The influence of through-thickness reinforcement geometry and pattern on delamination of fiber-reinforced composites: Part II - Modeling

John Botsis, Véronique Michaud, Georgios Pappas
Elsevier, 2017
Article

Résumé

This article reports modeling techniques and results based on experimental Mode I characterization of a layered woven GFRP material with tufting reinforcement. Standard tuft and loop-less geometries are investigated on three different areal patterns to evaluate their effects on fracture resistance. The experiments reveal that tow/ply and tuft bridging phenomena are present during delamination. The numerical modeling comprises cohesive elements for tow/ply bridging and 1D connector elements for discrete tufts. The traction separation relations for the cohesive model are acquired using an inverse scheme based on strain measurements and the force-separation relations of tuft's failure mechanisms from uniaxial pulling tests on a reference tufting pattern. Results show that tuft's failure mechanism is strongly affected by tufting pattern and geometry while tow/ply bridging contributes 20-30% of the overall fracture energy. The energy absorbed by pull-out of loop-less tufts is approximately twice the fracture energy of standard ones with pull-out triggering much more extensive tow/ply bridging phenomena. (C) 2017 Elsevier Ltd. All rights reserved.

Official source

https://infoscience.epfl.ch/record/232168?ln=fr

À propos de ce résultat

Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.

Concepts associés

Chargement

Publications associées

Chargement

The influence of through-thickness reinforcement geometry and pattern on delamination of fiber-reinforced composites: Part I - Experimental results

John Botsis, Véronique Michaud, Georgios Pappas

This experimental study reports effect of tuft geometry i.e., standard tufts and loops milled-down, and tufting pattern on Mode I interlaminar fracture in a GFRP. Standard tufting geometry is responsible for an increase in fracture resistance by 3.5-6 times than the neat composite, depending on the areal pattern. Based on 5 mm squared tufting pattern on DCB specimens with a full-length release film at laminate mid-plane and non-tufted ones, the individual contribution to fracture of tufts and ply delamination are isolated. For this pattern their superposition matches well the total fracture resistance of the normal tufted composite. Extrapolation of this simple superposition to other patterns underestimates the experimental results. Experiments also show that delamination mechanisms in tufted composites are significantly affected by tuft geometry, with a 65% increased resistance for the loop-less against the standard tufted specimens. This is attributed to the large amount of energy required for tuft pull-out compared to tuft rupture during delamination, which also triggers more extended ply delamination.

Elsevier2018

Chargement

Fiber-reinforced polymer (FRP) composites are increasingly used in many load-bearing structures. Joints are the most critical structural elements as they often become the weak links in engineering structures. Bonding and bolting are two of the main joining techniques used for composite elements but, although bolted and bonded joints have their advantages and disadvantages, adhesively-bonded joints are often preferred. This is thanks to their better performance in terms of cost-effective structures with uniformly distributed stresses without the need for cutting fastener holes that result in high stress concentrations in the FRP adherends at the edges of the holes. In addition, due to the absence of stress concentrations, adhesively-bonded joints exhibit better fatigue behavior. However, the efficient and reliable use of adhesively-bonded joints requires a thorough understanding of their failure mechanisms under mixed-mode quasi-static and fatigue loading conditions. In real structural applications, failure of adhesively-bonded joints occurs due to crack initiation and propagation under varying mixed-mode ratios. Since it is difficult to describe this progressive fracture in a structural joint, fracture joints with constant mode-mixity ratios are used to determine the strain energy release rates for crack initiation and propagation. Subsequently, mixed-mode failure criteria based on the determined strain energy release rates can be established and used to model/predict the fatigue and fracture behaviors of structural joints. The main objectives of this thesis were to understand the fracture and fatigue behaviors of adhesively-bonded FRP fracture joints under Mode I, Mode II, and mixed-Mode I/II loading conditions and, consequently, develop mixed-mode static and fatigue failure criteria applicable to structural joints. To fulfill these objectives a combined experimental-analytical-numerical study was undertaken. Double cantilever beam (DCB-Mode I), end-load split (ELS-Mode II), and mixed-mode bending (MMB-Mode I/II) specimens were examined under quasi-static and constant amplitude fatigue loading. Analytical approaches were used to determine the strain energy release rates under quasi-static and fatigue loading, whereas phenomenological models were developed to characterize the fatigue and fracture behaviors of the joints. A new approach designated the “extended global method” was established and applied for the analysis of the quasi-static and fatigue experimental data and the fracture mode partitioning in the mixed-mode experiments. In parallel, non-linear finite element models were developed to simulate the quasi-static fracture behavior and investigate the effects of both the existing asymmetry and the observed fiber bridging on the fracture behavior of the examined joints. The bridging zone was modeled by zero-thickness cohesive elements. An exponential traction-separation description of the cohesive zone model was shown to appropriately model the fiber bridging and calculate its contribution to the fracture energy under quasi-static loading. The FE models were successfully employed for the separation of the fracture parameters, i.e. strain energy released at the crack tip (Gtip) and the amount of energy released along the crack-bridging zone (Gbr), in the quasi-static mixed-mode failure criterion for crack propagation. In addition to the studies on the quasi-static fracture behavior of the joints, a new phenomenological fatigue crack growth formulation for the modeling and prediction of Rratio effects on the Mode I fatigue behavior of adhesively-bonded joints was also derived. The model was used for prediction of the fatigue behavior of the examined joints under different R-ratios. The results proved that the model could accurately simulate the fatigue behavior of the examined joints and was also capable of predicting behavior exhibited under unseen loading conditions, i.e. the R-ratios used for estimating the model parameters were different from those used for validating its prediction capability. The experimental results and numerical analyses combined with the phenomenological models were used to establish mixed-mode static and fatigue failure criteria for crack initiation and crack propagation for the adhesively-bonded joints. The derived mixed-mode failure criteria can be used for simulating/predicting the crack initiation and progressive crack propagation in structural joints comprising the same adhesive and adherends.

EPFL2013

Chargement

Specimen thickness dependence of large scale fiber bridging in mode I interlaminar fracture of carbon epoxy composite

John Botsis, Joël Cugnoni, Ebrahim Farmand Ashtiani

Large scale bridging is an important toughening mechanism in composite laminates that depends on the specimen geometry as well as the constituent materials. In this work, an iterative approach is applied to identify the effect of specimen thickness on traction-separation behavior in the bridging zone. Double cantilever beam specimens (DCB) with embedded arrays of wavelength-multiplexed fiber Bragg grating (FBG) sensors are subjected to monotonic mode I fracture loading. As processed specimens with thicknesses h = 2, 4, 8 and 10 mm as well as specimens of h = 4 mm milled down from the 8 and 10 mm are tested. Non-homogeneous strain distributions in the vicinity of the interlaminar crack plane are locally monitored by means of embedded FBGs. The measured strain data are used to quantify the bridging tractions associated with each specimen thickness and consequently the energy release rate (ERR) due to the bridging. The results show that the bridging zone length and the ERR at the plateau level increase with specimen thickness while the results for all specimens with h = 4 are the same thus excluding any processing effects. Moreover, the maximum stress of traction-opening in the bridging zone and crack opening displacement at the end of the zone are independent of thickness. In contrast, the rate of tractions decay depends on the specimen thickness in that it decreases with thickness scaling. Thus the so-called bridging law is not a material parameter. The identified traction-separation relations are employed in a cohesive zone model to predict fracture of DCB specimens with different thicknesses. (C) 2014 Elsevier Ltd. All rights reserved.

Elsevier2015

Chargement

EPFL2013