Fatigue failure analysis of riveted fibre-metal laminate lap joints
Introduction
Glare™ fibre-metal laminate (FML) was first developed at Delft University - Netherlands, whose patent was filled in 1987. This FML is built up from high-strength aluminium alloy layers interspersed with and bonded to unidirectional or cross-ply tape arrays of high-strength GLAss fibre-REinforcing thermoset epoxy resin. Glare is especially attractive to the aeronautical industry by virtue of its exceptionally higher structural efficiency in terms of fatigue strength, not to mention its notably superior burn-through resistance, than concurrent monolithic Al-alloys sheets [1], [2], [3], [4], [5], [6], [7].
Several Glare-FML grades are now available, and the main difference among them is in the composite layer layup and the metal layers thickness [2], [5].
Considering that the upper fuselage and leading-edges of Airbus A380 aircraft were successfully built with Glare, riveted lap joints of this composite material are entitled to the fabrication and repair of commercial jet planes [8], [9], [10], [11]. Hence, the assurance of in-flight safety of new and repaired aircrafts depend upon the structural integrity of mechanically fastened FML joints.
Concerns regarding the need for reliable design methods for bonding composite laminates, long-term ageing behaviour of adhesively bonded joints, and difficulties in their non-destructive evaluation and repair have led to reluctance to use adhesives in primary aircraft structures [12], [13]. A common mixed approach to pressurized aircrafts is to utilize riveted and bonded joints. In this joining process, the adhesive distributes loads and acts as a sealant, while the mechanical elements (rivets) will ultimately sustain the applied stresses in case of adhesive failure [14]. Therefore, full understanding of fatigue fracture mechanisms in riveted joints is of utmost importance, especially in the case of new structural materials like FML.
Ryan & Monaghan [12] studied the mechanical failure of riveted Glare™ joints subjected to static loading and predicted that stress concentration on the stiffer metallic layers of Glare, as compared to juxtaposed epoxy resin/glass fibre layers, causes delamination in countersunk holes. Lazzeri [15] found that rivet failure governs the failure mechanisms of Glare joints up to a high number of fatigue cycles, but above this number, panel failure becomes dominant.
Heida & Platenkamp [8] inspected non-destructively small-scale riveted Glare joints containing notches-simulating fatigue cracks and concluded that sub-surface cracks larger than 4 mm are well detectable at a depth of 3 mm, but at 6 mm cracks were no longer discernible. It is worth mentioning that none of these studies focused on the details of operating fatigue mechanisms in such mechanically fastened joints.
Although Glare displays higher fatigue crack growth resistance than concurrent monolithic Al-alloy due the so-called crack bridging mechanism [1], [2], [3], [4], [5], [6], [7], stress concentration is 70% higher in countersunk rivet holes regarding to flat-bottomed, due to the so-called knife-edge condition [16]. Fatigue crack initiation life in mechanically fastened Glare joints is therefore considerably lower than monolithic Al-alloy joints with identical sheet thickness submitted to similar remote peak stress [16], [17]. Premature delamination during riveting process in Glare sheets with countersunk rivet holes reduces laminate stiffness giving rise to secondary bending deflection in lap shear joints, which favours further delamination. As a result, fatigue crack nucleation life can be severely shortened [14], [16], [17].
The role of squeezing force applied during riveting process on the in-service performance (essentially fatigue) of Glare and monolithic Al-alloys has also been object of research. Heida & Platenkamp [8] determined that when the riveting force is relatively high the cracks initiate outside the holes, grow tangentially around the holes and are then less detectable. Other studies on the subject [14], [16], [18], [19] showed the beneficial effect of high squeezing forces on fatigue initiation life (up to one order of magnitude) of riveted joints due to residual compressive stresses at the hole surface, reduction of empty spaces, fretting prevention and more effective load transfer paths.
According to De Rijck [14], to understand the failure mechanisms acting in engineering structures it is preferable to evaluate the variables independently. The researcher defends that laboratory-controlled conditions allow the identification of the role and preponderance of an individual variable in the failure events, though, of course, synergic, and off-setting effects may occur when acting simultaneously in-service conditions. De Rijck also supports basic specimen configuration and simplified testing approaches by emphasizing the unfeasibility to separate the effects resulting from several variables acting concurrently in more complex riveting patterns resembling in-service structural components (e.g. Lazzeri [15] and Lanciotti & Lazzeri [20]).
The open literature on fatigue behaviour of riveted lap joints of FML is still relatively scarce [8], [10], [12], [14], [15], [16], [18], [19], [20], [21], and so incomplete, as compared to traditional aircraft construction materials. Even more recent studies have not clarified the essential relationship between each individual joint component (sheet and rivet) and the acting fatigue failure mechanisms when two basic rivet arrangement, namely, aligned, and non-aligned to the loading direction are considered separately.
By utilizing Glare laminate architecture and its riveted joint configurations at their most basic levels, the present work intends to identify and correlate the underlining fatigue mechanisms leading to FML joint failure. Stress-life curves, along with extensive failure analysis of completely fractured test pieces provide information that can substantiate the analysis and understanding of fatigue failure behaviour of in-service multi-rivet lap joints, and, by consequence, enhance their design process [4].
Section snippets
Materials
Glare-5™ FML variation, manufactured by Comtek Advanced Structures™ (Canada), was tested. This 1.6 mm-thick FML is composed of two 0.5 mm-thick layers of 2024-T3 aluminium alloy bonded to unidirectional high-strength glass fibres (GF, S2 grade) / epoxy (EPX) matrix composite in a 0°/90°/90°/0° layer system (i.e., 2/1 architecture). Its macro/microstructure is illustrated in Fig. 1.
Though this grade of Glare was conceived for maximum impact performance (e.g., J-nose leading edge of A380
Riveted specimens
Two riveted specimen configurations were tested, with two rivets aligned or not to the loading line, as depicted in Fig. 2a. Such joint configurations, hereafter named respectively A and NA, were chosen due to their simplicity, thus facilitating visualization, monitoring and interpretation of fracture damage mechanisms, as earlier commented.
Loading direction corresponded to the rolling direction of the metal layers, i.e., maximum tensile strength axis of the Glare laminate (0°).
Fig. 2b presents
Monotonic tensile tests
Fig. 3(a, b) presents the faying surfaces (numbered 2 and 3 according to Fig. 2c) of a fractured A-joint. The same is made for a NA-joint in Fig. 3(c, d). Arrows indicate the correspondence between rivets and holes for both lap joint configurations.
According to Fig. 3(e), a stable steady-state failure process occurred in the A-joint. Fig. 3(a, b) reveals that it started by plastic straining of both countersunk rivet heads at the maximum load capacity and extended until the complete rupture of
Conclusions
This study investigated single shear lap joints of fibre-metal laminate aiming at understanding their tensile fatigue performance in terms of principal operating fracture mechanisms under constant amplitude loading. For this intent, rivet pattern and fibre-metal laminate layout were evaluated at their elementary level, namely, two rivets aligned, and non-aligned to the loading direction, which fastened simple 2/1 Glare™ laminate layup.
Main obtained results, briefly stated, were as follows:
- i.
The
CRediT authorship contribution statement
A.C.S.P. Tarpani: formal analysis, draft writing, editing, visualization, and review; T.A. Barreto: draft writing, editing, visualization, and review; J.R. Tarpani: conceptualization, methodology, investigation, data curation and validation, formal analysis, resources, supervision, and funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
To FAPESP - Sao Paulo Research Foundation (Processes 2005/54899-8, 2006/50134-0, 2010/09822-5, 2010/09823-1, 2012/09319-7, and 2013/27030-7) for financial support; Aviation Equipment Inc. (Costa Mesa, CA-USA) for gently providing the fibre-metal laminate Glare-5™ 2/1; EMBRAER S/A – Brazilian Aerospace Company for machining and assembling the riveted lap shear joints; Compoende Aeronautics Brazil for analogous and digital X-ray radiography inspections. NEMAF - Nucleus of Materials Testing and
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (32)
- et al.
Fatigue and damage tolerance issues of glare in aircraft structures
Int J Fatigue
(2006) - et al.
A review: fibre metal laminates, background, bonding types and applied test methods
Mater Des
(2011) - et al.
Failure mechanism of riveted joint in fibre metal laminates
J Mater Process Technol
(2000) Delamination buckling of fibre-metal laminates
Compos Sci Technol
(2001)- et al.
Mechanical behaviour of self-piercing riveted multi-layer joints under different specimen configurations
Mater Des
(2007) - et al.
Fatigue behaviors of self-piercing rivets joining similar and dissimilar sheet metals
Int J Fatigue
(2007) - et al.
Fretting behaviour of self-piercing riveted aluminium alloy joints under different interfacial conditions
Mater Des
(2006) - et al.
Fretting wear in self-piercing riveted aluminium alloy sheet
Wear
(2003) - et al.
Observation, analysis and prediction of fretting fatigue in 2024–T351 aluminum alloy
Wear
(1998) - et al.
Fatigue crack initiation in fibre-metal laminate Glare 2
Mater Sci Engng
(1997)
Towards application of fibre metal laminates in large aircraft
Aircraft Eng Aerospace Technol
Glare technology development 1997–2000
Appl Compos Mater
Fatigue and damage tolerance of Glare
Appl Compos Mater
An historic overview of the development of fibre metal laminates
Appl Compos Mater
Glare design aspects and philosophies
Appl Compos Mater
Cited by (10)
Effect of laser shock peening without protective coating on surface integrity of titanium-based carbon-fibre/epoxy laminates
2023, Optics and Laser TechnologyApplications of data-driven approaches in prediction of fatigue and fracture
2022, Materials Today CommunicationsCitation Excerpt :Currently, fatigue life prediction of composite under variable amplitude loading is mainly focused on the polymer composites. Multiaxial in-plane and out-of-plane fatigue in cross-ply, unidirectional, and other layup configuration (e.g., woven and hybrid composites) have been investigated over the years [76–80]. The life prediction models can be divided into three categories: (a) macro property degradation models, (b) damage models, and (c) fatigue models.
An experimental investigation on press joining mechanism and bearing behavior of single-lap composite riveted joints
2023, Journal of Composite MaterialsEffect of bolt tightening on the fatigue behavior of GLARE double shear lap joints
2021, Journal of Composite Materials