Experiments on peeling

Abstract The force required to peel an aluminum (or nickel) ribbon glued to a rigid glass plate with a polyethylene or polyvinyl acetate was determined. It proved to be smaller than expected (from a theory, partly new) both for adhesives which were almost Hookean solids and for those whose stress‐strain curve could approximately be represented as stress = const. (strain) 0.5 . In the latter group, the difference between the theory and the experiment was due to stress concentrations at the right‐hand and the left‐hand edges of the adhesive layer; these stress concentrations formed because, when a pull was applied, the adhesive contracted in the directions perpendicular to that of pull. As a consequence of this effect, the peeling force W 0 was not proportional to the width w of the ribbon; sometimes, a linear relation W 0 = aw + b seemed to be valid. On the other hand, near‐Hookean adhesives of a low total elongation required a peeling force proportional to w ; the stress concentration dangerous for these materials occurred at the boundary between the adhesive and the bent ribbon at the sharp bend. The damage to the brittle adhesive caused by the ribbon deformation could be reproduced without bending, namely by extension of the ribbon beyond its yield strength.