Additives are not essential components of the concrete mix, however, they are of great importance when there is a need to modify the characteristics of this material, in such a way, it adapts to the conditions of the project and/or the requirements of the builder. Sometimes, the use of some admixture may be the only means to obtain the required result, for example, to give resistance to the material against freezing and thawing, to accelerate or retard the concrete setting time, or to obtain exceptionally high strengths.

Accelerators were the first type of admixture to be used in the construction of reinforced concrete structures [1]. These additives have been used, mainly, to advance the stripping formwork and its subsequent construction processes, however, in most cases, their use produces an increase in the shrinkage suffered by the concrete when it dries, which must be compensated with effective curing.

According to the theory of concrete technology, the main effects of accelerators on the properties of concrete are to modify its setting time and mechanical strength [2]. The setting is the process of hardening and loss of plasticity of concrete or mortar, produced by the drying and recrystallization of metal hydroxides, resulting from the chemical reaction between water and metal oxides present in the cement clinker. The effect of accelerators on the plastic state of concrete is the reduction of the initial and final setting times. The degree of this reduction varies depending on the amount of accelerator, the temperature of the concrete and the ambient temperature [3].

Once it has finished setting, the concrete begins to form stronger bonds which produce its ability to resist different mechanical stresses [4]. The effect of the accelerator in this hardened state occurs mainly at early ages, and it is reduced later.

Calcium chloride (CaCl₂) has historically been the most widely used accelerator, but due to its chlorinated composition, it has drawbacks when used in concrete elements that contain reinforcing steel since it can cause corrosion, for this to happen, the threshold of chloride initiators of corrosion inside the concrete must be exceeded, which is in an interval between 0.2 and 2.0 kg/m³ of concrete, according to the salinity of the environment [5]. On the other hand, when large amounts of chloride are present, the concrete tends to retain more moisture, which also increases the risk of corrosion. It is important to keep in mind that chlorides can enter the concrete by addition during its manufacture or by exposure in coastal environments [6]. Because of the above, building codes restrict the use of calcium chloride for durability.

Due to the above, in recent decades, there has been a growing interest in the use of chloride-free accelerators, such as calcium nitrite Ca(NO₂)₂ [7]. It has been observed that chloride-free accelerators are effective in reducing the setting time, but in some cases, they experience a tendency to reduce strength in the first three weeks. This phenomenon depends mainly on the type of cement and aggregate used [8]. In general, these additives tend to be more expensive than those containing calcium chloride.

This study was conducted in the Yucatan Peninsula, located south of the Tropic of Cancer. The predominant climate in the region is classified as warm sub-humid, with few variations in temperature between the seasons, and with rains in summer [9], this type of climate is common in most of the countries of Central America and the Caribbean, as well as in several countries of the world located at similar latitudes. The average maximum temperature in this region is 36˚C (97˚F) and occurs in May, the average minimum temperature is 16˚C (61˚F) and occurs in January, and the average annual temperature is 26˚C (79 ˚F) [10]. In this type of climate, conditions have been presented that favor the rapid evaporation of water from the mixtures, which in turn causes cracking due to plastic contraction in the concrete, this is of great importance for the control of using additives in concrete [11].

The concrete aggregates used in the study region are the product of crushing limestone rocks formed from calcareous Cenozoic sediments. Its chemical composition is approximately 80% calcium carbonate, 15% magnesium carbonate and small amounts of other compounds (silicon dioxide, iron oxide, sulfates, nitrates, and chlorides) [12]. These aggregates have main characteristics as their high absorption and many fine particles, the foregoing, together with the angular shape of the particles, favors greater water requirements in the mixtures, and therefore of cement. Some studies have proven that mixtures containing fine limestone experience slower setting speed, compared to other concrete studied, this is probably because the reaction induced by calcium carbonate reduces the impact of the chemical additions of the accelerators [13].

The objective of this study was to verify the effect of using two accelerating additives, one with chlorides and the other free of chlorides, on the properties of concrete at early ages, in a context of a warm sub-humid climate.

The investigation was of an experimental type and consisted of the measurement of some properties of the concrete, in samples manufactured in the laboratory, using different doses of accelerating additives. In the preparation of the concrete, two ratios were used between water and cement measured by weight (W/C), one of them typically used for structural function and the other for reinforcement and confinement of masonry elements.

The dependent variables that were measured in the experiment, according with American Society for Testing and Materials (ASTMs) standards, were the following: Slump (ASTM C143) [14], Volumetric weight of fresh concrete (ASTM C138) [15], Air content (ASTM C231) [16], Initial and final setting times (ASTM C403) [17], and Mechanical resistance to axial compression (ASTM C39) [18].

The independent variables that were controlled in the experiment were the following:

• The W/C ratio took two values: 0.50, with which it was intended to achieve an f’c of approximately 300 kg/cm², and 0.70, with which it was intended to achieve an f’c of approximately 200 kg/cm².

• Type of accelerating additive, which took two categories: the first was an additive with chlorides (which will be abbreviated as CL) and the second was a chloride-free additive (which will be abbreviated as CLF).

• Dose of the accelerators, which took three values: 0.5%, 1.0% and 1.5%, calculated in relative terms as a percentage of the weight of the cement used in the mixture. These three dosages are within the range recommended by accelerator manufacturers for use in warm weather.

• The curing treatment took two categories: wet, by means of saturation by immersion in a pool, to favor the hydration of the cement (ASTM C192) [19], and natural in the laboratory, trying to imitate the conditions that occur in the real context of many job sites, where the moist curing process or by means of membranes is generally omitted.

• The compressive strength test age took three values: 3, 7 and 28 days. This allowed us to observe the changes in this mechanical property during the period in which most of the cement hydration reactions take place.

Other variables that were measured were: the temperature and relative humidity measured in the laboratory at the time of unloading the concrete from the mixer, to know the climatic context in which the concrete tests were carried out in the fresh state.

The materials used were the following:

• CPC 30R type cement (Compound Portland Cement with a resistance of 30 N/mm²) that meets the specifications of the Mexican Standard NMX-C-414-ONNCCE [20] and is equivalent to Portland type I (ASTM C150) [21].

• Crushed stone aggregates of typical limestone from the north of the Yucatan Peninsula. Table 1 shows the main properties measured in the sample of aggregates used in the study.

• Two types of accelerating additives: CLF and CL, which according to their technical data sheets, comply with the ASTM-C 494 [22] standard. By means of a chloride analysis stoichiometry test, it was determined that the number of chlorides in the CL additive was 263.748 g/l.

The concrete samples studied were conducted as follows: for each W/C ratio studied, seven concrete manufacturing processes were carried out, six with an accelerating additive and one without an additive (control sample). For each combination of W/C, type, and dose of accelerator, 18 standard specimens (ASTM C39) were manufactured to carry out the resistance tests, which were divided into two groups randomly, and each of them was assigned one of the two curing treatments. Each of these groups of specimens was divided into three subgroups to randomly assign the strength test ages. In total, 252 concrete specimens were manufactured (126 for each W/C), of which 84 were tested in compression for each age.

Using the ACI-211 [23] concrete mix design method, the amounts of material for concrete manufactured without additive were obtained, which are shown in Table 2. From these amounts, the corresponding adjustments were made to add the amounts of additive in the three percentages already mentioned.

With the data obtained in the tests, an Analysis of Variance (ANOVA) was performed to evaluate the effect of different factors on the variables of interest, by comparing the means of the samples that resulted from the different levels of the factors. To do this analysis, the variance between the means of the groups and the variance within the groups were compared to determine if the groups can be considered part of a large population or separate populations with different characteristics.

The influence of the two factors: type and dose of accelerator were tested, in the variables initial and final setting times, and concrete resistance to compression. The curing type factor was also tested in the strength of concrete made with the CLF admixture, the most currently recommended.

The Fisher-Snedecor F distribution was used with k-1 degrees of freedom between groups and n-k within groups, where k is the number of samples and n is the total number of observations. The F statistic is defined as the ratio of the mean squares between groups and within groups. The root means squares are defined as the sum of squares divided by the degree of freedom. The sum of squares is calculated as the sum of the squares of the differences from the mean [24].

To reject the null hypothesis, which establishes that there is no difference between the means of the samples, a minimum level of significance of 0.05 was defined, in such a way that probabilities equal to or less than this are interpreted as the factor causing significant differences in the means of the samples studied.

In all cases studied, a decrease in slump was obtained as the dose of additive increased. For almost all combinations, the decrease in slump was in a range between 4 and 6 cm, the exception being the mixtures with W/C of 0.50 and accelerator CL, in which decreases in slump were measured in a range between 6 and 9 cm [25] reported that the use of chloride-free accelerators did not have a notable effect on the fluidity of fresh concrete, which was not confirmed in the present study.

The technical sheets of the manufacturers of the two accelerators used in this study recommend a dose of 1.0% for hot weather, to obtain a 50% decrease in the initial and final setting times. This reduction was not achieved in any mixture. The greatest reductions in the setting times were obtained with the W/C of 0.50, accelerator CL and a dose of 1.5%, for this combination, there was a decrease of 20% in the initial setting and 23% in the final setting. The previous behavior could be attributed to the use of limestone aggregates, which, as mentioned by [13], concretes with this type of stone experience lower setting speeds compared to other types of aggregate, probably due to the induced reaction of carbonate calcic. For their part [26] found that chloride-free accelerators have a lower performance as an accelerating agent in the setting reaction compared to those containing chlorides, which was observed in the present study.

On the other hand, the acceleration in the strength gain of concrete is key to advancing the formwork stripping time of the construction elements and their commissioning. The most critical case corresponds to the flexural elements, which can generally be dismantled after 14 days and put into service after 28 days. With the measured data of the compressive strength, in Table 11 and Table 12 the number of days in which these two events could occur is presented, using interpolations. For example, for the concrete with W/C of 0.50, accelerator CL and a dose of 1.5%, after 5 days the resistance was obtained that corresponded to 14 days in the concrete without accelerator, and at 16 days the resistance corresponded to 28 days. In general, formwork stripping and commissioning could be carried out in approximately half the time for the 1.5% doses.

In relation to the curing treatment, in this study, as in others carried out in the same region, wet curing did not increase the resistance of the concrete with respect to natural curing. Reference [27] has attributed this effect to the high

porosity of the aggregates in the region, which could produce internal curing, have also attributed the lower strength of wet-cured concrete to the effect of semi-saturation of the specimen at the time of performing the destructive compression tests. According to the results of this study, the use of accelerators did not cause a synergistic effect with moist curing in the phenomenon of maturity in concrete.

In the context of the warm region climate:

The two types of accelerators decreased the slump of the mixture, the effect being greater in the accelerator with Chlorides (CLs).

The two types of accelerators decreased the initial and final setting times, with a greater effect on the concrete final setting. However, the reductions contained in the technical sheets of the products were not reached.

The compressive strength at early ages increased with the use of the two accelerators, being more notable for the mixture rich in cement with a W/C of 0.50.

With the data obtained, the significant difference attributable to the accelerators in the setting times was statistically tested, but, in general, the difference in resistance was not confirmed, since the changes in this variable were different for the two W/C.

Thanks to the Laboratory of Materials personnel at the School of Engineering of the Autonomous University of Yucatan, México.

The authors declare no conflicts of interest regarding the publication of this paper.

Solís, R.G., Hay, D.R. and Zaragoza, J.N. (2022) Effect of Accelerant Additives in Concrete with Limestone Aggregate in Warm Weather. Journal of Building Construction and Planning Research, 10, 140-154. https://doi.org/10.4236/jbcpr.2022.104007