A Criterion for Joints Failure D. Leguillon Laboratoire de Modélisation en Mécanique, CNRS UMR7607, Université Pierre et marie Curie, 8 rue du Capitaine Scott, 75015 PARIS, France. ABSTRACT: The author has proposed a fracture criterion for crack nucleation at stress concentration points (notches, corner points, …). It involves both strength and toughness of the material and is consistent with the Griffith criterion for a crack. It is employed herein to predict failure of joints. Stress concentration occurs at the interfaces between the joint and the substrates at the end of the joints and it is a privileged site for crack nucleation. Butt and scarf joints are studied with and without thermal residual stresses. The role of a spew fillet in a corner is discussed. INTRODUCTION An important issue in the reliability of joints is the prediction of their load bearing capacity. Failure often occurs at the ends of the joints, at corners in lap joints, at straight edges in butt or scarf joints. They are stress concentrations locations because of the elastic mismatch between the components of the structure and the joint itself which is often much more compliant. The usual Griffith criterion is inefficient to predict crack initiation at these points. As emphasized by the author in recent papers [1,2], the differential form of the energy criterion must be replaced by an incremental one. This necessary condition for fracture must be completed by the stress criterion. Both are necessary conditions but neither one nor the other is sufficient. They give respectively a lower and an upper bound for admissible crack lengths. The consistency between these two bounds provides a criterion for crack initiation, which coincides with the Griffith one for a pre-existing crack starting to grow. It writes in an Irwin-like form: c k k t , (1) where k is the intensity factor of the most significant elastic singular term (characterizing the stress concentration). Its critical value ck is expressed as
RkJQdWJsaXNoZXIy MjM0NDE=