*3.1. Fresh State Properties*

Fresh grout properties for different limestone filler/glass microsphere ratios are presented in Table 4.


**Table 4.** Fresh grout properties.

The wet densities of fresh grouts confirm that the glass microspheres are an efficient constituent used to reduce the grout's density. As expected, the lowest wet density was obtained for the grout GM100 (0.82 g/cm3), where only glass microspheres were used as filler material, and the highest value was obtained for the grout LS100 (1.73 g/cm3), where the limestone filler was the only mineral admixture.

The wet densities of grouts with composed limestone–glass microspheres filler lie between these two limits, their average values being that of 1.51 g/cm3, 1.31 g/cm3, and 1.14 g/cm3 for the LS67-GM33, LS50-GM50, and LS33-GM67 grout, respectively. The replacement of the limestone filler of a relatively high density (2.75 g/cm3) with the same volume of glass microspheres of an extremely low typical density (0.125 g/cm3) is the main parameter that governs the grout's weight reduction. The water content, which decreases with the increase in the glass microspheres' volume (Table 3), is one additional parameter that influences the volume of prepared grout. Another is the packing density of solid particles in the suspension; grain size distribution of the glass microspheres is much coarser compared to the limestone filler. When considering the reduced weight of the grout, the two additional parameters—water content and packing density—need to be considered as well. Based on obtained results, it is possible to conclude that the reduction of the grout's wet density with the incorporation of a relatively high volume of glass microspheres could be an effective method for reducing the weight of the grout, when the re-attachment of large plaster detachments needs to be carried out.

The mini-slump-flow value of the fresh grout evaluates its flowability under the action of self-weight. It is a measure for fresh grout consistency, which is often related to the grout's workability. In this study, the workability was evaluated via injection of the grout through a 10-mL syringe while applying minimum pressure [1], since the test method was set to reflect the conditions on the conservation site. The water content of the grout was reduced with the increasing volume of glass microspheres (Table 3), to obtain the same workability of the grouts. It is highly likely that with this, yield stress and viscosity of the lime paste (lime + water + PCE-SP) were increased as well. Comparing the consistency of the LS67-GM33 grout to that of the LS100 grout, a considerable increase in the slump-flow value could be observed. Slump-flow values of the LS50-GM50 and LS33-GM67 grouts were the same as that of the LS100 grout; these two compositions show the same workability as the LS100 grout and, consequently, also the same slump-flow value. Mini-slump-flow value is often related to the paste's yield stress τ0; paste is a generic name for the mixture of binder, filler particles smaller than 0.1 mm and water, and can also contain a chemical admixture. The value of τ<sup>0</sup> increases with an increase in paste density and decreases with the mini-slump-flow value increase. If we assume that only the paste's own weight is controlling the phenomena, the equation proposed in [33] is the following: <sup>τ</sup><sup>0</sup> = C·ρ/SF5. In this equation, C represents a constant that includes gravity and volume of the paste, ρ is the paste's density, and SF is the mini-slump-flow value. The equation can be used to explain the influence of the glass microspheres on the rheological properties of the grouts. It is clear that by replacing the fine limestone filler with coarser glass microspheres, the grout's yield stress (τ0) decreases, despite the increase in yield stress and viscosity (μ) of the lime paste in the grout. For the constant SF value, the yield stress decrease is higher for grouts with a higher content of glass microspheres. The effect of glass microspheres on the rheological properties of the grout seems to be similar to that of air bubbles that are produced using an air-entraining agent (chemical admixtures) in cement paste. For the LS50-GM50 grout, the obtained standard deviation of the test results was

relatively high. Visual observation of the grout spreads revealed segregation between solid particles during the test; heavier particles settled to the bottom and lighter particles (the glass microspheres) accumulated on the fresh mixture's surface. A poor packing density of solid particles in the LS50-GM50 grout could be responsible for the observed behaviour. The lowest slum-flow value was measured for the GM100 composition. A high water-content reduction of 20% in the GM100 composition was needed to obtain the required workability of the grout; it appears that complete elimination of the limestone filler particles significantly changed the rheology of the lightweight grout. From these results, we can conclude that there is no clear relationship between the workability test and the mini-slump-flow test results.

The grouts with limestone-filler content representing 50% or more of the total filler content (LS100, LS67-GM33, LS50-GM50) showed a higher level of final bleeding, which ranged between 1.5% and 1.7%. In the mixtures where the prevailing part of the filler was composed of glass microspheres (LS33-GM67 and GM100), the final bleeding was between 0.1% and 0.6%. In all tested grouts, the final bleeding was lower than the standard limit value of 2% (EN 447 [34]; Table 1). These final bleeding values alone, however, are not enough when assessing the stability of the lightweight grout; important information can be provided by visual inspection of the sample appearance, as was the case for the GM100 grout (Figure 3), where it's lowest final bleeding of 0.1% was due to the fact that a big part of the bleed water was trapped between two layers of the tested sample. The trapped water was not considered when calculating the bleeding value; such behaviour of the hydrated lime grout was observed for the first time. It appears that local internal segregation of glass microspheres, bleed water, and (possibly) lime particles happened in the test sample. Internal segregation of particles was difficult to prove due to the same white colour of the lime, limestone filler, and glass microspheres.

**Figure 3.** Water trapped between the two layers of the grout sample for the GM100 mixture.

The results of water-retention capacity range between 78% and 84% for all tested grouts (Table 4). Although the reference grout mixture (LS100) seems to possess the highest water-retention capacity, the incorporation of glass microspheres did not significantly reduce this fresh grout property. Due to its high water retention, the grout resists releasing water into the highly porous media with high absorption capacity through which it flows. Consequently, the plugging of the grout inside the plasters can be prevented, and its drying shrinkage can be efficiently reduced [5]. The highest water retention was measured in the LS100 grout, which lacked the glass microspheres; this could be due to a lower content of free water, although this composition was prepared with the highest water content. The limestone filler particles are much finer than glass microspheres, and their shape is the same as the shape of crushed limestone aggregate grains. Thus, a significant reduction of free water content inside the LS100 grout can be attributed to a much higher surface at the same volume of particles (a spherical

shape results in the lowest surface at a particular volume) and a higher ability of the limestone particles to capture water by adsorption and absorption. Another influencing parameter is the ability of the filler to increase the packing density of the grout's solid particles, reducing the free water content. Ince et al. [35] showed that the filler with an appropriate granular composition could optimally fill the voids within the grout matrix. As a result, less free water would be available in the mixture during the suction action provided by porous plaster; the free water could be easily removed from the grout.

The results of the stability test are presented in Figure 4. For the grouts LS100, LS67-GM33, LS33-GM67, and GM100 the differences in the grout's wet density between the bottom and the top level of the testing column are low and equal to 0.01 or 0.02 g/cm3. All measured densities are also in agreement with density values given in Table 4. Therefore, these grouts can be evaluated as stable. When observing the grout LS50-GM50, segregation of particles was noted. The highest density was present in the bottom third of the column (1.42 g/cm3), while the lowest was in the top third of the column (1.35 g/cm3). The interparticle forces in this composition were not strong enough to maintain a homogenous suspension of particles along the column height. Therefore, a higher percentage of the limestone filler particles settled towards the bottom of the column, while a larger amount of the light glass microspheres was able to rise towards the surface. The same behaviour was also observed during the mini-slump-flow test of the LS50-GM50 grout. According to Rickerby et al. [7], the spherical morphology of glass microspheres and their coarser grain-size distribution may worsen the packing density of the composite filler. It seems that this was the case for the LS50-GM50 composition. Injection grouts have to possess sufficient stability/homogeneity after mixing, during the whole injection process, and while setting is taking place. If the mixture segregates during the process of injecting or setting, the consolidation of air pockets cannot be successful. In their study, Miltiadou-Fezans and Tassios [36] concluded that, for each grout, the critical water-to-solids ratio resulting in segregation depends on an acceptable degree of instability, the specific surface of solids, and the percentage of superplasticiser used. Based on the trapped water detected following the bleeding test (Figure 3), high stability of the fresh GM100 grout is an unexpected result. One possible explanation for the fresh properties measured in the GM100 composition is a distributed segregation of solid particles; along the entire column, there can be a local settlement of lime-binder particles, as well as flowing of the glass microspheres towards the internal-bleed water surface.

**Figure 4.** Results of the stability tests: measured grout's wet density from the bottom, middle, and top level of the testing column.

Figure 5 and Table 5 show the results of the drying shrinkage test inside the dry or prewetted mortar cup. From the results, it is clear that the resistance of the grout to drying shrinkage and, thus, to the formation of the separation ring and cracks inside of the grout, highly depends on the filler composition used. The lowest cracking was observed for grouts LS100 and GM100, where a separation ring with a thickness of only 0.5 mm was formed in the dry cups; in the prewetted cups, the 0.5 mm separation ring was only formed in the LS100 composition. On the other hand, compositions with composed limestone particles and glass microspheres filler showed a weaker resistance to drying shrinkage; this was also expressed through the formation of cracks inside of the grout, observed in the LS67-GM33 (dry and prewetted mortar cup) and LS33-GM67 (dry mortar cup) compositions. It can be concluded that the combination of limestone particles with high density and modulus of elasticity and glass microspheres with extremely low density and modulus of elasticity, induces additional differential deformations in the grout that result in reduced resistance to the formation of cracks. On the other hand, a reduction of the water-to-binder ratio through the increase of the glass microspheres content decreases the sensitivity of the grout to shrinkage. These two influencing parameters with opposite effects are responsible for the observed response to drying shrinkage in each particular grout. Additionally, prewetting the mortar cups seems to be more efficient for compositions where the glass microspheres content in the filler is 50% or higher.

**Figure 5.** Grout mixture after drying in the dry and prewetted mortar cups.


**Table 5.** Drying shrinkage in dry and prewetted mortar cups.

The separation ring between the mortar cup and the grout and/or the cracks in the grout might indicate an excessive water content in the mixture, which could weaken the bond between the grout and the plaster layers and reduce the grout strength [17].

The grout mixtures GM100 and LS50-GM50 did not meet the requirements set for fresh grout properties. Due to their resistance to segregation not being high enough, we did not determine the injectability and hardened properties for these two mixtures.

The injectability curves of the grout mixtures LS100, LS67-GM33, and LS33-GM67 are given in Figure 6 for the prewetted and dry crushed lime mortar columns. From these curves, it can be noted that the glass microspheres have an essential influence on the ability of the grout to be injected into detached plaster; the increase in the volume of the microspheres decreases the injectability of the grout. The results obtained are not in line with the results of studies carried out by Zajadacz and Simon [4] and Rickerby et al. [7], where glass microspheres improved the injectability of tested grouts. The authors concluded that the improvement is due to the spherical morphology and small particle size of the glass microspheres. However, in [4] there is no information regarding detailed grout composition and mixing procedure, and the composition of earthen grout used in [7] is not comparable with the hydrated lime grout used in our study.

**Figure 6.** Injectability curves of the grout mixtures LS100, LS67-GM33, and LS33-GM67 for prewetted (left) and dry (right) lime mortar columns.

The results also demonstrated that prewetting of the crushed lime mortar improved the injectability of all three tested grouts. The ability of the grouts to be injected was classified following the proposal of Biçer- ¸Sim¸sir and Rainer [37]. The mixtures LS100, LS67-GM33, and LS33-GM67 were classified as E (easy) when prewetted crushed mortar column was used. Additionally, the LS100 and LS67-GM33 mixtures were classified as E (easy) and F (feasible), respectively, and the LS33-GM67 mixture as D (difficult), when dry mortar column was used.

Lower bleeding and higher water-retention capacity of the LS100 mixture (Table 4) are the main influencing parameters responsible for better injectability of the grout not containing the glass microspheres.
