The art of adhesion
The influence of surface texturing and surface chemistry on adhesive bonding performance
In an earlier edition we had already reported on the adhesive bonding in the automotive industry and the associated challenges regarding the longevity of the bonded joint under corrosive load. Here, the influence of the chemical surface treatment of the final surface was discussed. In this issue, we focus on further factors contributing to an improvement in the bonding performance. The role of the adhesive/substrate interface was reflected by several adhesion theories: mechanical, adsorption, diffusion, and electrostatic theories. However, each approach is more suitable for a specific sample and/or experimental condition than being capable to explain the underlying mechanism in a comprehensive way. In view of this knowledge, the present work on final sheet material was initiated with the aim to examine firstly, the influence of surface texturing (e.g., electrical discharge texturing, mill-finish, etc.…) and secondly, the contribution of surface chemistry on the adhesive bonding performance. The findings from this study strongly indicate the presence of a combined effect of mechanical and chemical surface characteristics on the bonding strength and performance, respectively. In the past considerable efforts were made to gain a deeper understanding of the factors contributing to an improvement of the bonding performance. In this sense the study on fast hardening 6xxx aluminum sheet material equipped with different surface finishing in terms of surface structuring was manufactured and the bonding characteristics were probed.
Theories and mechanisms of adhesion
In the following the most prominent theories and mechanisms of adhesion shall be introduced, whereas each approach addresses different surface characteristics affecting adhesion as illustrated in Figure 1. By far the oldest-established hypothesis is the mechanical interlocking theory proposing that surface irregularities (e.g., cavities, pores…) of the adherend contribute to an increased surface. In this context different shapes of surface irregularities shown in Figure 1a are discussed in literature [1]. Consequently, the wet adhesive penetrates the micro-structured surface and serves as an anchoring after curing. According to this theory the surface topography strongly impacts the adhesive strength. Another mechanism is described within the chemical bonding concept (Figure 1b). Reactive sites within the adhesive and across the surface of the adherend results in the formation of bonds over the entire adherend/adhesive-interface. The force of the created chemical bonds (covalent, ionic, hydrogen and van der Waals bonds) depends on their interaction energies and hence enters the final bonding strength. There are further models explaining the mechanism of adhesion: the thermodynamic, the diffusion and the electrostatic theory. This study was initiated, considering the fact, that none of these theories fully met the demands to explain the phenomenon of adhesion separately.
Samples and surface texturing
After this short excursus in adhesion theories the focus shall be placed on the investigated material and surface texturing, respectively. Sheet material from AA6016 alloy manufactured with elemental limits according to EN 573-3 and a final thickness of 1 mm was employed for this study. Finally, the material was equipped with four differently structured surfaces (Figure 2). In automotive industry commonly a mill-finish (MF) or an electrical discharge texturing (EDT) finishing is applied within the final cold roll pass. Via EDT texturing, micro-structures such as cavities, dimples, or grooves are induced to finally yield a random pattern over the entire surface.
The major benefit of the surface texturing is that the generated isolated pockets enable an improved distribution and storage of the lubricant, which in turn helps to optimize subsequent processes at the customer site such as deep drawing or stretch-forming. However, there are numerous other surface texturing techniques such as Topocrom® (TPC) and Pomini digital texturing™ (PDT) that are fully integrable in the industrial process and therefore are suitable for mass production. 3D-imaging of the investigated surfaces (Figure 2) illustrates the strong differences in their anisotropy, peak density, number of dimples/valley and maximum height differences that are assumed to finally affect the adhesive bonding performance.
Adhesive bonding testing
Probing the adhesive bonding performance requires the following practice: specimen preparation was carried out according to DIN EN 1465: the lap shear specimen consists of two cleaned metal strips with dimensions of 100 mm x 25 mm each bonded together over an area of 250 mm² with an epoxy-based resin. A subsequent curing process completes the specimen preparation. In total ten lap shear joints of this type are required, whereas five parallel samples are taken as a reference and the others undergo an aging under artificial atmosphere. The required parameters of the natural salt spray test (NSS) were adjusted in agreement with DIN EN ISO 9227 and the duration was set to 240 hours. Finally, the results of tensile shear testing of reference and aged samples enable the determination of the reduction in force. This is presented in the following discussion as remaining strength [%].
Results and discussion
In this section, the adhesive bonding performance is discussed with referring to the above mentioned four surface textures. As shown in Figure 3, the tensile shear strength of the reference (unaged) specimens is not greatly affected by the surface texturing or the wetting behavior. On the contrary, taking an exposition to an artificial aging of 240 h into account the data suggest that the surface topography indeed enters the bonding performance. Whereas the TPC and PDT variants reach a remaining strength of almost 80%, the MF and EDT ones significantly exhibit a reduced performance. From these findings the question arises, whether a correlation between surface roughness parameters and bonding performance is conceivable. Consequently, the mechanical interlocking fosters adhesive bonding for this selective adhesive/adherend combination, which in turn opens new possibilities in customization. Next, the focus is placed on the characterization of the textured surfaces. In the framework of standard release testing roughness parameters are determined via tactile measurement: the arithmetic means roughness Ra [µm], the maximum height of profile Rz [µm] and the peak count Rpc [1/cm]. The remaining strength determined from tensile testing is applied to the addressed surface parameters and displayed in Figure 4a and 4b. Despite a notable spread of data, a valuable trend indication for the bonding quality is given by the roughness parameters Ra and Rz reflected by values of 0,87 and 0,92 for the coefficient of determination. Conversely, the peak density Rpc as can be seen on Figure 4b seems not to be beneficial as a tracer for the bonding property. Additionally, the 3D-parameters, the mean dale area Sda [µm²] and the mean dale volume Sdv [µm3] were included since these parameters are useful in the description of the motif-like surface features. In brief, the mean dale area Sda refers to the average area of a closed valley/dale, which exhibits a minimum point and is defined by a fictitious line. The mean dale volume Sdv reflects the average enclosed volume of the valley/dale [2]. Both parameters were obtained from 3D optical microscopy images recorded on a MarSurf CM mobile system. The data in Figure 4c and 4d obviously reveal that no linear relation between the remaining strength and the topographical parameters Sda and Sdv can be expected anymore. Instead, a logarithmic regression seems to be favorable. Considering the coefficients of determinations, the mean dale volume Sdv is the preferred parameter to be included in the discussion of relevant parameters for estimating the bonding behavior. To address the logarithmic relation between remaining strength and Sdv, it can be assumed that for Sdv values below ~300 µm³ a strong influence due to the topography is conceivable, whereas for increasing Sdv values the steep performance tends to flatten
Based on these results, a model describing the bonding behavior due to structuring was developed. A schematic representation of the model is given in Figure 5 and was adapted from [3]. Since the aging in adhesive joints is responsible for the drop in force, a closer look on the mechanism itself is essential: The diffusion of Cl- ions at the adhesive/adherend interface starts (step (1) in Figure 5) followed by the initiation of corrosion (2). As the growth of corrosion products proceeds the (local) displacement of the adhesive is triggered and the gap formation is fostered (3). Finally, the adhesive bond is weakened and a loss in performance is to be expected. Generally, the proposed mechanism of aging is applicable to smooth and rough surfaces. Nevertheless, rough surfaces have some beneficial effects that eventually are helpful towards a better bonding performance. From literature [4] it can be learnt that a rough boundary between adhesive and substrate provides a prolonged path to lateral diffusion. Consequently, the process of initiation and growth of corrosion products is delayed compared to smooth surfaces. Another positive aspect of rough surfaces is the increase in area enabling a higher number of bonds between two material per unit area. Furthermore, structured surfaces provide an improved and an increased density of mechanical interlocking sites [5].
Summarizing the presented results, a clear influence of the surface topography is evident for the investigated adhesive/adherend system substantiating the mechanical interlocking theory. On the other hand, the influence of the surface chemistry was addressed in the beginning of this article post. Taking this aspect into account, one must face the fact that each production step alters the surface conditions and hence, the surface chemistry. In this sense, the chemical pre-treatment after solution heat treatment is of interest. It comprises a pickling procedure to ensure the removal of organic contaminants as well as inorganic species. Subsequently, a rinsing step is carried out. Firstly, to stop the pickling process and secondly, to guarantee the elimination of unwanted residues. In this study, the impact of rinsing on adhesive bonding was investigated in more detail. For that purpose, the rinsing was optimized. The results are shown in Figure 6 for the EDT structured surfaces revealing a drastic increase in remaining strength for the improved variant compared to the initially processed type. For evaluating the differences in chemical conditions, X-ray photoelectron spectroscopy (XPS) was performed on both EDT surfaces. Elemental concentrations were derived from high-resolution data as well as the identification of the chemical groups (Table 1). Additionally, the oxide layer thickness was calculated after Strohmeier [6].
| variant | AI2p metal [at.%] | AI2p oxide [at.%] | C1s C-C/C-H [at.%] | C1s C-O [at.%] | C1s carboxy/carbonat [at.%] | O1s [at.%] | residues [at.%] | oxid layer thickness [nm] |
|---|---|---|---|---|---|---|---|---|
| initial | 3,5 | 12,8 | 51,3 | 30,5 | 1,9 | 4,4 | ||
| improved | 2,8 | 17,3 | 28,2 | 5,3 | 1,6 | 42,0 | 2,8 | 5,5 |
From Table 1 it is shown that both surfaces reveal an oxide layer in comparable thickness, but it is clear to see that the improved rinsing drastically reduces the amount of carbon. In particular, the C-C/C-H contributions that are traced back to unwanted contamination on the surface due to insufficient rinsing.Coming back to the initial question, if the mechanical interlocking or the chemical (bonding) theory has a greater impact on improving the adhesion, a clear statement can be given at least for the introduced material system (epoxide/aluminium): surface texturing and hence, mechanical interlocking can be estimated to play a role, since an increase of ~15% in remaining strength was observed between the EDT and the PDT variant. However, the crucial factor is undoubtedly ascribed to the surface chemistry as the increase in remaining strength was ~30%.
Conclusion
This study was intended to clarify the major factors affecting the adhesive bonding performance of an epoxide/6xxx-aluminium system. Initially, four differently structured surfaces were probed with the result that an increase in roughness, leads to an improved bonding. In this context, the roughness parameters as Ra, Rz as well as the topography parameter Sdv were identified as potential candidates for a correlation with the remaining strength. Furthermore, the observed differences in bonding behavior were elucidated in the framework of a proposed model. These findings suggest a certain contribution of mechanical interlocking on the bonding. In a second step, the elemental and chemical surface conditions were included in the study. Especially the rinsing step, which completes the chemical pre-treatment process, was determined as the major factor in improving the adhesive bonding independently of the surface structuring. The beneficial effect of the rinsing is seen in the removal of unfavorable contaminants and residues from the pickling step. To summarize, both approaches, the mechanical interlocking as well as chemical bonding theory, shall be considered as adhesion properties, even though the major influence can be expected from the surface chemistry.
Customer benefits
Due to increasing customer requirements, the optimization of surface properties is becoming even more important in materials research. The aim is to refine surfaces in functional, and often also decorative terms. The challenges here are enormous as the article clearly states: A product should almost always fulfil several surface properties at the same time, such as joinability, corrosion resistance, lubricant storage capacity, gloss level, etc., for it to be suitable for a single application.
Findings from optimization programs such as this one, can be transferred to any other product groups and thereby serve to improve further surface properties and contribute to not only maintaining but also further increasing the high product quality.
The ongoing development of functional and unique surfaces that go beyond customer requirements, is AMAG's top priority.
References
[1] Naat N, Boutar Y, Naïmi S, Mezlini, da Silva LFM (2021) J Adhes 99:166-258[2] Handbook for the MarSurf CM mobile [3] Scharnagl F, Greunz T, Hafner M (2021) contribution to the 55. Metallographie-Tagung[4] Leidheiser H, Wang W, Igetoft L (1983) Prog Org Coat 11:19-40[5] Cavezza F, Boehm M, Terryn H, Hauffmann T (2022) Metals 10:730[6] Strohmeier BR (1990) Surf Interface Anal 15:51-56
Comments
Add a comment