Moisture Content Influence on the Thermal Conductivity and Diffusivity

al

ckssaa

b

h i g h l i g h t s

he therngs incwith wm corred on th

experimental results show that the lightening of concrete by wood shavingslation capacity by decreasing the thermal conductivity and diffusivity; how-

The composite woodconcrete has been studied a lot these last

wood bers are naturally degradable [2], which is not negligible inthe current context of waste limitation. The reinforcement of con-crete by wood bers gives a composite material which can be usedin many applications such as oor formwork, suspended ceilings,screeds and interior masonry blocks .

Several studies have focused on the use of wood as ash in con-crete [35]. Based on its physical, chemical, and microstructural

products, roller-compacted concrete pavements (RCCPs), materialst al. [6] study theexaminincture. The

are encouraging and indicate the feasibility of producing aweight concrete. Bederina and Quneudec [7], Ziregue [8that the addition of wood shavings in concrete improves itmal insulation performance, while decreasing its compressivestrength. However, it is possible to ensure a compromise betweenthe compressive strength and the thermal conductivity for obtain-ing a structural insulation concrete.

Nevertheless, this material can present a high hygroscopicbehavior due to the strong absorption of wood particles. When thismaterial has being exposed to the conditions of temperature (T)and relative humidity (RH), which are variable between the interior

Corresponding author. Tel.: +212 0667540261.

Construction and Building Materials 48 (2013) 104115

Contents lists available at

B

evE-mail address: [email protected] (D. Taoukil).years because wood bers have many advantages: low cost,healthier production processes for the production of compositeswith different forms, the renewal and the recycling [1]. In addition,

for road base, and blended cements. Coatanlem edurability of wood chipping concrete mixture bypressive strength, exural strength and microstru0950-0618/$ - see front matter 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.conbuildmat.2013.06.067g com-resultslight-

] shows ther-Keywords:Wood concreteThermal conductivityThermal diffusivityMoisture contentHygrothermal effect

ever these proprieties are strongly dependent on water content. The thermal conductivity increasesrapidly with water content. Its experimental evolution with water content was conrmed by the com-parison with three theoretical models. The values of the thermal diffusivity depend on the used count-ing model. The results stemming from three most used models are compared between them, and theyshow that in general the thermal diffusivity presents a maximum corresponding to a water contentvalue Wm.

2013 Elsevier Ltd. All rights reserved.

1. Introduction properties, the authors reported that wood ash has a signicant po-tential for use in low and medium strength concrete, masonryAvailable online 23 July 2013been examined herein. Theincreases the thermal insu The inuence of moisture content on t Lightening the concrete by wood shavi Thermal conductivity increases rapidly Thermal diffusivity presents a maximu The values of thermal diffusivity depen

a r t i c l e i n f o

Article history:Received 29 November 2012Received in revised form 30 May 2013Accepted 17 June 2013mal proprieties of woodconcrete is studied.reases its thermal insulation capacity.ater content.sponding to a water content value Wm.e used counting model.

a b s t r a c t

The aim of the work reported in this paper is to determine the inuence of moisture content on thethermal proprieties of woodconcrete composite, i.e. thermal conductivity and thermal diffusivity.The wood shavings have been incorporated, without any preliminary treatment, into a sandcementmixture. Five formulations containing different percentage of shavings have been prepared and haveHTW Berlin, University of Applied Sciences, Ingenieurwissenschaften I, Berlin, GermanyMoisture content inuence on the thermof woodconcrete composite

Driss Taoukil a,, Abdelmajid El bouardi a, Friedrich Sia Energetic Laboratory, Physics Department, Faculty of Sciences of Tetuan, Abdelmalek E

Construction and

journal homepage: www.elsconductivity and diffusivity

b, Abdelaziz Mimet a, Hassan Ezbakhe a, Taib Ajzoul a

di University, Morocco

SciVerse ScienceDirect

uilding Materials

ier .com/locate /conbui ldmat

and the exterior of the building, heat and moisture transfers occurwithin the medium. An important interaction develops betweenthe heat and the moisture transfers and the different transportproperties [9]. Indeed, the thermal, phonic and mechanical proper-ties of the material are modied. It is thus crucial to quantify thevariation of these properties with temperature and moisture con-tent of the sample. But, this is not considered in most studies treat-ing the use of this type of material in building. In this context, thisstudy has been conducted to measure the effect of moisture con-tent on the thermal conductivity and diffusivity of the woodconcrete composite.

2. Materials

2.1. Basic materials

The wood shavings used are a sawmill waste (Fig. 1). They correspond to thespecies generally used in the carpentry work in Morocco [10]. The granulometricanalysis by sieving, established by a method identical to that proposed by theFrench standard NF P18-560 [11], demonstrates that the used wood shavings havea particle size between 8 mm and 20 mm. The sand is sea one of granulometry 0/5and of bulk density on the order of 1500 kg/m3. This sand is used without any wash-ing procedure. Indeed, the test of sand equivalent made according to the standardNF EN 933-8 [12] gives the value 81%, which shows that it is a clean sand with a

The constituents are mixed in a mixer at slow speed on the order of 50 r min1,

models used in this work are as follows:

3.1. Krischer and Kroll model [21]

The Krischer and Kroll model is a combination between the se-rial and the parallel models. The authors suppose that the volumefraction h of layers oriented perpendicular to the direction of heatow is arranged in series with the complementary fraction (1h) oflayers oriented parallel to the direction of heat ow. Firstly, theauthors suppose that the uid phase consists only of gas (air ingeneral). Afterwards, they extend their model for a uid phase con-sisting of gas and liquid (water in general). The apparent thermalconductivity is given by a harmonic mean of the series and parallelmodels:

D. Taoukil et al. / Construction and Buildin order to obtain a good homogenization of the components. The durations of thedifferent phases of mixing must be sufciently long to allow a good homogenizationand short enough to avoid leaving too much water to evaporate in ambient air.Thus, three phases are distinguished:

Mixing cement and sand: 3 min. Adding pre-wetted wood shavings: 3 min. Adding mixing water and mixing: 5 min.low percentage of clay nes, suiting perfectly for high quality concretes [13]. Thecement used is Portland cement CPJ 35 equivalent to CEM II 22.5 and whose tech-nical characteristics are in accordance with the Moroccan standard NM 10.01.004[14]. The used mixing water is a tap water.

The concretes prepared are based on a mass ratio of 2/3 of sand and 1/3 of ce-ment. The specimens are made with a mass ratio of water to cement (W/C) of 0.6.The study was conducted on ve formulations containing different proportions ofwood shavings.

2.2. Implementation of fresh concreteFig. 1. General aspect of the used wood shavings.2.3. Preparation of specimens

The homogenized mixture is then introduced into parallelepiped molds; twodimensions of molds were made according to the type of tests to be realized. As re-gards the thermal aspects, the molds of dimension (27 27 2 cm3) were used. Formechanical strength, the molds (16 4 4 cm3) were used. These different sizesare related to measuring devices (thermal box, mechanical press) whose dimen-sions are imposed. The use of the same manufacturing process, regardless of themold, allows to work on the same material.

The specimens are preserved before and after turning out into the room test at acontrolled temperature and humidity (Ta = 20 C, RH = 60%). These conditions corre-spond to a standard climate and allow to reproduce real conditions of the use of thematerial.

2.4. Formulations

The elaborate formulations, containing various rations in wood shavings, arepresented with their physico-mechanical properties in Table 1.

The density and the porosity are strongly inuenced by the addition of woodshavings. Thus, a progressive decrease in density and an increase in porosity are ob-served. Indeed, the sand concrete, due to its small granularity, is lighter than ordin-ary concrete. The introduction of wood shavings reduces it more. The higher theproportion of wood shavings used the lighter the sand concrete will be [7,16,17].This lightness is due, on the one hand, to the porous structure of wood, on the otherhand, to an additional porosity led in the matrix when the water absorbed by thewood shavings evaporates during the drying.

The increase in shavings content reduces the mechanical strength of concrete.This decrease is due, on the one hand, to the low mechanical strength of the inclu-sions, and on the other hand, to the increase of porosity.

The water absorption capacity is very sensitive to the addition of wood shav-ings. Thus, there is a signicant increase in Wsat. Indeed, wood particles have a verymarked brous aspect. The canals which lead the sap give this material a highporosity communicating via orices. This leads to a strong hygroscopic character[18]. It is noted that the concrete sand, itself, is characterized by an absorptioncapacity higher than that of ordinary concrete [19], and that the addition of woodincreases it further. Similar results were obtained for hemp-concrete [20].

3. Thermal conductivity models

Many models have been proposed in the literature to predictthe effective thermal conductivity of three-phase mixtures. The

Table 1Principal physico-mechanical properties of samples.

Sample Mix0 Mix1 Mix2 Mix3 Mix4

Mass ratio of wood shavings tocement (%)

0 6 15 24 30

Dry density (kg/m3) 2142 1914 1779 1547 1495Porosity (%) 21 24 31 44 55Compressive strengths (MPa)a 27 16 13 9 7Volumetric water content at the

saturation Wsat (%)20 22 29 42 53

a Compressive strength measured according to the European standard EN 196-1[15].

ing Materials 48 (2013) 104115 105k 11hk==

hk?1

cm

ting resistance R

BoxB Sample E

Cryostat K100 cm

45 cm

Exchanger H

Building Materials 48 (2013) 104115Cables links

Measurement and control console

Terminal block 35

35 cm

HeaCentral data acquisition

Computer

Cold enclosure A

106 D. Taoukil et al. / Construction andk== 1 PTks PT Wka Wkl : Parallel model 2

k? 11PTks

PTWka Wkl: Series model 3

With ks is the thermal conductivity of the solid phase; ka the ther-mal conductivity of air; kl the thermal conductivity of liquid water;PT the total porosity and W volumetric water content.

Many authors use this model to predict effective thermal con-ductivity in a wetted porous media: Carson [22] in porous food,Bal et al. [23] in laterite based bricks, etc. The effective thermalconductivity predicted by the model was in good agreement withthe experimental data.

According to Krischer and Kroll [21], the thermal conductivityof the solid matrix ks and the parameter h are determined by solv-ing a system of two unknowns:

hdryks 1=kdry 1=k==1=k? 1=k== : Dry stat W 0 4

200 cm

Fig. 2. Experimental setup of the box method.

y = 0.4196e0.048xR = 0.9852

y = 0.3426e0.0367xR = 0.991

y = 0.2897e0.0271xR = 0.9975 y = 0.2478e

0.0218x

R = 0.9965

y = 0.2559e0.0155xR = 0.988

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

0 5 10 15 20 25 30 35 40 45 50 55

Ther

mal

con

duct

ivity

(W/m

C)

Volumetric water content (%)

Mix0 Mix1 Mix2

Mix3 Mix4

Fig. 3. Thermal conductivity as function of water content of the various samples.

Table 2Report of the thermal conductivities ksat/kdry of the studied samples.

Sample Mix0 Mix1 Mix2 Mix3 Mix4

ksat/kd 2.76 2.31 2.20 2.54 2.44

Table 3Parameters ks and h of different samples.

Sample Krischer and Krollmodel

Chaudhary andBhandari model

Woodside and Messmermodel

ks (W/m C)

h ks (W/m C)

h ks (W/m C)

Mix0 1.449 0.200 1.65 0.476 0.862Mix1 1.071 0.193 1.375 0.480 0.779Mix2 0.734 0.140 0.811 0.342 0.845Mix3 0.699 0.110 0.765 0.295 1.367Mix4 0.643 0.050 0.71 0.173 3.529

0.8

ndu

Experimental Points

Build0.6

0.7

mal

co0.9

1

1.1

1.2

ctiv

ity (W

/m C

)

Krischer and Kroll ModelChaudhary and Bhandary ModelWoodside and Messmer Model1.3

D. Taoukil et al. / Construction andhsatks 1=ksat 1=k==1=k? 1=k== : Saturation stat W PT 5

hdry(ks) the value of h at the dry state and hsat(ks) the value of h at thesaturation state. kdry and ksat are the thermal conductivities mea-sured respectively at the dry state and at the saturation state.

The system of equations in h and ks is difcult to solve. In thisstudy, the determination of these unknowns is done using New-tons method, which consists in determining by successive ap-proaches the value of ks for which:

hdryks hsatks 6The values used for the thermal conductivity of air and that of

liquid water are: ka = 0.026 W/m C and kl = 0.58 W/m C [24].

0.3

0.4

0.5

0 2 4 6 8 10

Ther

Volumetric w

Fig. 4. Comparison between the experimental evolution of the thermal conduc

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 2 4 6 8 10

Ther

mal

con

duct

ivity

(W/m

C)

Experimental PointsKrischer and Kroll ModelChaudhary and Bhandary ModelWoodside and Messmer Model

Volumetric wa

Fig. 5. Comparison between the experimental evolution of the thermal conducing Materials 48 (2013) 104115 1073.2. Chaudhary and Bhandari model [25]

The apparent thermal conductivity, according to the Chaudharyand Bhandari model, is estimated by a geometric mean of the ser-ies and parallel models:

k kh?k1h== 7Experimental results of the effective thermal conductivity ob-

tained by Sugawara [26] on calcareous sandstone partially lledwith water are compared with values calculated by this model.The calculated values agree well with the experimental results.

In this study, it is proposed that the parameters ks and h, relativeto this model, is determined by following the same approach asthat in the case of the Krischer and Kroll model. The system ofequations in h and ks will be given by:

12 14 16 18 20 22

ater content (%)

tivity with water content and that simulated by theoretical models: Mix0.

12 14 16 18 20 22

ter content (%)


ity

Build0.4

0.5

rmal

con

duct

iv0.6

0.7

0.8

(W/m

C)


108 D. Taoukil et al. / Construction andhdryks Lnkdry=k==Lnk?=k== 8

hsatks Lnksat=k==Lnk?=k== 9

3.3. Woodside and Messmer model [27]

The authors calculated the values of thermal conductivity ofsaturated rocks (two-phase system), using a weighted geometricmean relationship of the two phases. The model is in good agree-ment with the experimental results. Chaudhary and Bhandari[25] have generalized this model for three-phase porous medium.The thermal conductivity is given by:

k k1PT s kPTWa kWl 10

0.2

0.3

0 2 4 6 8 10 12 14

The

Volumetric wa


0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5 10 15 20

Ther

mal

con

duct

ivity

(W/m

C)

Volumetric wa


Fig. 7. Comparison between the experimental evolution of the thermal conducing Materials 48 (2013) 104115In this study, it is proposed that the parameter ks, relative tothis model, is determined through the value of thermal conductiv-ity in the dry state. It has as expression:

ks Ln kdry=k

PTWa kWl

1 PT 11

4. Test methods

The technique used to measure the thermal conductivity iscalled Boxes method (Fig. 2). It presents the advantage of a verysimple implementation and the measurement precision is compa-rable to that obtained by conventional methods (hot wire, hot disk,etc.). This technique has been developed in the laboratory of Ther-mal and Solar Studies of the Claude Bernard University-Lyon I in

16 18 20 22 24 26 28 30

ter content (%)


25 30 35 40 45

ter content (%)


Build0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

Ther

mal

con

duct

ivity

(W/m

C)


D. Taoukil et al. / Construction andFrance [2830]. The Boxes method allows to measure thermal con-ductivity of materials tested in permanent regime by realizing anenergy balance of the system. The measuring principle is basedon achieving a permanent unidirectional heat ow through thesample, supposed homogeneous and without internal generationof heat, by creating a temperature gradient between its two faces.Indeed, the sample of parallelepiped form is placed between twoenclosures, one of which is heated and the other is cooled, in sucha way that the lateral ows are negligible. Once the permanent re-gime is established, the thermal conductivity is given by the Fou-rier law:

k qeSDT

12

q is the heat ow through the sample (W); e the thickness of thesample (m); S the area of the faces of the sample perpendicular tothe ow lines (m2) and DT temperature difference between thetwo faces of the sample.

0

0.1

0 5 10 15 20 25

Volumetric wa


1

1.5

2

2.5

3

3.5

4

4.5

0 5 10 15 20 25

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

Volumetric wa

Fig. 9. Thermal diffusivity as function of water coning Materials 48 (2013) 104115 109The thermal diffusivity was measured in the transition regime.The Flash method was used [31]. Its Principle is to produce a heatpulse of short duration t0 (Dirac pulse) on one face of the sampleand recording the temperature evolution on the other face, notirradiated, as function of time. The exploitation of the experimentalthermogram obtained by this manner and the theoretical thermo-gram stemming from the modeling allows to estimate the thermaldiffusivity a. In this work, existing theoretical models in the litera-ture are used to identify the thermal diffusivity by means of theFlash method.

For measuring the moisture inuence on the thermal conduc-tivity and diffusivity, the used variable is the volumetric water con-tent W. This is done by practising successive weightings of thematerial since the partial saturation state until the dry state. Thus,the rst wet measure is obtained after immersion of the material,for a few days, in water until its mass remains constant for 24 h(mass variation

Build1.5

2

2.5

3

3.5

4

4.5

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

110 D. Taoukil et al. / Construction andoperations are done in the ambient air with the measure of the cor-responding values of thermal conductivity and diffusivity. Then,the drying is continued in a ventilated and regulated oven at80 C until the dry state. This is for obtaining certain continuityof the curves which give the variation of k and a as function of vol-umetric water content.

5. Results and discussion

5.1. Thermal conductivity

The evolution of the thermal conductivity as function of volu-metric water content of the various samples is illustrated onFig. 3. Two remarks are to be underlined:

The addition of wood shavings in the concrete reduces the ther-mal conductivity of the composite and consequently increasesits thermal insulation capacity. This behavior is related to the

10 5 10 15 20 25

Volumetric wa

Fig. 10. Thermal diffusivity as function of water cont

1

1.5

2

2.5

3

3.5

4

0 5 10 15 20 25

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

Volumetric wa

Fig. 11. Thermal diffusivity as function of water coMix0 Mix1 Mix2 Mix3 Mix4

ing Materials 48 (2013) 104115insulating character of wood shavings, which have a low ther-mal conductivity. The thermal conductivity of a materialdepends on those of aggregates which constitute it. The lowerthe thermal conductivity of inclusions, the more the materialis insulating. In addition, the increase in porosity decreasesthe density of the composite and consequently its thermal con-ductivity. This dependence is always veried on materials withmineral matrix and vegetable bers according to Alrim et al.[32] which used a clayey matrix or, according to Khedari et al.[33] which used matrix based on portland cement. The valuesobtained for thermal conductivity make that the studied com-posite is a competitive material with lightweight insulatingconcretes. However, it is shown that increasing the heat insula-tion capacity is achieved at the expense of compressivestrength. In fact, the compressive strength decreases with theincorporation of wood shavings (p. 2.4). The knowledge of theresistance level sufcient for a given use is therefore crucialfor the optimal exploitation of the insulation quality. Anyway,

30 35 40 45 50 55

ter content (%)

ent for the different samples: Digiovanni model.

30 35 40 45 50 55

ter content (%)

Mix0 Mix1 Mix2 Mix3 Mix4

ntent for the different samples: Yezou model.

66x +

0,132

BuildYezou: y = -4E-05x4 + 0,0015x3 - 0,0212x2 + 0,14R2 = 0,9794

Digiovanni: y = -3E-05x4 + 0,0011x3 - 0,0183x2 + R2 = 0,996

3

3.2

3.4

3.6

3.8

4

4.2

4.4

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

D. Taoukil et al. / Construction andthe compressive strength values obtained in this study are com-patible with the use of the elaborated materials as load-bearinginsulating concrete, according to the functional classication ofRILEM [34]. Then, this composite should nd it an importantuse in constructions (oor formwork, suspended ceilings, scre-eds, interior masonry blocks. . .).

The thermal conductivity increases rapidly with water content.Table 2 gives, for every sample, the report thermal conductivityin the saturation state ksat than that in the dry state kdry. Itappears that the thermal conductivity measured at the satura-tion state is two to three times greater than that measured inthe dry state. Numerous works [21,35,36] bring to light thisdependence (conductivity/water content) on other buildingmaterials.

The results of the calculation of the parameters ks and h relativeto the three proposed simulation models are included in Table 3.The fraction h of the porous medium arranged perpendicular to

Parker: y = -2E-05x4 + 0,0009x3 - 0,0159x2 + 0,1359xR2 = 0,9925

2.80 2 4 6 8 10

Volumetric wat

Fig. 12. Comparison between the thermal diffusivity giv

Yezou: y = -8E-07x4 + 0,0003x3 - 0,0092x2 + 0R2 = 0,8381

Digiovanni: y = -2E-05x4 + 0,0009x3 - 0,0192x2R2 = 0,9557

Parker: y = -2E-05x4 + 0,0009x3 - 0,0164x2 + 0,R2 = 0,98912.5

2.6

2.7

2.8

2.9

3

3.1

3.2

3.3

3.4

0 2 4 6 8 10

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

Volumetric w

Fig. 13. Comparison between the thermal diffusivity giv 3,4171

1x + 3,6616

Yezou ModelDegiovanni ModelParker Model

ing Materials 48 (2013) 104115 111the direction of heat ow is very low, which allowed to deduct thatthe layers of the studied materials are mostly arranged in parallelto the direction of heat ow. The values of ks and h calculatedthrough Chaudhary and Bhandari model are superior to thosestemming from the Krischer and Kroll model. There is generally astrong deviation of the values of ks determined with Woodsideand Messmer model compared to those determined with the othertwo models.

The comparison between the experimental evolution of thethermal conductivity with water content and that simulated bythe proposed theoretical models is illustrated on Figs. 48. TheKrischer and Kroll model approach better measurement results.The mean relative error between experimental results and thosegiven by this model is between 2.4% (Mix1) and 4.9% (Mix3). TheChaudhary and Bhandari model is still acceptable, with a mean rel-ative error between 5.57% (Mix1) and 7.6% (Mix3). The good agree-ment of these two models with the experimental points is due tothe effect that the domain, bounded by the assemblies of phases

+ 3,6982

12 14 16 18 20 22er content (%)

en by Parker, Digiovanni and Yezou models: Mix0.

,0899x + 2,6064

+ 0,1329x + 2,7928

0994x + 3,1364

12 14 16 18 20 22

ater content (%)



YD

3 Degiovanni Model

Build2.2

2.3

2.4

2.5

2.6

2.7

2.8

2.9

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

Parker Model3.1Yezou Model

112 D. Taoukil et al. / Construction andin series and in parallel, includes necessarily all properties ofundamaged real materials, whatever the morphology and the dis-tribution of its phases. Indeed, the series model represents the low-er bound of the apparent thermal conductivity of the medium, andthe parallel model represents the upper bound. The values simu-lated by the Woodside and Messmer model are relatively far awayfrom experimental measurements except near the dry state. Theagreement near the dry state is expected because this model hasbeen found to hold good for two-phase systems [27]. The disagree-ment outside the dry state region can be attributed to the failure totake account in this model of the structure and distribution of thephases in the medium.

5.2. Thermal diffusivity

The thermal diffusivity a characterizes the velocity at which theheat propagates in a material. The lower its value, the more the

P

2

2.1

0 2 4 6 8 10 12 14Volumetric w

Fig. 14. Comparison between the thermal diffusivity giv

Yezou:

Digiova

Parker

1.5

1.6

1.7

1.8

1.9

2

2.1

2.2

2.3

0 5 10 15 20

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

Volumetric w


Fig. 15. Comparison between the thermal diffusivity givezou: y = 1E-05x4 - 0,0006x3 + 0,0101x2 - 0,0073x + 2,1005R2 = 0,9943

igiovanni: y = 6E-06x4 - 0,0004x3 + 0,0058x2 + 0,0071x + 2,3713

ing Materials 48 (2013) 104115heat takes the time to pass through a material. Thus, the insulationcapacity of a material depends not only on the value of k but alsoon the heat transfer velocity translated by a.

The study of the thermal diffusivity as function of moisture con-tent is carried out in the same experimental conditions and withthe same water contents than in the case of thermal conductivity.A comparison between the results obtained by three countingmodels of experimental thermograms is also conducted:

Parker model [31]: the thermal diffusivity is given by:

a 0:139 e2

t1=213

Degiovanni model [37]: the expressions of thermal diffusivityare given as follows:

a2=3 e2 1:150=t5=6 1:250 t2=3t25=6

14

R2 = 0,9871

arker: y = 4E-06x4 - 0,0002x3 + 0,0027x2 + 0,0267x + 2,5292R2 = 0,946

16 18 20 22 24 26 28 30ater content (%)


y = 1E-06x4 - 0,0002x3 + 0,0041x2 - 0,0094x + 1,5456R2 = 0,9934

nni: y = 1E-06x4 - 0,0001x3 + 0,0028x2 + 0,0027x + 1,6546R2 = 0,998

: y = -1E-08x4 + 1E-05x3 - 0,0013x2 + 0,0447x + 1,7751R2 = 0,9719

25 30 35 40 45

ater content (%)


1=2t25=6

a5=6 t5=6 to=2 0:713 t5=6 to=2 1:812 t5=6 to=2 1:037 17

07x4

3E-0

-07x

5

ic w

y giv

Builda e2

0:4032 t1=2 to=2 2

0:1103 t1=2 to=2

0:2027" #a1=3 e2 0:617 t5=6 0:862 t1=3t25=6

16

where ti/j is the time corresponds to the ratio i/j of the maximumtemperature.

Yezou model [38]: the terms of thermal diffusivity are given by:

e2 t1=2 to=2 2 t1=2 to=2 " #a e2 0:761 t5=6 0:926 t1=2 15

Yezou: y = 3E-

Digiovanni: y =

Parker: y = -3E

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

0 5 10 15 20 2

Ther

mal

diff

usiv

ity (.

10-7

m2 /s

)

Volumetr


Fig. 16. Comparison between the thermal diffusivit

D. Taoukil et al. / Construction and1=2 t1=2 to=2 t5=6 to=2 t5=6 to=218

For each model, the value of the thermal diffusivity is calculatedas being the arithmetic mean of the given expressions.

Figs. 911 recapitulate the curves of the variation of the thermaldiffusivity with water content for the ve simples, using respec-tively Parker, Digiovanni and Yezou models. The Figures conrmsthat the incorporation of wood shavings in concrete decreases itsthermal diffusivity. This result is expected and it is in agreementwith the results found for thermal conductivity. The lightening ofconcrete by wood shavings has a positive effect on its thermalinsulation capacity.

It is noted that both ascending and descending trends of therelationship between diffusivity and water content have been re-ported. In fact, a maximum of thermal diffusivity is observed forthe values of water content Wm, about 6% for Mix0, 5% for Mix1,17% for Mix2, 26% for Mix3 and 19% for Mix4. This phenomenonhas been observed on other materials [35,39,40] and can be ex-plained by the combined effects of thermal conductivity and volu-metric heat capacity qc according to the equation:

a kqc

19

In the moisture range [0,Wm], the increase of the thermal diffu-sivity with moisture content can be explained by the fact that thevalue of k increases faster than that of qc. In the moisture range[Wm, Wsat], the decrease of the thermal diffusivity with moisturecontent can be explained by the fact that the value of k increasesslower than that of qc.

The thermal diffusivity stemming from the three aforesaidmodels are clearly far away from each other (Figs. 1216). This isnormal, because of the difference between the basic hypotheses re-tained in each model. Indeed, for the development of their model,Parker et al. used cylindrical samples, and suppose that the follow-ing conditions are veried: homogeneous, isotropic and opaquesample, constant thermophysical properties, perfectly isolatedsample, very short photothermal pulse uniformly distributed onthe front face of the sample, unidirectional heat transfer. This mod-el, of a simple and practical application, neglects several aspects ofthe physical reality. Indeed, the heat loss on the different faces ofthe sample is never null and the thermal pulse is never of negligi-

- 3E-05x3 + 0,0008x2 + 0,012x + 1,1604R2 = 0,8741

7x4 - 4E-05x3 + 0,0008x2 + 0,0152x + 1,2201R2 = 0,8951

4 + 3E-05x3 - 0,0016x2 + 0,041x + 1,4674R2 = 0,9419

30 35 40 45 50 55

ater content (%)

en by Parker, Digiovanni and Yezou models: Mix4.ing Materials 48 (2013) 104115 113ble duration. Degiovanni model has the advantage to take into ac-count thermal losses on all faces of the cylindrical sample withoutthe need to evaluate them. For the Yezou model, the author usedparallelepiped samples. It takes into account the irradiation timet0 and the heat losses on the front and back faces of the sample.Lateral losses are neglected. However, it is possible to estimate thatYezou model reects well the thermal behavior of the studiedmaterials. This is justied by the fact that the author has used con-ditions similar to our experimental conditions (parallelepipedsamples, thermal excitation of no negligible duration t0). In gen-eral, this model is very suitable for building materials. Its precisionis between 2% and 4%.

6. Conclusion

This paper is a contribution to the general problem of the sus-tainable development, the improvement and the control of theproperties of woodconcrete composite. Thus, this work has aimedto study the inuence of moisture content on the thermal proper-ties of this type of material.

It is shown that even if the lightening of concrete by wood shav-ings engenders a signicant increase of the thermal insulationcapacity, the elaborated materials present high water absorption.The hygrothermal study allowed to demonstrate the importantapproximation made when only the thermal characteristics ofwoodconcrete composites are considered independently of their

translate well the evolution of the thermal conductivity as function

mogram recorded on the nonirradiated face of sample. However,

Buildthe curves which translate the evolution of the diffusivity as func-tion of water content present a maximum correspond to a watercontent Wm.

The inuence of temperature was not taken into account in thisstudy, while it is known that it plays an important role in the ther-mal properties of wet porous mediums [43]. Therefore, it is neces-sary to conduct a systematic work to study the inuence oftemperature on the thermal conductivity and diffusivity of thedeveloped woodconcrete composites.

Acknowledgments

The authors would like to thank the Ph. D student Kaoutar Baz-zar and the Professor Adil Alaoui Hadi from the Mechanical andCivil Engineering Laboratory of the Faculty of Sciences and Tech-nologies of Tangier, for their contribution in realizing some tests.

References

[1] Mahmoud T. Etude de matriaux minraux renforcs par des bres organiquesen vue de leur utilisation dans le renforcement et la rparation des ouvragestels que les ponts. Thse de Doctorat en Sciences, INSA de Lyon, France; 2005.

[2] Trouy-Triboulot MCh, Triboulot P. Matriau bois-structure et caractristiques.Technique de lIngnieur CB1(925);2001. p. 126.

[3] Fehrs JE. Ash from the combustion of treated wood: characterisation andmanagement options. In: Proceedings of the national bioash utilisationconference. Portland, USA; April 30 1996.

[4] Naik TR, Kraus RN. Use of industrial products in cement-based materials. In:Proceedings of the international congress creating with concrete. Dundee,Scotland, UK; September 610 1999.

[5] Siddique R. Utilization of wood ash in concrete manufacturing. Res ConserRecycling 2012;67:2733.

[6] Coatanlem P, Jauberthie R, Rendell F. Lightweight wood chipping concretedurability. Constr Build Mater 2006;20(9):77681.

[7] Bederina M, Quneudec M. Characterization of river-dune sand concretelightened by addition of wood shavings. In: Proceedings of the secondinternational conference on sustainable construction materials andtechnologies. Universit Politecnica delle Marche, Ancona, Italy; June 28302010.

[8] Ziregue A. Valorisation de coproduits industriels pour le dveloppement dematriaux de construction lgers: caractrisation physico-mcanique etof water content. Indeed, these two models take into account sev-eral aspects of the physical reality of the material. Nevertheless,the proposed method for determining the parameters h and ksmakes that the simulation is strongly dependent on the qualityof experimental measurements.

The study of the thermal diffusivity shows that its values de-pend on the model used for the counting of the experimental ther-hydrous state. Indeed, the presence of water within the studiedmaterials modies considerably their thermophysical characteris-tics. The differences observed, especially between the thermal con-ductivity of a dry material and that of a wet one, implyconsequences which can be very signicant during the establish-ment of the thermal balances of buildings. The treatment of woodshavings can be envisaged in next works to decrease the waterabsorption capacity, and therefore reduce the perturbation of ther-mal properties upon a change of the humidity inside or outside thebuilding. Bederina et al. [41] have shown that the treatment of thewood shavings reduced considerably the water absorption of con-crete lightened by wood shavings. Ledhem et al. [42] have shownthat the processing of wood shavings with oil intended for disposalallows signicantly decreasing the water absorption of the shav-ings as well as that of woodcement schistous nes composite.

The comparison of the experimental points of conductivity withthose deduced from existing theoretical models in the literatureshows that the KrisherKroll and ChaudharyBhandari models

114 D. Taoukil et al. / Construction andthermique dun bton calcaire lger aux copeaux de bois. In: Proceedings ofthe Colloque Mditerranen sur les Matriaux de Gnie-civil (CMEDIMAT).Oran, Alger; Decembre 67 2005.[9] Belhamri A, Rebahi K, Chikhi A, Bennamoun L. Etude et caractrisation destransferts thermiques dans des matriaux de construction locaux. In:Proceedings of the 13mes Journes Internationales de Thermique. Albi,France; 2830 Aot; 2007.

[10] Rihani M, Chedad K, Herrar R. Les rsidus de transformation du bois et lesdchets des abattoirs de viandes rouges au Maroc. Morocomp (Life TCY05/MA000141), Annex 4, Universit Chouaib Doukkali, Facult des Sciences ElJadida, Maroc; 2006.

[11] NF P18-560. Analyse granulomtrique par tamisage; 1990.[12] NF EN 933-8. Essais pour dterminer les caractristiques gomtriques des

granulats Partie 8: valuation des nes quivalent de sable; 1999.[13] Dreux G, Festa J. Nouveau guide du bton et de ses constituants. 8eme edition

Eyrolles 1998, 3eme tirage; 2007.[14] NM 10.1.004. Liants Hydraulique-Ciments-Composition, spcications et

critres de conformit; 2003.[15] NF EN 196-1. Mthodes dessai des ciments, Partie I. Dtermination des

rsistances mcaniques; 1995.[16] Aouadja FZ, Mimoune M, Laquerbe M. Etude exprimentale sur les btons

base de rsidus de bois. Revue Algrie-Equipement 1995;18:247.[17] Bederina M, Marmoret L, Mezreb K, Khenfer MM, Bali A, Queneudec M. Effect

of the addition of wood shavings on the thermal conductivity of the sandconcretes-experimental study and modeling. Constr Build Mater2007;21(3):6628.

[18] Tamba S, Voumbo LM, Wereme A, Gaye S, Sissoko G. Durabilit des btons base de copeaux de bois. J des Sci 2007;7(4):6772.

[19] Benaissa A. Dformations diffres dun bton de sable. Thse de Doctorat,Universit de Bordeaux, Laboratoire Rgional des Ponts et Chausses, France;1992.

[20] Cerezo V. Proprits mcaniques, thermiques et acoustiques dun matriau base de particules vgtales: approche exprimentale et modlisationthorique. Thse de Doctorat en Gnie Civil, INSA de Lyon, France; 2005.

[21] Krischer O, Kroll K. Technique du schage. Berlin: Centre Technique desIndustries Arauliques et Thermiques (CETIAT). Traduction Springer-Verlag;1963.

[22] Carson JK. Prediction of the thermal conductivity in pouros foods. Ph. D thesisin food engineering, Massey University, Palmerstone north, New Zealand;2002.

[23] Bal H, Jannot Y, Gaye S, Demeurie F. Measurement and modelisation of thethermal conductivity of a wet composite porous medium: laterite based brickswith millet waste additive. Constr Build Mater 2013;41:58693.

[24] Bousssaid S. Caractrisation physico-chimique, thermophysique et mcaniquedu matriau terre utilis en construction au nord du Maroc. Thse de Doctoraten Sciences, Universit Abdelmalek Essadi, Facult des Science de Ttouan,Maroc; 2001.

[25] Chaudhary DR, Bhandary RC. Heat transfer through a three phase porousmedium. Beit J Appl Phys 1968;2(1):8157.

[26] Sugawara A. Effect of moisture on thermal conductivity of porous materials. JAppl Phys Jpn. 1961;30(12):899906.

[27] Woodside W, Messmer JH. Thermal conductivity of porous media. I-Unconsolited sands. J Appl Phys 1961;32(9):168898.

[28] Mourtada A. Comportement thermique des mortiers disolation extrieur dubtiment. Thse de Docteur-Ingnieur, INSA de Lyon et UCB Lyon I, France;1982.

[29] Menguy G, Laurent M, Moutarda A, Leveau J. Cellule de mesure descaractristiques thermophysiques des matriaux E1700. Socit WeberBroutin, Bulletin technique: Deltalab; 1986.

[30] Mourtada A. Caractrisation thermique des parois opaques et transparentes dubtiment. Thse de Doctorat dEtat, Universit Claude Bernard Lyon I, France;1988.

[31] Parker WJ, Jenkins RJ, Butler CP, Abbott GL. Flash method of determiningthermal diffusivity-heat capacity and thermal conductivity. J Appl Phys1961;32(9):167984.

[32] Alrim K, Ledhem A, Douzan O, Dheilly RM, Queneudec M. Inuence of theproportion of wood on the thermal and mechanical performances of claycementwood composites. Cem Concr Compos 1999;21(4):26976.

[33] Khedari J, Suttisonk B, Pratinthong N, Hirunlabh J. New lightweight compositeconstruction materials with low thermal conductivity. Cem Concr Compos2001;23(1):6570.

[34] RILEM. Functional classication of lightweight concrete. Recommendations ofRILEM LC2, vol. 11;1978. p. 28183.

[35] El bouardi A. Etude en rgime stationnaire et dynamique des propritsthermophysiques de matriaux poreux humides non saturs utiliss en gniecivil. Thse de Doctorat dEtat, Universit Mohammed V, Rabat, Maroc; 1991.

[36] Laurent JP, Guerre-Chaley C. Inuence de la teneur en eau et de la tempraturesur la conductivit thermique du bton cellulaire autoclav. Mater Struct1995;28(8):46472.

[37] Degiovanni A. Conductivit et diffusivit thermique des solides. Techniques delIngnieur; 1994. R 2850.

[38] Yezou R. Contribution ltude des proprits thermophysiques des matriauxde construction cohrents et non cohrents. Thse de Docteur-Ingnieur, INSAde Lyon, France; 1978.

[39] Foures JC, Javelas R, Perrin B. Caractristiques thermiques des matriaux deconstruction: dtermination, variation en fonction de la teneur en eau. Revue

ing Materials 48 (2013) 104115Gnrale de Thermique 1981;230:1118.[40] Meukam P, Jannot Y, Noumowe A, Kofane TC. Thermophysical characteristics

of economical building materials. Constr Build Mater 2004;18(6):43743.

[41] Bederina M, Laidoudi B, Gouilleux A, Khenfer MM, Bali A, Quneudec M. Effectof the treatment of wood shavings on the physico-mechanical characteristicsof wood sand concretes. Constr Build Mater 2009;23(3):13115.

[42] Ledhem A, Dheilly RM, Queneudec M. Reuse of waste oils in the treatment ofwood aggregates. Waste Manage 2000;20(4):3216.

[43] Azizi S, Moyne C, Degiovanni A. Approche exprimentale et thorique de laconductivit thermique des milieux poreux humides-I. Experimentation. Int JHeat Mass Transfer 1988;31(11):230517.

D. Taoukil et al. / Construction and Building Materials 48 (2013) 104115 115

Moisture content influence on the thermal conductivity and diffusivity of woodconcrete composite1 Introduction2 Materials2.1 Basic materials2.2 Implementation of fresh concrete2.3 Preparation of specimens2.4 Formulations

3 Thermal conductivity models3.1 Krischer and Kroll model [21]3.2 Chaudhary and Bhandari model [25]3.3 Woodside and Messmer model [27]

4 Test methods5 Results and discussion5.1 Thermal conductivity5.2 Thermal diffusivity

6 ConclusionAcknowledgmentsReferences

Moisture Content Influence on the Thermal Conductivity and Diffusivity

Documents

Transcript of Moisture Content Influence on the Thermal Conductivity and Diffusivity