RHEOLOGICAL STUDY ON SUPERPLASTICIZER/CEMENT

COMPATIBILITY AT HIGH TEMPERATURES

F. Curto, S.Carr, A. Lolli, F. Surico, R. Saccone

Synopsis: Although polycarboxylate superplasticizers represent a

substantial improvement in comparison with the lower effective-
ness and retention of the workability offered by naphthalene sul-
phonate and lignosulphonate, they are still believed not fully satis-
factory in many applications, as in the case of ready mix concrete
produced at very high temperatures.
The compatibility between Arabic region produced cements and
different class of superplasticizers was investigated using a shear
stress/shear strain controlled rheometer, adopting flow and oscilla-
tory procedures.
A new class of polymers with extended workability properties
(WES) was found to be effective in imparting and retaining low
viscosity in cementitious paste even at high temperature.
Workability retention of WES, PCE and BNS at high temperature
(35C) was investigated and demonstrated. Measurements and re-
sults obtained on the rheometer were further validated using mini-
slump evaluations of flow as well as concrete mixes with the Ara-
bic region materials.

Keywords: high temperature, rheology, superplasticizer, workabil-

ity.

All of Mapei S.p.A., R&D Laboratories, Milan, Italy

Fabio Curto graduated in chemical engineering at the University
of Calabria, and after working for an engineering Company of
treatment plants design he joined the Mapei group in 2005. Since
then, he works as researcher for Mapei at the R&D Rheology Lab.

Stefano Carr graduated in chemical engineering at Milano Poli-

tecnico, and after working for 6 years for the Montedison group in
polymer research, he joined the Mapei group in 1995. He is cur-
rently responsible for the research on tile adhesives and selflevel-
lers at the Mapei Central Laboratories in Milano, where he also
heads the rheology lab.

Francesco Surico graduated in industrial chemistry at Milan State

University. He leads the research group for the development of
new admixtures for concrete in Mapei central laboratories in Mi-
lan. He is co-author of several papers on concrete technology and
admixtures for concrete as well as three international patents for
the invention of new admixtures for concrete.

Alberto Lolli graduated in chemical engineering at Milano Poli-

tecnico. He is co-operating with Mapei as expert of concrete tech-
nology and applicative tests at the Admixture for Concrete depart-
ment in the central R&D laboratory in Milan.

Roberto Saccone graduated in civil engineering at Rome Univer-

sity in 1982. After working for some of the biggest Italian and in-
ternational companies spanning the cement, admixture and con-
crete productions and markets, he presently works in Mapei Con-
crete Admixtures Department as Export Technological Market De-
velopment Manager.

INTRODUCTION

Arabic peninsula countries represent a severe challenge for con-

crete production, transportation and casting due to their extremely
high environmental temperature. Although polycarboxylate super-
plasticizers (PCE) represent a substantial improvement in compari-
son with the lower effectiveness and retention of the workability
offered by naphthalene sulphonate, melamine sulphonate and lig-
nosulphonate [1, 2], they are still believed not fully satisfactory in
some applications. When transportation of concrete by truck mix-
ers from the mixing plant to the job site takes long time in the
highly congested traffic of urbanised areas, especially with very
high environmental temperatures even polycarboxylate superplas-
ticizers may be ineffective in preventing slump-loss. It is some-
times necessary to restore the initial workability at the job site, be-
fore casting the concrete, either by addition of extra amount of ad-
mixture, or by the addition of more water to the mixture. The latter
operation, which is known as concrete retempering, increases the
water content of the concrete mixtures (water to cement ratio) and
causes detrimental effects on the mechanical properties and the du-
rability of hardened concrete. It is therefore a concern for most
people committed to quality of materials and structures to design
concrete in order to have consistent flow ability and W/C during
all transport time and placing operations.
Most of the currently available admixtures suffer from poor slump
retention in hot climates and, even the most effective ones are still
believed to impart significant increase of viscosity to cementitious
mixes with consequent difficulties in placing and pumping opera-
tions.
In this paper rheology was used to provide information about the
flow and stiffening behavior of cement pastes and in particular on
the compatibility between cement and superplasticizers i.e. con-
cerning the efficiency of different admixtures over time at tem-
perature of 35C.
The compatibility and efficiency of different classes of water re-
ducing admixtures with Arabic peninsula cement and aggregates at
high ambient and concrete temperature (35C) was then further
validated by laboratory tests of mini slump and concrete mixes.
 EXPERIMENTAL

Materials
Superplasticizers Four commercially available superplasticizers
were used, two of them based on PolyNaphataleneSulphonate
(PNS1, PNS2), the other two being a PolyCarboxylate (PCE1) and
a novel type of Polycarboxylate based polymer with improved
workability retention ability [3] (WES1). The characterization of
the superplasticizers used is shown in table 1.
Cement An Arabic Portland cement (CEM type I according to
ASTM C150 04) was used, whose composition and characteristics
are reported in table 2.
Aggregates All crushed aggregates used for concrete mixes were
from the valley region of the Kingdom of Saudi Arabia. Their par-
ticle size distribution is plotted in Fig. 1.

Test methods
Rheological Tests Flow and oscillatory rheological tests were
performed using a TA Instruments rheometer model ARG2. The
dosages of all superplasticizers ranged from 0,8 to 1,5% expressed
as mass of product by the mass of cement. The cement pastes were
prepared using 200g of cement and water enough for w/c = 0,35
after calculation of water contribution of the superplasticizers. The
cement and the superplasticizer solution were first mixed by hand
for 1 minute and then for 2 minutes by mechanical mixing (IKA
labortechnik RW20 ) at 450 rpm. To perform the tests after 30 and
60 minutes, the cementitious paste was mixed for 30 second by
hand and then for 1 minute with the laboratory mixer at 450 rpm.
The mixing water temperature was kept at 35C in a thermostatic
bath and so was the cementitious paste while waiting to perform
the 30 and 60 minutes tests. [4] The test temperature was kept at
35 C by a Peltier plate. The geometry used was plate-plate with
upper plate dimension of 40 mm. The gap was set at 1000 m and
during the tests a solvent trap was used to limit evaporation. The
measurement tests were performed using a flow and oscillatory
procedure [5]. The flow test was performed by applying a shear
rate increasing the shear rate from 0 to 50 s-1. This procedure al-
lows to derive the samples flow curve from which it was possible
to plot the viscosity ( ) and to evaluate the yield stress, the viscosi-
ty being the ratio between shear stress and shear rate. The yield
stress is the energy required by the cement paste to start flowing
and corresponds to the maximum viscosity in the flow of the sam-
ple.
The oscillatory test, a stress sweep, consisted in applying a stress
from 0.001 to 1000 [Pa] at 1 [Hz] of frequency and measuring G.
Usually this curve allows to identify the linear domain of the sam-
ple behavior, as a plateau of G, before critical stress of , called
c, at which G drops, usually, of several order of magnitude.

Mini-slump Tests The pastes were prepared using 300g of ce-

ment and water enough for w/c 0,35 after calculation of water con-
tribution of the superplasticizers. The dosages of all superplasticiz-
ers ranged from 0,8 to 1,5% expressed as mass of product by the
mass of cement. The cement and the superplasticizer solution were
first mixed by hand for 1 minute and then by 2 minutes of mechan-
ical mixing (IKA labortechnik RW20 ) at 450 rpm. The water used
was hot enough to get the grout temperature equal to 35C. A pre-
cisely level Plexiglas plate was placed on a table, the mini cone
(bottom internal diameter = 44mm, upper internal diameter = 20
mm, Height = 59,5 mm) was then placed on the plate and filled
with the paste. The cone was then raised rapidly enabling the paste
to spread. The diameter of the grout was measured along two per-
pendicular directions and the two values were averaged. The paste
was returned into the beaker, homogenized for 5 seconds and cov-
ered. The system was then kept at 35 C for testing after 30 and 60
minutes.
Concrete Tests Normal grade concrete mixes were designed to
have w/c of 0,50 while high grade ones were prepared at w/c of
0,38. For both grades the target Slump value was set to 210 10
mm using the necessary amount of superplasticizers.
The proportions of both mixes are given in table 3. The concrete
mixes were prepared in a forced action mixer with interchangeable
mixing plastic tubs (Collomix mod. XM3-900). All materials (ag-
gregates, cement and water) were stored at 35 C overnight prior to
use as well as the plastic tube containing the mixed concrete was
covered with a plastic sheet and stored at 35 C while waiting for
30, 60 and 90 minutes measurements.

RESULTS AND DISCUSSION

Rheological Tests
The estimated viscosity values of the samples at yield point and
different superplasticizer concentrations, measured at 0, 30 and 60
minutes are showed in figure 2a, 2b and 2c. A flow curve, in which
shear stress is measured as a function of shear rate, is often used as
a mean of characterizing paste rheology. A structurally meaningful
rheological parameter, the yield, sometimes defined as the mini-
mum energy to be imparted to a sample to initiate flow, can be es-
timated by the stress at which the viscosity is maximum. Oscillato-
ry measurements are even more significant to give indication about
the system structure [6, 7], indeed they allow to identify the elastic
vs. viscous nature of the sample. Flocculated suspensions, in which
particles form a three-dimensional network, typically show a beha-
vior marked by a high yield in flow curve characterizations and
elastic behavior at low stress in oscillatory test. Cement pastes
without superplasticizers at low w/c levels are flocculated, while
pastes with superplasticizers are expected to be partially or fully
dispersed, depending on the combination of cement and superplas-
ticizer. A typical rheological response of cement paste in oscillato-
ry shear G vs. shear stress is illustrated in Fig. 5 and a typical flow
curve in Fig. 6, these two measurements have been taken at 0 mi-
nutes with every dosage of the admixtures listed. The pastes added
with PNS1 and PNS2 showed an high viscosity value and this cha-
racteristic was maintained up to 60 minutes. Only high dosages of
PNS1 were effective in reducing the viscosity of the paste, but still
not sufficient to retain the system dispersed for 30 and 60 minutes.
These results are confirmed by oscillatory measurements (Figs. 4a,
4b, and 4c) in which the maximum value of elastic modulus (G) is
plotted. This indicates a cement flocculated system, as the conse-
quence of a very high modulus (higher than 105 [Pa]). In these
measurements WES1 and PCE1 admixtures imparted to the system
a good initial dispersion effect, but only WES1 maintained the sys-
tem dispersed for 60 minutes (figs. 4b, and 4c). PNS1 and PNS2
were less effective in keeping the system fluid of both WES1 and
PCE1 even at higher dosages.

Mini-slump Tests
Tests results are shown in Fig. 7. They indicate, in accordance with
rheological measurements, that PNS1 and PNS2 show less effec-
tiveness when compared to PCE1 and WES1. When low dosages
were used the initial paste flow diameters were quite similar for
PNS1 and PNS2, but, while PNS1 became effective at dosage of
1,5%, PNS2 showed consistent flow ability and increasing reten-
tion of flow when increasing the dosage. PCE1 became effective
for both initial flow and retention of workability from 1,2%,
WES1 imparted high fluidity and outstanding retention of flow
even when dosed at 0.6% by weight of cement.

Concrete Tests
The results showed in Table 4 clearly indicate that PNS1 was una-
ble to retain a slump value higher than 160 mm for more than 30 or
60 minutes. PNS2 and PCE1 showed similar behavior with PCE1
needing slightly higher dosages than PNS2 but being slightly faster
in mechanical strength development. With both normal grade and
high grade concrete mixes WES1 required less dosage to achieve
the target slump of 210 mm at equal w/c compared to the other su-
perplasticizers and was able to impart an excellent workability re-
tention over 90 minutes even at 35C.

CONCLUSIONS

According to the rheological measurements, the mini slump tests

and the concrete mixes showed it can be said that WES1 superplas-
ticizer can be advantageously used to impart an excellent retention
of workability in cementitious systems even at high temperature.
Its higher efficiency compared to PNS1, PNS2 and PCE1 was
showed in both normal grade and high grade concrete mixes pre-
pared and kept at 35 C using Arabian peninsula materials. It was
further demonstrated that even when low w/c and high dosages of
superplasticizer were present, the viscosity of the cementitious
mixes was not increasing when WES1 and PCE1 were used.
 REFERENCES

1. Collepardi M., Coppola L., Cerulli T., Ferrari G., Pistolesi C.,
Zaffaroni P., Queck F., Zero Slump-loss Superplasticized
Concrete, Proceedings of 18th Conference on Our World of
Concrete and Structures, Singapore, 25-27 August 1993, Vo-
lume XII, pp. 73-80.
2. Kinoshita M., Suzuki T., Soeda K., Nava T., Properties of
Methacrylic Water-Soluble Polymer as a Superplasticizer for
Ultra-Strength Concrete, Proceedings of 5th CANMET/ACI
International Conference on Superplasticizers and Other
Chemical Admixtures in Concrete, Ed. V. M. Malhotra, Rome,
Italy, 1997, pp.143-162.
3. T. Cerulli, P. Clemente, G. Ferrari, M. Gamba, A. Lolli, D. Pi-
noci, F. Surico, L. Badesso, Workability Extenders Super-
plasticizers, Proceedings of 8th CANMET/ACI International
Conference on Superplasticizers and Other Chemical Admix-
tures in Concrete, Ed. V. M. Malhotra, Sorrento, Italy, 2006,
pp. 263-272.
4. M. Nehdi, S. Al Martini, Effect of temperature on oscillatory
shear behavior of Portland cement paste incorporating chemi-
cal admixture , Journal of materials in civil engineering, De-
cember 2007,pp 1090-1100.
5. S. Al Martini, M. Nehdi, Effect of chemical admixtures rhe-
ology of cement paste ad high temperature, Journal of ASTM
International, Vol.4, No.3, Paper ID JAI100554.
6. C.T. Chen, L.J. Struble, H. Zhang, Using dynamic rheology
to measure cement admixture interctions, Journal of ASTM
International, March 2006, Vol.3, No.3, Paper ID JAI12787.
7. Y. Chen, R.P.Bright, Compatibility of superplasticizers with
cement, First American Drymix Mortar Conference admmc
one, Charlotte, NC, USA, June 04, 2008, pp. 44-52.
 Table 1. Characteristics of the superplasticizers used.

Designation Chemical Solid Density

in this paper Nature (%) (g/cm3)

PNS1 Polynaphtelenesulphonate 43 1,23

PNS2 Polynaphtelenesulphonate 40 1,20

PCE1 Polycarboxylate 27 1,09

Workability Extender
WES1 40 1,10
Superplasticizer
Table 2. Composition and characteristics of the cement used.

Cem Type I (ASTM C150 04)

SiO2 % 19.55
Al2O3% 5.87
Fe2O3% 3.15
CaO% 61.72
MgO% 1.47
K2O % 0.01
Na2O % 0.32
SO3 % 2.69
P2O5 % 0.15
TiO2 % 0.32
Cr2O3 % 0.13
MnO % 0.07
SrO % 0.28
Cl- % 0.07
L.O.I. % 2.76
Total % 98.56
C3 S % 46.0
C2 S % 27.0
C3 A % 8.2
C4AF % 6.9
Mean particle size m 16.06
 Table 3. Proportions of concrete mixes

Normal High
Strength Strength
w/c 0,50 0,38
Cement kg/m3 330 420
Washed Sand kg/m3 415 400
Unwashed Sand kg/m3 415 400
Gravel kg/m3 270 240
Gravel kg/m3 350 340
Gravel kg/m3 490 470

Table 4. Properties of fresh and hardened concrete mixes

Compressive
Admixture Slump (mm) Strength at
Cement Air 20C (MPa)
W/C
Dos. kg/m3 %
Type % 0 30 60 90 1d 7d 28d
bwc
PNS1 1,3 326 0,50 3,3 205 140 80 20 21,7 29,2 37,7
PNS2 1,1 328 0,50 2,2 210 170 120 95 16,9 28,7 38,4
PCE1 1,3 325 0,50 2,0 210 185 140 95 20,1 30,1 39,5
WES1 1,0 325 0,50 3,2 210 200 200 180 17,7 27,9 38,8
PNS1 1,6 416 0,38 2,3 210 180 90 35 16,9 42,5 51,2
PNS2 1,4 415 0,38 2,9 200 160 140 75 9,5 46,8 55,9
PCE1 1,5 417 0,38 2,3 210 180 110 60 16,5 43,6 53,7
WES1 1,3 412 0,38 1,9 210 230 200 180 10,3 44,6 57,8
Fig. 1. Cumulative Kept (%) vs. Diameter (mm) of Arabian
aggregates
tim e = 0 m in

1,0E+07
1,0E+06
Viscosity [Pa*s]

1,0E+05
1,0E+04
1,0E+03
1,0E+02
1,0E+01
1,0E+00
0,8 1 1,2 1,4 1,6
superplasticizers [%]
a) 0 min
tim e = 30 m in
Viscosity [Pa*s]

0,8 1 1,2 1,4 1,6

superplasticizers [%]
b) 30 min
tim e = 60 m in
Viscsoity [Pa*s]

0,8 1 1,2 1,4 1,6

superplasticizers [%]
c) 60 min
Fig. 2 Viscosity curves
Legend: = WES 1 = PCE 1
* = PNS 1 = PNS 2
 tim e = 0 m in

1,0E+04

1,0E+03

1,0E+02
Yield [Pa]

1,0E+01

1,0E+00

1,0E-01

1,0E-02
0,8 1 1,2 1,4 1,6
superplasticizers [%]
a) 0 min
tim e = 30 m in
Yield [Pa]

0,8 1 1,2 1,4 1,6

superplasticizers [%]
b) 30 min
tim e = 60 m in
Yield [Pa]

0,8 1 1,2 1,4 1,6

superplasticizers [%]
c) 60 min

Fig. 3 Yield curves

Legend: = WES 1 = PCE 1
* = PNS 1 = PNS 2
tim e = 0 m in

1,0E+08
1,0E+07
1,0E+06
max G' [Pa]

1,0E+05
1,0E+04
1,0E+03
1,0E+02
1,0E+01
1,0E+00
1,0E-01
0,8 1 1,2 1,4 1,6
superplasticizers [%]
at 0 min
tim e = 30 m in
max G' [Pa]

0,8 1 1,2 1,4 1,6

superplasticizers [%]
at 30 min.
tim e = 60 m in
max G' [Pa]

0,8 1 1,2 1,4 1,6

superplasticizers [%]
c) at 60 min

Fig. 4 Elastic modulus curves

Legend: = WES 1 = PCE 1
* = PNS 1 = PNS 2
 Leg- 1= 2 = 1.0 3 = 1.2 4 =
end: 0.8% % % 1.5%

Fig. 5 PNS2:stress sweep curves at 0 min

Fig. 6 PNS2:stress sweep curves at 0 min

Fig. 7 Mini slump test results