The strength of an adhesive contact in the presence of interfacial defects

Abstract

Adhesive contacts which possess a dominant stress concentration, such as at the contact edge in spherical junctions or at the detachment front in a peeling film, are well studied. More complex adhesive junction geometries, such as mushroom-shaped fibrils in bioinspired micropatterned dry adhesives, have exhibited a complex dependence of adhesive strength on the presence of interfacial defects within the contact. This has led to the emergence of statistical variation of the local behavior among micropatterned sub-contacts. In order to examine the interplay between geometry and interfacial defect character in control of the adhesive strength, the model system of a stiff cylindrical probe on an elastic layer is examined. Both experiments (glass on PDMS) and cohesive zone finite element simulations are performed, with analytical asymptotic limits also considered. The thickness of the elastic layer is varied to alter the interfacial stress distribution, with thinner layers having a reduced edge stress concentration at the expense of increased stress at the contact center. The size and position of manufactured interfacial defects is varied. It is observed that for the thickest substrates the edge stress concentration is dominant, with detachment propagating from this region regardless of the presence of an interfacial defect within the contact. Only very large center defects, with radius greater than half of that of the contact influence the adhesive strength. This transition is in agreement with analytical asymptotic limits. As the substrate is made thinner and the stress distribution changes, a strong decay in adhesive strength with increasing center defect radius emerges. For the thinnest substrate the flaw-insensitive upper bound is approached, suggesting that this decay is dominated by a reduction in the contact area. For penny-shaped defects at increasing radial positions, the adhesive strength for the thinnest substrates becomes non-monotonic. This confirms an intricate interplay between the geometry-controlled interfacial stress distribution and the size and position of interfacial defects in adhesive contacts, which will lead to statistical variation in strength when defects form due to surface roughness, fabrication imperfections, or contaminant particles.