Thygesen et al. 2011 Cellulosic fibres

This article was downloaded by: [DTU Library]On: 02 September 2011, At: 01:17Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Natural FibersPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/wjnf20

Cellulosic Fibers: Effect of Processing onFiber Bundle StrengthAnders Thygesen a , Bo Madsen b , Anne Belinda Bjerre a & HansLilholt ba Biosystems Division, Risø National Laboratory for SustainableEnergy, Technical University of Denmark, Roskilde, Denmarkb Materials Research Division, Risø National Laboratory forSustainable Energy, Technical University of Denmark, Roskilde,Denmark

Available online: 01 Sep 2011

To cite this article: Anders Thygesen, Bo Madsen, Anne Belinda Bjerre & Hans Lilholt (2011):Cellulosic Fibers: Effect of Processing on Fiber Bundle Strength, Journal of Natural Fibers, 8:3,161-175

To link to this article: http://dx.doi.org/10.1080/15440478.2011.602236

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching and private study purposes. Anysubstantial or systematic reproduction, re-distribution, re-selling, loan, sub-licensing,systematic supply or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand or costs or damages whatsoever or howsoever caused arising directly or indirectlyin connection with or arising out of the use of this material.

http://www.tandfonline.com/loi/wjnf20

http://dx.doi.org/10.1080/15440478.2011.602236

http://www.tandfonline.com/page/terms-and-conditions

Journal of Natural Fibers, 8:161–175, 2011Copyright © Taylor & Francis Group, LLCISSN: 1544-0478 print/1544-046X onlineDOI: 10.1080/15440478.2011.602236

Cellulosic Fibers: Effect of Processing on FiberBundle Strength

ANDERS THYGESEN,1 BO MADSEN,2 ANNE BELINDA BJERRE,1

and HANS LILHOLT2

1Biosystems Division, Risø National Laboratory for Sustainable Energy, Technical Universityof Denmark, Roskilde, Denmark

2Materials Research Division, Risø National Laboratory for Sustainable Energy, TechnicalUniversity of Denmark, Roskilde, Denmark

A range of differently processed cellulosic fibers from flax andhemp plants were investigated to study the relation between pro-cessing of cellulosic fibers and fiber bundle strength. The studiedprocessing methods are applied for yarn production and includeretting, scutching, carding, and cottonization. There was a mono-tonically decreasing relationship between the strength and thenumber of processing steps, which was well fitted by an expo-nential regression line. The reduction factor was determined tobe 0.27, indicating that the fiber bundle strength was on averagereduced by 27% per processing step at the applied conditions. Nolarge changes in cellulose content and crystallinity were observed,so the reduction in strength must be explained by other changesin the fiber ultrastructure. Altogether, the study presents a quan-titative basis for reduction in strength of cellulosic fibers due toprocessing.

KEYWORDS flax, hemp, cellulose, tensile strength

INTRODUCTION

Cellulosic fibers are of interest for reinforcement of composite materials dueto their good mechanical properties, low density, and renewability (Mohantyet al. 2005). Knowledge about their ultrastructure, chemical composition, and

Address correspondence to Anders Thygesen, Biosystems Division, Risø NationalLaboratory for Sustainable Energy, Technical University of Denmark, P.O. Box 49, DK-4000Roskilde, Denmark. E-mail: [email protected]

161

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11

162 A. Thygesen et al.

mechanical properties are therefore of importance. The cellulose functionsas the reinforcing component in the fibers, as it is organized as alignedcrystallinemicrofibrils embedded in lignin andhemicellulose. Interesting typesof cellulosic fibers include flax (Linum usitatissimum) and hemp (Cannabissativa L.) due to their high cellulose content in the range 55%–70% (Baley 2002;Thygesen et al. 2007) and their related high strength in the range 300–1800MPa (Lilholt and Lawther 2000; Sankari 2000; Baley 2002).

The general expectation is that processing of materials leads to disor-dered regions in the normally ordered materials structure (Cousins 1976,1978; Salmén 2004). Such regions of disorder are often called defects to indi-cate that they have a detrimental effect on mechanical properties. A numberof studies have been performed with microscopic techniques for quantifica-tion of defects (or dislocations) in cellulosic fibers (Thygesen and Hoffmeyer2005; Thygesen et al. 2006), but the detailed correlation between amount ofdefects and fiber strength is still to be investigated.

Usage of cellulosic fibers in high-performance composite materials oftenrequires orientated fibers, which can be obtained from yarn (Thygesen et al.2007). This requires processing of the fibers into yarn. The initial rettingstep involves microbial degradation of pectin, which might result in chem-ically induced defects in the fibers (Meijer et al. 1995). After retting of theplant stems, the shives are broken into small pieces by scutching and thefiber bundles are separated from the shives (Hobson et al. 2001). The fiberbundles are then carded using brushes to straighten the fibers. Finally, thefiber bundles are made suitable for spinning by cottonization, in which thelarger fiber bundles are split up into smaller ones. The processes, scutching,carding, and cottonization, might result in mechanically induced defects.

The purpose of this study is to investigate the correlation between pro-cessing and fiber strength for a range of differently processed cellulosicfibers. The number of processing steps is used as an independent experi-mental variable to quantify the effect of processing. Ideally, the experimentsin this study should be performed with individual cellulosic fibers in orderto investigate the fiber strength relation to processing but it is not practi-cal to use and test single fibers; therefore, more practically convenient fiberbundles were used. In addition to the fiber bundle strength, the chemicalcomposition and the cellulose crystallinity of the fibers are measured.

MATERIALS AND METHODS

Fibers

The investigated flax and hemp fibers are presented in Table 1. The sam-ples were sorted according to number of processing steps. The four samplesof flax fibers were supplied from Ekotex (Poland). These flax fiber sam-ples (F5, F1, F2, and F3) were processed by dew retting, scutching, and

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11

Processing Effect on Cellulosic Fiber Strength 163

TABLE 1 Overview of the Investigated Flax and Hemp Fibers. The Fibers are SortedAccording to the Number of Processing Steps

Fiber Fiber density Processingtype Code Form Suppliers1 (g/cm3) steps2 (N )

Flax F5 Sliver Ekotex 1.56 ± 0.01 3: Dr, Sc, CaF1 Loose fibers Ekotex 1.56 ± 0.01 4: Dr, Sc, Ca, CoF2 Loose fibers Ekotex 1.53 ± 0.04 4: Dr, Sc, Ca, CoF3 Loose fibers Ekotex 1.56 ± 0.02 5: Dr, Sc, Ca, Co, Co

Hemp H5 Loose fibers DJF 1.58 ± 0.01 0H7 Loose fibers DJF 1: WrH8 Loose fibers DJF 1: CrH9 Loose fibers DJF 1.58 ± 0.01 1: PrH3 Loose fibers BaFa 1.59 ± 0.03 2: Dr, ScH2 Loose fibers BaFa 1.55 ± 0.02 3: Dr, Sc, ScH4 Sliver Ekotex 3: Dr, Sc, Ca

Notes: 1Suppliers: Ekotex (Poland); DJF (Faculty of Agricultural Sciences, University of Aarhus, Denmark);BaFa (Badische Naturfaseraufbereitung, Germany).2Processing steps: Dr = dew retting, Cr = fungal retting with C. subvermispora, Pr = fungal retting withP. radiata Cel 26, Wr = water retting, Sc = scutching, Ca = carding, Co = cottonization.

carding. In addition, mechanical cottonization was performed on F1, F2,and F3 fiber samples, where F3 was cottonized twice. In the applied cot-tonization method, the fibers were mechanically broken and cut to obtainshorter fibers (2–5 cm). Seven samples of hemp fibers were investigated.The H5, H7, H8, and H9 fiber samples were produced from hemp plantsgrown at Flakkebjerg Research Station, University of Aarhus, Denmark. H5fibers were manually peeled from the un-retted stems, i.e., these fibers wereconsidered not processed. H7, H8, and H9 fibers were retted by three meth-ods: water retting (H7), fungal retting by Ceriporiopsis subvermispora (H8),and fungal retting by Phlebia radiata Cel 26 (H9) (Thygesen et al. 2007).The H3, H2, and H4 fiber samples were supplied by BaFa (Germany) andEkotex (Poland). These samples were processed by dew retting and scutch-ing. Extra scutching was performed to obtain H2. Carding was performed toobtain H4, and therefore these fibers are in the form of a sliver.

Chemical Composition and Cellulose Crystallinity

A gravimetric method was used to determine the contents of wax, water-soluble components, pectin, lignin, hemicellulose, and cellulose in thecellulosic fibers (Browning 1967; Thygesen et al. 2005b). The method con-sists of five steps: (1) Wax was soxlet extracted in chloroform, (2) waterextractives were extracted in water, (3) pectin was extracted in 30 g/L etylen-diamin-tetra-acetic acid, (4) lignin was oxidized and extracted in 33 g/LNaClO2 + 6 g/L CH3COOH, and (5) hemicellulose was extracted in 120 g/LNaOH + 20 g/L H3BO3. The remaining residue, which is unaffected by

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


the extraction steps, is cellulose. The mineral content was determined byincineration of 0.5g sample at 550◦C for 3 h.

Determination of crystallinity of the fibers was done by X-ray pow-der diffraction. The raw fibers were milled in a knife mill and side loadedinto the sample holder to get random fiber orientation. The diffractogramswere recorded in reflection mode with a Philips PW1820/3711 diffractometerwith Bragg-Brentano scattering geometry, copper Kα-radiation and automaticdivergence slits. Data were collected in the 2θ -range 5◦–60◦ with a step sizeof 0.02◦ and a counting time of 20 s per step (Thygesen et al. 2005b).The diffractograms were divided into crystalline diffraction and amorphousbackground diffraction with the Rietveld refinement method (Rietveld 1969).The method used the atom coordinates for cellulose Iβ with data fromNishiyama et al. (2002). The sample crystallinity was calculated assumingthat the background comprised the amorphous contribution to the diffrac-togram. Cellulose crystallinity (Xcel) was defined as the crystalline part ofthe cellulose, and it was calculated from the measured sample crystallinity(X sample) and cellulose content (W cel) as Xcel = X sample/W cel.

Strength of Fiber Bundles

Tensile strength tests of fiber bundles were done with a gauge length of3.0 mm and strain rates of 0.167 × 10−3/s and 1.67 × 10−3/s using anInstron 5566 with pressley clamps (Stelometer 654 from Zellweger Uster)(Thygesen et al. 2005a). Based on previous measurements, it was found thatthe two strain rates resulted in similar strength values (unpublished data).The temperature was controlled at 23 ± 2◦C and the relative humidity wascontrolled at 50% ± 10%. Following fracture of a fiber bundle, the two pieceswere weighed (wf). The length of the tested fiber bundles (lf) (clampinglength + gauge length) was 14.8 mm. The cross-sectional area of the fiberbundles (S) was thereby determined:

S (mm2) = wf (mg)

ρf (g/cm3) · lf (mm), (1)

where ρ f is the fiber density, which was measured by pycnometry (Pratten1981) for the different fiber types (see Table 1). The fiber bundle strength(σb) was calculated as σb = F/S, where F is the fracture force. Fiber bundleswith cross-sectional areas within the range 0.005–0.4 mm2 were tested.

RESULTS AND DISCUSSION

Chemical Composition and Crystallinity of Fibers

The chemical composition of the investigated flax and hemp fibers is shownin Table 2. Overall, the fibers contained 64%–78% cellulose, 12%–17%

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


TABLE 2 Chemical Composition and Cellulose Crystallinity of the Investigated Flax and HempFibers

Chemical composition

Fiber Cellulose Hemicellulose Lignin Pectin Residuals Crystallinitycode (% w/w) (% w/w) (% w/w) (% w/w) (% w/w) (% w/w)

F5 68.7 12.9 4.7 4.0 9.7 96.9F1 73.1 12.2 4.5 2.3 7.9 95.5F2 75.5 14.0 1.7 3.5 5.2 88.3F3 74.5 13.2 2.1 2.2 8.0 100.0H5 64.4 14.4 3.4 7.3 10.5 93.8H7 73.8 11.8 5.0 2.5 6.9 92.7H8 72.4 13.8 4.9 4.0 4.9 86.9H9 78.4 13.0 2.4 1.9 4.3 84.1H3 65.9 14.9 4.6 4.0 10.6 98.2H2 63.7 17.3 4.1 4.2 10.7 100.0H4 73.4 14.6 4.3 2.4 5.3 90.7

100

Cel

lulo

se c

onte

nt [%

w/w

]

80

60

40

20

00 1 2

Processing steps N3 4 5

Flax fibresHemp fibres

FIGURE 1 Cellulose content versus number of processing steps for flax and hemp fibers.

hemicellulose, 2%–5% lignin, and 2–7% pectin. The average cellulose con-tent for all the cellulosic fibers was 71% ± 5%. Figure 1 shows the cellulosecontent as a function of the number of processing steps N . As assessed fromFigure 1, the cellulose content was neither consistently affected by fiber typenor by processing as based on the low slope coefficient of 0.7 ± 1.0 % ver-sus N and the low correlation coefficient R2 of 0.05. The average cellulosecontents were 70% ± 6% for the hemp fibers and 73% ± 3% for flax fibers.

The cellulose crystallinity of the investigated flax and hemp fibers isshown in Table 2. Figure 2 shows the cellulose crystallinity as a function ofnumber of processing steps. The cellulose structure matched the monoclinicunit cell of cellulose Iβ containing two cellobiose chains. For both flax andhemp fibers, the unit cell dimensions was in the range of 7.87–7.94 Å in the

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


100

Cel

lulo

se c

ryst

alin

ity [%

w/w

]

80

60

40

20

00 1 2


Flax fibresHemp fibres

FIGURE 2 Cellulose crystallinity versus number of processing steps for flax and hemp fibers.

a-direction, and in the range of 8.19–8.27 Å in the b-direction perpendicularto the chain axis. These values agree well with previous studies of cellulose(Sugiyama et al. 1990; Horii 2001; Nishiyama et al. 2002). The average cel-lulose crystallinity for all the cellulosic fibers was 93% ± 5%. As assessedfrom Figure 2, the crystallinity of cellulose was neither consistently affectedby fiber type nor by processing as based on the low slope coefficient of1.5% ± 1.1% versus N and the low correlation coefficient R2 of 0.19. Theaverage cellulose crystallinity was 92% ± 6% for hemp fibers and 95% ± 5%for flax fibers.

Harsh mechanical processes, such as ball-milling, are known to reducethe cellulose crystallinity of plant fibers (Howsmon and Marchessault 1959),which indicates that the cellulose microfibril structure is getting disordered.Chemical and physical processes can increase the cellulose content in plantfibers by extraction of hemicellulose, pectin, and lignin (Nykter et al. 2008).The fibers in the present study were exposed to mechanical processes only(after the initial retting process). It was found that these processes did notsignificantly change the chemical composition and cellulose crystallinity ofthe fibers.

Strength of Fiber Bundles

The fiber bundle tensile strength σb for flax and hemp fiber samples as afunction of the cross-sectional area S is presented in Figure 3. Plots of thedata showed that σb decreased against S. This trend is general for materialswith defects, and is normally analyzed statistically with the Weibull-model(Weibull 1939; Zweben 1995), leading to the following relation for strengthversus test-volume:

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


Fib

re b

undl

e st

reng

th σ

b [M

Pa]

Fib

re b

undl

e st

reng

th σ

b [M

Pa]

Fib

re b

undl

e st

reng

th σ

b [M

Pa]

(a)

(b)

(c)

F5F1F2F3

H5

H3H2H4

H7H8H9

1000

800

600

400

200

1200

0

1000

800

600

400

200

1200

0

1000

800

600

400

200

1200

00.0 0.1 0.2 0.3 0.4 0.5

Cross sectional area S [mm2]

FIGURE 3 Fiber bundle strength versus cross-sectional area for four flax fiber samples andseven hemp fiber samples. Each data point represents one repetition of failure strength mea-surement. The curves were established by power law regression of bundle strength versusarea for each sample.

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


TABLE 3 Fiber Bundle Strength of the Investigated Flax and Hemp Fibers. The Power LawConstants σb0 and α are used to fit the Relationship Between Fiber Bundle Strength and Cross-Sectional Area of the Fibers (Equation 2). The Fiber Bundle Strength (σb) is Calculated for aCross-Sectional Area of 0.1 mm2

Sample σb0 for S0 = 1 mm2 (MPa) α σb for S = 0.10 mm2 (MPa)

F5 331 ± 45 −0.16 482 ± 24F1 160 ± 19 −0.32 334 ± 18F2 176 ± 27 −0.29 342 ± 25F3 118 ± 29 −0.30 238 ± 28H5 1053 ± 260 0.02 998 ± 45H7 267 ± 166 −0.37 629 ± 75H8 470 ± 176 −0.23 798 ± 58H9 801 ± 115 −0.01 812 ± 42H3 399 ± 87 −0.16 581 ± 44H2 286 ± 65 −0.14 396 ± 30H4 384 ± 49 −0.12 503 ± 24

σb = σb0 ·(

V

V0

)α

= σb0 ·(

S

S0

)α

, (2)

which, since the gauge length was constant at 3 mm, could be written interms of the cross-sectional area, where S0 is a reference area of 1 mm2, σb0 isthe fiber bundle strength at the reference area, and α is the slope coefficientdetermined in Figure 3. In accordance with Equation 2, power law regres-sion lines were fitted to the experimental data and the related constantsα and σb0 are presented in Table 3. The fiber bundle strength decreasedagainst cross-sectional area for both flax and hemp fibers (with only twocases, out of 11, of nearly constant strength). The average of the α-valueswas −0.19 with a standard deviation of 0.12. Decrease in fiber strength ver-sus cross-sectional area has been also found in tests on single flax fibers(Baley 2004).

For correlating the fiber strength and processing steps, it was decidedto analyze σb at a fixed cross-sectional area of 0.1 mm2. For flax fibers, thestrength of bundles at S = 0.1 mm2 was 482 MPa for the un-cottonized F5,338 MPa for the one-step cottonized F1 and F2, and 238 MPa for the two-stepcottonized F3. Thus, the fiber bundle strength was consistently decreased bycottonization. For hemp fibers, it could also be observed that processingresulted in lower fiber bundle strength. The strongest hemp fibers were theuntreated H5 with fiber bundle strength of 998 MPa. The retting processesresulted in reduction of strength to 798 MPa for H8, 812 MPa for H9, and629 MPa for H7 fibers. The dew retted and scutched H3 and H2 fibers hadfiber bundle strength of 581 and 396 MPa, respectively, where the lowerstrength of H2 fibers can be related to the two times of scutching. Thestrength was 503 MPa in the dew retted, scutched, and carded H4 fibers.

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


1000

Fib

re b

undl

e st

reng

th [M

Pa]

800

600

400

200

00 1 2

Flax fibresHemp fibresExponential reg.


FIGURE 4 Fiber bundle strength versus number of processing steps for flax and hemp fibersdetermined from the curves in Figure 3 at a cross-sectional area of 0.1 mm2.

By comparing the strength of the H4 and the H2 fibers, it was indicatedthat carding is resulting in less-strength reduction than scutching. The cor-relation between the effective fiber bundle strength (σb) and the number ofprocessing steps (N ) is presented in Figure 4. This figure showed that thedata follow the same trend for hemp and flax fibers with no clear differencebetween the two types of cellulosic fibers. All the data were consideredtogether and fitted by an exponential curve.

Figure 5a shows the established values of fiber bundle strength of thefiber samples as a function of their cellulose content by volume. The mea-sured cellulose contents by weight (Table 2) were converted to contents byvolume by use of the density for crystalline cellulose (1.62 g/cm3) and thedensity of the residual compounds lignin and hemicellulose (1.43 g/cm3).In Figure 5a, the fiber samples were sorted in groups according to the num-ber of processing steps. More than one data point was available for thegroups of 1, 3, and 4 processing steps. The data in these groups indicatedthat the fiber bundle strength is increased when the cellulose content isincreased. A similar increase in fiber strength versus cellulose content hasbeen previously shown for cellulosic fibers in composite materials (Thygesenet al. 2007). The establishment of linear relationships in Figure 5a was moti-vated by the law of mixtures for the strength (versus fiber volume fraction)of a composite with (long) aligned fibers in a (weaker) matrix, under theassumption of equal strains in fibers and matrix as used by Thygesen et al.(2007). The regression lines are not well determined individually; however,since they are related by the number of processing steps, they supporteach other.

This iso-strain model was used for the cellulosic fibers consisting ofaligned cellulose microfibrils in a weak matrix of hemicellulose and lignin.

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


1500

(a)

(b)

0 steps1 step2 steps3 steps4 steps5 stepsRegression lines

Experimental lines (0–5 steps)Keller et al. (2001)Regression line (Keller)

1000

500

Fib

re b

undl

e st

reng

th σ

b [M

Pa]

Fib

re b

undl

e st

reng

th σ

b [M

Pa]

0

1500

1000

500

00.0 0.2

Cellulose volume fraction Vcel

0.4 0.6 0.8 1.0

FIGURE 5 Fiber bundle strength versus cellulose volume fraction. In (a) the strength valuesfor flax and hemp were analysed in combination and sorted according to number of process-ing steps, which appear acceptable according to Figure 4. The lines have a fixed intercept of40 MPa corresponding to an approximated strength of hemicellulose and lignin. The lowergraph (b) shows data points for fiber bundle strength of decorticated fibers from hemp plants(Keller et al. 2001).

This was acceptable, since the cellulosic microfibrils are well aligned in thecell wall of individual fibers, with the main fraction of microfibrils havingmicrofibril angles in the range 0–5◦ from the fiber axis. The establishmentof the linear relationships in Figure 5a requires that the cellulose content isshown by volume fractions, and that the intercept of the lines at a cellulosecontent of zero equals the strength of the non-cellulosic matrix. A value of40 MPa for the (hemicellulose + lignin) matrix was used based on a previous

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


study of natural fiber composites (Thygesen et al. 2007). In addition to thelinear relation with the cellulose content, Figure 5a indicates that the fiberbundle strength was consistently decreased when the number of processingsteps is increased. Thus, it was established that the number of processingsteps is an appropriate sorting criteria.

In a previous study of hemp plants by Keller et al. (2001), the cellulosecontent and the fiber bundle strength of the hemp fibers were investigated.The applied hemp stems were processed five times by scutching to extractthe fibers. The data of fiber bundle strengths (in Figure 4 of Keller et al.(2001)) were measured for fiber bundles with an average cross-sectionalarea of 0.78 mm2. The power law function in Equation 2 was used to cal-culate the corresponding values of fiber bundle strengths at S = 0.1 mm2

using the present α-value of -0.19. These fiber bundle strength values wereplotted against the cellulose volume fraction in Figure 5b. The data of cel-lulose weight fractions (in Figure 8 of Keller et al. 2001) were converted tovolume fractions (V cel) as described previously. It could be observed thatthese data follow the same linear trend between fiber bundle strength andcellulose content as shown in Figure 5a, and that the fitted line agrees wellwith the line established in the present study for the samples exposed toone processing step. Thus, it was demonstrated that the bundle strength ofcellulosic fibers measured in two separate studies could be compared basedon their history of processing steps.

In Figure 5a, at the endpoints of the lines, the effective fiber bundlestrength at 100% cellulose was determined to be 1598, 1015, 900, 683,456, and 315 MPa for the six levels of processing steps, respectively. Thecorrelation between the effective fiber bundle strength (σcel) and the num-ber of processing steps (N ) is presented in Figure 6. The trend line is a fittedexponential line:

σcel = σcel,0 · (1 − r)N , (3)

where σcel,0 is the effective fiber bundle strength of unprocessed fibers, r is aconstant reduction factor per processing step. It was conservatively assumedthat the effect of different processing steps was same, i.e., N was used as anumerical variable. The effective fiber bundle strength of unprocessed fibers(σcel,0) was determined to be 1564 MPa. The constant reduction (r) in theeffective cellulose strength was determined to be 0.27, which means thatσcel in average was reduced by 27% per processing step with the appliedprocessing techniques.

Previously, a comparison between differently processed flax fibers hasbeen performed by Bos et al. (2002). One group of fibers was hand peeled,which is assumed to be a very mild processing method (i.e., no effect onstrength to be expected), and showed strength for single fibers of 1834 MPa.Another group of fibers were scutched and showed strength of 1522 MPa.

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


1500

1000

Effe

ctiv

e fib

re b

undl

e st

reng

th a

t10

0% c

ellu

lose

σce

l [M

Pa]

500

00 1 2


FIGURE 6 The effective fiber bundle strength at 100% cellulose, determined by extrapolationof the curves in Figure 5a, versus number of processing steps. The line was fitted to the datapoints by exponential regression.

These values give a reduction factor of r = 0.17 for the scutching step, andthis is in the same range as the average reduction factor found in the presentstudy for a range of different processing methods. The different mechanicaland chemical processing methods were thereby similar on the point thatdefects were formed in the fibers. The measured chemical composition andcellulose crystallinity of the investigated fibers were shown not to be consis-tently affected by processing, and the nature of the defects must thereforebe identified by other physical and chemical characteristics. These includechemical defects such as delamination between the layers in the cell wall inthe H9 fibers observed by transmission electron microscopy (Thygesen et al.2005a) and mechanical defects such as kink bands (Thygesen et al. 2006).

CONCLUSIONS

A range of differently processed flax and hemp fibers was investigatedfor the relation between processing and fiber bundle strength. Initial mea-surements of the cellulose content and cellulose crystallinity of the fibersshowed that these parameters were approximately constant versus numberof processing steps with average values of 71% and 93%, respectively.

The measured fiber bundle strength was shown to decrease when thecross-sectional area of the bundles was increased. This negative correlationwas fitted with a power law regression line with an average α-value of−0.19. The fiber bundle strength at a fixed bundle area of 0.1 mm2 was usedto analyze the relationship between processing and fiber bundle strength.The established fiber bundle strengths in the range from 200 to 1000 MPa

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


were grouped according to the six levels of processing steps that the fibershave been exposed to. The slope of the established linear relationships inFigure 5 was consistently decreased when the number of processing stepswas increased. The effective fiber bundle strengths at 100% cellulose wereplotted as a function of the number of processing steps and were wellfitted by an exponential regression line. The estimated reduction in effectivestrength for the fiber bundles was 27% per processing step.

Altogether, by using the number of processing steps as a sorting crite-rion, the study presents a quantitative basis for the reduction in strength ofcellulosic fibers due to their processing.

ACKNOWLEDGMENTS

The authors wish to thank Prof. Claus Felby (Faculty of Life Sciences,University of Copenhagen) for usage of the equipment for fiber bundletests, and Dr. Kenny Ståhl (Department of Chemistry, Technical Universityof Denmark) for usage of X-ray diffraction facilities for crystallinity deter-minations. Technical support is acknowledged to Tomas Fernqvist andLise Lotte Berring at the Technical University of Denmark. The EU com-mission is acknowledged for supporting the BIOCOMP project (EU 6thFramework Programme, IP No. 515769: “New Classes of EngineeringComposite Materials from Renewable Resources”).

REFERENCES

Baley, C. 2002. Analysis of the flax fibres tensile behaviour and analysis of the tensilestiffness increase. Composites Part A: Applied Science and Manufacturing 33:939–948.

Baley, C. 2004. Influence of kink bands on the tensile strength of flax fibers. Journalof Materials Science 39: 331–334.

Bos, H. L., M. J. A. Van den Oever, and O. C. J. J. Peters. 2002. Tensile and compres-sive properties of flax fibres for natural fibre reinforced composites. Journal ofMaterials Science 37: 1683–1692.

Browning, B. L. 1967. Methods of wood chemistry. New York: IntersciencePublishers, John Wiley.

Cousins, W. J. 1976. Elastic-modulus of lignin as related to moisture-content. WoodScience and Technology 10: 9–17.

Cousins, W. J. 1978. Youngs modulus of hemicellulose as related to moisture-content. Wood Science and Technology 12: 161–167.

Hobson, R. N., D. G. Hepworth, and D. M. Bruce. 2001. Quality of fibre separatedfrom unretted hemp stems by decortication. Journal of Agricultural EngineeringResearch 78: 153–158.

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


Horii, F. 2001. Structure of cellulose: Recent developments in its characterization. InWood and cellulosic chemistry, ed. D. N. Hon and N. Shiraishi, 83–107. NewYork: Marcel Dekker AG.

Howsmon, J. A., and R. H. Marchessault. 1959. The ball-milling of cellulosefibers and recrystallization effects. Journal of Applied Polymer Science 1:313–322.

Keller, A., M. Leupin, V. Mediavilla, and E. Wintermantel. 2001. Influence of thegrowth stage of industrial hemp on chemical and physical properties of thefibres. Industrial Crops and Products 13: 35–48.

Lilholt, H., and J. M. Lawther. 2000. Natural organic fibres. In Comprehensive com-posite materials, ed. A. Kelly and C. Zweben, 303–325. Amsterdam, Netherlands:Elsevier.

Meijer, W. J. M., N. Vertregt, B. Rutgers, and M. van de Waart. 1995. The pectincontent as a measure of the retting and rettability of flax. Industrial Crops andProducts 4: 273–284.

Mohanty, A. K., M. Misra, and L. T. Drzal. 2005. Natural fibres, biopolymers, andbiocomposites. New York: Taylor & Francis.

Nishiyama, Y., P. Langan, and H. Chanzy. 2002. Crystal structure and hydrogen-bonding system in cellulose Iá from synchrotron X-ray and neutron fiberdiffraction. Journal of the American Chemical Society 124: 9074–9082.

Nykter, M., H.–Kymäläinen, A. B. Thomsen, H. Lilholt, H. Koponen, A.-M. Sjöberg,and A. Thygesen. 2008. Effects of thermal and enzymatic treatments and har-vesting time on the microbial quality and chemical composition of fibre hemp(Cannabis sativa L.). Biomass & Bioenergy 32: 392–399.

Pratten, N. A. 1981. The precise measurement of the density of small samples.Journal of Materials Science 16: 1737–1747.

Rietveld, H. M. 1969. A profile refinement method for nuclear and magneticstructures. Journal of Applied Crystallography 2: 65–71.

Salmén, L. 2004. Micromechanical understanding of the cell-wall structure. PlantBiology and Pathology 327: 873–880.

Sankari, H. S. 2000. Comparison of bast fibre yield and mechanical fibre proper-ties of hemp (Cannabis sativa L.) cultivars. Industrial Crops and Products 11:73–84.

Sugiyama, J., T. Okano, H. Yamamoto, and F. Horii. 1990. Transformation of Valoniacellulose crystals by an alkaline hydrothermal treatment. Macromolecules 23:3196–3198.

Thygesen, L. G., J. Bilde-Sørensen, and P. Hoffmeyer. 2006. Visualisation of dislo-cations in hemp fibres: A comparison between scanning electron microscopy(SEM) and polarized light microscopy (PLM). Industrial Crops and Products 24:181–185.

Thygesen, A., G. Daniel, H. Lilholt, and A. B. Thomsen. 2005a. Hemp fibermicrostructure and use of fungal defibration to obtain fibers for compositematerials. Journal of Natural Fibers 2: 19–37.

Thygesen, L. G., and P. Hoffmeyer. 2005. Image analysis for the quantifi-cation of dislocations in hemp fibres. Industrial Crops and Products 21:173–184.

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11


Thygesen, A., J. Oddershede, H. Lilholt, A. B. Thomsen, and K. Ståhl. 2005b. On thedetermination of crystallinity and cellulose content in plant fibres. Cellulose 12:563–576.

Thygesen, A., A. B. Thomsen, G. Daniel, and H. Lilholt. 2007. Comparison of com-posites made from fungal defibrated hemp with composites of traditional hempyarn. Industrial Crops and Products 25: 147–159.

Weibull, W. 1939. A statistical theory of the strength of materials. Ingeniørs-ventenskap Akademiens Handlingar 151: 1–47.

Zweben, C. 1995. Simple, design-oriented composite failure criteria incorporatingsize effects. In Proceedings of ICCM-10, Whistler, B.C. Canada, vol 1, 675–682.Boca Raton, FL: CRC Press

Dow

nloa

ded

by [

DT

U L

ibra

ry]

at 0

1:17

02

Sept

embe

r 20

11

Thygesen et al. 2011 Cellulosic fibres

Documents

Transcript of Thygesen et al. 2011 Cellulosic fibres