Elsevier

Composite Structures

Volume 279, 1 January 2022, 114741
Composite Structures

Cyclic behaviour modelling of GFRP adhesive connections by an imperfect soft interface model with damage evolution

https://doi.org/10.1016/j.compstruct.2021.114741Get rights and content

Abstract

In this paper a methodology for studying the mechanical behaviour of complex structures made of GFRP structural members joined by means of adhesive connections and subject to variable intensity loads is presented. The fundamental equations are derived by an asymptotic approach applied on composite structures made of two elastic solids bonded together by a third thin one represented by a thin layer with a nonlinear behaviour. The adhesive layers are considered micro cracked according to the Kachanov’s assumptions.

Within this framework, to calibrate the parameters of the imperfect interface model, the mechanical properties and damage evolution of an epoxy adhesive have been experimentally evaluated under cyclic loadings. The experimental tests have been performed on aluminium cylinders considering different thicknesses of adhesive. The experimental results evidence how the adhesive thickness influences the strength, stiffness and consequently the initial damage parameter (initial cracks) of the bonded connections.

Finally, the robustness and accuracy of the imperfect interface model is demonstrated by the excellent comparison with experimental results of a GFRP hollow column to built-up beam adhesive connection, under static, cyclic and fatigue loads using a description of damage evolution.

Introduction

In the last decades, fibre reinforced material (FRP) have become more attractive as alternative to traditional construction materials thanks to their excellent mechanical properties such as high tensile strength and resistance to aggressive environments, high strength to weight ratio, simple and rapid installation time [1], [2], [3], [4], [5].

The availability on the market of continuous FRP profiles with constant cross-section such as hollow sections, angle, I-beams and channels, suitable for construction of frame structure, has been guaranteed during the years by the pultrusion process.

In this framework, one of the most promising application in the field of complex structure, is the realization of hollow column to built-up beam connection by using structural adhesives, becoming a valid alternative to the classical bolted connection.

Nowadays, in the civil engineering field, the connections in FRP structures are commonly made using bolted connections excluding a priori the bonding technique. Some FRP profiles manufacturers [6] and design guidelines [7] stipulates that bonded connections should not be allowed for primary load bearing components being the lack of knowledge about and experience with their performance the main reason for their prohibition [6].

Indeed, the most important feature that characterizes the adhesive connections is the absence of holes: the stresses are more uniformly distributed over the bonded surface avoiding the presence of high stress concentration that could damage the fibres and increase the risk of moisture penetration in members.

This lead to consider these type’s connections particularly suitable for the construction of structures in aggressive environments, such as the wind installations in the offshore structures in marine environment.

In this case, the performance of adhesive connections can be influenced by the presence of leading causes of degradation, such as vibrations (due to the waves) represents certainly an aspect to be deeply investigated.

Recently, to improve the confidence that nowadays limits the use of this technology, several authors have undertaken experimental and numerical investigations.

Encouraging results, both from experimental and numerical point of view on the beam-column adhesive connections made in GFRP materials have been obtained by some authors [8], [9], [10], [11], [12], [13].

To investigate the strength, stiffness and fatigue strength of adhesive connections, the response under quasi-static, cycling and fatigue loading have been tested. Furthermore, the connection response under static load has been compared with the behaviour of an analogous bolted connection to demonstrate the better overall performance of the former.

Therefore, it born the need to develop a methodology based on an imperfect interface approach [13] to improve the knowledge of the behaviour of GFRP structures characterized by such connections.

In this paper, a predictive numerical model of the cyclic behaviour has been presented and successfully compared to experimental results.

The model is derived by an asymptotic analysis, due to the thickness of the adhesive layer, of a composite structure made of two elastic solids bonded together by a third thin one, which has a nonlinear behaviour. The adhesive is micro cracked by adopting a Kachanov-type assumption [14], [15]. More in detail, the Kachanov’s theory considers an adhesive layer microcracked characterized by the following assumptions: there is no interaction between the cracks, the stress vector along the crack is assumed constant, and in the stress field, the effect of crack edge is ignored. The present model is able to take into account several adhesive parameters such as thickness, porosity, presence of an initial damage and to describe the damage evolution.

In order to quantify the mechanical parameters such as stiffness of adhesive, initial damage and damage evolution, some experimental tests on adhesive layer in cyclic load condition have been undertaken. The experimental evidence underlines the progressive evolution on the damage at varying of cycle load. Based on the experimental results, the interface model is enhanced.

Finally, we have implemented the model in a finite element software and demonstrated its powerful comparing numerical and experimental results.

Section snippets

GFRP hollow column to built-up beam adhesive connection: Experimental test

In a previous research program deeply described in [11], some of authors have experimentally investigated the behaviour of full scale GFRP connections under static and cyclic load. The investigated connections join a tubular column made of a commercially available hollow GFRP profile with square cross section (120x120x6 mm) and two U-profiles (160x48x8x5 mm) ranged together in the form of a built-up beam. Both the beam and the column are 500 mm long. The members were joined together using an

Imperfect interface model

The proposed imperfect interface model has been developed coupling the homogenization technique and the asymptotic approach within the small perturbation framework [17], [18], [19], [20], [21]. The model considers unilateral contact condition and includes the damage development of the interface.

Following the approach introduced in [22], [23], the thin adhesive interphase placed between the two adherents is considered to be a microcracked material subject to a degradation process.

In detail, the

Characterization of the cyclic behaviour of the adhesive

In order to calibrate the damage parameters of the interface model aforementioned according with Kachanov’s theory, several experimental tests, designed ad hoc, have been conducted at Mechanical and Acoustics Laboratory (LMA) in Marseille (France).

The specimens (Fig. 8) have been constituted by two cylindrical parts in aluminium with diameter equal to 18 mm. The total length of each cylindrical specimen is 120 mm.

Three different adhesive thicknesses have been tested, equal to 1, 5 and 10 mm,

Validation of the theoretical model

The imperfect model presented in Section 3, has been implemented in the commercial finite element software COMSOL Multiphysics. To assess the robustness and accuracy of the model, several simulations have been undertaken on the geometry of cylindrical specimens here presents, and of hollow column to built-up beam adhesive connections. Finally, comparisons between numerical and experimental results have been showed.

Conclusion

The cyclic behaviour of a structural adhesive available on the market has been evaluated by means of experimental tests performed on the universal testing machine. The parameter identified by the experimental evidence have permitted to model a GFRP hollow column to built-up beam adhesive connection under cyclic load by using an advanced imperfect interface model.

The results of the study support the following conclusions:

  • 1)

    The thickness of adhesive layer influences the interface mechanical

Data availability

The data required to reproduce these findings are all reported in the paper.

CRediT authorship contribution statement

M. Lamberti: Investigation, Software, Validation, Data curation, Writing – original draft, Writing – review & editing, Funding acquisition, Project administration. : . A. Maurel-Pantel: Supervision, Investigation, Methodology, Writing – review & editing. F. Lebon: Supervision, Methodology, Writing – review & editing, Funding acquisition. F. Ascione: Supervision, Investigation, Writing – review & editing.

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.

Acknowledgement

This project has received funding from the European Union Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 843218-ASSO (Adhesive connection for Secondary Structures in Offshore wind installations).

References (23)

Cited by (10)

  • A family of models of hard and soft interfaces with damage

    2023, Theoretical and Applied Fracture Mechanics
  • Testing and Theoretical Investigation of the Mechanical Behavior of Adhesively Bonded Butt Joints Under Cyclic Torsional Loading

    2024, Tianjin Daxue Xuebao (Ziran Kexue yu Gongcheng Jishu Ban)/Journal of Tianjin University Science and Technology
View all citing articles on Scopus
View full text