The cement CEM I 52.5N-SR5 used for the different mortar formulations in accordance with the requirements of the NF EN 197-1 standard was from the Lafarge cement plant (France). It had a Blaine surface of 3800 cm²/ g (38) according to the standard EN196-6 and an absolute density of 3.13 g / cm³(38) according to the standard NFP18-555. The mineralogical composition of the clinker used is presented in Table 1.

We used in our study two types of additions: the calcined mud and fly ash. The results obtained by the laser granulometry technique of these additions are shown in Figure 1.

The laser granulometry analysis was carried out using the CILAS 1180 granulometer. The samples were also subjected to wet analysis. The granular distribution of the grains of the calcined mud showed a higher proportion of the fine particles compared to the fly ash. It can clearly be noted that the proportion of 95% of the particles of the calcined mud was less than 50 microns in comparison with 75% for the fly ash, which confirms the great fineness of treated mud.

An alluvial siliceous sand was used, of particle size 0/4, from the quarry of La Repennelais (France), with a density of 2.58 g/cm³ (NFP 18-555) and a water absorption coefficient of 1.2 % (38) follows normative rules NF P 18-554 .

A superplasticizer Sika Viscocrete Tempo 11 was used, of a new generation multi-purpose water reducer based on acrylic copolymer, with a density of 1.06 ± 0.01 and a solid content between 28.5 and 31.5 %.

As an integral part of a SCC, self-compacting mortars (SCEMs) may serve as a basis for the design of concrete since the measurement of the rheological properties of SCCs (41). According to the procedure recommended by EFNARC committee (European Federation for Specialist Construction Chemicals and Concrete Systems) (42), the measure technique helped to characterize rheological properties of SCCEM. All mortar mixtures were designed to give a slump flow diameter of 20 ± 1 cm instead of 25 ± 1 cm to avoid bleeding and segregation. In the production of SCEMs, the mixing process was kept constant to supply the same homogeneity and uniformity in all mixtures (41).

After leaving the cone, the final diameter reached by the slump flow slab is measured.

The initial and final setting times of the self-compacting concrete equivalent mortars prepared were measured using a classical automatic VICAT apparatus, according to standard NF EN 480-2.

The tests were carried out on prismatic specimens of dimensions 4x4x16 cm³, previously preserved in water as soon as they were demolded, until the crushing deadline (28 days) in order to determine the flexural (tensile) strength, using the three-point bending test, and the compressive strength according to the standard EN 196-1.

The main objective of the thermal analysis was to follow the evolution of bound and free water content of the binder pastes. The compositions of these pastes are given in Table 7. It is worth reminding that bound water can be considered as an important and reliable indicator of hydration progress and microstructure modification (43). A crushed sample, with particle size less than 2.5 mm (about 4 g), was placed in a calcined and tared crucible. Let M_i be the initial mass of the paste. The crucible was then introduced into the oven at the temperature of 80 °C to dry the sample. Each weighing was carried out after allowing the crucible, holding the paste, to cool down for 15 minutes inside a desiccator containing silica gel. Let M_{80 °C} be the final dry mass of paste (44). The relative mass loss corresponding to free water is then given by [1]:

After that, the crucible containing the dry ground matter was introduced into the oven at 105 °C, for 2 hours; then, it was allowed to cool for 15 minutes inside a desiccator containing silica gel; next, it was weighed. The operation was repeated, but with the oven temperature set first at 400 °C and then at 600 °C, for 1 hour, for the two temperatures. The relative loss of mass at 600 °C is then given by [2]:

The operation was repeated in the oven where the temperature was previously set at 1000 °C. The relative loss of mass between 600 °C and 1000 °C is therefore equal to [3]:

This test consisted of measuring the percentage of voids, inside the mortar mass, which are connected to the surface, in accordance with standard NF P 18-459. Porosity is measured on prismatic specimens of dimensions 4x4x16 cm³, at 28 days. The porosity accessible to water, ε, which is expressed as a percentage of the volume, is given by the following relation [4]:

Where:

M_water is the mass in grams of the half specimen immersed in water

M_air is the mass in grams of the soaked half-specimen weighed in air

M_dry: the mass in grams of the dry half-specimen

The present study was carried out on mortar samples of dimensions 4x4x16 cm³, in accordance with the French standard NF P 15-433. The autogenous shrinkage and total shrinkage measurements were carried out from 24 hours after the beginning of malaxation. The specimens used to investigate the autogenous shrinkage were fully coated with adhesive aluminum foil to prevent any water exchange with the outside environment. All test pieces were then stored in a room with a controlled atmosphere. The measurements were carried out using a micrometer comparator, at a rate that is adapted to the kinetics of the observed phenomena. The difference between the total shrinkage and autogenous shrinkage allows obtaining the drying shrinkage.

In order to study the early-age cracking tendency of SCCEMs, it was decided to carry out restrained shrinkage tests using ring tests (43), according to the standard ASTM C.1581(45). The approach consisting of casting a mortar ring next to a steel ring was used to simulate the restrained shrinkage cracking (46). The test method involves casting a mortar ring around a steel ring. The geometry and testing device for specimens are described in Figure 3 (18). In this study, the upper surfaces of mortar rings were protected from desiccation by a double layer of aluminum adhesive, only the circumferential surface was exposed to drying (47). When the crack appeared on the surface of the mortar ring, its width was first measured using a microscope, then, a sensor was placed on the crack for measuring that same width. The evolution of the crack opening was recorded, every 15 minutes, over time.

All mixtures of mortar were made to give slump flow diameter of 20±1cm (avoiding segregation) which was acquired by adding the used admixture (41). The slump flow test results presented in table 6 and figure 4, show clearly that SCEMs present satisfying values, between 19.25 and 20cm. It was noticed that the use of calcinated vase increased slightly slump flow diameter of mixtures. In addition, it should be noted that all the mortars tested are located in relatively narrow ranges of values, given the uncertainties recognized for this test (48).

The measurement results of the setting times of the different concrete equivalent mortars are presented in Figure 5. It was easy to observe that the mineral additions (calcined mud and fly ash) had a very light effect on the setting times of the formulated mortars as compared with the SCEMC with a maximal gap of 25 minutes.

This may be explained by the fact that the fineness of the mixture, which is attributed rather to calcined mud (8310 cm²/g) and fly ash (4334 cm²/g) than to cement (3800 cm²/g), allowed having a larger area of contact with water. In addition, the setting time increased is attributed to the low hydration heat of mortars containing additions, because of the lower clinker content. Indeed, Jian. S et al. (49) studied the effect of mineral additions, such as fly ash, silica fume and blast furnace slag, with different percentages, on the hydration heat of mortars. They found out that the endothermic hydration process of mortars was greatly influenced by the partial replacement of cement with one of these additions (50). Thus, the delay in setting times during introduction of the studied additions may be due to the low hydraulicity of the fly ash and calcined mud. It is also noted that the replacement of Portland cement with fly ash involved a delay in the initial and final setting time of the obtained mortar as compared to that of mortar containing calcined mud. The time delay can be attributed to several factors, such as the high fineness of calcined mud, higher superplasticizer content of SCEM CM as compared to that of SCEM FA, low hydration heat, and slow hydration kinetics of SCEMCM at young age.

The evolution of the compressive strength, at 28 days, as a function of the mineral additions is illustrated in Figure 6. According to the results obtained the SCEMC displayed better mechanical performances than those of SCEMCM and SCEMFA.

Depending on the nature of the addition, the calcined mud SCEM developed better strengths as compared to the fly ash SCEM. This can be explained by the physical effect of the calcined mud as a result of great finesse (8310 cm² /g) compared to the fly ash (4334 cm²/g). Indeed, Sabir and al 2001 have shown that finesse, the morphology of the particles is also an influential parameter on the intensity of the Pozzolanic activity (51). In addition, the activity index of the calcined mud was 0.93 at 28 days compared to 0.79 at 28 days of fly ash, which was a sign of a strong Pozzolanic activity.

In addition, this can also due to the latent property of hydration of cements containing fly ash (43). The kinetics of the development of strengths in SCEMCM may be explained by the activity of mud and by the fact that the pozzolanic action became sensitive from the early age, this action broke out and started fixing the lime released by cement during its hydration. As a result, the newly formed calcium silicate hydrates (CSHs) will contribute to strength (52).

For better visualizing the evolution of the compressive strength, we compared the various ratios of the SCEM containing the mineral additions (calcined mud and fly ash) with the SCEMC, as illustrated in Figure 7. This fraction was greater for calcined mud SCCEM than for fly ash SCCEM. At 28 days, the compression strengths of fly ash SCCEM and calcined mud SCCEM were smaller by 5% and 2%, respectively, than that of control SCCEM.

The 28-day flexural strengths were shown graphically in Figure 8. It is well known that the factors that affect the compressive strength evolution of mortars also influence their flexural strength evolution (14). However, this was not true for our mortars because calcined mud SCCEM had the smallest tensile strength value. Neville (1996) indicated that there was a direct proportionality between splitting tensile strength and compressive strength of the concrete strength (53). Nevertheless, the two flexural strengths of mortars with the two additions (calcined mud and fly ash) remain close to each other.

According to the results obtained, control SCCEM displays higher flexural strength than those of calcined mud SCCEM and fly ash SCCEM.

For better visualizing the evolution of the flexural strength, we compared the various ratios of the SCEM containing the mineral additions (calcined mud and fly ash) with the SCEMC, as shown in figure 9. The fraction of fly ash SCCEM was greater than that of calcined mud SCCEM. At 28 days, the flexural strengths of fly ash SCCEM and calcined mud SCCEM were smaller by 17% and 22%, respectively, than that of control SCCEM.

Figures 10 display the results of measurements of the quantities of free water and bound water, for the three cement pastes, after 28 days of dry cure. The results showed a lower bound water content for the mortar paste containing fly ash compared to calcined mud and reference paste, which indicated less hydration and therefore the likely development of a smaller amount of hydrates at young age, and so less compressive strength in comparison with the compressive strength of SCEMC and SCEMCM. The results obtained indicated that the bound water content was greater in paste containing calcined mud as compared to the other two. This means that hydration was more important, and therefore a larger amount of hydrates was very likely to develop. The binder containing calcined mud had the highest free water content; it was close to that of the control binder. These results were consistent with those relating to porosity accessible to water. Indeed, it was clearly noted that the porosity of control SCEM was lower, then came that of fly ash SCEM. The measured free water content allowed an estimation of the porosity of the mortars studied. This free water occupied part of the porous network (44), so the quantity of free water in each mortar depends on the porous structure of this mortar.

Figure 11 showed the results for CO₂ content in the three cement pastes, after 28 days of dry cure. These results indicated that the CO₂ content was lower in the mixtures containing calcined mud than in those containing fly ash. The state of carbonation also plays a major role. In the case of Portland cement materials, carbonation caused a decrease in the diffusion coefficient which was well correlated with the decrease in porosity (54). In contrast, the CO₂ content of fly ash paste increased while its porosity and bound water content decrease. The increase in carbonation for fly ash with less porosity compared to calcined mud may be related to differences in the porous structure. The use of fly ash gives a denser porous structure, but at the same time a portion of Ca(OH)₂ was consumed by the hydration of fly ash, which lead to a high depth of penetration (38).

The carbonation depth also depended on the nature of the addition, in particular the CaO content of the fly ash. The higher this content, the lower the carbonation depth (38). This can be explained by the low percentage of CaO (3%) in the fly ash compared to the calcined mud (17%).

According to Audenaert et al. (2007), the incorporation of fly ash had no noticeable effect on improving the strength of SCCs to carbonation (55). Ongoing studies, such as capillary absorption and permeability may to corroborate these results.

Figure 12 showed the porosity results obtained for the three concrete equivalent mortars under study, after 28 days of dry cure. It can be seen that the porosity of concrete equivalent mortar containing fly ash was lower than that of calcined mud concrete equivalent mortar, and the accessible water porosity of fly ash SCEM is 8% higher than that of control SCEM. In addition, the use of calcined mud as a partial substitute for Portland cement lead to an increase in water porosity of 17.2%, at 28 days, with respect to that of control SCEM. These results were in agreement with those relating to the free water content. Comparison between the results relating to the mortars studied allows concluding that the porosity accessible to water increases in mortars containing mineral additions as partial substitute for cement (38). This increase in porosity during the introduction of calcined mud and the fly-ash can be due to the least pozzolanicity and hydraulicity of these additions in short-term (28 days). In addition, the pozzolanicity or the hydraulicity of the fly-ash can reduce the porosity in long-term if the mortar is preserved wet (56).

The restrained shrinkage tests were performed during a 49-day period. The formwork removal took place 24 hours after the manufacture of the SCEMs. Figure 16 showed the evolution of crack widths.

Flexural stresses developed under the effect of the restrained shrinkage; cracks appear when these stresses exceed the flexural strength of the mortar (64).

When monitoring the restrained shrinkage using the ring test under desiccation conditions, the SCEM CM and SCEMC were the first to crack, as illustrated in Figure 16. The SCEMCM cracked first at 2.71 days after formwork removal, followed by SCEMC at 2.81 days, and by SCEMFA at the age of 3.98 days. From Figure 16, crack openings due to the restrained shrinkage are determined using a sensor. The results obtained clearly showed the age of the appearance of the crack as well as the evolution of its opening, for all the SCEMs under study. Cracks can occur when the flexural stress due to the restrained shrinkage created within mortar exceeds the flexural strength of the material (43). At early ages, this flexural strength increased steadily and the constraints were generated by volume changes (18). Improvement in flexural strength of the concrete provided resistance against shrinkage cracking (65).

The SCEM CM and SCEMC cracked earlier than the SCEMFA. However, the crack openings in the control concrete equivalent mortar were smaller than those in the mortars containing mineral additions. The SCEMC crack presented an aperture that was 1,2 and 1,3 times lower than those of SCEM FA and SCEM CM, respectively. This difference was certainly due to the low porosity of SCEMC. The differences between widths of cracks observed in this study should be related to the differences in the fracture energy, flexural strength, and stress developed in mortar rings at early age (43). The incorporation of 22% of fly ash in mortar delays the appearance of cracks. This result can be explained by the high flexural strengths and the low compressive strengths of SCEM FA in comparison with the SCEM CM. The flexural strength of the mortar containing the calcined mud clearly intervened in the sensitivity to cracking, where the SCEMCM underwent higher flexural stresses despite its higher compressive strength, it cracked the first, i.e. the stress increased with early crack. Also, one can refer to its higher drying shrinkage and its weak autogenous shrinkage, since drying shrinkage was one of the major parameters responsible for cracking (66).

The results relative to bound water content showed that there was an indirect correlation between the amount of bound water in mortar and the age of cracking. In fact, the increase in bound water lead to a decrease in the age of cracking, thus increasing the risk of cracking. In effect a normal consistency test was performed by the Vicat apparatus, on a paste of 100% of calcined mud was in normal consistency with a ratio water / mud of 0.45 whereas a paste of 100% of cement was in normal consistency with an E / C equal to 0.26. (67), in addition to the great finesse of this mud which explained the high demand for water for this mud and therefore a high bound water content leading to the risk of an early crack. We suggest a setting retarder to reduce the risk of cracking. Ongoing studies, such as capillary absorption, ATG and XRD analysis, SEM observations that corroborate these results.