In the elaboration of the mortar prismatic specimens, the following materials were used: cement, aggregate (natural and recycled concrete aggregates), water and additive.

The description of the mortars used in these tests has the following coding [1]:

where N shows aggregate type, that could be NA for natural aggregate and RA for concrete recycled aggregate; D shows relation between cement and aggregate by dry-weight proportion, letter X describes relation between components 1:3 and letter Y describes relation between components 1:4; F refers to the fibre types that could be: SF for steel fibre, FPP for polypropylene fibre and FB for fibreglass. Finally, percentage sign % shows the fibre percentage added to the mortar mix.

The preparation of all mixes was developed using the same technique and equipment, following the standard UNE-EN 196-1 (29). Prismatic specimens elaborated with recycled concrete aggregate were prepared using 1% of superplasticizer/additive over the weight of cement, complying with the limits established by the manufacturer. According to the specifications, this additive reduces water content of the mix maintaining mortar workability and improving its properties such as hardness, durability and shrinkage.

Due to the format and characteristics of polypropylene fibers, manual separation of this material prior to its incorporation to the mortar mixes was performed. In this way, homogeneous distribution of this material in fabricated samples was achieved.

Dosage by dry-weight used for the different prismatic specimens is shown in Table 3 where their quantities and relations are indicated.

Water content of these specimens was experimentally set in order to achieve a plastic consistency, in other words, to achieve a diameter of mortar in the range of 175±10 mm, according to the standard UNE-EN 1015-2:1998 (30). Because of this, it is necessary to increase the water/cement ratio while recycled aggregates are used due to their higher water absorption during the mix process.

The experimental study was divided in two phases. In the first phase main characteristics of recycled concrete aggregates were studied and, subsequently, some tests have been carried out in order to study the fibres influence in this type of mortars. Also, some tests were carried out to obtain the compressive and flexural strength of the mortars to characterize the mechanical properties of the mortar mix containing fibres.

Shrinkage measuring was performed with a manual comparator that is shown in Figure 1(a). This comparator is able to measure the changes in length of the prismatic specimens during each drying phase, since they have been unmoulded, following the standard UNE 80-112-89 (31), and measuring in prismatic specimens of 25 x 25 x 287 mm. Thus, shrinkage is expressed as a proportion ∆L_nd (%) of its initial length given by the following equation [2]:

where L₀ is the inner length of the mould, M_nd is the comparator measure after n days and M_1d is the comparator measure on the first day.

On the other hand, this drying shrinkage is associated with a moisture loss of the specimens. This variation in its moisture content was measured with a capacitive sensor of the authors’ own design joined to an auto-oscillating circuit. This sensor is shown in figure 1(b). To obtain the maximum current in the circuit, its impedance must be minimum, condition that appears when the circuit is in resonance. So, it is only necessary to obtain the resonance frequency value to obtain the condenser capacitance, which depends on the moisture content of the specimen, because it serves as dielectric, changing the environment permittivity. This is given by the equation [3]:

where f, is the resonance frequency of the measurement circuit, in Hz; L is the induction of the coil, in H; and C is the condenser capacitance in F, that depends on the area of the confronted plates A^, on the distance between plates d which is the same that the specimens’ thickness and on the material’s permittivity ε that, in our case, is the mortar specimens and it varies depending on their moisture content.

The shrinkage measurement period was 180 days since the specimen had been removed from its mould till the last measurement.

Flexural and compressive strength tests were carried out using standard specimens RILEM with the dimensions 40x40x160 mm, after being cured in the curing chamber for 28 days under the temperature of 20°C and following the standard UNE-EN 1015-11:1999 (32) which regulates flexural and compressive strength tests on mortars.

Recycled concrete aggregates characterization was carried out following the standard UNE-EN 13139:2002 (33). Used recycled aggregates comply with the size restrictions and the maximum limit of fineness modulus established for masonry mortars. The results of physical characteristics for recycled concrete aggregates (RCA) and natural aggregates are shown in Table 4.

As it can be seen in Table 2, the density of recycled concrete aggregate is lower than the natural aggregate density, so the mortars elaborated with this type of aggregates have lower density compared to the traditional mortars. On the other hand, water absorption values of the recycled aggregates are much higher than the values obtained by the natural ones. That is a general characteristic of the recycled aggregates and, mainly, this is caused by the amount of mortar adhered to the aggregates in CDW (13). In turn, due to its higher water absorption, recycled mortar shrinkage is much higher than one of the mortar elaborated with natural aggregate, as recycled mortar needs a higher intake of mixing water.

One of the other characteristics that differentiates recycled aggregate from natural aggregate is its higher value of friability coefficient. This higher deterioration of recycled aggregate has a negative impact over the mechanical characteristics of recycled mortar.

To determine particle size distribution of aggregate, a sieving method was used, according to the standards UNE-EN 933-2:1995 (38) and UNE-EN 933-1:2012 (34). Particle size distribution is one of the most important characteristics studied in aggregates (39), being directly relevant to the characteristics of elaborated mortar such as workability, hardness, durability and compactness. Figure 2 shows size distribution curves of two used aggregates, where the continuous size distribution curve can be observed.

In order to complete the characterization of the recycled concrete aggregate (RCA), X-ray diffraction and X-ray fluorescence methods were used. The obtained results are shown in Table 5 and Figure 3.

The obtained results of X-ray fluorescence method for the recycled aggregate type show that there is a high level of silicates SiO₂ (68.2%), followed by a high percentage of Al₂O₃ and CaO. Apart from this, SO₃ content (0.338%) does not exceed the limit of 1% established by the standard UNE-EN 13139:2002 (33) for the aggregates used in mortar fabrication, as higher levels of this chemical harm the mechanical characteristics of the elaborated mortars reducing considerably their durability.

The crystalline structure and the purity of the specimens were analysed by X-ray diffraction (XRD) and it can be noticed that the X-ray diffraction pattern was measured using a Cu-Ka radiation (λ=1.540598 Å) by a diffractometer equipped with a monochromator of graphite. Data were collected at 300K in steps of 0.05°, measured in a range of 10°≤ 2θ ≤ 110° taken every 10s with a current of 20 mA and a 40 kV voltage.

As it can be seen in Figure 3, the main crystalline phases obtained with the X-ray diffraction method are quartz and calcite because they show a higher intensity after the test. Other founded phases are gypsum, sanidine and albite.

The obtained results of shrinkage for both reference mixes and recycled mortar mixes, that incorporate three different types of fibres, can be seen in Figures 4 and 5, corresponding to 1:3 and 1:4 dosages, respectively.

As it can be observed in Figures 4 and 5, the shrinkage produced in recycled mortars is much higher than that of traditional mortars, according to the theoretical predictions due to its higher water absorption. In the case of natural mortar mix, the shrinkage is higher in the mortar with dosage 1:3; this is related to the higher amount of cement in the mix. On the other hand, in the mix elaborated with recycled aggregates, the dosages with higher aggregate content (1:4) are the mixes that present higher shrinkage, due to the greater influence in this case of the used aggregate type with respect to the amount of used cement.

Moreover, it can be seen that the use of fibres reduces significantly the shrinkage results in recycled mortars. In both dosages 1:3 and 1:4, the mixes which incorporate fibreglass up to 0.2%, were those that have lower shrinkage reduction rate in case of recycled mortars with 30% and 20%, respectively. This decrease of shrinkage results is more important at the ages that are superior to 30 days, as until reaching this point, all specimens performance is similar. Among all these used fibres and percentages, the recycled mortars with 0.2% of polypropylene fibre, are the ones which present the best shrinkage results in both dosages, reaching values similar to the natural aggregates, and with a decrease of 40% in shrinkage in case of dosage 1:3 and 30% in case of dosage 1:4.

To analyse the obtained shrinkage results an analysis of variance (ANOVA) was carried out to determine if the added fibres and the used dosages have influence on the characteristics of recycled mortars. Moreover, a multi-range test was carried out to analyse the significant differences among the fibre types, their percentages and their components relation. The results of the ANOVA analysis and multi-range test are shown in Tables 6 and 7.

As it is shown in Table 6, both the fibre types and the used dosages obtained p-values far below 0.05, so it can be concluded that both factors have a great influence on the mortars shrinkage. In the multi-range test a significant difference between the polypropylene fibre (in 0.2% proportion) and the rest of fibres, with the exception of steel fibre (in 1% proportion) can be observed. In the multi-range test of the used dosages the results show important differences between both components relations, being the dosage 1:4 the one that presents better results.

Moreover, the moisture losses of the specimens at different ages were measured with a capacitive sensor using both mortar dosages, associating this variation with the shrinkage produced in each dosage. As an example, some results are shown in Figures 6 and 7, without including all the curves because of the great number of obtained values. Table 8 shows all the equations of the obtained curves and their R² coefficients.

As it can be observed in Figures 6 and 7, the moisture loss is more important in the early days, due to the exothermic reactions produced by the setting process of cement. This moisture loss, in its turn, is much more important in the mortars elaborated with recycled aggregates than that of traditional mortars, in accordance with the shrinkage values, which were higher in case of recycled concrete aggregates. The addition of fibres improves significantly the behaviour of these mortars, reducing the moisture loss. Moreover, it can be appreciated that, at late ages, the moisture loss is virtually zero, while the specimen shrinkage continues along the mortar life since it is placed.

In order to complete this study of recycled mortars with fibres addition, flexural strength, compressive strength and capillary water absorption tests were carried out. Mechanical properties of recycled mortars are in general poorer that those of traditional mortars, due to poorer properties of recycled aggregates, especially absorption. As shown in diverse studies, when the percentage of recycled aggregates rises, mechanical properties decrease (5, 7, 12, 21), being interesting the possibility to improve these properties using fibers in order to obtain higher ratio of recycling with better characteristics. Generally, the results of this paper show a better mechanical behavior of mortars elaborated with fiber compared to recycled mortars without this addition, however, there are no significant differences in terms of improvement of mechanical properties depending on the type of used fiber. The mixes that contain fiberglass obtained an average improvement of flexural strength of 15.86%. In terms of compressive strength, the average improvement is of 11.27%, being more representative for 1:4 dosages. Steel fibres, improve flexural and compressive strength results up to 15% on average. However, recycled mortar mixes elaborated with polypropylene obtained the poorest results, obtaining some slightly lower values compared to the reference mixes RA-X and RA-Y.

In terms of capillary water absorption test (Table 9), higher absorption coefficient was obtained due to higher content of occluded air in fresh mortars presented by recycled mortars compared to traditional mortars as shown in other papers [40]. In case of fibers incorporation, slight decrease of this coefficient is not significant, that is why it cannot be concluded that incorporation of fibers produces better behavior of the material in this respect. Furthermore, there is no standard that establishes a maximum value for this characteristic of masonry mortars, however it is convenient to take it into consideration in order to avoid water filtrations in walls bases and renderings placed in external facings.

Other observed patterns, is that the increasing percentage of added fibre does not improve flexural and compressive strength of recycled mortars, as their workability decreases causing a worse compaction, which is an essential characteristic in the mechanical capacities of these materials. The results are shown in Figures 8 and 9.