Influencia del transporte de agua en morteros de revestimiento , en la resistencia a la adherencia Influence of water transportation inside a mortar / block system on bonding resistance behavior

In the present study was performed an evaluation of the water transport in the mortar/block system, with the objective of linking the behavior of the bond strength by perpendicular tensile stress, and with water movement and porosity of mortar. This was done by producing two mortars with aggregates of different size distribution. Properties were determined in fresh state and bond strength was measured when hardened. The water transport profile was constructed by resistive sensors for measuring moisture. Measurements were done in different coats of mortars applied on a concrete substrate and the ceramic substrate. Masonry blocks used as a substrate were characterized according to their physical properties related to water absorption and porosity. The profile leads to the transport of water affects the porosity of mortar and showed that there is a direct relationship between the percentages of water being transported at different litters, with resistance to adhesion. Was checked also the not existence of relationship between the water retention test mortar and total water transported in the mortar-block system.


Introduction
From the very first moment after applying mortar on the masonry unit, an important change takes place in the inner material structure.This is due the transportation of water inside the mortar.Due to capillary absorption, water moves towards the porous structure and towards the exposed surface due to water absorption process.After its application, the substrate is responsible for mortar water loss, because of its water absorption capacity.
As far as the mortar is concerned, water loss at the moment it gets in contact with the masonry block is quite intense.As long as the water transportation process takes place, capillary absorption tensile stress in the contact surface decreases (with the humidity percentage increase on the masonry block surface).Water availability to be transported in the mortar also decreases (de Souza et al., 2012).We shall consider that water flow does not go alone in the movement of pores structure created by absorption.This is because there are capillary strengths, physical absorption phenomena, as well as water setting due to reactions taking place due to binders (concrete and lime-stone) used to elaborate hydration products.Therefore, in the masonry block, water suction takes place when the resulting actions of this factor set are in the mortar block direction.Some of the main equations leading to partially explain such phenomenon are described by Hall (1986Hall ( , 1994)).
Mortar water release at the initial moments (some hours after mixing) and at early ages (first 7 days) is extremely important for the development of the coating system properties.It is also responsible for the origin of pathologies (Pereira and Bauer, 2013), such as cracking due to mortar plastic retraction (Silva and Bauer 2009;Silva et al., 2009).Retraction is caused by water release due to evaporation and due to masonry unit capillary absorption.Retraction may lead to mortar cracking.Veiga (1998) and then Silva (2009) evaluated the effect produced by retraction and they linked mortars mechanical behavior with cracking development.Similarly, Pereira (2007) carried out a detailed research on water release due to evaporation, linking this phenomenon to a de-bonding mechanism.
Other experimental techniques used to determine water transportation in fresh concrete have been employed intending to explain this phenomenon, such as neutron transmission (Groot, 1993), nuclear magnetic resonance imaging (Brocken et al., 1998), permeameter test, absorption test and pressure plate tests (Hendrickx et al., 2010).
As far as masonry mortars are concerned, Carasek (1996) demonstrated the importance of water transportation for the substrate, evidencing the creation of hydration products (mainly ettringite) on ceramic blocks pores, together with the importance of mortars pores and block pores for the development of bonding.Recent researches, developing a micro-structural analysis of the mortar/block interface, demonstrate that the size, shape and orientation of ettringite crystals in a given area, determine at a large extent the bonding resistance of the rendering mortar (Junior and Gomes, 2009).In other cases, the increased bonding resistance is attributed to the utilization of a water repellent additive, which repels water from the interface so that spaces are filled up with fine aggregates, thus hardening this area (Costa and John, 2013).Other authors consider that fine aggregate percentage in the mortar is the most influencing factor, because of the effect it provokes on water/cement ratio and on mortar porosity (Miranda and Selmo, 2006).These parameters can be associated with the mortar ability to transport water towards substrate.Recent researches proved that filler type (Martínez et al., 2013) is a determinant factor of this and other properties, such as water retention and mortar retraction.
This research determines how and the volume of water transported from the mortar towards the substrate.It also connects bonding resistance behavior of different masonry mortars with water transportation in the mortar/block system.Furthermore, it was proved that there is no relationship between the mortar water retention test and the volume of water transported in the mortar/block system.

Materials and dosages
Rendering mortars were prepared by using Portland cement with lime-stone filler (CP II-F-32, made in Brazil, equivalent to CEM II/A-L), also including air hydrated lime aggregate and two sands from natural origin.Above materials are usually employed for the elaboration of rendering mortars and they are easily obtained in the region where the research was conducted.Table 1 shows the chemical composition of binders.
Table 2 shows the physical properties of bonding agents used for mortar calculation and dosages.
The aggregates employed correspond to average size sands from alluvial deposits, classified under NBR 7211 (ABNT, 1993).Their properties are shown on Table 3.These two types of sand were selected because of their grain size differences, which might influence water transportation.They were used to elaborate mortars with similar water/aggregate ratio and identical fine aggregates content, which are one of the main parameters determining variations of mortars properties.In such a case, parameters will be constant; so as to determine the influence the sand fineness module has on the mortar water transportation towards the substrate; as well as the influence derived from the type of substrate.
Materials mixing ratios were defined in accordance with a dosage research for rendering mortars used for external walls.Ratios are based on the regulation NBR 13755 (ABNT, 1996), varying from 1:0.5:5 to 1:2:8 (cement: hydrated limestone: moisture average size sand, in bulk).Table 4 shows dosages employed to elaborate the two types of mortars.WE observe there are slight differences in the lime content, which are necessary to maintain the mixtures with the same amount of fine material, as sand after passing through the sieve #200 has not the same material percentage.A consistency higher than 220 mm was worked out, so as to guarantee machinability in accordance with the mixing process and pouring technique.Apart from this machinability criterion, the amount of water employed had to guarantee that both mortars could have similar values for water retention te sts, so as to analyze differences between such tests; measuring water transportation by using humidity sensors.
In addition to mortars, this research employed two types of blocks commonly used for structural masonry: ceramic and concrete blocks.The selection of such material, used as substrate, has a great importance in this research because its porous structure, surface texture and absorption characteristics determine the development of mortar properties, in the mortar/block system.Table 5 shows the properties of blocks under study.Results are in accordance with the evaluation of blocks properties, at the age of 8 months.Blocks of this age were selected so as to avoid changes of properties in concrete blocks.Changes take place at early ages as the result of cement paste hydration reactions, which might influence the water transportation towards the substrate, as well as affecting the bonding resistance results.
Figure 1 shows the results from mercury-intrusion porosimetry developed on the two blocks types.We observe that ceramic block has a volume approximately three times higher than concrete block.These values are in accordance with water absorption results indicated on Table 5.On the other hand, the pores volume of interval 5 to 1000µm, which can be classified as big capillary pores, is considerably greater in the concrete block.This explains the higher value of initial absorption rate the concrete block has, if compared to the ceramic block.In the research developed by Rato (Rato 2006), this behavior was compared among mortars.The same research yielded a direct relation between the total absorbed water and the available pores volume, as well as a direct relation between biggest size pores and the initial water absorption by capillarity.

Experimental programme
For the experimental stage, mortar specimens were elaborated, 4x4x6cm, so as to determine mortars properties, according to procedure described by the ABNT NBR 13276:2005.In order to measure water transportation, ceramic blocks and concrete blocks were coated by using a device which is a "freefall box".The rendering mortar is poured from a standard distance (50cm) in freefall, with fixed impact energy when the mortar makes contact with the block surface, in accordance with the procedure described by Paes (2004).
So as to control water transportation in mortars covering the blocks, humidity resistant sensors were developed, in order to obtain the evolution profile of water transported from fresh mortar to the porous substrate, as described by Paes (2004).After pouring mortar on blocks, this arrangement was placed inside a hermetically sealed container keeping relative humidity close to 100%, as shown on Figure 3.This procedure was carried out to quantify water transportation by suction in the block, thus avoiding water loss from the mortar due to evaporation.Recordings for the evaluation of water transportation from the fresh mortar were carried out during the first nine hours (540 minutes), totalizing 25 humidity recordings from each litter.In the first half an hour, recordings were made every minutes, later they were recorded every 10, 30 and 50 minutes.During the total 9-hour period, an important portion of water had already been transported to the substrate (generally, more than 65%).Independently from the rest of variables involved, this enables us to establish a correlation between such behavior and the coating performance, taking into account that the two mortars were elaborated using the same binder agents and; that they were exposed to similar conditions during recordings and after those were completed.From mortars obtaining similar bonding resistance and water transportation values, at the age of 28 days, different samples were extracted so as to determine porosity, by using mercuryintrusion porosimetry, with the purpose of explaining their behavior.

Results from mortars characterization
Table 6 shows the results from the mortar evaluation in fresh and hardened conditions.We shall observe that consistency values are higher than 220mm diameter on the fluidity device, as previously defined by this study.Mortar MB happened to be more fluid, due to a higher amount of water to compensate the higher content of lime-stone aggregate, which is necessary to maintain the total fine aggregate content under steady condition, thus guaranteeing similar results from water retention tests.As it was previously defined, we confirmed that tests related to water retention, such as the funnel test and paper filter test showed similar results for the two mortars.This case indicates that conditions mortars undergo to retrieve water from their are not to provoke differences in the behavior of these mortars (differing by sand fineness module and w/c ratio).For these test conditions, it seems to be that the utilization of an equal amount of fine aggregates in mortars was the factor that influenced this property the most.Although mechanical properties of mortar B are slightly higher than mortar A, they can be considered as equal, since both classify under the Brazilian standard as mortar type P3-R2, in regards to compressive resistance and tensile strength, respectively.

Water transportation
By using the humidity sensor recordings, the amount of water transported was classified into three water flow regimes: Regime (R1) from 0 to 5 minutes, Regime 2 (R2) from 5 to 60 minutes, Regime 3 (R3) from 60 to 540 minutes, in accordance with the research developed by Paes 2004.Table 7 shows the results of water transportation for both mortars, on different substrates of the four defined litters.According to results obtained for water transportation, we observe that the greatest differences are caused by the type of substrate where mortars are placed.When the same mortar as well as B) is applied on the concrete block substrate, there is greater water transportation than when applied on ceramic block substrate.The water transportation index shows that transportation speed in concrete block is quite intense for regimes 1 and 2: the litters close to the interface.This is due to the higher initial absorption rate developed by the concrete block compared to the ceramic block; because the concrete block has interconnected pores of bigger size and higher volume (see critical porosity on Figure 1).We observe that the amount of transported water is not directly related with total porosity and, consequently, it is not connected with the substrate total absorption rate; because water flow depends on the pores size and their interconnection, as bigger size pores increase capillary absorption tension, which facilitates water movement.
Tabla 7. Resultados de la medición del transporte de agua en las camadas de los morteros por tipo de sustrato This differentiated behavior on identical in different substrates, under the same application conditions (2-3) has been studied by diverse research jobs (31)(32)(33).In this way, and although it was not the purpose of this research, it is worthwhile mentioning that the substrate surface treatment involves properties differences on the mortar applied on the same type of substrate using different treatments.
When comparing mortars (A and B) behaviors, applied in the same substrates, there are differences on water transportation values for the case of concrete block.In this case, it is clear that mortar B, in all litters, shows greater water flow, which increases in from the surface litter towards the interface litter.As mortars dosage keeps water/aggregate rate constant with similar amount of fine aggregates, the main difference lies in the grain size distribution of the sand employed.We observe on Table 3, that the sand employed for mortar B is thicker, therefore, authors consider that the factor conditioning the highest water flow in mortar B is the grain size distribution, which is determined by the sand fineness module in this case.Greater size grains facilitated water transportation; this behavior was not observed during water retention tests.It was also observed that mortar water retention was practically identical, thus leading us to the conclusion that having different absorption profiles for a given substrate reflects that water retention test -in the way it is performed -is not capable of identifying the actual behavior of water transportation on mortars subjected to suction.
The initial absorption rate is reduced in the ceramic block, while absorption takes longer, and it can be higher than in the concrete block.There are no significant differences on the water transportation values for mortars A and B; when they are applied on the ceramic block values are quite similar, which is produced by the decrease of initial absorption rate of the ceramic block in comparison to the concrete block.

Bonding resistance by perpendicular tensile stress
In the result obtained for bonding resistance there was an effective contribution made by water transportation (towards the substrate).Such result cannot only be explained by bonding or water transportation results or different factors involved in the system.It was proven that there is a direct relationship between these two factors, which are supported by other studies (Aldo et al., 2013;Peas, 2004;Pereira, 2007;Pereira and Bauer, 2013).
Figure 4 shows the results of bonding resistance by perpendicular tensile stress.In the four cases the prevailing cracking was cohesive (by mortar).The figure shows the average values and the interval corresponding to a probability of 95%.
We observe that in both mortar types (A and B) the bonding resistance was sinificantly higher when they were applied on the concrete substrate, in comparison to the ceramic substrate.It was proven that there is a direct relationship between property and the total amount of transported, being in all higher in the concrete block for the same type of mortar.By comparing mortars, it was also proven that in the case of concrete substrate, mortar B transported more water towards the substrate because the sand employed in the elaboration had a higher fineness module.Mortar B obtained higher values for bonding resistance, while the ceramic substrate did not show significant values.Above agrees with the proximity values of water transportation obtained in this case, where the maximum difference found in the total water transportation for mortar litters was 6%, which corresponds to the surface litter.

Conclusions
The main conclusions reached from results analysis on mortars properties, water transportation values and mortars bonding resistance in the two substrate types are the following: • By recording humidity percentage using resistant sensors, placed in the rendering coating in fresh condition, it is possible to determine the amount of water transported by the mortar, in the very first hours after its application.
• The case under study showed a direct relationship between transported water in different litters and the mortar bonding resistance.
• The amount of transported water by the same mortar in different substrates depended on the size and volume of the substrate critical porosity; therefore, in all cases the highest values of transported water were obtained from rendering blocks.The higher fineness module sand employed with mortar B seems to be responsible for higher values of transported water and bonding resistance obtained by mortar B compared to mortar A in the concrete block.
• Results proved that there is no relationship between the mortar water retention test and the total amount of transported water in the mortar/block system.

Acknowlegments
Authors wish to thank for the logistic and financial support provided by the following organizations:

Figura 2 .
Figura 2. Posición de los sensores para la medición del transporte de agua Figure 2. Arrangement of Sensors for measuring water transportation

Figura 3 .
Figura 3. Equipamiento para la medición del transporte de agua en el mortero Figure 3. Device used to measure water transportation in the mortar

Table 6 .
Physical properties of bonding agents

Table 7 .
Results from water transportation recordings on mortar litters per type of substrate Upper: Upper litter Inter-sup: Camada intermedia superior/ Inter-upp: intermediate upper litter Inter-inf: Camada intermedia inferior/ Inter-lower: intermediate lower litter Inter-fas: Camada de la interfase/ Interface: interface litter