You are on page 1of 8

Fracture Mechanics of Concrete and Concrete Structures -

Recent Advances in Fracture Mechanics of Concrete - B. H. Oh, et al.(eds)

2010 Korea Concrete Institute, Seoul, ISBN 978-89-5708-180-8

Recent advances on self healing of concrete

E. Schlangen, H. Jonkers, S. Qian & A. Garcia

Delft University of Technology, Microlab, Delft, Netherlands

ABSTRACT: In this paper an overview is given of new developments obtained in research on self healing of
cracks in cement based materials and asphalt concrete. At Delft University various projects are running to
study self healing mechanisms. The first project that is discussed is Bacterial Concrete, in which bacteria are
mixed in concrete, that can precipitate calcite in a crack and with that make concrete structures water tight
and enhance durability. In a second project hybrid fiber reinforced cementitious materials are studied that can
mechanically repair cracks when they occur. The last project described in this paper is on the raveling of po-
rous asphalt concrete and how to heal this damage by incorporating embedded microcapsules or steel fibers.
The state of the art results in all projects show that self healing is not just a miracle, but materials can be de-
signed for it.

1 INTRODUCTION On the other hand, materials can also be designed

to have a self healing capacity (Schlangen & Joseph
In the past, quite some investigations on the topic of 2008). Then we classify them as autonomic materi-
self healing of concrete have been conducted. als, which can again be subdivided in passive and
Neville (2002) gives a useful overview of his litera- active modes. A passive mode smart material has the
ture search in this field. He puts the practical signifi- ability to react to an external stimulus without the
cance of autogenous healing in the reduction of wa- need for human intervention, whereas an active
ter transport through cracks, for example in concrete mode smart material or structure requires interven-
water pipes. Neville also concludes from his litera- tion in order to complete the healing process.
ture research that there is no agreement between dif- Van Breugel (2007) discusses the performance of
ferent studies about what happens inside the crack structures with elapse of time. Gradual degradation
when self healing occurs and therefore further re- occurs until the moment that first repair is urgently
search would be useful. The early research on self needed. The durability of concrete repairs is often a
healing of concrete mainly focused on water retain- point of concern. Very often a second repair is nec-
ing structures or reservoirs where leakage through essary only ten to fifteen years later. Spending more
cracks was the main issue (Edvardsen 1999, Reinhardt money initially in order to ensure a higher quality
& Joos 2003). In the research of Ter Heide et al often pays off. The maintenance-free period will be
(2005) and Granger et al (2006) the main focus was longer and the first major repair work can often be
regaining mechanical properties of cracks in early postponed for many years. The extreme case would
age concrete by ongoing hydration of cement parti- be that no costs for maintenance and repair have to
cles. Ter Heide (2005) gives a nice overview of dif- be considered at all because the material is able to
ferent causes of autogenic healing (Fig. 1), in which repair itself.
a materials has already by nature the ability to heal In the last 5 years the design of materials with
itself. healing ability is becoming more and more popular
in a wide range of materials and applications (Van
der Zwaag 2007 and Gosh 2008). For cement based
materials different method can be found in literature.
In the first type of approaches encapsulated sealants
or adhesives are used (Dry 2000). The adhesives can
be stored in short fibre (Li et al. 1998, Qian et al.
2009) or in longer tubes (Nishiwaki et al 2006, Jo-
seph 2008, Joseph et al. 2008). Another approach is
Figure 1. Possible causes of self healing: (a) formation of cal- incorporating an expansive component in the con-
cium carbonate or calcium hydroxide, (b) sedimentation of par- crete which starts to expand and fill voids and cracks
ticles, (c) continued hydration, (d) swelling of the cement- when triggered by carbonation or moisture ingress
matrix, (Ter Heide , 2005).
J = D ( h,et
(Hosoda T )al.h 2007, Sisomphon et al. 2009). Using (1) explicitly
sity accounts
was mainly due toforthethelimited
evolution of hydration
life-time (hours to
bacteria to stimulate the self healing mechanism is areaction
few days) andof SF content. This enzymatic
the (urease-based) sorption isotherm
an alternative but promising
The proportionality technique
coefficient studied
D(h,T) is atcalled
dif- reads viability of the applied bacterial species. In
ferent groups (Bang et al. 2001, Jonkers
moisture permeability and it is a nonlinear function & Schlan- the present study the application of alkali-resistant
of the2007, De Muynk
relative humidity2008).
h and temperature T (Baant endospore-forming bacteria to enhance the self-
& Najjar 1972). The moistureadvances
In this paper some recent in threerequires
mass balance ongo- healing capacity
we (hcompressive of concrete

c s ) = G ( c s )
is investigated. Tensile-
+ 1

that the variation in time of the water mass perFirst
projects at Delft University are discussed. unit and , ,
strength characteristics
, 1

of reference

c c )hare

volume Bacterial
of concreteConcrete
(wateris content
described w) bein equal
whichto the (no bacteria added) and bacterial e concrete10

main focus of
divergence is sealing of cracks
the moisture flux and
J thus blocking of tified (Jonkers & Schlangen 2009a). Furthermore,
)h in concrete
the path to the reinforcement in order to improve the the viability ofKbacteria (g
c 10
c s e
) calcite poten-

= JSecond topic is on promoting the healing is quantified and, (finally, precipitation
, 1


of damage in fibre reinforced materials by adding tial of bacterial concreteis demonstrated by ESEM

different compnents to the mixture. The third project analysis. An example is given in Figure 2 where cal-

is onThehealing of asphalt
water content concrete,
w can where as
be expressed twothesmart
sum wherecrystals
cite the first term by
formed (gelbacterial
isotherm) represents are
precipitation the
of the evaporable water we (capillary water,porous
of healing are studied. In one case water physically bound (adsorbed) water and the second
vapor, andcontaining
adsorbed awater)rejuvenator
and thearenon-evaporable
used to start term (capillary isotherm) represents the capillary
the passive autonomic
(chemically bound) healing
water process.
wn (MillsIn the 1966,
other water. This expression is valid only for low content
case steel wool fibres are mixed
Pantazopoulo & Mills 1995). It is reasonable tothrough the bitumen of SF. The coefficient G1 represents the amount of
assumecan thatbetheheated with an
evaporable waterinduction machine.
is a function of water per unit volume held in the gel pores at 100%
This internal heating can repair damage
relative humidity, h, degree of hydration, c, and in the mate- relative humidity, and it can be expressed (Norling
degree Thisofissilica
thenfumean active autonomic
reaction, s, i.e.healing
h, , ) Mjornell 1997) as
= age-dependent sorption/desorptionto isotherm
an external stimulus is needed startc thes
(Norling Mjonell 1997). Under this assumption and c c+ ks s
G ( c s ) = k vg (5)
by substituting Equation 1 into Equation 2 one 1
c vg s
2 SELF HEALING PROJECTS where kcvg and ksvg are material parameters. From the
w h w maximum amount of water per unit volume that can
2.1 + ( D h ) =
Bio-concrete e & + we & + w& (3) fill all pores (both capillary pores and gel pores), one
h t h c
s n
can calculate K1 as one obtains
In this study the potential cof bacteria s to act as a self-
healing agent in concrete is investigated. Although
the ideawtoe/huseis bacteria
the slopeand of the sorption/desorption
integrate them in the

g c c h
isotherm matrix
concrete (also may called
odd at first,capacity).
it is not from The w 0.188 s + 0.22 s G
c s 1 e 1

agoverning equationviewpoint.
(Equation Bacteria
3) must be completed (6)
0 1

microbiological naturally oc- K ( c s ) =

by appropriate boundary and initial
cur virtually everywhere on earth, not only on its conditions. 1

g c c h

The relation
but also between the amount
e.g. in of evaporable
surface deep within, sediment and e 1

rock relative
a depth of morehumidity
than 1iskm. called adsorption
Various species
of so-calledifextremophilic
measured with increasing
bacteria, i.e. bacteriarelativity
that The material parameters kcvg and ksvg and g1 can
humidity and desorption isotherm
love the extreme, are found in highly desiccated en- in the opposite be calibrated by fitting experimental data relevant to
case. Neglecting
vironments such astheir difference
deserts, but also(Xiinside
1994),and in free (evaporable) water content in concrete at
the following,
even sorption
in ultra-basic isothermwhich
environments will be canused with
be con- various ages (Di Luzio & Cusatis 2009b).
sidered to both sorption to theandinternal
desorption conditions.
concrete envi-
By the way, if the hysteresis
ronment. Typical for many desiccation- and/or al- of the moisture 2.2 Temperature evolution
isotherm would
kali-resistant be taken
bacterial into isaccount,
species two different
their ability to form
relation, evaporable
endospores. water vs relative
These specialized humidity,
cells which must
are char- Note that,
Figure at earlyobserved
2. Biominerals age, since the chemical reactions
by ESEM.
be used according
acterized by an extremelyto the sign of the variation
low metabolic activity, of are
the associated with cement hydration and SF reaction
relativity humidity. The shape
known to be able to resist high mechanically- and of the sorption areExtensive
exothermic, the temperature
results of this studyfield is not uniform
are published else-
isotherm forinduced
chemically HPC is stresses
influenced andbyaremanyviable parameters,
for peri- for non-adiabatic
where (Jonkers et systems
al. 2009b).evenToifdate
the environmental
the main con-
ods of up tothose that influence
200 years. In some extent and rate
previously of the
published temperature
clusions is constant.
of this Heat conduction
ongoing research can ex-
are that the be
studies application and,of inbacteria
turn, for determine
cleaningpore of described in concrete, at least for temperature
periments done in this study show that alkaliphilic not
structure and pore size distribution
concrete surfaces (De Muynck et al. 2008) and (water-to-cement exceeding 100C (Baant
endospore-forming & Kaplan in1996),
bacteria integrated the con-by
ratio, cement
strength chemical
improvement composition,mortar
of cement-sand SF content,
(Bang Fouriers
crete law,canwhich
matrix readsprecipitate calcium carbon-
etcuring time and
al. 2001) wasmethod,
reported. temperature, mix additives,
Although promising re- ate minerals. Water, needed for the activation of en-
sults In the literature
reported, the various formulations
major drawback of thecan be
latter q = T can enter the concrete structure through
dospores, (7)
found to describe the sorption isotherm
studies was that the bacteria and compounds needed of normal freshly formed cracks. Furthermore, for mineral pre-
for mineral(Xiprecipitation
et al. 1994).could However,
only bein applied
the present ex- cipitation,
where q activeis thecellsheatneedflux,
an organic
T is substrate that
the absolute
paper the
ternally on semi-empirical
the surface of the expression
after crack- by can metabolically
temperature, and beisconverted to inorganic carbon
the heat conductivity; in this
Norling Mjornell
formation had occurred. (1997) Thisismethodological
adopted because neces-it that can subsequently precipitate with free calcium
Proceedings of FraMCoS-7, May 23-28, 2010
to calcium carbonate. Free calcium is usually present J = Dcarbonate
concrete some calcium ( h , T ) h will be formed due
in the concrete matrix, but organic carbon is not. In to the reaction of CO2 present in the crack ingress
the first experiments organic carbon was applied ex- water with Portlandite The(calcium hydroxide)
proportionality present D(h,T)
ternally as a part of the incubation medium, while in the concrete matrix
moisture permeability and it is are-nonlinea
according to the following
ideally it should also be part of the concrete matrix. action: of the relative humidity h and temperature
In that case only external water is needed to activate & Najjar 1972). The moisture mass balanc
the concrete-immobilized bacteria which can then CO2 + Ca(OH)2 thatCaCO 3 + H2O in time of the water mas
the variation
convert organic carbon present in the concrete ma- volume of concrete (water content w) be eq
trix to calcium carbonate and by doing so seal freshly The amount ofdivergence
calcium carbonate
of the moistureproduction
flux J in
formed cracks. this case in only minor due to the limited amount of
Therefore it was decide to design a new strategy CO2 present. As Portlandite
J is a rather soluble min-
in which a two component biochemical healing agent eral in fact most of it=present

on the crack surface

composed of bacterial spores and a suitable organic
will dissolve and diffuse out of the crack into the
bio-cement precursor compound is used. Both the overlying water mass. TheSubsequently,
water contentaswmore can be 2 is
COexpressed a
spores and the food are immobilized in reservoir po- present in the overlying water, dissolved
of the evaporable waterPortlandite
we (capillary wa
rous expanded clay particles. In this way the spores will as yet precipitate
vapor,in andthe form of calcium
adsorbed water)carbon-
and the non-e
and bacteria are also protected during the production ate but somewhat(chemically
away from thebound) crack itself,
wateras can wn (Mil
and hardening of the concrete and will survive be seen in FigurePantazopoulo
3A. The self&healing process Itin is reas
Mills 1995).
longer until the moment that self healing is needed. bacterial concreteassume
is muchthat moretheefficient
evaporable due water
to the is a fu
active metabolic relative
conversion humidity, h, degree of by
of calcium lactate hydration
the present bacteria:
degree of silica fume reaction, s, i.e. we=w
= age-dependent sorption/desorption
2 CaCO
Ca(C3H5O2)2 + 7O(Norling 3 + 5CO
Mjonell 2 + 5H2O
1997). Under this assum
by substituting Equation 1 into Equati
This process does not only produce calcium car-
bonate directly but also indirectly via the reaction of
on site produced COw2 ewith h
Portlandite present w on the
e does w
e &
crack surface. In the
latter +case,
h t
( DPortlandite
h h ) =

&c + not s + w

dissolve and diffuse away from the crackc surface,s
but instead reacts directly on the spot with local bac-
terially produced where
CO2 towadditional
e/h is thecalcium
slope ofcarbon-
the sorption/
ate. This process isotherm
results in efficient
(also called crack sealing
moistureas capac
can be seen in Figure 3B. equation (Equation 3) must be
The conclusionbyofappropriate
this work boundary
is that theand proposed
initial conditi
Figure 3. Cracked concrete specimens containing porous ag- two component biochemical
The relation healing agent composed
between the amount of e
gregates with food only (A) (top) and food plus bacteria (B) of bacterial sporeswater
and aand suitable organic
relative bio-cement
humidity is called
(bottom). The left figures are before and the right figures are precursor compound, both immobilized
isotherm if measuredinwith reservoir
after the healing.
porous expanded humidity
clay particles, represents aisotherm
and desorption promis- in th
Concrete disks are prepared containing the po- ing bio-based and case.thus sustainable
Neglecting theiralternative
difference to (Xi et al.
rous aggregates filled with food (lactate) only and strictly chemical or
following, basedsorption
healing agents.
isotherm will be
with food and bacteria. The specimens are cured for reference to both sorption and desorption c
56 days and then tested in a deformation controlled 2.2 By the way, if the hysteresis of the
tensile splitting loading to crack them partially. Af- isotherm would be taken into account, two
Self Healing fibre concrete

ter this cracking the specimens are placed in a per- Besides permeability enhancement,
relation, evaporablemany water research-
vs relative humi
meability test setup in which water is applied at one ers also looked into the mechanical
be used according toproperty the signrecov-
of the varia
side of the specimen for 24 hours. After the healing ery as a result ofrelativity
self-healing humidity.
in concrete Thematerials.
shape of the
the cracks are examined under the microscope and As suggested by isotherm
many previous for HPCstudiesis influenced
(Edvardsen by many p
the results are shown in figure 3. Also the perme- 1999, Ter Heide especially
2005, Reinhardt those that& Joosinfluence
the and
ability of the healed specimens was determined. crack width of thechemical
concrete reactions
material was and,foundin toturn,be determ
These results are discussed in Jonkers et al. (2009b). structuretoand
critical for self-healing takepore sizeThe
place. distribution
require- (water-
The outcome of this study shows that crack healing ment of crack widthratio, tocement
promotechemical composition,
self-healing falls SF
in bacterial concrete is much more efficient than in roughly below 200 curing
m and timepreferably
and method, lowertemperature,
than 50 mix
concrete of the same composition but without added m (Edvardsen etc.).1999),Inespecially
the literature for various
biochemical healing agent. The reason for this can based on ongoingfound to describe
hydration of cement. the Yetsorption
in prac-isotherm
be explained by the strictly chemical processes in tice, such small concrete
crack width (Xi etisal.very 1994). However,
difficult to in th
the control and additional biological processes in the paper
achieve consistently the semi-empirical
in normal concrete structures, expressionif pro
bacterial concrete. On the crack surface of control not possible at all.Norling Mjornell (1997) is adopted b

Proceedings of FraMCoS-7, May 23-28, 2010

J =To Dachieve
( h, T )h controlled tight crack width, a new
(1) explicitly accounts for the evolution of hydration
class of fiber reinforced strain hardening cementi- reaction and SF content. This sorption isotherm
proportionality termed as Engineered
coefficient D(h,T) Cementi-
is called reads
tious Composites (ECC) has been
moisture permeability and it is a nonlinear function developed by Li Average
etof the
al. relative
(1998) humidity
and continuously evolved Tover
h and temperature the
(Baant crack width:

last 15 years. ECC has been deliberately
& Najjar 1972). The moisture mass balance requires engineered we (h c s ) = G1 ( c , s )1 1
~15 m
that themicromechanics
variation in time theory of the waterto possess
mass perself- unit
, ,

controlled crack width that
volume of concrete (water content w) be equal does not depend ontosteel
the e

c c
divergence of the or structural
moisture flux dimensions.
J Instead, the )h
fibers used in ECC are tailored to work with a mor- (g
K ( c s ) e
c 10
tarmatrix in order to suppress localized brittle frac-
, 1

in= favor
J of distributed microcrack damage, even

when the composite is tensioned to several percent

The water ECC content
with crack w can width as low as as
be expressed 30 themicron
sum where the first term (gel isotherm) represents the
of the evaporable water we (capillary water, crack
been made. Given the well controlled water physically bound (adsorbed) water and the second
vapor, and Li etadsorbed
al (1998) haveand
water) investigated
the non-evaporablethe self- term (capillary isotherm) represents the capillary
(chemically behavior of ECCwater
bound) under wa nnumber (Millsof expo-1966, water. This expression
is valid only for low content
width: ~25
sure conditions. In their
Pantazopoulo & Mills 1995). It is reasonable experiments, deliberately to of SF. The coefficient 1 represents the amount of
width:G~60 m
assume that ECC specimens water
the evaporable were exposed
is a function to vari- of water per unit volume
m held in the gel pores at 100%
ous commonly
relative humidity, encountered
h, degreeenvironments,
of hydration,including c, and relative humidity, and it can be expressed (Norling
of silica fume andreaction,
submersion, wetting
s, i.e. we=weand (h,dry-
Mjornell 1997) as
= cycles,
age-dependentand chloride ponding. The mechanical
sorption/desorption isotherm
(NorlingtransportMjonell properties
1997). Undercan bethis largely recovered,
assumption and c c+ ks s
G ( c s ) = k vg (5)
especially for ECC specimens
by substituting Equation 1 into Equation 2 one preloaded to below 1
c vg s
obtainstensile strain. Besides the small crack width, the
low water/binder ratio in addition to the large where kcvg and ksvg are material parameters. From the
w h
of fly ash in their w
also helps pro- maximum amount of water per unit volume that can
e self healing via
+ ( D h ) = e
&c + e
&s + w&nand poz- (3) fill all4.pores
Figure ESEM (both capillary
images pores
of cracked and samples
SHCC gel pores),
h t
zolanic activities. h can calculate K1 as one obtains
c s
Self-healing behavior of pre-cracked SHCC (strain
where we/h cementitious
is the slope composites)
of the is also a main
sorption/desorption 2. The observations under ESEM andXEDS
g h
isothermmade topic
(also in called
the Microlab moistureat Delft University.
capacity). The firm that the w microcracks
s + submerged
s G e in water

c 10
c were

SHCC with local waste materials (blast fur- c


healed mainly with calcium carbonate.



governing equation (Equation 3) must be completed (6)
0 1

nace slag andboundary

by appropriate limestoneandpowder is investigated
initial conditions. suggests
K ( c s ) =that the healing products

grow from both
(Qian et al. 2009). Four-point bending testsevaporable
are used g the h

faces of the crack towards middle of the crack.
The relation between the amount of e c c

waterprecrack SHCC beam
and relative humidity specimens
is called deflected up to
adsorption This may be explained by the relatively high con- 1

2.4 mm with subsequent curing

isotherm if measured with increasing relativity in water and air for centration of calcium hydroxide
The material parameters kcvg and nearksvgtheandcrackg1sur-
28 days. The sample submerged
humidity and desorption isotherm in the opposite in water shows face via diffusion process from
be calibrated by fitting experimental data relevant the bulk cementi- to
greatly enhanced their
case. Neglecting deflection capacity
difference (Xi aset well as stiff-
al. 1994), in tious material and the fractal
free (evaporable) water content in concrete at surface which may
ness recovery due
the following, to the isotherm
sorption healing products
will be presented
used with serve
various as ages
the calcite nucleation
(Di Luzio & Cusatis sites.2009b).
in the microcrack,
reference to both sorptionwhile and thisdesorption
is not theconditions.
case for 3. Self-healing behavior in SHCC heavily de-
specimens cured in air. The
By the way, if the hysteresis of the moisture ESEM and XEDS ob- pends on the availability of unhydrated cement and
servation further confirmed the
isotherm would be taken into account, two different finding in mechani- 2.2 Temperature
other supplementary cementitious materials. Low
cal tests.evaporable
relation, The mechanical water vsproperties are recovered
relative humidity, must water/cementitious
Note that, at early age, material
sinceratio and high reactions
the chemical percent-
after healing in an environment
be used according to the sign of the variation where water isofpre- the age of cementitious material appear
associated with cement hydration and SF reaction to promote self-
sent and thehumidity.
relativity cracks areThe filledshape
with reaction
of the products,
sorption healing behavior.
are exothermic, the temperature field is not uniform
see Figurefor
isotherm 4. HPC is influenced by many parameters, for4.non-adiabatic
Microcrack with systemssmallerevencrackif thewidth as in the
The following
especially those that conclusions
influencecan be drawn
extent and rate based on
of the SHCC
temperaturemixtures is preferable
is constant. Heatasconduction
far as continuous
can be
this investigation (Qian et al.
chemical reactions and, in turn, determine pore 2009): hydration-based self-healing is
described in concrete, at least for temperature concerned, as it notre-
1. For and
structure specimens
pore size submerged
distributionin water, the deflec-
(water-to-cement quires much less healing products
exceeding 100C (Baant & Kaplan 1996), by to fill the crack
tion capacity after self-healing
ratio, cement chemical composition, SF content, can recover about and it is law,
Fouriers muchwhich
for the healing products to
curingtotime 105% and compared with those mix
method, temperature, virgin speci-
additives, grow from both faces of the crack to get connected.
etc.). In the literature various formulations can air
while this ratio is about 40% to 60% for be q =Two
T points that still need to be improved
found specimens.
to describeFurthermore,
the sorptiontheisotherm stiffnessofofnormalinitial are the healing of larger cracks and the curing (7) of
concrete stage (Xiofetself-healed
al. 1994). specimenHowever, isinmuch larger
the present specimens in air. In figure 4 it can be seen that small
paper the with that of the airexpression
semi-empirical cured specimen proposed due byto where with
cracks q isa width
the heat
less thanflux,15 Tmisaretheno problem absolute
the presence of healing products
Norling Mjornell (1997) is adopted because it formed inside the temperature, and is the heat
to heal. However cracks around 60 m in width conductivity; in can
crack and strengthened the bridging fiber. only partially be filled with healing products. To re-
Proceedings of FraMCoS-7, May 23-28, 2010
duce the crack width further the effect of adding mi- A second option J =to Dpromote
( h, T )h self healing in a dry

crofibers to the PVA-fibre reinforced cementitious environment which is under investigation is the use
composite is investigated (Antonopoulou 2009, of Super Absorbent The Polymers (SAP) in thecoefficient
proportionality mix. The D(h,T)
Tziviloglou 2009). The microfibers used are steel SAPs are filled with
moisture permeability and it isproc-
water during the mixing a nonlinea
wool or rockwool fibres with a length of 2 mm and ess and form in such
of the a way waterhumidity
relative pocketshinand thetemperature
an average diameter of 8 micron. These small fibres crete that can be&used Najjarfor1972).
The of the cement
moisture mass balanc
help to distribute the cracks in the cement matrix and thus self healing
that the variation in timeSAPs
in a later stage. of the are water mas
even more and thus the result is smaller cracks, known as additivevolumeto mitigate autogenous
of concrete (watershrinkage
content w) be eq
which improves the self healing capacity. in concrete (Jensen & Hansen
divergence of 2001). The self
the moisture fluxheal-
The second negative point is that the self healing ing capacity of the SHCC is already improved by
mechanism of ongoing hydration only works if wa- adding these SAPs in specimens that are cracked
ter is present. To promote self healing also in a dry and subsequently stored = inJ water (Antonopoulou

environment two approaches are under investigation. 2009, Tziviloglou 2009). However these SAPs can

Hollow plant fibres, due to large storage volume for also work for specimens The water stored in air.w The
content can be water-
expressed a
liquid, can potentially be used for a new self healing pockets probablyofarethe emptied during or shortly
evaporable water we (capillary after wa
concrete system. The idea is to use the plant fibre as the first hydration.
vapor,When andthe materialwater)
adsorbed cracksandat the a non-e
a reservoir for healing agent, which can be water to later stage no water is left anymore.
(chemically bound)But water
after some wn (Mil
get ongoing hydration or which can also be a glue. rain on the structure the SAPs
Pantazopoulo & Mills 1995). located in the It is reas
Once a crack occurs, the healing agent will follow or cracked zone are assume
again filled thatandthethen slowly release
evaporable water is a fu
diffuse toward the crack and eventually heal the the water for the self healing mechanism. This
relative humidity, h, degree of hydration seems
crack. From the investigation, it was discovered that to be a realistic and
degree practical
of silicascenario which is cur-
fume reaction, s, i.e. we=w
self healing of concrete cracks would be potentially rently under investigation.
= age-dependent sorption/desorption
feasible using coated wood fibres (or actually fibre To optimize the(Norling
self healing mechanisms
Mjonell in SHCC
1997). Under this assum
bundles) that are filled with healing agent. The fibre materials a model by is developed (Schlangen et al.
substituting Equation 1 into Equati 2009)
bundles have a diameter of around 200m and a that is uses discreteobtains
fibers and can simulate distributed
length of about 10 mm (see also Sierra Beltran & cracking and ductile behaviour as shown in figure 6.
Schlangen, 2010). For the proposed self-healing sys- w h
tem to work, it is crucial that the fibres must be bro- e + ( D h) = we w
&c + e &s + w
ken to deliver the healing agent. In this investiga- h t h
c s
tion, the wood fibre bundles were first coated with
polysiloxane coating, then filled with a fluorescent where we/h is the slope of the sorption/
dye solution and finally sealed properly. The treated isotherm (also called moisture capac
fibre bundles were then fractured. The coated wood governing equation (Equation 3) must be
bundles tend to fail in a delamination mode (see fig- by appropriate boundary and initial conditi
ure 5) along their length and negative pressure force a) b) c) d)

The relation between the amount of e

caused by the sealed ends is not an issue, as it might water and relative humidity is called
be with continuous or short fibres that fail in a brittle isotherm if measured with increasing
way with a single crack plane (Joseph 2008). As a humidity and desorption isotherm in th
result, healing agent could be released from the

Figure 6. Simulationscase.
of SHCC material with
Neglecting theirfibres containing
difference (Xi et al.
splintered fibre bundles into the damaged areas healing agent. the following, sorption isotherm will be
where it subsequently repairs. referenceoftothe both sorption and
In figure 6 an example
By the way, simulations
if the of adesorption
ten-of thec
sile test and a four-point
isotherm bending
would teststaken
be are shown.
into heal-In two
the model of figure 6a and 6b only fibres with
ing agent are used.relation,
be Figure
6a shows the
according to
relative humi
of the varia
and figure 6b shows the fibres
relativity from which
humidity. The healing
shape of the
agent is released.isotherm
In figurefor 6c HPC
and 6d a hybrid
istoinfluenced mate-
by many p
rial containing mechanical
especially fibresthat
those reduce crack
influence extent and
width and distribute cracks reactions
chemical as well as and, healingin fibres
turn, determ
containing a healing agent and
structure are pore
used.size It isdistribution
shown that (water-
in the case distributed
ratio, cracking is
cementarechemical obtained much
composition, SF
more fibres in the material
curing time and activated
method, and the mix
healing liquid is used
etc.). inIna the
much more efficient
literature various way.
Figure 6e shows found a bendingto test of athespecimen
describe sorption con-
taining fibres with mechanical
concreteare(Xiused fibres
et al. and healing
agent. The simulations
paper the to 1994).
investigate the inproth
Figure 5. ESEM picture of a fibre-bundle broken in a delami- amount of fibresNorlingand the Mjornell
amount of(1997) healingisagentadopted b
nation mode. that is optimal.
Proceedings of FraMCoS-7, May 23-28, 2010
D (h, T )hPorous Asphalt Concrete
J = Healing
2.3 (1) explicitlyis shown
scheme accountswith forthetheeffect
of theofcapsules
After some years, asphalt binder is degraded by en- reaction
the asphaltand SF content.
concrete. When aThis cracksorption
close to aisotherm
The proportionality
vironmental coefficient
factors, especially due D(h,T) is called
to UV-radiation reads in the material of the road, the capsule will
from the sun, until it loses the ability to bind function
moisture permeability and it is a nonlinear the sur- break and the maltenes will be in contact with the bi-
of theparticles
face relative together.
humidity This h andresults
in cracks T (Baant
which tumen around. Then, by diffusion
both, maltenes and
& Najjar
allow 1972). The
damaging moisture
moisture intomass balancepavement
the lower requires damaged bitumen
we (rejuvenated
h c s ) = Gandwill
( )
mixed. The bitumen
+ will

that thecreating
levels, variationsurfacein timeroughness,
of the water pot massholes,per unit
degra- be , ,
c thes crack

easily closed.
c )h (4)
c capsules

volume of concrete (water content

dation and eventual structural failure. At present, w ) be equal to the With this autonomic method
e of using
1 the
divergence of the moisture flux
there are no solutions to close cracks in the pave- J self healing rate is increased a lot compared to the
(g of c )h

ment. Occasionally, when signs of ageing are visi- autogenic self healing capacity
K ( )e c asphalt.
c s
, 1

ble, J that protects asphalt surfaces from (2) en-

t degradation and moisture penetration is
applied to the surface. Other times, asphalt rejuvena- where the first term (gel isotherm) represents the
tors,Thewith water
capability w can of be expressed
changing theaschemical
the sum physically bound (adsorbed) water and the second
of the evaporable water w (capillary
composition of bitumen, aree applied to the surface. water, water term (capillary isotherm) represents the capillary
All theseandprocedures
adsorbed can water) and the
increase thenon-evaporable
lifetime of as- water. This expression is valid only for low content
phalt for severalbound) years before wn (Millsor recon-
waterrehabilitation 1966, of SF. The coefficient G1 represents the amount of
struction & Mills
is required, but 1995).
they have It the
is reasonable
disadvantage to water per unit volume held in the gel pores at 100%
assume that the evaporable water
that they only work in the first centimeters from the is a function of relative humidity, and it can be expressed (Norling
surface humidity,
can reduce h, degree
sliding of hydration, c, and
resistance. Mjornell 1997) as
It is of silica fume
generally known reaction, s, i.e. roads
that asphalt we=wecan (h,heal
= themselves,
by age-dependent but it sorption/desorption
is a slow process at isotherm ambient
(Norling Mjonell
temperature, and it1997). Under ifthisthere
only works assumption
is no trafficand c c + k s s showing the working of(5)
G ( c7.Schematic
Figure ) = k representation en-
by substituting s
capsulated oil to healccracksvgin sasphalt concrete.
circulation on theEquation
road. It is 1also intowellEquation
known that 2 one
the 1

obtains of healing increases when the material is

subjected to a higher temperature during the rest pe- whereMany and ksvg areexplain
kcvgresearches material how parameters. From the
to make capsules,
riodw(Bonnaure et al. 1982).
w Therefore maximum
but asphaltamountconcreteofiswater
a realperharshunit environment
volume that can for
e to+ investigated e &increasew
e &aof project was
&n self-
c + s + wthe
started ( D h) = the (3) fill all
the pores (both
capsules capillary
to survive. porescapsules
These and gel should
pores), one
h t h can calculate
capsulate 1 as one hydrocarbures
veryKviscous obtains based oil, they
healing rates of the road, which c is as
s a good method
to increase the lifetime of the pavement. As a simpli- should not react with bitumen, and they should resist
where wbitumen
fication, e/h is the slopebeof the
could sorption/desorption
considered as a very the mixing process with the aggregates and the
g h

tumen andw the compaction

s + s Gon the
e road at about
dense (also a crack
oil; when called appears
moisture in it,capacity).
it will closeThe by 0.188
c 0.22



itself, but itequation
will do much (Equationfaster3)ifmust be completed
the liquid behav- 180oC. Besides, they
should not be so resistant that

K ( c never
they ) = break. To solve this, maltenes have been
by appropriate
ior of bitumenboundaryis increased. and That
be done by in- 1
g h

The relation
creasing between or
its temperature thebyamount mixingof with evaporable
a less encapsulated in very porous e c sand
1 c and

covered with a
composite made of a thermoresistant resin and very

water oil.
dense and relative humidity is called adsorption
In this paper,if measured
two fairlywith new increasing relativity
ideas are presented: fineThe sand. Research
material to provekcthe
parameters self shealing capac-
vg and k vg and g1 can
humidity and desorption isotherm
Induction heating of asphalt concrete and microcap- in the opposite ity of this new material incorporating
be calibrated by fitting experimental data theserelevant
capsules to
case. filled
sules Neglecting
with a their
healing difference
agent (Garcia(Xi,
2009a). in isfree
(evaporable) water content in concrete at
the following,
Both of them tosorption
increase isotherm
the self-healing will beratesusedofwithas- various ages (Di Luzio & Cusatis 2009b).
phalt to both
concrete andsorption
hereby, and desorption
the lifetime of theconditions.
road. 2.3.2 Induction Heating
By the way, if the hysteresis of the moisture The basis for the second approach that is followed is
2.3.1 would be taken into account, two different
Capsule-method 2.2 Temperature
heating the asphaltevolution
with induction energy to increase
relation, evaporable
Bitumen can be considered water vsasrelative
a two humidity,
phase material must its
Note that, at early age,first
healing rate. The sinceprerequisite
the chemical of induction
be used
with according
a liquid phase, to the maltenes,
called sign of theand variation of the
a solid phase, heating is that the heated
associated with cement hydration and SF material must be conduc-
called humidity.With
asphaltenes. The time,shapetheofliquid the phase
sorption is tive. In many previous
are exothermic, studies itfield
the temperature has isbeen demon-
not uniform
isotherm for
oxidized, HPC is influenced
disappearing and causing by many parameters,
asphalt to be- strated how it is possible
for non-adiabatic systemstoeven makeifasphalt or concrete
the environmental
come dry andthosebrittle.
that influence
To avoidextent and rate ofhave
this, maltenes the conductive by adding electrically
temperature is constant. Heat conduction can conductive fillers
been reactionsapplied
traditionally and, on in theturn,road determine
surface once pore and fibers (Garcia et al. 2009b).
described in concrete, at least for temperature notThe second prereq-
signs and pore
of ageing startsizeappearing.
distribution The(water-to-cement
problem is that uisite
exceeding is that100C
these fillers
(Baant and &fibers are connected
Kaplan 1996), by in
this cement chemical
of treatment composition,
is superficial, withSFwhat content,
only closed-loop
Fouriers law, circuits.
which In Figure 8 a schematic repre-
the firsttime and method,
centimeters fromtemperature,
the surfacemix are additives,
affected. sentation is given of the system in which inductive
To In the
solve this,literature
it was thought variousthat formulations
the optimum canway be energy
q = isT used for the healing of asphalt concrete.
First a microcrack appears in the bitumen. If enough (7)
found to describe the sorption
of adding maltenes to the road would be by mixing isotherm of normal
concrete filled
capsules (Xi etwith al. 1994).
maltenes However,
with theinasphalt the present
con- volume of conductive fibers or fillers is added they
paper With the semi-empirical expression proposedover by where
will is the heatcircuits
formq closed-loops flux, allT aroundis thetheabsolute
crete. this, aging effects could be avoided
the complete Mjornell
depth (1997) is adoptedIn because
of the pavement. Figure 7 ita crack. Then, if this magnetically susceptiblein and
temperature, and is the heat conductivity; this
electrically conductive material is placed in the vi-
Proceedings of FraMCoS-7, May 23-28, 2010
cinity of a coil, eddy currents are induced in the D (h, T )hhealing and reloading
these samples afterJ =6heating,
closed-loops circuits, with the same frequency of the cycles. The samples were frozen at -20 C during the
magnetic field. Heat is generated through the energy tests to avoid creep; The so theproportionality
test specimenscoefficient
had brittle D(h,T)
lost when eddy currents meet with the resistance of fracture (the elastic moisture permeability and it is a innonlinea
modulus is clearly appreciated
the material and, finally, bitumen is melted and the the curves). During of the therelative
and temperature
crack is closed. heating) it could be seen that the cracks disappeared.
& Najjar 1972). The moisture mass balanc
In Figure 9, the that resistance of the samples
the variation in time ofafter the the
water mas
fifth healing is about 70 % of the original
volume of concrete (water content one. Be-w) be eq
sides, the elastic modulus
divergence is very similar,
of the moisturebut the
after the ultimate strength is steeper in the healed
samples. This is logic =because the healed zone is a
weak zone where all fibers
the are broken, so the

sample is more brittle in that section.

The water content w can be expressed a

of the evaporable water we (capillary wa
and adsorbed water) and the non-e
(chemically bound) water wn (Mil
In this paper selfPantazopoulo
healing techniques & Mills in three
It is reas
Figure 8. Schematic representation showing the system of us- ent materials are assume
discussed. that the evaporable wateris is a fu
The first application
ing fibres heated with induction energy to heal cracks in as- using Bacteria to relative
precipitate calcite inh, cracks
humidity, degreeinofcon- hydration
phalt concrete. crete. With this method
degree relatively
of silica fume largereaction,
cracks insre- , i.e. we=w
inforced concrete=can age-dependent
be filled. The method does not
This research is being conducted in three steps. lead to strength improvements
(Norling Mjonell 1997). Under but
of the structure, this assum
First, the optimum combination of conductive fibers by filling the crack, by substituting Equation 1 intois Equati
the path to the reinforcement
and fillers was chosen; then, the temperature reached blocked. Herewith the ingress of liquids and ions
depending on the electrical conductivity was found. that start reinforcement corrosion is stopped and
Finally to prove the healing, the samples will be re- thus the durability of w h
the structure is enhanced. w
peatedly broken in three point bending and heated thermore this method e is useful for water eretaining
& e & + w
+ ( D h ) = c +
and healed again. It has been found that electrically h t h
structures. Cracks can be filled in this way cand leak-s s
conductive fibers are much more efficient that fillers age can be stopped. Especially in underground struc-
when increasing the conductivity and that there is an tures were repair whereis difficult or isimpossible
the slope Bacterial
of the sorption/
optimum in the electrical conductivity, related to a future.we/h
concrete has a bigisotherm (also called moisture capac
certain volume of fibers, above which it is difficult In the secondgoverning
application SHCC(Equation
equation materials3)are must be
to increase the conductivity. It has been discovered studied, which have by already a high
appropriate potential
boundary and for self conditi
also that although it is not necessary that the sample healing because of The theirrelation
small crackbetween widths.
the New of e
is conductive to heat it, the optimum volume of fi- additions, like microfibers
water and and SAPshumidity
relative even promote is called
bers for heating coincides with the optimum volume this self healing capacity further.
isotherm if measured with increasing
of fibers to have the maximum conductivity. The third application
humidityisand fordesorption
asphalt concrete isotherm in in th
which the self healing capacity is enlarged
case. Neglecting their difference (Xi et using
encapsulated oil the andfollowing,
micro-steel fibres. isotherm
sorption The latterwill be
approach has beenreference
proven to to both
in the laboratory
and desorption c
and will be applied Byinthe a real roadifin the the Netherlands
100 %

way, hysteresis of the

in 2010. isotherm would be taken into account, two
0,1 70 %
relation, evaporable water vs relative humi
be used according to the sign of the varia


relativity humidity. The shape of the

e 0,08

isotherm for HPC is influenced by many p


Financial support especially

from the Delft thoseCentre for Materials

that influence extent and

chemical reactions and, inNovem
and Senter turn, determ
0,02 IOP (
structure and pore size distribution (water-
for this work is gratefully
ratio, cement acknowledged.
chemical composition, SF
curing time and method, temperature, mix

Figure 9. Force-deformation diagrams of bitumen specimen

with steel-wool fibres loaded in 3 point bending, in 6 cycles of etc.). In the literature various formulatio
loading-heating and reloading. REFERENCES found to describe the sorption isotherm
Antonopoulou, S. Self concrete
healing in(Xi ECCetmaterials
al. 1994). However,
with high con- in th
Figure 9 gives an indication of the healing capacity paper the semi-empirical expression pro
of the bitumen. The figure shows the load-deformation tent of different microfibers and micro particles, MSc The-
Norling Mjornell
sis, Delft University of Technology, 2009. (1997) is adopted b
curves of four samples and the stress-strain curves of
Proceedings of FraMCoS-7, May 23-28, 2010
D (h, TGalinat,
J = S.S.,
Bang, ) h J.K., Ramakrishnan, V., 2001. Calcite pre- (1) explicitly
Joseph,, accountsA.D.
C., Jefferson, forandtheLark, evolution of hydration
R.J., Lattice modelling
cipitation induced by polyurethane-immobilized Bacillus reaction and SF content. This sorption isotherm
of autonomic healing processes in cementitious materials.
pasteurii. Enzyme Microb. Technol. 28, 404-409. WCCM8 / ECCOMAS 2008, Venice, Italy, (2008).
The proportionality
Bonnaure, F.P., Huibers, A.H., coefficient
Boonders, A., D(h,T) is called
A laboratory in- Li, V.C., Lim, Y.M. and Chan, Y.W., Feasibility study of a
vestigation permeability
of the influence and ofit rest
is a periods
nonlinearon thefunction
fatigue passive smart self-healing cementitious composite. Com-
of characteristics
the relative humidity of bituminous h and temperature
mixes, Journal of the T (Baant
Associa- posites Part B: Engineering, vol. 29(B), pp. 819-827,
& Najjar of Asphalt
1972). Paving Technologists,
The moisture mass 51, (1982),
balance 104-128.
requires (1998).
De Muynck, W., Debrouwer, D., De Belie, N.,mass
Verstraete, W., we (h c T.,s )Mihashi,
Nishiwaki, = G ( c H., ) B-K. & Miura, K. Develop-
s Jang,
the Bacterial
variation in time of the water perdura-
, , , 1

c )h selective
carbonate precipitation improves the ment of self-healing system
eforAdvancedc Concrete

bility ofof cementitious
concrete (water content
materials. Cementw) be equal toRes.
& Concrete the heating around crack. Journal

38, 10051014. of the moisture flux J nology, Vol. 4, No. 2 (2006) 267-275.
Dry, C. M. 2000. Three designs for the internal release of seal- Neville A. (2002).KAutogenous
( ) e healing
1 A concrete
c miracle?
ants, adhesives, and waterproofing chemicals into concrete Concrete International, c November
s 2002. 1

tow =reduce (2)

J permeability. Cement and Concrete Research Reinhardt H.W., Joos M. (2003). Permeability and self-healing
pp. 1969-1977. of cracked concrete as a function of temperature and crack
Edvardsen C. (1999). Water permeability and autogenous heal- width.the
where Cement
(gel Research,
isotherm) 33(7), 981-985. the
ing crackscontent w canACI
in concrete. be Materials
as the96(4),
sum Qian , S., Zhou, J., de Rooij, M.R., Schlangen,,represents
the evaporable water we (capillary water, water
of 448-454. physically
van Breugel, bound (adsorbed)
K. 2009, Self-Healing andE.,theYe,ofsecond
waterBehavior G. and
vapor, A,andSchlangen,
adsorbed E &water)
Ven, M.and van the
de (2009a). Two ways
non-evaporable term (capillary
Hardening isotherm)
Cementitious represents
Composites the capillary
Incorporating Local
of closing
(chemically cracks on asphalt
bound)heating. concrete pavements:
waterKey wengineering(Millsmicrocap-
1966, water.
WasteThis expression
Materials. Cementis and
only forComposites,
low content 31
sules and induction
Pantazopoulo & Mills 1995). It
reasonable to of SF. The coefficient G1 represents the amount of
(2009) 613-621.
417-418, 573-576. Schlangen,
water per E.,Qian,
unit volume S. & held
Liu, H. in 2009, gelSimulation
theMaterial.pores of100%
assume that the evaporable water
Garcia, A, Schlangen, E., Ven, M van de & Liu, Q (2009b). is a function of Healing
relative Capacity of Hybrid Fibre 2nd
conductivity h, degree
of asphaltofmortar
hydration, c, con-
containing and tional humidity,
Conference and it can
on Self be expressed
Healing Materials, (Norling
degree of silica fume reaction, s, i.e. we=we(h,materi-
ductive fibers and fillers. Construction and building c,s) Mjornell
2009. 1997) as
= als,age-dependent
23, 3175-3181. sorption/desorption isotherm
S., Loukili A., Pijaudier-Cabot G., Behloul M. (2006).
Schlangen, E. & Joseph, C (2008). Self-healing processes in
concrete. In cSK Ghoshs (Ed.), Self-healing materials: fun-
(Norling Mjonell
Self healing of cracks1997). Under from
in concrete: this aassumption
model materialand to G damentals,
c s k vg c cstrategies
( ) = design
, + k and
vg s s applications (pp. 141-182). (5)
by usual
substituting Equation of
concretes. Proceedings 1 the
international2 sym-
one 1
Weinheim: Wiley-vch.
posium on advances in concrete through science and engi- Sierra Beltran M.G.s & Schlangen, E. Fibre-matrix interface
neering. Quebec city, Canada, 11-13 September 2006. where kcvg and
properties vg arefibre
in akwood material
reinforced parameters. FromThis
cement matrix. the
w h
S.K. (ed), Self-healing w
materials;wfundamentals, design maximum amount of water per unit volume that can
conference, 2010.
strategies and applications,
+ ( D h ) = e & +2008.e & + w&
fill all pores K.,(both
opurolu, O. and
capillary poresFraaij,
andA.L.A., Durability
gel pores), one
h A, t Kishi,, T. h Arita,H. andc Takakuwa,
s Y., nSelf heal- of Blast-Furnace Slag
calculate K1 asSolution
canofluorophosphate Mortars
one obtains Subjected to Sodium Mon-
c s
ing of crack and water permeability of expansive concrete. Curing, Proc. 4th International
1st international conference on self-healing materials. Conference on Construction Materials: Performance, Inno-
we/hHolland,is the (2007).
slope of the sorption/desorption vations and Structural Implications, Nagoya, Japan, 2009.
g h



isotherm O.M. &
(also Hansen,
called P.F. moisture
Water-entrained cement-based
capacity). The Ter Heide, N.w Crack healing
s + in hydrating econcrete. MSc-thesis,
s s GThe
Delft University of cTechnology,
0.188 0.22 1
materials: I. Principles and theoretical background, Cem. Netherlands, 2005.
governing equation (Equation 3) must be completed (6)
0 1

Concr. Res. 31 (4) (2001) 647 654. Ter
K (Heide,
) =N., Schlangen, E. & van Breugel,
K. Experimental
by appropriate
Jonkers, H & Schlangenboundary E. and initial conditions.
in Schmets A.J.M. & van der 1 c ,
Study of Crack Healing of g Early


h Cracks, In Proceed-

The relation
Zwaag, between the
S. (eds) Proceedings amount
of the of evaporable
First International Con- ings Knud Hjgaard e conference
c c on
1 Advanced Cement-

water andonrelative
ference humidity
Self Healing is called
Materials, 18-20adsorption
April 2007, Based Materials, Technical University of Denmark, June
isotherm Zee, The Netherlands,
with Springer 2007.
increasing relativity 2005.
The material parameters and ksvgwith
kcvgmaterials andlow content
g1 can
humidity HM and& Schlangen,
desorption E. (2009a).
isotherm Bacteria-based
in the self-
opposite Tziviloglou,
be ofcalibrated E. Self-healing
by fitting in ECC
experimental data relevant to
healing concrete. International journal of restoration
case. Neglecting
buildings their difference
and monuments, (Xi et al. 1994), of
15(4), 255-265. in different microfibres and micro-particles, MSc Thesis,
freeTU (evaporable)
Delft, 2009 water content in concrete at
the following,
Jonkers, sorption
HM, Thijssen, A, isotherm
Muijzer, G,will be used Owith
Copuroglu, & various
Van agesK.,(DiIsLuzio
Breugel, there & Cusatisfor2009b).
a market self-healing cement-
Schlangen, to both sorptionApplication
E. (2009b). and desorption conditions.
of bacteria as self- based materials?, 1st international conference on self-
the way, agent forif thethedevelopment
hysteresisof sustainable concrete.
of the moisture healing materials. Noordwijk, Holland, 2007.
2.2 der Temperature evolution
Ecological engineering, 1-6.
isotherm would be taken into account, two different Van Zwaag, S. (ed), Self healing materials : an alternative
Joseph, C. Experimental and numerical study of fracture and approach to 20 centuries of materials science, Dordrecht,
relation, evaporable
self healing water materials.
of cementitious vs relativePhD humidity, must
thesis, Cardiff Note that, at early
Netherlands, Springer,age,2007.
since the chemical reactions
be University,
used according (2008). to the sign of the variation of the associated with cement hydration and SF reaction
relativity humidity. The shape of the sorption are exothermic, the temperature field is not uniform
isotherm for HPC is influenced by many parameters, for non-adiabatic systems even if the environmental
especially those that influence extent and rate of the temperature is constant. Heat conduction can be
chemical reactions and, in turn, determine pore described in concrete, at least for temperature not
structure and pore size distribution (water-to-cement exceeding 100C (Baant & Kaplan 1996), by
ratio, cement chemical composition, SF content, Fouriers law, which reads
curing time and method, temperature, mix additives,
etc.). In the literature various formulations can be q = T (7)
found to describe the sorption isotherm of normal
concrete (Xi et al. 1994). However, in the present where q is the heat flux, T is the absolute
paper the semi-empirical expression proposed by temperature, and is the heat conductivity; in this
Norling Mjornell (1997) is adopted because it

Proceedings of FraMCoS-7, May 23-28, 2010