Studies of strength development at polymer-polymer interfaces are examined and applications to welding of similar and dissimilar polymers are considered. The fracture properties of the weld, namely, fracture stress, u, fracture energy, GrC, fatigue crack propagation rate daldN, and microscopic aspects of the deformation process are determined using compact tension, wedge cleavage, and double cantilever beam healing experiments. The mechanical properties are related to the structure of the interface via microscopic deformation mechanisms involving disentanglement and bond rupture. The time dependent structure of the welding interface is determined in terms of the molecular dynamics of the polymer chains, the chemical compatibility, and the fractal nature of diffuse interfaces. Several experimental methods are used to probe the weld structure and compare with theoretical scaling laws. Results are given for symmetric amorphous welds, incompatible and compatible asymmetric amorphous welds, incompatible semicrystalline and polymer-metal welds. The relevance of interface healing studies to thermal, friction, solvent and ultrasonic welds is discussed.

SynopsisThe kinetics of craze growth and craze healing were studied by dark-field optical microscopy in monodisperse molecular weight polystyrene (PS) that varied in molecular weight from 88,OOO to 1,334,000. The following observations were made. (1) G , , the virgin growth rate, decreased rapidly with increasing molecular weight until M , -200,000 and then remained constant. (2) G , decreased with increasing craze density. (3) The growth rates of approaching craze tips decreased when the craze tips overlapped, and the effect was less for crazes whose parallel growth paths were greater than 40 pm apart. (4) Complete craze healing was ohserved by comparison of the nucleation times, T~, and growth rates, G , , of healed individual crazes with the craze kinetics of the virgin sample. (5) The extent of healing was characterized using four cases in which 7 and C were measured as a function of healing time, temperature, constant stress, and molecular weight. (6) Craze healing times were found to increase with molecular weight and were analyzed in terms of the modified molecular weight of the craze zone. (7) Significant bond rupture was determined to occur during crazing by comparison of healing times with stress relaxation and diffusion data. (8) Craze healing studies provide insight into both crack healing and fracture of glassy polymers.form.'--' However, Fellers and Keelo and Koltisko" found that the stress a t which crazes appeared for specimens in uniaxial tension fracture tests was independent of molecular weight. Although Fellers et al.'" and Lainchbury and Bevis9 found the craze density to decrease for molecular weights below 100,OOO. Koltisko et al." found no effect of molecular weight on craze density.