- Email: [email protected]
epoxy composites
Industrial Crops & Products 130 (2019) 25–33
Contents lists available at ScienceDirect
Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop
Compressive strength of flax fibre bundles within the stem and comparison with unidirectional flax/epoxy composites
T
Christophe Baleya, Camille Goudenhoofta, Patrick Perréb, Pin Lub, Floran Pierreb, ⁎ Alain Bourmauda, a b
Univ. Bretagne Sud, UMR CNRS 6027, IRDL, F-56100 Lorient, France LGPM, CentraleSupélec, Université Paris-Saclay, Centre Européen de Biotechnologie et de Bioéconomie (CEBB), 3 rue des Rouges Terres, 51 110 Pomacle, France
A R T I C LE I N FO
A B S T R A C T
Keywords: Buckling Composite materials Compressive strength Flax fibres Stem Unidirectional composite
Flax (Linum usitatissimum L.) fibres are commonly used as reinforcement of composite materials. Nevertheless, literature shows that the compressive strength of flax-based composites is rather modest compared with materials reinforced by synthetic fibres. The present article investigates the compressive strength of flax fibre bundles both within the stems and in unidirectional (UD) composites. In this way, an optimised arrangement of fibre bundles inside the plant is assumed. Damage mechanisms are found to be similar in the stem and within flaxbased UD materials, namely by buckling of fibre bundles, a typical failure mechanism of UD composites. Inside the stems, this phenomenon is highlighted by nanotomography, which underlines the key role of the woody core in the buckling resistance of the plant. For UD, failure can also be studied by scanning electron microscopy (SEM). The same ranges of average compressive strength values are estimated for flax fibre bundles, being 206 MPa within the stem and 242 MPa within UD composites. Finally, this study highlights that, if a flax stem is an optimised natural structure, the compressive strength of flax fibre bundles seems to be a limiting factor for structural applications of flax-based composite materials.
1. Introduction Flax (Linum usitatissimum L.) fibres have been used for centuries for textile applications. Their use as reinforcement of composite materials started more recently, in the 1940s (de Bruyne, 1939). This latest application is the subject of many research works, the main purposes of flax-based composite development being for instance the reduction of weight of manufactured parts or the decrease of their environmental impacts. The term of flax fibres used in the present work refers to elementary fibre cells located at the periphery of the stem, forming a fibrous ring composed of fibre bundles (Baley et al., 2018a; Goudenhooft et al., 2017). Flax fibres have a key role of supporting tissues due to noteworthy mechanical properties. Thus, they ensure the plant stability and allow it to present a remarkably slender structure, assimilated to a composite material composed of the epidermis as an external protection, a unidirectional ply of flax fibres at the periphery and a xylem part as the innermost porous core (Baley et al., 2018a). The flexural mechanical characterisation of a flax stem demonstrates that these plant fibres greatly contribute to the flexural stiffness of the stem (Gibaud et al., 2015). For example, taking into account a ⁎
dried stem, flax fibres were proven to ensure about 70% of the stem stiffness (Réquilé et al., 2018). In the case of dried hemp stems, the fibrous part only contributes to about 50% of the plant stiffness. The outstanding properties of flax fibres, and more specifically the length of elementary cells and associated mechanical properties (Baley and Bourmaud, 2014), are the results of an intrusive fibre development in a first place (Snegireva et al., 2010), which is followed by the thickening of the fibre cells through the deposition of cellulose (Rihouey et al., 2017). This latter process provides fibres exhibiting a very thick secondary cell wall and small lumen (Charlet et al., 2009). Within the stem, pectic junctions bond the fibres together and ensure the bundle arrangement. This structure gives a first model of a composite material at the bundle scale. In this latter, the stresses are transferred to the reinforcing fibres through a polymer matrix, made of pectic junctions, mainly composed of homogalacturonans and rhamnogalacturonans (Andème-Onzighi et al., 2000). This matrix can be found in two different areas: the middle lamella (i.e. between the primary cell walls of two neighbouring fibres) and the tri-cellular junctions (i.e. in the corner between three fibres) (Baley et al., 2014). Unfortunately, the compressive strength of plant stems is poorly studied in literature but it has been evidenced by Mattheck (1995) that hollow
Corresponding author. E-mail address: [email protected] (A. Bourmaud).
https://doi.org/10.1016/j.indcrop.2018.12.059 Received 19 June 2018; Received in revised form 14 December 2018; Accepted 18 December 2018 0926-6690/ © 2018 Elsevier B.V. All rights reserved.
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Table 1 Properties of unidirectional composites reinforced by flax fibres provided in literature, highlighting longitudinal compressive and tensile properties. Matrix
Fibre volume fraction (%)
Compressive modulus (GPa)
Compressive strength (MPa)
Tensile modulus (GPa)
Tensile strength (MPa)
References
Phenol formaldehyde Phenol formaldehyde
˜70 ˜75
41.4 48.3
163 182
41.4 48.3
310 414
Epoxy Epoxy Epoxy
43.9 40 51
24.7 15.1 30.3
136 137 127
22.8 23.9 31.4
318 223 287
(de Bruyne, 1939) (Livingston Smith and Mech, 1945) (Liang et al., 2015) (Van Vuure et al., 2015) (Mahboob et al., 2017)
compressive strength, which remains lower than 140 MPa
stems are detrimental to stability due to Brazier buckling (Brazier, 1927). Oppositely, when stems exhibit thick-walled structures, the failure is preferentially attributed to mechanical limit of tissues components (Leblicq et al., 2015). In case of flax fibres, despite remarkable longitudinal mechanical performances, these latter evidence poor transverse properties and especially stiffness (Bourmaud and Baley, 2009), which makes them very sensitive to transverse or compressive stress. This sensibility may be a significant drawback if processing steps such as pulling-out or scutching are too aggressive. Thus, despite this remarkable structure and tissue organisation of plant fibre reinforcement, the resistance to compressive strength of flaxbased composites is rather limited, whatever the type of matrix or the fibre volume fraction, when compared to their resistance to tensile strength (Roussière et al., 2012). Table 1 shows the longitudinal compressive mechanical properties of different unidirectional (UD) of flaxreinforced composites (Bos et al., 2004; de Bruyne, 1939; Liang et al., 2015; Livingston Smith and Mech, 1945; Mahboob et al., 2017; Van Vuure et al., 2015). For all the data sets given in Table 1, the compressive strength is lower than 200 MPa. In addition, it is always poorer than the tensile strength (which is about two times higher). Therefore, this mechanical parameter is of significance for the improvement of flax biocomposites. For comparison, aramid fibre may have an apparent compressive strength down to only 10–15% of the tensile strength (Kozez et al., 1995; Leal et al., 2007). Numerous parameters can influence the compressive behaviour of composite materials, such as fibre volume fraction, the fibre and matrix properties, the interface bond strength, the porosity, the fibre misalignment, as well as the difference in Poisson ratio between fibre and matrix (de Bruyne, 1939; Liang et al., 2015; Livingston Smith and Mech, 1945; Mahboob et al., 2017; Van Vuure et al., 2015). In the case of a unidirectional ply under longitudinal compression, the failure takes place due to fibre kinking; this result was demonstrated for natural fibres (Poulsen et al., 1997) as well as synthetic ones (Garland et al., 2001). A previous article (Baley et al., 2018b) presents a detailed investigation of the longitudinal compressive behaviour of composites reinforced by unidirectional plant fibre tapes. We study and discuss the influence of the fibre type (flax or jute), polymer (PP, PP/MAPP, PA11, epoxy or acrylic), volume fraction and the quality of interfacial shear strength (IFSS) with a maleic anhydride grafted PP (MAPP). The results obtained for composites reinforced by unidirectional flax fibres show that:
• The fibre nature (flax or jute) also play an important role. Compared
to flax, the use of jute fibres leads to lower mechanical properties (in tensile and in compression).
The present work investigates the compressive behaviour of flax fibre bundles within the stem. In this way, fibres are considered under an optimal situation within the stem, i.e. cohesive and oriented in the longitudinal direction of the stem, leading to a model of natural composite structure. First, loading is performed through flexural bending of the stems. Then, the induced damaging and failure mechanisms of fibre bundles are observed by nanotomography. Complementary experiments are performed on several flax varieties to evaluate the influence of the varietal selection work on the bundle compressive strength. In addition, the compressive strength at break of fibre bundles within the stems is compared with the ones of unidirectional plies. Finally, the estimation of the role in compression of the xylem, also called woody core, is given. 2. Material and methods 2.1. Plant material – flax stems and fibres All flax plants were provided by Terre De Lin (a flax cooperative based in Normandy, France). Two recent as well as two ancient textile flax varieties were studied. The recent varieties are Eden (registered in 2009) and Bolchoï (registered in 2014); they were selected by Terre De Lin and are both currently commercialised. Liral Prince, from a selection made in 1944 by the Linen Industry Research Association (L.I.R.A, Ireland)(Goudenhooft et al., 2017) and Ariane, registered in 1978 by the Coopérative Linière de Fontaine Cany (France) are the two ancient varieties studied. They are not commercialised anymore. Plants were cultivated in Saint-Pierre le Viger (Normandy, France) in 2016. Conventional seeding rates and culture conditions were used (Bert, 2013). Entire plants were pulled out at maturity (Goudenhooft et al., 2017) and subsequently dried under a standard atmosphere (temperature of 23 ± 2 °C and relative humidity of 50 ± 4%) until stabilisation of the stem weight. Some of the dried stems were waterretted (in order to facilitate the extraction of fibres for tensile tests) then dried as well. Prior to bending and tensile tests, stems and fibres were conditioned under the previously mentioned standard atmosphere for at least 24 h. Water contents of both stems and fibres were measured to be 7.8 ± 0.4% after conditioning.
• for small strains, the compressive modulus is closely similar to the tensile modulus. • the compressive strength is always lower than the tensile strength. • In compression, the failure mechanism involves the microbuckling • •
2.2. Three-point bending test on stem Three-point bending tests were performed on intact flax stem samples. Additional stems were carefully peeled, i.e. fibres were peeled off, and tested similarly in order to examine only the woody part of the samples. All stem samples were extracted from the middle part of the plants i.e. from the location of maximum fibre content (Lefeuvre et al., 2015). Three-point bending tests were carried out following the protocol
of flax fibres (as seen in composites reinforced by carbon, glass or organic fibres). The enhancement of IFSS using PP/MAPP or by selecting an appropriate matrix leads to better compressive properties. Nevertheless, a large change of polymer matrix or an increase in the fibre volume fraction does not imply a drastic increase of 26
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
2.4. Nanotomography
proposed by Réquilé et al. (2018), under the standard atmosphere (previously mentioned). Tests were performed on a universal MTS tensile testing machine equipped with a 50 N capacity load cell and the displacement rate was fixed at 0.1 mm/s. Experiments were conducted on 20 mm flax stems (and stem after peeling), using a span length of 18 cm in order to provide a length-todiameter ratio high enough to avoid shear effects (Réquilé et al., 2018). First, by considering both the entire stem and the woody core as a uniform plain beam with a circular cross section, the displacement y in the middle of the sample section read as (Timoshenko, 1947):
y=
FL3 48EIsample
The samples were scanned with a nanotomograph EasyTom XL 150/ 160 manufactured by RX-solutions (Annecy, France). The nano-focus Xray sources (an open nano-focus tube with a tungsten filament) and a CCD detector of 2004 × 1336 pixels were selected to obtain a voxel size of 1.7 μm, the best possible compromise to scan the entire stem diameter. The region of interest of the flax stem (Bolchoï variety) after bending was glued on the top of the 2 mm-carbon rod of the sample holder. This protocol allows the flax stem to be close enough to the Xray tube to obtain this resolution. These conditions impose an acquisition time of about 3 h.
(1)
2.5. Scanning electron microscopy (SEM)
Where F is the force, L the span length, (EIsample) the bending stiffness either of the stem or the woody core, Isample the quadratic momentum of the sample and E its apparent bending modulus. The stem or woody core bending stiffness for small strains is then obtained by applying the following formula:
EIsample =
dF L3 dy 48.
Composite samples were embedded in an epoxy resin before polishing. They were then metallised with gold before being observed under a JEOL JSM 6460LV scanning electron microscope. 2.6. Stem anatomy and bundle thickness
(2) Stem portions were taken at the middle height of the flax plants. To facilitate the cut, the portions were first immersed in water at least 24 h. Then, they were embedded in elder marrow and semi-thin sections were cut transversally using a razor blade and a hand microtome to clamp the sample. Optical images were taken under the microscope and used to estimate the bundle thickness.
Where dF/dy is the slope of the linear part (small displacements) of the force-displacement curve. The maximal force Fmax damaging the sample is determined for each test; the corresponding maximum bending moment Mmax at the centre of the sample is estimated by Eq. (3):
Mmax =
Fmax L 4
(3)
2.7. Unidirectional composites
At mid-length of the beam, the maximal strain εmax is expressed by Eq. (4):
εmax =
Mmax D EI 2
The matrix used is an epoxy resin (Axson, Epolam 2020). Its tensile Young’s modulus, strength and strain at break are 3.10 ± 0.01 GPa, 78 ± 6 MPa, and 3.1 ± 0.5%, respectively. The reinforcement consists of a tape of unidirectional fibres without any twist (Flax-Tape, Lineo®) (Khalfallah et al., 2014) with an area weight of 200 g/m². The processing of flax (Eden variety) tapes begins by decorticating and separating the bundles from the flax stems. The tapes are then cleaned by scutching to remove the remaining shives and woody core of the stems. The flax bundles are aligned to the input of the manufacturing machine and moistened by a water mist and continuously dried through an infrared oven. To maintain the cohesion of the parallel fibers, the process is based on the reactivation of the pectin cement during spraying a water mist. As in a stem, the unidirectional ply is reinforced by bundles (the bundles are not cut off). The unidirectional laminates were obtained by impregnating and then pressing to eliminate voids (Labtech© 50 T moulding press). Fibre volume fractions from 20% to 52% were measured by image analysis from SEM observations on transverse sections after polishing. After hardening for 24 h at 25 °C, samples were post-cured (3 h –40 °C; 2 h –60 °C; 2 h –80 °C; 5 h −100 °C). For compressive characterisation and after preliminary tests, the thickness of samples is set at 2.5 mm. For the test, two gauges were bonded, one on each face, in order to monitor the compressive behaviour.
(4)
In the case of the peeled stem, the sample is considered as a homogeneous material having a linear mechanical behaviour (validated Hooke’s law). The strength is thus calculated as the multiplication of the maximal strain by the apparent elastic modulus determined through bending tests. However, for the entire stem, a sandwich composite beam is considered, having an outer unidirectional ply on the outermost side (assembly of fibres in the form of bundles forming an external fibrous ring). In this case, the product EIsample of previous equations should be expressed as Eq. (5):
EIs =
π (Ef (Ds4 − Dc4 ) + Ec Dc4 ) 64
(5)
Where D is the diameter and E the elastic modulus with indices f, s and c standing for fibres, stem and woody core respectively. The elastic modulus of the fibres Ef is measured by tensile tests. In the present case, the fibre behaviour is simplified as being elastic linear. Thus, the bundles compressive strength at break σfmax expressed by Eq. (6):
σf max = εmax Ef
(6)
3. Results 2.3. Tensile tests on elementary flax fibres 3.1. Mechanical properties of elementary flax fibres Tensile tests on elementary fibres were performed according to NFT 25-501-2 standard. Fibres were manually extracted from retted stems and bonded onto a paper frame with a gauge length of 10 mm. The fibre samples were conditioned for at least 24 h under standard atmosphere. The diameter of each fibre was determined as the average of 6 measurements taken along the fibre using optical microscope. The samples were finally clamped on a universal MTS tensile testing machine equipped with a 2 N capacity load sensor using a crosshead speed of 1 mm/min. For each batch of fibres, a minimum of 70 fibres was tested.
In order to analyse the compressive behaviour of flax stems, it is necessary to know the mechanical properties of the elementary fibres composing their bundles. More particularly, the Young’s modulus is required for the studied batches. Table 2 presents the average tensile properties of elementary fibres extracted from the related stem batches according to the stem variety. These performances can be compared with the work of Baley and Bourmaud (2014) which have analysed numerous batches of 27
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Table 2 Tensile mechanical properties of elementary fibres of the four flax varieties. Variety
Fibre diameter (μm)
Young’s modulus (GPa)
Strength at break (MPa)
Strain at break (%)
Bolchoï Eden Ariane Liral Prince
22.7 16.6 17.4 21.8
52.9 54.9 43.6 40.9
1012 ± 283 1056 ± 310 960 ± 288 860 ± 268
2.12 2.22 2.39 2.39
± ± ± ±
3.9 2.8 2.6 2.4
± ± ± ±
12.2 10.5 9.6 10.2
± ± ± ±
0.59 0.62 0.49 0.62
Fig. 2. Nanotomography of the damaged area of a flax stem after a bending test. A local buckling is visible, due to excessive compressive strength.
elementary fibres, tested under a similar standard atmosphere and the same protocol than the present study. Average mechanical properties available in the mentioned literature are a Young’s modulus reaching 52.5 ± 8.6 GPa, a strength at break of 945 ± 200 MPa and a strain at break of 2.07 ± 0.45% for an average fibre diameter of 16.8 ± 2.7 μm. One can notice that, despite slightly lower tensile Young’s modulus and strength at break for the oldest varieties Liral Prince and Ariane, the diameters and mechanical properties of elementary fibres measured in the present work belong to the same ranges as literature values (Baley and Bourmaud, 2014). These similarities confirm the good reproducibility of flax fibres mechanical properties.
span length. The stem damage actually happens on the face in compression, and is often located close to the central support of the bending device. This observation is consistent with a three-point bending test, for which the bending moment is maximum at mid-length, which confirms the hypothesis of a behaviour very similar to pure bending. 3.3. Failure mechanisms of flax stems during bending tests Observations of a damaged stem through nanotomography highlight the mechanism of local buckling in the part of the stem in compression (Fig. 2). In addition, the nanotomographic investigations enable to visualize a longitudinal section of the stem within the damaged zone (Fig. 3.(A)). Thanks to these observations, successive damages can be identified within a stem with, first, the local buckling of a fibre bundle together with the formation of a kink-band (Fig. 3.(A)). The latter phenomenon is also visible on elementary fibres while bending (Baley, 2002), but also for wood under compressive strength (Benabou, 2008; Zaune et al., 2016) as well as for synthetic fibres such as UHMWPE (Attwood et al., 2015) or aramid (Deteresa et al., 1988). Then, the detachment of the bundle from the woody core is visible in a small part of the stem. Finally, the local buckling of other cell types, namely conduction cells within the woody core, is also visible in Fig. 3. This latter figure also shows that the cohesion between the fibre bundles and the woody core is a key parameter to the resistance of fibre bundles to local buckling. Other elements are involved in the compression behaviour of the fibre bundles such as the middle lamellae ensuring the bonding between the fibres within a bundle and between the bundles of the other cells, if the latter are not damaged by the retting step; moreover, the cortex ensuring the load transfer between the bundles is also a major contributor of bundles compression resistance (Baley et al., 2018a). However, one can notice that nanotomographic images do not allow to distinguish elementary fibres given the absence of density difference between fibres within the bundle. To facilitate the examination of the previous figure, Fig. 3.(B) shows the organisation of the flax stem in transverse cross-section (Bolchoï variety). The composite structure of the stem is highlighted, with two main constitutive parts: fibre bundles located at the periphery of the stem and the woody core of the plant (mostly xylem) located between the fibres and the central cavity (Catling and Grayson, 1982; Evert, 2006; Evert and Eichhorn, 2013; Gorshkova et al., 2010; His et al., 2001). For more information on the schematic organization of a stem, the reader may consult the following references: for a stem section (Catling and Grayson, 1982; Réquilé et al., 2018) and for a longitudinal section (Catling and Grayson, 1982). For a thin-walled tube solicited in bending, the risk of local buckling is a function, among other parameters, of the ratio of the external diameter of the tube and its thickness (Delaplace et al., 2008). Namely, for a given diameter, the smaller the wall thickness is, the higher the instability exists. Similarly, a flax stem is comparable to a thin-walled tube containing a core, i.e. a circular sandwich structure with an inner
3.2. Mechanical behaviour of flax stems and woody cores during bending tests The bending stiffness of flax stems is evaluated by three-point bending tests. Fig. 1 presents the force-displacement profiles for both the flax stem (with a diameter of 2.10 mm) and the woody part (with a diameter of 1.89 mm). Similar elastic and linear behaviours are exhibited at the beginning of the test for both the stem and the woody core, which allow us to define the flexural modulus of the stems based on Eq. (2). Considerably lower values are highlighted for the woody core of the stem, for both modulus and maximal strength compared to an entire stem. These results confirm the previously published ones (Réquilé et al., 2018), and underline the key role of fibre part in stem stability. Flax stems can be assimilated to cylindrical tubes of small diameter, taking into consideration the thickness of the fibre ring. Therefore, the bending behaviour of these stems can be described as analogous to those of tubes with thin walls (Leblicq et al., 2015). Considering the large L/D of the samples taken for the bending tests, the loss of linearity of the curve is explained by a large displacement of the central support just before breakage (Fig. 1). In addition, the ovalisation and squashing effects are negligible for all the performed tests thanks to the use of a pertinent
Fig. 1. Typical stress-strain curve for a three-point bending test on flax stem (whole) and woody core (after stem peeling) for Bolchoï plants (span length of 180 mm). 28
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Fig. 3. (A) Nanotomography image of the longitudinal section of a flax stem after the bending test, in the damaged area. The local buckling of the bundle is visible, as well as other cell types. The apparition of a kink-band is also highlighted. (B) Transverse cross-section of a stem of Bolchoï exhibiting thick fibre bundles of 100 μm on average.
a comparison between stems whose diameters range between 1.69 and 2.84 mm. The authors decided to use samples having dispersed diameters in order to verify the method. When estimating the apparent flexural modulus of the stem (Fig. 4.(B)), the results show that it is also slightly influenced by the stem diameter even though the dispersion of the results remains modest (23.8 ± 2.8 GPa) for a natural material such as a flax stem. For each stem sample, the diameter (D), the span length (L), the bending stiffness (EI) and the maximal force damaging the sample (Fmax) are measured or calculated. In addition, the average modulus of elementary fibres was previously determined (Ef = 52.9 GPa, see Table 1). From this data and through related equations, Table 3 gives the resulting stem apparent bending modulus, the maximal compressive strain at break (εmax) and the bundles compressive strength at break (σfmax). In order to illustrate the scattering of the results, Fig. 5 presents the compressive strength at break of fibre bundles as a function of the stem diameter. An absence of correlation between the mentioned properties is noticeable, i.e. there is no link between the compressive failure strength and the stem diameter. Moreover, the values are a little scattered (206 ± 26 MPa) despite the difficulty to precisely achieve the stress value leading to local buckling (shown in Fig. 1) in the part of the stem under compression. The maximal damaging force was used for calculations, but it is possible that the local buckling appears due to an accumulation of damages. This scenario is well described in literature for unidirectional plies reinforced by synthetic fibres solicited in compression (Garland et al., 2001). In this case, the mechanism of kinkband formation is described as a succession of several steps; namely, the fibre breaks first, followed by the subsequent development of a damaged zone. Finally, the kink-band is formed from the shear instability, originated and visible in the damage zone.
porous material. This core has a great contribution to the resistance to local buckling.
3.4. Mechanical behaviour of Bolchoï stems during a three-point bending test For this point, Bolchoï, which is a recent, performing and one of the most cultivated varieties is considered. Fig. 4.(A) highlights a good correlation between the diameter of a flax stem and its bending stiffness measured by three-point bending tests. This piece of information brings
3.5. Mechanical behaviour of peeled stems of Bolchoï during a bending test Complementary, peeled stems of Bolchoï (i.e. samples are deprived of fibre bundles and considered as only composed of the woody core) are submitted to three-point tests under identical conditions than the Table 3 Stem properties obtained from three-point bending tests on Bolchoï stems. Mean values Stem diameter (mm) Bending stiffness (N.m²) Apparent bending modulus (GPa) Compressive strain at break (%) Compressive strength at break (MPa)
Fig. 4. (A) Correlation between the flax stem diameter and the bending stiffness. (B) Correlation between the flax stem diameter and the apparent bending modulus. 29
2.11 ± 0.34 0.025 ± 0.015 23.8 ± 2.8 0.39 ± 0.05 206 ± 26
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Fig. 5. Compressive strength at break of the fibre bundles as a function of the stem diameter.
Fig. 7. Compressive strain at break of the woody core as a function of the sample diameter.
entire stems already tested. The main aim of this analysis is to achieve the strain capacity of the woody core who is often considered as a no structural part of the stem. This component of the stem is constituted of cells of short length, which are responsible for the conduction of the raw sap whereas phloem cells contribute to the elaborated sap conduction. These conductive cells have a great lumen size and thin cell walls, as shown by tomography (Fig. 3.(A)). Furthermore, the biochemical composition of woody cells differs from the fibre ones (Evon et al., 2018), mainly in terms of cellulose content but also of non-cellulosic polymer fraction and nature. On average, when expressed in percentage of cellulose in the dry matter, the woody core reaches 45.6% and fibres 78.7% (Evon et al., 2018). The mechanical behaviour of a peeled stem was shown in Fig. 1. Additionally, such as for entire stems, the bending stiffness of samples with scattered diameters is analysed. Interestingly, as similarly highlighted for complete stems, there is a good correlation between the bending stiffness of the woody core and the sample diameter (Fig. 6). By measuring the bending stiffness of the peeled stems, it is possible to calculate the apparent bending modulus of each sample if considering the woody core as a homogeneous material. In addition, the compressive strain at break obtained from the maximal force damaging the samples (similar hypothesis than for entire stems) is presented in Fig. 7. For the woody core, there is a slight correlation between the strain at break and the sample diameter.
Table 4 Woody core properties obtained from three-point bending tests on peeled samples of Bolchoï. Mean values Stem diameter (mm) Bending stiffness (N m²) Apparent bending modulus (GPa) Compressive strain at break (%) Compressive strength at break (MPa)
2.08 ± 0.28 0.009 ± 0.005 8.5 ± 4.7 0.69 ± 0.05 59 ± 6
Table 4 gives the mechanical properties resulting from three-point bending tests on peeled samples. In comparison with entire stem samples, one can notice the superior strain at break of the woody core (0.69% ± 0.05) compared with fibre bundles (0.39% ± 0.05, values presented in Table 3). Through its greater strain capacity, the contribution of the woody core in limiting the buckling risk of flax stems under compression is confirmed. However, the compressive strength at break of this porous structure is rather low (59 MPa on average) compared to fibre bundles (206 MPa in Table 3). Nevertheless,when compared to the compressive strength at break of wood, estimated using the wood density (Domone and Illston, 2010), a mean value of 45 MPa is expected for wood when using the flax woody core density of 0.455. This later average apparent density of flax woody core was measured (0.455 ± 0.113 g/cm3) for flax in a previous study (Réquilé et al., 2018). Finally, for an equivalent density, the compressive strength at break of the woody core belongs to the same range of values as for wood, and is even somewhat superior.
3.6. Bending mechanical behaviour of flax stems of different varieties The first flax variety, previously studied, is Bolchoï; it is a recent one having a high fibre content. This later parameter is a major criterion of selection for flax breeders. Thus, one can wonder what would be the influence of the fibre content, i.e. of the variety, on the compressive resistance of flax bundles. In a former study (Goudenhooft et al., 2017), the authors investigated the impact of the varietal selection on the stem architecture (namely the plant anatomy) thanks to an analysis of 4 different varieties cultivated during the same year, in the same location and under the same cultural conditions. These varieties, even though cultivated the same year for the study, had been registered in 1944 (Liral Prince), 1978 (Ariane), 2009 (Eden) or 2011 (Aramis). This article (Goudenhooft et al., 2017) highlighted the evolution of the fibre content over selection, namely increased fibre contents for more recent varieties. Fig. 3.(B) presents a transverse stem section of a more recent
Fig. 6. Bending stiffness of peeled stems (assimilated as the woody core) as a function of the sample diameter. 30
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Table 5 Influence of the flax variety and fibre bundle thickness on the compression resistance of the stems. Variety
Stem diameter (mm)
Fibre bundles thickness (μm)
Bending stiffness (N m²)
Apparent bending modulus (GPa)
Compressive strain at break (%)
Compressive strength at break (MPa)
Bolchoï Eden Ariane Liral Prince
2.11 2.02 2.11 2.10
102 ± 18 99 ± 16 70 ± 18 73 ± 13
0.025 0.018 0.020 0.019
23.8 20.8 18.2 18.4
0.39 0.38 0.43 0.45
206 207 189 183
± ± ± ±
0.34 0.23 0.34 0.29
± ± ± ±
0.015 0.006 0.012 0.009
variety, Bolchoï, having thick fibre bundles and a high fibre yield. Based on image analysis (pictures not shown) and three-point bending tests, Table 5 presents the results and links between the compression resistance of the stem and the thickness of their respective fibre bundles. Thus, a slight increase of the compressive strength at break is visible with increasing bundle thickness in the plant. In addition, as visible on the cross-section of Bolchoï (Fig. 3.(B)), the fibrous ring is continuous along the stem periphery, which decreases the risk of buckling. This result is true for recent varieties, whereas ancient ones exhibit gaps between neighbouring bundles (Goudenhooft et al., 2017). Moreover, the elastic modulus of elementary fibres is greater for recent varieties (Table 1), which stiffens the stems.
± ± ± ±
2.8 2.2 1.7 2.4
± ± ± ±
0.05 0.02 0.04 0.04
± ± ± ±
26 8 16 15
these results such as: i) The bundle division and distribution in a ply cross section (the composite microstructure); ii) The fibre orientation inside the composites. Indeed, flax fibres have a limited length and it is not possible to apply a tensile load during the process to obtain a perfect orientation; iii) The bonding between fibre and matrix with this batch of fibres. Nevertheless for this constituents, the good adhesion between flax and epoxy is well described in the literature (Le Duigou et al., 2014); iv) The possible evolution of the cell wall properties during this transformation cycle. The compression test is complex and the protocol influences the results (Hodgkinson, 2000). In addition, between the stems and the industrial flax fibres there are several steps (retting, scutching, combing, manufacturing of unidirectional preform, etc.) which can influence the properties of the cell walls.
3.7. Compressive behaviour of unidirectional composites Fig. 8 shows the buckling of flax fibres inside a unidirectional laminate loaded in compression. This is a standard behaviour for composite materials. Complementary, Fig. 9 shows the longitudinal compressive strength of flax/epoxy composites as a function of fibre volume fraction. The strength increasing with the fibre volume fraction illustrates a reinforcing effect. The maximum measured strength is 115 MPa for a fibre volume fraction of 50.9%. This limited value confirms the literature data presented in Table 1 for composites reinforced by flax fibres. Otherwise, this value is significantly lower than that measured for glass-fibre reinforced composites (Lo and Chim, 1992), e.g. in the case of glass E/ epoxy UD with 45% of fibre volume fraction, the longitudinal compressive strength is 621 MPa (Gibson, 2012b). Obviously, other parameters could be taken into account to refine
4. Discussion - correlation between the longitudinal compressive strength of flax/epoxy unidirectional and the compressive strength of bundles within a stem Furthermore, by an inverse approach and using Fig. 9, the compressive strength leading to the fibres micro-buckling can be estimated through different approaches. These approaches aim at comparing the properties of flax stems with unidirectional composites, rather than providing a micromechanical model to predict the compressive strength of flax bundles.
Fig. 8. Typical kinking failure in a flax-epoxy after a compressive load. 31
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Fig. 9. Compressive strength as a function of fibre volume fraction of unidirectional flax/epoxy composite.
First approach
Table 6 Estimated compressive strength of flax fibres from unidirectional composite tests and stem bending tests.
Fibre compressive strength (σfL C) can be estimated from the compressive strength of the unidirectional composite (σUD L C) using the following relationship (Kozez et al., 1995):
σfL C =
σUD L C Vf
Compressive strength with fibre buckling (MPa) UD first approach UD second approach Stem Bolchoï Stem Eden Stem Ariane Stem Liral Prince
(7)
The simple hypothesis is that the compressive strength is linearly proportional to the fibre volume fraction (Vf). This hypothesis is validated for glass fibres (Lo and Chim, 1992). With the set of results of this article, the longitudinal compressive strength of flax fibres is estimated to 260 ± 59 MPa (average of the measured values by taking all the specimens).
In a second approach, the contribution of the matrix is considered. In this case, the fibre buckling strain is much lower than the strain at break of the epoxy. The longitudinal compressive strength of a unidirectional (σUD L C) can be approached using a mixture law: (8)
With (Vf) the fibre volume fraction, (σfL C) the compressive strength causing the fibre buckling and (σm*) the stress in the matrix during the fibres buckling. Considering an elastic and linear behaviour, the Eq. (8) becomes:
σUD L C = Vf *σfL C + (1 − Vf )*εfLC *Em
(9)
With (Em) the matrix modulus (here Em = 3 GPa) and (εfLC) the strain causing fibre buckling. This last value is determined by bending tests on stems (0.4% ± 0.05). The longitudinal compressive strength of flax fibres can be estimated by Eq. (10):
σfL C =
59 44 26 8 16 15
Is the increase of the compressive strength of UD composites reinforced by flax fibres possible? By impregnating the cell walls with a melamine formaldehyde resin (MF) (there is penetration of MF into the cell walls) before impregnation with an epoxy resin, Bos et al. (2004) show that the compressive strength at break of composites increases, but the tensile strength drops down. Furthermore, the use of vegetable fibres is often justified to reduce environmental impacts, and a life-cycle analysis will be necessary to know the consequences of this cell wall treatment.
σUD L C − (1 − Vf )*εfLC *Em Vf
± ± ± ± ± ±
regarding UD (260 MPa and 242 MPa respectively), and a difference of less than 15% between average values of flax varieties (from 183 MPa to 207 MPa for Liral Prince and Eden, respectively). The estimated strength is slightly lower for the old varieties (Liral Prince and Ariane) due to their low fibre content. The poor compression strength of flax fibres in a unidirectional composite is also observed within the stems. This is a limit for the use of flax for the reinforcement of a polymer, as in the case of organic fibres (Kozez et al., 1995). A comparison with wood is also possible. In fact, wood is a composite material composed of an assembly of fibres mainly oriented along the axis of primary growth. The mechanism leading to the compressive break of wood is a kink banding phenomenon (Poulsen et al., 1997), such as observed in unidirectional composites reinforced by flax. The compressive behaviour of wood is linearly proportional to the specific gravity (Domone and Illston, 2010). This relationship allows the modelling of structures, and remains valid whether the wood is green or air dried. Thus, (Gibson, 2012a) estimated the axial compressive strength of a wood cell wall as 120 MPa, which corresponds to a densified wood with no lumen. Greater values are obtained for flax, mainly due to higher cellulose content and lower microfibrillar angle.
Second approach
σUD L C = Vf *σfL C + (1 − Vf )*σm*
260 242 206 207 189 183
(10)
With the set of results of this article, the longitudinal compressive strength of flax fibres is estimated to 242 MPa ± 44 MPa. For comparison, Table 6 shows the estimated compressive strength of flax fibres from unidirectional composite tests and stem bending tests. The estimated values (Table 6) are of the same order of magnitude, with a difference of less than 10% between the two approaches 32
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
5. Conclusion
and Use of Synthetic Materials for Aircraft Construction. Flight 77–79. Delaplace, A., Gatuingt, F., Ragueneau, F., 2008. MÉCANIQUE DES STRUCTURES Résistance des matériaux. Dunod. Deteresa, S.J., Porter, R.S., Farris, R.J., 1988. Experimental verification of a microbuckling model for the axial compressive failure of high performance polymer fibres. J. Mater. Sci. 23, 1886–1894. https://doi.org/10.1007/BF01115735. Domone, P., Illston, J., 2010. Construction Materials their Nature and Behaviour, fourth edition. Spon Press. Evert, R., 2006. Esau’s Plant Anatomy, third edition. John Wiley & Sons. Evert, R., Eichhorn, S., 2013. Raven Biology of Plants. W. H. Freeman and Company. Evon, P., Barthod-Malat, B., Grégoire, M., Vaca-Medina, G., Labonne, L., Ballas, S., Véronèse, T., Ouagne, P., 2018. Production of fiberboards from shives collected after continuous fiber mechanical extraction from oleaginous flax. J. Nat. Fibers 0, 1–17. https://doi.org/10.1080/15440478.2017.1423264. Garland, B.D., Beyerlein, I.J., Schadler, L.S., 2001. The development of compression damage zones in fibrous composites. Compos. Sci. Technol. 61, 2461–2480. https:// doi.org/10.1016/S0266-3538(01)00176-2. Gibaud, M., Bourmaud, A., Baley, C., 2015. Understanding the lodging stability of green flax stems; the importance of morphology and fiber stiffness. Biosyst. Eng. 137, 9–21. https://doi.org/10.1016/j.biosystemseng.2015.06.005. Gibson, L.J., 2012a. The hierarchical structure and mechanics of plant materials. J. R. Soc. Interface 9 (76). Gibson, R., 2012b. Principles of Composites Materials Mechanics. Taylor & Francis. Gorshkova, T., Gurjanov, O., Mikshina, P., Ibragimova, N., Mokshina, N., Salnikov, V., Ageeva, M., Amenitskii, S., Chernova, T., Chemikosova, S., 2010. Specific type of secondary cell wall formed by plant fibers. Russ. J. Plant Physiol. 57, 328–341. https://doi.org/10.1134/s1021443710030040. Goudenhooft, C., Bourmaud, A., Baley, C., 2017. Varietal selection of flax over time: evolution of plant architecture related to influence on the mechanical properties of fibers. Ind. Crops Prod. 97, 56–64. https://doi.org/10.1016/j.indcrop.2016.11.062. His, I., Andème-Onzighi, C.C., Morvan, C., Driouich, A., 2001. Microscopic studies on mature flax fibers embedded in LR white: immunogold localization of cell wall matrix polysaccharides. J. Histochem. Cytochem. 49, 1–11. Hodgkinson, J., 2000. Mechanical Testing of Advanced Fibre Composites. Woodhead Publishing Ltd. and CRC Press LLC. Khalfallah, M., Abbès, B., Abbès, F., Guo, Y.Q., Marcel, V., Duval, A., Vanfleteren, F., Rousseau, F., 2014. Innovative flax tapes reinforced Acrodur biocomposites: a new alternative for automotive applications. Mater. Des. 64, 116–126. https://doi.org/10. 1016/j.matdes.2014.07.029. Kozez, V., Jiang, H., Mehta, V., Kumar, S., 1995. Compressive behavior of materials: part II. High performance fibers. Mater. Res. Soc. 10, 1044–1061. Le Duigou, A., Kervoelen, A., Le Grand, A., Nardin, M., Baley, C., 2014. Interfacial properties of flax fibre–epoxy resin systems: existence of a complex interphase. Compos. Sci. Technol. 100, 152–157. https://doi.org/10.1016/j.compscitech.2014. 06.009. Leal, A., Deitzel, J., Gillespie, J., 2007. Assessment of compressive properties of high performance organic fibers. Compos. Sci. Technol. 67 (67), 2786–2794. Leblicq, T., Vanmaercke, S., Ramon, H., Saeys, W., 2015. Mechanical analysis of the bending behaviour of plant stems. Biosyst. Eng. 129, 87–99. https://doi.org/10. 1016/j.biosystemseng.2014.09.016. Lefeuvre, A., Bourmaud, A., Baley, C., 2015. Optimization of the mechanical performance of UD flax/epoxy composites by selection of fibres along the stem. Compos. Part A: Appl. Sci. Manuf. 77, 204–208. https://doi.org/10.1016/j.compositesa.2015.07.009. Liang, S., Gning, P.-B., Guillaumat, L., 2015. Quasi-static behaviour and damage assessment of flax/epoxy composites. Mater. Des. 67, 344–353. https://doi.org/10.1016/j. matdes.2014.11.048. Livingston Smith, S., Mech, M., 1945. A survey of plastics from the viewpoint of the mechanical engineer. Inst. Mech. Eng. 29–43. Lo, K.H., Chim, E., 1992. Compressive strength of unidirectional composites. J. Reinf. Plast. Compos. 11, 838–896. Mahboob, Z., El Sawi, I., Zdero, R., Fawaz, Z., Bougherara, H., 2017. Tensile and compressive damaged response in flax fibre reinforced epoxy composites. Compos. Part A: Appl. Sci. Manuf. 92, 118–133. https://doi.org/10.1016/j.compositesa.2016.11.007. Mattheck, C., 1995. Biomechanical optimum in woody systems. In: Gartner, B. (Ed.), Plant Stems: Physiology and Functional Morphology. Academic Press Limited, London, UK, pp. 27–52. Poulsen, J.S., Moran, P.M., Shih, C.F., Byskov, E., 1997. Kink band initiation and band broadening in clear wood under compressive loading. Mech. Mater. 25, 67–77. https://doi.org/10.1016/S0167-6636(96)00043-9. Réquilé, S., Goudenhooft, C., Bourmaud, A., Le Duigou, A., Baley, C., 2018. Exploring the link between flexural behaviour of hemp and flax stems and fibre stiffness. Ind. Crops Prod. 113. https://doi.org/10.1016/j.indcrop.2018.01.035. Rihouey, C., Paynel, F., Gorshkova, T., Morvan, C., 2017. Flax cellulosic fibres: how to assess the non-cellulosic polysaccharides and to approach cell wall supramolecular design. Cellulose 24, 1985–2001. Roussière, F., Baley, C., Godard, G., Burr, D., 2012. Compressive and tensile behaviours of PLLA matrix composites reinforced with randomly dispersed flax fibres. Appl. Compos. Mater. 19, 171–188. https://doi.org/10.1007/s10443-011-9189-8. Snegireva, A.V., Ageeva, M.V., Amenitskii, S.I., Chernova, T.E., Ebskamp, M., Gorshkova, T.A., 2010. Intrusive growth of sclerenchyma fibers. Russ. J. Plant Physiol. 57, 342–355. https://doi.org/10.1134/s1021443710030052. Timoshenko, S.P., 1947. Strength of Materials - Part 2 - Advanced Theory and Problems. 9th printing. van, Nostrand. . Van Vuure, A.W., Baets, J., Wouters, K., Hendrickx, K., 2015. Compressive properties of natural fibre composites. Mater. Lett. 149, 138–140. https://doi.org/10.1016/j. matlet.2015.01.158. Zaune, M., Stampanoni, M., Niemz, P., 2016. Failure and failure mechanisms of wood during longitudinal compression monitored by synchrotron micro-computed tomography. Holzforschung 70 (2). https://doi.org/10.1515/hf-2014-0225.
This work illustrated the compression buckling mechanisms of flax fibre bundles within a stem. The analysis of the measured values made it possible to estimate the stress causing bundle instability within a flax stem. The stem organisation influences the behaviour; i.e. the compression strength of the bundles was slightly lower for the older varieties, these latter having a smaller fraction of fibres. In addition, thanks to longitudinal compressive tests on unidirectional laminates, the limited reinforcement of flax fibres was illustrated. By an inverse method, the theoretical compressive strength of flax fibres was estimated. Both approaches led to the same orders of magnitude. Despite an optimal arrangement of the fibres in the stems and a strong fibre cohesion, the fibre compression strength remains limited. This assessment is a limit of the use of flax fibres as polymer reinforcement, especially when significant compressive properties are required. The central part of the stem (woody core) exhibits an apparent strain higher in compression despite the presence of large lumens, which can be explained by its conductive function. This core tissue significantly contributes to the compressive strength of the stem as in a structure sandwich. To go further, future works will be carried out on living plants for a better understanding of their strengthening mechanisms. Acknowledgements The authors would like to thank Dr. Marine Lan for part of the experimental tests, the French Ministry of Research and Innovating Technologies, OSEO, Région Bretagne for their financial support. The authors also acknowledge the 3D-BioMat project that is co-financed by the Grand Reims (31%) and the European Union by 48.7% (i.e. 50% of eligible expenditure). Europe is committed to the Grand Est with the European Regional Development Fund. The 3D-BioMat project with a total budget of 965,000€ is hosted by the Centre Européen de Biotechnologie et de Bioéconomie (CEBB 51110 Pomacle, France) for a three-year period (from 01/05/2016 to 30/04/2019). References Andème-Onzighi, C., Girault, R., His, I., Morvan, C., Driouich, A., 2000. Immunocytochemical characterization of early-developing flax fiber cell walls. Protoplasma 213, 235–245. https://doi.org/10.1007/bf01282161. Attwood, J.P., Fleck, N.A., Wadley, H.N.G., Deshpande, V.S., 2015. The compressive response of ultra-high molecular weight polyethylene fibres and composites. Int. J. Solids Struct. 71, 141–155. https://doi.org/10.1016/j.ijsolstr.2015.06.015. Baley, C., 2002. Analysis of the flax fibres tensile behaviour and analysis of the tensile stiffness increase. Compos. Part A: Appl. Sci. Manuf. 33, 939–948. Baley, C., Bourmaud, A., 2014. Average tensile properties of French elementary flax fibers. Mater. Lett. 122, 159–161. https://doi.org/10.1016/j.matlet.2014.02.030. Baley, C., Le Duigou, A., Bourmaud, A., Davies, P., Nardin, M., Morvan, C., Le Duigou, A., Bourmaud, A., Davies, P., Nardin, M., Morvan, C., 2014. Reinforcement of polymers by flax fibers: role of interfaces. Bio-Based Composites for High-Performance Materials. CRC Press, pp. 87–112. https://doi.org/10.1201/b17601-7. Baley, C., Goudenhooft, C., Gibaud, M., Bourmaud, A., 2018a. Flax stems: from a specific architecture to an instructive model for bioinspired composite structures. Bioinspir. Biomim. 13, 026007. https://doi.org/10.1088/1748-3190/aaa6b7. Baley, C., Lan, M., Bourmaud, A., Le Duigou, A., 2018b. Compressive and tensile behaviour of unidirectional composites reinforced by natural fibres: influence of fibres (flax and jute), matrix and fibre volume fraction. Mater. Today Commun. 16, 300–306. https://doi.org/10.1016/j.mtcomm.2018.07.003. Benabou, L., 2008. Kink band formation in wood species under compressive loading. Exp. Mech. 48, 647–656. https://doi.org/10.1007/s11340-007-9098-9. Bert, F., 2013. Lin Fibre: culture et transformation. Arvalis. Bos, H.L., Molenveld, K., Teunissen, W., van Wingerde, A.M., van Delft, D.R.V., 2004. Compressive behaviour of unidirectional flax fibre reinforced composites. J. Mater. Sci. 39, 2159–2168. https://doi.org/10.1023/b:jmsc.0000017779.08041.49. Bourmaud, A., Baley, C., 2009. Rigidity analysis of polypropylene/vegetal fibre composites after recycling. Polym. Degrad. Stab. 94, 297–305. https://doi.org/10.1016/j. polymdegradstab.2008.12.010. Brazier, L., 1927. On the flexure of thin cylindrical shells and other ‘Thin’ sections. Proc. R. Soc. A: Math. Phys. Eng. Sci. 116 (773), 104–114. Catling, D., Grayson, J., 1982. Identification of Vegetable Fibres. Archetype Press, London. Charlet, K., Jernot, J.P., Gomina, M., Bréard, J., Morvan, C., Baley, C., 2009. Influence of an Agatha flax fibre location in a stem on its mechanical, chemical and morphological properties. Compos. Sci. Technol. 69, 1399–1403. de Bruyne, N.A., 1939. Plastic Progress - Some Further Developments in the Manufacture
33
Contents lists available at ScienceDirect
Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop
Compressive strength of flax fibre bundles within the stem and comparison with unidirectional flax/epoxy composites
T
Christophe Baleya, Camille Goudenhoofta, Patrick Perréb, Pin Lub, Floran Pierreb, ⁎ Alain Bourmauda, a b
Univ. Bretagne Sud, UMR CNRS 6027, IRDL, F-56100 Lorient, France LGPM, CentraleSupélec, Université Paris-Saclay, Centre Européen de Biotechnologie et de Bioéconomie (CEBB), 3 rue des Rouges Terres, 51 110 Pomacle, France
A R T I C LE I N FO
A B S T R A C T
Keywords: Buckling Composite materials Compressive strength Flax fibres Stem Unidirectional composite
Flax (Linum usitatissimum L.) fibres are commonly used as reinforcement of composite materials. Nevertheless, literature shows that the compressive strength of flax-based composites is rather modest compared with materials reinforced by synthetic fibres. The present article investigates the compressive strength of flax fibre bundles both within the stems and in unidirectional (UD) composites. In this way, an optimised arrangement of fibre bundles inside the plant is assumed. Damage mechanisms are found to be similar in the stem and within flaxbased UD materials, namely by buckling of fibre bundles, a typical failure mechanism of UD composites. Inside the stems, this phenomenon is highlighted by nanotomography, which underlines the key role of the woody core in the buckling resistance of the plant. For UD, failure can also be studied by scanning electron microscopy (SEM). The same ranges of average compressive strength values are estimated for flax fibre bundles, being 206 MPa within the stem and 242 MPa within UD composites. Finally, this study highlights that, if a flax stem is an optimised natural structure, the compressive strength of flax fibre bundles seems to be a limiting factor for structural applications of flax-based composite materials.
1. Introduction Flax (Linum usitatissimum L.) fibres have been used for centuries for textile applications. Their use as reinforcement of composite materials started more recently, in the 1940s (de Bruyne, 1939). This latest application is the subject of many research works, the main purposes of flax-based composite development being for instance the reduction of weight of manufactured parts or the decrease of their environmental impacts. The term of flax fibres used in the present work refers to elementary fibre cells located at the periphery of the stem, forming a fibrous ring composed of fibre bundles (Baley et al., 2018a; Goudenhooft et al., 2017). Flax fibres have a key role of supporting tissues due to noteworthy mechanical properties. Thus, they ensure the plant stability and allow it to present a remarkably slender structure, assimilated to a composite material composed of the epidermis as an external protection, a unidirectional ply of flax fibres at the periphery and a xylem part as the innermost porous core (Baley et al., 2018a). The flexural mechanical characterisation of a flax stem demonstrates that these plant fibres greatly contribute to the flexural stiffness of the stem (Gibaud et al., 2015). For example, taking into account a ⁎
dried stem, flax fibres were proven to ensure about 70% of the stem stiffness (Réquilé et al., 2018). In the case of dried hemp stems, the fibrous part only contributes to about 50% of the plant stiffness. The outstanding properties of flax fibres, and more specifically the length of elementary cells and associated mechanical properties (Baley and Bourmaud, 2014), are the results of an intrusive fibre development in a first place (Snegireva et al., 2010), which is followed by the thickening of the fibre cells through the deposition of cellulose (Rihouey et al., 2017). This latter process provides fibres exhibiting a very thick secondary cell wall and small lumen (Charlet et al., 2009). Within the stem, pectic junctions bond the fibres together and ensure the bundle arrangement. This structure gives a first model of a composite material at the bundle scale. In this latter, the stresses are transferred to the reinforcing fibres through a polymer matrix, made of pectic junctions, mainly composed of homogalacturonans and rhamnogalacturonans (Andème-Onzighi et al., 2000). This matrix can be found in two different areas: the middle lamella (i.e. between the primary cell walls of two neighbouring fibres) and the tri-cellular junctions (i.e. in the corner between three fibres) (Baley et al., 2014). Unfortunately, the compressive strength of plant stems is poorly studied in literature but it has been evidenced by Mattheck (1995) that hollow
Corresponding author. E-mail address: [email protected] (A. Bourmaud).
https://doi.org/10.1016/j.indcrop.2018.12.059 Received 19 June 2018; Received in revised form 14 December 2018; Accepted 18 December 2018 0926-6690/ © 2018 Elsevier B.V. All rights reserved.
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Table 1 Properties of unidirectional composites reinforced by flax fibres provided in literature, highlighting longitudinal compressive and tensile properties. Matrix
Fibre volume fraction (%)
Compressive modulus (GPa)
Compressive strength (MPa)
Tensile modulus (GPa)
Tensile strength (MPa)
References
Phenol formaldehyde Phenol formaldehyde
˜70 ˜75
41.4 48.3
163 182
41.4 48.3
310 414
Epoxy Epoxy Epoxy
43.9 40 51
24.7 15.1 30.3
136 137 127
22.8 23.9 31.4
318 223 287
(de Bruyne, 1939) (Livingston Smith and Mech, 1945) (Liang et al., 2015) (Van Vuure et al., 2015) (Mahboob et al., 2017)
compressive strength, which remains lower than 140 MPa
stems are detrimental to stability due to Brazier buckling (Brazier, 1927). Oppositely, when stems exhibit thick-walled structures, the failure is preferentially attributed to mechanical limit of tissues components (Leblicq et al., 2015). In case of flax fibres, despite remarkable longitudinal mechanical performances, these latter evidence poor transverse properties and especially stiffness (Bourmaud and Baley, 2009), which makes them very sensitive to transverse or compressive stress. This sensibility may be a significant drawback if processing steps such as pulling-out or scutching are too aggressive. Thus, despite this remarkable structure and tissue organisation of plant fibre reinforcement, the resistance to compressive strength of flaxbased composites is rather limited, whatever the type of matrix or the fibre volume fraction, when compared to their resistance to tensile strength (Roussière et al., 2012). Table 1 shows the longitudinal compressive mechanical properties of different unidirectional (UD) of flaxreinforced composites (Bos et al., 2004; de Bruyne, 1939; Liang et al., 2015; Livingston Smith and Mech, 1945; Mahboob et al., 2017; Van Vuure et al., 2015). For all the data sets given in Table 1, the compressive strength is lower than 200 MPa. In addition, it is always poorer than the tensile strength (which is about two times higher). Therefore, this mechanical parameter is of significance for the improvement of flax biocomposites. For comparison, aramid fibre may have an apparent compressive strength down to only 10–15% of the tensile strength (Kozez et al., 1995; Leal et al., 2007). Numerous parameters can influence the compressive behaviour of composite materials, such as fibre volume fraction, the fibre and matrix properties, the interface bond strength, the porosity, the fibre misalignment, as well as the difference in Poisson ratio between fibre and matrix (de Bruyne, 1939; Liang et al., 2015; Livingston Smith and Mech, 1945; Mahboob et al., 2017; Van Vuure et al., 2015). In the case of a unidirectional ply under longitudinal compression, the failure takes place due to fibre kinking; this result was demonstrated for natural fibres (Poulsen et al., 1997) as well as synthetic ones (Garland et al., 2001). A previous article (Baley et al., 2018b) presents a detailed investigation of the longitudinal compressive behaviour of composites reinforced by unidirectional plant fibre tapes. We study and discuss the influence of the fibre type (flax or jute), polymer (PP, PP/MAPP, PA11, epoxy or acrylic), volume fraction and the quality of interfacial shear strength (IFSS) with a maleic anhydride grafted PP (MAPP). The results obtained for composites reinforced by unidirectional flax fibres show that:
• The fibre nature (flax or jute) also play an important role. Compared
to flax, the use of jute fibres leads to lower mechanical properties (in tensile and in compression).
The present work investigates the compressive behaviour of flax fibre bundles within the stem. In this way, fibres are considered under an optimal situation within the stem, i.e. cohesive and oriented in the longitudinal direction of the stem, leading to a model of natural composite structure. First, loading is performed through flexural bending of the stems. Then, the induced damaging and failure mechanisms of fibre bundles are observed by nanotomography. Complementary experiments are performed on several flax varieties to evaluate the influence of the varietal selection work on the bundle compressive strength. In addition, the compressive strength at break of fibre bundles within the stems is compared with the ones of unidirectional plies. Finally, the estimation of the role in compression of the xylem, also called woody core, is given. 2. Material and methods 2.1. Plant material – flax stems and fibres All flax plants were provided by Terre De Lin (a flax cooperative based in Normandy, France). Two recent as well as two ancient textile flax varieties were studied. The recent varieties are Eden (registered in 2009) and Bolchoï (registered in 2014); they were selected by Terre De Lin and are both currently commercialised. Liral Prince, from a selection made in 1944 by the Linen Industry Research Association (L.I.R.A, Ireland)(Goudenhooft et al., 2017) and Ariane, registered in 1978 by the Coopérative Linière de Fontaine Cany (France) are the two ancient varieties studied. They are not commercialised anymore. Plants were cultivated in Saint-Pierre le Viger (Normandy, France) in 2016. Conventional seeding rates and culture conditions were used (Bert, 2013). Entire plants were pulled out at maturity (Goudenhooft et al., 2017) and subsequently dried under a standard atmosphere (temperature of 23 ± 2 °C and relative humidity of 50 ± 4%) until stabilisation of the stem weight. Some of the dried stems were waterretted (in order to facilitate the extraction of fibres for tensile tests) then dried as well. Prior to bending and tensile tests, stems and fibres were conditioned under the previously mentioned standard atmosphere for at least 24 h. Water contents of both stems and fibres were measured to be 7.8 ± 0.4% after conditioning.
• for small strains, the compressive modulus is closely similar to the tensile modulus. • the compressive strength is always lower than the tensile strength. • In compression, the failure mechanism involves the microbuckling • •
2.2. Three-point bending test on stem Three-point bending tests were performed on intact flax stem samples. Additional stems were carefully peeled, i.e. fibres were peeled off, and tested similarly in order to examine only the woody part of the samples. All stem samples were extracted from the middle part of the plants i.e. from the location of maximum fibre content (Lefeuvre et al., 2015). Three-point bending tests were carried out following the protocol
of flax fibres (as seen in composites reinforced by carbon, glass or organic fibres). The enhancement of IFSS using PP/MAPP or by selecting an appropriate matrix leads to better compressive properties. Nevertheless, a large change of polymer matrix or an increase in the fibre volume fraction does not imply a drastic increase of 26
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
2.4. Nanotomography
proposed by Réquilé et al. (2018), under the standard atmosphere (previously mentioned). Tests were performed on a universal MTS tensile testing machine equipped with a 50 N capacity load cell and the displacement rate was fixed at 0.1 mm/s. Experiments were conducted on 20 mm flax stems (and stem after peeling), using a span length of 18 cm in order to provide a length-todiameter ratio high enough to avoid shear effects (Réquilé et al., 2018). First, by considering both the entire stem and the woody core as a uniform plain beam with a circular cross section, the displacement y in the middle of the sample section read as (Timoshenko, 1947):
y=
FL3 48EIsample
The samples were scanned with a nanotomograph EasyTom XL 150/ 160 manufactured by RX-solutions (Annecy, France). The nano-focus Xray sources (an open nano-focus tube with a tungsten filament) and a CCD detector of 2004 × 1336 pixels were selected to obtain a voxel size of 1.7 μm, the best possible compromise to scan the entire stem diameter. The region of interest of the flax stem (Bolchoï variety) after bending was glued on the top of the 2 mm-carbon rod of the sample holder. This protocol allows the flax stem to be close enough to the Xray tube to obtain this resolution. These conditions impose an acquisition time of about 3 h.
(1)
2.5. Scanning electron microscopy (SEM)
Where F is the force, L the span length, (EIsample) the bending stiffness either of the stem or the woody core, Isample the quadratic momentum of the sample and E its apparent bending modulus. The stem or woody core bending stiffness for small strains is then obtained by applying the following formula:
EIsample =
dF L3 dy 48.
Composite samples were embedded in an epoxy resin before polishing. They were then metallised with gold before being observed under a JEOL JSM 6460LV scanning electron microscope. 2.6. Stem anatomy and bundle thickness
(2) Stem portions were taken at the middle height of the flax plants. To facilitate the cut, the portions were first immersed in water at least 24 h. Then, they were embedded in elder marrow and semi-thin sections were cut transversally using a razor blade and a hand microtome to clamp the sample. Optical images were taken under the microscope and used to estimate the bundle thickness.
Where dF/dy is the slope of the linear part (small displacements) of the force-displacement curve. The maximal force Fmax damaging the sample is determined for each test; the corresponding maximum bending moment Mmax at the centre of the sample is estimated by Eq. (3):
Mmax =
Fmax L 4
(3)
2.7. Unidirectional composites
At mid-length of the beam, the maximal strain εmax is expressed by Eq. (4):
εmax =
Mmax D EI 2
The matrix used is an epoxy resin (Axson, Epolam 2020). Its tensile Young’s modulus, strength and strain at break are 3.10 ± 0.01 GPa, 78 ± 6 MPa, and 3.1 ± 0.5%, respectively. The reinforcement consists of a tape of unidirectional fibres without any twist (Flax-Tape, Lineo®) (Khalfallah et al., 2014) with an area weight of 200 g/m². The processing of flax (Eden variety) tapes begins by decorticating and separating the bundles from the flax stems. The tapes are then cleaned by scutching to remove the remaining shives and woody core of the stems. The flax bundles are aligned to the input of the manufacturing machine and moistened by a water mist and continuously dried through an infrared oven. To maintain the cohesion of the parallel fibers, the process is based on the reactivation of the pectin cement during spraying a water mist. As in a stem, the unidirectional ply is reinforced by bundles (the bundles are not cut off). The unidirectional laminates were obtained by impregnating and then pressing to eliminate voids (Labtech© 50 T moulding press). Fibre volume fractions from 20% to 52% were measured by image analysis from SEM observations on transverse sections after polishing. After hardening for 24 h at 25 °C, samples were post-cured (3 h –40 °C; 2 h –60 °C; 2 h –80 °C; 5 h −100 °C). For compressive characterisation and after preliminary tests, the thickness of samples is set at 2.5 mm. For the test, two gauges were bonded, one on each face, in order to monitor the compressive behaviour.
(4)
In the case of the peeled stem, the sample is considered as a homogeneous material having a linear mechanical behaviour (validated Hooke’s law). The strength is thus calculated as the multiplication of the maximal strain by the apparent elastic modulus determined through bending tests. However, for the entire stem, a sandwich composite beam is considered, having an outer unidirectional ply on the outermost side (assembly of fibres in the form of bundles forming an external fibrous ring). In this case, the product EIsample of previous equations should be expressed as Eq. (5):
EIs =
π (Ef (Ds4 − Dc4 ) + Ec Dc4 ) 64
(5)
Where D is the diameter and E the elastic modulus with indices f, s and c standing for fibres, stem and woody core respectively. The elastic modulus of the fibres Ef is measured by tensile tests. In the present case, the fibre behaviour is simplified as being elastic linear. Thus, the bundles compressive strength at break σfmax expressed by Eq. (6):
σf max = εmax Ef
(6)
3. Results 2.3. Tensile tests on elementary flax fibres 3.1. Mechanical properties of elementary flax fibres Tensile tests on elementary fibres were performed according to NFT 25-501-2 standard. Fibres were manually extracted from retted stems and bonded onto a paper frame with a gauge length of 10 mm. The fibre samples were conditioned for at least 24 h under standard atmosphere. The diameter of each fibre was determined as the average of 6 measurements taken along the fibre using optical microscope. The samples were finally clamped on a universal MTS tensile testing machine equipped with a 2 N capacity load sensor using a crosshead speed of 1 mm/min. For each batch of fibres, a minimum of 70 fibres was tested.
In order to analyse the compressive behaviour of flax stems, it is necessary to know the mechanical properties of the elementary fibres composing their bundles. More particularly, the Young’s modulus is required for the studied batches. Table 2 presents the average tensile properties of elementary fibres extracted from the related stem batches according to the stem variety. These performances can be compared with the work of Baley and Bourmaud (2014) which have analysed numerous batches of 27
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Table 2 Tensile mechanical properties of elementary fibres of the four flax varieties. Variety
Fibre diameter (μm)
Young’s modulus (GPa)
Strength at break (MPa)
Strain at break (%)
Bolchoï Eden Ariane Liral Prince
22.7 16.6 17.4 21.8
52.9 54.9 43.6 40.9
1012 ± 283 1056 ± 310 960 ± 288 860 ± 268
2.12 2.22 2.39 2.39
± ± ± ±
3.9 2.8 2.6 2.4
± ± ± ±
12.2 10.5 9.6 10.2
± ± ± ±
0.59 0.62 0.49 0.62
Fig. 2. Nanotomography of the damaged area of a flax stem after a bending test. A local buckling is visible, due to excessive compressive strength.
elementary fibres, tested under a similar standard atmosphere and the same protocol than the present study. Average mechanical properties available in the mentioned literature are a Young’s modulus reaching 52.5 ± 8.6 GPa, a strength at break of 945 ± 200 MPa and a strain at break of 2.07 ± 0.45% for an average fibre diameter of 16.8 ± 2.7 μm. One can notice that, despite slightly lower tensile Young’s modulus and strength at break for the oldest varieties Liral Prince and Ariane, the diameters and mechanical properties of elementary fibres measured in the present work belong to the same ranges as literature values (Baley and Bourmaud, 2014). These similarities confirm the good reproducibility of flax fibres mechanical properties.
span length. The stem damage actually happens on the face in compression, and is often located close to the central support of the bending device. This observation is consistent with a three-point bending test, for which the bending moment is maximum at mid-length, which confirms the hypothesis of a behaviour very similar to pure bending. 3.3. Failure mechanisms of flax stems during bending tests Observations of a damaged stem through nanotomography highlight the mechanism of local buckling in the part of the stem in compression (Fig. 2). In addition, the nanotomographic investigations enable to visualize a longitudinal section of the stem within the damaged zone (Fig. 3.(A)). Thanks to these observations, successive damages can be identified within a stem with, first, the local buckling of a fibre bundle together with the formation of a kink-band (Fig. 3.(A)). The latter phenomenon is also visible on elementary fibres while bending (Baley, 2002), but also for wood under compressive strength (Benabou, 2008; Zaune et al., 2016) as well as for synthetic fibres such as UHMWPE (Attwood et al., 2015) or aramid (Deteresa et al., 1988). Then, the detachment of the bundle from the woody core is visible in a small part of the stem. Finally, the local buckling of other cell types, namely conduction cells within the woody core, is also visible in Fig. 3. This latter figure also shows that the cohesion between the fibre bundles and the woody core is a key parameter to the resistance of fibre bundles to local buckling. Other elements are involved in the compression behaviour of the fibre bundles such as the middle lamellae ensuring the bonding between the fibres within a bundle and between the bundles of the other cells, if the latter are not damaged by the retting step; moreover, the cortex ensuring the load transfer between the bundles is also a major contributor of bundles compression resistance (Baley et al., 2018a). However, one can notice that nanotomographic images do not allow to distinguish elementary fibres given the absence of density difference between fibres within the bundle. To facilitate the examination of the previous figure, Fig. 3.(B) shows the organisation of the flax stem in transverse cross-section (Bolchoï variety). The composite structure of the stem is highlighted, with two main constitutive parts: fibre bundles located at the periphery of the stem and the woody core of the plant (mostly xylem) located between the fibres and the central cavity (Catling and Grayson, 1982; Evert, 2006; Evert and Eichhorn, 2013; Gorshkova et al., 2010; His et al., 2001). For more information on the schematic organization of a stem, the reader may consult the following references: for a stem section (Catling and Grayson, 1982; Réquilé et al., 2018) and for a longitudinal section (Catling and Grayson, 1982). For a thin-walled tube solicited in bending, the risk of local buckling is a function, among other parameters, of the ratio of the external diameter of the tube and its thickness (Delaplace et al., 2008). Namely, for a given diameter, the smaller the wall thickness is, the higher the instability exists. Similarly, a flax stem is comparable to a thin-walled tube containing a core, i.e. a circular sandwich structure with an inner
3.2. Mechanical behaviour of flax stems and woody cores during bending tests The bending stiffness of flax stems is evaluated by three-point bending tests. Fig. 1 presents the force-displacement profiles for both the flax stem (with a diameter of 2.10 mm) and the woody part (with a diameter of 1.89 mm). Similar elastic and linear behaviours are exhibited at the beginning of the test for both the stem and the woody core, which allow us to define the flexural modulus of the stems based on Eq. (2). Considerably lower values are highlighted for the woody core of the stem, for both modulus and maximal strength compared to an entire stem. These results confirm the previously published ones (Réquilé et al., 2018), and underline the key role of fibre part in stem stability. Flax stems can be assimilated to cylindrical tubes of small diameter, taking into consideration the thickness of the fibre ring. Therefore, the bending behaviour of these stems can be described as analogous to those of tubes with thin walls (Leblicq et al., 2015). Considering the large L/D of the samples taken for the bending tests, the loss of linearity of the curve is explained by a large displacement of the central support just before breakage (Fig. 1). In addition, the ovalisation and squashing effects are negligible for all the performed tests thanks to the use of a pertinent
Fig. 1. Typical stress-strain curve for a three-point bending test on flax stem (whole) and woody core (after stem peeling) for Bolchoï plants (span length of 180 mm). 28
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Fig. 3. (A) Nanotomography image of the longitudinal section of a flax stem after the bending test, in the damaged area. The local buckling of the bundle is visible, as well as other cell types. The apparition of a kink-band is also highlighted. (B) Transverse cross-section of a stem of Bolchoï exhibiting thick fibre bundles of 100 μm on average.
a comparison between stems whose diameters range between 1.69 and 2.84 mm. The authors decided to use samples having dispersed diameters in order to verify the method. When estimating the apparent flexural modulus of the stem (Fig. 4.(B)), the results show that it is also slightly influenced by the stem diameter even though the dispersion of the results remains modest (23.8 ± 2.8 GPa) for a natural material such as a flax stem. For each stem sample, the diameter (D), the span length (L), the bending stiffness (EI) and the maximal force damaging the sample (Fmax) are measured or calculated. In addition, the average modulus of elementary fibres was previously determined (Ef = 52.9 GPa, see Table 1). From this data and through related equations, Table 3 gives the resulting stem apparent bending modulus, the maximal compressive strain at break (εmax) and the bundles compressive strength at break (σfmax). In order to illustrate the scattering of the results, Fig. 5 presents the compressive strength at break of fibre bundles as a function of the stem diameter. An absence of correlation between the mentioned properties is noticeable, i.e. there is no link between the compressive failure strength and the stem diameter. Moreover, the values are a little scattered (206 ± 26 MPa) despite the difficulty to precisely achieve the stress value leading to local buckling (shown in Fig. 1) in the part of the stem under compression. The maximal damaging force was used for calculations, but it is possible that the local buckling appears due to an accumulation of damages. This scenario is well described in literature for unidirectional plies reinforced by synthetic fibres solicited in compression (Garland et al., 2001). In this case, the mechanism of kinkband formation is described as a succession of several steps; namely, the fibre breaks first, followed by the subsequent development of a damaged zone. Finally, the kink-band is formed from the shear instability, originated and visible in the damage zone.
porous material. This core has a great contribution to the resistance to local buckling.
3.4. Mechanical behaviour of Bolchoï stems during a three-point bending test For this point, Bolchoï, which is a recent, performing and one of the most cultivated varieties is considered. Fig. 4.(A) highlights a good correlation between the diameter of a flax stem and its bending stiffness measured by three-point bending tests. This piece of information brings
3.5. Mechanical behaviour of peeled stems of Bolchoï during a bending test Complementary, peeled stems of Bolchoï (i.e. samples are deprived of fibre bundles and considered as only composed of the woody core) are submitted to three-point tests under identical conditions than the Table 3 Stem properties obtained from three-point bending tests on Bolchoï stems. Mean values Stem diameter (mm) Bending stiffness (N.m²) Apparent bending modulus (GPa) Compressive strain at break (%) Compressive strength at break (MPa)
Fig. 4. (A) Correlation between the flax stem diameter and the bending stiffness. (B) Correlation between the flax stem diameter and the apparent bending modulus. 29
2.11 ± 0.34 0.025 ± 0.015 23.8 ± 2.8 0.39 ± 0.05 206 ± 26
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Fig. 5. Compressive strength at break of the fibre bundles as a function of the stem diameter.
Fig. 7. Compressive strain at break of the woody core as a function of the sample diameter.
entire stems already tested. The main aim of this analysis is to achieve the strain capacity of the woody core who is often considered as a no structural part of the stem. This component of the stem is constituted of cells of short length, which are responsible for the conduction of the raw sap whereas phloem cells contribute to the elaborated sap conduction. These conductive cells have a great lumen size and thin cell walls, as shown by tomography (Fig. 3.(A)). Furthermore, the biochemical composition of woody cells differs from the fibre ones (Evon et al., 2018), mainly in terms of cellulose content but also of non-cellulosic polymer fraction and nature. On average, when expressed in percentage of cellulose in the dry matter, the woody core reaches 45.6% and fibres 78.7% (Evon et al., 2018). The mechanical behaviour of a peeled stem was shown in Fig. 1. Additionally, such as for entire stems, the bending stiffness of samples with scattered diameters is analysed. Interestingly, as similarly highlighted for complete stems, there is a good correlation between the bending stiffness of the woody core and the sample diameter (Fig. 6). By measuring the bending stiffness of the peeled stems, it is possible to calculate the apparent bending modulus of each sample if considering the woody core as a homogeneous material. In addition, the compressive strain at break obtained from the maximal force damaging the samples (similar hypothesis than for entire stems) is presented in Fig. 7. For the woody core, there is a slight correlation between the strain at break and the sample diameter.
Table 4 Woody core properties obtained from three-point bending tests on peeled samples of Bolchoï. Mean values Stem diameter (mm) Bending stiffness (N m²) Apparent bending modulus (GPa) Compressive strain at break (%) Compressive strength at break (MPa)
2.08 ± 0.28 0.009 ± 0.005 8.5 ± 4.7 0.69 ± 0.05 59 ± 6
Table 4 gives the mechanical properties resulting from three-point bending tests on peeled samples. In comparison with entire stem samples, one can notice the superior strain at break of the woody core (0.69% ± 0.05) compared with fibre bundles (0.39% ± 0.05, values presented in Table 3). Through its greater strain capacity, the contribution of the woody core in limiting the buckling risk of flax stems under compression is confirmed. However, the compressive strength at break of this porous structure is rather low (59 MPa on average) compared to fibre bundles (206 MPa in Table 3). Nevertheless,when compared to the compressive strength at break of wood, estimated using the wood density (Domone and Illston, 2010), a mean value of 45 MPa is expected for wood when using the flax woody core density of 0.455. This later average apparent density of flax woody core was measured (0.455 ± 0.113 g/cm3) for flax in a previous study (Réquilé et al., 2018). Finally, for an equivalent density, the compressive strength at break of the woody core belongs to the same range of values as for wood, and is even somewhat superior.
3.6. Bending mechanical behaviour of flax stems of different varieties The first flax variety, previously studied, is Bolchoï; it is a recent one having a high fibre content. This later parameter is a major criterion of selection for flax breeders. Thus, one can wonder what would be the influence of the fibre content, i.e. of the variety, on the compressive resistance of flax bundles. In a former study (Goudenhooft et al., 2017), the authors investigated the impact of the varietal selection on the stem architecture (namely the plant anatomy) thanks to an analysis of 4 different varieties cultivated during the same year, in the same location and under the same cultural conditions. These varieties, even though cultivated the same year for the study, had been registered in 1944 (Liral Prince), 1978 (Ariane), 2009 (Eden) or 2011 (Aramis). This article (Goudenhooft et al., 2017) highlighted the evolution of the fibre content over selection, namely increased fibre contents for more recent varieties. Fig. 3.(B) presents a transverse stem section of a more recent
Fig. 6. Bending stiffness of peeled stems (assimilated as the woody core) as a function of the sample diameter. 30
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Table 5 Influence of the flax variety and fibre bundle thickness on the compression resistance of the stems. Variety
Stem diameter (mm)
Fibre bundles thickness (μm)
Bending stiffness (N m²)
Apparent bending modulus (GPa)
Compressive strain at break (%)
Compressive strength at break (MPa)
Bolchoï Eden Ariane Liral Prince
2.11 2.02 2.11 2.10
102 ± 18 99 ± 16 70 ± 18 73 ± 13
0.025 0.018 0.020 0.019
23.8 20.8 18.2 18.4
0.39 0.38 0.43 0.45
206 207 189 183
± ± ± ±
0.34 0.23 0.34 0.29
± ± ± ±
0.015 0.006 0.012 0.009
variety, Bolchoï, having thick fibre bundles and a high fibre yield. Based on image analysis (pictures not shown) and three-point bending tests, Table 5 presents the results and links between the compression resistance of the stem and the thickness of their respective fibre bundles. Thus, a slight increase of the compressive strength at break is visible with increasing bundle thickness in the plant. In addition, as visible on the cross-section of Bolchoï (Fig. 3.(B)), the fibrous ring is continuous along the stem periphery, which decreases the risk of buckling. This result is true for recent varieties, whereas ancient ones exhibit gaps between neighbouring bundles (Goudenhooft et al., 2017). Moreover, the elastic modulus of elementary fibres is greater for recent varieties (Table 1), which stiffens the stems.
± ± ± ±
2.8 2.2 1.7 2.4
± ± ± ±
0.05 0.02 0.04 0.04
± ± ± ±
26 8 16 15
these results such as: i) The bundle division and distribution in a ply cross section (the composite microstructure); ii) The fibre orientation inside the composites. Indeed, flax fibres have a limited length and it is not possible to apply a tensile load during the process to obtain a perfect orientation; iii) The bonding between fibre and matrix with this batch of fibres. Nevertheless for this constituents, the good adhesion between flax and epoxy is well described in the literature (Le Duigou et al., 2014); iv) The possible evolution of the cell wall properties during this transformation cycle. The compression test is complex and the protocol influences the results (Hodgkinson, 2000). In addition, between the stems and the industrial flax fibres there are several steps (retting, scutching, combing, manufacturing of unidirectional preform, etc.) which can influence the properties of the cell walls.
3.7. Compressive behaviour of unidirectional composites Fig. 8 shows the buckling of flax fibres inside a unidirectional laminate loaded in compression. This is a standard behaviour for composite materials. Complementary, Fig. 9 shows the longitudinal compressive strength of flax/epoxy composites as a function of fibre volume fraction. The strength increasing with the fibre volume fraction illustrates a reinforcing effect. The maximum measured strength is 115 MPa for a fibre volume fraction of 50.9%. This limited value confirms the literature data presented in Table 1 for composites reinforced by flax fibres. Otherwise, this value is significantly lower than that measured for glass-fibre reinforced composites (Lo and Chim, 1992), e.g. in the case of glass E/ epoxy UD with 45% of fibre volume fraction, the longitudinal compressive strength is 621 MPa (Gibson, 2012b). Obviously, other parameters could be taken into account to refine
4. Discussion - correlation between the longitudinal compressive strength of flax/epoxy unidirectional and the compressive strength of bundles within a stem Furthermore, by an inverse approach and using Fig. 9, the compressive strength leading to the fibres micro-buckling can be estimated through different approaches. These approaches aim at comparing the properties of flax stems with unidirectional composites, rather than providing a micromechanical model to predict the compressive strength of flax bundles.
Fig. 8. Typical kinking failure in a flax-epoxy after a compressive load. 31
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
Fig. 9. Compressive strength as a function of fibre volume fraction of unidirectional flax/epoxy composite.
First approach
Table 6 Estimated compressive strength of flax fibres from unidirectional composite tests and stem bending tests.
Fibre compressive strength (σfL C) can be estimated from the compressive strength of the unidirectional composite (σUD L C) using the following relationship (Kozez et al., 1995):
σfL C =
σUD L C Vf
Compressive strength with fibre buckling (MPa) UD first approach UD second approach Stem Bolchoï Stem Eden Stem Ariane Stem Liral Prince
(7)
The simple hypothesis is that the compressive strength is linearly proportional to the fibre volume fraction (Vf). This hypothesis is validated for glass fibres (Lo and Chim, 1992). With the set of results of this article, the longitudinal compressive strength of flax fibres is estimated to 260 ± 59 MPa (average of the measured values by taking all the specimens).
In a second approach, the contribution of the matrix is considered. In this case, the fibre buckling strain is much lower than the strain at break of the epoxy. The longitudinal compressive strength of a unidirectional (σUD L C) can be approached using a mixture law: (8)
With (Vf) the fibre volume fraction, (σfL C) the compressive strength causing the fibre buckling and (σm*) the stress in the matrix during the fibres buckling. Considering an elastic and linear behaviour, the Eq. (8) becomes:
σUD L C = Vf *σfL C + (1 − Vf )*εfLC *Em
(9)
With (Em) the matrix modulus (here Em = 3 GPa) and (εfLC) the strain causing fibre buckling. This last value is determined by bending tests on stems (0.4% ± 0.05). The longitudinal compressive strength of flax fibres can be estimated by Eq. (10):
σfL C =
59 44 26 8 16 15
Is the increase of the compressive strength of UD composites reinforced by flax fibres possible? By impregnating the cell walls with a melamine formaldehyde resin (MF) (there is penetration of MF into the cell walls) before impregnation with an epoxy resin, Bos et al. (2004) show that the compressive strength at break of composites increases, but the tensile strength drops down. Furthermore, the use of vegetable fibres is often justified to reduce environmental impacts, and a life-cycle analysis will be necessary to know the consequences of this cell wall treatment.
σUD L C − (1 − Vf )*εfLC *Em Vf
± ± ± ± ± ±
regarding UD (260 MPa and 242 MPa respectively), and a difference of less than 15% between average values of flax varieties (from 183 MPa to 207 MPa for Liral Prince and Eden, respectively). The estimated strength is slightly lower for the old varieties (Liral Prince and Ariane) due to their low fibre content. The poor compression strength of flax fibres in a unidirectional composite is also observed within the stems. This is a limit for the use of flax for the reinforcement of a polymer, as in the case of organic fibres (Kozez et al., 1995). A comparison with wood is also possible. In fact, wood is a composite material composed of an assembly of fibres mainly oriented along the axis of primary growth. The mechanism leading to the compressive break of wood is a kink banding phenomenon (Poulsen et al., 1997), such as observed in unidirectional composites reinforced by flax. The compressive behaviour of wood is linearly proportional to the specific gravity (Domone and Illston, 2010). This relationship allows the modelling of structures, and remains valid whether the wood is green or air dried. Thus, (Gibson, 2012a) estimated the axial compressive strength of a wood cell wall as 120 MPa, which corresponds to a densified wood with no lumen. Greater values are obtained for flax, mainly due to higher cellulose content and lower microfibrillar angle.
Second approach
σUD L C = Vf *σfL C + (1 − Vf )*σm*
260 242 206 207 189 183
(10)
With the set of results of this article, the longitudinal compressive strength of flax fibres is estimated to 242 MPa ± 44 MPa. For comparison, Table 6 shows the estimated compressive strength of flax fibres from unidirectional composite tests and stem bending tests. The estimated values (Table 6) are of the same order of magnitude, with a difference of less than 10% between the two approaches 32
Industrial Crops & Products 130 (2019) 25–33
C. Baley et al.
5. Conclusion
and Use of Synthetic Materials for Aircraft Construction. Flight 77–79. Delaplace, A., Gatuingt, F., Ragueneau, F., 2008. MÉCANIQUE DES STRUCTURES Résistance des matériaux. Dunod. Deteresa, S.J., Porter, R.S., Farris, R.J., 1988. Experimental verification of a microbuckling model for the axial compressive failure of high performance polymer fibres. J. Mater. Sci. 23, 1886–1894. https://doi.org/10.1007/BF01115735. Domone, P., Illston, J., 2010. Construction Materials their Nature and Behaviour, fourth edition. Spon Press. Evert, R., 2006. Esau’s Plant Anatomy, third edition. John Wiley & Sons. Evert, R., Eichhorn, S., 2013. Raven Biology of Plants. W. H. Freeman and Company. Evon, P., Barthod-Malat, B., Grégoire, M., Vaca-Medina, G., Labonne, L., Ballas, S., Véronèse, T., Ouagne, P., 2018. Production of fiberboards from shives collected after continuous fiber mechanical extraction from oleaginous flax. J. Nat. Fibers 0, 1–17. https://doi.org/10.1080/15440478.2017.1423264. Garland, B.D., Beyerlein, I.J., Schadler, L.S., 2001. The development of compression damage zones in fibrous composites. Compos. Sci. Technol. 61, 2461–2480. https:// doi.org/10.1016/S0266-3538(01)00176-2. Gibaud, M., Bourmaud, A., Baley, C., 2015. Understanding the lodging stability of green flax stems; the importance of morphology and fiber stiffness. Biosyst. Eng. 137, 9–21. https://doi.org/10.1016/j.biosystemseng.2015.06.005. Gibson, L.J., 2012a. The hierarchical structure and mechanics of plant materials. J. R. Soc. Interface 9 (76). Gibson, R., 2012b. Principles of Composites Materials Mechanics. Taylor & Francis. Gorshkova, T., Gurjanov, O., Mikshina, P., Ibragimova, N., Mokshina, N., Salnikov, V., Ageeva, M., Amenitskii, S., Chernova, T., Chemikosova, S., 2010. Specific type of secondary cell wall formed by plant fibers. Russ. J. Plant Physiol. 57, 328–341. https://doi.org/10.1134/s1021443710030040. Goudenhooft, C., Bourmaud, A., Baley, C., 2017. Varietal selection of flax over time: evolution of plant architecture related to influence on the mechanical properties of fibers. Ind. Crops Prod. 97, 56–64. https://doi.org/10.1016/j.indcrop.2016.11.062. His, I., Andème-Onzighi, C.C., Morvan, C., Driouich, A., 2001. Microscopic studies on mature flax fibers embedded in LR white: immunogold localization of cell wall matrix polysaccharides. J. Histochem. Cytochem. 49, 1–11. Hodgkinson, J., 2000. Mechanical Testing of Advanced Fibre Composites. Woodhead Publishing Ltd. and CRC Press LLC. Khalfallah, M., Abbès, B., Abbès, F., Guo, Y.Q., Marcel, V., Duval, A., Vanfleteren, F., Rousseau, F., 2014. Innovative flax tapes reinforced Acrodur biocomposites: a new alternative for automotive applications. Mater. Des. 64, 116–126. https://doi.org/10. 1016/j.matdes.2014.07.029. Kozez, V., Jiang, H., Mehta, V., Kumar, S., 1995. Compressive behavior of materials: part II. High performance fibers. Mater. Res. Soc. 10, 1044–1061. Le Duigou, A., Kervoelen, A., Le Grand, A., Nardin, M., Baley, C., 2014. Interfacial properties of flax fibre–epoxy resin systems: existence of a complex interphase. Compos. Sci. Technol. 100, 152–157. https://doi.org/10.1016/j.compscitech.2014. 06.009. Leal, A., Deitzel, J., Gillespie, J., 2007. Assessment of compressive properties of high performance organic fibers. Compos. Sci. Technol. 67 (67), 2786–2794. Leblicq, T., Vanmaercke, S., Ramon, H., Saeys, W., 2015. Mechanical analysis of the bending behaviour of plant stems. Biosyst. Eng. 129, 87–99. https://doi.org/10. 1016/j.biosystemseng.2014.09.016. Lefeuvre, A., Bourmaud, A., Baley, C., 2015. Optimization of the mechanical performance of UD flax/epoxy composites by selection of fibres along the stem. Compos. Part A: Appl. Sci. Manuf. 77, 204–208. https://doi.org/10.1016/j.compositesa.2015.07.009. Liang, S., Gning, P.-B., Guillaumat, L., 2015. Quasi-static behaviour and damage assessment of flax/epoxy composites. Mater. Des. 67, 344–353. https://doi.org/10.1016/j. matdes.2014.11.048. Livingston Smith, S., Mech, M., 1945. A survey of plastics from the viewpoint of the mechanical engineer. Inst. Mech. Eng. 29–43. Lo, K.H., Chim, E., 1992. Compressive strength of unidirectional composites. J. Reinf. Plast. Compos. 11, 838–896. Mahboob, Z., El Sawi, I., Zdero, R., Fawaz, Z., Bougherara, H., 2017. Tensile and compressive damaged response in flax fibre reinforced epoxy composites. Compos. Part A: Appl. Sci. Manuf. 92, 118–133. https://doi.org/10.1016/j.compositesa.2016.11.007. Mattheck, C., 1995. Biomechanical optimum in woody systems. In: Gartner, B. (Ed.), Plant Stems: Physiology and Functional Morphology. Academic Press Limited, London, UK, pp. 27–52. Poulsen, J.S., Moran, P.M., Shih, C.F., Byskov, E., 1997. Kink band initiation and band broadening in clear wood under compressive loading. Mech. Mater. 25, 67–77. https://doi.org/10.1016/S0167-6636(96)00043-9. Réquilé, S., Goudenhooft, C., Bourmaud, A., Le Duigou, A., Baley, C., 2018. Exploring the link between flexural behaviour of hemp and flax stems and fibre stiffness. Ind. Crops Prod. 113. https://doi.org/10.1016/j.indcrop.2018.01.035. Rihouey, C., Paynel, F., Gorshkova, T., Morvan, C., 2017. Flax cellulosic fibres: how to assess the non-cellulosic polysaccharides and to approach cell wall supramolecular design. Cellulose 24, 1985–2001. Roussière, F., Baley, C., Godard, G., Burr, D., 2012. Compressive and tensile behaviours of PLLA matrix composites reinforced with randomly dispersed flax fibres. Appl. Compos. Mater. 19, 171–188. https://doi.org/10.1007/s10443-011-9189-8. Snegireva, A.V., Ageeva, M.V., Amenitskii, S.I., Chernova, T.E., Ebskamp, M., Gorshkova, T.A., 2010. Intrusive growth of sclerenchyma fibers. Russ. J. Plant Physiol. 57, 342–355. https://doi.org/10.1134/s1021443710030052. Timoshenko, S.P., 1947. Strength of Materials - Part 2 - Advanced Theory and Problems. 9th printing. van, Nostrand. . Van Vuure, A.W., Baets, J., Wouters, K., Hendrickx, K., 2015. Compressive properties of natural fibre composites. Mater. Lett. 149, 138–140. https://doi.org/10.1016/j. matlet.2015.01.158. Zaune, M., Stampanoni, M., Niemz, P., 2016. Failure and failure mechanisms of wood during longitudinal compression monitored by synchrotron micro-computed tomography. Holzforschung 70 (2). https://doi.org/10.1515/hf-2014-0225.
This work illustrated the compression buckling mechanisms of flax fibre bundles within a stem. The analysis of the measured values made it possible to estimate the stress causing bundle instability within a flax stem. The stem organisation influences the behaviour; i.e. the compression strength of the bundles was slightly lower for the older varieties, these latter having a smaller fraction of fibres. In addition, thanks to longitudinal compressive tests on unidirectional laminates, the limited reinforcement of flax fibres was illustrated. By an inverse method, the theoretical compressive strength of flax fibres was estimated. Both approaches led to the same orders of magnitude. Despite an optimal arrangement of the fibres in the stems and a strong fibre cohesion, the fibre compression strength remains limited. This assessment is a limit of the use of flax fibres as polymer reinforcement, especially when significant compressive properties are required. The central part of the stem (woody core) exhibits an apparent strain higher in compression despite the presence of large lumens, which can be explained by its conductive function. This core tissue significantly contributes to the compressive strength of the stem as in a structure sandwich. To go further, future works will be carried out on living plants for a better understanding of their strengthening mechanisms. Acknowledgements The authors would like to thank Dr. Marine Lan for part of the experimental tests, the French Ministry of Research and Innovating Technologies, OSEO, Région Bretagne for their financial support. The authors also acknowledge the 3D-BioMat project that is co-financed by the Grand Reims (31%) and the European Union by 48.7% (i.e. 50% of eligible expenditure). Europe is committed to the Grand Est with the European Regional Development Fund. The 3D-BioMat project with a total budget of 965,000€ is hosted by the Centre Européen de Biotechnologie et de Bioéconomie (CEBB 51110 Pomacle, France) for a three-year period (from 01/05/2016 to 30/04/2019). References Andème-Onzighi, C., Girault, R., His, I., Morvan, C., Driouich, A., 2000. Immunocytochemical characterization of early-developing flax fiber cell walls. Protoplasma 213, 235–245. https://doi.org/10.1007/bf01282161. Attwood, J.P., Fleck, N.A., Wadley, H.N.G., Deshpande, V.S., 2015. The compressive response of ultra-high molecular weight polyethylene fibres and composites. Int. J. Solids Struct. 71, 141–155. https://doi.org/10.1016/j.ijsolstr.2015.06.015. Baley, C., 2002. Analysis of the flax fibres tensile behaviour and analysis of the tensile stiffness increase. Compos. Part A: Appl. Sci. Manuf. 33, 939–948. Baley, C., Bourmaud, A., 2014. Average tensile properties of French elementary flax fibers. Mater. Lett. 122, 159–161. https://doi.org/10.1016/j.matlet.2014.02.030. Baley, C., Le Duigou, A., Bourmaud, A., Davies, P., Nardin, M., Morvan, C., Le Duigou, A., Bourmaud, A., Davies, P., Nardin, M., Morvan, C., 2014. Reinforcement of polymers by flax fibers: role of interfaces. Bio-Based Composites for High-Performance Materials. CRC Press, pp. 87–112. https://doi.org/10.1201/b17601-7. Baley, C., Goudenhooft, C., Gibaud, M., Bourmaud, A., 2018a. Flax stems: from a specific architecture to an instructive model for bioinspired composite structures. Bioinspir. Biomim. 13, 026007. https://doi.org/10.1088/1748-3190/aaa6b7. Baley, C., Lan, M., Bourmaud, A., Le Duigou, A., 2018b. Compressive and tensile behaviour of unidirectional composites reinforced by natural fibres: influence of fibres (flax and jute), matrix and fibre volume fraction. Mater. Today Commun. 16, 300–306. https://doi.org/10.1016/j.mtcomm.2018.07.003. Benabou, L., 2008. Kink band formation in wood species under compressive loading. Exp. Mech. 48, 647–656. https://doi.org/10.1007/s11340-007-9098-9. Bert, F., 2013. Lin Fibre: culture et transformation. Arvalis. Bos, H.L., Molenveld, K., Teunissen, W., van Wingerde, A.M., van Delft, D.R.V., 2004. Compressive behaviour of unidirectional flax fibre reinforced composites. J. Mater. Sci. 39, 2159–2168. https://doi.org/10.1023/b:jmsc.0000017779.08041.49. Bourmaud, A., Baley, C., 2009. Rigidity analysis of polypropylene/vegetal fibre composites after recycling. Polym. Degrad. Stab. 94, 297–305. https://doi.org/10.1016/j. polymdegradstab.2008.12.010. Brazier, L., 1927. On the flexure of thin cylindrical shells and other ‘Thin’ sections. Proc. R. Soc. A: Math. Phys. Eng. Sci. 116 (773), 104–114. Catling, D., Grayson, J., 1982. Identification of Vegetable Fibres. Archetype Press, London. Charlet, K., Jernot, J.P., Gomina, M., Bréard, J., Morvan, C., Baley, C., 2009. Influence of an Agatha flax fibre location in a stem on its mechanical, chemical and morphological properties. Compos. Sci. Technol. 69, 1399–1403. de Bruyne, N.A., 1939. Plastic Progress - Some Further Developments in the Manufacture
33