You are on page 1of 9

Thermal conductivity and diffusivity of wood

B. M. Suleiman, J. Larfeldt, B. Leckner, M. Gustavsson


Summary Transient simultaneous measurements of thermal conductivity and
diffusivity of Swedish wood have been performed with the plane source technique
on oven-dry hardwood (birch) samples at room temperature and at 100 C. The
inuences of temperature, density, porosity and anisotropy on thermal
conduction were investigated. The measurements were done in longitudinal
(parallel to the grain) and transverse (intermediate between radial and tangential)
directions. As the temperature increased from 20 to 100 C, the thermal
conductivity of each sample increased slightly for both longitudinal and
transverse directions. The effect of density and porosity on the thermal
conductivity may be related to the presence of other scattering mechanisms such
as voids and cell boundaries. It seems that the dominant mechanism of heat
transfer across the cell lumina in these types of wood is the heat conduction
through the voids. An attempt was made to explain the behaviour of the effective
thermal conductivity by adopting a model based on the ratio between heat
conduction in parallel and serial layers of gas, liquid, and solid phases.
Introduction
Thermal properties of wood are needed in applications such as fuel conversion,
building construction and other elds of industry (Kamke and Zylkowski 1989).
Wood is a ligno-cellulosic material (Avramidis and Lau 1992). Due to tem-
perature variations, the crystalline structure of cellulose chains may be altered
resulting in permanent loss in strength and considerable changes in physical
behaviour including its ability to conduct heat.
The purpose of this study is to investigate the relation between thermal
properties of wood and the various factors affecting these properties. The more
Wood Science and Technology 33 (1999) 465473 Springer-Verlag 1999
Received 7 May 1997
B. M. Suleiman (&)
Department of Physics, Chalmers University of Technology,
S-412 96 Gothenburg, Sweden
J. Larfeldt
1
, B. Leckner
1
, M. Gustavsson
2
1
Department of Energy Conversion,
2
Department of Thermo and Fluid Dynamics,
Chalmers University of Technology,
S-412 96 Gothenburg, Sweden
We thank A. Green from the Department of Energy Conversion
at Chalmers University of Technology for the technical assistance
in preparing the sample holder and producing the samples
465
important inuencing factors are species, density, moisture content, direction of
heat ow (anisotropy), inclination of grain, and relation of volume or thickness of
the sample to moisture content. The study is focused on three of these, namely:
density, porosity and anisotropy. Density is one of the most important factors. It
exerts a considerable inuence on the transmission of heat through wood. Poro-
sity has its effect via air and/or water present in the voids and the distribution of
these voids within the medium. Anisotropy has a pronounced effect, either on the
type of species or on the cells and bers that constitute the samples.
Most of the published thermal conductivity measurements on wood samples
were made with standard hot-plate apparatus (steady-state method), in which the
samples were placed and left under constant conditions for a sufcient length of
time to insure a uniform temperature gradient throughout the sample. The
temperatures of the test surfaces were then recorded, and the rate of heat ow was
calculated from the electric input to the heating element, see i.e., MacLean (1941)
and Skaar (1988). In the present work, the recently developed Transient Plane
Source (TPS) technique, cf. e.g. Suleiman et al. (1997), is used to perform
simultaneous measurements of thermal conductivity and diffusivity. The mea-
surements are performed on oven-dry hardwood samples at room temperature
and at 100 C. These measurements are done in the longitudinal (parallel to the
grain) and the transverse direction (a combination of both radial and tangential
directions to the annual rings). Six samples of hardwoods (birch trees) were taken
from a site at the north-east of Gothenburg.
Experiments
Several thermal conductivity measurements have been reported using the tran-
sient plane source (TPS) technique. Suleiman (1994) gives a complete description
of the experimental capability regarding precision/accuracy and reproducibility
of the measured data of various applications. Yet, a brief description of the main
features and the principles of this technique is highlighted below.
The TPS technique uses a resistive heating pattern (TPS element) cut from a
thin sheet of metal and covered on both sides with thin layers of an insulating
(mica) material. The TPS element/sensor is used as both heat source and tem-
perature sensor, i.e. in a similar manner as the wire in the transient hot-wire
method.
To ensure good thermal contact, in the experimental arrangement the sensor is
clamped between the sample halves that consist of two identical cylinder-shaped
pieces, each having a diameter of 30 mm and a thickness of 10 mm.
The experiment is performed by recording the voltage/resistance variations
over the TPS-element, while its temperature is slightly raised (by 1 K) by a
constant electrical current pulse.
The time-dependent resistance of the TPS element during the transient
recording can be expressed, in rst approximation, as
R(t) = R
0
[1 aDT(t)[ (1)
where R
0
(~5 W at room temperature) is the resistance of the TPS element before
the transient recording has been initiated, a is the temperature coefcient of
resistance (TCR) for the TPS element (TCR ~ 4.0 10
)3
K
)1
at room tempera-
ture), and DT(t) is the time-dependent temperature increase of the TPS element.
The theory of the method is based on a three-dimensional heat ow inside the
sample. However, the samples were cut in a specied direction to maintain
466
unidirectionally measured properties and to minimise any additional contribu-
tion of heat ow from the perpendicular directions to the assigned direction of
measurement (see next section). By cutting the sample in a specied direction, it
can be regarded as a unidirectional innite medium, provided the time of the
transient recording is ended before the thermal wave reaches the boundaries of
the sample and produces edge effects. The assessment of the temperature increase
DT(t) in the heater depends on several factors, such as the power output in the
TPS element, the design parameters of the sensor, and the thermal transport
properties of the surrounding sample. For a disk-shaped sensor, the thermal
conductivity and diffusivity can be obtained from DT(t) that is given by the
following equation (Gustafsson 1991),
DT(s) = P
0
(p
3=2
ak)
1
D(s) (2)
Here, P
0
is the total output power, k the thermal conductivity of the sample, and a
the radius of the sensor. D(s) is the theoretical expression of the time dependent
temperature increase, which describes the conducting pattern of a disk-shaped
sensor, assuming that the disk consists of a number of concentric ring sources.
For convenience, the mean temperature change of the sensor is dened in terms
of the non-dimensional variable s, where s = (jt/a
2
)
1/2
or s = (t/h)
1/2
, t is the time
measured from the start of the transient heating, h = a
2
/j is the characteristic
time, and j the thermal diffusivity of the sample. Fitting the experimental values
of DT(t) from Eq. (1) to the theoretical values of DT(s) in Eq. (2) will yield all
desired information.
Preparation and description of samples
The wood samples investigated are hardwood from birch trees used for energy
conversion purposes (approximate age of 1015 years) with a stem diameter of
approximately 100 mm. The samples were cut from the stem at one meter above
ground level. The samples were prepared from different stems, each sample
consisting of two disks (30 mm in diameter and 10 mm thickness). Six samples
were selected, so that sufcient uniformity of direction of grain, density and other
physical characteristics could be obtained. The two disks required for each test
sample were cut from successive portions of the same stem to obtain as many
identical and uniform characteristics as possible. To study the inuence of bre
direction on thermal conduction, the disks were cut parallel (three samples) or
perpendicular (three samples) to the bre direction. Their surfaces were sanded
and equilibrated to the same testing laboratory conditions and then the mea-
surements were performed in the directions mentioned above. All samples were
oven-dried at a temperature of 120 C until they reached a constant mass. At a
temperature of 120 C, both the free and the bound water in wood are released,
and thus the samples were completely dry. On the basis of oven-dry masses and
volumes, the average values of the apparent densities were calculated for each
sample at the time of measurements. To prevent presence of any moisture, the
samples were wrapped in Al-foil during the measurements and kept in plastic
containers at all other times.
Wood is a porous material and porosity is a parameter which is likely to affect
the magnitude as well as the temperature dependence of the thermal properties.
This porosity r can be estimated using
r = 1 (q
ave
=q
th
) (3)
467
where q
ave
is the average apparent density of the sample, q
th
is the assumed
theoretical density of a compact solid wood free from voids. Its value is assumed
to be 1500 kg/m
3
(Kollmann and Cote 1968).
Results
The mean values of ve runs of k and j at 20 and 100 C, respectively, are
represented in Tables 1 and 2 for both longitudinal and transverse directions.
The accuracy was within 35% for the conductivity and within 810% for the
diffusivity measurements.
These tables also show the corresponding average apparent densities, q
ave
, and
porosities, r, for the six different samples. According to the estimated porosity
given in the Tables, and considering the fact that the samples have similar
thermal conductivity behaviour in terms of their thermal proles, the longitu-
dinal (axial) samples would be expected to have somewhat higher k-values than
the transverse samples as shown in the Tables (to be discussed).
Figure 1a and b give the variations of the thermal conductivity k in both
directions as a function of temperature and density. The thermal conductivity
was higher at higher temperatures with a higher slope for the axial sample at
100 C. The thermal conductivities as function of porosity are given by the upper
axis. They exhibit the same behaviour as the density, but with negative slopes as
given by Eq. (3): the conductivity decreases as porosity increases. Although
the few data points in the gure show the expected general trend for conductivity,
the gure does not explain why, at xed temperature, there is a difference in
the slope as density increases or porosity decreases (see Discussion).
Table 1. Thermal conductivity and diffusivity in the longitudinal direction for the six
samples, at 20 C and 100 C. The porosity r is dened by Eq. (3) using the corresponding
listed densities
Temp. (C) q
ave
(kg/m
3
) r k (W/m C) j (mm
2
/s)
21 680 0.547 0.323 0.513
567 0.622 0.293 0.496
543 0.636 0.291 0.496
100 680 0.547 0.370 0.307
567 0.622 0.309 0.322
543 0.636 0.318 0.357
Table 2. Thermal conductivity and diffusivity in the transverse direction for the six
samples, at 20 C and 100 C. The rest of symbols as in Table 1
Temp. (C) q
ave
(kg/m
3
) r k (W/m C) j (mm
2
/s)
21 680 0.547 0.214 0.418
473 0.689 0.196 0.537
443 0.705 0.177 0.449
100 680 0.547 0.250 0.335
473 0.689 0.244 0.470
443 0.705 0.207 0.373
468
Discussion
Thermal conductivity studies have been carried out by several research groups,
(e.g. Ratcliffe 1964; Steinhagen 1977; Chia et al. 1985). The present measurements
are in good agreement with these literature data for similar wood types. In
general, however, the thermal properties of wood are rather limited and mostly
concern thermal conductivity. Furthermore, data on other associated parameters,
such as thermal diffusivity and specic heat, are scarce. Due to the essential
demand for thermal conductivity data in industry and research, the discussion
will be focussed on this parameter and its relation to temperature, microstructure
(density and/or porosity) and anisotropy.
1. Inuence of temperature on thermal conduction
The present values of the thermal conductivity behave similarly to recent
published data for light-medium hardwood samples (Chia et al. 1985; Raznjevic
1995). As temperature increases from 20 to 100 C, conductivity slightly increases
up to 14 and 24% in longitudinal and transverse directions, respectively.
Although there is an agreement between the thermal proles of all samples, the
variations of conductivities of similar samples at each xed temperature have a
unique pattern. Presumably, this results from the type of grain, which differs with
each sample, and has a corresponding effect on the thermal conduction.
The values of k at 100 C are higher than those at 20 C for all samples. This
is contrary to the general expectation of air lling the voids in the wood, since
the density of air decreases as temperature increases, thus leading to lower heat
conduction through the voids. The higher values at 100 C may be due to an-
other mode of heat transfer involving the detailed microstructure of cell-wall
substance (see Sect. 4), rather than a simple heat transfer by conduction. Fur-
thermore, variations of the values of k at 100 C are higher than those at 20 C,
which may be due to the natural composition of wood as a ligno-cellulosic
material; at the cellular level, the crystalline structure of cellulose chains may be
altered at higher temperatures resulting in strength loss and changes in its
ability to conduct heat.
Fig. 1. Thermal conductivity as a function of temperature, density and porosity (upper
axis) in (a) the longitudinal and (b) the transverse directions
469
2. Inuence of the microstructure on thermal conduction
Wood is a porous material having a cellular substance with cavities (voids) lled
by air. Typically, porosity ranges from 0.5 to 0.8. Voids due to porosity serve
as scattering centres for phonons, and they take up a fraction of the heat con-
duction volume of the material leading to a lower thermal conductivity. From the
present data related to density and porosity, the following can be deduced: Firstly,
a density variation of 20% could alter the conduction by 15% as it is the case
for the longitudinal direction at 20 C. Secondly, the presence of other heat
conduction obstacles in addition to voids, such as rays and cell boundaries could
also affect the conduction process. However, it seems that conduction through
voids is the dominant factor inuencing heat conduction in this type of wood
samples. Peculiarities in bulk structure or inherent chemical composition of the
different samples appear to have no inuence upon thermal conduction, other
than that due to the unique microstructure of each sample.
3. The inuence of anisotropy on thermal conduction
Due to the anisotropic nature of wood there is no dispute about direction
dependency of thermal conductivity in both hard and soft woods, see for example
the review by Steinhagen (1977). Thermal conductivity in the longitudinal
direction (parallel to the grain) is greater than conductivity in the transverse
direction. This may be attributed to the orientation of the molecular chains within
the cell wall. The long-chain linear polymers (cellulose) that comprise the cell wall
are arranged in bundles called microbrils. These microbrils are most closely
aligned with the longitudinal axis of the cell. Obviously, the thermal conduction is
higher along the length of a microbril (parallel direction) than across a series of
microbrils (series direction), see Parrott and Stuckes (1975).
There might be some debate about the direction dependence of the transverse
conductivity (the difference in conductivity between the tangential and radial
directions with respect to the annual rings). Radial conductivity may be higher
than tangential conductivity and this may be attributed to the inuence of the
radially oriented wood rays. According to Steinhagen (1977), it appears that the
ratio of the tangential versus radial conductivity is primarily determined by the
volume of the ray cell in hardwoods and by the latewood volume in softwoods.
However, in our case, no appreciable difference would be expected between the
radial and tangential conductivities, since our samples have a uniform structure
throughout the annual rings. Furthermore, our samples are cut from young trees,
i.e. the samples are mostly free from latewood.
On the other hand, our data show a signicant difference between the longi-
tudinal and transverse directions. This difference is expected to be larger than the
measured one. The reason for this is that in order to satisfy the theoretical
principle of the TPS method and to maintain unidirectional measured properties,
we have cut the samples in a specied direction and this will not completely
eliminate any contribution of heat ow from the perpendicular directions to the
assigned direction of measurement. Thus, the resultant measured thermal con-
ductivity in the present case is an averaged value which is lower or higher than the
actual value in the direction of measurement.
4. Theoretical model
Various models of heat conductivity are given in the literature, see references
given by Grnli (1996). In this study, we will use one (Kollmann and Cote 1968) of
these models. This model gives the effective thermal conductivity of a brous
470
structure by using a weighting bridge-factor (n) between two limiting cases. The
rst case is when conduction takes place in parallel layers of gas, liquid and
solid phases (k
P
) and the second one when it takes place in serial layers (k
^
) in
relation to the heat transfer direction. Furthermore, the model includes the
inuence of the radiation in the gas phase. According to this model the effective
thermal conductivity (k
eff
) for pre-dried wood can be written as:
k
eff
= nk
P
(1 n)k
l
(4)
k
P
= (1 r)k
fiberP
r(k
g
k
rad
) (4a)
k
l
=
1
(1r)
k
fiberl

r
(k
g
k
rad
)
(4b)
where r is the porosity, k
berP
is the thermal conductivity of the cell wall substance/
bers along the grain, k
ber^
is the corresponding conductivity perpendicular to
the grain, and k
g
is the gas (air) conductivity. The radiative thermal conductivity
k
rad
at the considered temperatures is very small and can be neglected.
A plot of the effective thermal conductivity (k
eff
) for various values of the
bridge-factor is shown in Fig. 2. This plot was generated by applying Eq. (4) and
using k
g
= 0.0258 W/m C for dry air, k
berP
= 0.766 W/m C and k
ber^
=
0.430 W/m C for the wood bers at a density of 1500 kg/m
3
. The values of k
berP
and k
ber^
have been used by Grnli (1996) and were taken from mean values at
room temperature reported by Maku (1954) and by Siau (1984).
Our data in the longitudinal direction at 20 C lie in the range n = 0.81.0 and
at 100 C between n = 0.9 and slightly above n = 1.0. The data in the trans-
verse direction at 20 C are in the range n = 0.50.7 and at 100 C between
n = 0.6 and n = 0.9.
At 20 C, our values for both directions are in agreement with literature
data. Grnli (1996) obtained n in the range from 0.8 to 1.0 for the longitudinal
Fig. 2. Effective thermal conductivity according to Eq. (4) for different n values as a
function of density and/or porosity (upper axis). The symbols of the data points are the
same as the corresponding ones in Fig. 1
471
direction and in the range from 0.35 to 0.6 for the transverse direction. Kollmann
and Cote (1968) found n = 1 for the longitudinal direction and n = 0.58 for
the transverse direction.
At 100 C, our values are slightly higher than the range predicted by the
model at 20 C. A comparison at 100 C was not possible since k
berP
and k
ber^
are not known at this temperature. However, according to this model, at 100 C
neither an increase in gas phase conductivity nor the radiative conductivity can
explain the net increase in the effective conductivity. This measured increase in
k
eff
is most likely due to the increase in the conductivities (i.e. k
berP
, k
ber^
) of
the cell wall substance. Therefore, the model is expected to give slightly higher
values of k
eff
if the conductivities (k
berP
, k
ber^
) of the cell wall substance
increase as the temperature increases. Furthermore, the model indicates an in-
crease in effective conductivity as the density/porosity increases/decreases which
agrees with our investigation regarding the inuence of microstructure on heat
conduction in wood.
It should be noted that there are some constraints imposed upon using this
model, such as a limited number of samples, using mean values for k
berP
and
k
ber^
and limiting the analysis to room temperature data. Such constraints
may introduce a degree of uncertainty in the conclusions. However, there is a
tendency towards good agreement between the measured and calculated thermal
conductivities. Furthermore, in this attempt of comparison, we have applied
this model for these six samples, just to examine the possibilities of determin-
ing the bounds for estimating the effective thermal conductivity.
5. Thermal diffusion in wood
The thermal diffusivity is a third important physical parameter that expresses the
response of the medium to a thermal perturbation. Therefore, measurements of
this parameter may reveal the dynamics behind the various conduction mecha-
nisms and the effect of microstructure on heat conduction in wood. The few
reports published on thermal diffusivity of wood have been described in details
in the review by Steinhagen (1977). Our data are higher than these reported data.
This might be related to the measuring conditions of our samples, since they
were completely dry (both free and bound water were released). The data re-
ported by Steinhagen were measured on green volume basis, and an adjustment
factor is needed to compensate the difference in moisture content and density of
the samples. To match the conditions of our samples, a larger value of the
adjustment factor is needed for these reported data. In such case, the comparison
between these data and those of the present investigation would show good
agreement.
The relation between diffusivity and heat conduction is expressed by three
parameters, thermal conductivity (k), thermal diffusivity (j) and specic heat (c)
which are related through j = k/qc, where q is the density. Since, in this case,
density and specic heat may be considered as direction-independent (isotropic)
properties, one would expect the isotropic dependence of the thermal diffusivity
to follow the same proportionality as that of thermal conductivity. However, this
type of proportionality is not manifested by the present measurements.
Finally, it should be pointed out that, due to the inherent variability of wood, a
larger number of measurements may be needed to reveal the relation between
density, porosity and anisotropy and their effects on thermal conduction. How-
ever, the present measurements show a degree of correlation between these
factors and their inuence on heat transfer in these species.
472
Conclusions
As the temperature increases from 20 to 100 C, the thermal conductivity slightly
increases in both longitudinal and transverse directions. The effect of density and
porosity on the thermal conductivity may be related to the presence of other
scattering mechanisms, such as voids and cell boundaries. It seems that porosity
(conduction through voids) is the dominant inuencing factor for the heat
conduction. By cutting the samples in a specied direction we can use the TPS
method to measure unidirectional thermal properties and detect a signicant
difference in the thermal conductivity between the longitudinal and the transverse
directions. Furthermore, the results are supported by prediction of the model
proposed by Kollmann and Cote (1968) in spite of the various assumptions
involved.
References
Avramidis S, Lau P (1992) Thermal coefcients of Wood Particles by Transient Heat-Flow
Method. Holzforschung 46(5): 449453
Chan WC, Kelban M, Krieger B (1985) Modelling and experimental verication of physical
and chemical processes during pyrolysis of a large biomass particle. Fuel 64: 15051513
Chia LHL, Chua PH, Lee EEN (1985) A Preliminary Study on The Thermal Conductivity
and Flammability of WPC Based on Some Tropical Woods. Radiat. Phys. Chem. 26(4): 423
432
Grnli MG (1996) A Theoretical and Experimental Study of The Thermal Degradation of
Biomass, PhD Theses, Department of Thermal Energy and Hydropower, The Norwegian
University of Science and Technology
Gustafsson SE (1991) Transient plane source techniques for thermal conductivity and
thermal diffusivity measurements. Rev. Sei. Instrum. 62(3): 797804
Kamke FA, Zylkowski SC (1989) Effects of wood-based panel characteristics on thermal
conductivity. Forest. Prod. J. 39(5): 1924
Kollmann FFP, Cote WA (1968) Principles of Wood Science and Technology I, Solid Wood.
Springer-Verlag, NewYork: 246250
MacLean JD (1941) Thermal Conductivity of Wood. Heat. Pip. & Air Cond. 13: 380391
Maku T (1954) Studies on Heat Conduction in Wood, Kyote University, Bulletin of the
Wood Research Institute No. 13: 180
Parrott JE, Stuckes AD (1975) Thermal conductivity of solids. Pion Limited, London
Ratcliffe EH (1964) A Review of Thermal Conductivity Data-1. Wood July: 4951
Ratcliffe EH (1964) A Review of Thermal Conductivity Data-2. Wood Aug.: 4649
Ratcliffe EH (1964) A Review of Thermal Conductivity Data-3. Wood Sep.: 5054
Raznjevic K (1995) Handbook of Thermodynamic Tables. Begell House, Inc., New York: 29
Rowley FB (1933) The Heat Conductivity of Wood at Climatic Temperature Differences.
Heat. Pip. & Air Cond. 5: 313323
Siau JF (1984) Transport Processes in Wood, Springer-Verlag, New York
Skaar C (1988) Wood-water relations. Springer-Verlag, New York: 283
Steinhagen HP (1977) Thermal properties of wood, green or dry, from )40 to +100 C:
A literature review. USDA Forest Service General Technical Report FPL-9, Forest Products
Laboratory, Madison, WI
Suleiman BM, Karawacki E, Gustavsson M, Lunden A (1997) Thermal properties of lithium
sulphate. J. Phys. D: Appl. Phys. 30: 25532560
Suleiman BM (1994) Ph.D. Thesis. Physics Department, Chalmers University of Technol-
ogy, Gothenburg, Sweden
Wangaard FF (1940) Transverse Heat Conductivity of Wood. Heat. Pip. & Air Cond. 12:
459464
473

You might also like