The effect of tightening torque on the strength of composite-metal hybrid joints

In the manufacturing and assembly process of aircraft composite-metal hybrid structures, tightening torque is one of the important factors that control assembly quality and affect connection strength. This paper experimentally studies the mechanical properties of composite-metal single-bolt single-lap joints under different tightening torques, reveals the impact mechanism of different tightening torques on the manufacturing and assembly of composite-metal hybrid structures, and also establishes a finite element analysis model of hybrid connection structure considering the influence of drilling and assembly processes. The results show that the ultimate load of the composite-metal connection structure increased with the increase of tightening torque. When the tightening torques were 3 N·m, 6 N·m, and 9 N·m, the ultimate failure loads of the composite-metal hybrid joints were 12.9 kN, 13.8 kN, and 14.6 kN respectively.


Introduction
Composite materials are widely used in the aerospace field due to their advantages such as high specific strength, specific modulus, good fatigue performance, good corrosion resistance, strong structural designability, and integrated molding.As the proportion of composite materials used in aircraft structures increases, composite and metal hybrid connection structures will inevitably be used in aircraft structure design.In aircraft structures, hybrid structures usually use mechanical joints, and the nonlinear changes in structural strength caused during the manufacturing and assembly process will complicate the strength prediction of hybrid structures [1].
Ireman et al. [2] generated a design allowable value chart based on the method of applying computers to predict the extrusion strength value under non-uniaxial extrusion bypass load.The stress distribution in the thickness direction of the laminate near the nail holes was obtained by using a three-dimensional finite element model.Since interlaminar stress was three-dimensional (3D) and always occurred near the free boundary of composite materials, a 3D finite element model was required to accurately design the stress distribution of bolted joints.
Based on the finite element model, the authors [3] can predict the relative displacement between the two sides of the specimen, but the experimental measurement results of the displacement were quite different from the FEM results.In addition, the strain energy measured through experiments can intuitively reflect secondary bending.Failure load occurs when the strain-load curve becomes nonlinear.Differences in experimental testing may be due to deviations caused by non-linearity and friction effects between the two plates of the sample.The result showed that the finite element model was in good agreement with the experimental strains.Another conclusion was also drawn that the 3D finite element 2 model is an effective method to study the thickness direction performance of laminate joints.
Sun et al. [4] studied the effect of clamping area on connection stiffness and failure load for singlenail double-shear laminate (T800G/3900-2 carbon/epoxy) bolted connection structures, and compared and analyzed three different connection forms.The results showed that the failure load of bolted connections with and without washers was higher than that of pin connections without laterally restrained pin connections.In general, the failure load of the connector increased with the increase of the clamping area.Park HJ [5] conducted a study on five layups of mechanically connected (pinconnected and bolted) carbon/epoxy (HT145/RS122) laminates.When the surface layer of the laminate was a 90 o layer, it helped to increase the laminated extrusion strength; if the surface layer was a 0 o layer, it reduced both the final extrusion strength and the laminated extrusion strength.
McCarthy MA et al. [6] conducted a large number of experiments to study the effect of gap amounts (0%, 1%, 2%, 3%) on the stiffness and extrusion of single-nail (including convex and countersunk) single-lap composite connection structures.Wei JC et al. [7] studied the effects of different interference fit sizes on bearing strength and fatigue life through finite element analysis and experiments.
Gray et al. [8] proposed a global modeling method for composite bolt connections in large-scale modeling of aircraft and other applications.Liu FR et al. [9] also studied the mechanical properties of single-nail single-lap and single-nail double-lap joints under the secondary bending effect.They found that because the contact area of single-shear joints and laminates decreases faster than double-shear joints during load application, therefore single-shear joints will fail faster than double-shear joints.
Therefore, this paper carries out experimental research on typical hybrid connection structures considering different tightening torques during assembly, and based on the comparison between simulation results and experimental results, the correction and calibration of the finite element simulation model can be achieved, which can be used for composite-metal single nail and single shear hybrids.These works can provide the experimental basis and technical support for the design of composite-metal single-bolt and single-shear hybrid connection structures.

Specimens and experimental procedures
The tensile test of the composite-metal hybrid structure adopts the composite-aluminum alloy single bolt single lap connection method and designs different tightening torques as test variables to conduct static strength tests to obtain the hybrid structure connection strength.The specimens are shown in Figure 1.A static strength tensile test without the influence of tightening torque is carried out under dry environmental conditions at room temperature.The test methods, procedures, measurements, and data are carried out according to the ASTM D 5961 standard.First, the specimen was installed on the bending clamp to eliminate the secondary bending effect during the stretching process, and an extensometer was used to measure the deformation of the bearing hole.Then they assembled the specimen and clamped it into the chuck of the testing machine, adjusted the clamping position to ensure that the specimen was parallel to the loading block of the lower chuck of the testing machine, tightened the upper chuck, and placed one side of the sample close to the lower chuck on the limit strip to ensure axial loading of the specimen.Finally, the position of the lower chuck was adjusted to align with the 40 mm long dash line of the sample and clamped.The test status after installation is shown in Figure 2.

Results and discussion
The failure morphology of composite-metal single-lap structure specimen with different tightening torques is shown in Figure 3.The failure mode of the bearing hole was mainly a shearing failure of the composite laminate.The failure location was mainly around the bolt holes of the composite laminate.The cracks in the laminate extended along the 45° direction.This was because during the tensile process, the specimen was subjected to shear force, and the contact area between the bolt and the hole was small, resulting in a serious stress concentration phenomenon.At the same time, the failure strength of aluminum alloy was much greater than that of composite materials, so the composite laminate first reached the failure strength and broke.Therefore, due to the anisotropy of the composite laminate and the setting of the plying direction, the cracks generated when the laminate fails and damage extended along the 45° direction.4.During the test, the 3 N•m moment group sample suffered premature bolt failure during the loading process, causing the curve to suddenly drop the load.This could be due to the small preload force applied to the sample and the nut failing to fully compress the washer.This caused the bolt to undergo bending deformation during the stretching process, reaching its limit bearing load in advance and causing fracture failure.In addition, the 6 N•m and 9 N•m moment group specimens had larger pre-tightening force and the bolts did not undergo severe bending deformation.Therefore, their load-bearing capacity and tensile displacement were larger, and their main failure mode was the shearing failure of composite laminates.
The curves were two inflection points in Figure 4. First, before the first inflection point, the curve was in the linear growth stage, and the tensile load increased rapidly with the increase of displacement.This was due to the composite material around the connecting hole was extruded and deformed under the combined effect of pre-tightening and bolted connection.The increase in material density led to an increase in stiffness in the area around the hole.In addition, the tangential friction between the connecting plates of two different materials also shared part of the external load, so the slope of the curve was larger before the first inflection point.As the tightening torque increased, the slope of the curve also rose slightly at this stage.
Secondly, between the second inflection point and the first inflection point, the slope of the curve became smaller and the load growth rate slowed down.This was due to the forced extrusion of the materials around the holes of the laminate, and the compressive stress in the bearing area gradually increased, causing composite small-scale local damage to the material, and material properties deteriorated, resulting in a gradual decrease in structural stiffness, as well as slight sliding between the plates.The external load between the upper and lower plates also gradually decreased, so the slope of the load curve decreased significantly.At the same time, as the tensile displacement increased, the tensile load also gradually increased.This was due to the different characteristics of CFRP laminate materials and traditional metal materials.Due to the multi-phase structural material characteristics of composite materials, the damage to the laminate was a gradual degradation process, and the failure of the weak point would cause the load to be redistributed.Therefore, a wider area around the hole would help disperse the external load and improve the structural bearing capacity.
Finally, as the damage area expanded, it ultimately affected the structural bearing capacity until it reached the limit strength of the structure, which was the second inflection point in the curve.This was because the specimen had been seriously damaged during the loading process.As the damage to the specimen continued to expand, the load would drop sharply until the loading of the testing machine was terminated.In addition, the structure took a long time and had a large displacement from damage degradation to finally reaching the ultimate bearing capacity.On the one hand, the elastic deformation of the composite-metal single-lap joint was relatively large.On the other hand, under the action of the IOP Publishing doi:10.1088/1742-6596/2720/1/0120545 secondary bending moment, the contact area between the hole and the bolt had experienced a process of full contact around the hole-local contact-full contact again, and the generated local displacement would be relatively large.
The laminate was meshed into an eight-node linear hexahedral reduction integral unit (C3D8R), in which the mesh around the bolt-hole was refined.Since the secondary element was not very accurate in simulating contact problems, a linear element was used.However, when subjected to bending loads, linear fully integrated elements were prone to shear self-locking problems due to the rigidity of the elements, which can be solved by linear reduced integration elements such as C3D8R.At the same time, the hourglass problem was also introduced, which would lead to severe mesh distortion in the model, so hourglass control should be carried out to ensure the accuracy of the results.For simplifying the model and reducing the contact pairs, one end of the hexagonal head bolt was simplified to a circular end with the bolt envelope circle, and the eight-node linear reduction integration unit (C3D8R) was still used.The mesh around the bolt hole was refined, and 1152 units were arranged along the hole circumference, with a total of 18432 elements and 9696 bolt units.A meshing diagram of the hybrid joint model is shown in Figure 5.As shown in Figure 6, the tightening torque caused tensile stress in the thickness direction of the bolt-hole, and the composite material under the washer produced compressive stress, and with the increase of tightening torque, the stress peak also increased.The compression effect could effectively improve the load capacity of the structure.In the loading process, the compression effect could enhance the friction between the contact surfaces, and the friction force could carry a large part of the load, at the same time the damage was mainly concentrated in the bolt-hole contact area, Therefore, the diversion of friction to the load can effectively delay the damage initiation time of the material.To further analyze and study the effect of different tightening torques on the load-bearing capacity and tensile strength of composite/aluminum alloy single-shear structures, a finite element analysis model of hybrid structures considering different tightening torques was established.The relative interference between the bolts and the hole circumference in the model was 0%, and the tightening torque is set to 3 N•m, 6 N•m, and 9 N•m according to the test.Combining the test and finite element analysis results (as shown in Table 1), when the tightening torque was 3 N•m, the experimental value was 12.9 kN and the simulation value was 16.5 kN; when the tightening torque was 6 N•m, the experimental value was 13.8 kN and the simulation value was 16.6 kN; when the tightening torque was 9 N•m, the experimental value was 14.6 kN and the simulation value is 16.7 kN.According to the simulation results (as shown in Figure 7), it could be seen that the numerical simulation limit load was larger than the test limit load.This was because the simulation process was based on numerical simulation under completely ideal conditions, and its structural damage was smaller than the actual assembly process.Therefore, the numerical simulation results of the hybrid connection structure were high.Figure 7 shows the load-displacement curve of the static strength test under numerical simulation.Comparative analysis with the test showed that in the growth stage of the load curve, a higher tightening torque corresponded to a larger curve growth slope.That is, under the same tensile displacement, the greater the tightening torque of the specimen, the greater the tensile load it bears.On the one hand, the clamping force made the upper and lower plates of the joint contact each other, and the tangential friction force on the contact surface between the plates effectively shared part of the tensile load.The clamping force effectively inhibited the initiation and proliferation of cracks in the area around the hole, improving the structural load-bearing capacity.On the other hand, the tightening torque provided "lateral support" to the connection structure, which reduced the impact of secondary bending effects and improved the structural stiffness in the area around the hole.However, the effect provided by the tightening torque did not increase infinitely.The strength of the composite material along the thickness direction was mainly provided by the resin matrix, which was much smaller than that of the carbon fiber material.Therefore, an excessive tightening torque would crush the composite material before it was subjected to tensile load, causing damage and failure and reducing the bearing capacity of the specimen.

Conclusion
To obtain the effect of assembly factors on the static strength of hybrid structures, experiments, and finite element simulation analysis were conducted to consider the influencing factors of different tightening moments.The ultimate load of the composite/aluminum alloy connection structure increased with the increase of tightening torque, and the maximum failure loads were 12.9 kN, 13.8 kN, and 14.6

Figure 1 .
Figure 1.Composite-metal single-bolt single-lap connection specimens.According to the specimen condition adjustment requirements in the ASTM D 5961 standard, the specimen was placed under laboratory-standard environmental conditions for 48 hours.Environmental chambers, humidifiers, and other equipment are used to maintain the laboratory temperature at laboratory standard environmental conditions (temperature: 23±2℃; relative humidity: 50±10%).

Figure 2 .
Figure 2. Static strength test of composite-metal hybrid single-bolt single-shear connection.A static strength tensile test without the influence of tightening torque is carried out under dry environmental conditions at room temperature.The test methods, procedures, measurements, and data are carried out according to the ASTM D 5961 standard.First, the specimen was installed on the bending clamp to eliminate the secondary bending effect during the stretching process, and an extensometer was used to measure the deformation of the bearing hole.Then they assembled the specimen and clamped it into the chuck of the testing machine, adjusted the clamping position to ensure that the specimen was parallel to the loading block of the lower chuck of the testing machine, tightened the upper chuck, and placed one side of the sample close to the lower chuck on the limit strip to ensure axial loading of the specimen.Finally, the position of the lower chuck was adjusted to align with the 40 mm long dash line of the sample and clamped.The test status after installation is shown in Figure2.

Figure 3 .
Figure 3.The failure mode of specimens.

Figure 4 .
Figure 4. Load-displacement curves of single-shear joints under different tightening torques.The static tensile load-displacement curves of the composite metal single-lap joint under different tightening torques are shown in Figure4.During the test, the 3 N•m moment group sample suffered premature bolt failure during the loading process, causing the curve to suddenly drop the load.This could be due to the small preload force applied to the sample and the nut failing to fully compress the washer.This caused the bolt to undergo bending deformation during the stretching process, reaching its limit bearing load in advance and causing fracture failure.In addition, the 6 N•m and 9 N•m moment group specimens had larger pre-tightening force and the bolts did not undergo severe bending deformation.Therefore, their load-bearing capacity and tensile displacement were larger, and their main failure mode was the shearing failure of composite laminates.The curves were two inflection points in Figure4.First, before the first inflection point, the curve was in the linear growth stage, and the tensile load increased rapidly with the increase of displacement.This was due to the composite material around the connecting hole was extruded and deformed under the combined effect of pre-tightening and bolted connection.The increase in material density led to an increase in stiffness in the area around the hole.In addition, the tangential friction between the connecting plates of two different materials also shared part of the external load, so the slope of the curve was larger before the first inflection point.As the tightening torque increased, the slope of the curve also rose slightly at this stage.Secondly, between the second inflection point and the first inflection point, the slope of the curve became smaller and the load growth rate slowed down.This was due to the forced extrusion of the materials around the holes of the laminate, and the compressive stress in the bearing area gradually increased, causing composite small-scale local damage to the material, and material properties deteriorated, resulting in a gradual decrease in structural stiffness, as well as slight sliding between the plates.The external load between the upper and lower plates also gradually decreased, so the slope of the load curve decreased significantly.At the same time, as the tensile displacement increased, the tensile load also gradually increased.This was due to the different characteristics of CFRP laminate materials and traditional metal materials.Due to the multi-phase structural material characteristics of composite materials, the damage to the laminate was a gradual degradation process, and the failure of the weak point would cause the load to be redistributed.Therefore, a wider area around the hole would help disperse the external load and improve the structural bearing capacity.Finally, as the damage area expanded, it ultimately affected the structural bearing capacity until it reached the limit strength of the structure, which was the second inflection point in the curve.This was because the specimen had been seriously damaged during the loading process.As the damage to the specimen continued to expand, the load would drop sharply until the loading of the testing machine was terminated.In addition, the structure took a long time and had a large displacement from damage degradation to finally reaching the ultimate bearing capacity.On the one hand, the elastic deformation of the composite-metal single-lap joint was relatively large.On the other hand, under the action of the

Figure 5 .
Figure 5. Finite element model diagram of a composite-metal hybrid connection structure.As shown in Figure6, the tightening torque caused tensile stress in the thickness direction of the bolt-hole, and the composite material under the washer produced compressive stress, and with the increase of tightening torque, the stress peak also increased.The compression effect could effectively improve the load capacity of the structure.In the loading process, the compression effect could enhance the friction between the contact surfaces, and the friction force could carry a large part of the load, at the same time the damage was mainly concentrated in the bolt-hole contact area, Therefore, the diversion of friction to the load can effectively delay the damage initiation time of the material.

Figure 6 .
Figure 6.Simulation of the ultimate loads at different tightening torques.Table 1.Comparison of test and simulation results under different tightening torques.

Figure 7 .
Figure 7. Numerical simulation results under different tightening torques.Figure7shows the load-displacement curve of the static strength test under numerical simulation.Comparative analysis with the test showed that in the growth stage of the load curve, a higher tightening torque corresponded to a larger curve growth slope.That is, under the same tensile displacement, the greater the tightening torque of the specimen, the greater the tensile load it bears.On the one hand, the clamping force made the upper and lower plates of the joint contact each other, and the tangential friction force on the contact surface between the plates effectively shared part of the tensile load.The clamping force effectively inhibited the initiation and proliferation of cracks in the area around the hole, improving the structural load-bearing capacity.On the other hand, the tightening torque provided "lateral support" to the connection structure, which reduced the impact of secondary bending effects and improved the structural stiffness in the area around the hole.However, the effect provided by the tightening torque did not increase infinitely.The strength of the composite material along the thickness direction was mainly provided by the resin matrix, which was much smaller than that of the carbon fiber material.Therefore, an excessive tightening torque would crush the composite material before it was subjected to tensile load, causing damage and failure and reducing the bearing capacity of the specimen.

Table 1 .
Comparison of test and simulation results under different tightening torques.