Influence of stitching parameters on the tensile performance and failure mechanisms of CFRP T-joints

2.1 Manufacture of stitched T-joint and tensile test

Each T-joint is stitched through the flange-skin interface on both sides of the web, with the stitch line positioned 10 mm from the web. Kevlar thread is employed to ensure better stitching effect. The web and two L-shaped stiffeners are each constructed from six plies of carbon fiber prepreg arranged in a [0/90]₃ stacking sequence. The reinforcement stitching involved three main variables: stitch method, fineness, and needle distance. As shown in Figure 1, four stitching methods are investigated, including improved lock stitch (LS), chain stitch (CS), flat stitch (FS), and temporary stitch (TS). The stitching is performed using threads of three different deniers (500D, 1,000D, and 1,500D) with three distinct needle distances (10, 15, and 20 mm).

Figure 1

Diagram of the stitch method: (a) improved LS, (b) CS, (c) FS, and (d) TS.

Specimens are fabricated via a hot-press molding process as shown in Figure 2. Stitched carbon fiber prepreg preforms are placed in a steel mold (a plate and two steel angles) lined with a release fabric, with bolts used for pressurization. Curing is performed in an oven at 150°C for 1 h, followed by natural cooling to ambient temperature before demolding. The fiber volume fraction of the specimen is 50%. The as-cured parts are then machined to the final dimensions of 200 mm × 35 mm × 90 mm (L × W × H) and a nominal thickness of 2 mm (Figure 2). The specific stitching parameters for each specimen configuration are listed in Table 1.

Figure 2

(a) Fabrication progress, (b) T-joints, and (c) specimens for Mode I interlaminar tests.

The tensile test of T-joint is carried out by CSS-44100 universal testing machine, and the fixture is fixed to the machine through the protruding block at the lower part, and the specimen is fixed by four screws to prevent it from sliding as shown in Figure 3. All the tensile tests are conducted at a speed of 1 mm·min⁻¹ at room temperature. During the tensile process, the deformation of T-joints is observed on the macroscopic scale with the help of a digital camera.

Figure 3

The experimental setup: (a) tensile test of T-joints and (b) Mode I interlaminar test.

2.2 Specimen preparation and mode I interlaminar test

In Figure 2(c), all the double cantilever beam (DCB) coupons are prepared by hot-press molding, with [0/90]_6s stacking sequence and nominal dimensions of 140 mm (l) × 25 mm (b) × 4 mm (h). An initial delamination of 52 mm in length is introduced by embedding a polyimide film at the mid-plane of one end.

To investigate the influence of various stitch methods on Mode I interlaminar fracture toughness, the tests are conducted by DCB specimens in accordance with the ASTM D5528-13 standard (32). The experimental setup is illustrated in Figure 3(b). The test for fracture toughness is conducted on a universal testing machine at a test rate of 1 mm·min⁻¹. Each test is terminated once the crack has fully propagated through the specimen. To facilitate the observation of crack propagation, a scaled paper ruler is adhered to the side of specimen. Three replicate tests are performed for each configuration. The Mode I interlaminar fracture toughness of the laminates is calculated by using the Modified Beam Theory according to the following equation:

where G _I is the Mode I interlaminar fracture toughness, or strain energy release rate (J·m⁻²); P is the applied load (N); δ is the load-point displacement (mm); b is the specimen width (mm); a is the effective crack length (mm); and Δ is the length of the crack at P _max.

3.1 Effect of stitch method on Mode I interlaminar fracture toughness tests

Investigations presented in Sections 3.1 and 3.2 were conducted on specimens fabricated with a standardized set of stitching parameters: a fineness of 500D and a pitch of 10 mm. This methodology was adopted to isolate the influence of various stitching patterns on the interlaminar toughness. Figure 4(a) displays the typical load–displacement curves for specimens with various stitching methods subjected to Mode I interlaminar fracture toughness tests. The curve trends are largely similar across different methods. Video recordings reveal crack initiation and propagation. The load reached its peak value as the crack front extended to the first stitch hole, which corresponds to a crack extension of approximately 5 mm. At this point, stitch failure occurred, leading to a subsequent drop in load. As the crack reached subsequent stitch holes, its growth was briefly arrested, producing slight load rises and two to three local peaks on the curve.

Figure 4

Comparison of various stitch methods in Mode I interlaminar test at (a) typical force–displacement curves and (b) G _I.

Figure 4(b) presents the G _I for all specimens, calculated using Eq. 1. All four stitching methods significantly enhanced the interlaminar fracture toughness. Notably, the LS method doubled the toughness, indicating that it is the most effective approach for improving the Mode I interlaminar performance. As observed in the morphologies of Figure 5, the interlaminar fracture surfaces of the unstitched and stitched specimens exhibit no significant differences, both showing rough areas characteristic of matrix/fiber debonding. This suggests that the enhancement in Mode I interlaminar fracture toughness is primarily attributed to the fracture of the stitches.

Figure 5

Failure morphology of specimens after Mode I interlaminar tests.

3.2 Effect of stitch method on tensile performance of T-joints

In Figure 6, it is worth noting that the curve is divided into three stages: first, the specimen is in the linear elastic deformation stage, and the load is evenly distributed over the skin structure. As the crack increases, the vertical crack begins to propagate along the web, and the crack expands in both directions. The stitched specimen exhibits a higher ultimate force than the unstitched one, demonstrating the strength-enhancing effect of the thread. Among the stitched samples, the method of LS shows the predominant mechanical property as shown in Figure 6(b).

Figure 6

Comparison of various stitch methods in tension test at (a) typical force–displacement curves and (b) maximum displacement and force.

Damage morphology shows rapid crack growth in the triangular area of NS, CS, and TS due to weak horizontal interfaces. It can be seen from the damage morphology that the crack propagation in the triangular area of NS, CS, and TS is rapid, indicating the insufficient strength of the horizontal interface. The crack in the NS specimen propagated more toward the vertical interface than other and a unilateral fracture occurred as shown in Figure 7. This indicates that the stitching strength at the horizontal interface inhibits the propagation of horizontal cracks and enhances the tensile properties.

Figure 7

Tensile failure morphology of T-joints with different stitch methods under three stages.

3.3 Effect of stitch fineness on tensile performance of T-joints

Mechanical curves in Figures 8 and 9 reveal the influence of stitch fineness on T-joint bearing performance. Under constant needle distance, the tensile load of T-joints increases as stitch fineness increases up to 1,000D. Stitch fineness exceeding 1,000D leads to a reduction in the bearing capacity of the specimen. As the fineness increases, the interlaminar failure of the T-joint becomes more obvious, which shows that the fineness of the stitch significantly affects the physical strength and failure mode. For example, the LS-1000-10 specimen achieves an ultimate load of 875.48 N, which shows an increase of 15.36% and 29.41% compared to the 500D and 1,500D specimens, respectively.

Figure 8

Comparison of various stitch fineness in tension test at (a)–(c) typical force–displacement curves and (d) and (e) their maximum displacement and force.

When stitch fineness is below the critical value, a moderate increase in diameter enhances mechanical anchoring and promotes a stable three-dimensional interlocking structure, thereby improving interlayer bonding while preserving carbon fiber continuity without significant damage. At this stage, the stitch–matrix interface area is appropriate to distribute load via shear stress while avoiding excessive local stress concentration. The undamaged fiber network works in tandem with the stitch so that the tensile load increases steadily with increasing fineness.

However, when the stitch fineness exceeds the critical value, the destructive effect of the stitch fineness on the structure begins to dominate. During the implantation process, the increase in the fineness of the stitch leads to a sharp increase in damage to the matrix. The originally continuous fiber force transmission paths are divided into isolated fragments, and these breaking points become fragile areas with stress concentration. Under tensile load, the cut fibers cannot share the stress, resulting in excessive concentration of the load towards the remaining fibers and threads. The microcracks around the stitch point rapidly expand under stress drive and communicate with the interface debonding area, ultimately triggering interlayer shear failure. At this time, the higher the fineness of the stitch, the stronger the coupling effect of fiber damage and interface failure, and the structural bearing capacity will decrease.

3.4 Effect of needle distance on tensile performance of T-joints

Stress bridging, wherein stitches span crack surfaces to transfer load and suppress opening, retards crack propagation and enhances fracture toughness. As shown in Figure 10, specimens with smaller needle distance exhibited a higher peak load. The core mechanism is attributed to an optimized failure mode governed by the enhanced network constraint effect of the denser stitches. Specifically, a reduced needle distance increases the density of mechanical anchoring points. Upon crack initiation in the triangular region, the dense network constrains the crack tip, alleviates stress concentration through interfacial shear transfer, and suppresses delamination propagation. As the crack grows, high-density stitches redistribute stress more uniformly, thereby delaying catastrophic failure.

Figure 9

Comparison of various needle distance in tension test at (a)–(c) typical force–displacement curves and (d) and (e) their maximum displacement and force.

Conversely, larger needle distance creates unreinforced regions that function as weak links. In these zones, delamination accumulates rapidly, initiating a cascade failure where adjacent stitches rupture sequentially as local stress exceeds their strength thresholds. This process culminates in the collapse of the load transfer path and a reduction in load-bearing capacity. The superior performance of specimens with smaller needle distance results from the combined effects of 3D constraint, crack suppression, and load redistribution, which maintain structural integrity and enhance peak load.

Experimental data indicate that for specimens such as the LS-500-10D, increasing needle distance from 10 to 20 mm resulted in a 22.91% reduction in ultimate load capacity, while simultaneously triggering significant exacerbation of crack propagation. This phenomenon originates from the essential function of threads in maintaining structural integrity by suppressing fiber delamination and crack advancement under tensile loading. Elevated needle distance induces non-uniform stress distribution among stitches, thereby weakening localized constraint effectiveness on fiber layers. Particularly in stress-concentrated regions like the deltoid zone, the accumulation of interlaminar shear stresses preferentially initiates delamination, subsequently promoting fiber fracture and crack propagation. Furthermore, the reduced stitch density compromises continuity in load transfer pathways, ultimately manifesting as diminished peak load capacity and degraded displacement ductility.

3.5 Damage mechanism of stitching enforcement

The SEM analysis established a clear relationship between the observed damage phenomena and the underlying failure mechanisms. First, the failed Kevlar threads displayed significant filament necking and fan-like fibrillation (Figure 10a–c). This ductile behavior is identified as a primary energy-dissipating and toughening mechanism. Furthermore, severe localized damage was observed around the stitch thread, including matrix/fiber crushing, matrix cavity, and torsional shear failure of the thread itself (Figure 10c). This damage is attributed to stress concentration and bridging effects of the stitch, which provide another significant mode of energy absorption. Finally, extensive fiber/matrix debonding was the dominant failure feature in the unstitched regions (Figure 10d). This observation confirms that without direct reinforcement, the inherently weak interlaminar interface dictates the final crack path, resulting in catastrophic structural failure.

Figure 10

SEM results of stitched T-joints.

Stress bridging refers to the mechanism in which reinforcement elements spanning the crack surfaces carry part of the applied load and restrain crack opening, thereby retarding crack propagation and enhancing fracture toughness.