Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
10 SUMMARY OF DISCUSSION: MECHANICS An interdisciplinary approach to adhesives research is gaining wide acceptance. This approach requires researchers from the fields of chemistry, materials science, and mechanics to work closely together to develop a basic understanding of the adhesion process. However, there exist many disciplines within the mechanics area that must work hand-in-hand to make effective progress. Nondestructive testing, stress analysis, finite element analysis, material modeling, experimental testing, and structural design are a few of these specialized disciplines. Within the mechanics area there are two research goals: design and basic understanding. Although these two goals are not in conflict, they do, in certain problem areas, require two somewhat different approaches to the problem. Basically, design research is oriented toward the application of adhesives to structures. This approach recognizes that every aspect of the adhesive problem cannot be incorporated into the design criteria because of the need to keep the design criteria as simple as possible. It is hoped that the major concerns will be accounted for and that a safety factor will be applied to cover the unknown. The design approach would, in general, consider the adhesive joint to be a single system and would be concerned with the response of that system as a whole. In contrast, the basic understanding approach is more scientific in nature, with emphasis placed on developing a complete understanding of each component of an adhesive joint and their interactions. However, such research is not easy; indeed, much of it, although very important, must be considered long-term. As one outcome, the workshop participants identified the need for standard reference materials; these do not necessarily have to be chemically well-characterized materials, as long as the properties are reproducible. Such materials would allow intercomparison of the results of different laboratories and clearer interpretation of the effects of design and application variables. It was agreed that this goal is attainable, even though care must be used to minimize the effect of aging during adhesive storage. 35
36 The following sections address five areas of general mechanics research: nondestructive evaluation, test methods, stress analysis, finite element analysis, and failure mechanics. Each section contains a short descriptive paragraph followed by a mention of specific research needs. NONDESTRUCTIVE EVALUATION The current state of the art in NOT of adhesive joints is such that most physical defects such as porosity, debonding, and voids can be found in a laboratory environment. The detection and interpretation of defects may require more than one test technique. The size of the largest flaw not found must be used with a fracture mechanics approach to determine critical loads as a function of defect location. A good NDE program requires the combination of NOT to find defects and a physical mathematical model (e.g. , fracture mechanics) to evaluate the significance. Therefore, NDT is used for quality control as well as to establish input data for design. At least five topics in the area of design were identified as calling for research. A long-standing problem is establishing the quality (strength) of a bond. Developing an NDE procedure to determine good or bad bonds should be regarded as potentially long-term. More recent interest has developed in automating NDT techniques for application to large-scale manufacturing and to adapting NDT techniques to field inspection. Pro- cedures for determining critical flaws must be done in conjunction with stess analysis, considering possible failure mechanisms. Similarly, work is also needed on the development of methods of quality control during manufacturing. Two topics in the area of basic understanding were emphasized as needing research. These were NDE techniques to study material degradation and damage growth in adhesive joints, and the development of methods to evaluate surfaces and adhesives separately prior to bonding. TEST METHODS Many specimens and test methods for evaluating adhesives already exist. In spite of this, there is a severe shortage of reliable basic material property data such as G, E, v, GIC, and GUC. Furthermore, recent developments in stress analysis techniques (i.e., finite element methods) have revealed that popular specimens, such as the single-lap shear, are quite complex and their results very misleading when used in a traditional manner. It is important for experimentalists and stress analysts to work together to design new test methods and specimens and to reevaluate current testing procedures. New test methods and specimens are needed to measure specific basic material properties. In addition, durability tests on adhesive joints, meaning resistance to fatigue, creep, exposure, etc., need to be developed. The specimens must be well analyzed and understood. They must be suited for environmental testing and, if possible, should be representative of an application. Also needed are new, small test specimen designs for adhesive material evaluation and screening.
37 STRESS ANALYSIS Currently, accurate stress analysis of adhesive joints is limited by the lack of reliable data on materials properties for the adhesive, the interphase region, and, in some cases, the adherends. Adhesive properties are, for the most part, time-, temperature-, and humidity-dependent. Thus the environment as well as the load plays a significant role in stress analysis. For design research, the stress analysis and accompanying failure criteria must be kept simple and efficient yet be accurate enough to be covered by a reasonable factor of safety. The basic understanding approach requires a more exacting knowledge of the material properties and stress distributions throughout the joint. When a given aspect of the adhesive joint (e.g., viscoelasticity) is found to have a significant effect on the joint stress distribution, it must be integrated into the design procedure. One of our deficiencies is the lack of a model to describe the stress-strain relation throughout the joint as a function of load, time, temperature, and humidity. It is recognized that such a model would be â¢ material dependent. The use of a scrim cloth effectively turns the adhesive into a composite. It would be desirable to have the capability of treating the adhesive as an orthotropic material. It would be desirable, also, to treat through stress analysis the relationship of neat adhesive properties to joint properties. Finally, our basic understanding of the effect of changes occurring during cure on residual stress is inadequate. FINITE ELEMENT ANALYSIS Finite element analysis (FEA) is a specific type of stress analysis. However, FEA has become a discipline of its own because of the sophistication involved with present numerical analysis techniques. The problems that can be accurately solved with FEA are virtually limitless compared to the number of closed-form solutions or strength-of-materials approximations. As powerful as FEA is, the results are only as good as the modeling of the material behavior and the geometric conditions. A number of excellent FEA computer programs have recently been developed specifically for analysis of bonded joints and are currently undergoing evaluation. However, none of these codes is as flexible as might be desired. Research would be assisted if flexible programs that can handle geometric nonlinear analysis, plasticity, viscoelasticity, 2-D or 3-D analysis, thermal effects, resin and interphase degradation, and moisture absorption and migration were available. These programs should be as cost-efficient as possible without sacrificing accuracy. The FEA must be able to model realistic material behavior. FAILURE MECHANISMS To design an efficient, durable adhesively bonded structure, maximum design stresses must be established. Such stresses are those associated with failure, for example, by long-term creep rupture, cyclic fatigue, or
38 static failure due to material degradation. Potential failure mechanisms and their associated stress and environmental dependence must be characterized for each adhesive prior to application. Fracture mechanics has been demonstrated to be a viable tool for characterizing and predicting static strength and cyclic debonding of adhesive joints containing initial debonding. Several viable nonlinear viscoelastic models exist for the prediction of long-term creep rupture. There is general agreement that emphasis should be placed on defining the threshold region for fatigue damage and creep rupture and then using these data as upper design limits. The failure mechanisms of several frequently used structural adhesives when subjected to severe environments needs to be established. Such determinations will also permit the development of an experimental data base, which is needed. One use of a long-time durability data base is to verify accelerated characterization test methods. The testing mentioned above will help define the relative influences of viscoelasticity (creep), cyclic fatigue, and static loading in severe environments, and also to characterize fatigue and creep interactions.