1. Introduction
In accordance with EU Directives, there is a need to improve the energy efficiency of buildings. In the process of improving energy efficiency in construction, appropriate technologies and materials are used [
1,
2]. One way is to reduce heat transfer through horizontal and vertical partitions. Porous materials have lower heat transfer coefficients than materials with a continuous structure. This is due to the closure of small portions of gas (air, carbon dioxide) in their volume. However, if the spaces containing gas are too large, convection will begin to occur in them, which will increase the heat transfer coefficient of a given building material. This is the basis for the use of foamed mortar in construction. However, the term foamed mortar is not widely used. The term foamed concrete is more often used, despite the fact that foamed concrete does not contain coarse aggregate, and often fine aggregate (sand). Foamed concrete (FC) is classified as lightweight concrete with density ranging from 280 to 1800 kg/m
3 [
3,
4,
5] and with a minimum of 20% of air-pores volume in the cementitious mix [
6,
7]. Due to the air pore structure of the cement matrix, foamed concrete is characterized by good properties thermal [
8,
9,
10,
11,
12,
13]. The thermal conductivity coefficient ranges from about 0.1 W/(m∙K) to 0.7 W/(m∙K) for foamed concrete with density ranging from 600 kg/m
3 to 1600 kg/m
3 [
14,
15,
16,
17,
18,
19]. For the appropriate composition of foamed concrete, a thermal conductivity coefficient of up to 0.048 W/(m∙K) can be achieved at a density below 500 kg/m
3 [
20,
21]. For example, Proshin et al. [
22] used polystyrene granulate additives as composite of foamed concrete mix and they obtained a thermal conductivity coefficient in the range from 0.06 W/(m∙K) to 0.16 W/(m∙K) for foamed concrete with a density of 200 kg/m
3 to 650 kg/m
3.
However, these were single research results for given compositions of foamed concrete mixes. The physical properties of foamed concrete are closely related to the composition of the foamed concrete mix. The basic components of a foamed concrete mix are cement and a foaming agent of protein or synthetic origin. The remaining components are the same as in the case of other cement composites, i.e. water and possibly aggregate, whereby there is no coarse aggregate, and often also fine aggregate. In addition, various additions, including those from recycling, may be used. Therefore, this study aims to assess the physical properties of hardened foamed mortar with polystyrene granulate addition. Compared to previous studies [
23,
24], foamed concrete was produced on the basis of light plaster mortar. In addition, since tests of such foamed mortar have not been conducted so far, the its mechanical properties and sorptivity were also assessed as part of this study. Moreover, the sorptivity of foamed concrete/foamed mortar are very rarely found in the literature. The authors found only a few literature items presenting the results of such studies, e.g. [
25,
26,
27]. However, this is particularly important due to the potential application of foamed concrete/foamed mortar.
3. Results and Discussion
In the first case, the effect of the addition of polystyrene granulate on the mortar properties (without the addition of a foaming agent) was analyzed. It can be seen that with the increase in the content of addition of polystyrene granulate, density (
Figure 6a), thermal conductivity coefficient (
Figure 6b) and sorptivity (
Figure 6e) were decreased, what is a positive phenomenon. However, at the same time, a reduction in mechanical properties was demonstrated (
Figure 6c,d). The largest decrease in values compared to base sample was obtained for sorptivity and mechanical properties (flexural and compressive strength), see
Table 3.
In the second case, the impact of the foaming agent content on the properties of the plaster mortar with different content of polystyrene granulate was analyzed. It can be seen that the effect of the foaming agent content on the analyzed physical properties of the foamed mortar with the addition of polystyrene granulate is not unambiguous, see
Figure 7.
Figure 7a presents the results of average densities of the hardened foamed mortar with addition of polystyrene granulate for foaming agent content of 2.0, 4.0 and 6.0% of cement mass. It can be observed that the density of foamed mortar decreased with increasing foaming agent content and this correlation is linear. The volume of foam commonly created air voids and resulted in lower density [
25]. These results are consistent with the observations for foamed concrete of the basic composition [
6,
7,
23,
24]. The situation is different in the case of other properties of foamed concrete. Usually, a decrease in the mechanical properties and in the thermal conductivity coefficient of foamed concrete is observed with a decrease in density, which is proportional to the increase in the foaming agent content [
3,
25]. In the analyzed case, the lowest thermal conductivity coefficient of foamed mortar was obtained for the highest foaming agent content, regardless of the content of polystyrene granulate. However, with a foaming agent content of 2% of the cement mass, this value was higher compared to the base sample of plaster mortar and base samples of plaster mortar with addition of polystyrene granulate, see
Figure 7b and compare to
Figure 6b.
Mechanical properties are directly related with density and at the same foaming agent content [
14,
23,
24,
25], while with decrease in foaming agent content (increase in density), flexural and compressive strength also increased [
3,
25]. It is related with the fact, that the addition of foaming agent decreased strength because the volume of foam created pores resulting in lower density [
24]. The obtained results showed that the flexural (
Figure 7c) and compressive strength (
Figure 7d) of foamed mortar was halved with a foaming agent content of 6% of cement mass compared to the base mortar.
Figure 7e presents the correlation between density and compressive strength for foamed mortar with different foaming agent content and polystyrene granulate addition. It can be seen that this correlation is exponential, which is consistent with the results of own [
24,
33] and other researchers [
25,
29,
34,
35,
36,
37] for other foamed concrete mixture compositions. It should be noted the high agreement of the correlation equation obtained in this study and in previous research results for foamed concrete with the basic composition [
22,
24,
33].
Moreover, an interesting observation is that sorption is lower with a higher content of foaming agent, regardless of the content of polystyrene granulate, see
Figure 7f. The lowest values were obtained for polystyrene granulate with the highest content analyzed. There are no results for comparison in the literature, therefore research in this area will be conducted for other compositions of foamed concrete mixes.
In order to determine the quantitative relationship between the change in the amount of the addition of polystyrene granulate (X1) and foaming agent (X2), and the compressive strength and bending strength, thermal conductivity coefficient, density and sorptivity of hardened foamed concrete, multiple regression was used. A linear model was used according to (1).
where:
bi – model parameter (regression coefficient) describing the impact of the i-th variable,
ε – random component (standard error of the estimation Se).
The verification of the models was carried out by assessing the significance of the regression function, regression coefficients and based on the analysis of residuals. The significance of the models was tested with the F test at the probability level of
p=0.05. The multiple correlation coefficient
R and the
R2 index, which determines how much of the variability of the dependent variable is explained by the remaining variables, were adopted as a measure of interdependence between one of the variables and the other variables treated together.
Table 4 presents the results of multiple regression.
In each case, the
F-values are greater than the critical values of the
F – Snedecor distribution at a significance level of 0.05 (
F(2,6)cr = 4.26). The significance of the models is also confirmed by the determined
p-values, which are less than the generally accepted value of 0.05. The correlation coefficients
R≈1 mean that there is a strong linear relationship between the variables. In the next step, the significance of the regression coefficients was assessed, see
Table 5.
It can be observed that in the case of testing the variation of density, compressive and flexural strength and sorptivity, the p-values for partial regression coefficients are significant (p < 0.05).
In the case of testing the variation of the thermal conductivity coefficient
λ, the independent variable - the content of the addition of polystyrene granulate X1 [%] - is insignificant. The independent variable X2 - the content of the foaming agent - is highly significant. However, it was observed that the variation of the dependent variable was explained by 58% (
Rcorrected = 0.58724843). Such a model may be acceptable in complex systems, in which other factors may have an influence. For the λ coefficient, a well-fitting model was obtained in the form of formula (2).
In the next step, the redundancy (lack of collinearity) between independent variables was determined. For each variable, the
R2 coefficient, tolerance, partial and semi-partial correlations were calculated. The semi-partial correlation (also called partial correlation) describes the relationship between two variables, while controlling for the impact of one or more additional variables. In order to illustrate trends and graphically represent the structure of the data, surface plots were made, see
Figure 8,
Figure 9,
Figure 10 and
Figure 11.
Figure 8 present the variation of the thermal conductivity coefficient
λ and the results of the redundancy test. Analyzing the results, it was found that the semi-molecular correlation for the variable X1 (content of the addition of polystyrene granulate) was very small and amounted to -0.02. This indicates a weak correlation of this variable with the dependent variable
λ. On the other hand, the variable X2 (content of the foaming agent) is highly correlated and explains 82% of the variation of the coefficient λ.
A high semi-molecular correlation of both independent variables (X1 and X2) was observed for the density, compressive and flexural strength as well as sorption of foamed concrete.
The trends in the changes in the density of foamed concrete are presented in
Figure 9. The semi-partial correlations between the independent variables and the density are comparable and amount to -0.943 for the foaming agent content (X2) and -0.919 for the content of polystyrene granulate addition (X1). This means that the changes in the content of these components affect the change in the density of foamed concrete to a comparable extent (the strength of the interaction of both variables is comparable, but the variables are inversely correlated).
Figure 9.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the density of foamed concrete γ [kg/m3] described by a linear model.
Figure 9.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the density of foamed concrete γ [kg/m3] described by a linear model.
Figure 10 presents a surface graph of the variation of compressive strength of hardened foamed concrete. During the redundancy test, it was found that the semi-partial correlations between the independent variables and density are diverse and amount to -0.871 for the foaming agent content (X2), and -0.767 for the polystyrene granulate addition content (X1). This means that the effect of the change in the amount of the foaming agent on the change in the compressive strength of foamed concrete is greater (the strength of the effect of this variable is greater than the change in the content of the polystyrene granulate additive). However, the variables are inversely correlated.
Figure 10.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the compressive strength of foamed concrete fc [MPa] described by a linear model.
Figure 10.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the compressive strength of foamed concrete fc [MPa] described by a linear model.
For the study of the flexural strength variation, a strong correlation was shown with the change in the foaming agent content (X2) - level 95%. The impact of the polystyrene granulate addition content is insignificant, see
Figure 11. The trends of the simultaneous influence of the change in the amount of the foaming agent X2 and the addition of polystyrene granulate X1 on sorptivity are presented in
Figure 12.
Figure 11.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the flexural strength of foamed concrete ftk [MPa] described by a linear model.
Figure 11.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the flexural strength of foamed concrete ftk [MPa] described by a linear model.
Figure 12.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the sorptivity nk [MPa] described by a linear model.
Figure 12.
The effect of the simultaneous addition of polystyrene granulate X1 [%] and foaming agent X2 [%] on the sorptivity nk [MPa] described by a linear model.
Analyzing the partial correlations, it can be stated that both variables are similarly correlated. This means a strong relationship between both independent variables X1 and X2 with sorptivity. Wherein the variable content of the foaming agent is inversely correlated.