Fatigue lifetime prediction of adhesively bonded joints: An investigation of the influence of material model and multiaxiality. (October 2017)

• Record Type: Journal Article
• Authors: Beber, V.C.; Fernandes, P.H.E.; Schneider, B.; Brede, M.; Mayer, B.
• Is Part Of: International journal of adhesion & adhesives. Volume 78(2017:Oct.)
• Page Start: 240
• Page End: 247
• Publication Date: 2017-10
• Subjects: Structural adhesives -- Finite element stress analysis -- Stress distribution -- Fatigue -- Joint design; Adhesion -- Periodicals; Adhesives -- Periodicals; Adhésion (Physique) -- Périodiques; Adhésifs -- Périodiques; Adhesie; Kleefstoffen; Adhesion; Adhesives; Periodicals; 668.3
• DOI: 10.1016/j.ijadhadh.2017.08.007
• Languages: English
• ISSNs: 0143-7496

Abstract

Abstract: In the present work the influence of material model and multiaxiality on the fatigue lifetime predictions of scarf and single-lap joints under uniaxial tension-tension cyclic loading is investigated. Lifetime predictions, from 10 3 to 10 6 cycles, were performed using a stress-life approach taking into account for the effect of multiaxiality by means of hydrostatic pressure ( p ) and von Mises stress ( q ). These stresses were calculated employing 2D-FEA with linear-elastic and elastoplastic (von Mises and Drucker-Prager) material models. Effective stresses were obtained using the theory of critical distances. Findings showed that material model affected significantly the multiaxiality distribution of the single lap joint (SLJ), especially at singularity regions. For scarf joints, this effect was noticeable for higher stresses. With the exception of the SLJ, by excluding the singularity regions close to the free edges, the value of multiaxiality was nearly constant and decreased with increasing bondline angles (higher shear effects). With regards to lifetime predictions, overall accuracy for the scarf joint was higher than for the SLJ. Linear-elastic material models provided satisfactory accuracy for the scarf joint, but not for the SLJ. Elastoplastic material models were able to improve predictions and to provide suitable accuracy for both joint configurations.