Study of the tensile mechanical properties and failure mechanisms of CFRP screwed joints

Abstract

Removable joint technology is commonly used in composite laminates for various load-bearing structures. However, existing research primarily focuses on bolted joints, there is relatively limited research on screwed joints in composite materials. This study investigates the influence of connected layer thickness and hole diameter on the tensile behavior of threaded joints in carbon fiber-reinforced polymer (CFRP) laminates. After fabricating different CFRP screwed joint specimens, tensile tests were conducted. The digital image correlation (DIC) technique captured the deformation process. The experiment results indicate a significant increase in load-bearing capacity with the increase in diameter. For instance, joints with an 8 mm diameter exhibited a load-bearing capacity of 10.82 kN. The increase in the connected layer thickness correspondingly enhanced the load-bearing capacity of the joint. The joint with a thickness of 7 mm had the highest load-bearing capacity of 8.83 kN. Besides, with the increase in the thickness of the connected layer, the failure mode transitioned from shear failure in the connected layer to screw pull-out. The tilt angle of the screw during the pull-out process also decreases with the increase in the connected layer thickness. Strain and out-of-plane displacement measurements under ultimate load conditions verify these observations.

Keywords

Screwed joints tensile properties failure analysis digital image correlation

Introduction

Composite laminates have favorable characteristics such as lightweight, high strength, and excellent corrosion resistance. Hence, they are widely used in fields such as aerospace, maritime, and manufacturing.^1–3 Although composites have replaced traditional metal materials in various applications, it is challenging to connect composite laminates to metal structures or other laminates in structural engineering. Joints are regarded the weakest link in a structure, often prone to damage and failure. Therefore, several researchers have conducted studies to explore the properties of composite joints.^4–6 Typical connection techniques include mechanical connections, adhesive bonding, and hybrid connections.^7–9 Mechanical connections are characterized by their disassembly capability and have been extensively explored in numerous studies.

Most studies on mechanical connections of composite laminates primarily focused on the tensile performance of bolted joints. For instance, Li et al.^10,11 evaluated the effect of the end-to-diameter ratio and width-to-diameter ratio, both set to 3, on the bearing strength of single‐bolt joints. The results showed that load-bearing performance of the joint linearly increased with an increase in the number of bolts. Chao et al.¹² developed a mathematical model to evaluate the effect of tolerance fit on the mechanical performance of composite material bolted connections. The research revealed that the maximum equivalent stress of interference-fit bolts was lower than the stress observed in fully-fit and clearance-fit bolts. Camanho¹³ explored the factors influencing the failure of composite bolted joints and proposed methods for predicting failure. Single lap bolt joints are widely investigated due to their simplicity and effectiveness. However, the configuration of the single lap bolt joint leads to a secondary bending effect due to the eccentric loading path.^14,15 Liu et al.^16,17 used spring elements instead of solid bolt models to predict the secondary bending effects occurring in bolted connections. They improved finite element models by incorporating the deformation of the bolt shank, head, and bore bearings. It is imperative to explore damage development and failure mechanisms for surface strains and out-of-plane deformations of joints. However, it is challenging to determine the full-field strain evolution within the joint using conventional strain gauges and extensometers. This challenge can be effectively circumvented using digital image correlation (DIC) technology.^18,19 Masoud et al.²⁰ used DIC technique to monitor the strain variations in the area beneath the bolt. The results revealed that bypass action caused an increase in strain whereas bearing action led to a decrease in strain.

Several modeling methods have been developed to analyze the mechanical properties of composite bolted joints. Camanho²¹ used the point stress or average stress model to predict the final failure and failure modes. Comparative analysis with the experimental data revealed that the model accurately predicted the failure modes of composite bolted joints. Zhao²² proposed the concept of additional stiffness associated with hole tensile deformation and used its calculation method to modify the deformation compatibility equation. This approach significantly enhanced the accuracy of predicting multi-bolt joints in composite material. Shan²³ used a novel progressive fatigue damage model and the nominal stress method to predict the fatigue failure of carbon fiber-reinforced polymer (CFRP) and aluminum joint plates. Belardi et al.^24–26 used spring elements as replacements for solid bolt models. Subsequently, they assessed the stiffness of the bolt based on the deformation of the bolt shank, head and hole bearings, while considering the load distribution around the bolt. Moreover, they introduced a new composite bolt element to replace one region of the original model. The new element comprised a set of radially aligned beams connected to the bolt and the surrounding area. Additionally, they used CBJE finite element modeling with customized stiffness to simulate and evaluate the mechanical properties of composite multi-bolt joints.

Although extensive research has been conducted on composite bolted joints, the presence of visible bolt heads and nuts on the plates leads to aesthetic and aerodynamic concerns. In addition, rivet joints pose drawbacks as they cannot be removed without damaging the structure.²⁷ In scenarios requiring low load capacity, directly screwing a screw into the threads offers superior removability than bolting. This study focuses on investigating screwed connections in composite materials. Richar²⁸ conducted tensile tests on CFRP screws after fixing the threads. The M5 and M8 threads exhibited good load capacity with an average of 4.42 kN and 6.56 kN, respectively. Zakharova²⁹ experimentally assessed the load-bearing capacity of threads fabricated from carbon-carbon composites at varying temperatures. The findings revealed a significant increase in the load-bearing capacity of the threaded joint between 1900°C and 2200°C. Jeong³⁰ investigated the impact of screw inclination angle on the sliding modulus and shear strength of screw connections in laminated timber-concrete composite materials. Freitas et al.³¹ highlighted that the coating of the tap and the tapping speed are crucial factors affecting the tapping force during the threading process of CFRP. Ahmet et al.^27,32 Explored the load carrying capacity of CFRP composite laminate screwed joints under various hole diameters through compression tests. Punching experiments on CFRP composite screwed joints revealed that the maximum failure load increases with an increase in hole diameter.

The current research primarily involves evaluation of the compression and impact resistance of CFRP screwed joints. Limited studies have explored the mechanical properties of CFRP screwed joints under tensile loading. A comprehensive evaluation of the tensile performance of CFRP screwed joints is crucial because they offer superior removability compared to bolted joints. In this study, tensile experiments were conducted on CFRP screwed joints, and the effects of hole diameter and varying connected layer thicknesses on the load-bearing capacity and failure modes of these joints evaluated. Moreover, the dynamic process of CFRP threaded joints under tensile loading was explored, mainly focusing on a detailed analysis of the thread pull-out failure process. Furthermore, the variations in the screw tilt angle during the thread pull-out process were investigated. The failure mode during the tensile process was verified by determining the surface principal strain field and the out-of-plane displacement field using the 3D-DIC technique.

Experimental design

Specimen preparation

In this study, the specimens consist of a carbon fiber-reinforced polymer (CFRP) laminate with a single through hole serving as the connected layer, and a CFRP laminate with screwed holes serving as the connecting layer. The connected layer is connected by a screw and connecting layer. The outermost layer of the laminate consists of woven prepreg material in a bias ply configuration, the internal fibers were stacked alternately at 0° and 90° degrees. The thickness of the basic CFRP layer is 0.15 mm, and the material properties are shown in Table 1, in which the intra-laminar properties were tested and provided by the manufacturer.¹² The laminates were cured through the autoclave molding method. Orthotropic CFRP laminates constructed with only 0° and 90° carbon fiber layers will have reduced strength along the 45° and 30° diagonal direction. However, the cost will also be reducer, and orthotropic CFRP laminates have achieved very effective use in practical applications.³³

Table 1.

Material properties of the basic CFRP layer.

Items	E₁₁ (GPa)	E₂₂ = E₃₃ (GPa)	G₁₂ = G₁₃ (GPa)	G₂₃ (GPa)	v₁₂ = v₁₃	v₂₃
Values	140	10	5.2	3.9	0.3	0.5

For the experiments on the effect of hole diameter on the screwed connection, the thickness of both the connecting and connected layers was 5 mm to ensure uniformity of the variables. Besides, when investigating the effect of the thickness of the connected layer on the joint, the thickness of the connecting layer with screwed holes was 5 mm, and the thickness of the connected layer was set to h, as shown in Table 2. There is currently no specific design standard tailored for CFRP screwed joint connections. Therefore, the design specimens in this study refer to the design specification ASTM-D5961/D5961 M test standard in bolted connections. The experimental process to ensure that the tensile force of the specimen parallel, thickness compensating shims of 2W × 50 × h are bonded at the end of the plate using Ergo1309 strong AB adhesive, and the specimen of CFRP screwed joints is shown in Figure 1(a).

Table 2.

Specimen names and parameters.

Groups	Thickness h (mm)	Diameter D (mm)	End E (mm)	Wide W (mm)
S1-6	1	6	18	18
S3-6	3	6	18	18
S5-6	5	6	18	18
S7-6	7	6	18	18
S5-4	5	4	12	12
S5-8	5	8	24	24

The specimens were manufactured using CNC-controlled machine tools. The pre-holes in the connecting layer need to be drilled before tapping. The pre-hole operations were performed using a tungsten steel bit, and the tapping operations were implemented with machine taps at a cutting speed of 25 m/min.³² After the tapping process use the tap wrench along the axis of the hole in the direction of slow rotation and move up and down. The tap can be smoothly rotated and no chip generation state indicates that the tapping process has been completed. Manual tapping is done to ensure that the powder generated during the tapping process can be discharged effectively to prevent damage to the screws. As shown in Figure 1(b) the screwed hole obtained after tapping.

Assign unique identifiers to all specimens, denoted as “S,” with the first digit signifying the thickness of the connected layer, the second digit representing the diameter of the screwed hole, and the third digit indicating the serial number. For instance, in the nomenclature, “S3-6-1” designates the initial specimen within the CFRP laminated plate with a thickness of 3 mm and a screwed hole diameter of 6 mm, as outlined in Table 2, providing the specimen parameters. Three specimens were prepared for each parameter. A total of 18 specimens were prepared for this test.

Experimental set up

The universal testing machine (Shimadzu, AGS-X-50kND) was used in this study at the loading rate of 1 mm/min. All specimens are tested on the same testing machine. Spray white matte primer on laminate laps before stretching, and spray the black scattering after the primer has completely solidified to form random patterns on the surface.^34–36

The Vic-Snap9 software was used to capture real-time surface out-of-plane displacement and surface principal strain in the frontal view of the screwed joint. Two 12 M industrial cameras were used to acquire the images and the sampling rate of the cameras was set to capture images every 500 ms. Photographs were captured at the beginning of the tensile test until failure of the specimen occurred. Post-processing analysis of all measurement results using digital image correlation (DIC) professional software Vic-3D9. Set the facet size and facet step to 50 and 7, respectively, based on the resolution and average scatter size of the image to ensure the accuracy of the calculation. The spot patterns of the experimental equipment and specimens are shown in Figure 2.

Results and discussion

Failure analysis

Effect of diameter on failure

Figure 3(a) shows the load-displacement characteristic curves for specimens of different diameters. Compared to the specimen with a hole diameter of 6 mm, the other specimens showed a sudden drop in load after reaching the ultimate load. The sudden drop in load for the 8 mm specimen is attributed to bearing failure. Bearing failure is developed by the compressive forces between the fastener and the hole surface. Fasteners create stress concentrations around holes, and a crack is initiated at the hole edge, which usually leads to a sudden catastrophic joint failure.³⁷ For a 4 mm specimen, a sudden drop in load due to the fracture of a screw can be equally catastrophic in engineering. By comparing the 6 mm and the above two specimens, it can be seen that the load decreases relatively slowly after reaching the ultimate load. When used in engineering, it will be detected in time before complete failure occurs. Besides, before the ultimate load is reached, the 6 mm and 8 mm screws are tilted by the force. This phenomenon results in screw embedded in the laminate leading to localized fiber breakage in the hole, so fluctuations occur before the ultimate load is reached. The 4 mm screw did not tilt and create load fluctuations because it fractured itself.

Figure 3(b) shows the average ultimate load plots for different apertures depicting a clear trend in the load carrying capacity. 8 mm hole diameter joints have a load capacity of up to 10.82 kN. Compared to joints with 4 mm and 6 mm diameters, there was an increase of 99% and 26%, respectively. The graph indicates that the ultimate load capacity of carbon fiber-reinforced polymer (CFRP) screwed joints significantly increases with the enlargement of screw diameters. This enhancement is due to the increase in diameter leading to an increase in contact area. This increases the friction and thus the load carrying capacity. On the other hand, the increase in diameter also increases the shear strength of the screw. So sometimes larger screw diameters result in a greater force being required to achieve failure.

Table 3 displays the ultimate failure outcomes for CFRP screwed joints under varying diameters when both the connected layer thickness and the connecting layer thickness are 5 mm. For the 4 mm diameter, the M4 screw experiences fracture failure due to shear between the two plates. No thread pull-out or damage by the connected layer was observed. This is mainly due to the smaller cross sectional area resulting from the smaller screw diameter. Thus, a larger stress per unit area is applied causing the screw to break more easily. On the contrary, the shear capacity of screws increases with increasing screw diameter. But it is relatively more likely to be pulled out due to the tilt. Screws were also pulled out when the hole diameter was increased to 8 mm. However, the final failure result was due to the bearing failure caused by the M8 screw head on the connected layer. Due to the increased contact area between the screw and the connected plate as a result of increasing the diameter of the screw. This increases the localized compression of the screw on the surrounding material during loading. So the possibility of localized stress concentrations and material damage increases. Therefore, for CFRP screwed joints where the thickness of the connecting and connected layers are equal, screws cause more damage around the hole in the connected layer as the hole diameter increases.

Table 3.

Failure results for different diameter sizes.

Diameter	4 mm	6 mm	8 mm
Connected layer
Connection layer
Result	Screw cutting	Screw pull-out	Bearing failure

Effect of connected layer thickness on failure

Typical load-displacement curves for different thicknesses of connected layers are shown in Figure 4(a), the trend of the curves is an upward and then relatively slow downward process. Due to the gradual nature of the thread pull-out process or the shear failure process of the connected layers, the damage to the material itself is rapid. Therefore, the load drop process is relatively slow and has better ductility compared to the material’s destruction. Figure 4(b) shows the average ultimate load for different thicknesses of connected layers. The joint with a thickness of 7 mm exhibited the highest load-bearing capacity, capable of withstanding a load of 8.83 kN. Compared to joints with thicknesses of 1 mm, 3 mm, and 5 mm, there were increases of 175%, 23%, and 2%, respectively. According to the overall trend of the average ultimate loads, the load-bearing capacity of the CFRP screwed joints increase with the thickness of the connected layer. The increase in connected layer thickness results in a larger contact area. This means that the screw is in contact with more fibers and resin during the stretching process causing the load to increase.

To observe the impact of varying connected layer thicknesses on the tensile properties of CFRP screwed joints, the ultimate failure results of specimens with different connected layer thicknesses are presented in Table 4. The experimental failure results in shear failure of the connected layer when the thickness of the connected layer is 1 mm and 3 mm. Due to the small thickness of the connected layer, its force area in the connection is small. This results in a rapid buildup of shear stresses in the intervening plates when tensile forces are applied. When these stresses exceed the shear strength of the material, shear damage occurs in the connected layers. When the thickness of the connected layer increases to 5 mm and 7 mm beyond the thickness of the connecting layer. Pulling out of the screw becomes the final failure result. Because the connected layer can withstand greater shear stresses when its thickness is increased, it is less likely to be damaged. Threads in the connecting layer do not provide sufficient supporting force, so screws pull out of connecting plates.

Table 4.

Failure results for different connected layer thicknesses.

Layer thickness	1 mm	3 mm	5 mm	7 mm
Connected layer
Connection layer
Result	Shear failure	Shear failure	Screw pull-out	Screw pull-out

While the bearing capacity of CFRP screwed joints increases with the thickness of the connected layer, the ultimate load error bars for the specimen group with a 3 mm connected layer are notably higher than those of other specimen groups. Table 5 details the two types of failure results for this group of specimens. Specimen 6-3-3 shows the screws pulling out after a portion of the connected layer has suffered a shear failure, while the other specimens have suffered a shear failure. The interpretation of these phenomena is that after the connected layer undergoes compression from the screw, the screw begins to press against the 90° fibers in the through-hole plate as displacement increases. The fibers continuously bend until fracture occurs.³⁸ However, during this process, the fibers accumulate under the screw due to bending without fracturing, creating resistance to the movement of the screw. When the accumulation of fibers becomes excessive, the resulting resistance is sufficient to prevent the inclined screw from further tearing through the through-hole plate, prompting the screw to once again undergo screw pull-out until failure occurs.

Table 5.

Two failure results for connected layer thickness of 3 mm.

3 mm specimen	S3-6-1	S3-6-3
Connected layer
Connection layer
Result	Shear failure	Screw pull-out

Analysis of screw pull-out process

For the specimen where screw pullout occurs, it is a staged process. For example, Figures 5–7 shows the load-displacement curve of the specimen with screw pullout and its screw pullout process. The graphs show that after reaching the ultimate load, screw pull-out begins to occur. During the load descent, as the screw screws initially separate from the material of the screwed hole, localized fiber damage or delamination may cause a temporary drop in the load-displacement curve. However, as the screws continue to engage and the next section of the screw supports these damages, the load may rise again. This staged damage process can result in fluctuations in the load-displacement curve.^39,40 During the screw pull-out process, the descent of the load curve noticeably quickens with the increase in connected layer thickness, as a thicker connected layer diminishes the screw’s angle of tilt during pull-out. This limitation of the screw’s inclination by a thicker connected layer suggests an enlarged contact area between the screw and the hole, which concurrently necessitates a higher pull-out force. Once the screw begins to pull out, the greater pull-out force may lead to a more rapid screw failure and pull-out.

As depicted in Figures 5–7, in the process from points a to b preceding screw pull-out failure, the screw becomes inclined due to torque, with the inclination at an angle θ. The screw pull-out process initiates only after point b is reached. During the pull-out process, damage due to local material compression can further enlarge the screw’s angle of tilt, and the clearance between the screws increases as they are extracted, providing the screw with additional room to move. Thus, the angle of screw inclination dynamically changes during the process of screw pull-out failure.

Figure 8(a) summarizes the trend of the screw’s torsional angle change with each screw extracted during the screw pull-out tests. It can be seen that for the specimens where screw pullout occurred, the screw tilt angles at the start of screw pullout (i.e., point b in the figure) were 29°, 21°, and 17° for specimens with connected layer thicknesses of 3 mm, 5 mm, and 7 mm, respectively. It can be observed that as the connected layer thickness increases, the angle of screw inclination at the ultimate load decreases. Increasing the thickness of the connected layer enhances the shear strength of the entire connection because the shear area of the screw increases. A thicker material layer can provide more resistance, thereby reducing the angle of inclination. On the other hand, the thicker connected layer increases the stiffness of the overall structure. The high stiffness resists torsion due to moments and helps to keep the screws horizontal, thus reducing screw tilt.

Figure 8(a) also reveals the dynamic changes in the screw’s angle of inclination during the screw pull-out process. As the screw is progressively extracted, the angle of screw inclination exhibits an increasing trend due to localized damage. However, this trend shows a particular non-linearity in the thinner connected layer, which is somewhat related to the partial destruction of the accumulated 90° fibers in the connected layer. The thickness of the connected layer plays a crucial role in maintaining the angle of screw inclination. Therefore, the stability and durability of the screwed connection can be significantly improved by rationally designing the thickness of the connected layer, which is of great value for improving the performance of CFRP structural connections in engineering applications.

As shown in Figure 8(b) demonstrates the screw tilt angle after failure of specimens with different thicknesses of connected layers. It can be observed from the figure that the screw tilt angle of the 1 mm specimen is much smaller than the other specimens. Because when the thickness of the connected layer is small, the connected layer is deformed by the shear force of the screw. The smaller force inhibits the tilt angle of the screw, so the tilt angle of a screw that experiences a shear failure is less than the tilt angle of a screw that experiences a thread pullout failure pattern. As the thickness of the connected layer increases, the connected layer can withstand greater shear forces and is less prone to shear failure. The screw is subjected to the secondary bending moment and tilted.⁴¹ The contact area of the screw with the connected layer rises with the thickness of the connected layer, resulting in an increase in the resistance during the tilting of the screw to inhibit the tilting of the screw. Therefore, for specimens in which thread pullout occurs, the tilt angle of the screw decreases as the thickness of the connected layer increases.

Digital image correlation-based analysis of principal strain field and out-of-plane displacement field

Effect of change in thickness of connected layer

This section describes in two parts the acquisition of surface principal strains and out-of-plane displacements by digital image correlation (DIC) under different connected layer thicknesses and diameters. In the previous section, it was discussed that when the failure mode is characterized as a screw pullout failure, it is clearly stated that the ultimate load is the critical point at which the screw begins to pull out. The other failure results of the joint are also closely related to the ultimate load, so the strain field and out-of-plane displacement field in this section are recorded when the load reaches the ultimate load. The collection of such data is critical to the behavioral patterns of joints under ultimate loads and has significant academic and meaningful applications for optimizing joint design and predicting their structure.

Figure 9 displays the primary surface strain fields at ultimate load for different connected layers, where shadowing near the screw head has led to a partial omission in the strain field’s capture by the industrial camera. Nevertheless, the data present a pronounced strain gradient around the screw head region. The figure indicates that strains are predominantly localized at the lower part of the hole, corresponding to the load-bearing side subject to the screw shank’s extrusion pressure. Additionally, strains are present at the upper part due to the screw’s tilt exerting extrusion. This tilt results in a tighter screw head connection with the upper connected layer, intensifying friction during tensile loading and contributing to incremental deformation. Post-tensile residual strains and material springback also manifest around the hole.⁴² In the thinner connected layers, the strain below the hole is more acutely affected by the layer’s thickness. The 1 mm and 3 mm thicknesses, strain originates below the hole and extends to the plate’s end. In contrast, the 5 mm and 7 mm thicknesses show a strain that also initiates below the hole but decreases progressively. This strain distribution pattern is deemed rational considering the failure modes observed post-ultimate load. In thinner layers, areas of maximum strain within the field prefigure the location of shear failure. Conversely, with increasing plate thickness, the strain field lessens correspondingly. These findings are crucial for predicting screw connection behavior across varying plate thicknesses and provide insights that are instrumental in refining the design of composite material structures.

Despite the application of in-plane tensile loads, the screwed joint demonstrates notable out-of-plane displacement behaviors, which are attributed to factors such as screw inclination and the potential for screw pull-out.⁴³ Figure 10 illustrates these displacements across specimens with varying thicknesses of the connected layers. Analysis of the failure across four different connected layer thicknesses shows that the specimen with a 1 mm thickness experiences direct shear failure upon reaching ultimate load, with its out-of-plane displacement markedly lessened due to minimized screw tilt when compared to other thicknesses. Consequently, discussions about the 1 mm thick specimen are omitted here. Further analysis indicates that the upper side of the hole displays out-of-plane deformation in the reverse direction, or negative deformation, resulting from the screw head’s downward force. The connected layer exhibits a propensity for tilt, with the lower side of the plate orienting outward and the stretching side inward, a tendency that intensifies as the plate thickness diminishes. This effect precipitates greater out-of-plane displacement in the vicinity of the screw due to its angled position. Notably, with a 3 mm through-hole plate, the out-of-plane displacement at the hole’s lower end peaks at 0.83 mm, signifying increases of 130% and 143% relative to 5 mm and 7 mm thick plates, respectively. Similarly, the upper end registers a displacement of −0.61 mm, surpassing the 5 mm and 7 mm plates by 103% and 117%. These observations indicate that a decrease in connected layer thickness enhances the compressive deformation on both sides of the hole due to screw compression. Furthermore, this suggests a corresponding rise in the screw’s torsional angle, aligning with previous studies findings. Such insights are academically pivotal for elucidating screw connection behaviors under tensile stress and serve as a valuable reference for the refined design of screwed joints in composite material structures.

Effect of changes in diameter size

Figure 11 presents the distribution of the surface principal strain field in specimens with varying hole diameters at ultimate load. The specimen with an 8 mm diameter hole exhibits the highest principal strain, measured at 0.076, which is a substantial increase compared to those with 4 mm and 6 mm diameters. For the 4 mm diameter joint, strain concentration is localized on the lower side of the hole, where the screw shaft exerts compression. As the hole diameter expands to 6 mm and 8 mm, this strain concentration extends to both the upper and lower sides of the hole, becoming especially pronounced at 8 mm, with noticeable deformation on the upper side. The pattern of the strain field is closely linked to the specimens’ failure modes, indicating a transition from screw fracture in smaller diameters to extrusion damage of the connected layer in larger diameters. This trend suggests that deformation adjacent to the hole escalates with increasing hole diameter.

Screwed single lap joints are known to exhibit considerable out-of-plane deformation at ultimate load, even in the presence of in-plane tensile forces. As depicted in Figure 12, this deformation varies with the diameter of the specimens at ultimate load. For a 3 mm diameter, the out-of-plane displacement near the hole is minimal, with the entire connected layer experiencing slight out-of-plane displacement 0.332 mm outward at the lower end and 0.196 mm inward at the upper end. At a 6 mm diameter, the connected layer’s out-of-plane displacement becomes more pronounced due to the torsional compression exerted by the screw, yet the maximum displacements remain at the edges closest to the boundary, at 0.55 mm and −0.405 mm. An 8 mm diameter markedly increases the maximum out-of-plane displacement, especially at the lower side near the hole, which reaches up to 4.16 mm. The failure of the 8 mm specimen under ultimate load is attributed to material extrusion damage and some degree of screw disengagement, leading to the substantial out-of-plane displacement observed around the screw hole, and this displacement aligns with the observed failure pattern.

Conclusions

Compared to traditional bolted connections, due to the absence of nuts on the plate, cosmetic and aerodynamic design requirements can be better met by a screwed connection. Besides, without using the nuts, a screw connection is weight saving and facilitates assembly when only one side of the construction is accessible.²⁷ Hence, in this research, tensile tests were conducted on carbon fiber-reinforced polymer (CFRP) laminates with screws introduced by tapping. An industrial camera provided real-time monitoring, while digital image correlation (DIC) analysis was utilized to assess the strain field and out-of-plane displacements, this study examined the effects of connected layer thickness and hole diameter on the tensile behavior and failure outcomes of composite screwed joints. The findings led to the following conclusions:

(1) Tensile testing revealed that enlarging the diameter in CFRP screwed joints substantially improves their load-bearing capacity. Specifically, a diameter of 8 mm resulted in a load capacity of up to 10.82 kN, which is a 99% and 26% increase over the capacities of 4 mm and 6 mm diameters, respectively. However, the use of an 8 mm and 4 mm diameter is generally not recommended in engineering applications due to the sudden material failure observed in these cases.

(2) When the thickness of the connected layer of CFRP screwed joints is elevated relative to the thickness of the connecting layer, the load carrying capacity is improved, 7 mm compared to 1 mm, 3 mm, 5 mm by 175%, 23%, 2%. Failure results from the connected layer shear failure to screw pullout failure, the tilt angle of the screw in the pullout process will gradually become larger. And with the increase of the thickness of the connected layer, the tilt angle of the screw at the beginning of the screw pullout decreases.

(3) For CFRP screwed joints with different connected layer thicknesses, the principal strain field at the ultimate load increases with the increase of the connected layer thickness. When the failure is a screw pullout, the out-of-plane displacement at the ultimate load decreases with the increase of the thickness, verifying the change of the tilt angle, so the effect of the connected layer thickness should be taken into account in the design.

(4) For CFRP screwed joints with different hole diameters, the strain field increases with increasing hole diameter, gradually spreading from one side of the hole to the perimeter of the hole and increasing as the hole diameter increases. The out-of-plane displacement of the joint surface also increases. Therefore, the effect of the hole diameter on the structure as a whole should be considered in the design.

Footnotes

Acknowledgments

This work was supported by Tianjin 131 Research Team of Innovative Talents (No. 201916), and Tianjin Diversified Investment Project in Applied Basic Research (No.23JCYBJC00100). The authors would like to acknowledge the editors and the anonymous referees for their insightful comments.

Author contributions

Renyu He: Conceptualization, Methodology, Writing – original draft. Tao Yang: Software, Validation. Sinan Liu: Writing – review & editing. Penchao Zhang: Writing - original draft. Chang Liu: Writing – review & editing, Data curation. Wenhui Yuan: Validation, Data curation. Yu Du: Supervision, Data curation.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Tianjin 131 Research Team of Innovative Talents (No. 201916), and Tianjin Diversified Investment Project in Applied Basic Research (No.23JCYBJC00100).

Ethical statement

ORCID iDs

Sinan Liu

Yu Du

Data availability statement

The data and materials used during the current study are available from the corresponding author on reasonable request. Data will be made available on request.*

References

Min

Liu

Cheng

, et al. Efficient manufacturing techniques and connection failure analysis of CFRP lifting lug for main bearing structure. Eng Fail Anal 2023; 151: 107420.

Gao

Liu

, et al. Hole formation mechanisms in double-sided laser drilling of Ti6Al4V-C/SiC stacked materials. J Mater Process Technol 2024; 325: 118307.

Camanho

Erçin

Catalanotti

, et al. A finite fracture mechanics model for the prediction of the open-hole strength of composite laminates. Compos Appl Sci Manuf 2012; 43(8): 1219–1225.

Tao

Chang

. Investigation of tensile property and strain distribution of drilled carbon fiber reinforced polyetheretherkrtone composites. J Compos Mater 2023; 57(12): 1999–2012.

Sajid

Karuppanan

Sallih

, et al. Role of washer size in mitigating adverse effects of bolt-hole clearance in a single-lap, single-bolt basalt composite joint. Compos Struct 2021; 266: 113802.

Catalanotti

Camanho

. A semi-analytical method to predict net-tension failure of mechanically fastened joints in composite laminates. Compos Sci Technol 2013; 76: 69–76.

Manalo

Mutsuyoshi

. Behavior of fiber-reinforced composite beams with mechanical joints. J Compos Mater 2012; 46(4): 483–496.

Tan

Wang

, et al. Experimental investigations of bolted, adhesively bonded and hybrid bolted/bonded single-lap joints in composite laminates. Mater Today Commun 2020; 24: 101244.

Mingming

Xiaoquan

Wenjun

, et al. Damage mode and load distribution of countersunk bolted composite joints. J Compos Mater 2021; 55(13): 1717–1732.

10.

Zhang

, et al. Theoretical simulation and experimental verification of C/SiC joints with pins or bolts. Mater Des 2014; 53: 1071–1076.

11.

Zhang

, et al. Design, preparation and properties of online-joints of C/SiC-C/SiC with pins. Compos B Eng 2013; 48: 134–139.

12.

Chen

Liu

, et al. Evaluation on the interval values of tolerance fit for the composite bolted joint. Compos Struct 2018; 206: 628–636.

13.

Camanho

Matthews

. Stress analysis and strength prediction of mechanically fastened joints in FRP: a review. Composites Part A-Appl S 1997; 28(6): 529–547.

14.

Olmedo

Santiuste

Barbero

. An analytical model for the secondary bending prediction in single-lap composite bolted-joints. Compos Struct 2014; 111: 354–361.

15.

Belardi

Fanelli

Vivio

. Analysis of multi-bolt composite joints with a user-defined finite element for the evaluation of load distribution and secondary bending. Compos B Eng 2021; 227: 109378.

16.

Liu

Yao

Zhao

, et al. An improved 2D finite element model for bolt load distribution analysis of composite multi-bolt single-lap joints. Compos Struct 2020; 253: 112770.

17.

Liu

Fang

Zhao

, et al. A failure-envelope-based method for the probabilistic failure prediction of composite multi-bolt double-lap joints. Composites B 2019; 172: 593–602.

18.

Sun

Liu

Zheng

, et al. On fracture characteristics of adhesive joints with dissimilar materials-An experimental study using digital image correlation (DIC) technique. Compos Struct 2018; 201: 1056–1075.

19.

Timbolmas

Bravo

Rescalvo

, et al. Digital image correlation and numerical analysis of CFRP-poplar timber interface subjected to modified single shear test. Compos Struct 2023; 320: 117188.

20.

Mehrabian

Boukhili

. 3D-DIC strain field measurements in bolted and hybrid bolted-bonded joints of woven carbon-epoxy composites. Compos B Eng 2021; 218: 108875.

21.

Camanho

Lambert

. A design methodology for mechanically fastened joints in laminated composite materials. Compos Sci Technol 2006; 66(15): 3004–3020.

22.

Zhao

Fang

Liu

, et al. A modified stiffness method considering effects of hole tensile deformation on bolt load distribution in multi-bolt composite joints. Composites Part B-Eng 2019; 171: 264–271.

23.

Shan

Zhao

Liu

, et al. Revealing the competitive fatigue failure behavior of CFRP-aluminum two-bolt, double-lap joints. Compos Struct 2020; 244: 112166.

24.

Belardi

Fanelli

Vivio

. Comparative analysis of FE modeling techniques for single-lap multi-column composite bolted joints. In: International conference on chemical science and engineering. Perlis, Malaysia, 10–21 November 2021, Springer Nature Singapore, pp. 121–128.

25.

Belardi

Fanelli

Vivio

. Theoretical definition of a new custom finite element for structural modeling of composite bolted joints. Compos Struct 2021; 258: 113199.

26.

Belardi

Fanelli

Vivio

. A novel composite bolted joint element: application to a single-bolted joint. Procedia Struct Integr 2018; 12: 281–295.

27.

Meram

Can

. Experimental investigation of screwed joints capabilities for the CFRP composite laminates. Compos B Eng 2019; 176: 107142.

28.

Zemann

. Manufacturing of threads direct into a carbon fibre reinforced polymer. Mater Today Proc 2016; 3(4): 1226–1229.

29.

Zakharov

Lyubchenko

Magnitskii

. Carrying capacity of a threaded pair made of spatially reinforced carbon-carbon composite. Russ Eng Res 2021; 41(9): 803–807.

30.

Jeong

Pham

Tran

. Development of predicting equations for slip modulus and shear capacity of CLT-concrete composite with screw connections. J Build Eng 2023; 71: 106468.

31.

De Freitas

Vieira

Filho

SLMR

, et al. Experimental investigation of tapping in CFRP with analysis of torque-tension resistance. Int J Adv Manuf Technol 2019; 104(1-4): 757–766.

32.

Can

Meram

. Dynamic behavior of screwed joints for CFRP composite laminate structures under impact loading. J Manuf Process 2022; 75: 232–242.

33.

Liu

Sabbrojjaman

, et al. Effect of bolt size on the bearing strength of bolt-connected orthotropic CFRP laminate. Polym Test 2023; 118: 107894.

34.

Montagne

Lachaud

Paroissien

, et al. Failure analysis of single lap composite laminate bolted joints: comparison of experimental and numerical tests. Compos Struct 2020; 238: 111949.

35.

Luo

Kong

Jin

, et al. Investigation of the 3D-DIC testing method for composite shell in a deep-water high-pressure environment. Thin-Walled Struct 2023; 190: 110962.

36.

Wei

Hao

Chen

, et al. Rapid assessment of out-of-plane nonlinear shear stress-strain response for thick-section composites using artificial neural networks and DIC. Compos Struct 2023; 310: 116770.

37.

Ekh

Schön

Zenkert

. Simple and efficient prediction of bearing failure in single shear, composite lap joints. Compos Struct 2013; 105: 35–44.

38.

Akbaş

ŞD

. Post-buckling analysis of a fiber reinforced composite beam with crack. Eng Fract Mech 2019; 212: 70–80.

39.

Mahendran

. Screwed connections in built-up cold-formed steel members at ambient and elevated temperatures. J Constr Steel Res 2022; 192: 107218.

40.

Hizam

Manalo

Karunasena

, et al. Effect of bolt threads on the double lap joint strength of pultruded fibre reinforced polymer composite materials. Construct Build Mater 2018; 181: 185–198.

41.

Zhu

Akkerman

Xiong

, et al. Evaluation of insert design on the performance of repaired composite-Ti alloy joints. Compos Struct 2019; 230: 111506.

42.

Zhang

Yang

, et al. An experimental study on mechanical response of single-lap bolted CFRP composite interference-fit joints. Compos Struct 2018; 196: 76–88.

43.

Zhao

Xin

Liu

, et al. Secondary bending effects in progressively damaged single-lap, single-bolt composite joints. Results Phys 2016; 6: 704–711.