Poro-microelasticity of wood: the relevance of the microfibril angle

K. Hofstetter, Ch. Hellmich & J. Eberhardsteiner

Vienna University of Technology (TU Wien), Institute for Mechanics of Materials and Structures, Vienna, Austria

ABSTRACT: Wood exhibits an intrinsic structural hierarchy. It is composed of wood cells, which are hollow
tubes oriented in the stem direction. The cell wall is built up by stiff cellulose fibrils which are embedded in a
soft polymer matrix. This structural hierarchy is considered in a four-step homogenization scheme, predicting
the macroscopic elastic behavior of different wood species from tissue-specific chemical composition and
microporosity, based on the elastic properties of nanoscaled universal building blocks. Special attention is paid
to the fact that the fibrils are helically wound in the cell wall at an angle of 0 30 , generally denoted as microfibril
angle. Consideration of this microfibril angle in the continuum micromechanics model for wood is mandatory
for appropriate prediction of macroscopic shear properties. The presented developments can be readily extended
to the prediction of poroelastic properties, such as Biot and Skempton coefficients.

1 INTRODUCTION Longitudinal direction L

Macroscopic mechanical properties of wood are char-
acterized by a wide variability and diversity (Bodig &
Jayne 1982; Kollmann & Ct 1968). This variabil-
ity results from differences observed at the macro-,
micro-, and ultrastructural scale.
Wood is a cellular material with an intrinsic struc-
tural hierarchy (Fengel & Wegener 2003; Kollmann &
Ct 1968). The wood cells are hollow tubes oriented
in the stem direction (longitudinal direction L). In
softwood, the cell diameter is typically around 20
40 m, while in hardwood, there exist additional cells
with up to 500 m diameters, forming interconnected
pipe-like structures called vessels. The cell wall is
made up of cellulosic fibers with diameters of about
50200 nm, embedded in a non-cellulosic matrix. The
fibrils are helically wound within the cell wall, at an
inclination angle (microfibril angle) between 0 and
30 towards the cell axis (see Fig. 1). Only the inner
core of a fibril is crystalline, while the surface region
is more or less amorphous. The matrix deposited
in the spaces between the cellulose is composed of
non-cellulosic polysaccharides (commonly denoted as
hemicelluloses), lignin, and extractives. In the wet
state, water molecules interpenetrate the matrix.
Predictions of the macroscopic mechanical behav-
ior require consideration of the wood microstructure.
The most common approach is that of cellular solids
(Gibson & Ashby 1997), which provides only very
poor estimates of the elastic behavior with errors of
more than 1000% between experimental values and
theoretical estimates (see Gibson & Ashby (1997),
p. 403, Fig. 10.12). More detailed descriptions of the
(cell axis, stem direction)
Microfibril angle
Figure 1. Fibrillar (micro-) structure of wood cell wall.
microstructure of wood, based on laminate theory for
the representation of the internal structure of the cell
wall (Yamamoto et al. 2002; Bergander & Salmn
2000), often require material parameters which lack
a clear experimental basis. This limits the applicabil-
ity of such models; in particular, it does not permit
model predictions for non-tested conditions.
Thus, the authors invested in more reliable model
predictions resting upon the method of continuum
micromechanics (Zaoui 2002; Suquet 1997). In a
first approach, the hierarchical structure of wood was
represented by a four-step homogenization scheme
(Hofstetter et al. 2005), which is shortly recalled in
Section 2. A good agreement between model pre-
dictions and corresponding experimental results in
terms of Youngs moduli in the longitudinal and the
transversal direction was obtained. This confirmed the
suitability of the chosen approach as valuable tool

149
Copyright 2005 Taylor & Francis Group plc, London, UK
for estimating the elastic properties of wood. Shear
properties, however, were reproduced insufficiently by
the micromechanical model, presumably due to the
assumption of a zero microfibril angle. In detail, the
axial straining of inclined fibers upon (macroscopic)
shear loading of the composite material was disre-
garded. The crucial influence of the microfibril angle
on the mechanical properties of wood has already been
stressed by several authors, e.g. Cave & Walker (1994),
Reiterer et al. (1999).
Herein, we focus on consideration of the fiber
inclination in the framework of continuum micro-
mechanics.
2 CONTINUUM MICROMECHANICS MODEL
2.1 Fundamentals of continuum micromechanics
In continuum micromechanics (Zaoui 2002; Suquet
1997), a material is understood as a micro-
heterogeneous body filling a representative volume
element (RVE) with characteristic length l, l >>d, d
standing for the characteristic length of inhomo-
geneities within the RVE (see Fig. 2). The homog-
enized mechanical behavior of the material, i.e. the
relation between homogeneous deformations acting
on the boundary of the RVE and resulting (average)
stresses, can then be estimated from the mechanical
behavior of different homogeneous phases (repre-
senting the inhomogeneities within the RVE), their
dosages within the RVE, their characteristic shapes,
and their interactions. Based on matrix-inclusion prob-
lems (Eshelby 1957; Laws 1977), an estimate for the
homogenized stiffness of a material reads as (Zaoui
2002).
where cr and fr denote the elastic stiffness and the
volume fraction of phase r, respectively, and I is the
fourth-order unity tensor. The two sums are taken over
all phases of the heterogeneous material in the RVE.
The fourth-order tensor P0r accounts for the charac-
teristic shape of phase r in a matrix with stiffness
C0 . Choice of this stiffness describes the interactions
between the phases: For C0 coinciding with one of
the phase stiffnesses (Mori-Tanaka scheme), a com-
posite material is represented (contiguous matrix with
inclusions); for C0 = Cest (self-consistent scheme), a
dispersed arrangement of the phases is considered
(typical for polycrystals). If a single phase exhibits
a heterogeneous microstructure itself, its mechanical
behavior can be estimated by introduction of RVEs
150
Copyright 2005 Taylor & Francis Group plc, London, UK
d
l
d2
l2
Figure 2. Multistep homogenization.
within this phase, with dimensions l2 << d, compris-
ing again smaller phases with characteristic length
d2 << l2 , and so on (see Fig. 2). This leads to a
multistep homogenization scheme. Such a procedure
should, in the end, provide access to universal phase
properties inherent to all wood tissues, at a sufficiently
low observation scale.
2.2 Continuum micromechanics model for wood
It follows from the aforementioned hierarchical orga-
nization that universal properties inherent to all wood
species may be identified only at an observation scale
below the cell wall. Its components, namely crystalline
cellulose, amorphous cellulose, hemicellulose, lignin,
and water together with extractives, are probable to
exhibit such universal properties and, thus, are candi-
dates for (elementary) phases of the micromechanical
model.
The interaction of the elementary components is
considered in four homogenization steps:
(I) Within an RVE of polymer network with 820 nm
characteristic length, hemicellulose, lignin, and
water are intimately mixed. The disorder and the
intense mixing of the chemical components in
the network motivate the use of a self-consistent
scheme with inclusions of spherical shape.
(II) Within an RVE of cell wall material with 0.5
1 m characteristic length, cylindrical fiber-like
aggregates of crystalline cellulose and of amor-
phous cellulose, exhibiting typical diameters of
20100 nm, are embedded in a contiguous poly-
mer matrix. The behavior of such a composite
material is suitably estimated by a Mori-Tanaka
scheme. The cylindrical inclusions are inclined
by the microfibril angle to the longitudinal direc-
tion L.
(III) Within an RVE of softwood with 100150 m
characteristic length, longitudinally oriented
cylindrical pores with characteristic diameters
of 2040 m, representing the cell lumens, are
embedded in a contiguous matrix built up by
the cell wall material of homogenization step II.
 Again a Mori-Tanaka scheme is used for the esti- x3 (longitudinal
mation of the stiffness of the composite material direction L)
softwood.
(IV) Within an RVE of hardwood with 24 mm char- x3'
acteristic length, cylindrical pores with charac-
teristic diameters of 400500 m, representing
vessels, are embedded in a contiguous matrix
built up by the softwood-type porous mate-
rial of homogenization step III. Stiffness esti- x2
mates for the composite material hardwood are
again obtained by application of a Mori-Tanaka x1
scheme.

For the application of Equation 1 to the stiff-
ness estimates at observation scales (I), (III), and
(IV), we refer to (Hofstetter et al. 2005). The second
homogenization step, being related to the microfibril
deviation from the longitudinal axis, deserves further
consideration.
3 CONSIDERATION OF NON-ZERO
MICROFIBRIL ANGLES
Figure 3. Orientation of a microfibril defined by two Euler
angles and in a Cartesian coordinate system Ox1 x2 x3 .
cellulose) and one polymeric matrix phase, i.e. for
r [crycel; amocel; polynet], for phase stiffnesses
ccrycel and camocel , for C0 = cpolynet , and for P0crycel =
polynet
P0amocel = Pcyl , resulting in

3.1 Treatment of inclined inclusions in the

framework of continuum micromechanics
The cellulose microfibrils are arranged in the wood
cell wall in several concentric layers around the cell
lumen (Fengel &Wegener 2003). The arrangement of
the microfibrils is approximately axisymmetric at the
observation scale of an entire cell (Bergander 2001).
The assumption of axisymmetry implies transverse
isotropy of the material behavior of the cell wall, as
in case of a zero microfibril angle.
polynet
The orientation of the microfibrils in a Cartesian The fourth-order tensors cr and Pcyl depend on
coordinate system Ox1 x2 x3 is defined by the Euler the orientation of the inclusions. They are generally
angles and (see Fig. 3). Provided that the x3 -axis defined in a Cartesian coordinate system Ox1 x2 x3
is aligned with the axis of the wood cell, the angle aligned with the orientation of the (cylindrical) inclu-
corresponds to the microfibril angle. sion (see Hellmich et al. (2004) and Hofstetter et al.
Consideration of the inclination of the microfib- (2005) for mathematical details). The transformation
rils in homogenization step II requires reformulation of their components to the reference system Ox1 x2 x3
of Equation 1 for inclusions with orientation distri- requires standard rules of tensor calculus. For a fourth-
butions. Given the axisymmetric arrangement of the order tensor L, the component-wise transformation
fibrils with respect to and their constant inclination from a base e i , i = 1, 2, 3, to a base ej , j = 1, 2, 3,
angle = , we obtain rotated by Euler angles and , reads as

where nim denotes the cosine of the angle between

the base vectors ei and e m . It is a component of the
transformation tensor n reading in matrix notation as

In order to estimate the stiffness of the cell wall

material, Equation 2 is specified for two (cylin- The integrals in Equation 2 are evaluated numerically,
drical) inclusion phases (crystalline and amorphous the integral being replaced by a sum.

151
Copyright 2005 Taylor & Francis Group plc, London, UK
Table 1. Microfibril angle (MFA) measurements for differ- water lumped together with extractives, cH2 Oext , of
ent wood species by means of direct methods. amorphous cellulose, camocel , and of crystalline cellu-
lose, ccrycel , [experimental set I, given in (Hofstetter et
Species Methods MFA [ ] Reference al. 2005)] and tissue and sample-dependent composi-
tion data in terms of volume fractions of the universal
Spruce PM 18.120.4 Marton & McGovern (1970)
constituents, fhemcel , flig , fH2 O , fext , famocel , and fcrycel ,
PM 25.5 Ollinmaa (1961)
SM 1317 El-Osta et al. (1972) and of volume fractions of lumen pores, flum , and
vessels in hardwoods, fves , [experimental set IIa, col-
Pine SM 440 Hiller (1964)
lected as in (Hofstetter et al. 2005), on the basis of a
PM 20 Mark (1967)
PM 1724 Tang (1973) degree of crystallinity of cellulose of 0.66 (Fengel &
PM 1227 Boyd & Foster (1975) Wegener 2003; Kollmann & Ct 1968)], are compared
PM 20 Harada (1965) with corresponding experimentally determined stiff-
Douglas SM 1328 El-Osta et al. (1972) ness values [experimental set IIb given in (Hofstetter
Fir PM 730 Erickson & Arima (1974) et al. 2005)].
IC 1030 Fioravanti (2001) The focus of the present validation is on the lon-
Hemlock SM 1722 El-Osta et al. (1972) gitudinal shear modulus, G13 = G31 = GL . Model esti-
mates for GL , based on non-deviating and deviating
microfibrils ( = 0 , 20 ), are compared with cor-
responding experimental results for 73 samples of
3.2 Experimental determination of the 14 different wood species (see Fig. 4). Each pair of
microfibril angle stiffness measurement and corresponding microme-
Methods to measure the microfibril angle include chanical prediction is indicated by a wood speciesspe-
direct and indirect techniques. The most popular cific marker. The solid lines indicate locations of
indirect method is X-ray scattering (wide angle perfect agreement between predicted and experimen-
X-ray scattering (Cave 1966; Evans 1999) and small tal stiffness values. The stiffening effect of inclined
angle X-ray scattering (Jakob et al. 1994; Reiterer et al. microfibrils results in a considerable increase of the
1998; Lichtenegger et al. 1999)), which derives the shear stiffness, leading to an improved agreement
microfibril angle from diffraction patterns of irradi- between model predictions and experimental results
ated cell wall sections. Computation of the microfibril for non-zero microfibril angles, as compared to a zero
angle from the measured radiation intensities requires microfibril angle investigated earlier by Hofstetter
several assumptions, concerning, e.g., the shape of the et al. (2005) (see Fig. 4(a)).
cell or the orientation of the cellulose crystallites with The agreement is quantified by mean values e and
respect to the cellulose microfibrils (Cave 1997). This standard deviations s of the normalized errors between
somehow limits the quantitative significance of such micromechanical stiffness estimates and experimental
test results. measurements, included in Figure 4. While e = 1.3%
Direct methods include different applications of can be regarded as almost perfect, also the correspond-
polarization microscopy (PM) (El-Hosseiny & Page ing standard deviation of 27.4% is fair and does not
1973; Donaldson 1991), staining methods (SM) indicate the necessity for employing individual set-
(El-Osta et al. 1972; Hiller 1964), or examination tings of the microfibril angle for different species or
of the direction of iodine crystals (IC) (Senft & even samples.
Bendtsen 1985) or soft rot cavities (Khalili et al. 2001). The elastic moduli in longitudinal and transversal
Providing direct access to the microfibril angle, these direction, EL and Etrans , are affected by the microfib-
methods provide data free of any manipulation during ril angle as well, though not to a comparably high
evaluation. Results of microfibril angle measurements extent as the shear modulus GL . While EL decreases
by means of direct methods indicate a mean angle of with increasing microfibril angle, Etrans remains
20 for the wood cell wall of different species (cf. almost constant. The reduction of EL with increas-
Table 1). ing microfibril angle starts very gently at angles close
to zero and continuously passes into a rapid drop at
angles above 1015 (cf. also Cave & Walker (1994)).
4 VALIDATION Error measures for the model predictions of the elastic
moduli obtained with zero and non-zero microfibril
The validation of the micromechanical model is based angle, in comparison with experimental results, are
on two independent sets of experimental data, see e.g. listed in Table 2, based on 118 samples of 16 differ-
Hellmich et al. (2004) and Hofstetter et al. (2005): ent species for EL and on 84 samples of 14 different
Stiffness estimates of the micromechanical model on species for Etrans . The excellent predictive capabilities
the basis of tissue-independent phase stiffness prop- of the model underline the role of crystalline cellulose,
erties of hemicellulose, chemcel , of lignin, clig , of amorphous cellulose, hemicellulose, lignin, and water

152
Copyright 2005 Taylor & Francis Group plc, London, UK
 1.5 1.5
e = 53.8% e = 1.3%
s = 25.5% s = 27.4%

1 1
[GPa]

[GPa]
pred

pred
GL

GL
0.5 0.5

0 0
0 0.5 1 1.5 0 0.5 1 1.5
exp exp
GL [GPa] GL [GPa]
(a) Microfibril angle = 0 (b) Microfibril angle = 20
pred exp
Figure 4. Comparison of predicted and measured longitudinal shear moduli GL and GL ;  Cedar, Douglas Fir,  Fir,

Hemlock,  Larch,  Pine, + Spruce,  Ash, Balsa, Beech,  Birch,  Maple, Oak,  Poplar.

Table 2. Error measures for elastic moduli EL and Etrans readily extended to the prediction of poroelastic prop-
depending on microfibril angle . erties, like Biot and Skempton coefficients of wood,
as shown for bone in Hellmich & Ulm (2004). Fur-
EL Etrans ther information on their potential relevance for wood
e (%) s (%) e (%) s (%) drying technology will be given at the conference.

= 0 20.8 28.3 0.7 32.7

= 20 8.2 25.5 2.2 32.6
REFERENCES

Ariault, J. & Sanchez-Palencia, E. (1977). Etude du com-

together with extractives as universal elastic building
blocks inherent to all wood species and tissues.
5 CONCLUSIONS
Experimental validation of a continuum microme-
chanics model for wood elasticity elucidates that lon-
gitudinal shear stiffness of wood is largely due to axial
straining of cellulose fibers inclined towards the longi-
tudinal stem axis. An experimentally evident microfib-
ril angle of 20 results in excellent micromechanics
model predictions, in particular, for elastic moduli in
longitudinal and transversal direction as well as for the
longitudinal shear modulus GL of a multitude of wood
samples of different softwood and hardwood species.
The small mean values of the corresponding relative
errors between model predictions and experimental
results, together with a fair standard deviation of these
errors, indicate that the microstructure of wood cells
can be suitably characterized by a single microfibril
angle valid for different wood species, at least with
respect to prediction of macroscopic elastic properties.
Based on the theoretical landmark contributions of
Ariault & Sanchez-Palencia (1977) and Chateau &
Dormieux (2002), the current developments can be
portement macroscopique dun milieu poreux satur
dformable [Study of macroscopic behavior of satu-
rated deformable medium]. Journal de Mcanique 16(4):
575603. in French.
Bergander, A. (2001). Local variability in chemical and phys-
ical properties of spruce wood fibers. Doctoral Thesis,
KTH Stockholm.
Bergander, A. & Salmn, L. (2000). Variations in transverse
fibre wall properties: Relations between elastic properties
and structure. Holzforschung 54: 654660.
Bodig, J. & Jayne, B. (1982). Mechanics of Wood
and Wood Composites. New York, USA: Van Nostrand
Reinhold.
Boyd, J. & Foster, R. (1975). Microfibrils in primary and
secondary wall growth develop trellis configurations.
Canadian Journal of Botany 53(23): 26872701.
Cave, I. (1966). Theory of X-ray measurement of microfibril
angle in wood. Forest Products Journal 16: 3742.
Cave, I. (1997). Theory of X-ray measurement of microfibril
angle in wood. Part I. The diffraction diagramm X-ray
diffraction by materials with fibre type symmetry. Wood
Science and Technology 37: 225234.
Cave, I. & Walker, J. (1994). Stiffness of wood in fastgrown
plantation softwoods: the influence of microfibril angle.
Forest Products Journal 44(5): 4348.
Chateau, X. & Dormieux, L. (2002). Micromechanics of satu-
rated and unsaturated porous media. International Journal
for Numerical and Analytical Methods in Geomechanics
26: 831844.

153
Copyright 2005 Taylor & Francis Group plc, London, UK
Donaldson, L. (1991). The use of pit apertures as windows to
measure microfibril angle in chemical pulp fibres. Wood
Fiber Science 23: 290295.
El-Hosseiny, F. & Page, D. (1973). The measurement of fibril
angle of wood fibers using polarized light. Wood Fiber 5:
208214.
El-Osta, M., Wellwood, R. & Butters, R. (1972). An improved
X-ray technique for measuring microfibril angle of conif-
erous wood. Wood Science 5(2): 113117.
Erickson, H. & Arima, T. (1974). Douglas-fir wood quality
studies part II: Effects of age and stimulated growth on
fibril angle and chemical constituents. Wood Science and
Technology 8: 255265.
Eshelby, J. (1957). The determination of the elastic field of an
ellipsoidal inclusion, and related problems. Proceedings
of the Royal Society London A241: 376396.
Evans, R. (1999). A variance approach to the X-ray diffrac-
tometry estimation of microfibril angle in wood. Appita
Journal 52: 283294.
Fengel, D. & Wegener, G. (2003). Wood. Chemistry, Ultra-
structure, Reactions. Remangen: Verlag Kessel.
Fioravanti, M. (2001). The influence of age and growth fac-
tors on microfibril angle in wood. In Proceedings of the 1st
Conference of the European Society of Wood Mechanics,
pp. 127134.
Gibson, L. & Ashby, M. (1997). Cellular Solids. Structure
and Properties (second ed.). Cambridge, UK: Cambridge
University Press.
Harada, H. (1965). Cellular ultrastructure of woody plants.
In W. Ct (Ed.), Ultrastructure and organization of gym-
nosprem cell walls, Syracuse, pp. 215233. Syracuse
University Press.
Hellmich, C., Barthlmy, J.-F. & Dormieux, L. (2004).
Mineral-collagen interactions in elasticity of bone ultra-
structure a continuum micromechanics approach. Euro-
pean Journal of Mechanics A/Solids 23: 783810.
Hellmich, C. & Ulm, F.-J. (2004). Drained and undrained
poroelastic properties of healthy and pathological bone: A
poro-micromechanical investigation. Transport in Porous
Media: available online at www.kluweronline.com.
Hiller, C. (1964). Correlation of fibril angle with wall thick-
ness of tracheids in summerwood of slash and loblolly
pine. Tappi 47(2): 125128.
Hofstetter, K., Hellmich, C. & Eberhardsteiner, J. (2005).
Develpoment and experimental validation of a contin-
uum micromechanics model for the elasticity of wood.
European Journal of Mechanics A/Solids: submitted for
publication.
154
Copyright 2005 Taylor & Francis Group plc, London, UK
Jakob, H., Fratzl, P. & Tschegg, S. (1994). Size and arrange-
ment of elementary cellulose fibrils in wood cells: A
Small-Angle X-Ray Scattering study of Picea Abies.
Journal of Structural Biology 113: 1322.
Khalili, S., Nilsson, T. & Daniel, G. (2001). The use of soft
rot fungi for determining the microfibrillar orientation in
the s2 layer of pine tracheids. Holz als Roh- und Werkstoff
58: 439447.
Kollmann, F. & Ct, W. (1968). Principles of Wood Science
and Technology, Volume 1. Berlin, Heidelberg, New York:
Springer Verlag.
Laws, N. (1977). The determination of stress and strain con-
centrations at an ellipsoidal inclusion in an anisotropic
material. Journal of Elasticity 7(1): 9197.
Lichtenegger, H., Reiterer, A., Stanzl-Tschegg, S. & Fratzl, P.
(1999). Variation of cellulose microfibril angles in soft-
woods and hardwoods a possible strategy of mechanical
optimization. Journal of Structural Biology 128: 257269.
Mark, R. (1967). Cell wall mechanics of tracheids (2 ed.).
New Haven, USA: Yale University Press.
Marton, R. & McGovern, S. (1970). Relation of the crystallite
dimensions of fibrillar orientation to fiber properties. In
The physics and chemistry of wood pulp fibers, pp. 153
158. Tappi stap No. 8.
Ollinmaa, P. (1961). Reaktiopuututkimuksai. Acta Forestalia
Fennica 72.
Reiterer, A., Jakob, H., Tschegg, S. & Fratzl, P. (1998). Spiral
angle of elementary cellulose fibrils in cell walls of picea
abies determined by small-angle Xray scattering. Wood
Science and Technology 43(5): 335345.
Reiterer, A., Lichtenegger, H., Tschegg, S. & Fratzl, P. (1999).
Experimental evidence for a mechanical function of the
cellulose microfibril angle in wood cell walls. Philosoph-
ical Magazine A 79(9): 21732184.
Senft, J. & Bendtsen, B. (1985). Measuring microfibrillar
angles using light microscopy. Wood Fiber Science 17:
564567.
Suquet, P. (Ed.) (1997). Continuum Mircomechanics. Wien,
New York: Springer Verlag.
Tang, R. (1973). The microfibrillar orientation in cellwall
layers of virginia pine tracheids. Wood Science and Tech-
nology 5: 181186.
Yamamoto, H., Kojima, Y., Okuyama, T., Abasolo, W. &
Gril, J. (2002). Origin of the biomechanical properties of
the fine structure of the multi-layered cell wall. Journal
of Biomechanical Engineering (ASME) 124: 432440.
Zaoui, A. (2002). Continuum micromechanics: Survey.ASCE
Journal of Engineering Mechanics 128(8): 808816.