On the modelling of the acoustical properties of hemp

concrete
Philippe GLÉ, Kirill Horoshenkov, Emmanuel Gourdon, Laurent Arnaud

To cite this version:

Philippe GLÉ, Kirill Horoshenkov, Emmanuel Gourdon, Laurent Arnaud. On the modelling
of the acoustical properties of hemp concrete. Société Française d’Acoustique. Acoustics 2012,
Apr 2012, Nantes, France. <hal-00811275>

HAL Id: hal-00811275

https://hal.archives-ouvertes.fr/hal-00811275
Submitted on 23 Apr 2012

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
Proceedings of the Acoustics 2012 Nantes Conference 23-27 April 2012, Nantes, France

On the modelling of the acoustical properties of hemp

concrete
P. Gléa, K. V. Horoshenkovb, E. Gourdona and L. Arnauda
a
Département Génie Civil et Bâtiment, Rue Maurice Audin 69518 Vaulx-en-Velin cedex
b
University of Bradford, Great Horton Road, BD7 1DP Bradford, UK
philippe.gle@entpe.fr

959
23-27 April 2012, Nantes, France Proceedings of the Acoustics 2012 Nantes Conference

Abstract
(a) (b)
Hemp concrete is an attractive alternative to tradi-
tional materials used in building construction. It has a
very low environmental impact and it is characterized
by high thermal insulation and sound absorption prop-
erties. The shape of hemp aggregates is parallelepiped
and individual particles in a hemp mix can be organized
in a plurality of ways. As a result, modeling of such ma-
terial is quite complicated.
This paper is focused on the fundamental under-
standing of the relations between the particle shape and Figure 1: Photographs showing mix of hemp particles
size distribution and the acoustical properties of the re- SA ((a) shows the characteristic dimension of the
sultant material mix. The sound absorption and the particles and (b) shows the intraparticle porosity)
transmission loss of various hemp aggregates is charac-
terized using laboratory experiments and three theoreti-
cal models. These models are used to relate the particle In the case of granular media with a particle size distri-
size distribution to the pore size distribution. It is shown bution and a non-spherical shape, not much has been
that the pore size distribution is one of the main charac- done.
teristic which controls the observed acoustical behavior. Studies investigated and modeled the effect of pore
size distribution. Arbitrary and log-normal pore size
distributions have been studied by [11] and [12], by ex-
1 Introduction tending the Biot theory [13, 14]. These authors showed
that the pore size distribution has a noticed effect on the
There is an increasing interest today for alternative acoustical properties, and it has also be shown that this
materials in buildings, in order to meet the expectations effect is greater than the change in pore shape. Besides,
of low environmental impact and multifunctionnal prop- in our case, it is necessary to understand the effect of
erties. Hemp concrete is a mix of plant particles (hemp the particle size distribution on this pore size distribu-
particles) with a binder (lime or cement), and is one tion. The progress concerning the modelling of the rela-
of these key materials. Its life cycle analysis [1] shows tionship between these parameters has been reviewed in
that 1 m2 of hemp concrete having a width of 26 cm [15] and a strong dependance is shown between porosity,
encasing a timber frame stores 35.5 kg carbon dioxide particle shape and particle size distribution.
over a reference period of 100 years, and proves the eco- Finally, the influence of the shape of particles on the
logical interest of this material. Moreover, as shown in acoustical properties is also a key question. Usually,
[2, 3, 4, 5] it features very interesting properties from non-spherical shape of particles is considered empirically
mechanical, thermal and acoustical points of view, so and using a shape factor [16]. Numerical works have
that it can qualify as a multifunctional material. been realized by imputing the shape and configuration
The acoustical behaviour of hemp concrete has al- of non-spherical particles, for instance with spiky parti-
ready been the object of several studies. The effect of cles [17]. However, parallelepiped shape such as hemp
the binder content in the formulation was observed in particles have not been investigated.
[3]. Furthermore, the effect of density, particles size dis- The article focuses on five different types of shiv
tribution, type of binder, and water content was suc- called SA to SE and having different origins. The aim
cintly discussed in [5]. A first approach has been devel- is to highlight the effect of particle size distribution,
oped to model its acoustical properties in [5], and this particle shape, and multi-scale porosity of shiv, on the
model has been improved to highlight the effect of the acoustical properties. It is organized in three sections.
multiple scales of porosity existing in the material in [6]. The direct characterization methods are first described.
The porous microstructure of hemp concrete is quite Then the modelling approaches used to model this ma-
complex, because it is composite and has a natural ori- terial are presented and compared. Some of the models
gin. Therefore, pores have dimensions distributed through enable to complete the characterization process in an
three scales, with the inter-particle pores having a size indirect way. These indirect characterized parameters
ranging between 1mm to 10mm (See Figure 1 (a)), are discussed to explain the acoustical behaviour of the
intra-particle pores having sizes varying between 10 and material. Finally, the effect of the shape and particle
60 μm (See Figure 1 (b)) [7, 8], and with intra-binder size distribution are discussed.
pores having a characteristic size of 1 μm. In [6], it
is shown that the contrast of permeability existing be-
tween the inter-particle pores and the intra-particle pores 2 Experimental results - Direct
is high enough so that only the inter-particle pores take
part into the acoustical dissipation of hemp concrete.
characterization
The other difficulty concerning hemp concrete, and
2.1 Particles size distribution of the shiv
shiv (loose hemp particles), is the parallelepiped shape
of the aggregates and their particle size distribution. In- The hemp particles were first characterized by a par-
deed, the acoustical properties of spherical aggregates ticles size distribution analysis. Hemp particles shape is
having same size is now well known and modelled [9, 10]. assumed to be parallelepiped. So, to describe perfectly
a bed of particles, one has to know the distribution of

960
Proceedings of the Acoustics 2012 Nantes Conference 23-27 April 2012, Nantes, France

length, width and thickness. This is particularly diffi-

−3
cult and could only be made using 3D tomography. SA (ρ = 100kg.m ) − mean = −1.55 log(mm) − std = 0.76 log(mm)
We developed in the lab a more accessible method 200
Data

Population
based on image analysis. This method first developped Log−Normal PSD
by Ceyte [18] enables to know the distribution of length 100
and width of the particles.
This data have been compared to log-normal distri- 0
butions, and a perfect agreement has been found as de- 0 0.5 1 1.5 2
Req (mm)
scribed by Figure 2.
SA (ρ = 150kg.m−3) − mean = −1.51 log(mm) − std = 0.71 log(mm)
2000 500 200
Data

Population
400
1500 Log−Normal PSD
100
Population

Population

300
1000
200

500
100 0
0 0.5 1 1.5 2
0 0 Req (mm)
0 20 40 60 80 100 0 5 10 15 20
A (mm2) L/l

500
Figure 3: Pore size distribution of shiv SA , population
400
400 and associated log-normal distribution
Population

Population

300 300

200 200
These frame densities were measured for the five shiv
100 100
on three different samples of mass Mshiv of 4 g to en-
0
0 5 10 15 20 25
0
0 2 4 6
sure the representativity of the sample. The volume of
L (mm) W (mm) the frame Vf rame of the shiv has been evaluated using
air porosimetry [19].
Figure 2: Particle size distribution of shiv SA , Among the tested shiv, the frame density ranges be-
population and associated log-normal distribution tween 870 and 1350 kg.m−3 . These values for frame
density have the same order as in [5]. This leads to
These results enable to compare clearly the mean size open porosity ranging between 85% and 95% decreasing
of the particles and it appears for each parameter that lineary as a function of the density of the bed.
the standard deviation does not vary a lot as a function
of the shiv size.
2.4 Air flow resistance
2.2 Pore size distribution of the shiv The air flow resistivity, noted σ, has been estimated
from the low frequency range of the imaginary part of
Using a similar method, the pore size distribution the dynamic density ρeq .
has then been investigated for different configurations of The validity of this estimate has been checked by per-
shiv. Photographs of shiv SA have been taken for two forming direct measurement, following standard ISO9053
densities, a low density packing with ρ = 100kg.m−3 on shiv SA and gives satisfactory results. The measured
and a high density packing with ρ = 150kg.m−3 . Then, resistivities range between 1000 and 9000 N.m−4 .s, and
an image analysis has been performed in Matlab to de- increase with density and for shiv having smaller parti-
tect the pores and to evaluate their equivalent radii (ra- cles such as SE .
dius of a circle having same area than the pores). The evolution of resistivity can be modelled as a
Figure 3 shows the pore size distribution. These pore function of the inter-particle porosity and the character-
size distributions have been modeled using a log-normal istic size of the shiv, as presented in [6], and using mod-
distribution and again, a very good agreement has been els derived from self consistent methods. So, concerning
found. So, one can wonder if a granular material, whose resistivity, shiv present same behaviour than packings
aggregates are characterised by a log-normal size dis- of spheres having the same characteristic size.
tribution, has a log-normal pore-size distribution. This
will be discussed in Section 3.
Besides it appears on this figure that an increase of 2.5 Tortuosity
the density lowers the equivalent radius of the pores and Finally, tortuosity has been evaluated using a flight
its standard deviation. However, it is important to no- time method. This method described in [20] consists in
tice that in this case, we considered two extreme values measuring the time delay between the incident and the
of density, and the standard deviation does not change transmitted waves using a 48 kHz ultrasound wave.
noticeably. We can observe that for all shiv, the tortuosity is
considerably high in comparison to spherical particles
2.3 Open porosity for which this parameter ranges usually between 1 and
1.5 [21]. Here the tortuosity of hemp particles ranges
The open porosity φ is a function of the density between 1.5 and 3.5, and increases in a linear way with
knowing the frame density of the particles ρf rame [5].

961
23-27 April 2012, Nantes, France Proceedings of the Acoustics 2012 Nantes Conference

density. Such values can be attributed to the paral-

lelepiped shape of the particles and to the particles size
distribution. ρeq (ω) ≈ ρinter (ω) (1)
Modelling the evolution of tortuosity of such pack- Keq (ω) ≈ Kinter (ω) (2)
ings is quite complicated since no model enables to take
into account a parallelepiped shape. Some numerical The models decribed in the following enable to com-
predictions could be done as for example in [17] by de- pute these intrinsic properties from acoustical parame-
scribing the exact geometry of the particles: this is one ters such as the porosity and the resistivity of the inter-
of the outlooks of the authors. In [6], an empirical model particle network.
is used and a shape factor accounts for the shape and
size distribution of the particles. 3.1.1 Traditional approach for porous media
Some general models have been developped and can
2.6 Acoustical properties be used to describe porous materials such as granular,
The different shiv were tested in a B&K type 4106 fibrous or foam materials. Here, we use Johnson et al.
impedance tube having a diameter of 10 cm in the fre- model [27] for the visco-inertial effects and Zwikker and
quency range of 150 - 2000 Hz. The loose particles Kosten model [28] for the thermal effects.
were deposited in the tube which was held vertically This approach has been chosen first since it enables
and shaken to get the desired density with a constant with a reduced number of parameters to predict with
sample thickness of 5 cm. The diameter of the tube and accuracy the physical phenomena happening into the
the thickness of the samples have been chosen to ensure porous medium. Besides, it has been presented and ap-
a low ratio between the characteristic sizes of the par- plied successfully to hemp shiv and hemp concretes in
ticles and the tested sample, so that the properties of several papers [5, 6].
the granular mixes are not affected by the shape of the
containing tube. 3.1.2 Mono-sized spherical granular material ap-
The shiv were characterized using a three micro- proach
phone method [22] to enable the measurement of the
The second model investigated in this study is ded-
intrinsic properties of the material: the equivalent dy-
icated to granular media, which is more adapted to the
namic density ρeq , which accounts for visco-inertial ef-
porous nature of shiv. This kind of materials have been
fects and the equivalent bulk modulus Keq , which ac-
extensively studied by Boutin and Geindreau. In [9],
counts for thermal effects. These two frequency-dependent
these authors developed a model to predict the acous-
properties describe fully the material, and enable to
tical properties of mono-sized spherical aggregates. A
compute sound absorption α and sound transmission
very good agreement has been found between this model
loss (T L) as a function of the thickness. Furthermore,
and numerical predictions for aggregates having a spher-
viscous and thermal parameters can be determined from
ical shape. Besides, it has already been used for packings
these intrinsic properties, using the indirect character-
of porous grains [29].
ization relationships developed by Olny and Panneton
This model is used here to investigate the differences
[23, 24].
between shiv and packings of spherical particles having
same size, and to discuss the effect of the shape and
3 Modelling approaches distribution of particles on the acoustical properties.

3.1 Description of the models 3.1.3 Log-normal pore size distribution approach
Acoustical properties of porous materials can be pre- Finally, a third model has been considered. This
dicted, under rigid frame hypothesis, from their intrinsic model is also developed for granular material but takes
properties ρeq and Keq [25]. into account the distribution of pore size. A general
As explained in [6], shiv is a packing of porous parti- model could be used to describe arbitrary pore size dis-
cles. So a double porosity approach is used to take into tribution [30]. For the case of our shiv, we use the Padé
account the properties of the intra-particle pores and of approximants of this general model, presented in [30].
the inter-particle pores. In [6], the analysis of the exper- This approach has been chosen to take into account
imental data enabled to show that the contrast of per- the log-normal distribution of pore size and particle size
meability existing between these two networks of pores existing in shiv and to investigate the effect of such dis-
is big enough, so that, in the tested frequency range, we tribution.
are in the case, described in [26], of a very high con-
trat of permeability with no pressure diffusion between 3.2 Indirect characterization of the inter-
inter-particle and intra-particle pores. particle parameters
As a result, the intra-particle network does not take
part into te acoustical dissipation and the intrinsic prop- The inter-particle porosity φinter has been evaluated
erties of shiv ρeq and Keq can be described accurately by by adjusting the Zwikker - Kosten model [28] on the
computing the intrinsic properties of the inter-particle measurement of the real part of the bulk modulus. It
network ρinter and Kinter as described by Equations 1 can then be modelled as a function of both the density
and 2. of the shiv and the apparent density of the particles
ρparticle .

962
Proceedings of the Acoustics 2012 Nantes Conference 23-27 April 2012, Nantes, France

Then, in order to model the dynamic density and well. This is clearly not the case for the mono-sized
the bulk modulus using Johnson et al. and Zwikker granular model.
and Kosten models, inter-particle tortuosity and viscous Indeed, all models give satisfactory results for the
length, respectively α∞inter and Λinter have been indi- normalised bulk modulus K/P0 , but for the normalized
rectly characterized from the real and imaginary parts dynamic density ρ/ρ0 , the real part is a lot underesti-
(ρeq ) and (ρeq ) of the dynamic density measurement mated using the mono-sized spherical model. As a re-
using the analytical method presented in [24]. sult, the sound absorption α predicted with this model
The values of tortuosity measured with this method is far from the experimental one.
are close to the previous ones which were obtained using This can be attributed to the fact that the shape
the direct ultrasound technique. However, it is possible of the aggregates is not taken into account within this
to see differences, especially for shiv SE and for high two-parameter model. It has been shown that tortu-
densities. In these cases, the ultrasound tortuosity is osity is very high for shiv in comparison to spherical
greater than the inter-particle tortuosity. This differ- particles, that is why the real part of dynamic density
ence can be explained by the existence of the two scales of shiv, especially sensitive to tortuosity, is greater than
of porosity as shown in [31]. the dynamic density of spherical particles.
It is also possible to see that inter-particle viscous 1
length is affected by the shape and size distribution of
particles. In [6], it is shown that shiv have smaller vis- 0.5

α
cous length than mono-sized spherical aggreagates of 0
200 400 600 800 1000 1200 1400 1600 1800 2000
same inter-particle porosity and characterstic size, and 10 0
that the existing models are not suitable for this mate-
Re(ρ/ρ0)

Im(ρ/ρ0)
5 −5
rial.
0 −10
200 400 600 800 1000 200 400 600 800 1000
f (Hz) f (Hz)
3.3 Deduction of standard deviation of 3 1
Re(K/P0)

Im(K/P0)
2
pore size distribution and effective 1
0

pore size from the log-normal ap- 0

200 400 600 800 1000
−1
200 400 600 800 1000
proach 10
f (Hz) f (Hz)
TL (dB)

For the log-normal approach, the standard deviation 5

has first been evaluated so that the modeled sound ab-
sorption fits the measured one. Values range between 0
100 200 300 400 500 600 700 800 900 1000
f (Hz)
0.5 and 1. This range corresponds to the order of mag-
nitude found experimentally in Section 2.2.
Compared to other media, these standard deviations Figure 4: Modelling of shiv SA − ρ = 150kg.m−3 using
are pretty high. In [32], glass beads, coustone and foam the general model (...), the mono-sized spherical model
have been modeled using the log-normal approach and (.−), and the log-normal model (−−), compared to
the characterized standard deviation range between 0.22 data (−).
and 0.44. In our case the shape and size distribution of
hemp particles is responsible for this big increase of the
standard deviation of pore size compared to glass beads.
For a media having identical circular pores, the re- 4 Effect of the shape and distri-
sistivity σ can be evaluated from the radius of the pores bution of particles on the acous-
Rpores using Equation 3 [12, 33], where lef f = Rpores
for circular pores. tical properties
8α∞inter η It is pretty hard to distinguish the effect of the shape
σ= 2 (3)
φinter lef f
and the effect of the particle size distribution on the
acoustical properties. This is partially due to the fact
In the present case with a log-normal distribution of that both interact on the pore size distribution.
pore size, the effective pore size lef f takes into account In Figure 4, the experimental sound absorption of
the standard deviation of the pore-size stdpores and we
 2  2(std 2 shiv is compared to the modelling in the case of a pack-
pores log(2))
have: lef f = Rpores e . ing of spheres having same characteristic size and same
The characteristic size lef f of the pores decreases inter-particle porosity. This is a first way to notify the
with the density. We also notice that lef f is very close effect of shape and size distribution of hemp particles
from Λinter which can be interpreted as an estimate of on the acoustical properties.
the size of interconnection between pores [27]. We constat for the different cases that the sound ab-
sorption of shiv is a lot better than mono-sized spherical
3.4 Ability of the models to describe shiv granular media at low frequencies, and that it is a bit
lower or equal to it at higher freqencies. This is ex-
The predictions of the three models are compared plained by the increased tortuosity, that moves the first
to data in Figure 4 for the example of shiv SA − ρ = absorption peak towards the lower frequencies.
150kg.m−3 . It appears that the general model and the Concerning sound insulation, the transmission loss of
log-normal model predict the acoustical properties very shiv is greater than the transmission loss of the equiva-

963
23-27 April 2012, Nantes, France Proceedings of the Acoustics 2012 Nantes Conference

lent mono-sized spherical granular media. [11] T. Yamamoto and A. Turgut. Acoustic wave propagation through
porous media wth arbitrary pore size distributions. Journal of
So, this comparison shows that for a same inter- the Acoustical Society of America, 83 (5):1744–1751, 1998.
particle porosity and characteristic size, shiv provides
[12] K-V. Horoshenkov. Control of traffic noise in city streets. PhD
a better transmission loss in the whole frequency range thesis, University of Bradford, 1996.
and better sound absorption for the low frequencies.
[13] M-A. Biot. Theory of propagation of elastic waves in a fluid-
This is interesting in the cases where shiv is used in saturated porous solid. I.Low-frequency range. Journal of the
a loose way in buildings, for instance in attics. Acoustical Society of America, 28:168–178, 1956.

[14] M-A. Biot. Theory of propagation of elastic waves in a fluid-

saturated porous solid. II.High-frequency range. Journal of the
5 Conclusion Acoustical Society of America, 28:179–191, 1956.

[15] A-B. Yu and R-P. Zou. Prediction of the porosity of particle

In this study, the acoustical properties of hemp par- mixtures. KONA Powder and particle, 16:67–81, 1998.
ticles, eco-friendly aggregates mainly dedicated to build-
[16] K. Attenborough. On the acoustic slow wave in air-filled granular
ing applications, are investigated. The acoustical mod- media. Journal of the Acoustical Society of America, 81(1):93–
elling of this material leads to several difficulties, such 102, 1987.

as the existence of pores at multiple scales, the size- [17] I. Malinouskaya, V-V. Mourzenko, J-F. Thovert, and P-M. Adler.
distribution and the parallelepiped shape of the parti- Random packings of spiky particles: Geometry and transport
properties. Physical Review E, 80, 011304:16, 2009.
cles.
It is first highlighted in this paper that particle-size [18] I. Ceyte. Béton de chanvre, définition des caractéristiques
mécaniques de la chènevotte, Travail de Fin d’Études. ENTPE,
and pore-size distributions can be perfectly described page 155, 2008.
by log-normal distributions. Besides, a clear influence
[19] P. Leclaire, O. Umnova, K.-V. Horoshenkov, and L. Maillet.
of shape and size-distribution of particles is observed on Porosity measurement by comparison of air volumes. Review
the acoustical parameters. For instance, the tortuosity of scientific instruments, 74 (3):1366–1370, 2003.
of shiv is very high compared to monosized spherical [20] J-F. Allard, B. Castagnede, M. Henry, and W. Lauriks. Evalua-
aggregates. tion of tortuosity in acoustic porous materials saturated by air.
Review of Scientific Instruments, 65:754–755, 1994.
Three modelling appproaches are finally compared
and enable to show the influence of shape and size- [21] C. Boutin and C. Geindreau. Estimates and bounds of dynamic
permeability of granular media. Journal of the Acoustical Soci-
distribution of particles on the acoustical properties. For ety of America, 124(6):3576–3593, 2008.
shiv, sound absorption is shifted towards the lower fre-
[22] T. Iwase, Y. Izumi, and R. Kawabata. A new measuring method
quencies while transmission loss is higher than for mono- for sound propagation constant by using sound tube without
sized spherical aggregates. any air spaces back of a test material. In Internoise 98,
Christchurch, New Zealand, page 4, 1998.

[23] X. Olny and R. Panneton. Acoustical determination of the pa-

References rameters governing thermal dissipation in porous media. Journal
of the Acoustical Society of America, 123(2):814–824, 2008.
[1] M-P. Boutin, C. Flamin, S. Quinton, and G. Gosse. Analyse du
cycle de vie : Compounds thermoplastiques chargés fibres de [24] R. Panneton and X. Olny. Acoustical determination of the pa-
chanvre et Mur en béton de chanvre banché sur ossature bois. rameters governing viscous dissipation in porous media. Journal
Rapport d’étude INRA Lille, réf MAP 04 B1 0501. 2005. of the Acoustical Society of America, 119(4):2027–2040, 2006.
[2] L. Arnaud and E. Gourlay. Experimental study of parameters in- [25] J-F. Allard. Propagation of sound in porous media. Applied
fluencing mechanical properties of hemp concretes. Construction Science, 1993.
and Building Materials, 28:50�56, 2011.
[26] X. Olny and C. Boutin. Acoustic wave propagation in double
[3] V. Cerezo. Propriétés mécaniques, thermiques et acoustiques porosity media. Journal of the Acoustical Society of America,
d’un matériau ďż˝ base de particules végétales: approche 114(1):73–89, 2003.
expérimentale et modélisation théorique. PhD thesis, Ecole doc-
torale MEGA, Lyon, 2005. [27] D-L. Johnson, J. Koplik, and R. Dashen. Theory of dynamic per-
meability and tortuosity in fluid-saturated porous media. Fluid
[4] D. Samri. Analyse physique et caractérisation hygrothermique Mechanics, 176:379–402, 1987.
des matériaux de construction: approche expérimentale et
modélisation numérique. PhD thesis, Ecole doctorale MEGA, [28] C. Zwikker, J. Van Den Eijk, and C-W. Kosten. Absorption of
Lyon, 2008. sound by porous materials. part III. Physica, VIII:1094–1101,
1941.
[5] P. Glé, E. Gourdon, and L. Arnaud. Acoustical properties of ma-
terials made of vegetable particles with several scales of porosity. [29] R. Venegas and O. Umnova. Acoustical properties of double
Applied Acoustics, 72:249–259, 2011. porosity granular materials. Journal of the Acoustical Society
of America, 130 (5):2765–2776, 2011.
[6] P. Glé, E. Gourdon, and L. Arnaud. Characterization and mod-
elling of the acoustical properties of hemp shiv and hemp con- [30] K-V. Horoshenkov, K. Attenborough, and S-N. Chandler-Wilde.
crete. Proceedings of SAPEM 2011, Ferrara, page 2, 2011. Padé approximants for the acoustical properties of rigid frame
porous media with pore size distributions. Journal of the Acous-
[7] C. Garcia-Jaldon, D. Dupeyre, and M.R. Vignon. Fibres tical Society of America, 104(3):1198–1209, 2007.
from semi-retted hemp bundles by steam explosion treatment.
Biomass and Bioenergy, 14 (3):251–260, 1998. [31] M. Barrande, R. Bouchet, and R.Denoyel. Tortuosity of porous
particles. Analytical Chemistry, 79(23):9115–9121, 2007.
[8] M. R. Vignon, C. Garcia-Jaldon, and D. Dupeyre. Steam ex-
plosion of woody hemp chènevotte. Int. J. Biol. Macromol., 17 [32] K-V. Horoshenkov and M-J. Swift. The acoustic properties of
(6):395–404, 1995. granular materials with pore size distribtion close to log-normal.
Journal of the Acoustical Society of America, 110 (5):2371–
[9] C. Boutin and C. Geindreau. Periodic homogenization and con- 2378, 2001.
sistent estimates of transport parameters through sphere and
polyhedron packings in the whole porosity range. Physical Re- [33] G. Pispola, K-V. Horoshenkov, and A. Khan. Comparison of
view E, 82-036313:18, 2010. two modeling aproaches for highly heterogeneous porous media.
Journal of the Acoustical Society of America, 121(2):961–966,
[10] O. Umnova, K. Attenborough, E. Standley, and A. Cummings. 2007.
Behavior of rigid-porous layers at high levels of continuous acous-
tic excitation: Theory and experiment. Journal of the Acoustical
Society of America, 114(3):1346–1356, 2003.

964