Лако-красочные материалы — производство

General considerations

The adhesive, which can be likened to plastic material, represents a low modulus interlayer and is likely to be the weakest link in a structural joint. Exceptions to this might be when one or both substrates are concrete subjected to forces other than compression, or if thin sheet metal adherends are involved. Design considerations generally involve the geometry of the bond, the selection of an adhesive and the necessary bonding process, a knowledge of the properties of the adhesive and of the adherends, and finally an analysis of the stresses that the bonded assembly is likely to encounter in service. Petronio(9) states that for maximum success joint design should follow several general principles, namely to:

(1) stress the adhesive in the direction of maximum strength (i. e. in compresion or in shear)

(2) provide for the maximum bond area

(3) make the adhesive layer as uniform as possible

(4) maintain a thin and continuous bondline

(5) avoid stress concentrations.

Assuming that the essential factors necessary to create a reliable joint as outlined in the introduction have generally been complied with, joint strength is related essentially to the mechanical properties of the adhesive with which it is constructed. Since most of the adherends united by structural adhesives are loaded in tension, the subsequent loads and hence stresses on the adhesive are then a function of the geometry of the joint. Unfortunately even in the simplest of joint configurations the stress state is complex, and Adams and Wake(5) give very detailed consideration to the nature and magnitude of stresses in joints.

Most of the adhesives used for structural purposes are relatively strong in shear, and weak in peel or cleavage. Bonds are therefore usually designed to place the adhesive in shear and to minimise these other stresses. Excluding compression, the four important types of stress to consider are shear, tension, cleavage and peel, and the resulting joint elastic stress distributions are shown schematically in Fig. 4.1. (Note that peel is the enemy of the joint designer and implies principally transverse tearing loads.) Kinloch(4) presents a useful chronological review of the approaches made in analysing the stresses in various kinds of bonded joint.

The lapped joint depicted in Fig. 4.1(a) is one of the most commonly occurring joints in practice and is therefore the configur­ation most often used for testing adhesives(10, 11). As shown, the problem is that the loads are not linear, and a bending moment must exist so causing the joint to rotate. In consequence, the adhesive layer will be subjected both to shear and to tearing stresses at the ends of the joint. The adherends, too, are subjected to shearing, stretching and bending. As the load is increased it is probable that both adhesive and adherend will become plastic in the highly-stressed regions. The double lap joint, whilst eliminating gross joint rotation, is really no more than a back-to-back arrangement of two single-lap joints and significant tearing stresses still exist.

Lapped joints, howsoever fastened, support most of the applied load as stress concentrations at the two ends of the joint, and it is for this reason that the strength of such joints depends largely upon their width and very little upon the overlap length. So far as relatively rigid lapped joints are concerned Gordon(12) observed that it makes little difference whether the joint is glued, nailed, screwed, welded, bolted or riveted. However, as the imposed load on bonded lap joints is raised the adhesive layer may fracture before it yields. This problem may be overcome to a certain extent by using flexibilised adhesives which yield considerably before they fracture. Alternatively, toughened adhesives may be employed to inhibit crack propagation through the bondline, and so prevent sudden fracture. These approaches are not without their drawbacks because flexibilised and toughened adhesives may creep readily under sustained load. Further complexities of the problem such as

(b) (d)

Fig. 4.1. Elastic stress distributions in various kinds of bonded joints, (a) Lap shear. (b) Butt-tensile and scarf, (c) Cleavage. (d) Peel.

the adherend surface rugosity, cure shrinkage, the flaw-sensitive nature of cured adhesives, and the adhesive in-bondline mechanical performance are also apparent. On the positive side, real bonded joints are inevitably made with spew fillets of adhesive at their extremities and, as studied extensively by Adams and co-workers

at Bristol University during the 1970s and 80s, this generally has the effect of reducing stress concentrations at the end of the overlap. The adherends may also be made less stiff at the ends of the joint by tapering, scarfing and so on, to smooth the transfer of stress and strain across the joint. Many such examples of good practice in joint design for ‘production’ applications are depicted in several texts(2, 5, 6, 7), and some of these are shown in Fig. 4.2. The scarf joint, though ideal in minimising stress concentrations, is seldom practicable.

The design of a structural adhesive joint intended to resist significant loads is clearly quite onerous. For instance, it is not permissible to assume uniform stress distributions because at ultimate conditions fracture is often experienced. With traditional methods of fastening the same stress concentrations will exist but the materials can normally be relied upon to yield and there is little need to be so concerned about transverse normal tensile stresses. In order to make reliable strength predictions, bonded joint design necessitates either some form of mathematical analysis, or else extrapolation from experimental data. Two such design philosophies are based on stress analysis and on fracture mechanics. It is assumed implicitly that other modes of joint failure such as impact, excessive deformation, creep and gross corrosion are considered in parallel with these design techniques.