INVESTIGATION THE EFFECTS OF GREEN MORTAR PARAMETERS USING ANALYTICAL METHODS

The various green mortar mixes have been used in this study using various percentages of waste glass powder (WGP), steel slag (SG) and Micro-silica fume(SF). The different properties of flow, density, ultrasonic pulse velocity (UPV), compressive and flexural strengths have been tested for such green mortar in the first phase of experimental work. The second phase deals with the regression analysis of such properties. Whereas, the analysis of the results have also been using the integrated AHP and TOPSIS methods for selection the best performance of the green mortar due to the ecological effects of such materials. The results showed that the use of 70%OPC+8%WGP+12%SG+10% SF indicated as the best performance in term of ecological impact compared with other mortar mixes. Also, the regression analysis using the integrated AHP and TOPSIS methods gives an effective strategy for the selection of the best mortar mix.


INTRODUCTION
Cement considers as most produced materials in the world. The annual global cement production ranges to 4.15 billion tons in 2016, and is expected to increase to 4.25 billion tons per year in 2030 (1). The production of this material needs the use of a huge amount of raw materials, energy and fossil fuels in addition to air and water (2,3,4). The pollutants generated and non-renewable resources consumed during cement industry make cement material has a negative impact on the environment (4). Although this industry causes the formation of wastewater, solid waste, and noise, the main environmental issues are associated with air emissions and energy consumption. The high amounts of the carbon dioxide (CO 2 ), nitrogen oxides (NO, NO 2 , N 2 O), sulfur oxides (SO2, SO3) and dust emissions in addition to the other air pollutants are released from cement manufacturing (5,6.7,8). Approximately 8% of global carbon dioxide (CO 2 ) emission is liberated from cement industry (8). Moreover, the production of one ton of cement releases 360 pounds of dust (9),requires about 1.597 metric ton of raw materials (9,10) and consumes a high amounts of electricity and thermal energy. Another various industrial processes (silicon metal, ferrosilicon and steel) also have a significant role on the environment impact. The accumulation of solid wastes generated as a byproduct of these industry is one of the reasons which lead to deteriorate the environment. In 2016, the global crude steel, silicon and ferrosilicon production has estimated at 1630, 2.7 and 6.4 million tons respectively (11). Production of one ton of the steel, silicon and ferrosilicon generates a high amount of solid waste like slag and silica fume dust respectively (12). Besides the accumulation of the non-biodegradable solid waste like waste glass in the landfills is the one of phenomena which has negative impact on the environ-ment. For reducing this environmental impact, reducing of raw materials and energy consumption during cement production, producing sustainable building material, saving in cement and recycling of waste products, many materials were blended with cement to make green building materials (13)(14)(15)(16). Waste glass powder (WGP), Steel slag (SG) and Silica fume (SF) are non-biodegradable materials and available as waste materials (17)(18)(19). Utilization these materials with cement can improve mortar flowability, early and long term strength and long term durability (20)(21)(22). Physical properties and chemical composition of these materials have a significant role on the properties of concrete. Many previous studies (23)(24)(25)(26) have concluded that the chemical composition and the particle size of WGP have governed its pozzolanic activity, smaller particles decrease alkali silica reaction and give higher strength. Several researchers (22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) have focused on the use of SG with cement to improve the density ,durability and strength of concrete at later ages. Many researchers (28)(29)(30)(31)(32) have studied the properties of mortar or/and concrete prepared from blended cement with supplementary cementitious materials (SCMs).Silica fume is the one of SCMs which can contribute to improve the mechanical properties of mortar and concrete at identical water-binder ratio (W/b) and replacement level (33)(34)(35). Several researchers have worked to improve accurate building material properties and predict new models. Multiple regression analysis is the one of statistical techniques which have been utilized to analysis properties of concrete and mortar, predict the good relationship between concrete or mortar properties and produce generalized results for new concrete or mortar before test. Therefore, the objectives of this research are to study the O N L I N E F I R S T multiple regression analysis for green mortar properties and to selecting the best green mortar based on its impact on the environment.

Materials
The physical and mechanical properties of Ordinary Portland Cement(OPC), waste glass powder (WGP) steel slag (SG) and silica fume (SF) are given in Table  1. The properties of superplasticizer are shown in Table  2. Tap water was utilized in all mixes. The sand used for preparing green mixes was prepared according to ASTM C778 (36). Superplasticizer was procured from Specialties construction chemicals factory, Jahra, Kuwait. Its commercial name is KUT PLAST PCE 600. It was used to enhance the workability of mixes.

Mix proportions
In this study, nineteen (19) mixes were prepared to produce green mortars, binder/sand ratio was 1:2.75, water + superplasticizer / binder ratio (W+SP/B) was 0.39, the mix proportions of these mixes are listed in Table 3.

METHODOLOGY Experimental method
Flow test was conducted for each mix according to ASTM C1437 (37

Collection of the estimated data resulting from industrials processes and computing of the amount of the resources consumed and pollutants generated from cement industry
The estimated data of solid waste accumulated, the resources consumed and the emissions resulted from cement industry Based on the evaluated data in 2016, the global cement production was reached to 4150 million tons (1). Raw materials such as Limestone, clay, Sand and iron ore are the main requirements for cement production and are consumed in the large quantity (42). According to previous studies (9,10), the production of the one ton of cement requires about 1597 kg of raw materials. Therefor the global amount of raw materials consumed in 2016 were evaluated at 6628 million tons in this study. During cement manufacturing, the electricity and thermal energy are used in sufficient level. The thermal energy is used during cement processing while electricity energy is used for grinding of raw materials of cement and clinker (43,44). The global amount of electricity energy and thermal energy consumed in 2016 was estimated at 91KWh and 3400 mega joules per one ton of cement and clinker produced respectively and the clinker to cement ratio was estimated at 0.66 (1). Furthermore, the global amount of thermal energy consumed per one ton of cement produced was calculated at 2244 mega joules in this research. Cement industry not only consumes non-renewable resources like raw materials, thermal energy and electricity energy but also releases air pollutants like CO 2 and dust which have a negative impact on the environment and human health (79). About fifty percent (50%) of Carbon dioxide (CO 2 ) released during cement industry is resulted from calcination process of limestone (CaCO 3 ) and another fifty percent (50%) is resulted from fuel combustion in cement kilns (35). In 2016, the global amount of carbon dioxide (CO 2 ) emitted per one tone of cement produced was reached to 530 kg (1). Thus, the total amount of the global carbon dioxide (CO 2 ) emitted from the total amount of cement produced in the same year was estimated at 2200 million ton. In addition, the global amount of dust emitted during cement industry was also estimated at 676 million ton in this research based on the production of the one ton of cement releases approximately 163 Kg of dust (9).   (1,9,(43)(44) O N L I N E F I R S T silicon and ferrosilicon production had evaluated at 1630, 2.7 and 6.4 million tons respectively in 2016 (11). Approximately, 400 kg of steel slag, silica fume has produced as a byproduct per one ton of steel, silicone and ferrosilicon manufactured respectively (12) Therefore, the total amount of steel slag, silica fume in 2016 was estimated at 652, 3.64 million tons in this study. In addition, the global amount of the municipal solid waste was estimated to 2017 million tons in the same year (48). And the waste glass powder had formed about 5% of the municipal solid waste (48).

The calculation of the amount of the resources consumed and air contaminants resulting from cement used in the preparation of the traditional mortar and non-reinforced green mortars
The quantity of non-renewable resources (raw materials, electricity and thermal energy) consumed and air pollutants (carbon dioxide and dust) emitted during cement industry was calculated in this study. The calculation was based on the amount of cement used for preparation of the traditional mortar and non-reinforced green mortars. The data in Table 5 was used in the calculation of the total quantities of raw materials, electricity energy, thermal energy, CO 2 and dust resulting from cement used in the conventional mortar and non-reinforced green mortar. In order to reduce the amount of cement used in the preparation of mortars, the solid waste (waste glass, steel slag and silica fume) was used as partial replacement of cement as listed Table 5. This table illustrated the amount of resources consumed and air pollutants generated from cement used to prepare the conventional mortar and non-reinforced green mortar.

RESULTS AND DISCUSSION
The results for green mortar mixes are shown in Table  6. The results showed that the acceptable ranges of strengths and other properties can also be obtained using 30% replacements of cement.

Regression analysis
For analysis of green mortar data, Multivariable linear regression analysis (MLRA) was employed. The purpose of this analysis is to explain the relationship between one dependent variable and two or more independent variables.  Table 4 (1,9, 43-44)  Where, Y is dependent variable; C is constant; b 1 , b 2 and b n are slopes associated with X a , X b and X n respectively. X a , X b and X n are independent variables and e is error. For prediction of the strength of the green mortar before preparation it, the strength was considered as dependent variable, while the proportion of WGP, SG, SF, SIF, HHF, SNF, UPV, Density and curing time were considered as independent variables in this study. The enter techniques in SPSS program were used to create the regression models. The data analysis was done using MLRA. The number of important statistical parameters were associated with the MLRA. Some of these parameters were coefficient of regression determination, model error, the significance level, the confidence level, the t-distribution and the F-distribution. Detailed explanations of these important parameters can be found in previous studies (81). As a final approach; regression models were created to predict the strength of green mortars. The statistical parameters of regression models were calculated at the 95% confidence level. Summary of regression models were shown in Table 7. Predictive models of the strength of green mortar were given below: Where, CS and FS is compressive strength (MPa) and flexural strength (MPa) of green mortar respectively, UPV and DS is the ultrasonic pulse velocity (km/sec) and density of green mortar respectively, WGP is the content of waste glass powder (%), SG is the content of steel slag (%), SF is the content of silica fume (%). To assess the validity of the predictive models mentioned above; the behavior of correlation (R), determination of coefficient (R2), the t-test, the F-test and Multicollinearity-test were considered. The statistical parameters of regression models of green mortar were shown in Table 7.

Model
No.  value for the correlation coefficient of predictive model does not necessarily to express the good performance of model (49). The value of determination coefficient (R2) does not establish the validity of predictive model unless the results of test for significance of regression, t-test and Multicollinearity-test show the consistency between the experimental results and predictive model. Therefore, test for significance of regression, t-test and Multicollinearity-test were considered. Test for significance of regression was carried out using analysis of variance (F-test).This test was helped to determine whether the regression line was the most suitable curve in representing the relationship between the sample data sets of two correlated variables (48). The null hypothesis was designated Ho=0, which means that no correlation exists between the two variables tested using analysis of variance. The analysis of variance produced two values for each model: an F-value, which indicates the degree to which the regression equation fits the data, and a second value that indicates the statistical significance of the F-value. In the case that the statistical significance of the F-value was less than 0.05 at the 95% confidence level, Ho=0 was refused, meaning that the relationship between depended variable and the target independent variable could be expressed as a linear or non-linear equation at the 95% confidence level (49). Otherwise, it was assumed that the relationship could not be represented as a regression model (50). Therefore, all the predictive models in Table 7 are considered to be valid due to the significance of the F-value was less than 0.05 for all these models. The t-test was used to examine the significance of the variables in each model at the 95% confidence level. By considering the degrees of freedom for each variable, a t-value calculated for the experimental data can be compared with a tabulated t-value. In case that the calculated t-value is greater than the tabulated value, the significance of t-values was less than 0.05 at the 95% confidence level and the variable is considered to be significant to the model (50).The t-significance value of all predictive models for green mortars (except Model 1, 3 and 6) was less than 0.05at the 95% confidence level.  Table 7. Multicollinearity is a statistical phenomenon in which there exists a perfect or exact relationship between the predictor variables. In the other words, it means two or more of the independent variables in a multiple regression model are highly correlated (25,34). This causes a problem in the interpretation of the regression results. Multicollinearity was tested using Tolerance(T) and the variance inflation factor (VIF). Tolerance is the amount of variability in one independent variable that is no explained by the other independent variables (56,63). Tolerance values less than 0.10 indicate to the presence of multicollinearity. The variance inflation factor (VIF)is defined as the inverse of Tolerance (1/T). In case that the (VIF) is more than 5, the multicollinearity has been presented. Tolerance values of all predictive models of green mortars were more than 0.1. Table 7 were less than 5, except Model-2 of green mortar. For this model, the variance inflation factor of UPV and Density was 6.085 and 6.085 respectively. Therefore, Model-2 of green mortar was considered invalid because of multicollinearity problems.

AHP and TOPSIS methodology Analytic Hierarchy Process (AHP)
The Analytic Hierarchy Process (AHP) is a multi-criteria decision making approach which organizes and analyzes complex decisions (50). It was developed by Thomas Saaty in the 1970, 1980. The decision problem in this approach is arranged in hierarchic structure (51). The arrangement is in the descending form from an overall goal to criteria, sub-criteria and alternatives in successive levels (52). In this paper, AHP was used to evaluate the weight of the thirteen (13) criteria of reinforced green mortars and non-reinforced green mortars. Five (5) of them were compressive strength, Flexural strength, UPV and Cost of reinforced green mortars and another eight (8) were Raw materials, Electricity energy, Thermal energy, CO 2 , Dust, Waste glass, Steel slag and Silica fume of green mortars. The criteria weights of green mortars were computed using the following general steps: Step 1: Conduct the comparison for two criteria at the same time with respect their impact on the mortar prepared. The comparison conducts based on the one common scale (adapted from Satty) that is displayed in Table  8. to build the Pair-wise comparison matrix (F). The Pair-wise comparison matrix (F) builds by asking questions to experts or decision makers like, which criterion is more important with regards to the decision goal. The answers to these questions will construct the matrix (F) as shown below:

Definition
where fij represents a quantified judgment on Ci/Cj with fii = 1 and fij = 1/fji for i, j = 1, . . .,m Step 2: Compute the sum ∑ i 4 =1 fij for each column in matrix (F). Then, divide (Fij) to computed sum according to Eq.1, the result will be matrix (X): Step 3: Calculate the average of each raw in the matrix (X) to obtain the weight (w) of each criterion.
Step 4: Check the consistency of the pairwise comparison matrix (F) by using the following steps: a. Construct matrix (Y) by multiplying the criterion weight (w) with pairwise comparison matrix (F).
Compute the consistency index (CI) according to Eq. 2.
The random consistency index (RCI) is obtained using

Technique for Order Performance by Similarity to Ideal Solution (TOPSIS)
TOPSIS is a simple and useful approach which is used to deal with the complex system related to making a best choice among several alternatives (54). It was developed by Ching-Lai Hwang and Yoon in 1981. The concept of this technique is based on the selecting the ideal alternative which has the shortest distance from the positive ideal solution and the farthest distance from the negative ideal solution (55). In this research, TOPSIS was used to select and rank the best of reinforced green mortar based on its properties. In addition this technique was also used to determine and rank the best and worst non-reinforced green mortar based on its impact on the environment. The ranking and determining of reinforced and non-reinforced green mortars was achieved using the following steps: Step 1: Construct the decision matrix (N) for ranking of the alternatives, the structure of matrix can be expressed as follow: where Ai represents the alternatives i, i = 1…., m; Cj represents the criteria that are required on which the alternative is judged, j=1…, n; Zij represents jth attribute, j=1, n related to i, the alternative; and Zij is the obtained value representing the performance rating of each alternative Ai with respect to each requirement Cj.
Step 2: Calculate the normalized decision matrix (V): The raw data can be normalized by utilizing Eq. 3. to produce the matrix (V). where i=1, 2, 3,...,m and j=1, 2, 3,…,n (3) Step 3: Compute the weighted normalized decision matrix (B) by multiplying the weights of criteria (w) with the normalized decision matrix (V). In this paper, the weights of criteria (w) were previously calculated based on AHP method.
Step 5: Compute the distance of all alternatives to the positive and negative ideal reference point (D+ and D-) by using Eq. 4 and Eq.5 respectively.
Step 6: Calculate the relative closeness coefficient (R) of each alternative to the ideal reference point by using Eq. 4. Then, conduct the outranking of the alternatives in descending order. The larger value of Ri indicates to the better performance of the alternative.

Performance of the regression models to predict strength parameters
The strength of green mortar can be predicted using regression models. For prediction the strength of such green mortar, it is possible to use Model-4, Model-5, Model-7 and Model-8 in Table 10. The comparison between the obtained strength from the experimental results and regression models was illustrated in Fig. 1.
The statistical parameters of valid regression models are shown in Table 10.

Impact assessment of the traditional mortar and green mortars on the environment
In order to assess the impact of the traditional mortar (it has not contained a partial replacement of cement) and green mortar on the environment, the integrated AHP and TOPSIS method was used. Based on the non-renewable resource consumed, air pollutants emitted during cement industry and the solid waste used as replacement of cement, the environmental impact of mortars was assessed in this study. The non-renewable resources consumed were raw materials, electricity energy and thermal energy required cement industry. Air pollutants were carbon dioxide and dust emitted during cement production. The solid wastes used as a partial replacement of cement were: waste glass, steel slag and silica fume. Therefore the eight criteria were considered. The evaluated weight of each criterion based on AHP method was listed in Table 11. In order to check the pair-wise comparison matrix, that was shown in Table 12. According to AHP method, the weight of these criteria (Raw materials, electricity energy, Thermal energy, CO2, Dust, Waste glass, Steel slag, and silica fume) was computed. The matrices, that are obtained through applying of AHP method, were shown in Table 13., the step 4 that was mentioned previously in the general steps of AHP method, was applied as shown in Table 14. Therefore, the calculated consistency index was 0.082041 and the random consistency index obtained from

CONCLUSIONS
The study shows very interesting analytical equations that indicate significant relationships between each of compressive and flexural strengths for green mortar with different parameters used in the production of green mortar mixes. Such parameters included the age of the mortar, density, UPV, SF, WGP, SG and SF. The statistical parameters of regression models were calculated at the 95% confidence level. Besides, based on integrated AHP and TOPSIS method have shown dramatic meth-ods or the selection of the most important mortar mix that exhibit the best performance in ecological effects. Thus, the mix M19 which contains 70%OPC+8%WG-P+12%SG+10% SF is classified as the best green mortar and the control mortar mix is the worst green mortar in terms of their impact on the environment.