Novel Design of a Bonded Lap Joint

Abstract

A new design of a single-lap joint was proposed and investigated experimentally. In this new design, load eccentricity as well as singular peel stresses in the joint interface were avoided. In fact, numerical calculations show that, in the new design, the peel stress becomes compressive in the joint end region, and the shear load is more evenly transferred over the length of the joint. Two adherend layups, that is, (0/90/0/90) 2s and (90/0/90/0) 2s, were considered. Experimental results show that the strength of the new joint is signie cantly higher than that of the conventional single-lap joint. It is believed that even higher strength can be obtained by optimizing the new design cone guration. ANY structures consist of a number of basic parts that are connected together to take up the load. The connections or joints are usually the weakest locations in the structure. Among the many technical challenges facing engineers using advanced com- posite materials is joining composite parts or composite to metallic components.Eventhoughtheuseofcompositematerialscangreatly reduce the number of parts, joining parts is still unavoidable. Tradi- tional joining methods for metals such as welding and soldering are apparently unsuitable for e ber-reinforced composites. Compared to mechanical fastening, adhesive bonding is attractive for joining e ber-reinforced composite structures because it reduces the local- ized stresses encountered when using bolts and rivets. Moreover, bolts and rivets cut the continuous reinforcing e bers, and, as a re- sult, may greatly reduce the overall load carrying capacity of the structure. Adhesively bonded joints are used in many different cone gura- tions, among which the most commonly used are single-lap joints, double-lap joints, scarf joints, and step-lap joints. 1 3 The single-lap jointisgenerallythesimplestandleastexpensiveofalljointstoman- ufacture. However, because of its intrinsic load eccentricity and the abruptchangeofload path,boththe normal (peel)and shearstresses are highly localized at the two joint ends; thus, the conventional lap jointsare not efe cient in load transfer. 4 Moreover,because theinter- facial normal stress is tensile and theoretically singular at the joint ends, it can initiate failure at these locations. Although tapering the adherend or adding adhesive e llets at joint ends can lead to some reduction of peel stresses in the lap joint, these improvements are, however, quite modest. 5;6 Structural adhesives have relatively poor resistance to peel or cleavage stresses. Therefore, an obvious step to improve joint strength is to reduce the magnitude of peel stress or to make it compressive. To obtain maximum joint efe ciency, the interfacial shear stress distributions should also be made more uniform. In this paper, a newwavy bonded lap joint is proposed and inves- tigated. With this new joint design, the interfacial normal stress is compressiveatthetwojointends,andtheinterfacialshearstressdis- tribution is less localized. Experimental results show that the wavy joint has much greater strength than the conventional lap joint. Conventional Single-Lap and New Wavy-Lap Designs For comparison, a two-dimensional sketch of the conventional single-lap joint is shown in Fig. 1. Only the case of identical ad- herends is considered in this study. The length of both outer ad- herends is L, the overlap (joint) length is l, and the thickness of adherend and adhesive are ta and tg, respectively. the waviness angles (Fig. 2) ® DD tan 1 .2ta=l 0 ) and round out sharp corners of the joint with an adequate radius R. The e nal joint shape selected is shown in Fig. 3. The shape chosen here is mainly for simplicity; other possible shapes may be as effective. Strictly speaking, the two adherends in this wavy joint are not in a collinear cone guration. The dashed lines in Fig. 2 show this. If a collinear cone guration is desired, then the thickness of the adhesive must be taken into account during the e rst design step. Because the thickness of the adhesive is very small compared with that of the adherends in this study, it is neglected during the design process. The values of the geometrical parameters for both conventional and wavy joints are given in Table 1. Even though l 0 <l, the overall bonding areas are the same for the two joints. This ensures that analysis and test results of these joints can be directly compared. TheonlydimensionthatisnotshownintheFigs.1and2isthewidth of specimen, which is d D25:4 mm. The material for adherends is AS4/3501-6 carbon/epoxy. Two different layups, (90/0/90/0) 2s and (0/90/0/90) 2s, are used. The 0-deg direction is parallel to the longitudinal (loading)directionofthejoint.Thereasonforchoosing these two layups is to study the difference between 90 -90 deg and 0-0 deg bonding interfaces. The material constants of AS4/3501-6 are listed in Table 2. The two adherends are bonded together with a layer of e lm adhesive FM73M, which is treated as an isotropic material.Theelastic constants forFM73Mare also listed in Table2.