Mean stress correction and fatigue failure criteria for hyperelastic adhesive joints

Abstract

ABSTRACT The work investigates failure criteria and mean stress correction approaches for the fatigue lifetime prediction of two hyperelastic adhesives (a polyurethane, PU, and a silicon-modified polymer, SMP). Fatigue experiments are carried under constant amplitude cyclic loading at RT and 40°C/60% r.h with butt- and the thick-adherend-shear-test-joints at three stress ratios R = −1, 0.1 and 0.5. Three mean stress correction approaches are evaluated: Goodman (static strength based), Schütz (mean stress sensitivity based) and Kujawski & Ellyin (parameter optimisation based). Fatigue failure criteria considered are: Drucker-Prager (linear-relation with the hydrostatic stress), Beltrami (quadratic relationship with the hydrostatic stress), and a multivariable nominal shear and tensile stresses criterion (data-based). The comparison is based on prediction accuracy (R-squared of master SN curves) and complexity of parameter determination. The highest values of R-squared were obtained by the Kujawski and Ellyin correction, followed by Schütz and then Goodman. However, the complexity of parameter determination follows an opposite trend with Goodman being the lightest approach. Failure criteria yielded comparable results with the multivariable criterion having the advantage of not dealing with FEA, but being limited to joints with nearly uniform stress distribution. Finally, compared to Drucker-Prager, the Beltrami criterion had a more robust parameter determination.