Multi-scale study of the adhesion between flax fibers and biobased thermoset matrices

Accepted Manuscript Multi-scale study of the adhesion between flax fibers and biobased thermoset matrices L. Marrot, A. Bourmaud, P. Bono, C. Baley PII: DOI: Reference:

S0261-3069(14)00362-8 http://dx.doi.org/10.1016/j.matdes.2014.04.087 JMAD 6478

To appear in:

Materials and Design

Received Date: Accepted Date:

7 March 2014 30 April 2014

Please cite this article as: Marrot, L., Bourmaud, A., Bono, P., Baley, C., Multi-scale study of the adhesion between flax fibers and biobased thermoset matrices, Materials and Design (2014), doi: http://dx.doi.org/10.1016/j.matdes. 2014.04.087

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Multi-scale study of the adhesion between flax fibers and biobased thermoset matrices a,b

b

a

b

L. Marrot , A. Bourmaud , P. Bono , C. Baley a

®

Fibres Recherche Développement , Technopole de l’Aube en Champagne,

Hôtel de Bureaux 2, 2 rue Gustave Eiffel, CS 90601, 10901 Troyes Cedex 9, France Corresponding author: [email protected], +33 (0)2 97 87 45 07 [email protected], +33 (0)3 25 83 41 90 b

Laboratoire d’Ingénierie et Matériaux de Bretagne (UBS), Université Européenne de Bretagne, Centre de Recherche Christiaan Huygens, Rue Saint Maudé, 56321, Lorient, France [email protected], +33 (0)2 97 87 45 18 [email protected], +33 (0)2 97 87 45 53

Abstract The environmental impact of composite materials made with a thermoset matrix can be reduced in two ways. First, glass fibers can be replaced by natural fibers. Second, petrochemical components from the matrix can be replaced by biobased renewable equivalents. The quality of the interface between the matrix and the fibers has a strong influence on the composite mechanical properties. In this study, tensile performances of flax fibers and commercially partly biobased epoxy and polyester matrices have been investigated and corresponding unidirectional composites were elaborated. Their mechanical performances are in accordance with fiber and matrices properties, taking into account fiber dispersion. Then, at the microscopic scale, the debonding test was used; a great adhesion between flax fiber and thermoset matrices was highlighted. Finally, tensile tests on ± 45° laminates were carried out to create an in-plane shear at the macroscopic scale. Interestingly, the results obtained at the macroscopic scale are well correlated to the ones given by the debonding test at the microscopic scale. Keywords: adhesion, flax fiber, biobased thermoset resin, in plane shear, composite

1

1 Introduction In order to develop high performance composite materials at a moderate cost, the choice of the matrix is oriented towards a thermoset matrix. The environmental impact of composite materials made with a thermoset matrix can be reduced in two ways. First, glass fibers can be replaced by natural fibers. A recent study showed that natural textiles materials are now suitable for engineering applications [1]. Amongst the natural fibers, flax fibers, in particular, have been investigated and have been shown to be competitive, owing to their good mechanical properties [2][3][4], low density, and wide availability in France [5]. In the stem, fibers are gathered into bundles and have a polygonal section. Fibers are linked to one another by a middle lamella constituted of pectins, hemicelluloses, lignins, and wax. Flax fibers are made up of two (primary and secondary) concentric cell walls with a void in the middle (lumen), and the secondary wall itself is divided into three layers. The external primary wall is mainly constituted of pectins [6] with some poor crystallized cellulose [7]. The thickest layer within the secondary wall (S2) consists of cellulose microfibrils which are well organized and oriented according to an angle called the microfibrillar angle (MFA). This layer plays an important role in the mechanical behavior of the fiber. The cellulose in flax cell wall is embedded in an amorphous matrix composed of hemicelluloses and pectins. Additionally, all the components which may be found in the fiber composition contain hydroxyl groups. They will play an important role in the linkage with the matrix. Beside the reinforcement, the second way to improve the environmental impact of thermoset composites is to replace petrochemical components from the matrix with biobased renewable equivalents. Amongst the thermoset resins, epoxy and polyester resins are the most commonly used for high-performance composites and their association with flax fibers allows to obtain efficient composites [8][9][10]. The main drawbacks for the use of thermoset resins are their difficult recyclability and the toxicity of their constituents. The polyester resin, usually a solvent, is the styrene, which is a hazardous air pollutant and volatile organic compound. During the epoxy resin common synthesis, an epoxy precursor reacts with an amine or acid anhydride hardener. In order to reduce the environmental impact of the thermoset resins made with petroleum reserves, alternative renewable resources are available. The renewable substitutes may be either the epoxy precursors or the hardeners. Epoxidized oils can be good candidates as epoxy precursors. Since 1996, natural triglyceride oils have been used as a basis for polymers, adhesives, and composite materials [11][12].

2

On the one hand, the epoxidized triglyceride oil can react with an amine or anhydride hardener to produce an epoxy matrix resin. On the other hand, in the polyester field, the triglyceride oil can be transformed into fatty acid monomers, which are alternatives to styrene because of their low cost and low volatility [13]. Biobased thermoset resins are often mentioned as poor mechanical property resins. Indeed, because of the low reactivity of epoxy groups and the tendency for intramolecular bonding, any epoxidized oil leads to poorly cross-linked materials with limited thermal and mechanical properties [14][15][16]. Once the matrix and the reinforcement have been selected, the adhesion between these two components is crucial to determine the composite performance. Indeed, a strong adhesion at the interface is needed for an effective transfer of stress and load distribution throughout the interface. In a thermoplastic matrix, the link between fiber and matrix is ensured by physical interactions whereas in a thermoset matrix, the adhesion results from chemical bonding between both components [17][18]. The microbond test [19] is commonly used to determine the strength of the interfacial bonding. The interfacial shear strength (IFSS) has been evaluated via this test by different authors for flax fibers / petrochemical epoxy systems [20][21][22] and flax fibers / petrochemical polyester systems [10]. Several macroscopic mechanical tests [23] are sensitive to the interfacial properties. These tests include transverse tensile, in-plane ± 45 tension, Iosipescu, and short beam shear tests. Concerning biocomposites, only a few studies have used these tests. Meredith et al. [24] used the short beam test on flax/epoxy pre-pregs. They found interlaminar shear strength (ILSS) between 10.7 and 23.3 MPa depending on the flax fabrics. Yongli et al. [25] prepared flax/glass fiber reinforced hybrid composites with a phenolic matrix, with varying fractions of flax and glass. They measured the ILSS by the short beam strength test for their different hybrid composites. The lowest ILSS (19.35 MPa) was found for the glass fiber reinforced composite; the flax fiber composite exhibited an ILSS of 24.45 MPa, and the hybrid glass/flax fiber composite obtained the highest ILSS at 31.12 MPa, owing to bridging between glass fibers and flax fibers. There are even fewer data regarding the shear properties measured by the in-plane shear test on ± 45 laminates. Baley [26] used the in-plane shear test on ± 45 flax/polyester and glass/polyester laminates. Le Duigou et al. [27] used the same test for flax/PLLA laminates to evaluate the influence of the thermal history during the fabrication process.

3

The purpose of this article is to establish whether partially biobased epoxy and polyester resins can compete with their petrochemical equivalent in flax-reinforced biocomposites. The mechanical properties of the reinforcement and biobased thermoset resins have been investigated by tensile testing. Unidirectional composites are manufactured by compression molding with several biobased matrices. Their properties are evaluated under tension. In addition, a multi-scale analysis of the adhesion between matrices and flax fibers is considered. At the micro scale, the IFSS is determined via the debonding test, and results are correlated to the ILSS obtained via the in-plane shear test on ± 45 laminates.

2 Experimental details 2.1 Flax fibers The flax fibers, harvested in 2009 and belonging to the Melina variety, were supplied by the La Calira company (Picardie, France). Fibers were dew-retted to help fiber extraction and then scutched by the flax producer. No further treatment was applied to the fibers. They were manually extracted for tensile and debonding tests. Under the form of unidirectional tapes of untwisted yarns, they were sewn 2

together with cotton thread (Figure 1). Their basic weight was around 250 g/m . They were also used for making the composite samples for the in-plane shear test (Figure 1).

2.2 Tensile testing on single fibers Single fibers were manually extracted and glued on cardboard supports with an elliptical window. Tensile tests on single flax fibers were carried out at a controlled temperature (23°C) and relative humidity (48%), and longitudinal mechanical properties (Young’s modulus, ultimate strength and failure strain) were determined. Due to the short fiber length (about 20–30 mm), a gauge length of 10 mm was chosen. The fiber was clamped on a universal MTS type tensile testing machine equipped with a 2 N capacity load cell and loaded at a constant crosshead displacement rate of 1 mm/min up to rupture. The determination of the mechanical properties was made in accordance with the XP T 25501-2 standard [28] which takes into account the compliance of the loading frame. The mechanical properties result from 71 fibers. Before the tensile test, the diameter of every fiber was measured with an optical microscope. The diameter is the mean of three measurements at different locations on the fiber.

4

2.3 Epoxy and polyester matrices The study focused on six commercialized epoxy and polyester matrices. Given that each main resin producer has its green range of matrix, only matrices with renewable carbon content higher than 50% were retained as “partially biobased matrix”. A petrochemical reference was also selected. Table 1 shows the matrices under study and their characteristics (bio-content, origin of bio-content and curing specifications). The bio-content is not always measured according to the same method. The U.S. standard ASTM: D6866-12, based on the carbon dating method, enables the determination of the renewable carbon content. The bio-content can also be calculated by the mass ratio between renewable components and the total matrix weight.

2.4 Samples preparation 2.4.1 Matrix plates For each matrix, a plate was cast between two steel sheets separated by a silicon join in order to obtain a 3-4 mm thick matrix plate. Once cast, the mold was set in an oven to be cured in accordance with the cycle advised by each resin manufacturer (see Table 1). Then, specimens were milled for the tensile mechanical tests.

2.4.2 Samples preparation for in-plane shear test Specimens for in-plane shear were prepared according to ASTM: D3518 with the different matrices reinforced by (±45) flax layers. The fiber volume content was around 27%. The shear test standard requires a [45/-45]ns stacking sequence with 2 < n < 4, and at least 8 reinforcement layers to limit tension-flexion coupling [29] and increase the interlaminar area. Laminates were made with 8 reinforcement layers under compression molding. Liquid resin has been set between each flax ply in order to guarantee a homogeneous stratification. Rectangular specimens were milled with the 3

following dimensions: (25 x 180 x 3) mm (Figure 1).

5

2.4.3 Unidirectional composites Composites were manufactured according to the molding process described by Coroller et al. [20]. Fibers were impregnated with the epoxy or polyester resin. They were then put into a 100 mm long 2

aluminium mold with a 6 x 2 mm section opened on each side. The small cross section of the mold induced the unidirectional resin flow during the compression and preferential orientation of the fibers in the composite. Three different fiber volume fractions (Vf = 20%, 30%, 40%) were systematically targeted. The control of fiber content is described below. Ten samples were required for each fiber volume content. The mold was set in an oven to be cured in accordance with the cycle advised by each resin manufacturer. After complete curing, fiber glass tabs were glued to each composite extremity with an Araldite adhesive in order to reduce the risk of breakage in the jaws during the tensile test.

2.5 Characterization 2.5.1 Debonding test The droplets were placed on the flax fibers using a single glass fiber which had been dipped in the liquid resin. Fibers with droplets were placed in an oven to cure according to the specifications of each resin supplier (Table 1). Microbond specimens were then checked under the microscope to control the droplet geometry. Samples with defects (kink bands on the fiber or lack of symmetry of the droplet) were systematically rejected. Besides being symmetrical, microdroplets needed to be smaller than 150 µm length, otherwise the fiber breaks when loaded. Droplet length and height and fiber diameter at both extremities were measured for the selected samples. Then, the flax fiber with the microdroplet was mounted in the shearing device and continuously observed under a microscope. The fiber was pulled out of the droplet while the latter was constrained by the knife edges. The loading rate during debonding was 0.1 mm/min. Force–displacement plots were recorded for each specimen, in order to determine the debonding force and the friction force. At least 20 specimens were tested for each matrix. The interfacial shearing rupture mode was checked by an observation of the debonded droplet.

6

2.5.2 Matrix and unidirectional composites tensile testing and in-plane shear test Tensile testing on both matrices and UD composites was carried out on the apparatus MTS Synergie RT1000 (MTS, Eden Prairie, MN, USA). The laboratory atmosphere was controlled at a temperature of 23°C and a humidity of 48%. Tensile testing followed EN ISO 527-5 [30] instructions. The loading speed was 1 mm/min. A MTS Extensometer was used with a nominal length of 25 mm. The tests were carried out at least five times for each specimen, and the results were averaged arithmetically. In-plane shear testing was carried out according to ASTM: D3518 on the tensile machine described above. The crosshead displacement rate was 2 mm/min up to rupture. A biaxial extensometer measured the sample width and length variation during the test. At least five samples were tested for each matrix. This test provides shear strength τ12, shear strain ɣ12 and shear modulus G12. G12 is determined from the slope of the plot of shear stress versus shear strain for ɣ12 between 0 and 0.2%. The equations (1), (2), (3), give the different relations for calculating G12, τ12 and ɣ12: G12 = τ12/γ12

(1)

τ12 = σx/2

(2)

σx is the applied strength. 0

0

γ12 = (εx -εy )

(3)

Under load, the fibers rotate and the rotation reaches 1° for an axial strain of 2% [27]. According to the ASTM: D3518 standard, if the shear strain is below 5%, the shear stress can be taken at the maximum value: σx = σmax. If not, the shear stress corresponding to a 5% shear strain is used: σx = σɣ=5%.

2.5.3 Density and fiber fraction in composites The density of fibers, matrices and composites was measured with MS-DNY-54 Toledo Mettler balance. If there are no porosities, the density can be deduced by the equation (4) by weighing the sample successively in air and in ethanol. ρsample = (Mair (ρair - ρethanol ))/(Mair - Methanol ) + ρethanol

(4)

The fiber volume content was deduced from density according to the equation (5).

7

Vf = (ρcomposite - ρmatrix)/(ρfiber - ρmatrix )

(5)

2.5.4 Differential scanning calorimetry Thermal analysis was performed using a Mettler Toledo DSC822 on solid and liquid matrix samples weighing around 10 mg. These were heated from 25 to 250 °C at a 10°C/min rate. By integrating the DSC slopes, heat of cure of the liquid resin (ΔHuncured) material and that of the partially cured resin (ΔHcured) were determined. The percentage of cure was calculated following the equation (6). % Cure= (∆Huncured - ∆Hcured)/∆Huncured

(6)

where ΔHuncured and ΔHcured are heat of cure of liquid and cured matrices respectively.

2.5.5 Scanning Electron Microscopy pictures The fibers were sputter-coated with a thin layer of gold in an Edwards Sputter Coater and then observed under a Jeol JSM 6460LV Scanning Electron Microscope (SEM).

3 Results and discussion 3.1 Mechanical analysis of flax fibers The flax fibers showed a longitudinal Young’s modulus of 54.7 ± 11.7 GPa. The tensile strength at failure was 856 ± 354 MPa and the tensile strain was 1.8 ± 0.8 %. These values are in accordance with the data usually found in the literature [2][31][32]. A reason for these high mechanical properties is the relatively small diameters of the flax fibers which were used (11.8 ± 2.9 µm on average). Indeed, Young’s modulus and tensile strength are dependent on the fiber diameter. The decrease of the tensile strength according to the fiber diameter is frequently observed on natural fibers, flax included [2][33][34][35][36][37]. The diameter dependent decrease is explained by the weakest link theory [38]. The larger the diameter is, the higher the probability that a critical defect leads to rupture. As it is often the case with natural fibers, the studied flax fiber tensile properties are scattered. A large number of parameters induce the scattering of the mechanical properties. Amongst the uncontrollable parameters are the position of the fiber in the stem [39][4], the fiber biochemical composition [40], the fiber microfibril angle, and the fiber diameter, as seen before. Another reason for the dispersion, besides the natural origin of flax fibers, is the uncertainty due to the fiber-section measurement [41]. Lefeuvre et al. [42] highlighted that this measurement was responsible for 32% of the standard

8

deviation of Young’s modulus and 23% for the strength at rupture. In addition to the high mechanical properties found here, these flax fibers exhibit a low density of 1.39 ± 0.03. Knowing that the density of glass fibers stands around 2.5, this comparatively low value makes flax fibers very attractive for the reinforcement of composites, for which, the weight has to be as light as possible. The properties of the matrices will now be considered.

3.2 Tensile mechanical performances of the matrix Matrices were cured according to the specifications described in Table 1. As these specifications were recommended by their respective manufacturers, they are not identical. However, the complete curing was checked by DSC analysis (see Table 1). Each matrix was cured at more than 94% after the curing specifications. Tensile plots are shown in Figure 2, and values of Young’s modulus, strength at failure and strain at failure are compared in Figure 3. The Young’s modulus is taken at the beginning of the curve (Figure 2), in the linear part for strain between 0,05 % and 0,25 %. The best performances in terms of Young’s modulus and tensile strength are attributed to the petrochemical matrix for both types of matrices. PetroEP exhibits Young’s modulus of 3.3 ± 0.1 GPa and a tensile strength of 77 ± 3 MPa. BioEPA and BioEPC resins show properties close together and almost as high as those of the petrochemical matrix, although their tensile strengths are slightly lower (52 ± 1 and 57 ± 5 MPa respectively). These results are not surprising because these two partially biobased matrices are based on the same components as PetroEP, i.e. epichlorhydrine and Bisphenol A, and use an amine hardener. The only difference is that BioEPA’s epichlorhydrine and BioEPC’s Bisphenol A are biobased. BioEPC also has pine oil in its epoxy precursor, which can explain the lower properties. Indeed, within the epoxidized oils, epoxy groups are located in the middle of the fatty chains (Figure 4). This position is not easily accessible for the hardener, because of the repulsion from the apolar carbon chains. This configuration explains the slow reticulation of the resins using epoxidized vegetal oil only as an epoxy precursor, and their lower tridimensional network density. Added to the tendency for intramolecular bonding, epoxidized oils have a low reactivity. Several authors [14][15][16] reported that they lead to materials with limited mechanical properties. BioEPA might have vegetal oil in its composition as well, but it is not specified by the manufacturer. On the contrary, the epoxy functions in the DGEBA (epoxy

9

precursor of the petrochemical matrix) are located at the extremities of the molecule and are easily accessible (Figure 5). Tensile strength of unmodified DGEBA cured with amine curing agents is about 62 MPa, and typical values of the modulus of elasticity are between 2.8 and 4.1 GPa [43]. BioEPB shows the lowest properties (E = 1.7 ± 0.2 GPa, σ = 31 ± 1 MPa). The use of an amide hardener requires explanation. Indeed, it is well known in chemistry that amides have a lower reactivity than secondary amines, which have themselves a lower reactivity than primary amines [44]. The lower reactivity of amides leads to a lower cross-linked matrix with lower properties. Nevertheless, amide hardeners remain interesting because they allow diversifying resources; they can be made from biomass (castor oil) and are generally nontoxic, unlike some aromatic amines. For each epoxy matrix under study, the tensile strain is superior to 3%, hence higher than the flax fiber tensile strain (around 2%). This is a favorable condition for using flax fibers as reinforcement in composites made with these matrices. On the contrary, the tensile strain of the polyester matrices under study is lower than the tensile strain of flax fibers, which may be problematic for a composite. The mechanical properties of composites are governed not only by the characteristics of their constituents, fiber and matrix, which have both been considered, but also by the interface. Now the interface at the microscopic scale will be examine.

3.3 Adhesion at the microscopic scale Vegetal fibers are mainly constituted of cellulose and hemicelluloses. These components have hydroxyls groups on their surface, which can be linked via hydrogen and Van der Waals bonds to amine and epoxy groups of the epoxy resins, and to ester groups of the polyester resins. The quantification of the fiber/matrix link corresponds to the adhesion measurement. The debonding test was chosen to determine the adhesion of the several flax/epoxy and polyester systems. The value of the debonding force Fmax enables the calculation of the interfacial shear strength (IFSS). The equation (7) describes how to calculate the IFSS according to the Miller method [19]. A uniform distribution of the stress along the fiber/matrix interface is assumed. IFSS=Fmax/(2πrL)

(7)

10

where r is the fiber radius at the droplet extremities, L is the droplet length, and the hypothesis of a circular section for the fiber is taken. The interfacial shear strength (IFSS) is determined in two ways: first from the mean of the IFSS values, and second from a linear regression of the plot of debonding force versus the sample bonded area. IFSS values for the various systems flax/epoxy and flax/polyester are shown in Table 2 and compared with the values found in the literature data. 2

For each system, both results of IFSS fit well, and the linear regression correlation is significant (R

>0.65). For flax/epoxy systems, the IFSS stands between 11.9 ± 4.1 MPa and 28.5 ± 6.4 MPa. These values are in accordance with the literature (IFSS between 16.1 [21] and 22.9 MPa [22] for a flax/epoxy system). The flax/BioEPC system shows the highest IFSS (28.5 MPa). Then, the flax/PetroEP system and the flax/BioEPA system have lower values of IFSS, close to one another (20.4 and 17 MPa respectively). Finally, the flax/BioEPB system has the lowest IFSS value (11.9 MPa). The adhesion is strongly dependent on the chemical bonding which takes place at the fiber/matrix interface. Thus, matrix components have a direct influence on microscopic adhesion. For both PetroEP and BioEPA resins, components are similar with a polyamine hardener, so unsurprisingly, their adhesions on a flax fiber are in the same range. The BioEPC resin manufacturer stresses the high hydroxyl groups content in its resin, thanks to the aliphatic amine hardener. Hydrogen bonds between hydroxyl groups from the matrix with those of the fiber can partly explain the high adhesion obtained for the flax/BioEPC system. Finally, the flax/BioEPB system had the lowest IFSS (11.9 MPa). Zinck et al [46] showed that the amine or anhydride nature of the hardener in an epoxy resin formulation had an influence on the interfacial properties with a glass fiber. The BioEPB resin uses an amide hardener whereas all the other epoxy resins under study use an amine hardener. The lower reactivity of amides compared to amines results in less rigid resins, which has consequences on the adhesion. The adhesion of the flax/epoxy system is slightly lower than the adhesion between glass and epoxy (29.3 [45] - 37.2 MPa [20]). However, flax fibers undergo no surface treatment to improve the adhesion, unlike glass fibers. The adhesion between flax and PetroUP gives a value in accordance with the literature (17.5 ± 7.3 MPa) compared with 14.2 ± 0.4 MPa for Le Duigou et al. [21]). The glass/polyester system shows almost the same IFSS (15.7 MPa) according to Baley et al. [45]). Unfortunately, microdroplets remained undercured with the biobased resin. Moreover, a large standard

11

deviation was noticed for the flax/polyester system. A source of scattering and undercuring is the method used to prepare polyester resin. Basically, to crosslink the polyester resins, a small amount of accelerator (0.3%) and catalyst (1.5%) is added to the resin. Even if the mixing is realized with 150 g of resin, a perfect stoichiometry cannot be guaranteed for every droplet. Epoxy resins are cured by adding about one third of hardener to the resin, which provides a more homogeneous mixture and droplets. However, some authors [47] have observed the vaporization of the hardener during the cure of carbon/epoxy microdroplets. Then, local variation in stoichiometry within the droplet is possible as a result of the mixing. Microscopic adhesions of the flax/epoxy and flax/polyester systems are satisfactory with petrochemical and partially biobased resins. Nevertheless, within a composite, other parameters, such as the shape factor, alignment or distribution of the reinforcement fiber enter into account for the interface problem. Additional tests are required to analyze the influence of the microscopic IFSS at the macroscopic scale, and to determine whether a good microscopic adhesion is enough to guarantee a good interface and to obtain high quality composites. The performance of unidirectional composites made with petrochemical and partially biobased matrices are investigated in the following section of the study.

3.4 Unidirectional composites Unidirectional (UD) composites were manufactured using flax fibers and the different petrochemical and biobased matrices. To provide the best properties, the reinforcement has to be as straight as possible, and homogeneously distributed in the matrix. As explained in the introduction section, in the stem, fibers are initially gathered into bundles of fibers, linked together by the middle lamellae. Once extracted, fibers are more or less separated and can be found in the form of unitary fibers or bundles in the composites. Bundles have lower longitudinal mechanical properties than single fibers [20][48][49]. In this study, only one type of fiber was used, and the influence of the type of matrix was investigated. The typical tensile stress–strain plot for UD composites is shown in Figure 6. The particular evolution of the plot, nonlinear at the beginning, has been reported and described by several authors [20][50][51]. UD flax fiber’s reinforced composites behave like the unitary flax fibers under tensile load, even if the composite behavior does not depend on the fibers only. This type of curve is specific to

12

2

composites of a small size (section of 6x2 mm ) manufactured by compression molding, where porosity is insignificant and fibers are well distributed. Figure 7 shows a UD composite cross section observed by SEM. No porosity is seen. Figures 8 a and b present the mechanical properties of the UD composites according to the fiber volume content. The rule of mixture is plotted for Young’s modulus and for the tensile strength at failure with the PetroEP properties following the equations (8) and (9). For polyester matrices, the rule of mixture for the tensile strength at failure is given by the equation (10). EL,UD = EL,f × Vf + (1-Vf) × Em

(8)

where EL,UD and EL,f are the mean Young’s modulus of the UD composite and the flax fibers respectively. For epoxy matrices, the mean tensile strain of the fibers is lower than the tensile strain of the matrix: εL,f < εm :

σUD = k1 × k2 × σL,f × Vf + (1-Vf) × σL,f × (Em/EL,f)

(9)

For polyester matrices, the mean tensile strain of the fibers is higher than the tensile strain of the matrix: εL,f > εm :

σUD = k1 × k2 × σm × (EL,f / Em) × Vf + (1-Vf) × σm

(10)

where k1 is the efficiency factor and k2 is the orientation factor. These two factors are introduced to balance the ideal law in which fiber orientation is optimal (straight), repartition of fibers is uniform and fibers are continuous. In this study, unidirectional fibers were used, so k2 equals 1. On the one hand, regardless of the matrix, the longitudinal Young’s modulus of the UD composites is close to the rule of mixture (see Figure 8 a and Table 3), though slightly below. Considering the standard deviation, there is no clear tendency to classify the matrices on the Young’s modulus of their composites. The contribution of fibers is dominating. On the other hand, for each type of composite, the tensile strength, plotted in Figure 8 b, is lower than the values which could be expected by the ideal rule of mixture in which k1 would be 1. These results below the rule of mixture can be explained in terms of several parameters. First, even if the quality of the composites is good, inhomogeneous zones can create early failures. Second, the strength is strongly dependent on the fiber

13

individualization, as highlighted by Coroller et al. [20]. Effective values of strength and values predicted by the ideal rule of mixtures are reported in Table 3. k1 is calculated for all the matrices. Efficiency factors stand between 0.55 and 0.84, except for three marginally weak values around 0.40. For sake of clarity, Table 3 focus only on one fiber volumic fraction for each batch of composites; nevertheless, the whole data are presented on Figure 8. The gap between real and ideal values expected by the rule of mixture can be explained by the poor separation and distribution of the fibers within the composite. The values of efficiency factors are in the same range as these found by Coroller et al. [20] (0.650.72). The level of individualization of the flax fibers in the UD composites can be seen in Figure 7. Most of the fibers are still gathered into bundles. Some areas are free of fibers, but there is no porosity. The separation of the fibers could be improved either by deepening the retting of the flax fibers, by hackling the fibers after scutching, or by applying a soft treatment to the fibers before resin impregnation. For example, Bourmaud et al. [31] made the extraction easier without any damage or biochemical structure modification by soaking the fibers in water for 72 hrs. Strength data show a high scatter, which makes the interpretation difficult. The composites realized with the petrochemical matrices (epoxy and polyester respectively) show the highest mean values of tensile strength. BioEPB presents the lowest mean tensile strength. Table 3 shows the results on UD composites with petrochemical matrices and some literature data as a comparison. When considering the flax/epoxy system, for the same fiber content, the results of this study fit well with the literature data in terms of Young’s modulus. In this study, the strength is almost twice as high as that found by Oksman [52] or Van de Weyenberg et al. [53]. For Vf = 32%, a tensile strength of 224 ± 18 MPa was obtained, compared with 132 ± 4.5 MPa for Oksman. For fiber content around 40%, the strength was 253 ± 9 MPa compared with 133 MPa according to Weyenberg. Such a difference may come from the process used to manufacture the UD composites and from the volume of the composite specimens. Indeed, it is generally assumed that every part of the material behaves in the same way, and so volume-averaged properties can be measured over the domain of the material tested. It is appropriate for properties such as the elastic modulus, whose value should not vary with specimen size. However, the strength is different because failure tends to initiate from defects in the material. Then strengths will be a function of the extreme values rather than the means of the distributions of local strengths, and can be expected to reduce with increasing the volume of material tested [55]. The quoted authors

14

manufactured their composites by resin transfer molding or autoclave under vacuum respectively. Although in this study the same process as Baley et al. [50] and Coroller et al. [20] was used (wet impregnation and compression molding), a lower strength was obtained, due to the poor fiber individualization and distribution, as explained above. According to the literature, glass/epoxy systems provide, for an equivalent fiber volume content, about the same Young’s modulus as flax/epoxy systems, but their tensile strength is twice as high (817 ± 35 MPa for Vf = 48% according to Oksman [52]). They also get longer before breaking. Concerning flax/polyester systems, the results are in accordance with the literature [54]. The large scatter of the results in terms of Young’s modulus and tensile strength on UD composites does not enable us to positively differentiate one matrix from another. In fact, the UD composite break is governed by the rupture of the fibers, which itself is very scattered (see Section 3.1). In the following section, the shear performances at the macroscopic scale will be considered.

3.5 In-plane shear test on ±45 laminates The ± 45 laminates were cured according to the specifications of the resin manufacturers (Table 1). Then, the results can be compared with the tensile performance of the matrix on the one hand, and with the microscopic adhesion found with the debonding test on the other hand. The in-plane shear test is sensitive to both the fiber/matrix interface and the matrix properties themselves [27]. Fiber volume content (Vf) for the composites manufactured with the various matrices for the in-plane shear test are given in Table 4, and the results obtained in this study compared with those in the literature data. Vf are found in the same range (24-27%) so the results can be compared together. The flax/epoxy systems show shear strengths between 35 and 48 MPa and shear modulus between 1474 and 2507. These results are in accordance with the literature [26]. The highest shear strength (48 ± 8 GPa) recorded for flax/ PetroEP system is as good as the shear strength reported by Baley [26] for glass/epoxy composites. The flax/polyester composites show shear strengths between 27 and 39 MPa. This is slightly higher than the values reported by Baley [26] for flax/polyester composites. For the epoxy systems, the best value of shear strength is attributed to flax/PetroEP and flax/BioEPC composites, then flax/BioEPA composites with very scattered values, and finally flax/bioEPB composites.

15

Figure 9 shows the shear strength of the ±45 laminates vs. the microscopic IFSS given by the microbond test. It is clear that the higher the IFSS, the better the shear strength. Nevertheless, the improvement of shear strength is not proportional to the improvement of IFSS. For composites with epoxy matrices, the shear strength to IFSS ratio is between 1.6 for BioEPC and 3.1 for BioEPB. Indeed, the in-plane shear test is sensitive to the fiber matrix adhesion, but also to the matrix properties itself. Since the matrix is different for each system, these two parameters enter into account for the resulting shear strength. In the literature, several authors tried to correlate mechanical properties measured by micro and macromechanical tests on various systems. Park et Kim [56] showed that a sizing agent increases the surface energy of glass fibers, which improves the ILSS and mode II fracture toughness of glass/polyester composites. Mäder [57] and Keusch et al [58][59]studied the influence of fiber sizing by following the evolution of IFSS measured by pull-out and the corresponding glass/epoxy and carbon/epoxy composite properties with transverse tension and short beam shear tests. The effect of IFSS on macroscopical shear properties has particularly been studied on carbon/epoxy systems. Drzal et Madhukar [60] investigated the relationships between fiber-matrix adhesion determined by fragmentation tests with the in-plane and interlaminar shear properties of graphite/epoxy composites. ±45° tension, Iosipescu, and short beam shear tests were conducted on three identical sets of composites differing only in their fiber-matrix interfacial shear strength. Their experimental results showed a strong sensitivity of the IFSS to both in-plane and interlaminar shear strengths. Unfortunately, the combination in the failure modes complicates interpretations. Only few authors carried out this type of two scale experiment on biocomposites. Van de Weyenberg [22] related the flax/epoxy IFSS measured by the debonding test to the transverse flexural strength found by the 3 point bending test. Le Duigou et al [27] correlated in-plane shear and mode I interlaminar fracture testing with microdroplet interface debonding for flax reinforced PLLA composites. Figures 10 and 11 present the fracture surfaces for the flax/BioEPB and flax/BioEPC laminate composites, which have, respectively, the lowest and the highest IFSS. For both systems the fracture surface analysis revealed that the failure mode is a combination of interfacial and matrix failure, with a predominance for the interfacial failure. Indeed, areas of the matrix with impressions of adjacent layers are typical of interfacial failure. However, the presence of debris at the fiber’s surface indicates a

16

matrix failure mode. A highly damaged fiber is visible in Figure 11. The higher level of adhesion between fiber and matrix in the BioEPC composite made the failure more violent. Damage induced in the composite during in plane shear test are complex due to the presence of both elementary fibers and fibers gathered into bundles, as evidenced in Figure 10. Consequently, failure can occur at several levels, in the matrix, at the fiber matrix interface or inside a fiber bundle. In the case of the microbond test, only one fiber embedded in the resin is implicated. Moreover, the microbond test does not involve the same volume ratio of fibers and matrix as in a macroscopic composite, and the stoichiometry of matrix components in a small volume (droplet) may be inhomogeneous. In spite of all these differences between the methods, the adhesion at the microscopic scale is correlated to macroscopic interfacial properties for epoxy and polyester composites reinforced with flax fibers.

4 Conclusions Composites were elaborated by using partly biobased thermoset resins reinforced with Melina flax fibers. Firstly, the matrices tensile properties were investigated and showed satisfactory mechanical performances. At the macroscopic scale, a good correlation was highlighted between the longitudinal Young’s modulus of the UD composites and those of the flax fibers, as expected from the rule of mixture. Due to the moderate fiber individualization level, UD tensile strengths were more scattered and lower than the expected values from the rule of mixture. The linkage between flax fibers and partly biobased resins was investigated at both microscopic and macroscopic scales. At the microscopic scale, the debonding test highlighted a satisfactory adhesion for every composite. In addition, in plane shear tests were carried out to analyze the influence of the interfacial bonding at the macroscopic scale. A strong sensitivity of the IFSS to the in-plane shear strength was highlighted. Finally, this study shows encouraging mechanical results for the use of partly biobased epoxy and polyester resins in plant reinforced composites. A forthcoming work will be dedicated to Life Cycle Assessments to justify the positive environmental impact of biobased epoxy and polyester resins compared with their petrochemical equivalent. References

17

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]

Misnon MI, Islam MM, Epaarachchi JA, Lau K. Potentiality of utilising natural textile materials for engineering composites applications. Mater Des 2014;59:359–68. Baley C. Analysis of the flax fibres tensile behaviour and analysis of the tensile stiffness increase. Compos Part Appl Sci Manuf 2002;33:939–48. Bourmaud A, Morvan C, Bouali A, Placet V, Perré P, Baley C. Relationships between microfibrillar angle, mechanical properties and biochemical composition of flax fibers. Ind Crops Prod 2013;44:343–51. Charlet K, Baley C, Morvan C, Jernot JP, Gomina M, Bréard J. Characteristics of Hermès flax fibres as a function of their location in the stem and properties of the derived unidirectional composites. Compos Part Appl Sci Manuf 2007;38:1912–21. ADEME. Evaluation de la disponibilité et de l’accessibilité de fibres végétales à usages matériaux en France. 2011. Gorshkova TA, Salnikov VV, Pogodina NM, Chemikosova SB, Yablokova EV, Ulanov AV, et al. Composition and Distribution of Cell Wall Phenolic Compounds in Flax (Linum usitatissimum L.) Stem Tissues. Ann Bot 2000;85:477–86. Zykwinska A, Thibault J-F, Ralet M-C. Competitive binding of pectin and xyloglucan with primary cell wall cellulose. Carbohydr Polym 2008;74:957–61. Di Bella G, Fiore V, Valenza A. Effect of areal weight and chemical treatment on the mechanical properties of bidirectional flax fabrics reinforced composites. Mater Des 2010;31:4098–103. Muralidhar BA. Study of flax hybrid preforms reinforced epoxy composites. Mater Des 2013;52:835–40. Baley C, Busnel F, Grohens Y, Sire O. Influence of chemical treatments on surface properties and adhesion of flax fibre–polyester resin. Compos Part Appl Sci Manuf 2006;37:1626–37. Khot SN, Lascala JJ, Can E, Morye SS, Williams GI, Palmese GR, et al. Development and application of triglyceride-based polymers and composites. J Appl Polym Sci 2001;82:703–23. Wool RP. Chemtech, Vol 29, 1999. Chemtech Vol 29 1999:44. Campanella A, Scala J, Wool RP. Fatty acid-based comonomers as styrene replacements in soybean and castor oil-based thermosetting polymers. J Appl Polym Sci 2011;119:1000–10. A. Shabeer SS. Physicochemical properties and fracture behavior of soy-based resin. J Appl Polym Sci 2007;105:656 – 663. Park S-J, Jin F-L, Lee J-R. Effect of Biodegradable Epoxidized Castor Oil on Physicochemical and Mechanical Properties of Epoxy Resins. Macromol Chem Phys 2004;205:2048–54. Earls J, White E, Lopez L, Lysenko Z, Dettloff M, Null M. Amine-cured omega-epoxy fatty acid triglycerides: Fundamental structure-property relationships. Polymer 2007;48:712–9. Parlevliet P, Bersee HEN, Beukers A. Residual stresses in thermoplastic composites—A study of the literature—Part I: Formation of residual stresses. Compos Part Appl Sci Manuf 2006;37:1847–57. Thomason JL, Yang L. Temperature dependence of the interfacial shear strength in glass reinforced polypropylene and epoxy composites, Montreal: 2013, p. 3518–26. Miller B, Muri P, Rebenfeld L. A microbond method for determination of the shear strength of a fiber/resin interface. Compos Sci Technol 1987;28:17–32. Coroller G, Lefeuvre A, Duigou A, Bourmaud A, Ausias G, Gaudry T, et al. Effect of flax fibres individualisation on tensile failure of flax/epoxy unidirectional composite. Compos Part Appl Sci Manuf 2013;51:62–70. Le Duigou A, Davies P, Baley C. Interfacial bonding of Flax fibre/Poly(l-lactide) bio-composites. Compos Sci Technol 2010;70:231–9. Van de Weyenberg I. Flax fibers as a reinforcement for epoxy composites. Katholieke Universiteit Leuven, 2005. Herrera-Franco PJ, Drzal LT. Comparison of methods for the measurement of fibre/matrix adhesion in composites. Composites 1992;23:2–27. 18

[24] Meredith J, Coles SR, Powe R, Collings E, Cozien-Cazuc S, Weager B, et al. On the static and dynamic properties of flax and Cordenka epoxy composites. Compos Sci Technol 2013;80:31–8. [25] Yongli Z, Yan L, Hao M, Tao Y. Tensile and interfacial properties of unidirectional flax/glass fiber reinforced hybrid composites. Compos Sci Technol 2013;88:172–7. [26] Baley C. Contribution à l’étude de matériaux composites à matrice organique renforcés par des fibres de lin 2003. [27] Le Duigou A, Davies P, Baley C. Macroscopic analysis of interfacial properties of flax/PLLA biocomposites. Compos Sci Technol 2010;70:1612–20. [28] AFNOR. EN X PT 25-501-2:2010 Reinforcement fibres — Flax fibres for plastics composites — Part 2: Determination of tensile properties of elementary fibres 2010. [29] Whitney JM. Bending-Extensional Coupling in Laminated Plates Under Transverse Loading. J Compos Mater 1969;3:20–8. [30] AFNOR. EN ISO 527-5:2009 Plastiques - Détermination des propriétés en traction - Partie 5: Conditions d’essai pour les composites plastiques renforcés de fibres unidirectionnelles 2009. [31] Bourmaud A, Morvan C, Baley C. Importance of fiber preparation to optimize the surface and mechanical properties of unitary flax fiber. Ind Crops Prod 2010;32:662–7. [32] Charlet K, Jernot JP, Gomina M, Bréard J, Morvan C, Baley C. Influence of an Agatha flax fibre location in a stem on its mechanical, chemical and morphological properties. Compos Sci Technol 2009;69:1399–403. [33] Hu W, Ton-That M-T, Perrin-Sarazin F, Denault J. An improved method for single fiber tensile test of natural fibers. Polym Eng Sci 2010;50:819–25. [34] Andersons J, Spārniņš E, Joffe R, Wallström L. Strength distribution of elementary flax fibres. Compos Sci Technol 2005;65:693–702. [35] Zafeiropoulos NE, Baillie CA. A study of the effect of surface treatments on the tensile strength of flax fibres: Part II. Application of Weibull statistics. Compos Part Appl Sci Manuf 2007;38:629–38. [36] Charlet K. Contribution à l’étude de composites unidirectionnels renforcés par des fibres de lin: relation entre la microstructure de la fibre et ses propriétés mécaniques. Caen, 2008. [37] Baley C, Bourmaud A. Average tensile properties of French elementary flax fibers. Mater Lett 2014;122:159–61. [38] Katz JI. Atomistics of tensile failure in fused silica : Weakest link models revisited. SPIE Proc. Ser., Boston: Society of Photo-Optical Instrumentation Engineers; 2000, p. 2–10. [39] Duval A, Bourmaud A, Augier L, Baley C. Influence of the sampling area of the stem on the mechanical properties of hemp fibers. Mater Lett 2011;65:797–800. [40] Alix S, Goimard J, Morvan C, Baley C, Schols HA, Visser RGF, et al. Influence of pectin structure on the mechanical properties of flax fibres: a comparison between linseed-winter variety (Oliver) and a fibre-spring variety of flax (Hermes). Pectins Pectinases 2009:87–98. [41] Marrot L, Lefeuvre A, Pontoire B, Bourmaud A, Baley C. Analysis of the hemp fiber mechanical properties and their scattering (Fedora 17). Ind Crops Prod 2013;51:317–27. [42] Lefeuvre A, Bourmaud A, Lebrun L, Morvan C, Baley C. A study of the yearly reproducibility of flax fiber tensile properties. Ind Crops Prod 2013;50:400–7. [43] Lee H, Neville K. Handbook of epoxy resins. McGraw-Hill; 1967. [44] Barrere C, Dal Maso F. Résines époxy réticulées par des polyamines : structure et propriétés. Oil Gas Sci Technol 1997;52:317–35. [45] Baley C, Grohens Y, Busnel F, Davies P. Application of Interlaminar Tests to Marine Composites. Relation between Glass Fibre/Polymer Interfaces and Interlaminar Properties of Marine Composites. Appl Compos Mater 2004;11:77–98. [46] Zinck P, Wagner H., Salmon L, Gerard J. Are microcomposites realistic models of the fibre/matrix interface? I. Micromechanical modelling. Polymer 2001;42:5401–13. 19

[47] Rao V, Herrera-franco P, Ozzello AD, Drzal LT. A Direct Comparison of the Fragmentation Test and the Microbond Pull-out Test for Determining the Interfacial Shear Strength. J Adhes 1991;34:65–77. [48] Bos HL, Van Den Oever MJA, Peters OCJJ. Tensile and compressive properties of flax fibres for natural fibre reinforced composites. J Mater Sci 2002;37:1683–92. [49] Moothoo J, Allaoui S, Ouagne P, Soulat D. A study of the tensile behaviour of flax tows and their potential for composite processing. Mater Des 2014;55:764–72. [50] Baley C, Le Duigou A, Bourmaud A, Davies P. Influence of drying on the mechanical behaviour of flax fibres and their unidirectional composites. Compos Part Appl Sci Manuf 2012;43:1226– 33. [51] Assarar M, Scida D, El Mahi A, Poilâne C, Ayad R. Influence of water ageing on mechanical properties and damage events of two reinforced composite materials: Flax–fibres and glass– fibres. Mater Des 2011;32:788–95. [52] Oksman K. High Quality Flax Fibre Composites Manufactured by the Resin Transfer Moulding Process. J Reinf Plast Compos 2001;20:621–7. [53] Van de Weyenberg I, Ivens J, De Coster A, Kino B, Baetens E, Verpoest I. Influence of processing and chemical treatment of flax fibres on their composites. Compos Sci Technol 2003;63:1241– 6. [54] Hughes M, Carpenter J, Hill C. Deformation and fracture behaviour of flax fibre reinforced thermosetting polymer matrix composites. J Mater Sci 2007;42:2499–511. [55] Wisnom MR. Size effects in the testing of fibre-composite materials. Compos Sci Technol 1999;59:1937–57. [56] Park S-J, Kim T-J. Studies on surface energetics of glass fabrics in an unsaturated polyester matrix system : Effect of sizing treatment on glass fabrics. J Appl Polym Sci 2001;80:1439–45. [57] Mäder E, Grundke K, Jacobasch H-J, Wachinger G. Surface, interphase and composite property relations in fibre-reinforced polymers. Composites 1994;25:739–44. [58] Keusch S, Queck H, Gliesche K. Influence of glass fibre/epoxy resin interface on static mechanical properties of unidirectional composites and on fatigue performance of cross ply composites. Compos Part Appl Sci Manuf 1998;29:701–5. [59] Keusch S, Haessler R. Influence of surface treatment of glass fibres on the dynamic mechanical properties of epoxy resin composites. Compos Part Appl Sci Manuf 1999;30:997–1002. [60] Drzal LT, Madhukar M. Fibre-matrix adhesion and its relationship to composite mechanical properties. J Mater Sci 1993;28:569–610.

20

FIGURES Figure 1. Sewn unidirectional flax tapes used to manufacture composite samples for the in-plane shear test

21

Figure 2. Stress-strain curves of the different matrices

80

Tensile Strength (MPa)

60

40

PetroEP BioEPA BioEPB PetroUP BioUP BioEPC

20

0 0

1

2

3

4

5

Tensile strain (%)

22

Figure 3. Tensile mechanical properties of the matrices

6 Young's modulus Strain Strength

80

60

4

3 40

Strength (MPa)

Young's modulus (GPa) / Strain (%)

5

2 20 1

0

0 PetroEP

BioEPA

BioEPB

BioEPC

PetroUP

BioUP

Matrices

23

Figure 4. Epoxy groups in the middle of the fatty chains within epoxidized linseed

24

Figure 5. Epoxy groups easily accessible at the extremities of DGEBA molecule

25

Figure 6. Typical tensile stress–strain curve for UD composites (shown here for UD with PetroEP resin and Vf=20%)

250

Tensile strength (MPa)

200

150

100

50

0 0,0

0,2

0,4

0,6

0,8

1,0

1,2

Tensile strain (%)

26

Figure 7. Level of individualization of the flax fibers in the UD composites (shown here with BioEPA matrix and Vf = 20%) observed by SEM

27

Figure 8. Young’s modulus (a) and strength at break (b) of the UD composites according to the fiber volume content. Dotted lines are the plotted equations of the rules of mixture.

a)

E (GPa)

30

20

PetroEP BioEPA BioEPB BioEPC PetroUP BioUP

10

0 0

10

20

30

40

50

60

Vf (%)

350 k1 = 0,7

b)

300

250

σ (MPa)

k1 = 0,4

200

PetroEP BioEPA BioEPB BioEPC PetroUP BioUP

150

100

50 0

20

40

60

Vf (%) 28

Figure 9. Shear strength of the ±45 laminates versus microscopic IFSS given by the microbond test

60

Composite Shear strength (MPa)

55

50

PetroEP

45 BioEPC BioEPA 40 PetroUP 35 BioEPB 30

25 0

5

10

15

20

25

30

35

40

IFSS (MPa)

29

Figure 10. Fracture surface of flax/BioEPB ±45 composites after in-plane shear test

Fiber impressions

Fiber bundle

30

Figure 11. Fracture surface of flax/BioEPC ±45 composites after in-plane shear test

Fiber damage

31

TABLES Table 1. Characteristics of the epoxy and polyester matrices Epoxy Commercial Epolam 2020

®

Greenpoxy 55

®

SuperSap

®

Epobiox STR 100

®

Name Simplified Name

PetroEP

BioEPA

BioEPB

BioEPC

Manufacturer

Axson

Sicomin

Entropyresin

Sandtech

Hardener

Polyamine

Polyamine

Polyamide

Aliphatic amine

-

55 ± 2

51

50-55

-

Biobased epichlorhydrine

Pine oil

Biobased Bisphenol A, pine oil

14

C biocontent

(ASTM D6866) Bio origin

1h at room T Curing specifications

24h at room T 2h at 120°C

1h at 82°C

+2h80°C

+24h at 40°C +5h140°C

ΔHuncured (J/g)

423

364

293

224

ΔHcured (J/g)

17

7

16

13

% Cure

96

98

95

94

Polyester Commercial Norsodyne® G703

Enviroguard® 93250

Simplified Name

PetroUP

BioUP

Manufacturer

CCP composites

Cray Valley

%weight biocontent

23

53

Curing specifications

24h at room T

24h at room T

+2h at 80°C

+2h at 80°C

ΔHuncured (J/g)

395

253

ΔHcured (J/g)

14

1

% Cure

96

99

Name

32

Table 2. IFSS results from this study compared with those of the literature Fiber/matrix system

IFSS (MPa)

EPOXY

Ref Mean value

Linear regression

R

Flax/PetroEP

20.4 ± 4.9

20.1

0.79

This study

Flax/BioEPA

17.0 ± 5.0

16.4

0.82

This study

Flax/BioEPB

11.9 ± 4.1

11.8

0.65

This study

Flax/BioEPC

28.5 ± 6.4

25.4

0.69

This study

Flax/EPOLAM 2020®

22.3 ± 2.1

[20]

Flax/epoxy

22.9 ± 7

[22]

Flax/ EPOLAM 2015®

16.1 ± 0.8

[21]

Glass/EPOLAM 2020®

37.2 ± 4.6

[20]

Glass/ EPOLAM 2015®

29.3 ± 2.4

[45]

Flax/PetroUP POLYESTER

2

17.5 ± 7.3

This study

0.87

This study

Flax/polyester

14.2 ± 0.4

[10][21]

Glass/polyester

15.7 ± 2.9

[45]

33

POLYESTER

EPOXY

Table 3. Estimated and real UD tensile mechanical performances from this study and comparison with the literature UD composites

Vf (%)

E (GPa)

E Mixture rule (GPa)

σ (MPa)

σMixture rule (GPa)

Efficiency factor k1

Strain (%)

Ref

Flax (Melina)/epoxy (PetroEP)

32 ± 2

18.4 ± 1.4

19.7

224 ± 18

308

0.69

1.2 ± 0.1

This study

Flax (Melina)/epoxy (PetroEP)

47 ± 2

28.3 ± 2.6

27.7

253 ± 9

433

0.55

0.9 ± 0.2

This study

Flax (Melina)/epoxy (BioEPA)

42 ± 2

23.1 ± 2.7

24.5

226 ± 32

383

0.56

0.9 ± 0.3

This study

Flax (Melina)/epoxy (BioEPB)

20 ± 3

11.6 ± 1.5

12.4

130 ± 15

195

0.63

1.1 ± 0.1

This study

Flax (Melina)/epoxy (BioEPC)

40 ± 3

20.0 ± 3.3

23.7

175 ± 25

371

0.43

0.9 ± 0.2

This study

Flax/epoxy

32

15 ± 0.6

-

132 ± 4.5

-

-

1.2

[52]

Flax/epoxy

40.4 ± 1.2

22.50 ± 1.51

-

328 ± 18

-

-

1.6 ± 0.2

[50]

Flax (Hermès)/epoxy

51 ± 2

26 ± 2.0

26.6

408 ± 36

582

0.69

1.3 ± 0.05

[20]

Flax (Ariane)/epoxy

40

28

-

133

-

-

-

[53]

Glass/epoxy

32 ± 0

25 ± 0.9

24.7

446 ± 12

818

0.69

2.3 ± 0.12

[20]

Glass/epoxy

48

31 ± 1

-

817 ± 35

-

-

2.8

[52]

Flax (Melina)/polyester (PetroUP)

28 ± 3

17.2 ± 1.6

18.1

185 ± 19

253

0.68

1.2 ± 0.1

This study

Flax (Melina)/polyester (PetroUP)

50 ± 2

27.3 ± 2.5

29.3

264 ± 2

408

0.62

1.0 ± 0.1

This study

Flax (Melina)/polyester (BioUP)

31 ± 1

18.8 ± 4.0

19.0

174 ± 76

202

0.84

1.0 ± 0.1

This study

Flax/polyester

57.6 ± 2.1

29.9 ± 1.8

-

304 ± 29

-

-

1.73 ± 0.10

[54]

E-Glass/polyester

42.4 ± 3.5

30.6 ± 2.2

-

695 ± 60

-

-

2.37 ± 0.36

[54]

34

PLA

POLYESTER

EPOXY

Table 4. In-plane shear performances from this study and comparison with the literature Fiber/matrix system

Vf (%)

τ12 (MPa)

G12 (MPa)

Ref.

Flax/PetroEP

27

48 ± 8

2405 ± 117

This study

Flax/BioEPA

27

40 ± 9

2507 ± 215

This study

Flax/BioEPB

27

35 ± 2

1474 ± 107

This study

Flax/BioEPC

27

45 ± 4

2026 ± 258

This study

Flax/Epoxy

32.5

46.2 ± 1.8

1679 ±121

[26]

Glass/Epoxy

-

52 ± 2

-

[26]

Flax/PetroUP

26

39 ± 4

2032 ± 117

This study

Flax/BioUP

24

27 ± 3

1839 ± 197

This study

Flax/polyester

33.5

20.5 ± 0.8

-

[26]

Flax/PLLA

30

22.6 ± 3.1

1989 ± 159

[27]

35

Multi-scale study of the adhesion between flax fibers and biobased thermoset matrices

Highlights •

A double analysis of the thermoset biocomposites fiber-matrix adhesion was performed

•

Debonding tests allowed to estimate the interface properties at the microscopic scale

•

Tensile and in plane share experiments were conducted on macroscopic specimens

•

A strong sensitivity of the IFSS to the in-plane shear strength was highlighted

36