This paper examines the need to incorporate a rate-dependent traction-separ
ation law in order to model quasi-static debonding between rubber and metal
. A pseudo-stress model was used to account for the nonlinear, multi-axial
and time-dependent nature of the filled rubber that was used in the experim
ents. The parameters for the traction-separation law were extracted on the
basis of measurements of load, crack length and crack opening displacements
in an opening mode experiment at one applied displacement rate. The form o
f the traction-separation law was consistent with observations of ligament
content on the metal fracture surface. Additional experiments were then con
ducted in opening mode at different rates and in mixed mode with positive a
nd negative shear so that comparisons with predictions from the calibrated
cohesive zone model could be made. The crack length history proved to be th
e most discriminating measure of the validity of the model, which was most
effective at higher loading rates. (C) 2001 Elsevier Science Ltd. All right
s reserved.