Crack growth in adhesives: Similitude and the Hartman-Schijve equation. (1st October 2021)

• Record Type: Journal Article
• Authors: Jones, R.; Kinloch, A.J.; Michopoulos, J.; Iliopoulos, A.P.
• Is Part Of: Composite structures. Volume 273(2021)
• Publication Date: 2021-10-01
• Subjects: Fatigue -- Delamination growth -- Cohesive crack growth -- Hartman-Schijve crack growth equation; Composite construction -- Periodicals; Composites -- Périodiques; 624.18
• DOI: 10.1016/j.compstruct.2021.114260
• Languages: English
• ISSNs: 0263-8223

Abstract

Highlights: This paper discusses the importance of similitude in analysis and in laboratory tests. Plotting da/dN versus ΔG yields curves that are a function of adhesive thickness. Hence two cracks with the same value of ΔG, and Gmax, will NOT grow at the same rate. The Hartman-Schijve crack driving force Δκ can overcome this shortcoming. FM94 crack growth tests at different temperatures lie on a single da/dN v Δκ curve. Abstract: Structural adhesives are widely used for joining composite components in many industries and crack growth in such materials is far more likely to occur when they are subjected to repeated cyclic loading than to monotonic loading. Whilst the Hartman-Schijve equation for fatigue crack growth (FCG) has been shown to hold for cohesive crack growth in adhesives under Mode I, Mode II and Mixed-Mode I/II loading, little attention has been paid to its ability to capture the effects of the thickness of the adhesive layer or the test temperature. The present paper examines FCG, that occurs cohesively through the adhesive layer, in two toughened epoxy adhesives typical of those used in the automotive and the aerospace industries. Firstly, it is established that when the FCG rate, da/dN, curves are expressed as a function of Δ√ G, or Δ G, where G is the energy release-rate, then the crack growth curves are a function of the thickness of the adhesive layer. It is then shown that this dependency vanishes when da/dN is expressed as a function of the crack-driving force, Δ κ, as defined by the Hartman-Schijve equation. Secondly, it is shown that similar observations are recorded when the effects of test temperature are considered. Therefore, it is suggested that the parameter Δ κ appears to be a valid similitude parameter.