Analyzing adhesively bonded joints for automotive applications.
Structural adhesives have been used successfully in adhesively bonded automotive joints for more than twenty years, with SMC/SMC and SMC/steel single lap joints most frequently used. Recent experience has shown that the bonding of SMC body panels to a steel frame (pioneered by General Motors) results in stiffening of the vehicle frame and consequent improvements in the quality of the ride. The stiffening is a result of load sharing by the SMC body panels - the loads being transferred from the frame to the panels by the structurally stiff adhesive bond. Studies have also shown that adhesive bonding is the most viable approach to the design of lightweight, stiff car structures. To achieve the maximum benefits of bonded joints, the car designer needs information on adhesive properties, stress distribution, strength, and durability.
Traditional methods of bond characterization have successfully used single lap shear and wedge tests to qualitatively measure bond quality. The lap shear test specimen shown in Fig. 1 is geometrically representative of the lap joints commonly used in automotive applications and is used to predict average lap shear strength. The combinations of maximum peel and shear stresses found in the lap shear test need to be properly interpreted and applied to actual applications. The wedge test is best in providing a qualitative indication of adhesion weakness, if it exists.
In most bonded joint applications, SMC has been shown to fail by delamination in the bond area, and this has become the de facto desired failure mode. However, with tougher and stronger SMC systems and at higher temperatures, the failure can migrate to the bond interface, i.e., adhesion failure or adhesive (cohesive) failure. The current industry philosophy is to consider adhesion failure as unacceptable based on the hypothesis that this is a low energy, fast propagation failure mode. However, fracture mechanics principles suggest that for the same geometry, if the failure loads are the same during propagation, the tendency to propagate would be the same regardless of failure mode. The sole use of delamination failure mode as an acceptability criterion may be conservative and too restrictive.
In order to develop a more detailed and quantitative approach applicable to higher performance design, this article considers some aspects of adhesive properties, analysis of test results, failure criteria, and evaluation of joint durability.
Adhesives for automotive applications have to be low-cost, tough, temperature resistant, readily processable, and able to bond to a variety of untreated or contaminated surfaces. Urethane and epoxy adhesive systems are commonly used.
Some adhesives are viscoelastic and continuously deform when subjected to constant loads (creep phenomenon). Tensile stress-strain data of viscoelastic adhesives are dependent upon specimen geometry and strain rate. Hence, tensile data representation should be standardized for comparison of adhesive response and design options.
Creep characterization. A comprehensive method for adhesive creep characterization is to conduct creep tests and fit the data to a mathematical model. A power law model including nonlinear stress and temperature factors is commonly used. A useful power law model is given by:
e = [[g.sub.s.g.sub.T]D + C (t/[a.sub.s.a.sub.T])sup.n]S
where: e = creep strain; s = applied constant stress; t = time; D, C, n = creep parameters; [g.sub.s, a.sub.s] = stress dependent factors (1 at low stress); and [g.sub.T,a.sub.T] = temperature dependent parameters (1 at room temperature).
The model, which is initially generated using data from a few tests, can then be used to predict the adhesive response for a broad range of stress and temperature conditions and can also be used in finite element and other structural analyses. The disadvantage of the power law model is that it predicts a continuously increasing creep strain with time whereas most materials tend to eventually stop creeping at low stress.
The tensile creep properties of an adhesive were studied using the creep model. Adhesive sheets were cast to a 0.125-in thickness, cut to dog-bone shape, and tested on a creep frame for one hour. Two load and unload cycles were used to mechanically condition the specimen. Because the specimens were thin, two extensometers were used on either side of the specimen to eliminate bending moments, which would occur if only one extensometer was mounted asymmetrically.
The room temperature creep responses at three stress levels are shown in Fig. 2. Numerical analysis was used to determine the creep coefficients. The stress dependent parameters are shown in Fig. 3. Increasing stress results in considerable nonlinearity. The effect of temperature on creep properties is being evaluated.
The creep response is useful in designing the minimum overlap required in single lap joints. An elastic low stress region (see Fig. 1) is desirable to prevent the high stress regions from deforming continuously and eventually leading to stress rupture. When a very low modulus elastomeric adhesive is used, there is uniformly high stress all across the overlap, which leads to continuous creep and eventual creep rupture.
Adhesive yield stress. In aircraft applications, bonded joints are designed such that at constant or fatigue loads, the maximum adhesive equivalent stress is less than the yield stress. In typical adhesives used in automotive applications, this value is not easily identifiable. Creep and recovery data may possibly be used to identify stress and environmental conditions at which adhesive irreversible deformation occurs.
Lap Shear Test
The single lap shear test (ASTM D 1002-72) is widely used to determine bond strength and quality. Nonuniform shear and peel stresses occur as shown in Fig. 1. SMC substrates generally fail in delaminations, though with improved SMC materials, adhesion and cohesive failure are observed.
Delamination or fiber tear is considered desirable. In tests, once peak load is reached, fracture occurs rapidly, independent of fiber tear or adhesion failure. Hence, the shear strength is more important than the failure mode. As to whether adhesion failure mode is an indication of propensity to rapid failure under service loads needs to be determined by testing pre-cracked specimens subjected to stable crack growth.
Stress analysis of lap shear joint. In automotive applications, the maximum shear and peel stresses can occur in combinations different from that in the testing. Hence, it is desirable to be able to relate the two sets of maximum stresses. Also, adhesive properties can affect the magnitude of the maximum stresses that initiate fracture by SMC delamination or other failure modes.
Within the linear elastic range, the Volkersen shear lag analysis can be used to predict maximum shear and peel stresses and the maximum bending stress in the adherend. The analysis requires a bending moment factor, which is a function of test specimen geometry and test machine compliances. A computer printout from a bond analysis software (Fig. 4) shows the input and output data in a lap shear analysis.
The maximum shear and peel stresses predicted for room-temperature SMC/SMC lap shear tests done using different adhesives are shown in Figs. 5 and 6. Strain-gaged specimens were used to measure the bending moment factor. All failures were by SMC delamination. A possible explanation for the decrease in lap shear strength with increasing adhesive modulus may lie in the prediction of higher maximum peel and sheer stresses for higher adhesive modulus, as shown in Fig. 5.
Failure criteria. Failure in the SMC is precipitated by a combination of shear, peel, and bending stresses - the relative magnitudes depending upon test or load conditions. The test results discussed above were used to develop a tentative failure criterion for the particular SMC used in the tests:
F = [[S.sub.b.sup.2]/11180 + [S.sub.p.sup.2]/3610 + [S.sub.s.sup.2]/1930]sup. 1/2
where: F = failure ratio, [is greater than or equal to] represents failure; [S.sub.b] = SMC bending stress; [S.sub.p] = maximum peel stress; and [S.sub.s] = maximum shear stress.
Other failure criteria can be used. More testing with controlled peel/shear combinations needs to be done in order to develop and verify failure criteria. The advantage of such failure criteria is that they can be used for different load/shear/bending stress combinations and also extended to environmental, creep, and fatigue load conditions by using strength reduction factors in the failure criterion expression.
Effect of bondline thickness. The lap shear analysis with an experimentally determined moment factor was used to study the effect of bondline thickness on maximum normal stress (Fig. 7). As the bondline thickness is increased, the load line offset also increases, resulting in higher peel stress, which negates to some extent the benefit of reduced maximum stress concentration. Hence, one cannot merely increase bondline thickness in adhesively bonded joints to increase strength. These results apply to the lap shear test geometry and not necessarily to configurations and loads found in applications.
Wedge and Peel Tests
The wedge test shown in Fig. 8 is widely used as a qualitative measure of adhesion. The test is easy to conduct though material requirements are high. It is good at detecting adhesion weakness at the interface, if it exists, with the adhesion weakness manifesting itself as interface failure mode. Normal stresses are induced in the adhesive and substrate. The peak load at fracture initiation, energy under the load deflection curve, and the amount of fiber tear on the fracture surface are used as quantitative measures of adhesion quality (Fig. 9).
Data scatter has been observed in the peak load and energy measures. The main reason is the structural nonlinearity during the test arising, as shown in Fig. 8, from the large geometric deflection and friction effects. The energy depends on the fracture mode, nonlinearity, and friction. In spite of problems with the quantitative measures, the test is still the best qualitative measure of adhesion at the SMC to adhesive interface.
A possible alternative to the wedge test is the peel test shown in Fig. 10. The test is similar to the double cantilever bending (DCB) test used for the testing of delamination and bonds for aircraft applications. A width-tapered specimen can be used to force constant fracture propagation load conditions. The advantage of width taper is that stable fracture growth occurs at the constant fracture load. This load could provide a direct measure of the fracture resistance.
Peel stress analysis. The peel load for initial fracture, which is governed by the maximum peel stress in the fracture region, is of interest to the designer. An analysis to predict this maximum peel stress was developed using a beam-on-an-elastic-foundation model (Fig. 10). The expression for the maximum peel stress is:
[S.sub.p] = [k.sub.eq.delta]
where: [S.sub.p] = peel stress; [k.sub.eq] = equivalent stiffness of adhesive and SMC elastic foundation; and [delta] = beam deflection (see Fig. 10). The detailed equations for [k.sub.eq] and [delta] are too lengthy to present here. The expression for [delta] is obtained from equations for beams on elastic foundation subjected to moment and shear load.
The analysis was used to predict peel stresses in the initial fracture region of wedge tests, neglecting large deflection geometric effects. The results, shown in Fig. 11, give the load required to generate unit normal stress as a function of adhesive and SMC modulus. Lower adhesive modulus results in higher wedge load to attain a given peel stress, which implies a higher initial fracture strength.
There is need for a convenient peel test for the evaluation of peel strength. Until then, the wedge test will be used as a qualitative measure of adhesion quality.
Durability of Bonded
The designer is concerned about bonded joint durability under creep (constant load) and fatigue load conditions. Aircraft industry experience has been that creep is more critical than fatigue, provided that the maximum adhesive stress does not exceed the yield stress. The reason is that at constant load the adhesive creep deformation can grow until rupture occurs, whereas in fatigue, the reversed cycling prevents cumulative creep deformation. However, a high mean load in fatigue can be detrimental. In automotive joints, cyclic peel/shear stresses can initiate failure in the SMC.
Lap shear fatigue response. Fatigue tests were done on SMC/SMC and SMC/steel adhesively bonded lap shear specimens. Their S-N curves are shown in Fig. 12. Delamination cracks formed and grew in the SMC substrates until failure occurred. The SMC/steel joint is stronger because steel is relatively much stiffer than SMC and helps spread shear transfer throughout the overlap rather than only at the ends. It is worth noting that the strength at [10.sup.6] cycles is as high as 30% of the static ultimate strength and that the failure mode is in the composite and not the bond interface. Failure modes and fatigue strengths need to be obtained at elevated temperatures.
Adhesively bonded joints have proven their worth in automotive applications. Increased use in structural applications will require the use of stress analysis and failure prediction. The lap shear test is relevant and will continue to be used in adhesive evaluation and to generate design data. The use of stress analysis methods and combined stress failure criteria for SMC provides additional information from the lap shear tests.
The wedge test continues to be a good test to detect adhesion weakness. Peel tests combined with stress analysis methods can provide a quantitative measure of joint performance and could resolve the issue regarding the necessity of selecting fiber tear over adhesion failure as the preferred failure mode.
Adhesively bonded joints stand up well to fatigue cycling. At room temperature, failures occur in the SMC substrate and not in the adhesive or interface.
PHOTO : FIGURE 1. In the single lap shear test, nonuniform peel and shear stresses are generated.
PHOTO : FIGURE 2. Room-temperature tension creep response of a typical adhesive.
PHOTO : FIGURE 3. The stress dependent creep parameters are used to fit creep data to a power law model.
PHOTO : FIGURE 4. Computer screen display of single lap shear stress analysis.
PHOTO : FIGURE 5. Stresses predicted by bond stress analysis and lap shear strength as a function of adhesive modulus.
PHOTO : FIGURE 6. SMC bending stresses predicted by bond stress analysis as a function of adhesive modulus.
PHOTO : FIGURE 7. Effect of bondline thickness on maximum stresses in a single lap shear joint.
PHOTO : FIGURE 8. The wedge test.
PHOTO : FIGURE 9. Wedge test load-deflection response and quantitative measures.
PHOTO : FIGURE 10. Double cantilever beam (DCB) geometries for adhesive bond peel test.
PHOTO : FIGURE 11. Effect of adhesive and SMC modulus on wedge loads to generate a given peel stress magnitude.
PHOTO : FIGURE 12. S-N curves for lap shear specimens with maximum stress as percent of static lap shear strength.
|Printer friendly Cite/link Email Feedback|
|Date:||Feb 1, 1990|
|Previous Article:||Recycling polystyrene, add value to commingled products.|
|Next Article:||Additives 1990.|