Experimental study of initial strengths and hygrothermal degradation of adhesive joints between thin aluminum and steel substrates

https://doi.org/10.1016/j.ijadhadh.2013.01.001Get rights and content

Abstract

Growing usage of lightweight materials such as Al and Mg alloys and composites in automotive body manufacturing has come to a point that bonding of dissimilar materials is a realistic problem to address. A significant issue related to the bonding of dissimilar materials is that the differences in substrate surface conditions, substrate chemical and physical properties often lead to bond failure at strength levels far less than the bond strength established by the adhesive manufacturer for a balanced joint. This research experimentally studied several important factors influencing initial shear strengths and hygrothermal degradation of adhesively-bonded single lap shear (SLS) joints. The effects of surface treatments such as lubrication and zinc coating on the substrates were first investigated. It was observed that even a very small change in the amount of lubricant applied to an aluminum alloy affected the initial shear strength. On the other hand, varying the amount of mill oil on a galvanized steel surface had little effect. Next, the comparative study of the initial joint strength between electro-galvanized (EG) steels and hot dipped galvanized (HDG) steels revealed that the two coatings exhibited no difference in terms of the initial strength. Also, various combinations of aluminum alloys and steel substrates were studied to observe the effect of substrate materials. It revealed that the strength of a dissimilar joint constructed of a strong substrate and relatively weak substrate fell below the strength of the like-material joint made of the relatively strong substrate, and was closer to the strength of the like-material joint composed of the relatively weak substrate. Ageing tests were performed on SLS joints at various temperatures with and without humidity. The shear strength barely changed after 60-days of exposure at various temperatures with room humidity, but degraded significantly at high temperature with high humidity.

Introduction

The increasing demand for energy-efficient vehicles has driven automakers to employ more and more lightweight materials such as high strength steel, aluminum alloys and magnesium alloys as well as composites in making automotive body components to achieve a reduced vehicle mass. The usage of mixed materials brings great challenges to the body assembly process. As a method which can be used to join virtually any material, adhesive bonding along with mechanical joining has been greatly favored in the assembly of dissimilar materials, such as metal to composites, in which conventional fusion welds are not applicable. When it is used in bonding two dissimilar metals with different electrical potentials, the adhesive serves as a natural barrier for preventing galvanic corrosion between the two metals. The adhesive bonding is now a mature technique in the automotive industry. Sealing the edges of bondline could be an engineering solution to resist the early weakening of the corrosion of zinc coated steels. Keep using some spot welds or rivets is another method to avoid the joint suddenly collapsing after the ageing. However, for reducing the number of spot welds and evaluating the performances of a used car, the durability of the bonding joints is a worth topic of study.

Many researchers have studied and modeled bonds between dissimilar materials. Early works are mainly motivated by the need in the aerospace industry to bond aluminum alloys and composites, while in the last 20 years studies have focused on automotive applications. Hart-Smith [1] looked at elastic–static stress distribution in balanced single lap shear (SLS) joints and unbalanced joints. He showed that the shear stress distribution along the length of the SLS joint had a pronounced trough, where most of the load was transferred through two effective zones near the ends of the joint, while the bond in the middle was lightly loaded. In the unbalanced SLS joint, there was a severe stress concentration on the weaker substrate compared with the bending stresses developed in a balanced joint. Seong et al. [2] investigated the effect of overlap lengths, substrate thicknesses and substrate materials on the failure load of adhesive bonds between aluminum alloys and composites. The study showed that the thicker the substrate, the higher the strength of a balanced joint, although the joint strength was not linearly proportional to the thickness of the substrate. Da Silva et al. [3] also studied the effect of materials, geometry and surface treatment on the SLS shear strength of balanced joints. For their test conditions, it was concluded that the shear strength increased with the overlap length, the substrate thickness and substrate yield strength; and decreased as the bondline thickness increased. Surface treatment such as coating had no effect. Crocombe et al. [4] developed two different closed form approaches incorporating moisture degradation by the Fickian diffusion model to determine moisture distribution in the joint. This approach currently only incorporated the degradation of the cohesive adhesive properties. Mallick et al. [5] used detailed finite element models to calculate peel, shear and longitudinal stresses in the SLS bonds between similar and dissimilar materials. He revealed that the maximum peel and shear stresses in the joints of dissimilar materials occurred at the lap end where the weak substrate extended. Also, the maximum stresses were close to the weak substrate surface in the bond thickness direction. This explained why the bond between the strong–strong substrates could sustain a higher load than the bond between the weak–strong substrates. As a summary, the previous studies revealed that the stresses were not uniform across the bond but peaked at the two ends in a SLS specimen, and the maximum stresses and bond strength were specific to the material properties of the substrates and their structure stiffness. It was noticed that almost all the above studies were done with cautious surface cleaning of substrates and targeted for aerospace applications. Automotive substrates contain various amounts and types of lubricants and have specific surface coatings such as the zinc coating in steels. The existence of lubricant or blank wash, which is commonly used on panels in automotive stamping processes, and the surface coatings may greatly affect the bonding characteristics and make the above conclusions inapplicable. Also, the bond behavior varies greatly with respect to each adhesive. For example, for an adhesive having a relatively weak strength, the corresponding joint strength may not be sensitive to the thickness of stiff substrates at all. Therefore, it is necessary to understand the effect of surface conditions on the bond strength and failure characteristics, and to conduct a study of the strength of the adhesive bonds with respect to the adhesive of interest and substrates of interest.

Research on temperature and humidity effects on adhesive bonds are found mostly pertaining to bonds between identical substrates. Grant et al. [6], [7] studied the temperature effect on the strength of adhesively bonded mild steel SLS and T-joints. For thin bond lines, the authors observed that the failure load was higher at lower temperatures. It was concluded that the failure criterion developed at room temperature was still valid at low and high temperatures, with the failure envelope moving up and down as the temperature changed. The effect of the humidity has been studied by many researchers as water is a substance which potentially replaces the link between the adhesive and substrate, thereby weakening the bond. Mubashar et al. [8] looked at moisture absorption–desorption effects in adhesive joints between aluminum AA2024-T3 substrates. It was found that the joint strength degraded with the absorption of moisture, but could recover if the moisture was desorbed. Meanwhile, it was noted that the dry joints failed cohesively and the wet joint failed adhesively. Sugiman et al. [9] investigated the effect of moisture on the static response of adhesively bonded single lap joints. The testing results showed that the mechanical properties degraded in a linear way with the moisture content. The residual strength after exposure decreased with increasing moisture content (exposure time) and tended to level off towards saturation. However, it also indicates that the effect of residual stresses due to thermal and swelling strains on the predicted static strength was not significant. Schroeder et al. [10] studied adhesive bonds between thick AA5754 being exposed to 50 °C and 90% R.H. for 200 days. Both lap shear coupons and coach peel coupons exhibited essentially no strength loss after 200-days of exposure. Datla et al. [11] carried out fatigue tests of adhesive-bonded asymmetric double-cantilever-beam aluminum specimens under various temperature and humidity conditions. Results showed that the temperature change under dry conditions had little effect on the fatigue threshold. Kinloch et al. [12] found that the hydration weakening of the oxide is a major failure mechanism for the epoxy-to-anodizing-aluminum alloy joints in an aqueous environment. Lafarie-Frenot et al. [13] found that practically no degradation occurred in environments with relative humidity lower than 60%. Kinloch [14] studied the dependence of the degradation rate on the temperature in a moist environment and found that the rate of degradation increased with increasing temperature. In summary, the aforementioned works show that both temperature and humidity play significant roles in the strength and durability of bonds between identical substrates. However, the study of the behavior of dissimilar material bonds under various temperature and humidity conditions is still limited and specific to the substrates and adhesives being tested. Crocombe et al. [15] studied the interfacial failure of adhesive joints for a range of degradation. A mixed mode interfacial rupture element was proposed with a traction–separation law. The two moisture dependent fracture parameters, fracture energy and tripping traction, were calibrated using a mixed mode flexure (MMF) test and finite element analyses. Liljedahl et al. [16] investigated the long-term durability of adhesively bonded dissimilar substrate joints exposed to humid environments. Failure of the joints was modeled with a cohesive zone model (CZM) approach where the governing parameters were determined from fracture mechanics test specimens saturated in a range of humid environments. In their study, the prediction of the joints using the 2D model was slightly higher than for the 3D model, which was expected as moisture also entered from the sides.

The failure mechanism of aged joints with various substrates has been studied and revealed by many researchers. Baldan [17] and Watts [18] summarized three main failure mechanisms as a result of the exposure of joints to water. They are hydrodynamic displacement of adhesive from substrates, adhesive plasticization, and corrosion at substrates. All these modes of failure were described as interfacial failure in their work, although some were actually cohesive failure due to extremely thin adhesive remaining on the substrates. Kinloch et al. [19] studied the durability of the steel/epoxy/aluminum-alloy dissimilar-substrate joints tested in the ‘wet’ environment. The results showed that degradation did not arise from residual stresses but from the occurrence of additional cathodic corrosion occurring in the dissimilar-substrate joints in water. Lu et al. [20] studied the static strength degradation of galvanized Dual Phase (DP) 600 steel joints in hot-wet condition. It was concluded that plasticization and crazing of the adhesive, as well as zinc oxide formation were responsible for static strength reduction. Kinloch et al. [21] studied the fracture energy, Gc, of metallic joints bonded with a rubber-toughened epoxy adhesive in various relative humidity and in water. At relatively low crack velocities, the value of Gc measured in the aqueous environment was relative low compared to that measured in relatively dry environment. At high crack velocities, the value of Gc was relatively high and independent of the environment. In the case of the steel–aluminum alloy joints, a corrosion process was seen to play a role in failure, and a cathodic disbondment mechanism occurred, which was leading to a predominantly interfacial failure.

There are many test standards and methods [22], [23], [24] to measure the strength of adhesive joints or assess their durability. To simulate service conditions of a car, ageing tests have often been performed under a combination of several standards. The tests consisted of periods of low temperature, periods of high temperature and high humidity, and a salt spray period. Examples are the VW P-1200 test and the climate test according to the German VDA 621-415 used in the automotive industry. However, Adams [25] concluded that there was no simple method of determining the long-term durability from accelerated tests, and that excessive temperature and humidity would trigger degradation mechanisms which were not representative. The extent and characteristics of environmental degradation in bonded joints cannot be generalized—they depend upon the specific bonded material system of interest. In general, accelerated tests tend to overestimate the reduction of joint strength. Adams argued that no accelerated procedure for degrading materials correlates perfectly with actual service conditions.

The research being reported is the study of several important factors influencing initial shear strengths and hygrothermal degradation of adhesively-bonded joints between aluminum alloys and galvanized steels. It explored how factors such as lubrication, surface coatings and substrate material properties affected initial bond strength between dissimilar thin substrates in automotive applications, and explains the mechanism of bond degradation when exposed to various constant environmental conditions. Its target application was the bonding of an aluminum skin panel to a steel skin panel or a steel frame. The steel could be either mild steel or high strength steel depending on the actual application. A good example would be the joining of an Al roof panel to a steel body side outer. Single lap shear (SLS) coupons, very representative of the bonding configuration between the roof and the body side outer, were adopted to evaluate apparent shear strength of the joints, where the apparent shear strength was defined as the pulling force divided by the bond area. In the following sections, the effects of surface lubrication and zinc coatings on initial shear strengths were experimentally studied. Also, various combinations of aluminum alloys and steel substrates were investigated to observe the effect of substrate materials. Finally, ageing tests were performed on SLS joints at various temperatures with and without humidity to study the mechanism of joint degradation.

Section snippets

Material

A single-part, heat-curing, crash-toughened epoxy structural adhesive was employed throughout this study. Based on its technical data sheet, the bulk tensile strength of the adhesive can reach beyond 30 MPa. There is only one adhesive used in the whole study. This adhesive can be spot-welded through and adhere to a wide variety of clean or oily material surfaces, so it is very fit for bonding metals in the body shop. It is one component heat curable epoxy based high performance impact resistant

Effects of surface treatments

Effects of surface treatments were investigated from two aspects: surface lubrication and surface coating. The amount of lubrication varies from sheet to sheet in automotive production; therefore, the effect of the lubrication was studied by applying various amounts onto the substrates to observe the influence on the joint strength. Next, the joint strength of substrates with either electro-galvanized zinc coating or hot dipped galvanized zinc coating were studied to compare the effect of

Conclusions

In this paper, the effect of lubricant amount and coating type on the initial strength was studied using the crash-toughened-structural-adhesive-bonded SLS joints among AA6111-T6 sheet, DP590 HDG sheet, GMW2 sheet with HDG or EG coating. It was concluded that the amount of mill oil on steel sheets caused very little change in the bond strength, but the more lubricant on the aluminum surface, the weaker the strength was. Therefore, the amount of lubricant on Al panels should be carefully

Acknowledgments

The authors would like to acknowledge J.C. Ulicny, J.F. Cantalin Jr., S. Hartfield-Wunsch, G. Song, K. Corbin and N. Irish of General Motors Company for their constructive advice during this study. The support provided by the Ministry of Education of China for Fan Zhang's visit to GM in conducting the documented adhesive bonding research is gratefully acknowledged.

The project was also sponsored in part by collaborative research project no. 2010THZ02-2 between Tsinghua University and General

References (32)

  • J.F. Watts

    Role of corrosion in the failure of adhesive joints

  • Z. Lu et al.

    Effect of moisture content in uncured adhesive on static strength of bonded galvanized dp600 steel joints

    Int J Adhes Adhes

    (2011)
  • R.D. Adams et al.

    Influence of local geometry on the strength of adhesive joints

    Int J Adhes Adhes

    (1987)
  • L.J. Hart-Smith

    Adhesive-bonded single-lap Joints, NASA technical report

    NASA CR

    (1973)
  • A.D. Crocombe

    Incorporating environmental degradation in closed form adhesive joint stress analyses

    J Adhes

    (2008)
  • P.K. Mallick et al.

    Analysis of single lap adhesive joints between magnesium and other structural automotive materials

    SAE Int J Mater Manuf

    (2011)
  • Cited by (54)

    • Structural monitoring of adhesive joints using machine learning

      2023, Advances in Structural Adhesive Bonding, Second Edition
    • Durability and accelerated characterization of adhesive bonds

      2023, Advances in Structural Adhesive Bonding, Second Edition
    View all citing articles on Scopus
    View full text