You are on page 1of 7

Cement and Concrete Research 95 (2017) 188–194

Contents lists available at ScienceDirect

Cement and Concrete Research

journal homepage: www.elsevier.com/locate/cemconres

Comparison of tensile and compressive creep of fly ash concretes in the


hardening phase
Anja Estensen Klausen a,b,⁎, Terje Kanstad a, Øyvind Bjøntegaard c, Erik Sellevold a
a
Norwegian University of Science and Technology (NTNU), Department of Structural Engineering, Richard Birkelandsvei 1a, 7491 Trondheim, Norway
b
SINTEF, Building and Infrastructure, 7465 Trondheim, Norway
c
Norwegian Public Roads Administration, 7037 Trondheim, Norway

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

Article history: Tensile and compressive creep tests of fly ash concretes in the hardening phase have been performed in the Tem-
Received 7 July 2016 perature-Stress Testing Machine (TSTM) at NTNU. All tests were performed under nominally identical conditions,
Received in revised form 11 January 2017 e.g. dimensions, curing conditions, measurement set-up and load-regime. Two concretes with 17% and 33% fly
Accepted 23 February 2017
ash (as % by weight of cement and fly ash content) were investigated. Contradictorily to most of the compressive
Available online xxxx
versus tensile creep comparisons found in the literature, both the investigated concretes showed similar creep
Keywords:
behaviour in compression and tension throughout the hardening phase. It was also seen that for the investigated
Concrete (E) time-span, the specific creep development over time was found to increase (i.e. soften) with increasing amount
Aging (C) of fly ash.
Basic creep (C) © 2017 Elsevier Ltd. All rights reserved.
Compression
Tensile properties (C)
Fly ash (D)

1. Introduction considerable significance when it comes to deformation and crack


width calculations.
The time-dependent stress response of concrete is a complex and Concrete creep has been found to be dependent on a number of fac-
much studied phenomenon which is usually described by creep or re- tors, e.g. load level, temperature, loading age, size of specimen and type
laxation. Creep and relaxation are closely connected physically, and in of stress (compressive versus tensile) [8]. Concrete creep is also depen-
concrete structures they often occur simultaneously. As often found in dent on concrete mix-design, such as constituents and w/b ratio. The
engineering practice, the term creep is in the following used to denote importance of tensile creep of concrete when cracking is to be consid-
both creep and relaxation of stresses. ered has been long recognized. However, only a few studies on early
Early age concrete is subjected to volume changes caused by autog- age tensile creep are available in the literature, and in addition, the re-
enous deformation (AD) and thermal dilation (TD). If these volume search results are rarely consistent [9,10]. For instance, while several
changes are restrained, stresses will start to develop in the concrete. To- studies conclude that (long time) tensile creep is larger than compres-
gether with AD and TD, creep is one of the key factors when it comes to sive creep [3,11–13], other studies presents contradictorily results [14,
restrained stress development and the corresponding early age crack 15]. Observations of similar compressive and tensile creep compliances
assessment of concrete structures. AD- and TD-induced stresses at have also been reported [16]. Experimental challenges connected to
early ages can be significantly influenced by creep. Studies have found tensile creep tests is probably one of the reasons for the limited amount
a reduction of restrained stresses by 30–50% due to the beneficial of reported tensile creep results.
creep behaviour of concrete [1–5]. On the other hand, temperature-in- A comprehensive experimental test program on fly ash (FA) con-
duced compressive creep at very early ages can contribute to an in- cretes has been performed at NTNU over the last years. The aim has
crease in the succeeding tensile stress development [6,7]. In addition been to determine decisive parameters for early age crack assessment
to its importance in early age crack assessment, creep is also of [5]. The present paper presents a set of results from this test program:
tensile and compressive creep tests performed under nominally identi-
cal conditions in the Temperature-Stress Testing Machine (TSTM) at
⁎ Corresponding author at: Norwegian University of Science and Technology (NTNU),
Department of Structural Engineering, Richard Birkelandsvei 1a, 7491 Trondheim,
NTNU. The investigated concretes contain the mineral additives silica
Norway. fume (SF) and fly ash (FA). The tests have been performed at different
E-mail address: anja.klausen@ntnu.no (A.E. Klausen). loading ages and subjected to 20 °C isothermal and sealed conditions.

http://dx.doi.org/10.1016/j.cemconres.2017.02.018
0008-8846/© 2017 Elsevier Ltd. All rights reserved.
A.E. Klausen et al. / Cement and Concrete Research 95 (2017) 188–194 189

2. Creep model as a new and improved measurement set-up to achieve the necessary
accuracy and robustness of the system [5].
In the current study, the time-dependent stress response of concrete The Dilation Rig and the TSTM are always run in parallel. The Dila-
has been modelled based on the theory of linear viscoelasticity for age- tion Rig measures free deformation, i.e. the sum of thermal dilation
ing materials. This theory, which is a simplification of the real material and autogenous deformation, of a sealed concrete specimen, while the
performance, implies that creep strains under a constant stress are line- TSTM measures the stress generation of a sealed concrete specimen
arly related to the stress level. This linearity has currently been de- through the hardening phase at a chosen degree of restraint. After the
scribed by the compliance function J(t,t′) modelled by the Double reconstruction the TSTM now is both load-controlled and deforma-
Power Law, Eq. (1), see [17]. tion-controlled, providing a unique possibility to test tensile and com-
pressive creep at the same stress level under the exact same test
1 h i
conditions [5]. When performing a creep test in the TSTM System, a
J ðt; t 0 Þ ¼  0  1 þ φ0 ∙t 0eq −d ∙ðt−t 0 Þp ð1Þ
Ec t eq load is applied the concrete specimen in the TSTM and then kept con-
stant while simultaneously measuring the length change. The resulting
In the above equation, t [days] is the concrete age, t′ is the concrete creep data can then be found by subtracting the stress-independent
age at which the actual stress was applied, Ec(teq′) is the E-modulus at strain measured in the parallel Dilation Rig from the length change mea-
teq′, teq′ is the equivalent age at t′, and φ0, d and p are creep model pa- sured in the TSTM. All creep tests were carried out in the following way:
rameters. The above described creep model is not unique; several alter-
native approaches are found in the literature [15,18–20]. The currently 1. Mixing and casting in the TSTM and the Dilation Rig (dummy)
used model was chosen based on previous experience at NTNU, as it 2. After casting, the TSTM specimen was allowed to move freely, while
has been found to be suitable and sufficiently accurate when evaluating the load was kept at 0.0 ± 0.02 MPa
creep test results. 3. At the chosen loading time, a load corresponding to 1.0 MPa com-
The E-modulus is very important when modelling creep, see Eq. (1). pressive or tensile stress was applied to the TSTM specimen. Immedi-
The age dependent development of the E-modulus has been modelled ately after applying the load, the software was programmed to let the
by Eq. (2), which is a modified version of CEB-FIP MC 1990 [21], see TSTM specimen move freely while maintaining the stress at 1.0 ±
[22,23]: 0.02 MPa
( " sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi!#)nE During all creep tests, the following data were measured: free defor-
  28−t 0
Ec t eq ¼ Ec28 ∙ exp s∙ 1− ð2Þ mation in the Dilation Rig, deformation in the TSTM, stress in the TSTM
t eq −t 0
as well as temperatures in the Dilation Rig, the TSTM and the room. The
creep strain was found by subtracting (for creep in compression) or
In Eq. (2), Ec(teq) represents the E-modulus as a function of equiva- adding (for creep in tension) the free deformation measured in the Di-
lent age teq. Ec28 is the E-modulus at 28 days, s and nE are curve-fitting lation Rig from the deformation measured in the TSTM.
parameters, and t0 = t0* is the start time for stress development [equiv- The TSTM can also be used to directly determine the E-modulus. At
alent time]. The parameter t0 was included in Eq. (2) by [22], while t0* the time of loading in the TSTM, deformation and load data are stored
was introduced later in the program CrackTeSt COIN [24,25]. 10 times a second. This feature provides a stress-strain relation during
the load application. The concrete E-modulus can then be determined
3. Experimental set-up based on the recorded data, providing an E-modulus representing the
specific load application. The E-modulus obtained from the given
All creep tests, both compressive and tensile, were performed in the TSTM equipment has been shown to give good agreement with the cor-
Temperature-Stress Testing Machine (TSTM) System at NTNU. The responding E-modulus determined from independent mechanical test-
TSTM System consists of a Dilation Rig, Fig. 1, and a Temperature-Stress ing [5]. Similar correspondence between E-moduli obtained from TSTM
Testing Machine (TSTM), Fig. 2. Both rigs are connected to a Tempera- tests and parallel methods has also been seen in other studies, e.g. [28].
ture-control System (Julabo FP45), providing an accurate control of
the concrete temperature during testing. The TSTM System is located 4. Experimental program
in a conditioned room which holds a constant temperature of 21 °C ±
0.5 °C. The TSTM System was built in 1995 and is well established and Two concretes were included in the creep test series: ANL FA (17%
documented, e.g. [26,27]. During the years 2009–2013 the TSTM System FA) and ANL FA + 16FA (33% FA), where the total fly ash content is
was reconstructed with new software for managing and logging as well given in parenthesis as % by weight of cement and fly ash content. The

Fig. 1. The Dilation Rig [mm].


190 A.E. Klausen et al. / Cement and Concrete Research 95 (2017) 188–194

Fig. 2. The TSTM [mm].

detailed concrete composition and mechanical properties at 28 days are batch, however, with new batches of aggregate and silica fume. A com-
given in Table 1. plete overview of the performed creep tests is given in Table 2.
The concretes were made with Portland-fly ash cement CEM II/A-V In order to maximize the practical relevance of the creep tests, the
42.5 N “Norcem Anlegg FA” (ANL FA), a water-to-binder ratio of 0.4 normal restrained stress-curves for the concretes when subjected to re-
and a cement paste volume of 292 l/m3. The fly ash content was in- alistic temperature curing conditions were used as background for
creased by replacing cement with fly ash 1:1 by weight, while keeping choosing time and magnitude of loading. Hence, a load of 1.0 MPa com-
the water-to-binder ratio and the cement paste volume constant. All pressive or tensile stress was applied and removed from the concrete
concretes contain 5% silica fume (by weight of cement + FA). specimen according to the stress development curve previously obtain-
Originally, two creep tests were performed for each concrete: one ed from the TSTM as illustrated in Fig. 3. As the applied stress was
compressive creep test loaded at 1 day and one tensile creep test loaded 1.0 MPa, the obtained creep strain curves also represent the compliance
at 9 days. This test series was denoted Test series 1. Due to the found sim- function J(t,t′).
ilarity between compressive and tensile creep behaviour for both con- Both compressive and tensile loads as well as compressive and ten-
cretes, the test series was decided repeated with opposite loading sile creep curves are presented with positive values in the following
ages, denoted Test series 2. However, due to capacity issues caused by figures.
the lower tensile strength, the tensile creep test specimens were loaded
at 4 days instead of after 1 day. The two test series were performed 12– 5. Test results and modelling
14 months apart. Both series were carried out with the same cement
Examples of load and deformations measured in the TSTM System
during a compressive and a tensile creep test is given in Fig. 4a) and
Table 1 Fig. 4b), respectively. In the figures, “TSTM” denotes the strain develop-
Concrete composition and mechanical properties at 28 days. ment measured for the loaded specimen in the TSTM, while the corre-
sponding strain measured for the unloaded specimen in the Dilation
Materials ANL FA ANL FA + 16FA
Rig (the dummy) is denoted “Dilation Rig (dummy)”. Fig. 5 shows the
Concrete composition [kg/m3] total AD development measured in the Dilation Rig under 20 °C isother-
Cement 365.3 284.3
FAcem 60.6 47.2
mal and sealed conditions for the two test series. The curves are zeroed
FAadded 0.0 71.1 at the starting time for stress development as determined in [5], i.e. 9.5
Silica 18.3 17.6 and 12.0 maturity hours for ANL FA and ANL FA + 16FA, respectively.
Free water 160.7 156.2 The E-modulus is very important when determining the creep
Sand 0–2 201.1 201.1
model parameters. For all performed creep tests, two E-moduli are pre-
Sand 0–8 740.2 740.2
Sand 4–8 275.0 275.0 sented: E0, which is the E-modulus found from mechanical testing and
Gravel 8–16 614.1 614.1
Admixture 2.01 1.56 Table 2
Measured values: fresh concrete Test program.
Air [%] 2.3 2.3
Density [kg/m3] 2400 2370 Concrete Creep in tension Creep in compression
Slump [mm] 175 180
Load Stress σ/ft Test Load Stress σ/fc Test
Binder composition (ratio)
age level σ length age level σ length
Total FA-content, FA/(cem + FA) 17% 33%
[days] [MPa] [%] [days] [days] [MPa] [%] [days]
Silica-content, Silica/(cem + FA) 5% 5%
w/b, kFA_added = 1.0 0.40 0.40 Test series 1
(w/b, kFA_added = 0.7) (0.40) (0.42) ANL FA 9 1.0 32 19 1 1.0 6 7
Mechanical properties at 28 days ANL FA + 16FA 9 1.0 38 33 1 1.0 9 8
fc28 [MPa] 71.2 53.6 Test series 2
ft28 [MPa] 3.55 3.05 ANL FA 4 1.0 37 28 9 1.0 2 31
Ec28 [GPa] 30.55 27.80 ANL FA + 16FA 4 1.0 45 8 9 1.0 2 40
A.E. Klausen et al. / Cement and Concrete Research 95 (2017) 188–194 191

4 4
Restrained stress development Restrained stress development
3 3
Compressive Compressive

Stress [MPa]

Stress [MPa]
2 Tensile 2 Tensile
Applied load
1 1

0 0

-1 -1
Applied load
-2 -2
0 7 14 21 0 7 14 21
Maturity [days] Maturity [days]
a) b)
Fig. 3. Applied load during compressive and tensile creep tests [37]; Test series 1 (a) Test series 2 (b).

120 1.2 120 1.2


Measured strain [10-6]

Load (compression) Load (tension)

Measured strain [10-6]


100 1.0 100 1.0
80 TSTM 0.8 80 0.8

Load [MPa]
Load [MPa]

60 0.6 60 TSTM 0.6


40 0.4 40 0.4
Dilation Rig
20 0.2 20 0.2
(dummy)
0 0.0 0 0.0
-20 -0.2 -20 Dilation Rig -0.2
(dummy)
-40 -0.4 -40 -0.4
0 5 10 0 7 14 21
Time [days] Time [days]
a) b)
Fig. 4. Measured deformation ANL FA Test series 1; Compressive creep (a) Tensile creep (b).

ETSTM, which is the E-modulus found directly from the TSTM, Table 3. A the method of least squares when combining both test results (creep
thorough comparison and evaluation of these two E-moduli are given in in compression and creep in tension). Secondly, the parameters φ0
[5]. For all tests, it was decided to use ETSTM when determining the creep and p were found by different approaches as described in the following:
model parameters as described above. This was considered the most
consistent solution as the entire compliance function then could be de- - Using the method of least squares for only the given creep test re-
termined from the same test. sult; i.e. creep in compression or creep in tension
The creep model parameters were determined from Test series 1 by - Using the method of least squares when combining both creep test
the following procedure: First, the parameter d was found by using results; i.e. creep in compression and creep in tension

0 0
Autogenous def. [10-6]

Autogenous def. [10-6]

-20 ANL FA +16FA -20


ANL FA +16FA
-40 ANL FA -40
ANL FA
-60 -60

-80 -80

-100 -100
0 7 14 21 28 35 42 0 7 14 21 28 35 42
Time [days] Time [days]
a) b)
Fig. 5. Autogenous deformation Test series 1 (a) and Test series 2 (b).
192 A.E. Klausen et al. / Cement and Concrete Research 95 (2017) 188–194

Table 3 higher ETSTM than the corresponding tests in Test series 1. This increase
E-modulus. in E-modulus may be a product of new batches of aggregate and/or sim-
Concrete Creep in tension Creep in compression ply an internal scatter in the test results. Additional compressive E-mod-
Loading age E0 ETSTM Loading age E0 ETSTM
ulus tests were thus performed, supporting an increase in E-modulus,
[days] [MPa] [MPa] [days] [MPa] [MPa] however not quite as much as indicated by the TSTM tests.
The creep parameters presented in Table 4 were deduced from Test
Test series 1
ANL FA 9 28,800 28,500 1 20,250 20,400 series 1. For both concretes, the creep model and appurtenant parame-
ANL FA + 16FA 9 25,650 26,000 1 15,850 17,700 ters were found to provide a very good description of the experimental-
Test series 2 ly achieved creep development, Fig. 6. It is noticeable that for both
ANL FA 4 26,650 28,400 9 28,800 32,800 concretes, very similar creep parameters were found for creep in com-
ANL FA + 16FA 4 23,650 27,800 9 25,650 30,500
pression and creep in tension. Actually, the individually deduced creep
parameters gave an almost identical modelled creep development as
the creep parameter set deduced from the compressive and tensile
Table 4 creep tests results combined, Fig. 6.
Determined creep parameters.
The creep parameters deduced from Test series 1 were further used
φ0 d p to model the creep developments for Test series 2 which were per-
ANL FA formed 12–14 months later with new loading ages, Fig. 7. For ANL
Compressive 0.68 0.32 0.26 FA + 16FA, the creep parameters from Test series 1 gave a virtually per-
Tensile 0.63 0.32 0.31 fect description of the experimental creep curves deduced from Test se-
Combined 0.67 0.32 0.28
ries 2. For ANL FA, these creep parameters provided an accurate
ANL FA + 16FA
Compressive 0.48 0.22 0.35 description of the experimentally deduced creep up to approximately
Tensile 0.53 0.22 0.31 14 days for the tensile creep and 19 days for the compressive creep,
Combined 0.49 0.22 0.33 while beyond this a deviation in modelled and experimental creep
was observed. This deviation was however considered relatively
small: only 8 microstrains after 28 and 31 days for the tensile and com-
pressive creep test, respectively. Consequently, for both concretes, com-
Consequently, for each concrete three sets of creep parameters were pressive and tensile creep loaded at both 1, 4 and 9 days, could be
determined: 1) creep in compression (“model – individual”), 2) creep in modelled by the same set of creep parameters, clearly implying similar
tension (“model – individual”) and 3) creep in tension and compression behaviour for compressive and tensile creep.
combined (“model – combined”). The deduced creep parameters are Specific creep curves versus time for ANL FA and ANL FA + 16FA are
summarized in Table 4, while the experimental and modelled specific given in Fig. 8 (a) and (b), respectively. The specific creep obtained
creep curves are presented in Fig. 6 (Test series 1) and Fig. 7 (Test series under compressive and tensile loading regimes can be directly com-
2). pared for the tests loaded at 9 days: For both ANL FA and ANL
FA + 16FA, these specific creep curves are very similar. For ANL FA, the
6. Discussion compressive creep beyond 19 days seems to flatten out. This behaviour
is hard to explain, but as previously discussed, the deviation is consid-
The Dilation Rig was found to provide very good reproducibility for ered to be rather small and it could be a matter of an unfortunate scat-
the given concretes for both test series, Fig. 5. In addition, good agree- ter-combination in the TSTM and the Dilation Rig results. An evaluation
ment was also seen between the two test series which were performed of the reproducibility of the TSTM System is given in reference [5]. Fig. 8
approximately 12–14 months apart. The results show a small reduction also indicates that the measured specific creep curves give good agree-
in AD (contraction) with increasing amount of fly ash. ment with the corresponding calculated curves.
As shown in Table 3, both E0 and ETSTM were consistently lower for Both [3,13] found that the creep behaviour in compression and ten-
the concrete with the higher fly ash content. This is as expected, as the sion were quite different for early age concretes containing mineral ad-
addition of pozzolanes is known to cause slower strength and stiffness ditives. While the initial creep and creep rate was higher for creep in
development than for ordinary Portland cement alone [29–31]. It was compression than for creep in tension, the compressive creep rate de-
further observed that all tests performed in Test series 2 obtained a creased more over time than the tensile creep rate. After a while, the

120 120

100 100
Compliance [10-6]
Compliance [10-6]

Creep in compression Creep in compression

80 80
Creep in tension
Creep in tension
60 60

40 40
Measured Measured
20 Model - individual 20 Model - individual
Model - combined Model - combined
0 0
0 7 14 21 28 35 42 0 7 14 21 28 35 42
a) Time [days] b) Time [days]

Fig. 6. Experimental and modelled creep strains for Test series 1: ANL FA (a) and ANL FA + 16FA (b).
A.E. Klausen et al. / Cement and Concrete Research 95 (2017) 188–194 193

120 120
Experimental Experimental
100 Model, combined 100

Compliance [10-6]
Model, individual

Compliance [10-6]
Model, individual Model, combined
80 Creep in 80
Creep in
tension
60 60 tension
Creep in
40 Creep in 40
compression
20 compression 20

0 0
0 7 14 21 28 35 42 0 7 14 21 28 35 42
Time [days] Time [days]
a) b)
Fig. 7. Experimental and modelled creep strains for Test series 2: ANL FA (a) and ANL FA + 16FA (b).

80 80
Compressive creep Compressive creep
70 70
Specific creep [10-6]

Specific creep [10-6]


Tensile creep Tensile creep
60 Model - combined 60 Model - combined
50 50
40 40
30 30
20 20
10 10
0 0
0 7 14 21 28 35 42 0 7 14 21 28 35 42
a) Time [days] b) Time [days]

Fig. 8. Specific creep; ANL FA (a) ANL FA + 16FA (b).

compressive and tensile creep curves would cross, leading to a higher - Both compressive and tensile creep tests were performed with the
tensile creep than compressive creep over time. Contradictorily, the re- same experimental equipment, the TSTM. Hence, all creep tests
sults from the current creep test series show very similar creep behav- were performed on specimens with exactly the same dimensions,
iour in both compression and tension throughout the hardening the same curing conditions and the same strain measurement set-up.
phase. The current compressive and tensile creep tests differ from [3, - For both compressive and tensile creep tests, the test specimens were
13] (as well as from most of the compressive versus tensile creep com- subjected to the same absolute load level, i.e. 1.0 MPa. This applied
parisons found in the literature) by the following conditions: stress level is considered to be within the linear area of viscoelastic

140
120 ANL FA +16FA
Compliance [10-6]

100
ANL FA
80 ANL FA +16FA
60
40 ANL FA
Measured
Creep in Creep in tension Model - individual
20
compression Model - combined
0
0 7 14 21 28 35
Time [days]
Fig. 9. Experimental and modelled creep strains for Test series 1: ANL FA and ANL FA + 16FA.
194 A.E. Klausen et al. / Cement and Concrete Research 95 (2017) 188–194

behaviour for all performed tests [3]. The time and magnitude of load- [7] Dirk Schlicke, Nguyen Viet Tue, Consideration of viscoelasticity in time step FEM-
based restraint analyses of hardening concrete, J. Mod. Phys. 4 (2013) (2013) 9–15.
ing was decided from experimentally achieved restrained stress [8] A.M. Neville, Creep of Concrete: Plain, Reinforced and Prestressed, North-Holland
curves from the TSTM. This was done in order to maximize the rele- publishing company, University of Leeds, England, 1970.
vance of the obtained creep parameters with respect to the restrained [9] Pierre Rossi, Jean Philippe Charron, Maléna Bastien-Masse, Jean-Louis Tailhan,
Fabrice Le Maou, Sandrine Ramanich, Tensile basic creep versus compressive basic
stress calculations of the TSTM. Contradictorily, for most of the com- creep at early ages: comparison between normal strength concrete and a very
pressive versus tensile creep comparisons found in the literature, a high strength fibre reinforced concrete, Mater. Struct. 47 (2014) (2014) 1773–1785.
load corresponding to a percentage of the compressive cube strength [10] Adrien Hilairea, Farid Benboudjemaa, Aveline Darquennes, Yves Berthauda, Georges
Nahas, Modeling basic creep in concrete at early-age under compressive and tensile
or the uniaxial tensile strength, respectively, have been used during
loading, Nucl. Eng. Des. 269 (2014) (2014) 222–230.
testing. [11] J.M. Illston, The creep of concrete under uniaxial tension, Mag. Concr. Res. 17 (1965)
(1965) 77–84.
[12] J.J. Brooks, A.M. Neville, A comparison of creep, elasticity and strength of concrete in
In the current study, creep was found to increase slightly (i.e. soften) tension and in compression, Mag. Concr. Res. 29 (100) (1977) 131–141.
with increasing fly ash content, Fig. 9. Similar observations were also [13] G.M. Ji, T. Kanstad, Ø. Bjøntegaard, E.J. Sellevold, Tensile and compressive creep de-
made by [32,33]. The opposite tendency, i.e. a decrease in creep (stiffen- formations of hardening concrete containing mineral additives, Mater. Struct. 46
(2012) (2012) 1167–1182.
ing behaviour) with increasing fly ash content, has been found by other [14] S. Hagihara, S. Nakamura, Y. Masuda, T. Uenishi, T. Okihashi, M. Kono, Mechanical
researchers, e.g. [34,35]. A comprehensive series of creep tests per- properties and creep behavior of high-strength concrete in early age, In the Pro-
formed by [36] showed, however, that while one type of fly ash would ceedings of the International Workshop, Sendai Japan, 2000.
[15] Alexander-W Gutsch, Viscoelastic behaviour of Early Age Concrete, IPACS Report,
reduce creep, another type of fly ash would slightly increase the creep
Luleå University of Technology, Sweden, 2001 (ISBN: 91-89580-38).
properties of the concrete. [16] Matthieu Briffaut, Farid Benboudjema, Jean-Michel Torrenti, Georges Nahas, Con-
The given concretes have also been subjected to restrained stress crete early age basic creep: experiments and test of rheological modelling ap-
proaches, Constr. Build. Mater. 36 (2012) (2012) 373–380.
tests under realistic temperature conditions in the TSTM [5]. The TSTM
[17] Z.P. Bazant, E. Osman, Double power law for basic creep of concrete, Mater. Struct. 9
tests have been back-calculated with the above deduced creep parame- (49) (1976) 3–11.
ters, providing a very good description of the measured stress develop- [18] Mats Emborg, Thermal Stresses in Concrete Structures at Early Ages(PhD-Thesis)
ment for the given fly ash concretes, which very much supports the Luleå University of Technology, Sweden, 1989.
[19] Zdenek P. Bazant, Prediction of concrete creep and shrinkage: past, present and fu-
validity of the obtained creep test results. Rather small changes of the ture, Nucl. Eng. Des. 203 (2001) (2001) 27–38.
creep parameters were found to have a considerable effect on the calcu- [20] G. De Schutter, Fundamental study of early age concrete behaviour as a basis for du-
lated stress curves, underlining the substantial importance of accurate rable concrete structures, Mater. Struct. 35 (2002) (2002) 15–21.
[21] CEB-FIP, CEB-FIP Model Code 1990: Design Code. CEB Bulletin No.203, Comité Euro-
creep parameters when conducting restrained stress analyses and International du Béton, Lausanne, Switzerland, 1991 (ISBN: 0727716964,
early age crack assessment of concrete structures. 9780727716965).
[22] Terje Kanstad, Tor Arne Hammer, Øyvind Bjøntegaard, Erik J. Sellevold, Mechanical
properties of young concrete: part II: determination of model parameters and test
7. Summary and conclusion program proposals, Mater. Struct. 36 (2002) (2003) 226–230.
[23] Øyvind Bjøntegaard, Basis for and practial approaches to stress calculations and
• A small reduction in autogenous deformation (contraction) with crack risk estimation in hardening concrete structures - state of the art, COIN Project
report 31 - 2011, SINTEF Building and Infrastructure, Trondheim, Norway, 2011.
increasing amount of fly ash was found for tests performed [24] Jan-Erik Jonasson, Peter Fjellström, Henrik Bäckström, Inverkan av variabel
under 20 °C isothermal and sealed conditions. härdningstemperatur på betongens hållfasthetsutvikling (the influence of variable
• The TSTM was able to perform tensile and compressive creep tests curing temperature on the strength development of concrete - only available in
Swedish), Bygg & Teknik 7 (2010).
under nominally identical conditions, e.g. dimensions, curing condi-
[25] JEJMS Concrete AB, CrackTeSt COIN, Luleå, Sweden, 2009–2012.
tions, measurement set-up and load-regime. [26] Øyvind Bjøntegaard, Thermal Dilation and Autogenous Deformation as Driving
• For the performed test series, both investigated concretes obtained Forces to Self-Induced Stresses in High Performance Concrete(PhD-Thesis) Norwe-
similar creep behaviour in compression and in tension. gian University of Science and Technology (NTNU), Trondheim, Norway, 1999, ISBN
82-7984-002-8.
• The specific creep development over time was found to increase [27] Øyvind Bjøntegaard, E.J. Sellevold, The Temperature-Stress Testing Machine
slightly (i.e. soften) with increasing amount of fly ash for the investi- (TSTM): Capabilities and Limitations. Advances in Concrete Through Science and
gated time-span. Engineering Northwestern University, Evanston, Illinois, 2004.
[28] Claude Boulay, Stéphanie Staquet, Brice Delsaute, Jérôme Carette, Michela Crespini,
Oumaya Yazoghli-Marzouk, Érick Merliot, Sandrine Ramanich, How to monitor the
Acknowledgements modulus of elasticity of concrete, automatically since the earliest age? Mater. Struct.
47 (2014) (2014) 141–155.
[29] A.M. Neville, Properties of Concrete, Pitman Publishing Corporation, New York, USA,
The article is based on work performed in the User-driven Research- 1972, ISBN 0 273 36150 3.
based Innovation project DaCS (Durable advanced Concrete Solutions, [30] Aage D. Herholdt, Chr.F.P. Justesen, Palle Nepper-Christensen, Anders Nielsen,
2015–2019) in addition to COIN (Concrete Innovation Centre, 2007– Beton-Bogen, Aalborg Portland, Denmark, 1979, ISBN 87-980816-0-8.
[31] Göran Möller, Nils Peterson, Paul Samuelsson, Betong-Handbok, AB Svensk
2014, http://www.sintef.no/en/projects/coin/coinp/#/, a Centre for Re- Byggtjänst, Stockholm, Sweden, 1982, ISBN 91-7332-060-9.
search-based Innovation established by the Research Council of Norway). [32] Guomin Ji, Cracking Risk of Concrete Structures in the Hardening Phase(PhD-Thesis)
Norwegian University of Science and Technology (NTNU), Trondheim, Norway,
2008.
References [33] Øyvind Bjøntegaard, Knut O. Kjellsen, Property development and cracking tendency
[1] Daniela Bosnjak, Self-induced Cracking Problems in Hardening Concrete in hardening concrete: effect of cement type and fly ash content, COIN Project Re-
Structures(PhD-Thesis) Norwegian University of Science and Technology (NTNU), port 40, SINTEF Building and Infrastructure, Trondheim, Norway, 2012.
Trondheim, Norway, 2000. [34] Katherine Kuder, Dawn Lehman, Jeffrey Berman, Gudmundur Hannesson, Rod
[2] Salah A. Altoubat, David A. Lange, Creep, shrinkage, and cracking of restrained con- Shogren, Mechanical properties of self consolidating concrete blended with high
crete at early age, ACI Mater. J. 98 (4) (2001) 323–331. volumes of fly ash and slag, Constr. Build. Mater. 34 (2012) (2012) 285–295.
[3] Dawood S. Atrushi, Tensile and Compressive Creep of Early Age Concrete: Testing [35] Yuan-Feng Wang, Yi-Shuo Ma, Lang Zhou, Creep of FRP-wrapped concrete columns
and Modelling(PhD-Thesis) Norwegian University of Science and Technology with or without fly ash under axial load, Constr. Build. Mater. 25 (2011) (2011)
(NTNU), Trondheim, Norway, 2003, ISBN 82-471-5565-6. 697–704.
[4] G. De Schutter, Applicability of degree of hydration concept and maturity method [36] Mohand L. Sennour, Ramon L. Carrasquillo, Creep and Shrinkage Properties in Con-
for thermo-visco-elastic behaviour of early age concrete, Cem. Concr. Compos. 26 crete Containing Fly Ash, The University of Texas at Austin, Austin, Texas, 1989
(2004) (2004) 437–443. (FHWA/TX-90+481-6).
[5] Anja Estensen Klausen, Early Age Crack Assessment of Concrete Structures: Experi- [37] Anja Estensen Klausen, Terje Kanstad, Øyvind Bjøntegaard, Updated Temperature-
mental Determination of Decisive Parameters(PhD-Thesis, ISBN: 978-82-326-1850- Stress Testing Machine (TSTM): introductory tests, calculations, verification and in-
7 [printed ver.] 978-82-326-1851-4 [electronic ver.]) Norwegian University of Sci- vestigation of variable fly ash content, Proceedings of CONCREEP 10, Vienna, Swit-
ence and Technology, Trondheim, Norway, 2016. zerland, 2015.
[6] F. Benboudjemaa, J.M. Torrenti, Early-age behaviour of concrete nuclear contain-
ments, Nucl. Eng. Des. 238 (2008) (2008) 2495–2506.

You might also like