Failure analysis of adhesively-bonded metal-skin-to-composite-stiffener: Effect of temperature and cyclic loading

Research output: Contribution to journalArticleScientificpeer-review

23 Citations (Scopus)
119 Downloads (Pure)

Abstract

The aim of this research is to analyse the failure of a Fiber Metal Laminate (FML) skin adhesively bonded to a Carbon Fiber Reinforced Polymer (CFRP) stiffener, under quasi-static loading at different environmental temperatures (−55 °C, Room Temperature RT and +100 °C) and under fatigue loading at RT. This bonded joint was tested using stiffener pull-off tests, which is a typical setup used to simulate full-scale components subject to out-of-plane loading. The failure sequence for all test conditions consist of: (1) damage initiation at the noodle of the CFRP stiffener; (2) damage propagation by delamination from the noodle to the stiffener foot; (3) detachment of the stiffener from the skin. Increasing the temperature, decreases the joint stiffness (40% when compared to RT) and decreasing the temperature decreases the maximum load (50% when compared to RT). The fatigue life initiation of the joint presents a very large scatter but the fatigue life propagation presents more stable results. The fatigue threshold (no damage) is reached at approximately 30% of the maximum load level. The fracture surfaces indicate a predominant inter and intra-laminar failure of the composite under mixed mode I/II. The CFRP stiffener is the weakest link of the bonded FML-skin-to-CFRP-stiffener both for static and fatigue loading.

Original languageEnglish
Pages (from-to)27-37
Number of pages11
JournalComposite Structures
Volume166
DOIs
Publication statusPublished - 15 Apr 2017

Keywords

  • composite-to-aluminium joints
  • fatigue
  • Keywords: Adhesive bonding
  • skin-to-stiffener connection
  • thermal loading

Fingerprint Dive into the research topics of 'Failure analysis of adhesively-bonded metal-skin-to-composite-stiffener: Effect of temperature and cyclic loading'. Together they form a unique fingerprint.

Cite this