while keeping the number of layers the same. For instance, the 2-layer twill flax hybrid SLJ (J/F-Twill2) exhibited a failure load approximately 28% lower than that of the 2-layer unidirectional flax hybrid SLJ (J/F-UD2), while the 3-layer twill flax SLJ (J/F-Twill3) had a failure load approximately 32% lower than that of the 3-layer unidirectional flax SLJ (J/F-UD3). Therefore, the flax fibre architecture is a parameter that influences the average failure load performance of hybrid SLJs, highlighting the importance of considering fabric characteristics in the design of structural composites, while the increase in the number of layers showed a limited impact under these specific conditions.

When comparing the hybrid SLJs with their respective neat joint counterparts, i.e., with Neat Flax UD and Neat Flax Twill, some similarities can be observed. For example, the hybrid joints with UD flax (J/F-UD2 and J/F-UD3) achieved 92% and 98% of the failure load of Neat Flax UD (4289.94 ± 149.51 N and 4576.62 ± 304.35 N vs 4684.54 ± 443.79 N, Table 5), demonstrating near-equivalent performance despite hybridization (<10%). Similarly, twill flax hybrids (J/F-Twill2 and J/F-Twill3) reached 97–99% of Neat Flax Twill strength (3080.27 ± 61.79 N and 3130.36 ± 136.48 N vs 3168.08 ± 133.78 N), suggesting that adding flax layers beyond two provides negligible gains (<3% difference).

To further assess the observed similarity between the hybrid and neat flax joints, a rigorous statistical validation was performed. A one-way ANOVA followed by Tukey’s post-hoc test was performed on the failure load data, considering the number of specimens per group as described in the methodology. In this analysis, a p-value <0.05 indicates that the investigated parameter (architecture or layer number) significantly affects the joint performance, while p > 0.05 suggests that any observed variations are within the expected experimental scatter. The statistical results confirmed that while the UD-hybrids (J/F-UD2 and J/F-UD3) showed a highly significant improvement over the Neat Jute baseline (p < 0.001), the differences between Neat Jute and the twill-hybrids were non-significant (p = 0.992 for J/F-Twill2 and p = 0.999 for J/F-Twill3). Furthermore, the analysis verified that increasing the number of flax layers from two to three provided no significant gains for either UD (p = 0.642) or Twill (p = 0.988) architectures. In contrast, when comparing hybrids with the same number of layers but different architectures (UD vs Twill), the differences were highly significant (p < 0.001), reinforcing that fabric geometry is the primary driver of performance. Finally, no statistically significant differences were found between the hybrid joints and their respective neat flax counterparts (p = 0.435 for J/F-UD2 vs Neat UD and p = 0.885 for J/F-Twill2 vs Neat Twill). These findings statistically validate that a hybridization strategy with only two external layers is sufficient to replicate the performance of neat flax joints, regardless of the architecture used.

The average Lap Shear Strength (LSS) was utilized to evaluate joint performance, as this normalized metric allows for a clearer observation of how the external flax fabric architecture influences the assembly, despite the inherent variations in substrate thickness and fibre content. As shown in Figure 10 and Table 5, the LSS results reveal a significant disparity between the two architectures: while UD-reinforced hybrids achieved a substantial increase in shear strength compared to the Neat Jute baseline, the twill-reinforced joints showed a performance plateau, with values nearly identical to the neat jute. This behavior suggests that the intrinsic efficiency of the joint is predominantly dictated by the fabric’s mesoscale geometry at the bonding interface rather than solely by global substrate parameters such as adherend thickness.OPEN IN VIEWER

This architectural influence is directly linked to the observed failure modes (see Figure 6). In the UD flax hybrids, the alignment of the fibres at 0° promotes a more stable stress distribution that constrains the crack path to the adhesive/resin interface, leading to light fiber-tear (LFT) failure. This mechanism allows the joint to reach higher LSS values (up to 14.65 MPa) by delaying the onset of severe structural damage. In contrast, the bidirectional structure of the twill fabric facilitates crack penetration into the laminate. The interlacing of warp and weft fibres creates regions of stress concentration that promote early delamination, regardless of the additional thickness provided by extra layers. ²⁴ Consequently, the twill hybrids exhibit a lower and more stagnant LSS (approx. 10 MPa), as the failure is governed by the premature interlaminar rupture of the twill architecture.

Therefore, the LSS highlights that, although substrate stiffness and thickness contribute to joint performance, the superior behavior of UD hybrids is predominantly associated with the more favorable failure mechanism promoted by the unidirectional architecture. Even with marginal differences in fibre content and thickness between the 2-layer and 3-layer configurations, the LSS remains primarily sensitive to the fabric type. This confirms that, for the hybrid jute/flax joints investigated in this study, the external layer architecture represents one of the most influential design parameters governing bonded joint strength and failure progression.

The observed similarity in the average failure load and LSS can be associated with the failure modes of the SLJ joints. As previously discussed, for both SLJ joint failure modes, there is a critical load at which plastic damage initiates in the adherend surface resin, causing the crack to develop, either toward the inner layers of the adherends (twill flax SLJ joints) or toward the resin/fibre and adhesive interface (UD flax SLJ joints). As observed, SLJ joints of each flax architecture type exhibited similar failure modes, demonstrating that failure always begins at the same load, remaining below the plastic damage threshold of the adherend. Therefore, even if the adherend properties are improved, this will not influence the critical load required for the initiation of plastic damage in the adherend surface resin. This explains why SLJ joints exhibited similar failure behaviors among their counterparts, with failure progressing predictably regardless of changes in adherend properties. Finally, this similarity in performance highlights the importance of considering other factors, such as material architecture and layer configuration, when selecting joint designs for specific applications.

To sum up, the experimental results demonstrate that the performance of hybrid jute/flax SLJs is governed predominantly by the external flax fabric architecture rather than by the mere increase in the number of reinforcement layers. While the addition of a third flax layer provided limited improvements, the use of unidirectional flax significantly enhanced joint efficiency by promoting more stable failure mechanisms and delaying delamination. These findings indicate that hybrid configurations with only two external flax layers can provide a favorable balance between structural performance, material usage, and cost, highlighting their potential for lightweight and sustainable bonded composite applications.