In uence of Surface Roughening of the Raw Material on the Lap Shearing Strength and Failure Behavior of Adhesively Bonded Aluminum Joints


 Surface roughening of the substrates before bonding plays a significant effect on improving the mechanical performance of the adhesively bonded joints, which are prevalently used for light-weighting vehicle bodies. In this study, the influence of surface roughening on the lap shearing strength and failure behavior of adhesively bonded aluminum sheet joints was investigated. Sandpaper grinding was employed for surface roughening, methods such as tensile testing microstructure observation, etc., was employed for evaluating the performance of the joints. The results showed that the lap shearing strength of adhesively bonded joints increased and then decreased with the surface roughness of the aluminum substrate. The maximum shearing strength of the joint bonded with grinded substrates was 30.4 MPa which was improved by 57.5% compared to that produced with un-grinded substrates. However, over roughening is harmful. When the surface roughness was too large, the failure mode of the joint turned from the mixed failure mode to interfacial failure mode, which decreased the strength of the joints. Related mechanisms were demonstrated. When the substrate surface was coarsened, the bonding area and the wettability of the adhesive on the surface were both increased, which promoted the beneficial mechanical interlock effect between the adhesive and the substrate. However, when the substrate surface was over roughened, defects such as voids, insufficient infiltrating of the adhesive, etc., were induced, which apparently increased the proportion of the interfacial failure area.

1 Introduction 8 Adhesively bonded aluminum structural parts got potential applications in recent 9 years in the design of advanced transportation tools, such as automotive, aircraft and 10 marine applications. Their exceptional performances include light-weighting, sealing, 11 sound insulating, high stiffness and corrosion resistance, etc. Compared to welding, 12 riveting and bolting, the residual stress and micro-cracks in the adhesively bonded joints 13 are much less. The operability and efficiency of the adhesive bonding technology are 14 much higher. For example, it can be operated in a narrow space without sophisticated 15 equipment and can be fully cured along with the paint baking process which is very 16 cost-effective. Therefore, the performances of such bonded and multilayered joints have 17 attracted great attention [1][2][3][4][5]. 18 The mechanical strength and failure behavior under diverse service conditions are 19 the most concerning issues for the adhesively bonded aluminum joints, especially when 20 they are used in vehicle bodies. The dangerous points should be known when engineers 21 design and verify the safety of the structure. They are mainly depended on the cured 22 aluminum substrate and the steel substrate were 2.05 μm and 1.98 μm. The maximum 1 strength of the adhesively bonded joints was 4.97 MPa and 6.78 MPa [12]. These 2 studies help engineers to realize that the relevance between the lap shearing strength of 3 the adhesively bonded joints and the surface roughness of the raw materials is not 4 linear. Nevertheless, the related mechanism of this phenomenon is still indistinct. 5 Since surface roughening of the substrate to some extent plays an active effect in 6 strengthening the mechanical behaviors of the adhesively bonded joints, researches 7 were further conducted to study the influence of the surface morphology. A quantity of 8 surface treatment methods was attempted to obtain different surface status, including 9 sandpaper grinding, plasma etching, sandblasting, chemical etching, anodizing, etc. 10 Through these, different types of grooves were obtained, which were found to be 11 effective in reducing stress concentration and improving the adhesion quality by 12 preventing entrapped air at the interface. Using sandpapers with different mesh 13 numbers, Ghumatkar et al found that the surface roughness for aluminum substrates 14 could be achieved in a wide range by the various grinding direction and duration [12]. 15 Da Silva et al verified the influence of grooves and scratches on the joint strength with 16 brittle adhesives and ductile adhesives comparatively. The results showed that scratches 17 or grooves produced by grinded aluminum substrate were harmless and even increase 18 interface force [13]. Wong  Hz Nd-YAG pulsed laser and generated an aperiodic concentration ring with constant 20 depth and diameter [14]. The results demonstrated that the bonding quality of the 21 adhesive joints was promoted apparently. Among the above methods, mechanical 22 technology, sandpaper grinding, for example, is still considered to be optimal for large- 1 scale industrial production, in view of the operability, efficiency, and economic cost. 2 Clarifying the mechanisms for the strengthening of the interface of the bonded 3 metal joints through surface treatment of the raw material is crucial for the processing 4 parameters designing. In the study carried out by Ghumatkar et al, they hold the opinion 5 that the improved adhesion quality resulted from mechanical interlocking between the 6 adhesive layer and metal surface [12]. Dong-Jun Kwon et al. investigated the influence 7 of the surface roughness on the interfacial adhesion strength of joints. Experimental 8 results showed that the mechanical anchoring effect of the joints was enhanced by 9 increasing the roughness of the aluminum surface [15]. J.P.B. van Dam et al. 10 investigated the influence of surface roughness on the interfacial adhesion of epoxy 11 adhesively bonded steel joints through different surface treatments. They revealed that 12 surface roughening has a significant impact on the strength in the joints tensile test and 13 the improved shear strength mainly resulted from the increased interfacial bond area 14 [16]. Even though, regarding the mechanisms, the consensus has not been reached. The 15 reason is the adhesive layer itself and the bonding surface are always the weakest part. the substrate surface. Surface chemical modification has also been proved to be an 9 available way to improve the bonding strength because the bonding state during the 10 adhesively bonded joint preparation process is involved. Surface roughening provides 11 the beneficial mechanical interlocking between the aluminum substrate and adhesive, 12 but over roughening would reduce the wettability of the substrates thus the strength of 13 the joints was weakened. Voids possibly occur at the bonding interface in this condition, 14 which accelerates the crack propagation and lowers the strength of the interfacial force. surface roughness was produced on the bonding surface of aluminum sheets by grinding 5 with sandpapers of different mesh number. The surface roughness, morphology and free 6 energy, wettability of the adhesive on the aluminum substrates with or without 7 roughening, lap shearing strength and interface failure models of the adhesively bonded 8 joints under shear loading were measured and compared. The contact angle test 9 equipment was used to investigate the relationship between the surface roughness and 10 surface free energy. The influence of surface roughness, microcosmic area and surface 11 free energy on the lap shearing strength was examined. The related mechanisms for the 12 failure behavior of adhesively bonded joints were systematically discussed. 13 2 Materials and methods 14 15 A commercial 6063 aluminum alloy in the state of the extruded sheet was used as 16 the raw material, which has excellent mechanical performance and is widely used in the 17 field of automobile manufacturing. Specimens with sizes of 100 mm in length, 3 3 Sandpaper grinding was adopted as the surface treatment of the aluminum sheets, 4 which were subdivided into six groups to investigate the mechanical behaviors of the 5 adhesively bonded joints for tensile testing, as showed in Table 1. Sandpaper mesh 6 numbers of P80, P180, P320, P600, and P800 were used. The sheets were grinded in a 7 single direction that vertical to the tensile load direction of the adhesive joints for tensile 8 testing, as showed in Fig. 1. The grinding time for each sheet was 10 minutes. In Table   9 1,Sixteen sheets were prepared for each group, and six sheets were used for the test 10 microcosmic surface area, surface free energy and surface roughness respectively, and 11 the remaining ten sheets were bonded into five single lap joints. The surface 12 characteristics were measured at ambient humidity of 42 ± 3% and room temperature of 13 25 ± 1 °C.

Raw materials
14 Table 1 The surface treatment of the raw material samples for adhesive bonding 15 Un-grinded P80 P180 P320 P600 P800 16  For each sample, 5 points as shown in Fig. 1 were selected for testing, and each point 7 was tested for three times. Three specimens of each parameter group as shown in Table   8 1 were measured, and the average value was regarded as the ultimate result. The surface 9 morphology characteristics and micro surface area of samples were obtained using a 10 VHX-2000C testing machine.

11
To accurately obtain the failure pattern and the interface failure area, the failure  Eq. (3) can be introduced: or Eq. (4): While the diiodomethane and the water contact angles on the samples were tested 2 at the room temperature, the components γ p s and γ d s of the aluminum substrates can be 3 counted according to the Owen-Wendt equations Eqs. (5) and (6): 14 Table 2 The components and its surface free energy of the measured liquids at room 15 temperature, in mN/m.

16
Testing liquids was beyond its melting point, were also tested for comparison. The same testing and 20 calculating methods as described above were adopted. The only difference was that the 1 testing platform was constantly heated and kept at the testing temperature. adhesively bonded joints were cured for half a minute in a drying oven at 180 ± 0.2 °C. 12 Lap shearing test which is a widely used joint test for automobile parts [27], was surface of all test joints was analyzed after failure. grinded simples (S-0), respectively. The R a and R z of specimens decrease as the mesh 1 number of sandpaper increases. In general, the smaller the sandpaper mesh number is, 2 the larger the particle size will be. When the sandpaper mesh number turned from P80 3 to P800, the average variation of R a and R z increased 1.54 μm and 9.81 μm, 4 respectively. 13.74 μm is 26.7 Mpa, which is 12.2 % lower than that of the joints made of the sheets 10 whose surface roughness is 6.08 μm.    As shown by Fig. 4, the picture shows that surface roughness significantly affects 11 the lap shearing strength of the adhesive bonded joints significantly. When the surface 12 roughness of the grinded aluminum substrates is 6.08 μm, the lap shearing strength is 13 57.5 % higher than that of the joint made of the sheets whose surface roughness is 2.88 14 μm. The reason is that when the aluminum substrates were grinded by sandpaper, 15 grooves, ridges and other uneven flats were produced on the substrate surface, as shown 16 in Fig. 5(d). It prevents the adhesive from detaching the aluminum substrates. In this 17 way, surface roughening improves the lap shearing strength of joints greatly. The 18 roughening grade also significantly affects the lap shearing strength of the joints. When 19 the aluminum substrate surface roughness made of the sheet is 6.08 μm, the lap shearing 20 strength is 30.4 MPa, which is 9.7 % higher than that of the joints made of the sheet 21 whose surface roughness is 3.93 μm. The reason is that when the aluminum substrates 22 were grinded by the coarse sandpaper, so that several hollows were formed on the 1 aluminum substrates surface. The grooves become wider and deeper, and the ridges 2 become wider and higher (illustrated in Fig. 5(b) and (d)). The greater the height 3 difference between the grooves and the ridges is, the more difficult to separate the 4 adhesive from the aluminum substrates will be [23]. It was found that the grinded  The anchor behavior mechanisms between the substrates and the adhesive indicate 3 that the grinding surface morphologies have a significant effect on the shearing strength.

4
As demonstrated in Fig. 6 and Fig. 7, surface roughening obviously increases the 5 interfacial area ratio. As a result, the number of bonding points at the contact area 6 between the adhesive and aluminum substrates increases remarkably. In addition, a 7 large number of anchors allow the adhesive to permeate the surface of the aluminum 8 substrate, and then the adhesive is interlocked with the substrate in the groove during 9 curing. Because the grinding process modifies the amount of interaction between the 10 substrates and adhesive, surface roughening increases the mechanical interlocking effect. 11 In Fig. 6, the fluctuation of un-grinded samples profile curve is lower than that of the  Fig. 6. It is obvious to see 23 that the cross-section of Fig. 7(a) shows a nearly straight line, and the cross-section of 1 Fig. 7(d) shows a sawtooth shape. The difference of two types of cross-sections leads to 2 different surface roughness. Combined with Fig. 4, the lap shearing strength of the 3 bonded joints made of grinded sheets is higher than that of joints made of un-grinded 4 sheets. This is related to the stronger anchor effect of the surface morphologies 5 originated from larger contact areas [13]. With larger anchor sizes, it is much more 6 difficult for the adhesive to separate from the aluminum substrates. Analyzing the other 7 samples were grinded by sandpaper, the same effect can be summarized when the 8 surface roughness values are below 6.08 μm.   Fig. 8 shows the change of microcosmic surface area for grinded sheets compared 5 to un-grinded sheets. It is obviously examined that the microcosmic surface area value 6 of samples grinded by sandpaper is apparently higher than that of un-grinded samples. 7 The microcosmic surface area of un-grinded sheets is 3.2×10 5 μm 2 . The higher surface 8 roughness of aluminum substrates is, the larger microcosmic surface area will be. 9 Besides, when the surface roughness is 3.93 μm and 13.74 μm, its microcosmic surface 10 area increment is 3 × 10 5 μm 2 and 19 × 10 5 μm 2 , respectively. These results indicate that 11 surface roughening can increase the microcosmic area of the aluminum substrate. 12 Furthermore, combined with Fig. 4, it is clear that the lap shearing strength of grinded 13 samples with a rougher surface is greater than that of un-grinded samples when the R z is 14 between 2.88 μm and 6.08 μm. This is related to that surface roughening improves 15 effective contact areas in grinded samples. Due to the existence of grooves and ridges 1 on the substrate surface, there is more micro surface area for bonding. Therefore, the 2 larger the effective contact area is, the weaker the stress concentration will form, and the 3 adhesively bonded joints are difficult to be damaged [13]. Similar trends were observed 4 in average roughness and microcosmic surface area when the R z is between 2.88 μm and 5 13.74 μm, which proves that the enhanced adhesion caused by mechanical grinding 6 resulting from the increasement of surface area. However, the lap shearing strength of 7 the joint is not a linear relationship with the microcosmic surface area.  11 In order to find out the reasons why the lap shearing strength of the joint does not 12 increase monotonically with the surface roughness increasing, the interfacial adhesion 13 situations of the joints with grinded aluminum substrates surface were investigated. 14 Wettability is an important index to evaluate interfacial adhesion situations, which can 15 help engineers to comprehend the interaction mechanism between the adhesive and 16 substrate. Wetting behavior of the adhesive on the aluminum sheets with different surface morphology, and roughness was analyzed based on the method reported in the 1 literature [28]. Fig. 9 shows the average water contact angles and diiodomethane contact 2 angles on the aluminum substrate surface after grinding with the different number mesh 3 sandpaper, respectively. As the surface roughness of aluminum substrate increases, the 4 surface contact angle decreases firstly and then increases gradually in the test liquid 5 using both water and diiodomethane. Furthermore, when the surface roughness of the 6 aluminum substrate is 2.88 μm, the water contact angle is 67.21°, and then it decreases 7 dramatically when the surface roughness turns from 2.88 μm to 6.08 μm, and further 8 drops to 42.25° when the surface roughness increases to 6.08 μm. After that, when the 9 surface roughness is 13.74 μm, the water contact angle grows quickly to about 53.94°.

10
The changing trend of the diiodomethane contact angle is the same as that of water. 11 When the surface roughness of the aluminum substrate is 2.88 μm, the diiodomethane 12 contact angle is 47.18°, It drops to 27.97° when the surface roughness increases from 13 2.88 μm to 6.08 μm. After that, the diiodomethane contact angle grows quickly to about The contact angles of the 1840C structural adhesive at 180 °C on the aluminum 6 substrate with different surface roughness values are shown in Fig. 10. It can be found 7 that the changing trend of adhesive contact angle is consistent with that of water and 8 diiodomethane. When the surface roughness increases from 2.88 μm to 6.08 μm, the 9 adhesive contact angle decreases from 66.2° to 55.4°. Meanwhile, the wettability of the 10 aluminum substrates increases with the surface roughness increasing. The fact is that the 11 interfacial area both in the substrate and adhesive is largely increased. When the surface 12 roughness of the substrate reaches to 6.08 μm, the aluminum substrate has the best 13 wettability. The infiltration of the adhesive on a metal substrate was affected by many 14 factors, for example, the geometry of the dimples, the surface free energy and the 15 rheology of the adhesive. For the adhesively bonded aluminum joints, even if the oils 16 and organic contaminants of the substrate surface are removed, the spreading of 17 adhesive on the aluminum substrates surface is difficult because the surface free energy of the substrates is very low, Therefore, the adhesive cannot infiltrate every corner of 1 the aluminum substrate surface.  Fig. 11 shows the influence of the 7 surface roughness on the surface free energy and its component calculated according to 8 Eq. (1) and Eq. (2). From these results, it seems that the surface free energy of the  2.88 μm is 45.89 mJ/m 2 which are 14.8% and 29.5% lower than those when the surface roughness is 3.93 μm and 6.08 μm, respectively. Meanwhile, the surface free energy of 1 the sample with a surface roughness of 13.74 μm is 56.69 mJ/m 2 , which is 12.9% lower 2 than that of the sample when the surface roughness is 6.08 μm. Thus, when the surface 3 roughness of aluminum sheets is 6.08 μm, it has the highest surface free energy. As   Table 3 and Fig. 11. The simplified zero diffusion 14 pressure is the premise for calculating the adhesion work in ambient air using Eq. (3). 15 Because the work of adhesion, W a , determines the status that adhesive penetrating into 16 the substrate, the bonding work has a significant impact on the diffusion of the adhesive. 17 When the bonding joints are heated to 180 ° C, this temperature is also the curing 18 temperature of the adhesive. The adhesive will form a specific contact angle on the The work adhesion increases with increasing surface roughness in the range from 5 2.88 μm to 6.08 μm. One factor improving the lap shearing strength is that the   Fig. 11. Surface free energy and its components for different aluminum substrates after grinding 2 3 The failure mode of adhesively bonded joints after the tensile test usually includes    It indicates that the surface roughening reduces the percentage of interface failure 12 significantly and enhances the lap shearing strength. In addition, the smaller the 13 interface failure percentage is, the stronger the adhesively bonded joints strength will 14 be. When the surface roughness is 6.08 μm, the adhesively bonded joints have the 3.5 Influence of over roughening on the failure behavior of the joints 5 As illustrated in Fig. 4, the lap shearing strength decreases apparently when the 6 surface roughness turns from 6.08 μm to 13.74 μm. In another word, over roughening 7 surface of raw materials is harmful to the interfacial adhesion state of joints cross-  Fig. 15(a) shows the mechanism diagram of the bonding status of the coarse 17 interface, indicating the samples grinded by sandpaper with the P80 and P180, whose 1 surface roughness is greater than that of the other three treatment methods and un- valid bond area and create stress concentration at the interfacial region. As a result, 12 when the aluminum substrate surface is grinded by P80 or P180, the surface is coarse, 13 and a lot of underfill areas of the joints will be formed, and an alternating solid-liquid 14 and gas-liquid interface will also gradually appear as shown in Fig. 14 and Fig. 15(a). It can be seen from Fig. 15 that the penetration of the epoxy on the grinded 6 aluminum surface has two different modes. Fig. 15 (a) illustrates the distribution of the 7 epoxy on the coarse substrate with surface roughness, for example the sheets grinded by 8 P80 and P180, while Fig. 15 (b) and (c) depicts the distribution of epoxy on the sheet 9 with moderate and fine roughness, representing the cross-section of adhesive joints 10 made of sheets which were grinded by P320, P600 and P800 and un-grinded sheets  It means that the adhesive is not entirely contacted with the aluminum sheets. The lap 10 shearing strength of adhesive joints with over roughening surfaces is very low because 11 of many internal voids and other defects in the joints. roughness of the aluminum substrate increased, the lap shearing strength of the joints 20 initially increased and then decreased. When the surface roughness of the aluminum substrate reached to 6.08 μm after roughening with sandpaper, a maximum lap shearing 1 strength of 30.4 MPa was obtained for the joint, improved by 57.5% compared with that 2 of the joint made by sheets with surface roughness of 2.88 m without roughening. 3 Beyond that, the lap shearing strength decreases with the roughness of the substrates 4 decreasing. 5 (2) The mechanism for the change of the lap shearing strength with the surface 6 roughness is related to the modification of the mechanical interlocking effect, and the 7 microcosmic area, and wettability of the adhesive on the substrate. When the surface 8 roughness of the aluminum substrate ranged from 2.88 m to 6.08 μm, the lap shearing 9 strength of the adhesively bonded joints was mainly increased by increased 10 microcosmic surface area and improved mechanical interlocking. When it is in the 11 range of 6.08 μm to 13.74 μm, the strength of the joints was mainly affected by the 12 wettability by which the surface energy is obviously decreased. Over roughening would 13 worsen the wettability of the aluminum substrate which leads to a weakened lap 14 shearing strength of joints.