EXPERIMENTAL CHARACTERISATION OF INJECTION GROUTS
1INCORPORATING HYDROPHOBIC SILICA FUME
2Luis G. Baltazar1*, Fernando M.A. Henriques2, Douglas Rocha3and Maria T. Cidade4 3
1 Departamento de Engenharia Civil, Faculdade de Ciências e Tecnologia, Universidade 4
NOVA de Lisboa, 2829-516 Caparica, Portugal. 5
2Departamento de Engenharia Civil, Faculdade de Ciências e Tecnologia, Universidade 6
NOVA de Lisboa, 2829-516 Caparica, Portugal. 7
3Departamento de Engenharia Civil, Faculdade de Ciências e Tecnologia, Universidade 8
NOVA de Lisboa, 2829-516 Caparica, Portugal. 9
4Departamento de Ciência dos Materiais e Cenimat/I3N, Faculdade de Ciências e 10
Tecnologia, Universidade NOVA de Lisboa, 2829-516 Caparica, Portugal. 11
* Corresponding author. Tel.: +351 21 2948580; E-mail addresses: luis.baltazar@fct.unl.pt 12
ABSTRACT 13
The present paper put forward a new hydrophobic silica fume and assesses their contribution 14
on the performance improvement of grouts for stone masonry consolidation. The experiments 15
were conducted using different dosages of hydrophobic silica fume with natural hydraulic 16
lime grouts in the presence of a polycarboxylate-based high range water reducer. Results 17
revealed that the effects of hydrophobic silica fume on properties of natural hydraulic lime 18
grouts optimise the use of ordinary silica fume. Remarkable rheological performance was 19
obtained in the presence of hydrophobic silica fume: the plastic viscosity and yield stress were 20
reduced compared with the ordinary silica fume. On mechanical strength aspects it was found 21
that hydrophobic silica fume slightly affect the flexural and compressive strength; however 22
the values obtained are suitable for the old stone masonry consolidation purposed. It was also 23
observed that water capillarity was substantially reduced, namely due to the water repellent 24
behaviour of hardened grouts. This study suggests that the promising effectiveness of this new 25
silica fume on injection grouts open the way to be used in many other applications. 26
Keywords: Masonry; Grout; Rheology; Natural hydraulic lime; Hydrophobic silica fume.
27
Manuscript R2 Click here to download Manuscript Revised Manuscript
INTRODUCTION 28
Stone masonry is a simple and durable constructive technique that is present in the large 29
majority of old buildings in many urban centres in Europe. However, these masonries exhibit 30
a non-monolithic behaviour which increases their vulnerability against several damaging 31
actions, like seismic events (Binda et al., 1997). Grout injection is an effective technique to 32
repair and consolidate stone masonry walls by improving the cohesion between the masonry 33
elements (Corradi et al., 2002; Chaudhry, 2007). Grouts should be well designed in order to 34
have adequate performance (i.e. considerable fluidity and penetrability). The use of ultra-fine 35
materials, such as silica fume, in cementitious materials is one of the main trends in 36
development of materials with improved performance (Artelt and Garcia 2008; Sobolev et al. 37
2016). 38
Silica fume is a by-product of the smelting process in the electrometallurgy industry, which 39
has been mainly used in cementitious materials as a binder replacement. The effect of silica 40
fume on cementitious materials is now well known. Several studies (Wei-Hsing, 1997; 41
Shannag, 2000) have shown that the properties of cementitious materials can be improved in 42
the presence of silica fume. Although, the use of silica fume should not be indiscriminate in 43
order to take full advantage of its properties. Previous studies have shown that silica fume can 44
have harmful effects on rheology of injection grouts (Park et al., 2005; Baltazar et al., 2013), 45
which result in grouts with low flowability. 46
Knowing that the decrease of grout workability is associated with the high water demand of 47
ordinary silica fume, the authors of this paper put forward a pre-treatment with 48
poly(dimethylsiloxane) (PDMS) solution on ordinary silica fume in order to get a silica fume 49
with hydrophobic behaviour. PDMS is a polymer belonging to the family of silicones. One of 50
the first applications of PDMS was in analytical chemistry, mostly in sample preparation 51
techniques (Seethapathy and Górecki, 2012). However, over the last decades, the applications 52
of PDMS have expanded into different areas, such as in the manufacture of textiles, personal 53
care products, or acting as additive to provide desirable properties in polish formulations, 54
water proofers and other surface treatments (Lecomte et al., 2013). 55
It is expected that the majority of fluidity and penetrability problems of injection grouts may 56
be easily overtaken with hydrophobic silica fume. This can be explained by the combined 57
effect of hydrophobicity and spherical shape of silica fume particles that cause a pure ball-58
bearing action between the bigger and elongated natural hydraulic lime (NHL) particles. The 59
choice of NHL as binder is a consequence of its compatibility with pre-existing materials in 60
old masonries. According to other authors (Ballantyne, 1996; Binda et al., 2006; Biçer-Şimşir 61
et al., 2009) the NHL binders are widely agreed to be the most compatible material for the 62
conservation of heritage buildings. It should be noted, however, that the major advantage of 63
hydrophobic silica fume over other materials (that also led to improved hydrophobicity) is 64
that besides to obtain hardened hydrophobic grouts, which improve their performance from 65
durability point of view, the fresh properties, namely the rheological ones, are expected to be 66
improved as well. 67
This work provides insight into the influence of hydrophobic silica fume on the rheological 68
properties of NHL-based grouts with different concentrations of silica fume (10%, 20% and 69
30% replacement by NHL mass). Furthermore, the effects of this new silica fume on the 70
stability, water absorption, porosity, pore size distribution, pozzolanic reactions and 71
mechanical strength of NHL-based grouts were also evaluated. 72
MATERIALS AND GROUTS COMPOSITION 73
Materials characteristics 74
A NHL from Secil-Martigança produced according with the European Standard EN459-75
1:2010 was used in the experiments, with nominal strength of 5 MPa and specific surface of 76
9400 cm2/g (blaine specific surface). The density of this NHL is of 2.7 g/cm3. The silica fume
77
used was an undensified silica fume produced by MAPEI. The specific surface of this silica 78
fume is of 17.5 m2/g (BET method). The particle size is comprehended between 2 and 30 μm,
79
80% being below 10 μm. The chemical properties of both NHL and silica fume are listed in 80
table 1. Polycarboxylate-based high range water reducers (HRWR) produced by BASF 81
according to ASTM C494-05 Type F was used. It has a specific gravity, pH, chloride content, 82
charge and solid content of 1.05, 8, <0.10%, anionic and 28-32%, respectively. The PDMS 83
used has a density, appearance, viscosity, and volatile content of 0.97 g/cm3, liquid/paste
84
aspect, 0.5 Pa.s and <1.0%, respectively. PDMS is a polymeric organic silicon composed of n 85
repeating monomer [SiO(CH3)2] units, with trimethylsilane groups as terminal units, as shown
86
in Fig. 1. Previous studies (Lecomte et al., 2007; Lucquiaud et al., 2014; Roos et al., 2008) 87
demonstrate that due to its high flexibility, low surface tension, resistance to ageing agents, 88
high hydrophobicity, chemical inertness and good gas permeability, PDMS became a material 89
of great benefit in the field of hydrophobic treatment of construction materials, such as 90
mortars, concrete, tiles, etc. In addition to the above characteristics, the water solubility and 91
reduced cost of PDMS, compared to that for the equivalent materials, make it appropriate for 92
the coating treatment of silica fume. 93
Grouts composition and silica fume pre-treatment 94
A comparison between two types of silica fume was made, namely ordinary silica fume and 95
the hydrophobic one. In order to carry out the hydrophobic treatment the following procedure 96
was adopted: first the PDMS was diluted in deionised water in a ratio of 1:7 (by weight) using 97
a high shear mixer. Then, the silica fume was immersed in the PDMS solution and stirred for 98
10 min, followed by a resting period of 30 min. After this, the silica fume was filtered and 99
placed in an oven at 60 ºC until constant mass was achieved. The temperature of 60 ºC was 100
adopted because it is the optimal temperature to evaporate the water from the impregnated 101
silica fume without damaging the polymer (Lötters, et al., 1997). Finally the dry silica fume 102
was ground and sieved in order to remove any lumps and to get a homogeneous powder. 103
Different dosages were tested for both types of silica fume (0%, 10%, 20%, and 30% as 104
replacement of NHL in weight percentage). All grouts were made with HRWR to maintain 105
constant water content, since this is the parameter that potentially provides more changes on 106
fresh and hardened properties of grouts. The HRWR was incorporated in weight percentage of 107
total NHL binder. The compositions of the NHL-based grouts are shown in table 2. The 108
HRWR dosage was kept constant at the value of 0.8% when the hydrophobic silica fume is 109
used as result of its reduced water demand (this will be further discussed in the next section). 110
Grouts were prepared at room temperature of 20 ± 2 ºC and a relative humidity of 60 ± 2 %. 111
For the preparation of grouts ordinary tap water was used. The silica fume was added and 112
mixed with the dry NHL separately to ensure a homogeneous distribution before being 113
transferred to the mechanical mixer. The mixture procedure adopted was the following: the 114
whole powder (NHL + silica fume) is added to 70 % of the total mix water and mixed for 10 115
min. The remaining water (with diluted HRWR) is added within 30 s (without stopping the 116
mixer). After all materials had been added, the mixture was maintained for an additional 3 117
min. Other studies (Aiad, 2003) have shown that the delay of 10 min on the HRWR addition 118
improves the effectiveness of its dispersing action. 119
Preliminary tests 120
A set of preliminary flow tests was made to assess the appropriate dosage of HRWR in order 121
to get grouts with satisfactory injectability. To do so, some grout’s injectability criteria 122
already established in the literature was taken into account (Miltiadou-Fezans and Tassios, 123
2012). The fluidity factor (FF) proposed by Miltiadou and Tassios (2012) and a targeted 124
porous medium with a representative diameter of orifices of 0.108 mm to be injected were 125
considered (Miltiadou-Fezans, 1990). Flow tests on Marsh cone with 6 mm nozze-diameter 126
was performed to determine the FF and, consequently, the HRWR dosage that leads to a FF 127
higher than the one associated to the nominal orifice width (Wnom) of 0.108 mm. Higher FF
128
values are associated with greater injectability. In the scope of these preliminary tests (table 2) 129
a HRWR dosage of 0.8% showed a satisfactory fluidity for the reference grout and those with 130
hydrophobic silica fume to be injected in the reference porous media. Regarding the grouts 131
containing ordinary silica fume it was necessary to increase the HRWR up to a content of 132 1.5% to obtain a proper FF. 133 EXPERIMENTAL PROCEDURES 134 Rheological measurements 135
Rheological properties were evaluated with a Bohlin Gemini HRnano rotational rheometer,
136
equipped with parallel-plate geometry. The diameter of the geometry was 40 mm and the gap 137
was set to 2 mm. To avoid the undesirable influence from mechanical histories, fresh samples 138
were subjected to a pre-shearing at an identical shear rate of 1 s-1 for 1 min and left standing
139
for an additional 1 min before measurements took place. Then, the sample was subjected to a 140
stepped ramp with shear rates from 1 to 100 s-1. Each shear rate was applied long enough in
141
order to ensure the attendance of the steady state. All grout samples were analysed with a 142
constant temperature of 20 ºC, maintained by means of a temperature unit control. According 143
to Baltazar et al. (2015) the NHL grout’s rheological behaviour can be modelled fairly well 144
using the Bingham model Eq. (1). 145
= + × ̇ (1)
146
where: is the shear stress (Pa), is the yield stress (Pa), is the plastic viscosity (Pa.s) and 147
γ̇ is the shear rate (s-1). In order to better characterise the shear-thinning behaviour of NHL
148
grouts, the power-law model Eq. (2) was fitted to the experimental data in order to get the 149
flow index. 150
= × ̇ (2)
151
where: is the consistency coefficient (Pa.sn), γ̇ is the shear rate (s-1) and n is the flow index
152
which characterises shear-thinning behaviour of grouts. 153
Bleeding 154
The bleeding test was based on ASTM C 940. After mixing, a 1000 ml glass graduated 155
cylinder was filled with 800 ml of grout and covered to prevent water evaporation. For each 156
grout, the thickness of the bleeding water was measured after complete sedimentation. The 157
final bleeding was calculated using the Eq. (3). 158
Final bleeding (%) = × 100 (3)
159
where Vwis the volume of bleed water (ml) measured 120 min after the end of the mixing and
160
Viis the volume of grout sample at beginning of test (ml).
161
Flexural and compressive strength 162
Flexural and compressive strength were determined with five samples of each grout 163
composition. The strength tests were done following standard EN 1015-11:1999. Flexural and 164
compressive strengths were obtained using a universal traction machine Zwick Z050 with a 2 165
kN load cell and a deformation rate of 0.2 mm/min and a load cell of 50 kN and deformation 166
rate of 0.7 mm/min for flexural and compressive test, respectively. 167
Water absorption by capillarity 168
The test of water absorption by capillarity was performed based on the European standard EN 169
1015-18:2002. For each grout, three samples (40 x 40 x 160 mm) previously cured for 28 170
days were used. The samples were vertically placed in a watertight box with water depth of 2 171
mm that was kept constant and the box was covered to maintain constant hygrothermal 172
conditions. The grout specimens were weighted after 5, 10, 15, 30 min and at each hour 173
during the first 6h of testing, and then weighted every 24 h. The capillary water absorption 174
coefficient was determined by the angular coefficient of the linear part of the capillary 175
absorption curve (i.e. absorbed water mass per unit of the surface vs. time). 176
Open porosity, density and pore size distribution 177
Open porosity was measured by total saturation with water of five samples of each grout 178
under vacuum and hydrostatic weighting based on the European Standard EN 1936:2006. 179
Samples were dried and placed under vacuum for 24 h, then maintained under vacuum but 180
immersed in water for another 24h and finally left for 24h immersed at ambient pressure. 181
After this procedure the samples were hydrostatically and saturated weighted. In order to 182
better understand the previous results, the fresh density was measured after mixing by 183
following the ASTM C138. Pore size distribution was conducted on grout samples using a 184
Micromeritics’ AutoPore IV 9500 Series mercury intrusion porosimeter (MIP). The pressure 185
required to intrude mercury into the sample’s pores is inversely proportional to the size of the 186
pores; in this study a measuring pressure ranging from 0.003 to 207 MPa was adopted. 187
Contact angle measurements 188
The grout wettability was characterised with measurements of the grouts contact angles 189
determined using a sessile drop method described by Leelamanie and Karube (2012), using a 190
Goniometer KSV instrument. Contact angle measurements were performed as follows: a 191
droplet of water was carefully dispensed onto the grout samples and an image of the droplet 192
making contact with the grout surface was captured by a video camera. The image was 193
subsequently processed with software to determine the contact angle. 194
Thermogravimetric analysis 195
The grout samples to be analysed by thermogravimetry (TG) were dried and ground. 1g 196
sample was used in all tests performed. TG analyses were performed using a NETZSCH 449 197
F3 Jupiter thermogravimetric analyser. The experimental conditions were: N2 gas dynamic
198
atmosphere (40 ml/min); heating rate (40 ºC/min) and an alumina top-opened crucible. The 199
samples were heated from room temperature to 1000 ºC at a constant rate. The TG test was 200
conducted at the maturity age of 28 days. 201
RESULTS AND DISCUSSION 202
Yield stress and plastic viscosity 203
Rheological properties of injection grouts are decisive parameters since they affect the fresh 204
performance and therefore the success of the consolidation operation. Thus, the yield stress, 205
plastic viscosity and flow index were determined to compare grouts to each other. The 206
knowledge of the yield stress enables to understand if a fluid will flow or not, since it 207
represents that threshold. This important property affects the flow behaviour of the grout and 208
its capacity to flow inside the masonry inner core. Plastic viscosity is associated with the flow 209
resistance once flow is initiated. Fig. 2 presents the evolution of the yield stress and plastic 210
viscosity, respectively, according to the silica fume used and its dosage. A smaller yield stress 211
and plastic viscosity is better than larger ones. From the rheological parameter results it is 212
clear that as the ordinary silica fume dosage increases the grout becomes less workable. Even 213
for the lowest dosage of silica fume (10 wt%) it can be observed that yield stress and plastic 214
viscosity increased regarding the reference grout (without silica fume). The addition of silica 215
fume leads to an increase of specific surface, causing higher adsorption of mixing water 216
resulting in a decrease of grout flowability and therefore a lower injectability (Toumbakari, 217
2002; Assaad and Daou 2014; Khayat et al., 2008). This behaviour is dependent on the solid 218
volume fraction, meaning that when the silica fume dosage exceeds a threshold value, and 219
according with the Krieger-Dougherty equation, an increase in the inter-particle friction 220
occurs (Struble and Sun, 1995; Phan et al., 2006). In other words, the introduction of small-221
sized silica fume particles is the source of additional surface area resulting in an increase of 222
contact forces among fine particles leading to an easier coagulation due to interparticle 223
interactions (Van-der-Waal’s interactions). The penetration capacity of such grouts is 224
significantly decreased, making its injection at low pressure rather difficult. Thus, for a 225
required penetrability performance, the water content of the grout must increase with 226
detrimental effects on the fresh grout stability and hardened properties of grout. 227
As mentioned, a good flowability (i.e. grouts with lower plastic viscosity and yield stress) is 228
preferred for injection purposes. The results illustrate that the rheological properties of grouts 229
containing silica fume with hydrophobic treatment are significantly improved when compared 230
to the ones of grouts containing ordinary silica fume. It is worth mentioning that grouts 231
containing hydrophobic silica fume were found to have lower plastic viscosity values than the 232
grout without silica fume. This is due to the fact that hydrophobic silica fume acts as fine 233
aggregates (with reduced absorption of water) whose small-spherical shape particles fill the 234
spaces made by the large and long shape particles of hydraulic lime binder imposing a ball 235
bearing effect, which reduce the friction forces between lime particles and therefore an 236
improvement of grout flowability. 237
Flow index 238
The flow index (n) is dimensionless and reflects the kind of flow behaviour (Cao et al., 2016). 239
If n > 1, the grout is shear thickening; while n < 1 the grout is shear thinning. For the case of a 240
Newtonian grout (n = 1). Fig. 3 shows the evolution of the flow index calculated from the 241
power-law model for different silica fume type and dosages. Most grout compositions tested 242
presented a shear-thinning behaviour, which means that n is lower than 1. A shear-thinning 243
behaviour means that if flow velocity decreases during grout injection it leads to a viscosity 244
increase (which is not desirable); thus n value should be closer to 1 in terms of injection 245
grouts. 246
For the grouts tested, when the dosage of silica fume increases, the n index decreases. Indeed, 247
the higher adsorption of water by silica fume explains this decrease. However, for grouts with 248
hydrophobic silica fume, fluidity index are always higher than the grouts with ordinary silica 249
fume. In fact, for grout containing 10 wt% of hydrophobic silica fume, flow index was found 250
to be very close to 1 which means a quasi Newtonian behaviour or in other words a grout 251
whose viscosity is not dependent on the injection pressure. Baronio et al. (1992) carried out 252
an experimental program with masonries of different dimensions with cracks and voids 253
irregularly distributed that showed that it was difficult to conduct the injections with a 254
constant pressure. A grout exhibiting quasi-Newtonian behaviour enables an easier flow 255
inside the porous medium, even when velocity decreases during injection, since viscosity will 256
not change. Thus, on the basis of these results, it can be concluded that the presence of 257
hydrophobic silica fume is useful to improve the rheological performance of injection grouts 258
as result of the combined effect of its spherical shape and water repellence. 259
Bleeding 260
The results obtained for the bleeding evaluation are presented in Fig. 4, where it can be seen 261
that all grouts are stable. According to Toumbakari (2002) and Lambardi (1985) a grout is 262
considered stable when the final bleeding is less than 5 %. Nevertheless, the results show 263
differences between grouts with ordinary and hydrophobic silica fume. Thanks to the 264
hydrophobic behaviour of the later the bleeding increases as consequence of higher amount of 265
free water available in the mixture. The best results were obtained by the grout with ordinary 266
silica fume which presented no bleeding. These results can be explained by a higher specific 267
surface and a higher interstitial water retention that is offered by the ordinary silica fume 268
compared to the hydrophobic one. 269
Mechanical strength 270
Compressive and flexural strength are good indicators of the quality of grout in hardened 271
state. Fig. 5 and 6 present flexural and compressive strength values respectively. From the 272
flexural strength results, it is clear that it increases when NHL is replaced by silica fume. A 273
significant increase of flexural strength is observed when 10 wt% silica fume is added to the 274
grout. However, the increase of 10 to 30 wt% in silica fume dosage does not present further 275
improvement. In regard to the grouts with hydrophobic silica fume, it is clear that the flexural 276
strength is always lower than the ones with ordinary silica fume. 277
Concerning compressive strength, it increases with increasing of ordinary silica fume dosage; 278
the explanation for this behaviour is based on the high amorphous silicon dioxide contend of 279
silica fume that reacts with the Ca(OH)2 and results in the formation of additional calcium
280
silicate hydrate structures (Shihada and Arafa, 2010; Khana and Siddique, 2011). From Fig. 6 281
it can be seen that the compressive strength values decrease with increasing dosage of 282
hydrophobic silica fume, which is probably caused by the fact that hydrophobic silica fume 283
has little or no pozzolanic reactivity. In addition, grouts in which NHL is replaced by 284
hydrophobic silica fume will not have enough silicate content (i.e. less NHL) to hydrate and 285
to reach the compressive strength value of the reference grout. Another explanation for the 286
mechanical strength loss is based on the increase of grout’s open porosity in the presence of 287
hydrophobic silica fume (see further section). Note, however, that the compressive strength is 288
less determinant that flexural strength to get a successful consolidation process since the latter 289
are responsible to promote the adherence between grout and masonry elements (Jorne et al., 290
2015). In addition, the typical compressive stress in old masonry is in range of 1 to 2.5 MPa 291
(Corradi et al. 2002; 2008) and all the analysed grout’s compositions have higher compressive 292
strength. Moreover, according to Valluzi et al. (2004) who assessed masonry walls through 293
compressive testing before and after grout injections, and despite the difference in the grout 294
compressive strengths, the strength gains in the walls were similar. 295
Open porosity 296
In order to better understand the mechanical strength results, the grout’s open porosity was 297
determined by the hydrostatic method. It should be highlighted, however, that the open 298
porosity of grouts with hydrophobic silica fume was also determined using the MIP method, 299
once in the hydrostatical method, the water repellence of grouts with hydrophobic silica fume 300
might block the entrance of water into some pores resulting in lower open porosity. As it can 301
be seen from Fig. 7 the values of open porosity obtained using the MIP method are higher 302
since the pressure used with mercury is able to overcome the surface energy of the 303
hydrophobic pore surfaces. Fig. 7 shows that there is a decrease of the grout porosity for 304
increasing dosage of ordinary silica fume. As expected the presence of silica fume decreases 305
the porosity, increasing the packing density, which explains the compressive strength 306
evolution. The same trend is not detected in hydrophobic silica fume, since when the 307
hydrophobic silica fume content increases the open porosity also increases. The authors 308
believe that this behaviour is due to combined effect of higher entrained air (as shown by the 309
fresh density values in Fig. 7) in the fresh grout, promoted by the hydrophobic silica fume, 310
and the lower amount of hydration products as result of the lower pozzolanic effect of this 311
kind of silica fume. 312
Thermogravimetry 313
The thermographs presented in Fig. 8 indicate that the peak of the decarbonation of CaCO3
314
(peak around 900 ºC) decreases with increasing replacement of lime by silica fume (this 315
occurs for both silica fume types). As expected the amount of CaCO3 decreases with
316
decreasing amount of lime available to hydrate. However, differences between both silica 317
fume types can be detected on the peak around 500 ºC (i.e. dehydroxylation of calcium 318
hydroxide). The derivative thermogravimetry (dTG) curves of the ordinary silica fume show 319
that this silica fume reduces the Ca(OH)2 content as illustrated by the lower Ca(OH)2 peaks
320
for higher silica fume dosage; this is attributed to the reaction of silica fume with Ca(OH)2
321
leading to the formation of C-S-H. In contrast, in grouts with hydrophobic silica fume, the 322
content of Ca(OH)2 is only slightly affected, which corroborates the results of the mechanical
323
strength in the previous section. 324
Water absorption by capillarity 325
The results of water absorption coefficient are presented in Fig. 9. According to this figure, 326
there is a decrease of water capillary coefficient with increasing of the ordinary silica fume 327
dosage. This comes from the fact that ordinary silica fume contributes to pore-size refinement 328
and matrix densification and, consequently, the capillary pores are reduced. A major decrease 329
until 10 wt% of silica fume can be seen. Above 10 wt% of ordinary silica fume there are no 330
significant differences in water capillary coefficient; notwithstanding a slight tendency to 331
decrease can be observed. These results indicate that there can be no advantage in using 332
higher dosages than 10 wt%. 333
Regarding the grouts containing hydrophobic silica fume, the water capillary coefficient 334
values are lower than the ones presented when the ordinary silica fume is used. The reason 335
behind this is that hydrophobic silica fume has turned the hardened grout also hydrophobic. 336
Those results can be explained by evaluating the wetting behaviour of hardened grouts 337
through contact angle measurements. As the grout samples have a flat and rigid surface the 338
water drops were placed directly over the surface of the sample. The results obtained for the 339
contact angle measurements are presented in table 3. These results are averages of four 340
measurements and the standard deviation is given between parentheses. It can be noted that 341
the contact angle increases in the presence of hydrophobic silica fume. In fact, the grout with 342
hydrophobic silica fume can also be considered as having hydrophobic properties (quasi non-343
wettable) since the contact angle is higher than 90º (see Fig. 10). This is in fact a promising 344
behaviour from a durability point of view, since typically water transport soluble salts in old 345
masonries which will induce destructive crystallization/dissolution cycles and in this case the 346
hydrophobic performance of the grouts prevents or minimises that risk. Furthermore it should 347
be noted that this hydrophobic behaviour does not compromise water vapour permeability. 348
Pore size distribution 349
The pore structure is a major component of the microstructure that affect mechanical strength 350
and capillarity of hardened cementitious materials (Perraton et al., 1988). Thus, the 351
microstructure of the grouts was examined using MIP to assess the influence of type and 352
dosage of silica fume on the pore size distribution, at the age of 28 days. The MIP results 353
(Fig. 11) show that there are differences among the grouts related to pore size distribution; 354
however the majority of pore volume in all compositions is in the range of capillary pores (i.e. 355
mesopores). As expected, the use of ordinary silica fume led to a pore diameter reduction. In 356
Fig. 11a the comparison of the reference grout (0% ref.) with the one with silica fume dosage 357
of 30wt% shows that there is a decrease of capillary pore size from 1 μm to 0.1 μm. This is 358
due to the fact that ordinary silica fume contributes to the formation of additional pozzolanic 359
products that fill the pore spaces and therefore refines the pore structure. These results are 360
consistent with the water capillarity shown in Fig. 9. In the case of hydrophobic silica fume 361
(see Fig. 11b) it also shifts the pore size distribution to smaller diameter; although this 362
reduction is much less pronounced than in the case of ordinary silica fume. It can be therefore 363
observed that the grout with 30wt% of hydrophobic silica fume presented some pores greater 364
than 1 μm (macropores range). The authors believe that these macropores might be assigned 365
to shrinkage cracks due to the low strength of this grout composition. 366
CONCLUSIONS 367
This paper focuses on the influence of hydrophobic silica fume and its dosage on the 368
properties of NHL-based grouts, namely rheological and mechanical properties, pozzolanic 369
reactivity, water transport and microstructure. From the results obtained, the following 370
conclusions can be deduced: 371
x The rheological properties of grouts containing hydrophobic silica fume are 372
significantly improved. The grout containing 10wt% of hydrophobic silica fume 373
showed the best rheological performance (lower plastic viscosity, lower yield stress 374
and flow index close to 1). 375
x The hydrophobic silica fume acts as fine aggregates, filling the spaces between the 376
long shape particles of hydraulic lime, and imposing a pure ball bearing effect, which 377
improves the grout flowability by reducing the friction forces between lime particles. 378
x Silica fume has a strong impact on the shear-thinning behaviour of NHL grouts: the 379
presence of hydrophobic silica fume leads to a more Newtonian behaviour. This can 380
be seen as an advantage for grout’s flow and injectability performance. 381
x The bleeding phenomenon slightly increases when hydrophobic silica fume is used, 382
however, all grout compositions tested had an acceptable stability i.e. bleeding less 383
than 5%. 384
x Mechanical strength is somewhat affected by the hydrophobic silica fume. 385
Notwithstanding, the compressive strength values obtained are still suitable, with 386
compressive stresses in the range of the ones (1-2.5 MPa) typically found in old stone 387
masonry. 388
x The thermogravimetric analyses indicates that hydrophobic silica fume does not fully 389
react with Ca(OH)2 in the presence of water to produce additional pozzolanic
390
products. 391
x The water capillary coefficient is remarkably reduced when hydrophobic silica fume is 392
used. The main reason behind this is that hydrophobic silica fume has turned the 393
hardened grout also hydrophobic, as showed by the contact angle measurements. 394
x Both types of silica fume tested lead to a reduction of pores size when dosage is 395
increased (with the exception of hydrophobic silica fume at 30wt%). Nevertheless, it 396
was observed that there is a decrease of the grout porosity with increasing dosage of 397
ordinary silica fume and an opposite trend for hydrophobic silica fume as a result of 398
entrained air during mixing. 399
This study suggests that the hydrophobic silica fume can be used to improve the performance 400
of NHL grouts for consolidation of old stone masonries. It is clear, however, that additional 401
characterisation of the contribution of this kind of silica fume in grout injection processes are 402
needed. Anyway, the results of this study open the way to the manufacture of other materials 403
like mortars, using the promising effectiveness of hydrophobic silica fume on improving 404
rheological behaviour and durability by granting remarkable water transport properties. 405
Acknowledgments 406
This work received support from the Fundação para a Ciência e Tecnologia – Ministério da Ciência, 407
Tecnologia e Ensino Superior (FCT/MCTES) through the scholarship SFRH/BPD/108427/2015. The 408
authors would like to acknowledge the support from the strategic project UID/CTM/50025/2013. 409
References 410
Aiad, I. (2003). “Influence of time addition of superplasticizers on the rheological properties of fresh 411
cement pastes”. Cem Concr Res 33: 1229–34. 412
Artelt, C. and Garcia, E. (2008). “Impact of Superplasticizer Concentration and of Ultra-Fine Particles 413
on the Rheological Behaviour of Dense Mortar Suspensions.” Cem Concr Res 38(5):633–42. 414
Assaad, J. J. and Daou Y. (2014). “Cementitious grouts with adapted rheological properties for 415
injection by vacuum techniques”. Cem and Conc Res 59: 43-54. 416
ASTM C138. (2008). “Standard Test Method for Density (Unit Weight), Yield, and Air Content 417
(Gravimetric) of Concrete”. ASTM International, West Conshohocken, PA, USA.
418
ASTM C494. (2005). “Standard Specification for Chemical Admixtures for Concrete”. ASTM 419
International, West Conshohocken, PA, USA. 420
ASTM C940. (2010). “Standard Test Method for Expansion and Bleeding of Freshly Mixed Grouts for 421
Preplaced-Aggregate Concrete in the Laboratory”, ASTM International, West Conshohocken,
422
PA, USA. 423
Ballantyne, A. (1996). “Stabilising the substrate: Grouting – the options. Anne Ballantyne’s 424
contribution on 28thOctober 1995”. Conservation News 60: 43–5.
425
Baltazar, L. G., Henriques, F. M. A., Jorne, F. and Cidade, M. T. (2013). “Performance improvement 426
of hydraulic lime based grouts for masonry consolidation – An experimental study”. WIT 427
Transactions on the Build Environment 131: 417-30.
428
Baltazar, L. B., Henriques, F. M. A. and Cidade, M. T. (2015). “Experimental study and modeling of 429
rheological and mechanical properties of NHL grouts”. J Mater Civ Eng : 430
10.1061/(ASCE)MT.1943-5533.0001320, 04015055. 431
Baronio, G., Binda, L. and Modena, C. (1992). “Criteria and methods for the optimal choice of grouts 432
according to the characteristics of masonries”. In: International workshop CNR- GNDT, 433
effectiveness of injection techniques for retrofitting of stone and brick masonry walls in
434
seismic areas, Milan, Italy, 30–31 March 1992, pp 139-57.
435
Biçer-Şimşir, B., Griffin, I., Palazzo-Bertholon, B. and Rainer, L. (2009). “Lime-based injection 436
grouts for the conservation of architectural surfaces”. Stud. Conserv. 54: 3-17. 437
Binda, L., Saisi, A. and Tedeschi, C. (2006). “Compatibility of materials used for repair of masonry 438
buildings: research and applications”. Fracture and Fail of Nat Build Stone 2006: 167–82. 439
Binda, L., Modena, C., Baronio, G. and Abbaneo, S. (1997). “Repair and Investigation Techniques for 440
Stone Masonry Walls.” Constr Build Mater 11(3):133–42. 441
Cao, Y., Zavattieri, P., Youngblood, J., Moon, R. and Weiss J. (2016). “The relationship between 442
cellulose nanocrystal dispersion and strength”. Constr Build Mater 119: 71-9. 443
Chaudhry, C. (2007). “Evaluation of grouting as a strengthening technique for earthen structures in 444
seismic areas: case study chiripa Theses (Historic Preservation)”, University of Pennsylvania, 445
USA. 446
Corradi, M., Borri, A. and Vignoli, A. (2002). “Strengthening techniques tested on masonry structures 447
struck by the Umbrian–Marche earthquake of 1997–1998”. Constr Build Mater 16(4):229–39. 448
Corradi, M., Tedeschi, C., Binda, L. and Borri, A. (2008). “Experimental evaluation of shear and 449
compression strength of masonry wall before and after reinforcement: Deep repointing”. 450
Constr Build Mater 22:463–72.
451
EN 1015-11. (1999). “Methods of Test for Mortars for Masonry - Part 11: Determination of flexural 452
and compressive strength of hardened mortar”. CEN, European Committee for
453
Standardization. 454
EN 1015-18. (2002). “Methods of Test for Mortar for Masonry - Part 18: Determination of Water 455
Absorption Coefficient due to Capillary Action of hardened Rendering Mortar”. CEN,
456
European Committee for Standardization. 457
EN 1936. (2006). “Natural stone test methods – determination of real density and apparent density, 458
and of total and open porosity”. CEN, European Committee for Standardization.
459
EN 459-1. (2010). “Building lime - Part 1: Definitions, specifications and conformity criteria”. CEN, 460
European Committee for Standardization. 461
Jorne, F., Henriques, F. M. A. and Baltazar, L.G. (2015). “Evaluation of consolidation of different 462
porous media with hydraulic lime grout injection”. J Cult Herit 16(4): 438-51. 463
Khana, M. and Siddique, R. (2011). “Utilization of silica fume in concrete: Review of durability 464
properties”. Resour Conserv Recy 57: 30–5. 465
Khayat, K. H., Yahia, A. and Sayed, M. (2008) “Effect of supplementary cementitious materials on 466
rheological properties, bleeding, and strength of structural grout”. Aci Materials Journal 467
105(6): 585-93. 468
Lambardi, G. (1985) “The role of cohesion in cement grouting of rock”. In the Proceedings of 15th 469
International Congress on Large Dams, Lausanne 1985, 235-60. 470
Lecomte, J. P., Thibaut, M. and Stammer, A. (2007) “Encapsulated silicone/silane mix enhances water 471
resistance of mortars”. European Coating Journal 12: 46-51. 472
Lecomte, J. P., Llado, D., Salavati, D., Rodrigues, G., Ribeiro, M. (2013) “Mortar protection. New 473
silicon resin-based hydrophobic powder for the dry mix market, European Coating Journal, 474
12: 88-91. 475
Leelamanie, D. A. L. and Karube, J. (2012). “Drop size dependence of soil-water contact angle in 476
relation to the droplet geometry and line tension”. Soil Science and Plant Nutrition, 58: 675-477
83. 478
Lötters, J.C., Olthuis, W., Veltink P. H. and Bergveld, P. (1997) “The mechanical properties of the 479
rubber elastic polymer polydimethylsiloxane for sensor applications”. J Micromech Microeng 480
7: 145–7. 481
Lucquiaud, V., Courard, L., Gérard, O., Michel, F., Handy, M., Aggoun, S. and Cousture A. (2014) 482
“Evaluation of the durability of hydrophobic treatments on concrete architectural heritage”. 483
Proceedings of Hydrophobe VII, 7th International Conference on water repellent Treatment of
484
Bulding Materials, Aedificatio Publishers: Freiburg, Germany: 29-38. 485
Miltiadou-Fezans, A. E. (1990). “Contribution à l’ étude des coulis hydrauliques pour la réparation et 486
le renforcement des structures et des monuments historiques en maçonnerie”. Doctoral de l’ 487
École National de Ponts et Chaussées, (PhD thesis in French). 488
Miltiadou-Fezans, A. and Tassios, T. P. (2012). “Fluidity of hydraulic grouts for masonry 489
strengthening”. Mater Struct 45: 1817–28. 490
Park, C. K., Noh, M. H. and Park, T. H. (2005). “Rheological properties of cementitious materials 491
containing mineral admixtures”. Cem Concr Res 3: 842–49. 492
Perraton, D., Aitcin, P. C. and Vezina, D. (1988). “Permeabilities of silica fume concrete. In: Whiting 493
D., Walitt A., editors. Permeability of concrete”. Am Concr Inst SP-108, pp 63-4. 494
Phan, T. H., Chaouche, M. and Moranville, M. (2006). “Influence of organic admixtures on the 495
rheological behaviour of cement pastes”. Cem Concr Res 36: 1807–13. 496
Roos, M., König, F., Stadtmüller, S. and Weyershausen, B. (2008) “Evolution of silicone based water 497
repellents for modern building protection”. Proceedings of Hydrophobe V, 5th International
498
Conference on water repellent Treatment of Building Materials, Aedificatio Publishers: 499
Freiburg, Germany: 3-16. 500
Seethapathy, S. and Górecki, T. (2012) “Applications of polydimethylsiloxane in analytical chemistry: 501
A review”. Analytica Chimica Acta 750: 48– 62. 502
Shannag, M. J. (2000). “High strength concrete containing natural pozzolan and silica fume”. Cem 503
Concr Comp 22: 399–406.
504
Shihada, S. and Arafa, M. (2010). “Effects of silica fume, ultrafine and mixing sequences on 505
properties of ultra high performance concrete”. Asian J Mater Sci 2: 137–46. 506
Sobolev, L., Lin, Z., Cao, Y., Sun, H., Flores-Vivian, I., Rushing, T., Cummins, T. and Jason Weiss, 507
W. (2016) “The influence of mechanical activation by vibro-milling on the early-age 508
hydration and strength development of cement”, Cem Concr Com 71: 52–62. 509
Struble, L. and Sun, G-K. (1995). “Viscosity of Portland cement paste as a function of concentration”. 510
Advn Cem Based Mater 2: 62–9.
511
Toumbakari, E-E. (2002). “Lime-pozzolan-cement grouts and their structural effects on composite 512
masonry walls”. PhD Thesis, Katholieke Universiteit Leuven, Belgium. 513
Valluzzi, M.R, Porto, F. and Modena, C. 2004. “Behavior and Modeling of Strengthened Three-Leaf 514
Stone Masonry Walls.” Mater Struct 37:184–92. 515
Wei-Hsing, H. (1997). “Properties of cement-fly ash grout admixed with bentonite, silica fume, or 516
organic fiber”. Cem Concr Res 27: 395–406. 517
Table 1. Chemical composition of hydraulic lime and silica fume Formula Hydraulic lime (%) Silica fume (%)
Al2O3 2.00 0.15 CaO 85.00 0.20 Fe2O3 2.00 0.03 Na2O - 0.05 MgO 1.00 0.30 MnO 0.03 -SiO2 8.00 97.00 SiC - 0.50 SO3 1.00 -SrO 0.05 -K2O 0.70 0.80 Free lime (%) > 15
Table 2. Compositions of tested NHL grouts and its fluidity factor Grout No. Amount of lime (%) Ordinary Silica fume (wt%) Hydrophobic Silica fume (wt%) w/b HRWR (wt%) FFmin= 1.2x103 (mm/s) for a Wnom= 0.108mm FF x103(mm/s) I 100 - - 0.5 0.8 2.2 II 90 10 - 0.5 1.0 1.6 III 80 20 - 0.5 1.3 1.5 IV 70 30 - 0.5 1.5 1.3 V 90 - 10 0.5 0.8 2.8 VI 80 - 20 0.5 0.8 2.7 VII 70 - 30 0.5 0.8 2.6
Table 3. The contact angle between the hardened grout samples and water Grout designation ߐSDM
Grout without silica fume 28.5º (2.1) Grout with ordinary silica fume = 20% 41.3º (2.6) Grout with hydrophobic silica fume = 20% 108º (1.4)
3ODVWLFYLVFRVLW\3DV <LHOGVWUHVV3D 6LOLFDIXPHGRVDJHZW <LHOGVWUHVV2UGLQDU\6) <LHOGVWUHVVK\GURSKRELF6) 3ODVWLF9LVFR2UGLQDU\6) 3ODVWLF9LVFR+\GURSKRELF6)
)ORZLQGH[ 6LOLFDIXPHGRVDJHZW 2UGLQDU\6) +\GURSKRELF6)
%OHHGLQJ 6LOLFDIXPHGRVDJHZW 2UGLQDU\6) +\GURSKRELF6) 6WDEOHLIWKHEOHHGLVOHVVWKDQ
)OH[XUDO VWUHQJWK 03D 6LOLFDIXPHGRVDJH 2UGLQDU\6LOLFDIXPH +\GURSKRELF6LOLFDIXPH
&RP SUHVVLYH VWUHQJWK 03D 6LOLFDIXPHGRVDJH 2UGLQDU\VLOLFDIXPH +\GURSKRELFVLOLFDIXPH
)UHVKGHQVLW\ JFP 2SHQ SRURVLW\ 6LOLFDIXPHGRVDJHZW )UHVKGHQVLW\ 2UGLQDU\6) )UHVKGHQVLW\ +\GURSKRELF6) 2SHQSRURVLW\ 2UGLQDU\6) 2SHQSRURVLW\ +\GURSKRELF6) 2SHQSRURVLW\ +\GURSKRELF6)0,3
G7* P JP LQ 7HPSHUDWXUH& D UHI Dehydroxylation of Ca(OH)2
G7* P JP LQ 7HPSHUDWXUH& E UHI
&DSLOODU\ZDWHUFRHILFLHQW NJP P LQ 6LOLFDIXPHGRVDJHZW 2UGLQDU\6LOLFD)XPH +\GURSKRELF6LOLFD)XPH
/RJ GLIIHUHQWLDO LQWUXVLRQ FP J 3RUHVL]HGLDPHWHUP D UHI
/RJ GLIIHUHQWLDO LQWUXVLRQ FP J 3RUHVL]HGLDPHWHUP E UHI
Figure 1 The structural units of PDMS
Figure 2. Evolution of yield stress (τ ) and plastic viscosity ( ) according to the dosage and type of silica fume
Figure 3. Evolution of the flow index (n) with the dosage and type of silica fume Figure 4. Final bleeding of each grout composition studied
Figure 5. Evolution of flexural strength of hardened grout with the dosage and type of silica fume studied
Figure 6. Evolution of compressive strength of hardened grout with the dosage and type of silica fume studied
Figure 7. Evolution of open porosity of hardened grout and fresh density with dosage and type of silica fume studied Figure 8. Themogravimetric analyis of grouts: (a) ordinary silica fume; (b) hydrophobic silica fume
Figure 9. Water absorption coefficient due to capillary of hardened grout for the different dosage and the silica fume studied
Figure 10. The contact angle between the hardened grout with hydrophobic silica fume and the water drop
Figure 11. Differential pore volume as function pores size diameter of grouts: (a) ordinary silica fume (b) hydrophobic silica fume