Cement and Concrete Research 42 (2012) 1286–1297

Contents lists available at SciVerse ScienceDirect

Cement and Concrete Research

journal homepage: http://ees.elsevier.com/CEMCON/default.asp

Evaluation of the effects of drying shrinkage on the behavior of concrete structures

strengthened by overlays
Y. Theiner ⁎, G. Hofstetter
Institute of Basic Sciences in Civil Engineering, University of Innsbruck, Technikerstraße 13, A-6020 Innsbruck, Austria

a r t i c l e i n f o a b s t r a c t

Article history: The present contribution focuses on the experimental evaluation of the effects of drying shrinkage on the be-
Received 18 October 2011 havior of concrete structures strengthened by overlays. To this end a comprehensive laboratory test program
Accepted 6 June 2012 is presented. Tests on thin concrete slices served for determining the water desorption isotherm and the ul-
timate drying shrinkage strains. The time and depth dependent mass water content distributions and the
Keywords:
evolution of the drying shrinkage strains were measured on concrete prisms and on larger brick-shaped con-
Concrete (E)
Drying (A)
crete specimens during two years of drying. After two years of drying the brick-shaped specimens were sup-
Humidity (A) plemented by a concrete overlay. Measurements of the mass water content distribution were continued
Shrinkage (C) during water jetting and subsequent wetting of the top surface. In addition, the shrinkage strains were
Mass water content recorded in the composite specimens during subsequent drying.
© 2012 Elsevier Ltd. All rights reserved.

1. Introduction Drying shrinkage of concrete is the time-dependent volume de-

crease due to moisture migration and moisture transfer to the envi-
Shrinkage is the time-dependent volume decrease of concrete ronment caused by a change in relative humidity. For normal
without the action of external loads. It is caused by chemical, thermal strength concrete it contributes the main part of the volume change
and hygral processes. Basically, four different types of shrinkage can without any actions of external loads [3].
be distinguished [1–4]: very early age shrinkage (also called capillary There exists a large body of literature regarding shrinkage of
or plastic shrinkage), autogenous shrinkage, carbonation shrinkage cementitious materials. Hansen et al. [11] investigated the influence
and drying shrinkage. of drying shrinkage of paste, of aggregate content and of modulus
The very early age shrinkage takes place in fresh concrete due to ratio of aggregates and hydrated cement on the ultimate drying
transfer of moisture from the surface to the environment by evapora- shrinkage of concrete. Chaallal et al. [12] compared experimental re-
tion and due to mass transfer from inside the concrete to the surface sults of drying shrinkage strains to estimated values based on ACI
[5,6]. Very early age shrinkage, which is commonly measured within [13] and Euro-International Committee of Concrete (CEB) [14] recom-
the first 24 h, can be avoided by an appropriate mix design and by mendations. The main impetus was to compare drying shrinkage
proper curing procedures. strains recorded for a large-scale concrete block with the shrinkage
Autogenous shrinkage refers to the effect of volume reduction strains computed on the basis of the ACI and CEB codes. Bissonnette
caused by cement hydration. Hence, it can be observed in sealed con- et al. [15] carried out shrinkage tests on specimens of cement pastes,
ditions without any moisture transfer from the concrete to the envi- mortars and concretes and evaluated the influence of different key
ronment. For normal strength concrete with a w/c-ratio greater parameters on drying shrinkage such as the size of specimens and rel-
than 0.4 volume changes due to autogenous shrinkage are small com- ative humidity. In [16] Baroghel-Bouny et al. presented equilibrium
pared to volume changes due to drying shrinkage [7–9]. and transfer moisture properties of high-performance cementitious
The reaction between carbon dioxide and the hardened cement materials and compared them with the respective properties of ordi-
paste results in carbonation shrinkage. The volume reduction caused nary cement pastes and concretes. The equilibrium moisture proper-
by carbonation shrinkage is very slow and is restricted to the periph- ties of the hardened materials are described by means of water
eral zone. It can be neglected compared to drying shrinkage [3,10]. sorption isotherms, whereas the experimentally determined drying
shrinkage strains are reported as function of relative humidity. Fur-
thermore, a thermodynamic model based on the capillary curve and
the intrinsic permeability of concrete was provided. After identifica-
tion of the intrinsic permeability, numerically predicted moisture
⁎ Corresponding author. Tel.: + 43 512 507 6587; fax: + 43 512 507 2908. profiles were compared with gamma-ray attenuation measurement
E-mail address: Yvonne.Theiner@uibk.ac.at (Y. Theiner). data. In [17] design-provisions for calculating the shrinkage and

0008-8846/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cemconres.2012.06.001
 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297 1287

creep strains of concrete from information available at the design were tested in the laboratory [28]. Parts of the test program were sim-
stage, namely, the 28-day concrete strength, the concrete strength ulated numerically. To this end a numerical model for concrete cracking
at loading, element size, and relative humidity, are presented. The was employed consisting of a rotating smeared crack model in combi-
combined effects of self-desiccation and “external” drying were stud- nation with a crack model based on the strong discontinuity approach
ied experimentally in [9] by submitting concrete specimens to drying and formulated within the framework of finite elements with embed-
at early age. Furthermore, an experimental study on “pure” drying ded discontinuities [29]. The numerical model and the numerical simu-
shrinkage, performed on a few-millimeter thick specimens of hard- lations are described in [30]. The load history, considered in the
ened cement pastes and concretes was presented. Ayano and numerical simulations, consisted of (i) loading of the original test
Wittman [18] described both an experimental approach for deter- model, which induces pre-damage due to cracking, (ii) unloading, ap-
mining the time-dependent relative humidity in the pore system of plication of a concrete overlay and (iii) loading of the strengthened
concrete and a numerical method for determining material parame- structure up to failure. Drying shrinkage of the concrete overlay was
ters on the basis of experimental data. The hygral diffusion coefficient taken into account in the numerical simulation by the commonly
was expressed as function of moisture content and as function of rel- employed engineering approach considering uniform shrinkage strains
ative humidity. Based on a phenomenological approach a model for as initial strains. The latter were measured on concrete specimens in the
the delayed deformations was proposed in [19]. Experimental studies laboratory, made from the concrete used for the overlay. Neglecting
highlighted the close links between delayed deformations under load drying shrinkage in the numerical simulation revealed the expected in-
and shrinkage deformations. In the authors opinion the analysis car- fluence of shrinkage of the concrete overlay on the structural behavior
ried out leads to consideration of shrinkage as creep due to capillary of the strengthened slab strip model. Expectedly, if drying shrinkage
stress. In [20] Videla et al. address the need to update existing con- was neglected, then cracking in the overlay was initiated at a higher
crete drying shrinkage prediction models to reflect the properties of load level and the measured maximum crack widths were under-
local materials. The experimental program was carried out using estimated in the numerical simulation. The employed engineering ap-
eight concrete mixtures, manufactured with commonly used cements proach for considering drying shrinkage of concrete overlays is
in Chile. In 2007 Baroghel-Bouny published two papers concerning characterized by the deficiency considering only a mean value of the
water vapor sorption experiments on hardened cementitious mate- shrinkage strains neglecting the dependence on the change of moisture
rials [21,22]. In the first part [21] a broad range of normal and high- distribution due to a change of relative ambient humidity.
performance hardened cement pastes and concretes were studied However, for an improved representation of the drying shrink-
under both laboratory and in-situ conditions. Water vapor desorption– age strains in composite structures detailed experimental data re-
adsorption experiments were carried out on very thin specimens by garding the evolution of both the moisture distribution and the
means of the saturated salt solution method. Furthermore, moisture drying shrinkage strains in the overlay and the adjacent substrate
profiles in structural elements exposed to various drying conditions concrete has to be determined. This task is the aim of the present
in the laboratory and in natural environments were assessed by contribution.
means of gamma-ray attenuation measurements. In order to evaluate
the influence of governing parameters in the second part [22] the
transport properties of various hardened cementitious materials 3. Experimental study
were studied experimentally and theoretically. In [23] Bissonnette
et al. described that warping and joint opening of concrete slabs in Although the shrinkage mechanisms in concrete members are
a controlled environment develop early and that the respective rate well known, the effects of drying shrinkage on the performance of
is proportional to that of drying shrinkage. The moisture changes concrete structures strengthened by concrete overlays have not
and strains induced by water desorption of several cement based ma- been fully clarified so far. Since drying shrinkage of a concrete overlay
terials were studied in [24]. Experiments were conducted on ordinary may have a considerable effect on the behavior of a strengthened con-
mortars and cement pastes with water-to-cement ratio of 0.5 and 0.8, crete structure [25–27], it is the aim of this contribution to present a
respectively. Finally, a numerical model, based on experimental por- detailed experimental investigation of the effects of drying shrinkage.
omechanical results, was proposed for predicting shrinkage for mate- Whereas in [26,27] the experimental study was restricted to the over-
rials submitted to drying. lay of composite specimens, in the present contribution the effects of
Drying shrinkage of a concrete overlay has a considerable effect on drying shrinkage will be investigated beginning with hardening and
the behavior of a strengthened concrete structure. In particular, the drying of the original specimen, during pre-treatment of the interface
shrinkage strains, which develop during the first days and weeks, are by water jetting and placement of the overlay as well as during hard-
important for bonded concrete overlays. In [25] a computer-based de- ening and drying of the overlay.
sign tool was presented which allows calculating the stress distribution Hence, in this contribution a comprehensive laboratory test pro-
in a bonded concrete overlay due to volume changes and restrained gram including (i) shrinkage tests conducted on thin concrete slices,
conditions at early age. In [26] extensive experimental measurements (ii) shrinkage tests conducted on concrete prisms, (iii) shrinkage
and numerical analyses were carried out in order to investigate the tests conducted on larger brick-shaped specimens with added con-
early age behavior of bonded concrete overlays. Within this time span, crete overlays as well as (iv) shrinkage tests conducted on reference
restrained shrinkage may result in tensile stresses, exceeding the early specimens will be presented in detail. The tests focus on normal
age tensile strength of the overlay and, hence, causing cracking of the strength concrete. On the basis of the mentioned experiments the
overlay. It was found that bonded concrete overlays with high- water desorption isotherm, ultimate drying shrinkage strains in
performance concrete mixtures have a strong tendency to early-age terms of relative humidity, depth dependent mass water content dis-
debonding. However, as shown in [27], the tensile stresses in the over- tributions and drying shrinkage strains in terms of drying time will
lay are reduced by stress relaxation in the overlay. Beushausen et al. [3] be determined. The measurement data will contribute to a better un-
used composite specimens for identifying fundamental characteristics derstanding of the role of drying shrinkage on the behavior of con-
of strain development in bonded concrete overlays, considering differ- crete structures strengthened by concrete overlays and will
ent interface textures and overlay materials. diminish the risk of shrinkage induced cracks in overlays. Hence,
In an applied research project, dealing with the experimental and the experimental evaluation of the effects of drying shrinkage on
numerical investigation of the structural behavior of pre-damaged con- the behavior of strengthened concrete structures is a first step to-
crete bridge decks strengthened by concrete overlays, large-scale ward the understanding of the complex structural response of con-
models, representing strips of a real bridge deck on the scale of 1:2, crete structures strengthened by overlays.
1288 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297

Table 1 All specimens were moist cured for 81 days after casting. Hence,
Composition of concrete mix (mass proportions). hydration can be assumed to be almost completed and autogenous
Composition shrinkage can be assumed to have attained its ultimate value at the
beginning of testing.
Cement type: CEM II AM (S-L) 32,5 R 340 kg/m³
w/c-ratio 0.50
Superplasticizer type: Glenium SKY 581 0.9% 3.1. Testing program
Aggregate size distribution 0/4 mm, 4/8 mm, 4/16 mm
45%, 30%, 25%
3.1.1. Shrinkage tests on concrete slices
For determining the water desorption isotherm and the drying
shrinkage strains at different values of relative ambient humidity (RH)
10 thin concrete slices with dimensions of 110 × 110 × 6 mm (Fig. 1)
Table 2 were submitted to drying at constant temperature of (20 ± 2)°C. Indi-
Selected material properties at different concrete ages.
vidual slices were obtained from a concrete prism with dimensions of
Material 14 days 28 days 56 days 112 days 365 days 110 × 110 × 400 mm by wet sawing. Starting at 100% RH, the RH was re-
properties duced step by step to 90%, 80%, 65% and 50%, respectively, and the
fcm,cube 50.0 N/mm² 55.0 N/mm² 62.0 N/mm² 74.0 N/mm² 83.0 N/mm² corresponding equilibrium mass water contents were determined.
fctm 4.3 N/mm² 4.8 N/mm² 5.0 N/mm² n.a. 5.6 N/mm² The mass water content, representing the ratio of the mass of water in
Ecm 35,164 37,483 39,224 41,188 41,422 the specimen to the mass of the dry specimen (desiccated at 105 °C),
N/mm² N/mm² N/mm² N/mm² N/mm²
was determined by weighing. Contrary to the present investigation, in
[16,21] the dry reference state for calculating the mass water content
was defined by the equilibrium state at 3% RH and 23 °C. However, in
[21] the consequences of different definitions of the dry reference
Shrinkage tests on thin concrete slices served for determining the state for determining the water desorption isotherm were emphasized.
water desorption isotherm as well as the drying shrinkage strains in Significantly higher water contents were reported in [21] when the
terms of ambient relative humidity. Multi-Ring-Sensors embedded specimens were desiccated at 105 °C. Constant mass of the specimen
in the concrete prisms were employed for determining the depth de- was assumed according to [31], i.e. when the mass changes are smaller
pendent moisture distribution. The test set-up for the shrinkage tests than 0.1% of the total mass in three subsequently conducted
on thin concrete slices and on concrete prisms was based on [16]. measurements.
However, contrary to the Multi-Ring-Sensors applied in the present Each slice was equipped with 8 measuring discs highlighted by red
investigation, in [16] the distribution of the moisture content was circles in Fig. 1. They served for measuring distances along the four
measured by gamma-ray attenuation. Furthermore, in the present in- symmetry lines of the concrete slices as shown in Fig. 1. To this end
vestigation the drying shrinkage strains were also measured on the an extensometer with a measuring accuracy of 0.003 mm and a
concrete prisms. rated measuring length of 100 mm was employed. Based on the elon-
According to [3,27] different shrinkage of the substrate and the gations, computed from the measured distances, the respective strain
overlay significantly influence the serviceability and durability of was determined. As the samples were of sufficient age that autoge-
the composite structure. Hence, in the present investigation Multi- nous shrinkage can be assumed to have attained its ultimate value
Ring-Sensors were arranged close to the interface between brick- (as described above), the deformations measured can be considered
shaped specimens and concrete overlays providing measurement as drying shrinkage strains.
data for moisture distribution and migration in composite specimens.
Comparison of the moisture distributions and shrinkage strains mea- 3.1.2. Shrinkage tests on concrete prisms
sured in concrete overlays and in reference specimens finally allowed 5 specimens with dimensions of 100 × 100 × 56 mm (Fig. 2) were
evaluating the restraint effects due to the interface between the over- equipped with Multi-Ring Sensors (MRS, Fig. 3), temperature detec-
lay and the original specimen. tors and a number of measuring discs. In order to obtain one-
The composition of the concrete mix is summarized in Table 1. Se- dimensional moisture migration the lateral surfaces were sealed and
lected material parameters, as the cube compressive strength fcm,cube, the top and bottom face of the samples were submitted to drying at
the direct uniaxial tensile strength fctm and the elastic modulus Ecm, a relative humidity of (65 ± 5)% and a temperature of (20 ± 2)°C.
determined at the age of 14 days, 28 days, 56 days, 112 days and Multi-Ring-Sensors (MRS) [32] were applied for determining the
365 days are provided in Table 2. The mentioned material parameters depth dependent moisture distribution of the drying samples by mea-
represent the respective mean values determined on 5 specimens. suring electrolytic resistances. Each MRS consisted of 8 measuring

Fig. 1. Concrete slices for determining the water desorption isotherms (left) and detail of a thin concrete slice with eight measuring discs and four measurement lengths for deter-
mining the drying shrinkage strains (right).
 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297 1289

Fig. 2. Concrete prisms with sealed lateral surfaces, equipped with MRS, temperature detectors and measuring discs (left), and detail of a concrete prism with measuring discs and
measurement lengths for determining the drying shrinkage strains (right).

points (MP) over a length of about 35 mm (Fig. 3). Calibration curves shaped specimens is not uniform along the depth, the measured
were used for determining the mass water content from the mea- shrinkage strains represent mean values along the measuring lengths.
sured electrolytic resistances. Fig. 5 shows the arrangement of four MRS (MRS 6 to MRS 9) used
The drying shrinkage strains of the concrete prisms were determined in the original concrete for one brick-shaped sample. Two of them
using an extensometer with a rated measuring length of 50 mm. The were placed consecutively starting from the concrete surface,
measurement lengths are shown in Fig. 2. Since during drying the mois- i.e. MRS 6 below MRS 7 and MRS 9 below MRS 8. Hence, the mass
ture distribution in the prisms is not uniform along the depth, the mea- water content distribution was determined to a depth of about
sured shrinkage strains represent mean values along the measuring 80 mm below the interface.
lengths until a uniform moisture distribution was attained throughout After more than two years of drying of the brick-shaped speci-
the specimen. mens the top surfaces of two specimens were prepared by high-
pressure water jetting, complying with the respective Austrian code
requirements for the surface conditions of concrete structures
3.1.3. Shrinkage tests on larger brick-shaped specimens with added strengthened by overlays [33]. The prepared top surfaces were cov-
overlays ered by wet cloth for 2 days until the concrete overlay was placed.
In addition to the tests on thin concrete slices and concrete prisms, These boundary conditions were chosen in order to reproduce the
shrinkage tests were performed on 3 brick-shaped concrete speci- strengthening procedure of existing concrete bridges in engineering
mens with dimensions of 800 × 300 × 300 mm, supplemented by con- practice as close as possible.
crete overlays with the dimensions of 800 × 300 × 90 mm on the top Similar to the lateral surfaces of the brick-shaped specimens also
surface (Fig. 4). The brick-shaped specimens as well as the concrete the lateral surfaces of the concrete overlays were sealed. The concrete
overlays were made of the same concrete type (see Table 1). After overlays were equipped with four Multi-Ring Sensors (MRS 21 to
sealing of the lateral surfaces, the top and the bottom surfaces of MRS 24 in Fig. 5) and a multitude of measuring discs (Fig. 5) for mea-
the specimens were submitted to drying at a relative humidity of suring distances parallel to the vertical edges of the specimen with
(65 ± 5)% and a temperature of (20 ± 2)°C for more than 2 years. rated measuring lengths of 50 mm and of 100 mm. In order to deter-
The brick-shaped specimens were equipped with Multi-Ring Sen- mine the mass water content distribution through the depth of the
sors, temperature detectors and a multitude of measuring discs. They concrete overlay two MRS were placed consecutively starting from
served for measuring distances parallel to the vertical edges of the the overlay surface, i.e. MRS 21 below MRS 22 and MRS 23 below
specimen with rated measuring lengths of 50 mm and 100 mm MRS 24. The top and bottom surfaces of the composite samples
(Fig. 4). Since during drying the moisture distribution in the brick-

Fig. 4. Large brick-shaped concrete specimen supplemented by a concrete overlay with

Fig. 3. Multi-Ring-Sensor (MRS) with the respective measuring points (MP) and measuring discs and measurement lengths for determining the drying shrinkage
elevations. strains.
1290 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297

Fig. 5. Arrangement of MRS and measuring discs in the brick-shaped specimen and in the concrete overlay.

were submitted to drying at a relative humidity of (65 ± 5)% and a 3.1.4. Shrinkage tests on reference specimens for the overlays
temperature of (20 ± 2)°C. In addition to the specimens mentioned so far, two so-called refer-
In [3,27] short-term and long-term tests on concrete beams with ence specimens for the concrete overlay with dimensions of
bonded concrete overlays were reported. They focused on the influ- 300 × 300 × 90 mm were cast (Fig. 6). Two MRS were embedded in
ence of different surface preparation methods on the micro- and each reference specimen and measuring discs were applied at the lat-
macro-roughness and on the resulting restraint effects at the inter- eral surfaces.
face. In particular, casting the top surface of the substrate concrete
against steel formwork for generating a very smooth interface,
sand-blasting for roughening the interface and the application of
notches at the interface for ensuring interlock between the substrate
and the overlay were investigated. After surface preparation, the top
surfaces were kept moist for 24 h, followed by drying for 30 to
60 min prior to the placement of the overlay. Since surface prepara-
tion by water jetting of the top surface and the evolution of the mois-
ture distribution were not investigated in [3,27], the results are not
comparable to the findings of the present study.

Fig. 7. Results of the tests on thin concrete slices: (a) Water desorption isotherm (mean
Fig. 6. Reference specimen for the concrete overlay, equipped with two MRS, measur- values and standard deviation), (b) time periods for achieving sorption equilibrium at
ing discs and measurement lengths. different values of relative humidity.
 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297 1291

Fig. 8. Mean values of the drying shrinkage strain measured on thin concrete slices.

One of these reference specimens was sealed at the lateral surfaces Fig. 10. Mean value and standard deviation of drying shrinkage strains measured at
and was submitted to drying at a relative humidity of (65 ± 5)% and a two sealed lateral surfaces of a concrete prism.
temperature of (20 ± 2)°C immediately after casting. The combined au-
togenous and drying shrinkage strains measured for this reference reference specimens with different curing conditions allows dis-
specimen can thus be compared to the respective measuring data of tinguishing between autogenous shrinkage strains on the one hand
the concrete overlay. Furthermore, comparison of the depth dependent and the combined autogenous and drying shrinkage strains on the
mass water content distribution determined for the concrete overlays other hand.
with the one determined for this reference specimen allows insight
into the moisture migration and distribution in composite specimens. 3.2. Test results
The second reference specimen was moist cured after casting.
Comparison of the measured shrinkage strains determined for both 3.2.1. Shrinkage tests on concrete slices
Fig. 7a shows the mean values and standard deviation of equilibri-
um mass water content for selected values of ambient relative hu-
midity determined from the 10 thin slices shown in Fig. 1. At the
end of moist curing for 81 days, i.e. at a RH of about 100%, the mean
value of the measured mass water content amounted to 6.4% (stan-
dard deviation 0.25%). At a RH of 65% the mean value of the mass
water content was reduced to about 3.5% (standard deviation 0.20%).
The time periods for achieving sorption equilibrium at different
values of ambient relative humidity are depicted in Fig. 7b.
The mean values of the drying shrinkage strains measured on the 10
thin concrete slices in the directions A–A, B–B, C–C, and D–D (Fig. 1) are
plotted in terms of ambient relative humidity in Fig. 8. There is no statis-
tically significant difference between the respective mean values com-
puted from the four measurement lengths A–A, B–B, C–C, and D–D. At
a relative humidity of 65% the mean value of the measured drying
shrinkage strain is about −0.49‰ (standard deviation 0.036‰).
The water desorption isotherm of Fig. 7a is of comparable shape as
the one for a similar concrete quality reported in [16]. However, be-
cause of the different definitions of the dry state, different mix param-
eters and a somewhat smaller w/c-ratio, the water desorption
isotherm in [16] is characterized by a lower mass water content.
E.g., at 65% RH a mass water content of about 2.5% is reported in
[16] compared to about 3.5% in Fig. 7a. Because of the mentioned dif-
ferences in the test procedures and in the mix parameters also the
drying shrinkage strains reported in [16] are smaller than those
shown in Fig. 8.

3.2.2. Shrinkage tests on concrete prisms

The shrinkage tests on prismatic specimens served for determining
the mass water content evolution and distribution in the samples
along the length of a Multi-Ring-Sensor (MRS) in terms of drying time.
On the basis of a calibration curve, relating the measured specific
electrolytic resistance to the mass water content [32], the 8 measured
values of the specific electrolytic resistance for each MRS were
converted to mass water contents as a function of depth. Fig. 9a
Fig. 9. Results of the tests on concrete prisms: Mass water contents at the 8 measuring
shows the mass water content, i.e., the ratio of the mass of water con-
points (MP) of a single MRS plotted (a) in terms of drying time and (b) along the along tained in a specimen to the mass of the dry specimen, determined at
the center line of the specimen for selected time instants during drying. the 8 measuring points (MP) of a single MRS, embedded in a concrete
1292 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297

Fig. 11. Arrangement of the MRS in the composite specimen and respective elevations.

prism, as a function of drying time. In Fig. 9a the numbers in parenthe- curing in Fig. 9a. The respective comparison reveals good agreement.
ses following a particular MP indicate its distance from the top surface Accordingly, in Fig. 9a the mass water content after about 950 days of
of the concrete prism. drying at 65% RH is close to the equilibrium mass water content of the
The mass water contents in Fig. 9a are shown, starting from the desorption isotherm at the respective RH, shown in Fig. 7.
end of moist curing (81 days after casting), for a drying period of In Fig. 9b the mass water content distribution along the MRS, char-
about 950 days. The mass water contents determined by the MRS acterized by a measurement length of about 35 mm, is plotted for se-
(Fig. 9a) can be compared with the respective values in Fig. 7. lected time instants during drying. The initial constant mass water
E.g., in Fig. 7 the equilibrium mass water content of the water desorp- content distribution at the end of moist curing gradually changed
tion isotherm determined at a RH of about 100% can be compared with progressing drying time to a distribution with decreasing values
with the respective value of mass water content at the end of moist toward the surfaces due to moisture migration and moisture transfer
to the environment. Furthermore, the curves in Fig. 9b mostly confirm
the expected symmetry of the mass water content distribution along
the center line of the specimen by comparing the measured values of
mass water content of the measuring points MP1 (45.5 mm) vs. MP8
(10.5 mm), MP2 (40.5 mm) vs. MP7 (15.5 mm), MP3 (35.5 mm) vs.
MP6 (20.5 mm), and MP4 (30.5 mm) vs. MP5 (25.5 mm).
The moisture content distribution and its evolution in terms of dry-
ing time can be compared with the results presented in [16] for cylindri-
cal samples with dimension of Ø160 × 100 mm. They were made of
normal strength concrete and were two years old at the beginning of
drying. For the latter the evolution of the distribution of the moisture
content, determined by gamma-ray attenuation measurements, is sim-
ilar to the one in Fig. 9b, however, with a somewhat smaller decrease of
the mass water content. The differences can be explained by the already
mentioned different mix parameters and the different definitions of the
dry reference state. The plotted distributions of the mass water content
in [16] indicate the difficulty of preserving the symmetry properties of
the test set-up in the test results.
The mean values of the drying shrinkage strains measured on the
Fig. 12. Mass water content distributions in the brick-shaped specimen for selected basis of the four measurement lengths, shown in Fig. 2, are depicted
time instants during two years of drying. in Fig. 10. After about 950 days of drying the mean value of the
 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297 1293

Fig. 14. Mass water content distributions in the brick-shaped specimen at the last day
of drying (t = 823 days), after surface preparation during subsequent wetting of the
top surface (t = 824 days), before placement of the overlay (t = 825 days) up to
294 days after placement of the overlay (t = 1119 days).

in [15] the evolution of the shrinkage strains, requiring about

200 days of drying for obtaining 90% of the ultimate shrinkage strain,
is considerably slower than for smaller concrete specimens.

3.2.3. Shrinkage tests on larger brick-shaped specimens

The depth dependent mass water content distribution in the larg-
er brick-shaped specimens was determined by means of MRS in a
similar manner as in the shrinkage tests on concrete prisms. The
lower part of Fig. 11 shows the arrangement of MRS 7 and MRS 6 in
the brick-shaped specimen of Fig. 5. In Fig. 12 the mass water content
distribution is plotted for selected times along the two consecutively
placed MRS 7 and MRS 6 in one of the brick-shaped specimens. The
mass water contents in the brick-shaped specimen, shown in
Fig. 13. Evolution of drying shrinkage strains (mean values and standard deviation) Fig. 12, were measured during drying until the top surface of the
measured at (a) the upper region, (b) the middle region and (c) the lower region of
specimen was prepared by high-pressure water jetting, as described
the brick-shaped specimen.
in Section 3.1.3. Ideally, at the end of moist curing (t = 81 days) the
measured mass water content should be uniform throughout the
measured shrinkage strains amounted to about − 0.52‰ (standard specimen. The measured deviations from the expected uniform distri-
deviation 0.003‰). bution can be attributed to different degrees of compaction of the
The evolution of the shrinkage strains, shown in Fig. 10, qualita- concrete through the specimen depth. The measured values of MRS
tively agrees quite well with the evolution of shrinkage strains 6 show a uniform decrease of the mass water content in the inner re-
reported in [15] for specimens with dimension of 50 × 50 × 400 mm gion of the brick-shaped concrete specimen, i.e. at distances from the
made of normal strength concrete with a w/c-ratio of 0.5. The latter surface larger than about 50 mm (Fig. 11), from initially 6.75% to 5.4%
were moist cured for 28 days, followed by drying at 48% RH. The ulti- after 1 year of drying. During the second year of drying the latter
mate shrinkage strain of − 0.65‰ was larger than the one in Fig. 10. value decreased to 5.0%.
The differences are the result of different mix properties, of a lower By contrast, in the near-surface region the mass water content
value of ambient RH and of different concrete ages at the beginning during drying, described by the measuring values of MRS 7 in
of drying. According to the shrinkage curves plotted in Fig. 10 and Fig. 12, decreases toward the top surface, resulting in a decrease of
1294 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297

and cracks are produced in the superficial zone extending to a

depth of several millimeters [34,35]. Hence, application of the calibra-
tion curve for determining the mass water content from the electro-
lytic resistances measured by the MRS is questionable for the
superficial zone of the brick-shaped specimen. For this reason the
mass water contents measured at measurement points MP8 and
MP7 of MRS 7 are indicated only by dots and are not connected to
the plotted distributions below the superficial zone.
Whereas the mass water content of about 5% measured at mea-
surement point MP1 of MRS 6 in the inner region of the brick-
shaped specimen remained almost constant during high-pressure
water jetting and subsequent wetting of the top surface of the
brick-shaped specimen, the measurement points of MRS 7, placed
closer to the top surface of the brick-shaped specimen, showed a dis-
tinctive change in mass water content. Due to water jetting and sub-
sequent wetting of the top surface of the brick-shaped specimen the
mass water content at MP6 of MRS 7 increased from 3.6% at the end
Fig. 15. Mass water content distributions in the added concrete overlay for selected
of drying (t = 823 days) up to about 6.5%.
time instants.
After placement of the overlay in the region close to the interface
the mass water content in the old concrete is decreasing by water mi-
the initial water content of 7.0% at the end of moist curing to 2.8% gration toward the interior of the brick-shaped specimen. In the inte-
near the surface after more than two years of drying. rior of the brick-shaped specimen a delayed increase of the mass
The mean values of the drying shrinkage strains, measured on the water content is observed (up to t = 951 days), followed by a de-
basis of the measurement lengths, shown in Fig. 4, are depicted in crease toward the original mass water content of 5%, measured before
Fig. 13. The shrinkage strains measured at the upper and lower region surface preparation and placement of the concrete overlay.
of the brick-shaped specimen attained the expected final value of For the concrete overlay the mass water content distribution is
about −0.5‰ (standard deviations of 0.03‰ and 0.05‰, respectively) shown for selected time instants in Fig. 15. The mass water contents
after about 600 days of drying at 65% RH. Expectedly, in the middle of the concrete overlay close to the interface, measured by the mea-
region of the specimen the evolution of the drying shrinkage strain surement points MP1 to MP3 of MRS 21, show a uniform decrease
(standard deviation 0.03‰) was somewhat delayed with respect to from about 7.2% after 28 days of drying to about 6.3% after 294 days
the upper and lower region. of drying, whereas the measuring points MP4 to MP8 of MRS 21 indi-
cate a slight decrease of the mass water content toward the surface of
the overlay. By contrast, the mass water content distribution in the
3.2.4. Shrinkage tests on larger brick-shaped specimens with added near-surface region, represented by the measuring values of MRS
overlays 22, is characterized by a steep gradient, resulting at MP8 of MRS 22
In Fig. 14 the mass water content distributions in a brick-shaped in a decrease of the mass water content from 5.7% after 28 days of
specimen are plotted for selected time instants along the consecutive- drying to 3.9% after 294 days of drying.
ly placed MRS 7 and MRS 6, starting from the last day of drying imme- The evolution of the drying shrinkage strains for the brick-shaped
diately before preparation of the top surface by high-pressure water specimen is shown in Fig. 16 in terms of the mean values measured at
jetting (t = 823 days), 24 h after water jetting (t = 824 days), imme- a lateral surface in the upper, middle and lower region, respectively,
diately before placement of the concrete overlay (t = 825 days) until as a function of drying time. The mean values were determined
294 days after placement of the concrete overlay (t = 1119 days). from four measurements along the length of the specimen shown in
From water jetting until placement of the overlay at t = 825 days Fig. 4. The red arrow in Fig. 16 indicates the time instant, when the
the top surface of the brick-shaped specimen was covered by wet top surface of the brick-shaped specimen was prepared by high-
cloth, protected by a plastic foil. pressure water jetting. Expectedly, Fig. 16 clearly shows that in the
High-pressure water jetting removes concrete from the top sur- lower region of the brick shaped specimen the evolution of the
face down to a depth of a few millimeters below the original surface shrinkage strains was not affected by surface preparation and place-
ment of the overlay. In the middle region of the specimen the

Fig. 16. Mean values of drying shrinkage strains measured at the upper, middle and Fig. 17. Mean value and standard deviation of shrinkage strains measured at a lateral
lower region of a lateral surface of the brick-shaped specimen. surface of the concrete overlay.
 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297 1295

exact positions of the MRS during placement and compaction of the

concrete. After 28 days of drying, moisture migration and moisture
transfer to the environment resulted in decreasing values of the mass
water content toward the surfaces. From 28 days to 294 days of drying,
in the near-surface regions the water content decreased from 6.0% to
4.1% at MP8 of MRS 31 and from 6.3% to 4.5% at MP1 of MRS 30,
respectively.
The evolution of the combined autogenous and drying shrinkage
strains, measured at a sealed lateral surface of the reference speci-
men, submitted to drying immediately after casting, is shown in
Fig. 19. The diagram represents the mean values determined from
the three measurements along the length of the reference specimen
shown in Fig. 6.
For the second reference specimen the mass water content distri-
bution is shown for selected time instants during moist curing in
Fig. 18. Mass water content distributions in the reference specimen submitted to dry- Fig. 20. The recorded values of MRS 32 and MRS 33 for the mass
ing immediately after casting. water content show a nearly uniform decrease from about 7.5% to
about 6.3%.
evolution of the drying shrinkage strains did not change significantly The mean values of shrinkage strains measured at a lateral surface
after surface preparation and casting of the concrete overlay. By con- of the reference specimen after 81 days of moist curing amounted to
trast, the evolution of the shrinkage strains in the upper region was about −0.08‰. The latter are autogenous shrinkage strains. As
strongly influenced. The increased mass water content near the sur- expected, they are small compared to the shrinkage strains measured
face of the brick-shaped specimen, indicated by the measurement on the reference specimen submitted to drying immediately after
values of MRS 7 in Fig. 14, resulted in a delayed increase in mass casting.
water content in the region below MRS 7 (according to the measure- In [9] the effects of combined self-desiccation and external drying
ment values of MRS 6 in Fig. 14). Thus, the shrinkage strains in the were investigated by submitting concrete specimens made of normal
upper region of the brick-shaped specimen decreased as shown in strength concrete with dimensions of 70 × 70 × 280 mm to drying at
Fig. 16. (65 ± 5)% RH one day after casting. Simultaneously, autogenous
The mean values of the shrinkage strains measured at a lateral sur- shrinkage strains were measured on sealed specimens. The reported
face of the concrete overlay are shown in Fig. 17. The diagram repre- autogenous shrinkage strain of about −0.12‰ is larger than the one
sents the mean values determined from the four measurements along of the present investigation. However, the total shrinkage strain
the length of the overlay, shown in Fig. 4. Because the concrete over- reported in [9] is of similar magnitude as the total shrinkage strain
lay was submitted to drying at a relative humidity of (65 ± 5)% and a measured on the reference specimen in the present investigation.
temperature of (20 ± 2)°C immediately after removing the formwork, The moisture distribution determined by means of capacitive probes
the strains shown in Fig. 17 represent the combination of autogenous on drying samples of normal strength concrete with dimensions of
and drying shrinkage strains. Furthermore, they are affected by the 150 × 150 × 300 mm in [9] can be compared with the mass water con-
restraint effects exerted from the old concrete to the overlay. tent distributions in the reference specimen submitted to drying im-
mediately after casting shown in Fig. 18. Qualitatively, the reported
3.2.5. Shrinkage tests on reference specimens decrease of mass water content in [9] agrees with the one plotted in
As described in Section 3.1.4, in addition to the added overlays the Fig. 18.
two so-called reference specimens were produced. For the reference
specimen submitted to drying immediately after casting, Fig. 18 depicts 3.3. Interpretation of the test results for the composite specimen
the mass water content distributions at selected time instants for the
two consecutively placed MRS 31 and MRS 30. Comparison of the mea- Fig. 21, which is composed of Figs. 15 and 14, from left to right
surement values of MRS 31 and MRS 30 reveals deviations from the depicts the distributions of the mass water content from the surface
expected symmetric distribution of the mass water content through of the overlay to the interior region of the original brick-shaped
the depth. They can be explained by difficulties in maintaining the specimen from the last day of drying (t = 823 days) of the brick

Fig. 19. Mean value and standard deviation of combined autogenous and drying
shrinkage strains measured at a lateral surface of a reference specimen, submitted to Fig. 20. Mass water content distributions in the reference specimen for selected time
drying immediately after casting. instants during moist curing.
1296 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297

Fig. 21. Mass water content distributions from the surface of the overlay to the interior region of the brick shaped specimen at selected time instants.

shaped specimen until 294 days after placement of the overlay observed in both the overlay and the superficial zone of the old con-
(t = 1119 days). Since the overlay was placed at t = 825 days, the crete. They result in the simultaneous evolution of shrinkage strains in
four curves for the mass water content in the overlay refer to overlay the overlay and re-increase of shrinkage strains in the old concrete
ages of 853 days–825 days = 28 days, 881 days–825 days = 56 days, close to the interface. The simultaneous evolution of shrinkage strains
951 days–825 days = 126 days and 1119 days–825 days = 294 days, in the overlay and the adjacent old concrete reduces restraint effects
respectively. Comparison of the peak value of the mass water content at the interface.
of 6.5%, measured in the old concrete at MP6 of MRS 7 after water jet- Comparison of the evolution of the mass water content for the ref-
ting and wetting of the top surface, with Fig. 12 reveals that the peak erence specimen, submitted to drying immediately after casting, in
value of 6.5% is close to the mass water content of 7% in the brick- Fig. 18 with the respective values for the near-surface region of the
shaped specimen at the end of moist curing. After the placement of concrete overlay in Fig. 15 shows similar behavior close to the surface.
the concrete overlay, water was migrating in the old concrete from Expectedly, in the interior region of the overlay the mass water
the wetted region at the interface toward the interior of the brick- content decrease is lower than in the reference specimen, since in
shaped specimen. This is indicated by the gradient of the mass water the latter also the bottom surface was exposed to drying.
content along MRS 7 and the slight gradient along MRS 6, which de-
creases only at later stages (t> 951 days, see also Fig. 14). The time- 4. Summary and outlook
dependent changes in mass water content of the old concrete resulted
in a delayed decrease of the shrinkage strains in the upper region of It was the aim of this contribution to investigate the effects of dry-
the brick-shaped specimen. This follows from Fig. 16 for t > 825 days, ing shrinkage on the behavior of concrete structures strengthened by
where the measured shrinkage strains in the upper region of the brick overlays. Normal strength concrete was used for both the original
shaped specimen refer to the measurement length ranging from concrete and the concrete overlay. Thin concrete slices served for de-
50 mm to 100 mm below the interface (as shown in Fig. 5), which cor- termining the shrinkage strains in terms of ambient relative humidity
responds to the range of MRS 6 in Fig. 21. and the water desorption isotherm, characterizing the equilibrium
In the overlay the time-dependent decrease of the mass water con- moisture properties of the selected concrete. Depth dependent mois-
tent, shown in Fig. 21, indicates water migration toward the surface of ture distribution profiles were determined by measuring electrolytic
the overlay. Since the decrease of the mass water content in the overlay resistances by means of Multi-Ring-Sensors, placed in concrete
is caused by both hydration and drying, the shrinkage strains, shown in prisms with sealed lateral surfaces for enforcing one-dimensional
Fig. 17, represent combined autogenous and drying shrinkage strains. In moisture transport. Furthermore, the drying shrinkage strains were
addition, they are influenced by restraint effects exerted from the inter- measured on the lateral surfaces of the concrete prisms. The measure-
face on the deformations of the overlay. Comparison of Figs. 17 and 19, ments by the Multi-Ring-Sensors were verified by comparing the
the latter depicting the combined autogenous and drying shrinkage equilibrium mass water contents, determined by means of step-by-
strains measured on the reference specimen submitted to drying imme- step desorption, with the respective mass water contents determined
diately after casting, reveals that the shrinkage strains in the concrete by the Multi-Ring-Sensors.
overlay are smaller than the ones in the reference specimen. However, In addition to the lab tests for determining the hygral and me-
the difference is small. This fact indicates relatively small restraint chanical properties of the employed concrete, tests on larger brick-
effects exerted from the old concrete on the deformations of the over- shaped specimens supplemented by concrete overlays were per-
lay. This desirable effect was achieved by thorough wetting of the formed. Multi-Ring-Sensors were placed in the near-surface region
superficial zone of the old concrete, which results in a decrease of the of the brick-shaped concrete specimen for monitoring the depth de-
shrinkage strains, related to the maximum value of the shrinkage pendent moisture distribution during drying for more than two
strains due to previous drying. During subsequent drying of the overlay, years, during surface preparation by high-pressure water jetting and
according to Fig. 21 decreasing values of the mass water content are subsequent wetting of the interface by wet cloth and after placement
 Y. Theiner, G. Hofstetter / Cement and Concrete Research 42 (2012) 1286–1297 1297

of the concrete overlay. Moreover, the shrinkage strains were [8] M.-H. Zhang, L. Li, P. Paramasivam, Shrinkage of high-strength lightweight aggre-
gate concrete exposed to dry environment, ACI Mater. J. 102 (2005) 86–92.
recorded in the brick-shaped specimens and in the added concrete [9] V. Baroghel-Bouny, J. Godin, Experimental study on drying shrinkage of ordinary
overlay. In the overlay the mass water content decreased due to hydra- and high-performance cementitious materials, Concr. Sci. Eng. 3 (2001) 13–22.
tion and drying. Both effects resulted in shrinkage strains. In the upper [10] M.G. Alexander, Deformation and volume change of hardened concrete, Fulton's
Concrete Technology, 8th edn, 2001, Midrand, South Africa.
zone of the old concrete, firstly wetting induced a delayed decrease of [11] W. Hansen, J.A. Almudaiheem, Ultimate drying shrinkage of concrete — influence
the shrinkage strains, referred to those from previous drying, followed of major parameters, ACI Mater. J. 84 (1987) 217–223.
by a re-increase of the shrinkage strains due to water migration to inte- [12] O. Chaallal, B. Benmokrane, G. Ballivy, Drying shrinkage strains: experimental
versus codes, ACI Mater. J. 89 (1992) 263–266.
rior regions of the old concrete. The simultaneous increase of shrinkage [13] ACI Committee 209, Prediction of creep, shrinkage, and temperature effects in
strains in the overlay and in the old concrete close to the interface re- concrete structures, Designing for Creep and Shrinkage in Concrete Structures,
duced restraint effects at the interface. This conclusion was confirmed SP-76, American Concrete Institute, Detroit, 1982, pp. 193–300.
[14] Model Code for Concrete Structures, Bulletin d'Information 124/125E, Comité
by similar evolutions of the combined autogenous and drying shrinkage
Euro-Internationale du Béton/Fédération Internationale de la Précontrainte,
strains in the overlay and in the reference specimen submitted to drying Paris, 1978.
immediately after casting. [15] B. Bissonnette, P. Pierre, M. Pigeon, Influence of key parameters on drying shrink-
Firstly, the findings of this investigation contribute to a better un- age of cementitious materials, Cem. Concr. Res. 29 (1999) 1655–1662.
[16] V. Baroghel-Bouny, M. Mainguy, T. Lassabatere, O. Coussy, Characterization
derstanding of the impact of surface preparation and of drying shrink- and identification of equilibrium and transfer moisture properties for ordinary
age on the behavior of concrete structures strengthened by concrete and high-performance cementitious materials, Cem. Concr. Res. 29 (1999)
overlays. 1225–1238.
[17] N.J. Gardner, M.J. Lockmann, Design provisions for drying shrinkage and creep of
Secondly, extensive experimental data is provided for calibrating normal-strength concrete, ACI Mater. J. 98 (2001) 159–167.
and validating numerical models for drying shrinkage. The test data [18] T. Ayano, F.H. Wittmann, Drying, moisture distribution, and shrinkage of
from the thin concrete slices and the concrete prisms can be used cement-based materials, Mater. Struct. - Sci. Rep. 35 (2002) 134–140.
[19] G. Pons, P. Munoz, G. Escadeillas, Determination of concrete total deformations
for the former task, whereas for the latter task the test results from under load as function of shrinkage value, ACI Mater. J. 100 (2003) 14–20.
the composite specimens can be employed. In a further investigation [20] C. Videla, J.P. Covarrubias, C. Masana, Updating concrete drying-shrinkage predic-
the results of the presented laboratory tests will be used for calibrat- tion models for local materials, ACI Mater. J. 101 (2004) 187–198.
[21] V. Baroghel-Bouny, Water vapour sorption experiments on hardened cementi-
ing and validating two models for drying shrinkage. The first one is a tious materials: Part I: essential tool for analysis of hygral behaviour and its rela-
simplified model, available in commercial FE-programs, which relates tion to pore structure, Cem. Concr. Res. 37 (2007) 414–437.
the shrinkage strain to the change in relative humidity by a shrinkage [22] V. Baroghel-Bouny, Water vapour sorption experiments on hardened cementi-
tious materials: Part II: essential tool for assessment of transport properties and
parameter [36]. The latter is based on a three-phase model for con-
for durability predictions, Cem. Concr. Res. 37 (2007) 438–454.
crete taking into account the physical origin of drying shrinkage, [23] B. Bissonnette, E.K. Attiogbe, M.A. Miltenberger, C. Fortin, Drying shrinkage, curl-
which is related to the increase of the capillary pressure in the con- ing, and joint opening of slabs-on-ground, ACI Mater. J. 104 (2007) 259–267.
crete with decreasing relative pore humidity during the drying pro- [24] T. Rougelot, F. Skoczylas, N. Burlion, Water desorption and shrinkage in mortars
and cement pastes: experimental study and poromechanical model, Cem.
cess [37]. Concr. Res. 39 (2009) 36–44.
Thirdly, the test results reveal possible improvements of the test [25] D.A. Lange, H.-C. Shin, A computer-based design tool for analysis of bonded con-
setup for the composite specimens. In the tests of the present investi- crete overlays, Concr. Sci. Eng. 3 (2001) 189–194.
[26] H.-C. Shin, D.A. Lange, Effects of shrinkage and temperature in bonded concrete
gation Multi-Ring-Sensors were not placed across the interface be- overlays, ACI Mater. J. 101 (2004) 358–364.
cause of the large risk of damage of the sensors during high- [27] H. Beushausen, M.G. Alexander, Failure mechanisms and tensile relaxation of
pressure water jetting and because of the questionable application bonded concrete overlays subjected to differential shrinkage, Cem. Concr. Res.
36 (2006) 1908–1914.
of the calibration curve for converting measured electrolytic resis- [28] J. Feix, A. Andreatta, C. Niederegger, G. Fritsche, G. Hofstetter, G. Niederwanger, Y.
tances to mass water contents in the superficial zone, in which cracks Theiner, T. Cordes, Composite constructions used for load-bearing structures and
are produced by water jetting. However, this region is of considerable roadways of bridges, Research Report, Published in Scientific Series of Research
(Straßenforschungsheft Nr. 589), Federal Ministry of Traffic, Innovation and
interest. Hence, future attempts should be aiming at measurement Technology, Vienna, Austria, 2010, in German.
data in the regions very close to the interface between the old and [29] Y. Theiner, G. Hofstetter, Numerical prediction of crack propagation and crack
new concrete. widths in concrete structures, Eng. Struct. 31 (2009) 1832–1840.
[30] Y. Theiner, G. Hofstetter, A crack model with delayed embedded discontinuities
for the numerical prediction of crack widths in concrete structures, Proceedings
Acknowledgment of the 9th International Conference on Computational Structures Technology
(CST2008), Athens, Greece, 2008, CD-ROM, 18 pp.
Financial support for this research project by the Tyrolean Science [31] EN ISO 12571, Hygrothermal performance of building materials and products —
determination hygroscopic sorption properties, ICS 91.100.01, 2000, in German.
Fund is gratefully acknowledged. [32] W. Brameshuber, M. Raupach, P. Schröder, C. Dauberschmidt, Non-destructive
determination of the water-content in the concrete cover using the
References multiring-electrode, Proceedings of the International Symposium
Non-Destructive Testing in Civil Engineering (NDT-CE), Berlin, Deutsche Gesell-
[1] P. Acker, Shrinkage stresses in fracture mechanics of concrete structures — from schaft für Zerstörungsfreie Prüfung, DGZfP, Berlin, 2003, 14 pp.
theory to applications, in: L. Elfgren (Ed.), Report of the Technical Committee [33] Guideline RVS 15.02.34, Bridges: design and detailing of concrete overlays, Draft
90-FMA Fracture Mechanics to Concrete-Applications, RILEM, Chapman and 20.01.2010, Austrian Research-Promotion Agency Road–Rail Traffic, 2010, in
Hall, London, 1989, pp. 155–161. German.
[2] R. Le Roy, F. de Larrard, Creep and shrinkage of high-performance concrete: the [34] B. Bissonnette, L. Courard, A.M. Vaysburd, A.N. Belair, Concrete removal tech-
LCPC experience, in: Z. Bažant, I. Carol (Eds.), Proceedings of the Fifth Internation- niques, Concr. Int. 28 (2006) 49–55.
al RILEM Symposium in Barcelona, E&FN Spon, London, 1993, pp. 499–504. [35] L. Courard, J.-F. Lenaers, Evaluation of saturation and microcracking of the super-
[3] H. Beushausen, Long-term performance of bonded concrete overlays subjected to ficial zone of concrete, Concrete Repair, Rehabilitation and Retrofitting II, Taylor &
differential shrinkage. Ph.D. thesis, University of Cape Town, South Africa, 2005. Francis, London, 2009, pp. 977–982.
[4] K.H. Khayat, W.J. Long, Shrinkage of precast, prestressed self-consolidating con- [36] A. Van Beek, E. Schlangen, Simulating the effect of shrinkage on concrete struc-
crete, ACI Mater. J. 107 (2010) 231–237. tures, in: V. Baroghel-Bouny, P.C. Aitcin (Eds.), RILEM Proceedings PRO 17 Shrink-
[5] P. Turcry, A. Loukili, Evaluation of plastic shrinkage cracking of self-consolidating age of Concrete, Shrinkage 2000, 2000, pp. 477–492.
concrete, ACI Mater. J. 103 (2006) 272–279. [37] D. Gawin, F. Pesavento, B.A. Schrefler, Hygro-thermo-chemo-mechanical model-
[6] N.K. Emberson, G.C. Mays, Significance of property mismatch in the patch repair ling of concrete at early ages and beyond; part I: hydration and hygro-thermal
of structural concrete. Part 1: properties of repair systems, Mag. Concr. Res. 42 phenomena, part II: shrinkage and creep of concrete, Int. J. Numer. Methods
(1990) 147–160. Eng. 67 (2006) 299–363.
[7] E. Tazawa, S. Miyazawa, Autogenous shrinkage of concrete and its importance in
concrete technology, in: Z. Bažant, I. Carol (Eds.), Proceedings of the Fifth Interna-
tional RILEM Symposium in Barcelona, E&FN Spon, London, 1993, pp. 159–168.