Effects of mineral admixtures on shear thickening of cement paste

Construction and Building Materials 126 (2016) 609–616

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Effects of mineral admixtures on shear thickening of cement paste Kunlin Ma a,b, Jin Feng a, Guangcheng Long a,b,⇑, Youjun Xie a,b a b

School of Civil Engineering, Central South University, Changsha 410075, China National Engineering Laboratory for Construction Technology of High-Speed Railway, Changsha 410075, China

h i g h l i g h t s
Influence of FA, LP and SL on shear thickening of cement paste are investigated.


ccrit, scrit, gmin and rheological index (n) are put forward to describe and compare shear thickening.


The addition of FA, LP and SL into cement paste make paste take on shear thickening easily.
Intensity sequence of shear thickening of mineral admixture to cement paste are SL, FA and LP.

a r t i c l e

i n f o

Article history: Received 4 May 2016 Received in revised form 17 August 2016 Accepted 20 September 2016

Keywords: Rheology Fly ash Slag Limestone powder Rheological parameter Shear thickening

a b s t r a c t Rheological behaviors are the essential workability characteristics of fresh concrete. Generally speaking, paste volume in self-compacting concrete (SCC) are no less than 340 L/m3. Therefore, rheology of paste could be mainly responsible for rheology of SCC. Mineral admixtures are important for modern concrete, taking great influence on concrete rheology. In this paper, an experiment was designed to investigate the shear thickening behaviors of cement paste replaced by fly ash (FA), slag (SL) and limestone powder (LP), respectively. Results show that plastic viscosity of paste discussed in this paper prominently decrease first and then increase with the shear rate increasing. Therefore, there exist critical shear rate (ccrit), critical shear stress (scrit) and minimum plastic viscosity (gmin) in rheological curves at the beginning of shear thickening taking place. The addition of FA, SL and LP into cement paste not only decrease the ccrit, scrit and gmin, resulting in cement paste easy to exhibit shear thickening, but also increase the rheological index (n), leading to shear thickening intensity augmenting. ccrit of cement paste with addition 30% SL drops to10 s1, displaying shear thickening response taking place easily compared to cement paste with the same FA and LP contents. Rheological index (n) of cement paste with SL is also larger than cement paste with FA and LP at the same mineral admixtures contents. The shear thickening intensity sequence of cement paste under the effects of mineral admixtures are SL, FA and LP, respectively. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Due to the characteristics of higher fluidity and segregation resistance of fresh concrete, lower porosity, higher strength and durability of hardened concrete, self-compacting concrete (SCC) is believed as a milestone achievement in modern concrete technology[1]. Mineral admixtures materials are indispensable compositions for modern concrete, especially for SCC. These mineral admixtures materials come from wastes and by-products of industry including fly ash (FA), ground granulated blast furnace slag (GGBS), and silica fume (SF) et al. Replacement part of cement with ⇑ Corresponding author at: School of Civil Engineering, Central South University, Changsha 410075, China. E-mail address: [email protected] (G. Long). http://dx.doi.org/10.1016/j.conbuildmat.2016.09.075 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.

mineral admixtures materials is an effective way of improving SCC properties and reducing carbon footprint of concrete production [2–4]. Rheological properties are the essential workability characteristic of SCC, which is very important for fresh SCC [5,6]. Shear thickening is an important aspect of concrete rheology, referring to the growth of plastic viscosity with shear rate increasing. Shear thickening is hard to avoid for SCC, but shear thickening is strongly unwanted in the process of mixing or pouring concrete for it can interfere with product quality and may even lead to damaging processing equipment. In some cases, the process of producing, mixing, transporting and pumping concrete would result in shear rate changing, giving rise to an obvious change of rheological behaviors and leading to shear thickening taking place in fresh SCC [7–9]. Generally speaking, paste volume in SCC are no less than 340 L per m3 [10,11], and paste mainly take on lubricating and

610

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

packing effects among aggregate. Therefore, rheological properties of paste play a key role in workability of SCC. Now, a large number of FA, SL and LP are used in SCC widely, resulting in rheological properties of SCC being susceptible. Replacement of cement with mineral admixtures will change the rheological behaviors of SCC greatly, including shear thickening [12]. Martin Cyr et al. pointed out [13] that the intensity of the shear thickening depended on the nature of mineral admixtures. It could be amplified (metakaolin), unchanged (quartz, fly ashes) or reduced (silica fumes). Yahia A carried out experiment and found [14,15] the addition of 8% silica fume particles reduced the shear thickening response. Diamantonis N et al. showed [16] that limestone (40%) could improve the rheological behavior of cement paste, and the synergy of limestone (20%) and fly ash (20%) could lead to higher packing density. Studies showed that there were a shear thinning stage and shear thickening stage in rheological curves of fly ash-cement paste. And replacement cement with FA makes paste strengthen shear thickening intensity [17,18]. The shear-thickening behavior observed with cement-based materials could be explained by either an order-disorder transition theory or cluster theory [19,20]. Barnes H A [21] studied shear thickening deeply and proposed that the severity of shear thickening depended on the particle concentration in proportion to the maximum packing fraction. The intensity of shear thickening continuously decreased as the mixture became more and more polydisperse. Shear thickening took place only when the suspensions were deflocculated. Particles that present irregular shape tend more easily to show a shear thickening behavior. All these studies above are useful for people to understand rheological properties of SCC. Practically, because of the different of shape, size, chemical compositions and physical structure, mineral admixtures will take on different effects on shear thickening of cement paste. As results, comparing and evaluating the effects of mineral admixtures on cement paste are necessary. In SCC, the addition of FA, SL and LP are more than other mixture admixtures. Therefore, the objectives of this study were to experimentally study the effects of FA, SL and LP on shear thickening of cementbased materials and put forward parameters to judge and compare, mainly aimed at comparing and evaluating the shear thickening behaviors of cement paste under such mineral admixtures. The significance of this study is to distinguish the roles of different mineral admixtures on rheological behaviors of cement-based materials and provide useful information for SCC design and preparation.

plasticizer was added into all paste so that the experimental conditions could be same. 2.2. Rheology test 2.2.1. Rheological curves A room with ambient temperature (25 ± 2) °C and relative humidity (70 ± 5)% was used to carry out this test. Co-axial cylinder rotary viscosimeter produced by ANTON PAAR Company with Rheolab QC type was used to determine the rheological curves of different paste. In order to keep mixing uniformity, electric mixer of one-phase was used to mix paste. First, water and superplasticizer were mixed together by electric mixer with 500 rpm over a time span of 1 min. Second, cement and mineral admixtures were added into the admixtures of water and superplasticizer with 1500 rpm over a time span of 2 min. After being mixed samples were immediately loaded to co-axial cylinder rotary viscosimeter for testing. The rheological curves were tested at 2 min after mixing. The shearing rate was then increased gradually from 1 S1 to 300 S1. Rheological equation and related rheological parameters were calculated according to mathematic fitting. 2.3. Methods for data analysis The shear-stress and shear-rate data came from experiment and the rheological parameters were fitted through Herschel-Bulkley (H-B) model. This model can be expressed by Eq. (1)

c ¼ 0; s < s0

s ¼ s0 þ gcn ; s P s0

ð1Þ

where s is shear stress (Pa); s0 is yield stress (Pa); g is plastic viscosity coefficient (Pas); c is shear rate (S1); n is rheological index. When s0 is equal to 0 and n is equal to 1, the fluid is called Newtonian fluid. When s0 is not equal to 0 and n is equal to1, the fluid is called Bingham fluid. When n is more than 1, the fluid is called dilatant fluid. Namely, shear thickening takes place. In shear thickening stage, the larger the rheological index (n) is, the stronger the intensity of shear thickening is. 2.4. Particle size distribution test Particle size distribution of cement, fly ash, limestone powder and slag were taken by auto laser particle size analyzer produced by Jinan Runzhi Science and Technology Ltd. Specific surface area of cement and other mineral admixtures were tested through BET method.

2. Materials and method 3. Results 2.1. Materials and admixtures proportion 3.1. Effect of mineral admixtures on rheological curves Ordinary Portland cement P.O 42.5 (C) satisfied Chinese standards GB 175 was prepared [22]. Qualified and densified FA, SL and LP were used. Fig. 1(a)(d) displayed the tested results of particle size distribution of C, FA, SL and LP. The chemical and physical properties of C, FA, SL and LP were given in Table 1. In order to disperse the cementitious materials and reduce agglomeration, a polycarboxylate-based superplasticizer with solid content 25.4% and specific gravity 1.09 was used. Mixing water (W) was deionized. In this study, five mineral admixtures contents, namely 0, 20%, 30%, 40% and 50%, each expressed as a percentage by mass of the total cementitious materials, were adopted for the design of the paste samples, respectively. Water to binder ratio was kept constantly in 0.28 for all, and a dosage of 0.4% (mass fraction) super-

Studies of rheological behaviors of cement paste addition of FA, SL and LP can provide useful information for understanding mechanism of controlling shear thickening response with different mineral admixtures concentration. Fig. 2(a)(c) show the curves of shear rate vs. plastic viscosity under the influence of FA, LP and SL. Fig. 3(a)(c) are the curves of shear rate vs. shear stress with the addition of FA, LP and SL. Fig. 2(a)(c) present that no matter what kind of mineral admixtures, with the growth of shear rate, the plastic viscosity decreases obviously first and then increases gradually. In other words, shear thinning takes place first and then shear thickening happens. At the beginning of shear thinning transforming to shear thickening, being points A,B,C,D and E presented in Fig. 2(a), points

611

60

3 40

2

20

1 1

10

100

0

100

5

80

4

60

3 40

2

20

1 0

1

0 100

10

Particle size (µm)

(b) Particle size distribution of FA 100

5

80

4

60

3 40

2

20

1 1

10

100

0

Differental content (%)

Particle size (µm)

(a) Particle size distribution of C 6

0

6

6

100

5

80

4

60

3 40

2

20

1 0

0

10

20

30

40

Cumulative content (%)

4

Differental content (%)

80

Cumulative content (%)

5

0

Differental content (%)

100

Cumulative content (%)

Differental content (%)

6

Cumulative content (%)

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

0 50

Particle size (µm)

Particle size (µm)

(c) Particle size distributionof SL

(d) Particle size distribution of LP

Fig. 1. Particle sizes distribution of raw materials.

Table 1 Chemical and physical properties of C, FA, SL and LP. No.

Ordinary cement (C)

Fly ash (FA)

Slag (SL)

Limestone powder (LP)

SiO2 Fe2O3 Al2O3 CaO MgO SO3 CaCO3 C3A C3S C2S C4AF Loss on ignition Specific gravity (kg/m3) Specific surface area (m2/kg) D[V,0.5] (lm) Compressive strength at 28d (MPa) Activity index at 1d (%) Active index at 28d (%)

20.3 3.2 4.8 62.3 4.2 2.75 – 7.3 56.2 15.8 9.7 2.4 3120 344 16.211 49.5 100 100

51.8 5.0 26.4 4.1 1.0 2.0 – – – – – 3.09 2450 486 8.249 – 85 95

35.5 6.6 16.1 36.9 12.9 1.02 – – – – – 1.05 2730 465 9.157 – 96 120

0.45 0.02 0.01 58.3 – – 97.8 – – – – – 2670 573 6.770 – 65 60

A,F,G,H and J presented in Fig. 2(b) and points A,K,L,M, and N presented in Fig. 2(c) and (d), the plastic viscosity of paste decrease to the lowest. Therefore, it is easy to find that there exists a minimum plastic viscosity (gmin) in the curve of shear rate vs. plastic viscosity, and the corresponding shear rate can be expressed as a critical shear rate (ccrit). After the shear rate is beyond ccrit, the paste plastic viscosity begin to develop along with the increasing in shear rate. Fig. 2(a)(b) illustrate that with FA and LP contents increasing, the gmin and ccrit of paste decrease greatly. Fig. 2(c) (d) show that when SL contents are about 20%30%, gmin and ccrit decrease to the lowest, but when SL contents reach to 40% and 50%, gmin and ccrit increase. Fig. 2(a)(d) also show that gmin and ccrit of paste with mineral admixtures are lower than plain cement paste (group A), showing that after addition of FA, SL and LP cement paste are easy to exhibit shear thickening.

Corresponding to ccrit and gmin in shear rate vs. plastic viscosity curves shown in Fig. 2, there are ccrit and critical shear stress (scrit) in shear rate vs. shear stress curves shown in Fig. 3. Points A’,B’,C’, D’ and E’ in Fig. 3(a), points A’,F’,G’,H’ and J’ in Fig. 3(b) and points A’,K’,L’,M’ and N’ in Fig. 3(c) show these points of ccrit and scrit. From the analysis above, it is found that addition of mineral admixtures take on prominently influence on rheological parameters of ccrit, gmin and scrit. Therefore, ccrit, gmin and scrit could be taken as parameters to describe and compare shear thickening of cement-based materials. Fig. 3(a)(c) also present the change of rheological index (n). Rheological index (n) can well describe the intensity of shear thickening. The increase of rheological index (n) means the growing shear thickening intensity of paste. It is found that rheological index (n) increases with FA,SL and LP content increasing. The details analysis and compare of mineral

612

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

0.5 0.4

A

0.3

B 0.2

D

0.1

(

E

0.0

, crit

C:LP=100:0 C:LP=80:20 C:LP=70:30 C:LP=60:40 C:LP=50:50

0.5 0.4

A

0.3

F

H G

0.2

J

0.1

) min

(

0.0 0

50

0.6

Plastic viscosity (Pa.S)

C

100

150

250

300

150

200

(a) addition of FA

(b) addition of LP

0.3

A

0.20

Fig.2(d)

0.2 0.1

(

0.0 50

100

Shear rate (s-1)

0.4

0

50

0

Shear rate (s-1)

C:SL=100:0 C:SL=80:20 C:SL=70:30 C:SL=60:40 C:SL=50:50

0.5

200

Plastic viscosity (Pa.S)

Plastic viscosity (Pa. S)

C:FA=100:0 C:FA=80:20 C:FA=70:30 C:FA=60:40 C:FA=50:50

Plastic viscosity (Pa. S)

0.6

0.6

100

150

200

crit

250

,

min

)

(

crit

,

min

crit

,

min

250

)

300

)

0.15

0.10

C:SL=80:20 C:SL=70:30 C:SL=60:40 C:SL=50:50

N

0.05

M K

0.00 0

L 50

100

150

200

300

Shear rate

Shear rate (s-1)

250

300

350

(s-1)

(d)

(c) addition of SL Fig. 2. Shear rate vs. Plastic viscosity.

admixtures on the influence of rheological index (n) are presented in Part 3.3. 3.2. Effects of mineral admixtures on ccrit, gmin and scrit Table 2 shows the rheological parameters, fitted from HerschelBulkley (H-B) model, of cement paste with different mineral admixtures contents. Results in Table 2 illustrate that the ccrit of plain cement paste is 148 s1, indicating plain cement paste is shear thinning paste until the shear rate increases to high rate. However, with the increase of mineral admixtures, ccrit decreases greatly, illustrating that cement paste are easy to take on shear thickening after addition of FA, SL and LP. From the studied results of rheological behaviors of cement paste replaced by FA, SL and LP, it is not hard to find the same effects of these mineral admixtures on cement paste are that they decrease ccrit, gmin and scrit of cement paste. However, there is a great difference for mineral admixtures to influence shear thickening parameters of cement paste. Fig. 4(a)(c) display the influence of mineral admixtures on shear thickening parameters including ccrit, gmin and scrit. There are close quadratic polynomial relationships between mineral admixtures contents and shear thickening parameters. Fig. 4(a)(c) show ccrit, gmin and scrit decrease along with FA and LP contents increasing. SL also decreases the rheological parameters compared to plain cement paste. However, ccrit, gmin and scrit decrease first and then increase with the growth of SL contents. When SL contents are about 20%30%, the ccrit, gmin and scrit drop to the lowest. As can be seen from Fig. 4(a), when mineral admixtures contents are under 40%, keeping the same mineral admixtures contents, ccrit of cement paste with LP is

almost the highest, but ccrit of cement paste with SL is almost the lowest. The ccrit of cement paste with 20%30% SL are nearly below 15 s1, which means shear thickening is very easy to take place. As can be seen from Fig. 4(b) and (c), in the same mineral admixtures contents, gmin and scrit of cement paste addition of SL are almost the lowest. All these indicate that addition of FA, LP and SL can make cement paste exhibit shear thickening easily. Moreover, SL can cause cement paste display shear thickening under very low ccrit, gmin and scrit condition, especially when SL contents are between 20%30%.

3.3. Effects of mineral admixtures on rheological indexes Rheological index (n) is another important parameter to evaluate and compare the rheological behaviors of Non-Newton fluid. In the process of shear thickening, large rheological index means that increasing same shear rate will give rise to large shear stress, leading to shear thickening intensity developing. Fig. 5 displays the influence of FA,SL and LP on the rheological index (n). It is also observed that with the growth of FA, SL and LP, rheological index (n) is on the rise. And there are good relations between mineral admixtures contents and rheological index (n). All the relative coefficients (R) are no less than 0.9. Fig. 5 also shows that when the paste keep the same mineral admixtures mass fraction, rheological index (n) of cement paste with SL are the largest, showing cement paste with SL take on larger shear thickening intensity compared to FA and LP. Therefore, the intensity sequence of shear thickening of mineral admixtures on cement paste should be SL, FA and LP, respectively.

613

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

120

80

(

60 40

E

crit

C

D

crit

100 n=1.348

)

n=1.526 n=1.516

A

B

C:LP=100:0 C:LP=80:20 C:LP=70:30 C:LP=60:40 C:LP=50:50 ( crit crit )

n=1.298

Shear stress (Pa)

100

Shear stress (Pa)

120

C:FA=100:0 C:FA=80:20 C:FA=70:30 C:FA=60:40 C:FA=50:50

20

80 60

G

H

40

F

n=1.298 n=1.312 n=1.381

A n=1.397

J

20

n=1.631

n=1.403

0

0 0

50

100

150

200

250

0

300

50

100

-1

150

200

250

300

-1

Shear rate (s )

Shear rate (s )

(a) addition of FA

(b) addition of LP

120

C:SL=100:0 C:SL=80:20 C:SL=70:30 C:SL=60:40 C:SL=50:50

Shear stress (Pa)

100 80

(

60 40 20

K

L

crit

M

crit

n=1.298

n=1.458

) n=1.672

A

N

n=1.681 n=1.769

0 0

50

100

150

200

250

300

-1

Shear rate (s )

(c) addition of SL Fig. 3. Shear rate vs. shear stress.

Table 2 Rheological parameters of cement paste with different mineral admixtures. No.

ccrit (s1)

gmin (Pa.s)

scrit (Pa)

n

R

C1 C-FA2 C-FA3 C-FA4 C-FA5 C-LP2 C-LP3 C-LP4 C-LP5 C-SL2 C-SL3 C-SL4 C-SL5

148 78.8 78 49 31 139 108 101 56.7 14.8 10 56.7 78.1

0.281 0.206 0.175 0.096 0.058 0.244 0.172 0.159 0.136 0.0163 0.013 0.0559 0.0772

41.6 16.3 13.7 4.8 1.9 33.9 18.5 16 7.72 0.24 0.13 3.17 6.03

1.298 1.348 1.526 1.516 1.631 1.312 1.381 1.397 1.403 1.458 1.627 1.681 1.769

0.9911 0.9922 0.9939 0.9947 0.9999 0.9962 0.9956 0.9972 0.9959 0.9974 0.9985 0.9982 0.9979

Note: C1 means plain cement paste; C-FA2C-FA5 mean cement paste addition of 20%50% FA (mass fraction); C-LP2C-LP5 mean cement paste addition of 20%50% LP (mass fraction); C-SL2C-SL5 mean cement paste addition of 20%50% SL (mass fraction).

4. Discussion 4.1. Theory about shear thickening In literatures, there were two theories to explain shear thickening. The first one is the formation of (hydro-) clusters theory [22–24]. In this theory, shear thickening could be the consequence of the formation of transient ’’hydrodynamic clusters’’ that jam the flow. These clusters are composed of compact groups of particles formed when the shear forces are sufficient to drive particles nearly into contact. Under these conditions, the increase in shear rate leads to an increase in the viscosity

as the clusters become larger and larger. The second one is called order-disorder transition theory [19,25,26]. This theory explains the mechanism of shear thickening is that particles are ordered into layers, shifts at a critical shear rate, to a disordered state where this ordering is absent. This less -ordered structure dissipates more energy while flowing due to particles ‘‘jamming”, and hence the viscosity increases. The critical shear rate corresponds to a state where the hydrodynamic forces are strong enough to overcome the repulsive interparticle forces that favor the well-ordered structure. Some studies show that shear thickening of cement paste can be well explained by means of clusters theory [8,9].

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

200

2

Y =145.72-3.21 X+0.018 X ,R=0.9691 2 Y =144.49-9.20 X+0.16 X ,R=0.9436 2 Y =148.16+0.11 X-0.037 X ,R=0.9637

180

Critical shear rate (s-1)

Minimum plastic viscosity (Pa.s)

614

160 140

LP

120 100 80 60 40

FA SL

20 0

10

0

20

30

40

50

0.40

2

Y =0.28-0.0029 X-3.32E-5 X ,R=0.9857 2 Y =0.27-0.016 X+2.65E-4 X ,R=0.9573 2 Y =0.28-0.0031 X+6.62E-7 X ,R=0.9379

0.35 0.30 0.25

LP

0.20 0.15

FA

0.10 0.05

SL

0.00 10

0

20

30

40

Mineral admixture (%)

Mineral admixture (%)

(a)

(b) Critical shear stress (Pa)

60

50

2

Y =41.15-1.34 X+0.011 X ,R=0.9849 2 Y =42.34-0.56 X-0.0027 X ,R=0.9524 2 Y =40.47-2.55 X+0.038 X ,R=0.9653

50 40

FA

30 20 10

LP

0

SL

-10 0

10

20

30

40

50

Mineral admixture (%)

(c) Fig. 4. Influence of mineral admixtures on shear thickening parameters.

2.0

Y=1.295+0.00969X R=0.9905 SL Y=1.274+0.00676X R=0.9461 FA

Rhological index

1.8

1.6 1.4 LP

1.2

Y=1.290+0.0024X R=0.9335 1.0 0

10

20

30

40

50

Mineral admixture (%) Fig. 5. Influence of mineral admixtures contents on rheological index (n).

4.2. Parameters influencing shear thickening There are several parameters influencing shear thickening behavior, including in particle size, particle shape, particle distribution, volume fraction and interaction between particles [27]. Suspensions that contain particles up to a few tens of micrometers can exhibit shear thickening obviously. Particles that present irregular shape tend more easily to show a shear thickening behavior. The intensity of shear thickening continuously decreases as the mixture becomes more and more polydisperse. The severity of shear thickening depends on the particle concentration in proportion to the maximum packing fraction. Generally, Shear thickening takes place only when the suspensions are deflocculated. In this paper, D[V,0.5] of FA,SL and LP are 8.249 lm,9.157 lm and 6.770 lm. And the specific surface of FA, SL and LP are 486 m2/kg, 456 m2/kg and 573 m2/kg (seen from Table 1), respec-

tively. Obviously, the sequence of particle size is SL, FA and LP. As far as particle distribution (seen from Fig. 1) concerned, particle distribution of FA and SL are almost the same wide distribution, but LP is narrow distribution. With regard to particle shape and constitution, FA is glassy state spherical particles containing active Si2O and CaO, SL is also glassy state but irregular shape containing active Si2O and CaO, and LP is irregular shape containing mainly a little active CaCO3. So there are some different from the shape and constitution of such three kinds of mineral admixtures. The density of FA, LP and SL used in this study are 2.45 g/cm3, 2.67 g/cm3 and 2.73 g/cm3, respectively, which are lower than that of cement. So when these mineral admixtures replace the same mass fraction of cement, the volume of paste especially the particle volume fraction in paste would change greatly. Studies show the shear thickening has great relations with the particle concentration in proportion to the maximum packing fraction. The growth of volume fraction in paste is a benefit for the development of shear thickening [15,23]. Fig. 6(a)(d) provide the influence of particle volume fraction on shear thickening parameters. Just as can be seen from Fig. 6(a)(c), as for FA and LP, with particle volume fraction increasing, shear thickening parameters including ccrit, gmin and scrit decrease greatly, which means that shear thickening of cement paste with FA and LP is easy to take place. However, as for SL, with particle volume fraction increasing, shear thickening parameters first decrease obviously and then increase. When SL content is about 30% the shear thickening parameters decrease to the lowest. When particle volume fraction is same, cement paste with SL displays lower ccrit, gmin and scrit compared to FA and LP. Fig. 6(d) presents that the increase in paste volume fraction results in rheological index (n) being on the rise, but the rheological index (n) of paste with SL increases faster than FA and LP. It shows that among the three mineral admixtures studied in this paper, the addition of SL into cement will decrease ccrit, gmin and scrit greatly and increase rheological index (n) obviously. These

615

Critical shear rate (s -1)

240

2

Y =20883-71605.6 X+61443.6 X ,R=0.979 2 Y =418118-1.5E6 X+1.41E6 X ,R=0.950 2 Y =-52840+200224 X-189049.3 X ,R=0.948

200 160

LP

120

FA

80 40 0 0.53

SL 0.54

0.55

0.56

0.57

Minimum plastic viscosity (Pa.s)

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

Particle volume fraction

0.40

2

Y =-13.8+58.09 X-59.13 X ,R=0.987 2 Y =651.8-2390.5 X+2191.5 X ,R=0.960 2 Y =37.8-130.1 X+111.9 X ,R=0.953

0.35 0.30 0.25

FA

0.20 0.15

LP

0.10 0.05

SL

0.00 0.53

1.8

Rhological index

Critical shear stress (Pa)

40

LP FA SL

Y=-13.23+27.21X R=0.9905

SL

0.54

0.55

0.56

0.57

Particle volume fraction

Y=-3.95+9.79X R=0.943

FA

1.6 1.4

LP

1.2

Y=-1.832+5.85X R=0.9306

0 0.53

0.57

2.0

2

20

0.56

(b) volume fraction vs. ηmin

Y =939-864 X-1494 X , R=0.9466 2 Y =11271.1-39470.7X+34562.6 X , R=0.9883 2 Y =80020-292111 X+266576 X ,R=0.9581

60

0.55

Particle volume fraction

(a) volume fraction vs. γcrit 80

0.54

1.0 0.53

0.54

0.55

0.56

0.57

Particle volume fraction

(c) volume fraction vs. τcrit

(d) volume fraction vs. n

Fig. 6. Influence of particle volume fraction on shear thickening parameters.

difference cannot be well explained by the influence of particle size, particle size, particle shape, particle distribution and volume fraction. Table 1 exhibits that the activity index of SL is the biggest among FA, SL and LP, indicating that SL contains more active particles of Si2O and CaO. Therefore, the shear thickening different between paste with PL and paste with FA and LP are most likely to come from the high activity of SL particles. However, the real mechanisms of particle properties to shear thickening still need to be further studied. 5. Conclusions In this study, the influence of FA, LP and SL on shear thickening of cement paste are investigated experimentally. Shear thickening parameters, including gmin, ccrit, scrit, and rheological index (n) are introduced in order to compare and evaluate the shear thickening degree and intensity. The results are as follows. (1) The increase in shear rate makes the plastic viscosity of cement paste decrease obviously first and then increase slowly. Therefore, in the curves of shear rate vs. plastic viscosity and shear rate vs. shear stress, there exist minimum plastic viscosity (gmin), critical shear rate (ccrit) and critical shear stress (scrit) at the beginning of shear thinning transforming into shear thickening. (2) FA, LP and SL give different influence on rheological behavior of cement paste. Plain cement paste displays shear thinning even at high shear rate. But replacement cement with FA, LP and SL prominently decrease ccrit, gmin and scrit, leading to cement paste easy to take on shear thickening. Rheological index (n) of shear thickening is on the rise with mineral

admixtures contents increasing, result in the development of shear thickening intensity. (3) Under the same mass fraction and particle volume fraction condition in cement paste, SL takes more prominent effects on shear thickening intensity than FA and LP, especially when mass fraction of slag is about 20%30%. The intensity sequence of shear thickening of mineral admixtures on cement paste should be SL, FA and LP, respectively. Compared to particle concentration in cement paste, particle properties would exhibit great influence on shear thickening.

Acknowledgements The authors highly acknowledge the financial support received from the National Key Basic Research and Development Plan of China (No. 2013CB036200), The National Natural Science Foundation of China (No. 51678569, 51678568) and Teacher Science Foundation of Central South University of China (No. 2014JSJJ013). References [1] P.-C. Aïcin, Cements of yesterday and today-concrete of tomorrow, Cem. Concr. Res. 30 (9) (2000) 1349–1359. [2] V. Corinaldesi, G. Moriconi, The role of industrial by-products in selfcompacting concrete, Constr. Build. Mater. 25 (8) (2011) 3181–3186. [3] A.K.H. Kwan, I.Y.T. Ng, Improving performance and robustness of SCC by adding supplementary cementitious materials, Constr. Build. Mater. 24 (11) (2010) 2260–2266. [4] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious materials, Cem. Concr. Res. 41 (12) (2011) 1244–1256.

616

K. Ma et al. / Construction and Building Materials 126 (2016) 609–616

[5] F. Rosquoët, A. Alexis, A. Khelidj, A. Phelipot, Experimental study of cement grout: rheological behavior and sedimentation, Cem. Concr. Res. 33 (5) (2003) 713–722. [6] K.H. Khayat, Viscosity -enhancing admixtures for cement-based materials-an overview, Cem. Concr. Compos. 20 (2) (1998) 171–188. [7] F. Curcio, B.A. DeAngelis, Dilatant behavior of superplasticized cement paste containing metakaolin, Cem. Concr. Res. 28 (5) (1998) 629–634. [8] D. Feys, R. Verhoeven, G.D. Schutter, Why is fresh self-compacting concrete shear thickening?, Cem Concr. Res. 39 (6) (2009) 510–523. [9] D. Feys, R. Verhoeven, G.D. Schutter, Fresh self-compacting concrete, a shear thickening material, Cem. Concr. Res. 38 (7) (2008) 920–929. [10] CH CCES02.Guide to design and construction of self-compacting concrete, 2004. [11] CH JGJ/T283. Technical specification for application of self-compacting concrete, 2012. [12] K.L. Ma, G.C. Long, Y.J. Xie, The properties of fresh SCC used for filling layer of slab ballastless track of high speed railway, in: C.J. Shi, Z.H. Ou, K.H. Khayat (Eds.), ‘Design, Performance and Use of Self-Consolidating Concrete’, Proc of the 3rd Int RILEM Symp. China: Xiamen, 2014, pp. 613–621. [13] Martin Cyr, C. Legrand, M. Mouret, Study of the shear thickening effect of superplasticizers on the rheological behaviour of cement paste containing or not mineral admixtures, Cem. Concr. Res. 30 (9) (2000) 1477–1483. [14] A. Yahia, Shear-thickening behavior of high-performance cement grouts— influencing mix-design parameters, Cem. Concr. Res. 41 (2) (2011) 230–235. [15] A. Yahia, Effect of solid concentration and shear rate on shear-thickening response of high-performance cement suspensions, Constr. Build. Mater. 53 (2) (2014) 517–521.

[16] N. Diamantonis, I. Marinos, M.S. Katsiotis, et al., Investigations about the influence of fine additives on the viscosity of cement paste for self-compacting concrete, Constr. Build. Mater. 24 (2010) 1518–1522. [17] K.L. Ma, G.C. Long, Y.J. Xie, Rheological properties of compound paste with cement-fly ash-limestone powder, J. Chin. Ceram. Soc. 41 (5) (2013) 582–586. [18] K.L. Ma, G.C. Long, Y.J. Xie, Effects of fly ash on shear thinning and shear thickening of cement paste, J. Chin. Ceram. Soc. 41 (9) (2014) 1209–1218. [19] A.A. Catherall, J.R. Melrose, R.C. Ball, Shear thickening and order–disorder effects in concentrated colloids at high shear rates, J. Rheol. 44 (1) (2000) 1–25. [20] R.L. Hoffman, Explanations for the causes of shear thickening in concentrated colloidal suspensions, J. Rheol. 42 (1) (1998) 111–123. [21] H.A. Barnes, Shear-thickening, ‘‘Dilatancy” in suspension of non-aggregation solid particles dispersed in Newtonian liquids, J. Rheol. 33 (2) (1989) 329–366. [22] CH GB175. Common Portland cement, 2007. [23] J.F. Brady, G. Bossis, Stokesian dynamics, Ann. Rev. Fluid Mech. 201 (1) (1988) 111–157. [24] G. Bossis, J.F. Brady, The rheology of brownian suspensions, J. Chem. Phys. 91 (3) (1989) 1866–1874. [25] R.L. Hoffman, Discontinuous and dilatant viscosity behavior in concentrated suspensions: I. Observations of a flow instability, Trans. Soc. Rheol. 16 (1972) 155–173. [26] W.H. Boersma, J. Laven, N. Steinh, Computer simulations of shear thickening of concentrated dispersions, J. Rheol. 39 (5) (1995) 841–860. [27] B.J. Maranzano, N.J. Wagner, The effect of particle size on reversible shear thickening of concentrated colloidal dispersions, J. Chem. Phys. 114 (23) (2001) 10514–10527.