Dissimilar Materials Joining of Carbon Fiber Polymer to Dual Phase 980 by Friction Bit Joining , Adhesive Bonding , and Weldbonding

In the present work, joining of a carbon fiber-reinforced polymer and dual phase 980 steel was studied using the friction bit joining, adhesive bonding, and weldbonding processes. The friction bit joining process was optimized for the maximum joint strength by varying the process parameters. Then, the adhesive bonding and weld bonding (friction bit joining plus adhesive bonding) processes were further developed. Lap shear tensile and cross-tension testing were used to assess the joint integrity of each process. Fractured specimens were compared for the individual processes. The microstructures in the joining bit ranged from tempered martensite to fully martensite in the cross-section view of friction bit-joined specimens. Additionally, the thermal decomposition temperature of the as-received carbon fiber composite was studied by thermogravimetric analysis. Fourier-transform infrared–attenuated total reflectance spectroscopy and X-ray diffraction measurements showed minimal variations in the absorption peak and diffraction peak patterns, indicating insignificant thermal degradation of the carbon fiber matrix due to friction bit joining.


Introduction
Achieving lightweight, multi-material auto body structures is a critical goal for the automotive industry to comply with government regulations (i.e., to improve fuel efficiency and reduce greenhouse gas emissions) [1,2].Currently, four types of lightweight materials-high-strength aluminum alloys, magnesium alloys, ultra-high-strength/advanced high-strength steels (AHSSs), and polymer composites (i.e., carbon fiber-reinforced polymers (CFRPs) and glass fiber-reinforced polymers)-have been identified as substitutes for current steel and/or cast iron auto body structures.The selection of lightweight materials should be carefully explored to produce multi-material structures while satisfying the structural stability and safety performance requirements of vehicles.
Although various combinations of the four candidate materials are possible for potential lightweight vehicles, one of the most interesting material combinations is CFRPs and AHSSs.This is because CFRPs have high mechanical strength with a high strength-to-weight ratio [3,4].Additionally, a high-strength dual-phase (DP) steel with good mechanical properties [5,6] could be used as a thinner-gauge substitute in existing vehicle structures, leading to substantial weight reductions.However, because of the physical and chemical dissimilarities of individual materials, the development of suitable joining technologies is essential for enabling multi-material auto bodies.Recently, extensive research has been conducted for joining carbon fiber composites to various metals (e.g., aluminum or magnesium alloys), using laser welding [7,8], friction stir blind riveting [9], friction spot joining [10][11][12], ultrasonic welding [13], adhesive bonding [14], friction lap welding [15,16], friction-based injection clinching joining [17], self-piercing riveting [18,19], and hybrid bonding (adhesive + mechanical fastening) [20][21][22].In addition, very limited work has been focused on joining carbon fiber composites to steel [8,14,23,24].However, the hole clinching and self-piercing rivet processes have been shown to be difficult to apply to AHSSs (σ TS > 780 MPa) [25][26][27].For this reason, no previous study has attempted to join carbon fiber composites to DP980, potentially because of the limitations of the joining mechanism for each individual process.Although adhesive bonding can be used for CFRPs and different grades of AHSSs, weak peel strength and environmental degradation are key concerns [28].Therefore, none of the joining technologies mentioned is an easy or the sole way to join a carbon fiber composite and DP980.
In the present work, the authors applied the friction bit joining (FBJ) process to spot-weld carbon fiber composites to DP980.Initially, various welding parameters were used to achieve the highest lap joint strength for FBJ specimens.Next, adhesive bonding and weldbonding (FBJ + adhesive) processes were further developed for the selected materials.The joint integrity for each process was assessed by lap shear tensile and cross-tension testing.Fractured samples were observed and compared for the individual processes.Vickers microhardness was measured on the cross-sectioned FBJ samples to characterize hardness profile changes due to the evolution of microstructures during the joining process.Optical and scanning electron microscopy were used to analyze the interfaces between the joining bit and the CFRP.Additionally, thermogravimetric analysis (TGA) was conducted to study the thermal decomposition temperature of the as-received CFRP.In a cross-sectioned FBJ specimen, Fourier transform infrared-attenuated total reflectance (FTIR-ATR) spectroscopy and X-ray diffraction (XRD) were used to study the thermal degradation of the CFRP close to the joining bit.

Materials and Methods
For the study, the authors purchased a commercially available carbon fiber composite with 2.0 mm-thick epoxy resin (TB Carbon PN.008, T g = 120 • C) laminate reinforced with 34-36 wt% of unidirectional carbon fibers (CP150NS, TB Carbon, Yangsan-si, Korea), which was used as a top sheet material.The physical properties of carbon fiber are summarized in Table 1.The stacking sequence of the carbon fiber layers followed the configuration [0 • /45 • /90 • ] with four plies.The tensile strength of carbon fiber composites was measured at 655 MPa.A 1.2 mm-thick DP980 was used as a bottom sheet material.Joining bits were fabricated using the American Iron and Steel Institute (AISI, Washington, DC, USA) 4140 alloy steel.The joining bit head was a hexagonal shape with 9.525 mm in width.The shank diameter and length of joining bit were 6.6 and 4.57 mm, respectively.
Nominal chemical compositions and mechanical properties of each material are provided in Tables 2  and 3.Each material was cut and prepared with coupon dimensions of 100 mm in length × 25 mm in width for lap joint specimens.The overlap area for the lap joint configuration was 25 × 25 mm 2 .A cross-tension specimen composed of carbon fiber composite and DP980 had coupon dimensions of 150 mm in length × 50 mm in width [25].The hole size in the cross-tension specimen was 20 mm.The overlap area for the cross-tension specimen was 50 by 50 mm 2 .Before the materials were joined, acetone and isopropyl alcohol were used to remove any dirt and grease on both material surfaces.An STA 449 F1 Jupiter simultaneous thermal analyzer (Netzsch Gerätebau-GmbH, Selb, Germany) was used to investigate the thermochemistry of the as-received carbon fiber composite.Differential scanning calorimetry (DSC) was coupled with thermogravimetry with the aim of determining incipient decomposition temperatures.Disc-shaped specimens taken from the as-received carbon fiber composite, weighing 61.98 mg with a diameter of 6.4 mm and a thickness of around 2 mm, were prepared to fit snugly and flush to the bottom of platinum crucibles supplied by Netzsch (Selb, Germany).During FBJ, the peak temperature in the joining bit can be expected to exceed the austenite transition temperature (A 3 ) of steels [29].For this reason, the maximum temperature was set at 1000 • C for the thermogravimetry measurements.The samples were heated at a constant ramp rate of 5 • C/min to 1000 • C in flowing ultra-high-purity argon gas at 100 cc/min.The change in weight and a microvolt signal, due to the temperature difference between the sample and the reference crucible, were recorded as functions of temperature.
FBJ is a two-stage joining process [30,31].In the first stage (Figure 1b), the joining bit cuts and plunges into the top material using a spindle speed of 2000 rev•min −1 and a plunge rate in the z-direction of 171.5 mm•min −1 .In the second stage (Figure 1c), the joining bit is bonded to the bottom steel sheet by frictional heat and z-direction axial force.Various welding conditions were tested to achieve the highest lap shear strength.After preliminary tests, the total plunge depth by the joining bit and joining feed rate were fixed at 4.32 mm and 171.5 mm•min −1 , respectively.Then, different joining speeds in the range from 1500 to 2500 rev•min −1 were used to achieve the maximum lap joint strength.Figure 2 is a schematic view of the lap joint-configured sample holder, tool holder, joining bit, and clamping parts used for the present work.An STA 449 F1 Jupiter simultaneous thermal analyzer (Netzsch Gerätebau-GmbH, Selb, Germany) was used to investigate the thermochemistry of the as-received carbon fiber composite.Differential scanning calorimetry (DSC) was coupled with thermogravimetry with the aim of determining incipient decomposition temperatures.Disc-shaped specimens taken from the asreceived carbon fiber composite, weighing 61.98 mg with a diameter of 6.4 mm and a thickness of around 2 mm, were prepared to fit snugly and flush to the bottom of platinum crucibles supplied by Netzsch (Selb, Germany).During FBJ, the peak temperature in the joining bit can be expected to exceed the austenite transition temperature (A3) of steels [29].For this reason, the maximum temperature was set at 1000 °C for the thermogravimetry measurements.The samples were heated at a constant ramp rate of 5 °C/min to 1000 °C in flowing ultra-high-purity argon gas at 100 cc/min.The change in weight and a microvolt signal, due to the temperature difference between the sample and the reference crucible, were recorded as functions of temperature.
FBJ is a two-stage joining process [30,31].In the first stage (Figure 1b), the joining bit cuts and plunges into the top material using a spindle speed of 2000 rev•min −1 and a plunge rate in the zdirection of 171.5 mm•min −1 .In the second stage (Figure 1c), the joining bit is bonded to the bottom steel sheet by frictional heat and z-direction axial force.Various welding conditions were tested to achieve the highest lap shear strength.After preliminary tests, the total plunge depth by the joining bit and joining feed rate were fixed at 4.32 mm and 171.5 mm•min −1 , respectively.Then, different joining speeds in the range from 1500 to 2500 rev•min −1 were used to achieve the maximum lap joint strength.Figure 2 is a schematic view of the lap joint-configured sample holder, tool holder, joining bit, and clamping parts used for the present work.An STA 449 F1 Jupiter simultaneous thermal analyzer (Netzsch Gerätebau-GmbH, Selb, Germany) was used to investigate the thermochemistry of the as-received carbon fiber composite.Differential scanning calorimetry (DSC) was coupled with thermogravimetry with the aim of determining incipient decomposition temperatures.Disc-shaped specimens taken from the asreceived carbon fiber composite, weighing 61.98 mg with a diameter of 6.4 mm and a thickness of around 2 mm, were prepared to fit snugly and flush to the bottom of platinum crucibles supplied by Netzsch (Selb, Germany).During FBJ, the peak temperature in the joining bit can be expected to exceed the austenite transition temperature (A3) of steels [29].For this reason, the maximum temperature was set at 1000 °C for the thermogravimetry measurements.The samples were heated at a constant ramp rate of 5 °C/min to 1000 °C in flowing ultra-high-purity argon gas at 100 cc/min.The change in weight and a microvolt signal, due to the temperature difference between the sample and the reference crucible, were recorded as functions of temperature.
FBJ is a two-stage joining process [30,31].In the first stage (Figure 1b), the joining bit cuts and plunges into the top material using a spindle speed of 2000 rev•min −1 and a plunge rate in the zdirection of 171.5 mm•min −1 .In the second stage (Figure 1c), the joining bit is bonded to the bottom steel sheet by frictional heat and z-direction axial force.Various welding conditions were tested to achieve the highest lap shear strength.After preliminary tests, the total plunge depth by the joining bit and joining feed rate were fixed at 4.32 mm and 171.5 mm•min −1 , respectively.Then, different joining speeds in the range from 1500 to 2500 rev•min −1 were used to achieve the maximum lap joint strength.Figure 2 is a schematic view of the lap joint-configured sample holder, tool holder, joining bit, and clamping parts used for the present work.A commercially available adhesive, 3M Scotch-weld Epoxy DP460, was used for the adhesive bonding and the weld bonding (FBJ + adhesive) processes.An applicator with a mixing nozzle that supplies a 2:1 mixing ratio of resin to a hardener was used.A 250 µm glass bead was applied in the joint area to maintain a constant adhesive bond line thickness.Then the adhesive was cured in an oven for 1 h at 60 • C, as recommended by the manufacturer.For weldbonding, FBJ was conducted on a cured specimen with the same welding process parameters.
Lap shear tensile tests were conducted using an MTS hydraulic test frame at a cross head rate of 10 mm•min −1 and at room temperature.Spacers (25 × 25 mm 2 ) were used to grip the samples to prevent bending of the specimen during the lap shear test.The gripping area on the both ends was 25 × 25 mm 2 .Similarly, cross-tension tests were conducted with the same cross head rate as used for the lap shear tensile tests and at room temperature.Each mechanical test was triplicated for each joining process.
For cross-sectional analysis, FBJ samples were cross-sectioned and then mounted in epoxy resin.Then the samples were finely polished using a common metallographic procedure.A 5% Nital solution was used to etch the sample surface to reveal the microstructure.An optical microscope (Nikon Epiphot, Nikon, Tokyo, Japan) and a high-resolution scanning electron microscope (SEM) (JELO 6500 FEG-SEM, Hitachi High-Technologies, Krefeld, Germany) were used to characterize the cross sections of FBJ specimens.A Leco LM100 AT (Leco, Saint Joseph, MI, USA) was used to measure the Vickers microhardness of FBJ specimens with the following conditions: a 500 g load, 13 s of dwell time, and 250 µm spacing.
XRD patterns of a thermally affected carbon fiber composite directly mated with DP980 and the as-received CFRP were recorded using a PANalytical X'Pert Pro-MPD powder diffractometer (Malvern Panalytical, Amelo, The Netherlands) equipped with a Si-based position-sensitive one-dimensional detector and a nickel-filtered copper Kα radiation source.For the measurements, X-ray was generated at 45 kV/40 mA at a beam wavelength of λ = 1.5416 × 10 −10 m (copper Kα radiation).The XRD measurements of the epoxy/carbon fiber multilayer were carried out in the reflection mode; hence, the XRD probed the surface morphology of the outer epoxy layer.
FTIR-ATR spectroscopy (Bruker Optik GmbH, Ettlingen, Germany) with a Bruker Lumos FTIR microscope in the ATR mode was used to probe the surface of the epoxy composite in close proximity to the joining bit.The spectra represented an average of 128 scans with a spectral resolution of 4 cm −1 .Spectra from five different spots with a 300 µm spacing over a 1000 µm distance were obtained using a line scan that originated at the interface between the joint bit and the carbon fiber composite.A spectrum for the as-received carbon fiber composite was also collected as a reference.

Thermogravimetry and Differential Scanning Calorimetry Measurements for the As-Received Carbon Fiber Composites
The results of the thermogravimetry-coupled DSC analysis of the as-received CFRP sample are plotted in Figure 3.The STA 449 F1 instrument and software use the fundamentals of heat-flux DSC to determine both reaction temperatures and the associated enthalpy (∆H rxn ). Figure 3 indicates the as-received CFRP tested is thermally stable up to 290.5 • C. The mass of the sample precipitously dropped immediately after the beginning of a sharp exothermic DSC signal.The sharp exothermic peak could be due to interactions among leftover elements that might occur during thermal degradation of the epoxy.

Lap Shear Tensile Testing and Fractorgraphy
The assembly for lap shear testing of the FBJ specimen is presented in Figure 4a.Lap shear tensile testing was conducted to determine the static mechanical strength of the FBJ joints under optimized welding conditions, as already summarized in Table 4.A summary of the average lap joint failure loads and lap shear strengths for each process is plotted in Figure 5.A maximum lap joint failure load of 6.4 kN was achieved for FBJ specimens.The average lap shear failure load was 6.0 kN when three samples were repeatedly tested.The average lap shear failure loads for adhesive-bonded and weld-bonded specimens were 14.8 and 13.3 kN, respectively.For weld-bonded coupons, a small reduction in the lap joint strength (approximately 10%), compared with the adhesive-bonded coupons, was attributed to thermal degradation of the cured adhesive on the periphery of the joining bit due to frictional heating.However, the lap joint failure load of the weld-bonded specimen was almost twice higher than the failure load for the FBJ-

Lap Shear Tensile Testing and Fractorgraphy
The assembly for lap shear testing of the FBJ specimen is presented in Figure 4a.Lap shear tensile testing was conducted to determine the static mechanical strength of the FBJ joints under optimized welding conditions, as already summarized in Table 4.

Lap Shear Tensile Testing and Fractorgraphy
The assembly for lap shear testing of the FBJ specimen is presented in Figure 4a.Lap shear tensile testing was conducted to determine the static mechanical strength of the FBJ joints under optimized welding conditions, as already summarized in Table 4.A summary of the average lap joint failure loads and lap shear strengths for each process is plotted in Figure 5.A maximum lap joint failure load of 6.4 kN was achieved for FBJ specimens.The average lap shear failure load was 6.0 kN when three samples were repeatedly tested.The average lap shear failure loads for adhesive-bonded and weld-bonded specimens were 14.8 and 13.3 kN, respectively.For weld-bonded coupons, a small reduction in the lap joint strength (approximately 10%), compared with the adhesive-bonded coupons, was attributed to thermal degradation of the cured adhesive on the periphery of the joining bit due to frictional heating.However, the lap joint failure load of the weld-bonded specimen was almost twice higher than the failure load for the FBJ-  A summary of the average lap joint failure loads and lap shear strengths for each process is plotted in Figure 5.A maximum lap joint failure load of 6.4 kN was achieved for FBJ specimens.The average lap shear failure load was 6.0 kN when three samples were repeatedly tested.The average lap shear failure loads for adhesive-bonded and weld-bonded specimens were 14.8 and 13.3 kN, respectively.For weld-bonded coupons, a small reduction in the lap joint strength (approximately 10%), compared with the adhesive-bonded coupons, was attributed to thermal degradation of the cured adhesive on the periphery of the joining bit due to frictional heating.However, the lap joint failure load of the weld-bonded specimen was almost twice higher than the failure load for the FBJ-only coupon, indicating the mechanical joint strength was improved because the load distribution was shared between the adhesive bonding and FBJ.Similarly, Squires et al. [29] observed an improvement in the mechanical joint strength for a highstrength aluminum alloy joined to steel by weld bonding.Di Franco et al. [32] also found increased joint strength when self-pierce riveting was combined with adhesive bonding to join a carbon fiber composite and aluminum sheet AA2024-T6.In addition, the presence of an adhesive layer in the joint area can effectively act as an insulator or a barrier against corrosion media, resulting in improvement in the corrosion resistance [33][34][35].For this reason, weld bonding of carbon fiber composites to lightweight metals may have the highest success rates for protecting against corrosion under environmental conditions and for improving mechanical joint performance.
Because of the different joining processes, joint dimensions, and material combinations, a direct comparison of FBJ, adhesive-bonded, and weld-bonded joint strength with data from the available literature is not possible.For this reason, the maximum lap shear strength for the FBJ coupon was calculated based on the shank diameter (6.6 mm) of the joining bit, yielding a value of 186 MPa.Maximum lap shear strengths of 27.3 and 22.5 MPa were calculated for the adhesive bonding and weld bonding cases, respectively, by using the overlap area (25 × 25 mm 2 ) in the lap joint.This normalized calculation has been used as a common practice [11,36].Table 5 allows readers to compare the results from this study with values for current state-of-the-art techniques for joining polymers to metals.Again, direct comparison of different processes would be not applicable, as previously explained, but general information of various joining techniques can be beneficial to readers.Overall, the results from this study show higher or comparable lap shear strengths, compared with the strengths demonstrated by other joining technologies.Similarly, Squires et al. [29] observed an improvement in the mechanical joint strength for a high-strength aluminum alloy joined to steel by weld bonding.Di Franco et al. [32] also found increased joint strength when self-pierce riveting was combined with adhesive bonding to join a carbon fiber composite and aluminum sheet AA2024-T6.In addition, the presence of an adhesive layer in the joint area can effectively act as an insulator or a barrier against corrosion media, resulting in improvement in the corrosion resistance [33][34][35].For this reason, weld bonding of carbon fiber composites to lightweight metals may have the highest success rates for protecting against corrosion under environmental conditions and for improving mechanical joint performance.
Because of the different joining processes, joint dimensions, and material combinations, a direct comparison of FBJ, adhesive-bonded, and weld-bonded joint strength with data from the available literature is not possible.For this reason, the maximum lap shear strength for the FBJ coupon was calculated based on the shank diameter (6.6 mm) of the joining bit, yielding a value of 186 MPa.Maximum lap shear strengths of 27.3 and 22.5 MPa were calculated for the adhesive bonding and weld bonding cases, respectively, by using the overlap area (25 × 25 mm 2 ) in the lap joint.This normalized calculation has been used as a common practice [11,36].Table 5 allows readers to compare the results from this study with values for current state-of-the-art techniques for joining polymers to metals.Again, direct comparison of different processes would be not applicable, as previously explained, but general information of various joining techniques can be beneficial to readers.Overall, the results from this study show higher or comparable lap shear strengths, compared with the strengths demonstrated by other joining technologies.Figure 6 shows fractography results for different specimens from lap shear tensile testing.For FBJ, the joining bit and DP980 were still intact, indicating good consolidated joining between the joining bit and DP980.The failure mode was found to be shear out-one of the common fracture modes (i.e., net tension, shear out, cleavage, and bearing) for composite joints-in cases, where the distance between the hole edge and the edge of the laminate was small [37].For adhesive bonding, delamination of the carbon fiber composite matrix was observed, indicating good adhesion between the carbon fiber composite surface and the adhesive.Mixed failure modes (i.e., shear out and delamination of carbon fiber composites) were observed for weld-bonded samples, as a result of combining two joining processes (i.e., FBJ and adhesive bonding).
Metals 2018, 8, x FOR PEER REVIEW 9 of 15 Figure 6 shows fractography results for different specimens from lap shear tensile testing.For FBJ, the joining bit and DP980 were still intact, indicating good consolidated joining between the joining bit and DP980.The failure mode was found to be shear out-one of the common fracture modes (i.e., net tension, shear out, cleavage, and bearing) for composite joints-in cases, where the distance between the hole edge and the edge of the laminate was small [37].For adhesive bonding, delamination of the carbon fiber composite matrix was observed, indicating good adhesion between the carbon fiber composite surface and the adhesive.Mixed failure modes (i.e., shear out and delamination of carbon fiber composites) were observed for weld-bonded samples, as a result of combining two joining processes (i.e., FBJ and adhesive bonding).

Cross Tension Testing and Fractorgraphy
Examples of final assembled cross-tension specimens joined by FBJ are shown in Figure 4b.Representative load and displacement curves from cross-tension testing of the individual processes are plotted in Figure 7.For FBJ, the peak failure load was found to be 2 kN; a maximum failure load of 4.82 kN was achieved for the adhesive-bonded specimen.Two peak failure loads (4.51 and 2.59 kN) were found for the weld-bonded specimen because of the nature of the two combined joining processes.Although the peak failure loads for adhesive-bonded (4.82 kN) and weld-bonded (4.51 kN) specimens were similar, different failure modes were seen from the load and displacement curves.The adhesive-bonded sample immediately failed at a displacement distance of 3.35 mm, whereas the weld-bonded specimen failed after reaching the second peak load.This is because the friction bit joint remained after the first failure of the adhesive-bonded coupon.It is worth mentioning that the second peak failure load for the weld-bonded sample was close to the maximum failure load of the FBJ-only sample.
Absorption energy is the ability to absorb energy under the mechanical testing condition, which is important for crash or dynamic performance for autobody structures.Absorption energy for each joining process was calculated by integration of load and displacement curved in Figure 7. Summary of peak failure load, absorption energy, displacement at failure for each joining process is presented in Table 6.Absorption energy of weld-bonded specimen is nearly twice higher than the absorption energy of adhesive bonded case.Therefore, the weld bonding process can provide improved structural safety performance compared with adhesive bonding and FBJ only.

Cross Tension Testing and Fractorgraphy
Examples of assembled cross-tension specimens joined by FBJ are shown in Figure 4b.Representative load and displacement curves from cross-tension testing of the individual processes are plotted in Figure 7.For FBJ, the peak failure load was found to be 2 kN; a maximum failure load of 4.82 kN was achieved for the adhesive-bonded specimen.Two peak failure loads (4.51 and 2.59 kN) were found for the weld-bonded specimen because of the nature of the two combined joining processes.Although the peak failure loads for adhesive-bonded (4.82 kN) and weld-bonded (4.51 kN) specimens were similar, different failure modes were seen from the load and displacement curves.The adhesive-bonded sample immediately failed at a displacement distance of 3.35 mm, whereas the weld-bonded specimen failed after reaching the second peak load.This is because the friction bit joint remained after the first failure of the adhesive-bonded coupon.It is worth mentioning that the second peak failure load for the weld-bonded sample was close to the maximum failure load of the FBJ-only sample.
Absorption energy is the ability to absorb energy under the mechanical testing condition, which is important for crash or dynamic performance for autobody structures.Absorption energy for each joining process was calculated by integration of load and displacement curved in Figure 7. Summary of peak failure load, absorption energy, displacement at failure for each joining process is presented in Table 6.Absorption energy of weld-bonded specimen is nearly twice higher than the absorption energy of adhesive bonded case.Therefore, the weld bonding process can provide improved structural safety performance compared with adhesive bonding and FBJ only.Fractography results for cross-tension-tested individual specimens are provided in Figure 8.For the FBJ specimen, the joining bit and DP980 remained in an intact condition, indicating good metallurgical bonding between them, whereas the carbon fiber composite was pullout.For adhesivebonded specimens, failure modes were observed, such as adhesive failure (i.e., fracture at the interface between the adhesive and the carbon fiber composite) and some delamination of the carbon fiber composite layer, as shown in Figure 8b.Finally, the weld-bonded specimens showed complex failure modes, including carbon fiber composite pullout, some adhesive failures, and some delamination of the composite matrix, as presented in Figure 8c.These mixed failure modes for weldbonded samples are thought to have resulted from the combined FBJ and adhesive bonding processes.

Cross Sectional Analysis of Firction Bit-Joined Specimen
The Vickers microhardness distribution in the joining bit and DP980 (Figure 9a) is mapped in Figure 9b, and microhardness profiles along the x direction with three dashed lines are shown in Figure 9c.The average Vickers microhardness values for the joining bit and DP980 were measured at 322 and 320 HV, respectively.In the joining bit, the Vickers microhardness increased up to 760 HV  Fractography results for cross-tension-tested individual specimens are provided in Figure 8.For the FBJ specimen, the joining bit and DP980 remained in an intact condition, indicating good metallurgical bonding between them, whereas the carbon fiber composite was pullout.For adhesive-bonded specimens, mixed failure modes were observed, such as adhesive failure (i.e., fracture at the interface between the adhesive and the carbon fiber composite) and some delamination of the carbon fiber composite layer, as shown in Figure 8b.Finally, the weld-bonded specimens showed complex failure modes, including carbon fiber composite pullout, some adhesive failures, and some delamination of the composite matrix, as presented in Figure 8c.These mixed failure modes for weld-bonded samples are thought to have resulted from the combined FBJ and adhesive bonding processes.Fractography results for cross-tension-tested individual specimens are provided in Figure 8.For the FBJ specimen, the joining bit and DP980 remained in an intact condition, indicating good metallurgical bonding between them, whereas the carbon fiber composite was pullout.For adhesivebonded specimens, mixed failure modes were observed, such as adhesive failure (i.e., fracture at the interface between the adhesive and the carbon fiber composite) and some delamination of the carbon fiber composite layer, as shown in Figure 8b.Finally, the weld-bonded specimens showed complex failure modes, including carbon fiber composite pullout, some adhesive failures, and some delamination of the composite matrix, as presented in Figure 8c.These mixed failure modes for weldbonded samples are thought to have resulted from the combined FBJ and adhesive bonding processes.

Cross Sectional Analysis of Firction Bit-Joined Specimen
The Vickers microhardness distribution in the joining bit and DP980 (Figure 9a) is mapped in Figure 9b, and microhardness profiles along the x direction with three dashed lines are shown in Figure 9c.The average Vickers microhardness values for the joining bit and DP980 were measured at 322 and 320 HV, respectively.In the joining bit, the Vickers microhardness increased up to 760 HV

Cross Sectional Analysis of Firction Bit-Joined Specimen
The Vickers microhardness distribution in the joining bit and DP980 (Figure 9a) is mapped in Figure 9b, and microhardness profiles along the x direction with three dashed lines are shown in Figure 9c.The average Vickers microhardness values for the joining bit and DP980 were measured at 322 and 320 HV, respectively.In the joining bit, the Vickers microhardness increased up to 760 HV because of the evolution of the microstructure from tempered martensite to fully martensite, as a result of rapid heating and cooling cycle during the joining stage.This result indicates that the peak temperature during the heating cycle was above the A 3 temperature of typical steels [29,30].because of the evolution of the microstructure from tempered martensite to fully martensite, as a result of rapid heating and cooling cycle during the joining stage.This result indicates that the peak temperature during the heating cycle was above the A3 temperature of typical steels [29,30].Figure 10 shows optical and magnified SEM images from the cross-sectioned FBJ specimen.In Figure 10a, consolidated bonding between the joining bit and DP980 is observed without any cracks.Figure 10b,c shows SEM images near the joining bit and carbon fiber composite.At the distance ranging from 0.5 to 1.0 mm away from the edge of the joining bit, the gap between the CFRP and the joining bit was filled by the CFRP matrix.In this region, carbon fibers near the joining bit were redirected by the rotational motion of the joining bit during the joining step, as seen in Figure 10d,e.This redirection of the carbon fibers was also seen in other studies, in which rotation of a rivet occurred [38].The remaining area was the base CFRP material, in which the carbon fibers were aligned as manufactured.
Based on the joint configuration and bonding mechanism, the CFRP matrix closer to the polymer and metal (i.e., joining bit and DP980) interface will be affected by frictional heating generated during the plunge and joining stages, as illustrated in Figure 1b,c.In the plunge stage, the joining bit cuts and plunges into the CFRP, resulting in increasing temperature in the CFRP.Because no outgassing was observed in the plunge stage during the FBJ process, the peak temperature generated in the carbon fiber composite was expected to be lower than the decomposition temperature of 290.5 °C for the composite.
During the joining step, great frictional heat is generated when the joining bit engages the steel substrate with a higher spindle rotational speed and a higher axial plunge load.Two scenarios of frictional heat conduction can be anticipated.First, frictional heat will radially diffuse into the steel substrate from the joining bit.Then, frictional heat will be conducted from the bottom steel sheet to carbon fiber composites when they directly contact the steel substrate, as a result of being clamped to it, as shown in Figure 2.However, the peak temperature of the steel substrate will decrease because of the large thermal mass from the fixture substrate.Figure 6a shows a thermally affected radial area of the carbon fiber composite, which whitened slightly.XRD was used to find any noticeable peak changes in the thermally affected radial areas of the CFRP and as-received carbon fiber composites.10 shows optical and magnified SEM images from the cross-sectioned FBJ specimen.In Figure 10a, consolidated bonding between the joining bit and DP980 is observed without any cracks.Figure 10b,c shows SEM images near the joining bit and carbon fiber composite.At the distance ranging from 0.5 to 1.0 mm away from the edge of the joining bit, the gap between the CFRP and the joining bit was filled by the CFRP matrix.In this region, carbon fibers near the joining bit were redirected by the rotational motion of the joining bit during the joining step, as seen in Figure 10d,e.This redirection of the carbon fibers was also seen in other studies, in which rotation of a rivet occurred [38].The remaining area was the base CFRP material, in which the carbon fibers were aligned as manufactured.
Based on the joint configuration and bonding mechanism, the CFRP matrix closer to the polymer and metal (i.e., joining bit and DP980) interface will be affected by frictional heating generated during the plunge and joining stages, as illustrated in Figure 1b,c.In the plunge stage, the joining bit cuts and plunges into the CFRP, resulting in increasing temperature in the CFRP.Because no outgassing was observed in the plunge stage during the FBJ process, the peak temperature generated in the carbon fiber composite was expected to be lower than the decomposition temperature of 290.5 • C for the composite.
During the joining step, great frictional heat is generated when the joining bit engages the steel substrate with a higher spindle rotational speed and a higher axial plunge load.Two scenarios of frictional heat conduction can be anticipated.First, frictional heat will radially diffuse into the steel substrate from the joining bit.Then, frictional heat will be conducted from the bottom steel sheet to carbon fiber composites when they directly contact the steel substrate, as a result of being clamped to it, as shown in Figure 2.However, the peak temperature of the steel substrate will decrease because of the large thermal mass from the fixture substrate.Figure 6a shows a thermally affected radial area of the carbon fiber composite, which whitened slightly.XRD was used to find any noticeable peak changes in the thermally affected radial areas of the CFRP and as-received carbon fiber composites.Figure 11 exhibits the typical amorphous phase of the epoxy [39].Note that the XRD patterns of the two zones are essentially identical, implying that the thermal degradation is ignorable.Next, heat conduction from the joint bit to the surrounding carbon fiber composite was considered.As mentioned earlier, the peak temperature of the joining bit during the joining step can exceed the A3 temperature, as evidenced by the Vickers microhardness measurements.The high temperature can lead to localized degassing of the polymer matrix adjacent to the joining bit due to the relatively short weld time (~0.8 s) and low thermal diffusivity (1.2-2.0 m 2 •s −1 ) of the carbon fiber composite [40,41].For this reason, the remaining polymer composite was not expected to suffer from   Next, heat conduction from the joint bit to the surrounding carbon fiber composite was considered.As mentioned earlier, the peak temperature of the joining bit during the joining step can exceed the A3 temperature, as evidenced by the Vickers microhardness measurements.The high temperature can lead to localized degassing of the polymer matrix adjacent to the joining bit due to the relatively short weld time (~0.8 s) and low thermal diffusivity (1.2-2.0 m 2 •s −1 ) of the carbon fiber composite [40,41].For this reason, the remaining polymer composite was not expected to suffer from Next, heat conduction from the joint bit to the surrounding carbon fiber composite was considered.As mentioned earlier, the peak temperature of the joining bit during the joining step can exceed the A 3 Metals 2018, 8, 865 13 of 16 temperature, as evidenced by the Vickers microhardness measurements.The high temperature can lead to localized degassing of the polymer matrix adjacent to the joining bit due to the relatively short weld time (~0.8 s) and low thermal diffusivity (1.2-2.0 m 2 •s −1 ) of the carbon fiber composite [40,41].For this reason, the remaining polymer composite was not expected to suffer from frictional heat.A cross-sectioned FBJ sample was subjected to FTIR-ATR spectroscopy to examine the surface of the epoxy composite in close proximity to the joining bit to determine if frictional heat from the joining process led to significant polymer degradation.Figure 12 shows characteristic absorption peaks for epoxy resin systems at five different locations and for a reference layer [42,43].Although there are minor variations in the data, they are qualitatively similar; it can be concluded that if any thermal degradation occurs during the joining process, it is insignificant.
Metals 2018, 8, x FOR PEER REVIEW 13 of 15 frictional heat.A cross-sectioned FBJ sample was subjected to FTIR-ATR spectroscopy to examine the surface of the epoxy composite in close proximity to the joining bit to determine if frictional heat from the joining process led to significant polymer degradation.Figure 12 shows characteristic absorption peaks for epoxy resin systems at five different locations and for a reference layer [42,43].Although there are minor variations in the data, they are qualitatively similar; it can be concluded that if any thermal degradation occurs during the joining process, it is insignificant.

Figure 1 .
Figure 1.A schematic of the friction bit joining (FBJ) process: (a) initial stage, (b) cutting and plunge stage, and (c) joining stage.

Figure 2 .
Figure 2. Schematic view of the lap shear joint-configured sample holder and the clamped parts used in this study.

Figure 1 .
Figure 1.A schematic of the friction bit joining (FBJ) process: (a) initial stage, (b) cutting and plunge stage, and (c) joining stage.

Figure 1 .
Figure 1.A schematic of the friction bit joining (FBJ) process: (a) initial stage, (b) cutting and plunge stage, and (c) joining stage.

Figure 2 .
Figure 2. Schematic view of the lap shear joint-configured sample holder and the clamped parts used in this study.

Figure 2 .
Figure 2. Schematic view of the lap shear joint-configured sample holder and the clamped parts used in this study.

Figure 3 .
Figure 3.The Thermogravimetry (TG) (green) and Differential Scanning Calorimetry (DSC) (blue) signals from thermal analysis of the as-received carbon fiber reinforced polymer (CFRP) sample showing an exothermic reaction accompanied by mass loss at 290.5 °C.

Table 4 .
FBJ welding process parameters for joining of CFRP to DP980.

Figure 3 .
Figure 3.The Thermogravimetry (TG) (green) and Differential Scanning Calorimetry (DSC) (blue) signals from thermal analysis of the as-received carbon fiber reinforced polymer (CFRP) sample showing an exothermic reaction accompanied by mass loss at 290.5 • C.

15 Figure 3 .
Figure 3.The Thermogravimetry (TG) (green) and Differential Scanning Calorimetry (DSC) (blue) signals from thermal analysis of the as-received carbon fiber reinforced polymer (CFRP) sample showing an exothermic reaction accompanied by mass loss at 290.5 °C.

Table 4 .
FBJ welding process parameters for joining of CFRP to DP980.

Table 4 .
FBJ welding process parameters for joining of CFRP to DP980.

Metals 2018, 8 ,
x FOR PEER REVIEW 7 of 15 only coupon, indicating the mechanical joint strength was improved because the load distribution was shared between the adhesive bonding and FBJ.

Figure 9 .
Figure 9. (a) Cross-sectional view of FBJ specimen; (b) Vickers microhardness mapping at cross section of FBJ specimen; (c) Vickers microhardness profile along the x direction.

Figure 9 .
Figure 9. (a) Cross-sectional view of FBJ specimen; (b) Vickers microhardness mapping at cross section of FBJ specimen; (c) Vickers microhardness profile along the x direction.

Metals 2018, 8 , 15 Figure 11
Figure 11 exhibits the typical amorphous phase of the epoxy [39].Note that the XRD patterns of the two zones are essentially identical, implying that the thermal degradation is ignorable.

Figure 10 .
Figure 10.(a) Macrograph of FBJ cross-section view.(b,d) SEM images of the left side of FBJ cross section (c,e).SEM images of the right side of FBJ cross section.

Figure 11 .
Figure 11.X-ray diffraction patterns measured the indicated area of lap joint.Zone A: thermally affected area shown in Figure 6a.Zone B: as-received CFRP.

Figure 10 . 15 Figure 11
Figure 10.(a) Macrograph of FBJ cross-section view.(b,d) SEM images of the left side of FBJ cross section (c,e).SEM images of the right side of FBJ cross section.

Figure 10 .
Figure 10.(a) Macrograph of FBJ cross-section view.(b,d) SEM images of the left side of FBJ cross section (c,e).SEM images of the right side of FBJ cross section.

Figure 11 .
Figure 11.X-ray diffraction patterns measured the indicated area of lap joint.Zone A: thermally affected area shown in Figure 6a.Zone B: as-received CFRP.

Figure 11 .
Figure 11.X-ray diffraction patterns measured the indicated area of lap joint.Zone A: thermally affected area shown in Figure 6a.Zone B: as-received CFRP.

Figure 12 .
Figure 12.FTIR spectra with different locations (colored dots in the image) measured at the interface between the joining bit and CFRP.

Table 1 .
Physical properties of carbon fiber.

Table 5 .
Summary of state of the art for joining polymers to metals.

Table 6 .
Summary of cross-tension testing for each joining process.

Table 6 .
Summary of cross-tension testing for each joining process.

Table 6 .
Summary of cross-tension testing for each joining process.