t2_18_4

8/4/2019 t2_18_4

1/23

Published by

Department of Civil & Mining EngineeringDivision of Structural Engineering

ISBN 91 - 89580 18 4 2001:18-4 SE

THERMAL PROPERTIES

OF CONCRETEVariations with the temperature and during the

hydration phase

Paolo Morabito 1

1ENEL.HYDRO Hydraulic and Structural Research Center

REPORT BE96-3843/2001:18-4

8/4/2019 t2_18_4

2/23

2

IPAC

S

Improved Production ofAdvanced Concrete Structures

THERMAL PROPERTIES OF CONCRETEVariations with the temperature and during

the hydration phase

Report No:2001:18-4

Author Paolo Morabito [email protected]

Address ENEL.HYDRO B.U. PIS Via Pozzobonelli 6, 20162 Milan, Italy

Task/Subtask no: T2/T2.3

Brite EuRam Contract No. BRPR-CT97-0437

Project no: Brite EuRam Proposal No. BE96-3843

Project title: IPACS - IMPROVED PRODUCTION OF ADVANCED

CONCRETE STRUCTURES

Project co-ordinator:

Betongindustri AB, Dr Mats Emborg

Partners: Betongindustri ABCementa ABSelmer ASATechnical University of DelftENELTechnical University of LuleNCC ABSkanska Teknik ABTechnical University of BraunschweigIsmesNorwegian Public Roads DirectorateElkem ASNorcem ASTechnical University of Trondheim

Date of issue of this report: 31 May 2001

Revised date: 31 May 2001

Project funded by the European Community under the Industrial & Materials Technologies

Programme (Brite-EuRam III)

8/4/2019 t2_18_4

3/23

3

IMPROVED PRODUCTION OF ADVANCED CONCRETE STRUCTURES - IPACS

Background

Research and practical experience show that the quality and lifetime of concrete structures largelydepend on the curing conditions in the concrete's early life, as inadequate curing leads to malfunctionand cracking. A major source of deleterious cracking already in the construction stage is theoccurrence of stresses in the hardening concrete due to restrained volume change related to hydrationtemperatures and shrinkage phenomena. It is thus of utmost importance, especially regarding new highperformance concrete, that the proper execution conditions are maintained throughout the constructionperiod by avoiding the premature cracking.

Objective of project

Main goal of IPACS is to evaluate, integrate and extend the existing knowledge about early ageconcrete crack prediction in engineering practice yielding the following benefits:Contractors and designers will have new and more reliable engineering instruments enabling them topredict and to optimise the technical effect and cost of alternative designs and execution procedures -all in the process of fulfilling the quality requirements set up by the owners or the community (codes).Reduced costs because of the present tendency to specify costly but unnecessarily rigorous crack

criteria will be avoided.Owners will have access to improved means of specifying and controlling desired qualityrequirements regarding serviceability and service life of their structures.Reduced maintenance costs and increase of service lifetime.

Main tasks and investigations in IPACS and output from the project:

Hydration and volume changes To acquire data for the modelling of properties of anumber of currently used concrete types.

Mechanical properties - Testing and modelling of mechanical properties.Behaviour of structures - Computer modelling of structural behaviour.Field tests - To check and improve the models of the previous tasks in full-scale tests.Expert System.

TheExpert System synthesises the results from the project into a robust engineering tool for planningand control of the production of concrete structures. It contains modules of varying simplicity, whichcan be used in all the phases of a construction project from pre-design to maintenance

Project Partners:

See earlier page

Project Co-ordinator:

Dr Mats Emborg Betongindustri AB (Heidelberger Zement North Europe) (SE)Dr Hans-Erik Gram/Mr Mats berg Cementa AB (Heidelberger Zement North Europe) (SE)

DisclaimerThe author/authors and producer of this report have used their best effort in preparing this report. These effortsinclude the development, research and testing of the theories and programs to determine their effectiveness. Theauthor/authors and producer make no warranty of any kind, expressed or implied, with regard to these programsor documentation contained in this report. The author/authors and publisher shall not be liable in any event forincidental or consequential damage in connection with, or arising out of, the furnishing, performance, or use ofthese programs.

Editorial/production supervision: Prof. Lennart ElfgrenCover design: Hans HedlundPrepress material: By report authorsPrinted and published by Lule University of Technology,

Department of Civil and Mining Engineering,Division of Structural EngineeringSE-971 87 Lule, Sweden

8/4/2019 t2_18_4

4/23

4

Table of Content

1 Introduction ......................................................................................................................... 7

2 The two-linear-parallel-probe method................................................................................. 7

3 Composition of the tested concretes.................................................................................. 10

4 Testing programme............................................................................................................12

4.1 Tests performed during the hardening stage............................................................... 12

4.2 Tests performed in hardened concrete samples under temperature variation............. 12

5 Test results against the temperature variation ................................................................... 12

5.1 Influence of temperature on thermal conductivity...................................................... 13

5.2 Influence of temperature on thermal diffusivity. ........................................................ 13

5.3 Influence of temperature on specific heat. .................................................................. 14

5.4 Modelling of the thermal properties against the temperature variation...................... 15

5.4.1 Thermal conductivity ........................................................................................... 15

5.4.2 Specific heat ......................................................................................................... 16

5.4.3 Thermal diffusivity............................................................................................... 16

6 Test results during the hydration stage.............................................................................. 17

6.1 Experimental tests on a pure cement paste sample .....................................................176.2 Experimental tests in concrete samples ...................................................................... 18

6.2.1 Thermal conductivity ........................................................................................... 19

6.2.2 Thermal diffusivity............................................................................................... 20

6.2.3 Specific heat ......................................................................................................... 20

7 Remarks............................................................................................................................. 21

8 References ......................................................................................................................... 23

8/4/2019 t2_18_4

5/23

5

Abstract

An experimental study on the thermal properties of hardening concrete is presented in thisreport. It takes into account the results of measurements of thermal conductivity and thermaldiffusivity and the results of specific heat calculated from the knowledge of the abovemeasured parameters.

The experimental testing method is described as well: based upon the linear heat sourcetheory, it requires the use of two special probes to be inserted into the sample.

The experimental programme was forwarded to test concrete mixtures with different kinds ofcement and aggregate.

The measurements were carried out from the pouring time of cylindrical samples and wereended up when hardened conditions were achieved.

The unavoidable temperature variations during the hydration have required the knowledge ofthe influence of the temperature on the thermal properties. This was studied as well byperforming experimental measurements on the same samples of concrete under different

levels of temperature.The whole results have been modelled by empirical relationships. They describe both thevariation of the thermal properties against the maturity age and the variation of the thermalproperties against the temperature in a range going from about 0C up to 100C.

8/4/2019 t2_18_4

6/23

6

Notation and symbols

Thermal conductivity, W/(mC)

D Thermal diffusivity, cm2/s

C Specific heat, kJ/(kgC)

20 Thermal conductivity at the reference temperature of 20C, W/(mC)

D20 Thermal diffusivity at the reference temperature of 20C, cm2/s

c20 Specific heat at the reference temperature of 20C, kJ/(kgC)

T Temperature, C

t Time, s

r Radial length, cm Bulk density, kg/m

3

Denotes the relative variation of thermal conductivity against the temperaturevariation, C-1

D Denotes the relative variation of thermal diffusivity against the temperaturevariation, C-1

c Denotes the relative variation of specific heat against the temperature variation, C-1

T Temperature change, C

te Maturity age or equivalent age of concrete, hTLPP Two Linear and Parallel Probe method

GHP Guarded Hot Plate method

8/4/2019 t2_18_4

7/23

7

1 IntroductionWithin the task # 2 of the IPACS project, an experimental research program was carried out

with the aim to determine the variations of the thermal properties of concrete with thetemperature and during the hydration phase.

The tests were performed by an innovative transient measuring technique based upon thelinear heat source theory. The method, the so-called Two Linear and Parallel Probe method(TLPP) allows to measure simultaneously the coefficients of thermal conductivity andthermal diffusivity.

The specific heat is determined according to the following relationship:

D

=c (1)

being:

c = specific heat, [kJ/(kgC)]; = thermal conductivity, [W/(mC)];

D = thermal diffusivity, [cm2/s]; = bulk density, [kg/m3].

The experimental research program has mainly taken into account limestone concretes mixedwith two different cement types, but additional measurements carried out during the hydrationof a pure cement paste sample and in a hardened sample of gravel concrete, used for the

construction of a sluice gate in Italy, are presented.The measurements performed during the hydration started right after the pouring ofcylindrical samples, 16 cm in diameter and 32 cm in height, and were stopped when nosignificant change in the thermal properties was detected. A complete set of tests took aboutup to 200 hours.

The effects of the temperature variations are analysed by performing TLPP tests in hardenedsamples placed inside a controlled climatic chamber by which a temperature variation fromabout 0C up to 100C is applied to the specimen.

2 The two-linear-parallel-probe methodThe testing method used to measure the thermal conductivity and diffusivity is the Two-Linear-Parallel-Probe method.

The method is based upon the transient theory of the linear heat source developed by Carslawand Jaeger (1959). According to that theory, the rate of the temperature rise at any point of aninfinite homogeneous medium heated by an infinite linear heat source is given from thefollowing relationship:

8/4/2019 t2_18_4

8/23

8

=tD

rt

Q

dt

rtdT

4exp

4),( 2

(2)

where:

Q = the heating power per unit length of the source, [W/cm];r = the radial distance of the point from the heat source [cm];

t = the time elapsed from the start of heating [s].The rate of the temperature rise is plotted in Figure 1 against the time; the graph shows thateq.(2) increases until a peak valueMis reached and then decreases going to zero at infinitetime.

Time

Rateoftemperaturerise

r

Figure 1 Rate of temperature rise against the time according to the transient linear

heat source theory.

It is possible to demonstrate that the thermal conductivity and diffusivity are in relationshipwith the peak value M and with the corresponding time tM according to the following

equations:

MtM

Q

=

)1exp(4 MtrD

=4

2(3)

The experimental set-up adopted to perform the test is given in Figure 2. Two thermal probes,4 mm in diameter and 300 mm in height, are inserted in a parallel way into cylindricalsamples having a diameter of 160 mm and a height of 320 mm. One probe is used as heatingprobe and is equipped with an electrical heating wire over the entire length; the other probe isthe temperature probe, it is usually spaced 2025 mm from the heating probe and is equipped

with a thermistor to measure the temperature.

8/4/2019 t2_18_4

9/23

9

Heating probe

Temperature probe

r

Thermistor

Sample

Figure 2 Experimental set-up to measure thermal conductivity and diffusivity.

The rate of the temperature rise, measured by the temperature probe, is computed from thestart of supply the heating probe and is fitted by eq. 2; the coefficientsMand tM are thusdetermined by a last mean square procedure.

A typical example of test carried out in a concrete sample is given Figure 3

0

0.001

0.002

0.003

0 200 400 600 800

Time [ s ]

dT/dt[C/s]

Experimental data

Best fit curve

Figure 3 Example of a TLPP test carried out on a concrete sample.

The TLPP test method is particularly suitable to be applied in damp and porous solids, like

concrete. The main features of the method are: little thermal gradients applied to the sample (less than 0.5 C/cm); very short duration of the test;

8/4/2019 t2_18_4

10/23

10

capability to perform tests during the hardening phase of concrete; suitability to test incoherent solids; use on site.

The reliability of the method has been verified through comparative measurements carried outby means of the standard guarded hot plate (GHP) method (ASTM C177-63) in samples of

reference materials and in fully dried concrete samples.The results of the comparative tests are given in Figure 4

0

0.5

1

1.5

2

2.5

3

0 0.5 1 1.5 2 2.5 3

Thermal conductivity measured by the GHP method, [W/(mC)]

Thermalconductivitymea

suredbytheTLPP

method,[W/(

mC)]

Pyrex sample

PTFE sample

Dry concrete 1 sample

Dry concrete 2 sample

Figure 4 Comparison between the standard guarded hot plate (GHP) method and the

TLPP method.

3 Composition of the tested concretesTwo concretes having the same composition and mixed with different cement types have beentested. The basic concrete composition is given in Table I. The two types of cement used for

each mixing are a blustfurnace cement, type 32.5 IIIA, and a Portland cement 42.5 IIA-L.Their chemical analyses are given in Table II.

8/4/2019 t2_18_4

11/23

11

Table I - Mix composition of the concrete.

Material Recipe

[kg/m3]

Density

[kg/m3]

Cement content (C) 300 300

Crushed limestone 0-1 mm 494






Total aggregate 1976

Water (W) 175 175

W / C 0.58

Total for 1 m3 of concrete 2451

Table II Chemical analyses of the cements.

Cement type 32.5 III A 42.5 II A-L

Ca0 [%] 49.26 65.27

SiO2 [%] 29.48 15.96

Al2O3 [%] 8.46 3.91

Fe2O3 [%] 1.50 1.87

K2O [%] 0.47 1.69

Na2O [%] 0.10 0.52MgO [%] 4.87 1.95

SO3 [%] 2.41 6.49

Cl [%] 0.140 0.094

PbO [%] - -

ZnO [%] - -

TiO2

[%] 0.32 0.22

Glow loss [%] 2.22 -

no solving rest [%] 0.50 -CO2 [%] - -

Mn3O4 [%] 0.17 0.072

S [%] 0.38 -

8/4/2019 t2_18_4

12/23

12

4 Testing programme4.1 Tests performed during the hardening stage.For each of the two concretes a cylindrical sample, having a diameter of 160 mm and a heightof 320 mm, was prepared.

Right after the mixing procedure the fresh concrete was cast in the moulds, compacted andthen subjected to a set of thermal conductivity and diffusivity measurements by the TLPPmethod.

The set of measurements in each sample started after about 0.5 hours from the beginning ofthe mixing procedure. To avoid loss of free water during the tests, which in turn causes

variation of the thermal properties due to the variation of moisture content, the samples wereproperly sealed.

Additionally, the thermal properties of concrete will depend on the temperature of the sampleunder test. Due to the development of heat of hydration, it is almost impossible to keepconstant the samples temperature. So, empirical relationships between thermal properties andtemperature were determined on the same samples after that their have reached completehydration; such relationships were used to reduce to the reference temperature of 20C thethermal conductivity and diffusivity measurements performed under variable temperature.

4.2 Tests performed in hardened concrete samples under temperature variation.The samples were placed inside a climatic cell that allows controlling the temperature and thehumidity. The temperature was varied from about 0C up to 100C at steps of 10 C. Incorrespondence of each step the samples were allowed to reach the temperature of the cell and this was checked by the temperature probe inserted into the specimen and then theTLPP test was performed.

5 Test results against the temperature variationThe influence of the temperature level was also investigated in completely hardened samplesof a concrete used for the construction of a sluice gate in Italy (Morabito, 2000), having anatural quartzy gravel as aggregate. The concrete mixing is given in Table III.

Table III Mixing composition of the concrete used for the construction of a sluice gate

in Italy.Material Content [kg/m3]Sand 0 4 mm 380Sand 0 8 mm 580Gravel 4 12.5 mm 440Gravel 8 25 mm 600

Pozzolanic cement type CEM IV-A 32.5 320Super plasticizer Sikament (1% of cement content) 3.2Water 190

8/4/2019 t2_18_4

13/23

13

With such an additional test it is possible to take into account both the effects of differentcement types as well different types of aggregate.

5.1 Influence of temperature on thermal conductivity.The results of the thermal conductivity measurements against the temperature are plotted in

Figure 5.They put in evidence that: the thermal conductivity decreases with the increase of the concrete temperature; the relationship does not seem to depend on the cement type; the decrease in thermal conductivity is more pronounced in limestone aggregate than in

natural gravel (0.15%C-1 against 0.06%C-1).

2.4

2.5

2.6

2.7

2.8

2.9

3

0 20 40 60 80 100

Temperature [C]

Thermalconductivity[W/(m

oC)]

Limestone concrete - Cem.32.5 IIIA

Limestone concrete - Cem.42.5 IIA-L

Gravel concrete

Figure 5 Thermal conductivity against the temperature variation for the three tested

concretes.

5.2 Influence of temperature on thermal diffusivity.The thermal diffusivity measurements are plotted in Figure 6 and lead to the same commentsmade for conductivity: the thermal diffusivity decreases with the increase of the concrete temperature; the relationship does not depend on the cement type; the decrease in thermal diffusivity is more pronounced in limestone aggregate than in

natural gravel (0.27%C-1 against 0.12%C-1).It is also to point out that the variations in thermal diffusivity are more pronounced than thosein conductivity.

8/4/2019 t2_18_4

14/23

14

0.9

1

1.1

1.2

1.3

1.4

1.5

0 20 40 60 80 100

Temperature [C]

Thermaldiffusivity,

[c

m2/s]in10-2

Limestone concrete - Cem.32.5 IIIALimestone concrete - Cem.42.5 IIA-LGravel concrete

Figure 6 - Thermal diffusivity against the temperature variation for the three tested

concretes.

5.3 Influence of temperature on specific heat.The specific heat was calculated according to eq. 1 from the experimental data of thermalconductivity and diffusivity.

The results are plotted in Figure 7. They put in evidence that: the specific heat increases with the increase of the concrete temperature; the relationship seems to be independent on the cement type; the variations of specific heat are only a bit more pronounced in limestone aggregate than

in natural gravel.

0.75

0.80

0.85

0.90

0.95

1.00

1.05

1.10

0 20 40 60 80 100

Temperature [C]

Specificheat[kJ

/(kgC)]

Limestone concrete - Cem.32.5 IIIALimestone concrete - Cem.42.5 IIA-LGravel concrete

Figure 7 Specific heat against the temperature variation for the three tested

concretes.

8/4/2019 t2_18_4

15/23

15

5.4 Modelling of the thermal properties against the temperature variationThe experimental results seem to put in evidence that thermal conductivity and specific heatcould be conveniently described by a linear relationship with the temperature.

For both of these parameters the following relationship has been adopted:

( )oXXX

TTo

oT =

(4)

being XT the thermal conductivity or the specific heat at the generic temperature T, theslope of the relationship to be determined from the experimental data, To a reference

temperature and Xo the thermal property at the reference temperature.

The reference temperature has been assumed to be equal to 20C so the coefficient in eq. 4represents the relative variation of the thermal property against the unit variation of

temperature from the reference temperature of 20C.

The variation of thermal conductivity with the temperature is thus described by the coefficient whereas the variation of the specific heat is described by the c coefficient.

As the measurements of thermal diffusivity exhibit a slight non-linear trend, its variations are

calculated by eq. 1 from the knowledge of and c.

5.4.1 Thermal conductivityIn Figure 8 the experimental results of thermal conductivity are plotted according to the eq. 4

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

-20 0 20 40 60 80

Temperature, (T-20) [C]

Relativevariationofthermalconductivity



Gravel concrete

Figure 8 Relative variation of thermal conductivity against the temperature variation.

8/4/2019 t2_18_4

16/23

16

The coefficient depends on the aggregate type and from a best fit on the experimental datathe following values can be assumed:

= 0.0006 [C-1

] for gravel aggregate;

= 0.0015 [C-1

] for limestone aggregate.

5.4.2 Specific heatIn Figure 9 the variations of specific heat are plotted according to eq. 4.

The corresponding c coefficients are:

c = 0.0007 [C-1

] for gravel aggregate;

c = 0.0016 [C-1

] for limestone aggregate.

-0.1

-0.05

0

0.05

0.1

0.15

-20 0 20 40 60 80


Relative

variationofspecificheat



Gravel concrete

Figure 9 Relative variation of specific heat against the temperature variation.

5.4.3 Thermal diffusivityThe relative variation of thermal diffusivity can be predicted from the following relationship,derived from eq. 1:

( ) ( )( )20T1

20T

D

DD

c

c

20

20T+

=

(5)

A comparison between the measured and predicted results is given in Figure 10.

8/4/2019 t2_18_4

17/23

17

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

-20 0 20 40 60 80


Relativevariationoftherm

aldiffusivity



Gravel concrete

Predicted

Figure 10 Relative variation of thermal diffusivity against the temperature variation.

6 Test results during the hydration stage6.1 Experimental tests on a pure cement paste sampleThe gradual transition of the cement paste from plastic to hardened material gives rise to avariation of the thermal properties in young concrete. Preliminary tests were performed on apure cement paste sample of Portland cement with a water/cement ratio of 0.4. The totaltemperature variation of the sample during the run time of the tests was of only about 3.5 C(see Figure 11) so no temperature correction was carried out on the measured results ofconductivity and thermal diffusivity.

21

22

23

24

25

26

27

0 5 10 15 20 25

Time [h]

Temperature[C]

Figure 11 Temperature variation of a Portland cement paste sample duringhydration.

8/4/2019 t2_18_4

18/23

18

The measurements result of thermal conductivity, diffusivity and specific heat are plotted inFigure 12 as ratio between the actual measurement X(t) and the corresponding one in thehardened condition XH. They put in evidence that: the thermal properties seem to reach a plateau level after about 20 hours; both thermal conductivity and thermal diffusivity increase during the hydration phase of

cement whereas the specific heat exhibits a slight decrease;

thermal conductivity and thermal diffusivity reach a plateau value after an increase ofabout 9% and 13%, respectively;

as a consequence of that, the specific heat, calculated according to eq. 4, decreases of onlyabout 3%.

Water exhibit, from one hand, a lower thermal conductivity and diffusivity than the mainsolid minerals and, from the other hand, a greater specific heat than solid minerals. As aresult, the variations of thermal properties in a hydrating cement paste are caused by thegradual transformation of the free water in bound water and by the consequent increase of thesolid/fluid ratio.

0.7

0.8

0.9

1.0

1.1

0 5 10 15 20 25

Age [h]

X(t)/XH

Thermal conductivity

Thermal diffusivity

Specific heat

H = 1.013 [W/(mC)]D H = 0.00340 [cm

2/s]

cH = 1.6 [kJ/(kgC)]

Figure 12 Change in thermal properties during the hardening phase of a pure

Portland cement paste sample.

6.2 Experimental tests in concrete samplesAs previously mentioned, such tests were carried out in limestone concretes mixed with twodifferent cement types. The tests, being performed during the hydration of the cement, aresubjected to unavoidable temperature variations. Such variations were measured as well andthe thermal conductivity, diffusivity and specific heat measurements were corrected for thetemperature changes according to the corresponding relationships described in chapter 5. It isassumed that the correlations between thermal properties and temperature, determined inhardened concretes, can be applied at any age of concrete. As the aggregate is by far the main

constituents of a concrete mixing and the experimental results have demonstrated that thevariation of the thermal properties with the temperature depends on the nature of aggregate,such an assumption is likely to be realistic.

8/4/2019 t2_18_4

19/23

19

The experimental results during the hydration stage are thus reduced to the referencetemperature of 20C and are plotted against the equivalent age. The latter has been calculatedtaking account of the temperature variations during the hydration stage and of theexperimental relationships of energy activation (Morabito, 2000) for the two tested cements,given by:

EA = 45.14 + 0.99 (20 T) for cement type 32.5 IIIA (6)EA = 58.80 + 1.43 (20 T) for cement type 42.5 IIA-L (7)

20

25

30

35

0 24 48 72 96 120 144

Age [h]

Tem

perature[C]

Cem ent type 32.5 IIIA

Cem ent type 42.5 IIA-L

Figure 13 Temperature variation during the hydration stage of the tested concretes.

6.2.1

Thermal conductivityThe ratio /20 between thermal conductivity at early age and the conductivity of the hardenedconcrete at the reference temperature of 20C is plotted in Figure 14 against the equivalentage for the two tested concretes.

0.7

0.8

0.9

1.0

1.1

0 24 48 72 96 120 144 168 192

Equivalent age [h]

(te)/20

Cem ent type 32.5 IIIA

Cem ent type 42.5 IIA-L

Figure 14 Variation of thermal conductivity against the equivalent age of hardening

concretes.

8/4/2019 t2_18_4

20/23

20

From a best fit on the experimental data, the conductivity variation can be described by thefollowing relationship:

( )et2.8

20

e e08.092.0t

+=

(8)

being 20 equal to 2.75 W/(mC) for both types of concrete.

6.2.2 Thermal diffusivityIn a similar way as for conductivity, the ratio D/D20 between the thermal diffusivity at earlyage and the diffusivity of the hardened concrete at 20C is plotted in Figure 15 against theequivalent age.

The rise of thermal diffusivity at early age can be described by the following equation:

( )et

6.6

20e e07.093.0D

tD

+= (9)

with D20 equal to 0.0126 cm2/s for the concrete with cement type 32.5 IIIA and 0.0129 cm

2/s

for the other tested concrete.

0.70

0.80

0.90

1.00

1.10

0 24 48 72 96 120 144 168 192

Equivalent age [h]

D(te)/D20

Cem ent type 32.5 IIIACem ent type 42.5 IIA-L

Figure 15 - Thermal diffusivity against the equivalent age of hardening concretes.

6.2.3 Specific heatThe specific heat has been calculated according to eq. 1 and is plotted in Figure 16.

8/4/2019 t2_18_4

21/23

21

The results put in evidence only a very slight decrease of this parameter during the hydrationstage so, for all the practical purpose, a constant value, equal to 0.21 kcal/(kgC) for the twotested concretes, can be assumed.

0.7

0.8

0.9

1.0

1.1

0 24 48 72 96 120 144 168 192

Equivalent age [h]

c(te)/c20

Cement type 32.5 IIIA

Cement type 42.5 IIA-L

Figure 16 - Specific heat against the equivalent age of hardening concretes.

7 RemarksSome relationships dealing with the variations of thermal properties of concrete, namelythermal conductivity, thermal diffusivity and specific heat, are proposed in this paper. Theyhave been obtained from best-fit analyses on experimental measurements carried out inconcrete samples. The tests have been performed by the Two-Linear-Parallel-Probe method, atransient method particularly suitable to measure thermal conductivity and diffusivity inconcrete samples.

The tests have taken into account the variations of the thermal properties at early ages, causedfrom the gradual transition from a plastic to a lytic material during the cement hydration, and

the variations of the thermal properties with the temperature.The temperature effects, examined in a limestone concrete and in a gravel concrete, have putin evidence that the thermal conductivity and diffusivity decrease as far as the temperatureincreases whereas the specific heat increases with the temperature rise. The variations, whichare more pronounced in the limestone concrete, can be conveniently described by thefollowing general relationship:

( )20T20X

20XTX =

8/4/2019 t2_18_4

22/23

22

where is a temperature coefficient that depends on the aggregate type and must beexperimentally determined.

To have an idea of the order of magnitude and the range of variation of, in Table IV thevalues of this coefficient for two tested concretes are summarised.

Table IV Temperature coefficients determined in the two tested concretes.Temperature coefficient [C

-1] Limestone concrete Gravel concrete

Thermal conductivity, -0.0015 -0.0006

Specific heat, c 0.0016 0.0007

Thermal diffusivity, D= c -0.0031 -0.0013

In Table IV, it has been assumed that the thermal diffusivity changes linearly with the

temperature and a D coefficient equal to chas been adopted. This rises from eq. 5

where the second term of the denominator can be neglected in respect to 1 for lowtemperature variations. At last, the temperature coefficients do not seem to be affected fromthe cement type and this is in agreement with the main role played from the aggregate.

The test results carried out during the hydration stage put in evidence a gradual increase ofthermal conductivity and diffusivity up to a plateau value corresponding to that one measuredin the hardened samples whilst the specific heat can be assumed to be constant during thehydration process.

The gradual transformation of the free water in bound water gives rise to a lytic material.Water conductivity and diffusivity are lower than conductivity and diffusivity of a lyticmaterial typically 0.6 W/m/C against 2.9 W/m/C for thermal conductivity and 0.00142

cm2

/s against 0.015 cm2

/s for thermal diffusivity, respectively and this explains the increaseof conductivity and diffusivity in hardening concretes. Additionally, a better process of heatconduction is to be ascribed to a solid than to a fluid and composite material because in thelatter the unavoidable thermal contact resistances between the different components canobstruct the process of heat conduction.

On the other hand, the specific heat of water is greater than the specific heat of a lytic material 4.186 kJ/kg/C against 0.75 kJ/kg/C - so a decrease in specific heat would be expectedduring the hardening stage. The experimental results have put in evidence a decrease in thepure cement paste of the order of only 3% but negligible variations are observed in theconcrete samples. Such a different behaviour is to be ascribed to the role of the aggregate

which, from one hand, it is not affected from physical/chemical changes during the hydrationstage and, from the other hand, its content in the concrete mixes is of the order of 80% against12% of cement content.

At last, the increases in conductivity and diffusivity in hardening pure cement paste sampleshave been of about 9% and 13%, respectively. The corresponding increases in concretesamples during the hydration are of the order of 78%. The effects of the aggregate are stillevident but to a lower extent than in the specific heat.

8/4/2019 t2_18_4

23/23

8 ReferencesCarslaw, H. S. and Jaeger, J. C. Conduction of Heat in Solids. 3rd edition, Clarendon Press,

Oxford 1959.

Morabito, P. Measurements of the thermal properties of different concretes. 11th

ECTP.Pion Limited, London 1989.

Lanciani, A. et al. Measurements of the thermophysical properties of structural materials inlaboratory and in situ: methods and instrumentation. High Temperature-High Pressure,vol. 21, pp. 391-400, 1989.

Lanciani, A. et al. The two-linear-parallel-probe method: a review. 12th ECTP. PionLimited, London 1993.

A.S.T.M. Thermal conductivity of materials by means of a guarded hot plate. ASTMSpecification C177-63, 1963.

Morabito, P. Determination of the apparent activation energy by adiabatic tests on concretesamples. IPACS Report, Task # 2, Sub-task # 2.2, 2000.

Morabito, P. Field test in Italy: sluice gate on the Brembo river. IPACS Report, Task # 5,Sub-task # 5.1, 2000.

t2_18_4

Documents