Additional Interface Corner Toughness Data for an Adhesively-Bonded Butt Joint;International Journal of Fracture: Letters in Fracture & Micromechanics

摘要

Additional Interface Corner Toughness Data for an Adhesively-Bonded Butt Joint Abstract. Over a period of 15 months, five sets of adhesively-bonded butt joints were fabricated and tested. This previously unreported data is used to assess the variability of measured interface corner toughness values, as well as the dependence of on surface preparation. A correlation between KC and the size of the adhesive failure zone is also noted. 1. Introduction. The observed decrease in the tensile strength of adhesively-bonded butt joints with increasing bond thickness is accurately predicted by a fracture criterion that assumes failure occurs at a critical value of the interface comer stress intensity factor, (Reedy and Guess, 1993, 1997). This approach is analogous to linear elastic fracture mechanics except here the critical stress intensity factor is associated with a discontinuity other than a crack tip (i.e., the point were the interface intersects the stress-free edge). For the case of a cylindrical butt joint formed by bonding aluminum adherends together with a thin epoxy adhesive layer, KC is defined by Set 1 was cured for 24 hours at 50 'C followed by an additional 24 hours at 40 'C. This cure schedule was chosen to minimize residual stress. The epoxy's glass transition temperature is 68 'C (dielectric measurement technique). Note that Sets 2-5 added an initial 24 hours at 28 'C step to the cure schedule to simplify handling. This modification of the cure schedule has minimal effect on the epoxy's properties. Compression tests of molded epoxy plugs cured with and without the 28 'C step yielded nearly identical stress-strain relations. All except one set of joints included joints of several different bond thickness (e.g., 0.25,0.5, and 1.0 mm). Note that when the K,C failure criteria holds, IQ should not depend on bond thickness. The joints were tested in a conventional, screw-driven load frame. The load train utilizes a chain linkage to minimize the effect of specimen misalignment. All joints were tested at room temperature and loaded at a crosshead displacement rate of 0.2 mm/s. 5. Observed correlation of the size of the adhesive failure zone and IQ A fractography examination of the failed butt joints showed that failure always initiates adhesively along a segment of the specimen periphery (i.e., at low magnification the initial failure zone appears to be interracial). Beyond this thumbnail-like region the failure is cohesive. There was a general tendency for the size of the adhesive failure zone to decrease as; increased. This was true for nominally identical joints, and also for joints with different surface preparations. Figure 1 plots results for all joints in Sets 1-5 except for the Set 1 joints with a 0.2 mm thick bond and a R;= 1 ;m surface roughness. There was no clearly definable region of cohesive failure in these nine joints. The characteristic size of the adhesive failure zone is taken to be E7Z (the failure looks roughly semielliptical, and this is the radius of a semicircle with the same area). Although this result pertains to a single adherend/adhesive pair, it does suggest that under certain circumstances, one might be able to estimate & and the failure stress since bond thickness is known, Eq. (1) by simply measuring the size of the adhesive.