265 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) ISSN (Print) 2313-4410, ISSN (Online) 2313-4402 © Global Society of Scientific Research and Researchers http://asrjetsjournal.org/ Determination of Appropriate Mix Proportions for the Kenyan Blended Portland Cement Concrete Production Victoria A. Okumua*, Stanley M. Shitoteb, Walter O. Oyawac aPan African University, Institute of Basic Sciences, Technology and Innovations (PAUISTI), Hosted at Jomo Kenyatta University of Agriculture and Technology (JKUAT), P.O. BOX 62,000-00200, NAIROBI, KENYA. bRongo University, P.O. BOX 103-40404, RONGO, KENYA. cCommission of University Education, P.O. BOX 54999-00200, NAIROBI, KENYA. aEmail: vokumu@gmail.com, bEmail: shitote@hotmail.com, cEmail: oyawaw@yahoo.com Abstract The Kenya’s vision 2030 seeks to address the rising needs of its population through infrastructure development. Reinforced concrete being the most commonly used construction material forms an integral part of this development strategy. The direct substitution of the ordinary Portland cements with the cheaper, lower strength, locally available blended Portland cements could be responsible for the production of poor quality concrete and contribute to the failure of several concrete buildings in the country. This paper presents findings of an experimental investigation on the appropriate mix proportions for the Kenyan blended Portland cement concrete. Key variables used in this study included the water/ cement ratio (x1), the cement/ total aggregates ratio (x2) and the fine aggregates/ coarse aggregates ratio (x3). The response was measured in terms of slump, compressive strengths at 7days, 14days and 28 days and density. Minitab 17 software was used in the design of experiments and results analysis based on Central Composite Design method. The investigation revealed that for a workable concrete with slump of ≥ 30mm, the appropriate mix ratios for the Kenyan blended Portland cement concrete are: 1:2.2:3.4 (w/c 0.6) for strength class C15 and 1:1.3:2.2 (w/c 0.5) for strength class C20. It was further noted that the different brands of blended Portland cement in the country had varying properties and thus produced concrete with different wet and hardened properties. None of the brands achieved the target design strength for strength class C25 and above. Therefore, the blended Portland cements may not be suitable for producing structural concrete strength class C 25 and above. Keywords: Appropriate mix ratios; Blended Portland cement; blended Portland cement concrete; central composite design; concrete strength class; target design strength. ------------------------------------------------------------------------ * Corresponding author. http://asrjetsjournal.org/ American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 266 1. Introduction The construction industry influences the social-economic development of any nation. At present, the Kenyan urban housing sector is characterized by inadequacy of affordable and decent housing, low level of urban home ownership, extensive and inappropriate dwelling units including slums and squatter settlements. Informal settlements house 60% of the urban population. To satisfy the urban housing needs, it is estimated that a total 200,000 housing units are required annually, yet only an estimated 35,000 units are produced [1]. Developers seek to meet this ever increasing demand for decent housing through constructions that include reinforced concrete residential buildings. However, in most cases, no difference is made in cement strengths resulting in the use of same mix proportions irrespective of the cement type and strength. Further, quality assurance/control mechanisms are often overlooked and so the quality of concrete produced may not be as designed. The inappropriate mix ratios, coupled with lack of trial mixes leads to production of concrete that do not meet the designed target strengths [2, 3]. This scenario is however different for the few developers (public and private) who employ qualified professionals to design, construct and supervise their buildings. Concrete mix design can be defined as the science of correct proportioning of concrete ingredients based on project requirements, to obtain the desired properties in plastic/wet as well in hardened stage [4]. Research has shown that the strength properties and other qualities of concrete depend on the mix design proportions; the type, content, and properties of ingredient materials, method of compaction, placing and curing [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]. When properly designed, mixed, placed, compacted and cured, concrete has good compression resistance and durability [6]. In the recent past, several reinforced concrete residential and commercial buildings have collapsed during construction and usage. This failure has been attributed to the poor quality of in situ concrete and the concrete technologies being implemented. [3, 5, 18]. All components within the mix design must be selected in such a way that the required properties of the final product are retained after the concrete mixture hardens [6]. In the fresh/ plastic state, workability is specified as the most important property while in the hardened state, compressive strength, density and durability are considered as the most important properties. The main parameters affecting the design of a concrete mixture are: type of cement, water/cement ratio, coarse aggregate/total aggregate ratio and total aggregate/cement ratio [7, 19]. Blended Portland cements exhibit a slower setting time and lower early strength development [20]. These cements however are the most commonly used type in Kenya and other developing countries due to their cheaper costs resulting from the local availability of the natural deposits of the pozzolanic materials used in their manufacture. Cement type and content, aggregate type and properties, age and curing conditions have also been reported to have a great effect on concrete strengths and durability [21]. Research has revealed that different mix design methods calculate the target mean strength and constituent ingredients mix proportions differently [22]. This experimental research was undertaken to determine the appropriate concrete mix proportions for blended Portland cement concrete production in Kenya. Locally available blended Portland cement concrete constituent materials were used during the study. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 267 2. Materials and Methods 2.1. Materials Properties 2.1.1. Aggregates Crushed stone aggregates from the Mlolongo quarry and ordinary river sand from the banks of river Ewaso Nyiro in Meru were used as coarse aggregates (CA) and fine aggregates (FA) respectively in the manufacture of concrete. The suitability of the aggregates for concrete production was ascertained through particle distribution in accordance to BS EN 1097-6-2013; and tests on their physical properties determined following the laid down procedures in their respective British standards: Specific gravity (BS 812-102:1995), Bulk density (BS 812- 2:1995), Water Absorption (BS 813-2:1995) and moisture content (BS 812-109:1990). The results were as summarized in Table 1 for fine aggregates and Table 2 for coarse aggregates. Table 1: Fine Aggregates Physical Properties SEIVE DESIGNATION WEIGT OF AGG. RETAINED % WEIGHT RETAINED CUMMULATIVE % RETAINED % PASSING mm g % % % 10 0 0 0 100.0 5 7 0.70 0.70 99.3 2.36 16.5 1.66 2.36 97.6 1.18 103 10.36 12.72 87.3 0.6 372.5 37.46 50.18 49.8 0.3 261 26.24 76.42 23.6 0.15 221 22.22 98.64 1.4 pan 13.5 1.36 100.00 0.0 Total 994.5 Physical Properties FM=2.41 Grading Zone II Fineness Modulus 2.41 Specific Gravity 2.63 water absorption 0.91% Moisture Content 0.73% Bulk Density 1564kg/m3 The aggregates physical and mechanical properties tests were done at the Jomo Kenyatta University of Agriculture and Technology (JKUAT) Civil Engineering laboratories. The results in Table 1 and Table 2 show that the aggregates were suitable for concrete production. The Fine aggregates grading curve was within Zone II envelope of the British standard and the fineness modulus was 2.41 which was within the recommended range of 2.0-4.0 [18]. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 268 Table 2: Coarse Aggregates Physical Properties SEIVE DESIGNATION (SIZE) Weight Of Agg. Retained Cumulative Wt. Retained % Weight Retained % Weight Passing mm g g % % 50 0 0 0.00 100.00 38.1 0 0 0.00 100.00 20 590 590 59.00 41.00 10 380 970 97.00 3.00 5 21 991 99.10 0.90 pan 9 1000 100.00 0.00 Physical Properties FM=2.551 AIV 12.06 ACV 22.27 Fineness Modulus 2.55 Specific Gravity 2.5 water absorption 1.25% Moisture Content 5.78% Bulk Density 1448kg/m3 2.1.2 Cement Portland Pozzolana Cement (PPC) is a type of Blended Cement which is produced by either inter-grinding Ordinary Portland Cement (OPC) clinker along with gypsum and pozzolanic materials in certain proportions, or grinding the OPC clinker, gypsum and pozzolanic materials separately and thoroughly blending them in certain proportions when producing concrete. Constituent materials that are permitted in blended Portland cements are artificial pozzolans (blast furnace slag, silica fume, and fly ashes) or natural pozzolans (siliceous or siliceous aluminous materials such as volcanic ash glasses, calcined clays and shale). In Kenya, Lime and natural pozzolanic materials such as volcanic ashes, tuffs and diatomaceous earths deposits are commonly used in the manufacture of blended Portland cements. The cement is produced in accordance to KS EAS 18-1: 2001 standard which is an adoption of the European Norm EN 197 cement standards [23]. The cements produced are blended cements in which cement replacement materials are added to the clinker at the time of grinding. The cements readily available in the Kenyan market are Portland Pozzolanic Cement (PPC) CEM II/B-P containing 21-35% natural pozzolana, Pozzolanic Cement (PC) CEM IV/A with 11-35% pozzolanic material, and Portland Limestone Cement (PLC) CEM II/A-LL with 6-20% limestone addition. A limited quantity of Ordinary Portland cement (OPC) CEM I is produced for specific uses [18], [24]. Normal cements are denoted N while rapid strength development cements are denoted R. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 269 Currently there are six cement manufacturing companies in Kenya producing different brands of blended Portland cements. The six companies have been coded in this study as company A to F. Company A and company D each produce two brands of blended Portland cements coded as A1 and A2 and D1 and D2 respectively. The cements setting time, compressive strength and consistency tests were based on BS EN196:2010 while the fineness tests were done using the Blaine apparatus and the 32 Micron residue methods. The tests were done at the Kenya Bureau of Standards Laboratories and the properties of the locally available blended Portland cements have been summarized in Tables 3. Table 3: Blended Portland Cement Physical and Mechanical Properties BLENDED PORTLAND CEMENT TYPE CEMENT FINENESS COMPRESSIVE STRENGTH (MPa) CONSISTENCY AND SETTING TIME (MIN) Cement Type Cement Brand Residue (32 Microns) % Blaine (Cm2/g) 2 Days 7 Days 28 Days Consiste ncy Initial Setting Time Final Setting Time CEM II/B- L 32.5R CEM A1 17.41 3856 20.2 31.4 46.9 27.6 182 251 CEM IV/B- P 32.5N CEM A2 16.58 3935 13.6 23.4 37.6 33.3 200 295 CEM IV/B- P 32.5N CEM B 17.55 4471 12.1 23.9 32.6 34.9 230 319 CEM IV/B- P 32.5R CEM C 21.98 4063 21.1 35.3 45.5 31.5 197 270 CEM II/B- P 32.5N CEM D1 22.98 3191 13.2 26.6 43.8 29.7 251 393 CEM II/B- P 32.5N CEM D2 21.86 3451 13.5 28.0 39.3 30.9 208 292 CEM II/B- P 32.5R CEM E 28.03 3034 10.3 24.9 40.1 30.56 201 290 CEM IV/B- P 32.5N CEMF 27.38 3918 14.0 25.0 32.3 30.2 215 319 The results indicate that other than CEM F which had a lower value of compressive strength <32.5MPa, and CEM B which exceeded the minimum requirement by only 0.1MPa, all the other brands of blended Portland cements met the requirements as stated in the KS EAS 18-1: 2001 Standard. CEM C had the best combination of ultimate compressive strength and fineness and thus was used during the study to develop the mix design proportions since cement strength and fineness influence the strength development of concrete. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 270 2.1.3 Water Tap water from Jomo Kenyatta University of Agriculture and Technology water treatment plant was used during the study in the mixing of concrete and curing of all the concrete specimens. 2.2. Design of Experiments and experimentation Central Composite Design (CCD) is a classified design for Response Surface Method (RSM) which is especially useful in sequential experiments because it is built on previous factorial experiments by adding axial and center points. CCD enables estimation of the regression parameters for second- order polynomial regression model for the response. They consist of cube points, center points and axial points. A factorial or fractional factorial design (2k or 2k-1 factorial points, where k is the number of factors) allow for the estimation of linear and interaction effects. Center points are used to check for curvature while axial (or star) points are used to estimate quadratic terms. Alpha (α) for axial points is the distance of each axial point from the center calculated by 42 k =α [25, 26, 27]. In this study, CCD was used to determine a quadratic response surface which has curvature and to predict factor levels that produce maximum or minimum response values for the composite material concrete. MINITAB 17 software was used to generate the concrete mixture proportions for experiments based on the Central Composite Design (CCD) method. Three variables namely; (i) Water/ Cement ratio as x1, [0.4, 0.5, 0.5], (ii) Cement / Total aggregates ratio as x2 [0.18, 0.22, 0.26] and (iii) Fine Aggregate / Coarse aggregates ratio as x3 [0.56, 0.6, 0.64] were used. The variables were mixed randomly to obtain a full factorial design at three levels and repeated three times yielding a total of 60 runs with 20 base factorial points, 24 cube points, 12 center points, 18 axial/ star points, 6 center points on the axial points and 3 blocks. The generated mixtures were then cast and tested experimentally and the response evaluated in terms of Slump as Y1, 7 days compressive strength as Y2, 14 Days compressive strength as Y3, 28 days compressive strength as Y4 and Density as Y5. The results were as shown in Table 4. Table 4: Design of Experiment based on CCD and Results of the experiments Run Order Pt Type Blocks x1 x2 x3 Y1 (mm) Y2 (MPa) Y3 (MPa) Y4 (MPa) Y5 (kg/m3) 1 -1 3 0.5 0.28532 0.6 119 17.6843 20.6667 24.8307 2426 2 0 3 0.5 0.22 0.6 5 25.3947 28.2600 32.8633 2466 3 -1 3 0.3367 0.22 0.6 0 31.8893 33.7037 37.5653 2425 4 0 3 0.5 0.22 0.6 60 17.7593 21.1257 26.6777 2436 5 -1 3 0.5 0.22 0.53468 20 19.4903 23.3520 25.6027 2432 6 -1 3 0.5 0.22 0.66532 16 24.7543 28.0393 31.9137 2471 7 -1 3 0.6633 0.22 0.6 178 9.5227 12.1453 13.8743 2403 8 0 3 0.5 0.22 0.6 22 20.3303 24.6477 28.3687 2451 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 271 9 -1 3 0.3367 0.22 0.6 0 14.7980 16.8950 22.2483 2297 10 -1 3 0.6633 0.22 0.6 145 10.0233 13.6190 16.8733 2395 11 -1 3 0.5 0.15468 0.6 0 22.1967 26.9337 30.7127 2449 12 -1 3 0.5 0.22 0.53468 46 18.1250 22.9447 26.9840 2437 13 -1 3 0.5 0.28532 0.6 79 21.5290 26.7977 29.3450 2418 14 -1 3 0.5 0.22 0.53468 25 21.3067 25.4707 29.0913 2479 15 0 3 0.5 0.22 0.6 40 22.5180 27.0080 28.6100 2442 16 0 3 0.5 0.22 0.6 34 20.9717 26.2637 28.3360 2449 17 -1 3 0.5 0.22 0.66532 17 24.2600 27.0837 31.2523 2469 18 -1 3 0.5 0.15468 0.6 0 27.6507 32.0383 32.9640 2475 19 -1 3 0.5 0.22 0.66532 19 22.2050 26.2387 29.2247 2451 20 -1 3 0.5 0.15468 0.6 0 22.5987 25.9257 28.5010 2507 21 -1 3 0.5 0.28532 0.6 48 22.5260 26.4633 28.9590 2165 22 0 3 0.5 0.22 0.6 36 18.1237 23.6153 26.4823 2483 23 -1 3 0.3367 0.22 0.6 0 21.0787 34.3950 40.2660 2405 24 -1 3 0.6633 0.22 0.6 127 11.7603 14.1250 16.2230 2419 25 0 1 0.5 0.22 0.6 19 20.0800 25.3320 28.0543 2459 26 1 1 0.4 0.26 0.64 0 30.0670 36.2663 40.0590 2471 27 1 1 0.6 0.26 0.56 179 13.3357 16.4277 20.3050 2329 28 1 1 0.6 0.18 0.64 54 18.0570 19.8903 24.6177 2458 29 1 1 0.4 0.18 0.56 0 37.0520 38.8603 45.0367 2497 30 1 1 0.4 0.26 0.64 0 35.4320 41.7900 44.2050 2491 31 0 1 0.5 0.22 0.6 29 17.7777 22.1710 26.6133 2426 32 0 1 0.5 0.22 0.6 24 21.4933 25.6683 31.0600 2444 33 0 1 0.5 0.22 0.6 26 22.2083 28.6603 31.9413 2437 34 1 1 0.6 0.18 0.64 59 17.0493 18.9420 22.9697 2458 35 0 1 0.5 0.22 0.6 15 21.6350 26.5200 30.2403 2469 36 1 1 0.6 0.18 0.64 47 20.8780 24.3963 25.9997 2463 37 1 1 0.6 0.26 0.56 162 14.2423 17.6640 22.1520 2404 38 1 1 0.4 0.18 0.56 0 31.3223 34.2660 37.1040 2484 39 1 1 0.4 0.26 0.64 0 34.0237 37.1100 41.8920 2504 40 1 1 0.4 0.18 0.56 0 31.1180 37.7493 40.6843 2484 41 1 1 0.6 0.26 0.56 171 16.4057 19.1530 20.7490 2451 42 0 1 0.5 0.22 0.6 16 21.9257 27.7750 30.0290 2458 43 0 2 0.5 0.22 0.6 9 25.3853 30.0360 31.4287 2467 44 0 2 0.5 0.22 0.6 16 23.3377 26.7687 30.4777 2465 45 1 2 0.6 0.26 0.64 189 13.3767 15.6147 19.7300 2431 46 0 2 0.5 0.22 0.6 15 25.0243 28.9647 31.9650 2489 47 0 2 0.5 0.22 0.6 11 24.5367 28.6610 31.9650 2474 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 272 48 1 2 0.6 0.18 0.56 39 14.5397 18.0910 21.4490 2411 49 1 2 0.4 0.26 0.56 0 29.8100 34.5507 40.1690 2502 50 1 2 0.6 0.26 0.64 173 14.0673 17.2430 21.1427 2377 51 1 2 0.6 0.18 0.56 92 12.2783 15.1760 18.6677 2389 52 1 2 0.6 0.26 0.64 192 11.3460 15.3810 16.5090 2372 53 1 2 0.4 0.18 0.64 0 32.2370 34.1743 39.7800 2457 54 1 2 0.4 0.26 0.56 2 32.5597 36.6003 45.5950 2488 55 0 2 0.5 0.22 0.6 15 25.7463 27.8867 32.0750 2467 56 1 2 0.4 0.26 0.56 9 29.9810 33.5277 40.3447 2466 57 1 2 0.4 0.18 0.64 0 32.7410 36.2083 38.9030 2471 58 0 2 0.5 0.22 0.6 12 18.1333 23.4363 29.1447 2459 59 1 2 0.4 0.18 0.64 0 30.7233 35.6497 39.7387 2463 60 1 2 0.6 0.18 0.56 53 15.9157 16.6587 20.6830 2451 2.3. Instrumentation and Testing Nine (9) 150mm by 150mm by 150mm concrete cubes were cast for each of the sixty (60) runs and slump test was used to evaluate the wet concrete response properties while three cubes were tested at 7, 14 and 28 days of curing each to evaluate the compressive strength development of the concrete and the density of the concrete. The compressive strength of concrete was investigated at 7, 14 and 28 days using the Universal Testing Machine with a loading capacity of 1500kN in accordance to BS 1881-116: 1983 as illustrated in Figure 1. (a) (b) (c) Figure 1: (a) Compressive strength testing machine, (b) slump test cones and (c) casted concrete cubes American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 273 3. Results and Discussion The analysis of the response based on the 60 runs of experiments carried out was done using Minitab 17 software. Each response was analyzed independently and the interaction effects of the various variables were also investigated. 3.1. The Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the Slump (Y1) The experiment resulted in slump values ranging from 0mm to 195 mm as shown in Table 4. The response was then analyzed to evaluate the influence of the different variables and their interactions on the slump of the concrete. The interaction effects of the different variables on the slump was also investigated and the results show that the interaction between the water /cement ratio (x1) and the cement/ total aggregates ratio (x2) had a significant effect in the slump while their interaction with the fine aggregates/ coarse aggregates ratio (x3) did not have a significant effect on the slump as shown in Figure 2. Figure 2: Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the Slump (Y1) 0.60.50.4 175 150 125 100 75 50 25 0 0.250.200.15 0.650.600.55 x1 Me an of Y1 x2 x3 Factorial plots for Y1 Fitted Means 300 150 0 0.650.600.55 0.250.200.15 300 150 0 0.60.50.4 300 150 0 x3 x2 * x1 x3 * x1 x1 * x2 x3 * x2 x1 * x3 x1 x2 * x3 x2 0.3367 0.5 0.6633 x1 0.15468 0.22 0.28532 x2 0.53468 0.6 0.66532 x3 Me an of Y1 Factorial plots for Y1 Fitted Means American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 274 The results were further analyzed to obtain the residual plots and the quadratic model for the effect of the three variables x1, x2 and x3 on the blended Portland cement slump and the results were as shown in Figure 3. From the results, as expected, slump was affected by both the water/ cement ratio and the cement/ total aggregates ratio in that in both cases, the higher the water/ cement ratio and cement content, the higher the slump and vice versa. The fine aggregates/ coarse aggregates ratio however had very minimum effect on the slump as the value of the slump remained almost constant at the different values of the aggregates ratios investigated. The contour plots for the three variables was then plotted to be used to derive the mix design ratios for the different values of slump for the blended Portland cement concrete production as shown in Figure 4. Figure 3: Residual plots and Model Building Report for the effect of the variables on the Slump Y1 20100-10-20 99 90 50 10 1 Residual Pe rc en t 200150100500 20 10 0 -10 -20 Fitted Value Re sid ua l 20100-10-20 16 12 8 4 0 Residual Fr eq ue nc y 605550454035302520151051 20 10 0 -10 -20 Observation Order Re sid ua l Normal Probability Plot Versus Fits Histogram Versus Order Residual Plots for Y1 Y1 = 1374.3 - 3541 X1 - 6274 X2 + 2350 X1^2 + 7117 X2^2 + 7874 X1*X2 Step Change Step P Final P 5 4 3 2 1 Add X2^2 Add X1^2 Add X1*X2 Add X2 Add X1 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1007550250 R-Squared(adjusted) % x3 x2 x1 6040200 Increase in R-Squared % x3 x2 x1 100500 R-Squared % X1: x1 X2: x2 X3: x3 Final Model Equation Model Building Sequence Displays the order in which terms were added or removed. Incremental Impact of X Variables Long bars represent Xs that contribute the most new information to the model. Each X Regressed on All Other Terms Gray bars represent Xs that do not help explain additional variation in Y. A gray bar represents an X variable not in the model. Multiple Regression for Y1 Model Building Report American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 275 Figure 4: The Interaction Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the Slump (Y1) 3.2. The Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the Early (7-Days) Compressive Strength (Y2) The experiment resulted in 7 days compressive strength values ranging from 9MPa to 34MPa as shown in Table 4. The response was then analyzed to evaluate the influence of the different variables and their interactions on the early strength gain of the blended Portland cement concrete as shown in Figure 5. From the results, the early strength gain was significantly affected by the water/ cement ratio (x1). The lower the water cement ratio, the higher the early compressive strength. The cement/ total aggregates ratio (x2) and the fine aggregates / coarse aggregates ratio (x3) on the other hand had a slight effect of the early strength gain. Between 0.2 to 0.25 cement / total aggregates ratio (x2), there was no effect on the 7 days strength while below 0.2, the strength increased with decrease in the ratio and above 0.25 the strength increased with increase in the ratio. There was, however, slight increase in strength with the increase in the fine aggregates/ coarse aggregates ratio (x3) as shown in Figure. 5. The interaction effects of the different variables on the 7 days compressive strength was also investigated and the results show that the interaction between the water /cement ratio and the cement/ total aggregates ratio and that of the cement/ aggregates ratio and the fine aggregates/ coarse aggregates ratio had an effect in the 7 days compressive strength while the interaction between the fine aggregates/ coarse aggregates ratio and the water x1 0.5 x2 0.22 x3 0.6 Hold Values x2*x1 0.60.50.4 0.28 0.24 0.20 0.16 x3*x1 0.60.50.4 0.650 0.625 0.600 0.575 0.550 x3*x2 0.280.240.200.16 0.650 0.625 0.600 0.575 0.550 > – – – – – – < 0 0 50 50 100 100 150 150 200 200 250 250 300 300 Y1 Contour Plots for Y1 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 276 cement ratio did not have a significant effect on the 7 days compressive strength as shown in Figure. 5. The results were then used to generate contour plots for the determination of the blended Portland cement concrete ratios as shown in Figure 6. Figure 5: Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the 7 Days Compressive Strength (Y2) Figure 6: The 7 days contour plots for the Interaction Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the 7 Days Compressive Strength (Y2) 3.3. The Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the Ultimate compressive strength at 28 days (Y4) 0.60.50.4 30 25 20 15 10 0.250.200.15 0.650.600.55 x1 M ea n of Y 2 x2 x3 Factorial plots for Y2 Fitted Means 30 20 10 0.650.600.55 0.250.200.15 30 20 10 0.60.50.4 30 20 10 x3 x2 * x1 x3 * x1 x1 * x2 x3 * x2 x1 * x3 x1 x2 * x3 x2 0.3367 0.5 0.6633 x1 0.15468 0.22 0.28532 x2 0.53468 0.6 0.66532 x3 M ea n of Y 2 Factorial plots for Y2 Fitted Means x1 0.5 x2 0.22 x3 0.6 Hold Values x2*x1 0.60.50.4 0.28 0.24 0.20 0.16 x3*x1 0.60.50.4 0.650 0.625 0.600 0.575 0.550 x3*x2 0.280.240.200.16 0.650 0.625 0.600 0.575 0.550 > – – – – – < 10 10 15 15 20 20 25 25 30 30 35 35 Y2 Contour Plots for Y2 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 277 The experiment resulted in 28 days compressive strength values ranging from 13MPa to 45MPa. The response was then analyzed to evaluate the influence of the different variables and their interactions on the 28 days strength gain of the concrete. From the results, as expected, the ultimate strength gain is highly affected by the water/ cement ratio (x1) as shown in Figure 7. The lower the value of the water cement ratio, the higher the value of the ultimate compressive strength. The cement/ total aggregates ratio and the fine aggregate / coarse aggregates ratio had a slight effect of the ultimate compressive strength gain. The interaction effects of the different variables on the 28 days compressive strength was also investigated and the results show that the interaction between the water /cement ratio and the cement/ total aggregates ratio and that of the cement/ aggregates ratio and the fine aggregates/ coarse aggregates ratio had a significant effect in the 28 days compressive strength while the interaction between the fine aggregates/ coarse aggregates ratio and the water cement ratio did not have a significant effect on the 28 days compressive strength as shown in Figure 7. Figure 7: Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the 7 Days Compressive Strength (Y4) Figure 8 (a): The ultimate compressive strength residual plots 0.60.50.4 40 35 30 25 20 15 10 0.250.200.15 0.650.600.55 x1 M ea n of Y 4 x2 x3 Factorial plots for Y4 Fitted Means 44 32 20 0.650.600.55 0.250.200.15 44 32 20 0.60.50.4 44 32 20 x3 x2 * x1 x3 * x1 x1 * x2 x3 * x2 x1 * x3 x1 x2 * x3 x2 0.3367 0.5 0.6633 x1 0.15468 0.22 0.28532 x2 0.53468 0.6 0.66532 x3 M ea n of Y 4 Factorial plots for Y4 Fitted Means 5.02.50.0-2.5-5.0 99 90 50 10 1 Residual Pe rc en t 403020 5.0 2.5 0.0 -2.5 -5.0 Fitted Value R es id ua l 630-3-6 20 15 10 5 0 Residual Fr eq ue nc y 605550454035302520151051 5.0 2.5 0.0 -2.5 -5.0 Observation Order R es id ua l Normal Probability Plot Versus Fits Histogram Versus Order Residual Plots for Y4 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 278 Figure 8 (b): The Ultimate Compressive Strength Model Building Report The results were then used to plot the residual plots for the ultimate compressive strength at 28 days and to generate the mathematical model for the ultimate compressive strength of the blended Portland cement concrete production as shown in Figure 8. The results were further used to plot the contour plots for the generation of the concrete mix design ratios for the different target compressive strengths at 28 days given in Figure 9. Figure 9: The 7 days contour plots for the Interaction Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the 28 Days Compressive Strength (Y4) Y4 = -66.9 - 21.9 X1 + 581 X2 + 180.8 X3 - 303.6 X1*X2 - 721 X2*X3 Step Change Step P Final P 4 3 2 1 Add X2*X3 Add X1*X2 Add X2 Add X3 Add X1 0.006 0.015 0.227 0.018 0.000 0.006 0.004 0.213 0.003 0.000 1007550250 R-Squared(adjusted) % x3 x2 x1 7550250 Increase in R-Squared % x3 x2 x1 100500 R-Squared % X1: x1 X2: x2 X3: x3 Final Model Equation Model Building Sequence Displays the order in which terms were added or removed. Incremental Impact of X Variables Long bars represent Xs that contribute the most new information to the model. Each X Regressed on All Other Terms Gray bars represent Xs that do not help explain additional variation in Y. A gray bar represents an X variable not in the model. Multiple Regression for Y4 Model Building Report x1 0.5 x2 0.22 x3 0.6 Hold Values x2*x1 0.60.50.4 0.28 0.24 0.20 0.16 x3*x1 0.60.50.4 0.650 0.625 0.600 0.575 0.550 x3*x2 0.280.240.200.16 0.650 0.625 0.600 0.575 0.550 > – – – – – < 15 15 20 20 25 25 30 30 35 35 40 40 Y4 Contour Plots for Y4 ULTIMATE COMPRESSIVE STRENGTH PLOTS American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 279 3.4. The Effect water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the Average Density of blended Portland cement concrete (Y5) The experiment resulted in average density values ranging from 2100kg/m3 to 2500kg/m3 as shown in Table 4. The response was then analyzed to evaluate the influence of the different variables and their interactions on the density of the concrete as shown in Figure 10. From the results, it was clear that the density is highly affected by the water/ cement ratio (x1) and the cement/ total aggregates ratio (x2) while the fine aggregates / coarse aggregates ratio had a slight effect of the average density. The higher the value of the water / cement ratio, the higher the value of the density up to 0.5 above which the higher the water/ cement ratio, the lower the density. The cement/ total aggregates ratio on the other hand affected the density in that the higher the ratio, the lower the density. Figure 10: Effect of water /cement ratio (x1), the cement/ total aggregates ratio (x2) and fine aggregates/ coarse aggregates ratio (x3) variables on the 7 Days Compressive Strength (Y4) 3.5. Model validation Model validation was done through repeat tests on the 28 days compressive strength with various target strengths. The same type of aggregates and CEM C cement was used to cast 20 runs repeated three times at different times 0.60.50.4 2.48 2.46 2.44 2.42 2.40 2.38 0.250.200.15 0.650.600.55 x1 M e a n o f Y 5 x2 x3 Factorial plots for Y5 Fitted Means 2.5 2.4 2.3 0.650.600.55 0.250.200.15 2.5 2.4 2.3 0.60.50.4 2.5 2.4 2.3 x3 x2 * x1 x3 * x1 x1 * x2 x3 * x2 x1 * x3 x1 x2 * x3 x2 0.3367 0.5 0.6633 x1 0.15468 0.22 0.28532 x2 0.53468 0.6 0.66532 x3 M e a n o f Y 5 Factorial plots for Y5 Fitted Means American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 280 and the results were as shown in Table 5. Table 5: 28 day’s compressive strength model validation test results for Blended Portland cement (CEM C) RUN x1 x2 x3 Cement Water Fine Aggregates Coarse Aggregates SLUMP Y1 (mm) 7 days (Y2) 28 days Strength (Mpa) Y4 kg kg kg kg mm Mpa ACTUAL Mpa TARGET Mpa 1 0.5 0.3 0.7 491.3 236.9 687.9 1033.9 86.0 22.5 30.1 30.0 2 0.5 0.2 0.6 405.3 202.6 690.8 1151.3 23.0 23.0 29.3 29.9 3 0.5 0.2 0.5 405.3 202.6 641.8 1200.3 37.0 23.4 30.9 28.5 4 0.5 0.3 0.6 489.5 244.8 643.4 1072.3 88.0 22.6 30.1 29.7 5 0.5 0.2 0.7 405.3 202.6 735.9 1106.2 35.0 24.4 33.0 31.4 6 0.5 0.2 0.5 432.4 216.2 627.6 1173.8 31.0 20.5 29.2 25.0 7 0.6 0.2 0.6 342.4 205.4 742.3 1159.9 34.0 16.7 25.0 24.4 8 0.5 0.2 0.5 405.3 202.6 641.8 1200.3 30.0 22.2 32.2 28.5 9 0.5 0.3 0.6 489.5 244.8 643.4 1072.3 92.0 21.3 29.1 29.7 10 0.5 0.2 0.6 405.3 202.6 690.8 1151.3 21.0 21.7 29.2 29.9 11 0.6 0.2 0.5 304.5 177.0 685.8 1282.7 11.0 19.4 25.1 20.0 12 0.6 0.3 0.6 449.9 269.9 621.1 1109.1 122.0 16.0 20.0 20.0 13 0.6 0.2 0.6 342.4 205.4 682.8 1219.3 33.0 16.0 20.5 20.3 14 0.6 0.3 0.6 449.9 269.9 675.2 1055.0 137.0 15.1 20.9 19.4 15 0.6 0.2 0.6 342.4 205.4 742.3 1159.9 40.0 12.5 20.3 24.4 16 0.7 0.2 0.5 301.8 197.3 679.7 1271.2 22.0 13.0 16.8 15.0 17 0.6 0.3 0.6 449.9 269.9 1730.2 1055.0 133.0 12.4 17.7 19.4 18 0.6 0.3 0.6 449.9 269.9 1730.2 1109.1 119.0 12.2 18.0 20.0 19 0.6 0.2 0.6 342.4 205.4 1902.2 1219.3 31.0 13.9 18.9 20.3 20 0.6 0.2 0.6 342.4 205.4 1902.2 1159.9 27.0 16.5 22.0 24.4 The same mix ratios for the 20 runs were then used to cast concrete using the same aggregates but varying the brands of blended Portland cements. Cements from all the six local cement companies were used to evaluate the suitability of the mix proportions. The results were then compared with the target 28 days compressive strength generated through the model and the results were as shown in Table 6. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 281 Table 6: 28 day’s compressive strength model validation test results for different brands Blended Portland cements SLUMP (mm) ACTUAL 28 DAYS COMPRESSIVE STRENGTH (MPa) MPa RUN S CE M A1 CE M B CE M C CE M D CE M E CE M F CE M A1 CEM B CE M C CE M D1 CE M E CE M F PREDICTE D 1 43 40 86 136 139 76 28.2 8 26.42 30.1 4 32.0 8 30.0 6 23.5 0 30.00 2 9 10 23 25 30 29 28.1 0 24.08 29.3 2 33.5 9 25.8 1 22.0 0 29.91 3 16 9 37 20 20 25 29.0 6 23.63 30.9 0 31.7 6 28.0 7 22.4 2 28.47 4 44 60 88 145 170 95 28.3 3 22.16 30.1 5 30.1 4 28.3 1 20.6 1 29.70 5 28 5 35 14 27 23 26.9 2 24.14 32.9 9 30.6 2 25.6 9 20.8 3 31.36 6 16 4 31 11 23 22 25.1 6 21.37 29.1 6 27.5 3 18.1 7 19.6 3 25.00 7 33 20 34 24 32 22 19.3 9 18.07 25.0 2 23.4 6 18.8 3 18.3 3 24.42 8 16 10 30 24 28 11 25.9 5 24.48 32.2 1 33.6 7 24.0 3 25.9 6 28.47 9 41 48 92 141 174 105 23.4 2 21.62 29.0 7 32.0 1 22.2 5 21.1 1 29.70 10 5 5 21 12 25 9 26.2 7 26.64 29.1 8 32.4 7 28.3 2 22.3 0 29.91 11 4 2 11 6 6 5 20.9 5 21.21 25.1 0 25.5 1 17.8 9 18.0 9 20.00 12 156 179 122 203 227 202 17.4 0 17.71 20.0 2 21.3 5 16.3 8 13.1 4 19.97 13 40 29 33 46 47 20 19.8 6 19.27 20.4 9 27.7 4 22.3 4 15.6 7 20.35 14 145 180 137 118 204 218 19.3 2 18.30 20.9 3 24.3 0 20.2 7 15.9 0 19.42 15 38 38 40 69 52 13 16.5 6 18.27 20.3 1 27.6 4 19.4 1 14.8 2 24.42 16 39 55 22 75 70 21 15.8 5 13.63 16.8 2 21.9 3 17.9 2 14.1 9 15.00 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 282 17 150 201 133 215 211 202 14.8 6 15.10 17.7 1 21.3 3 20.4 3 15.3 5 19.42 18 135 204 119 207 209 229 12.0 5 17.85 18.0 3 15.3 5 18.3 8 15.8 4 19.97 19 26 13 31 16 47 21 19.6 0 19.49 18.8 5 25.2 9 22.3 4 15.6 7 20.35 20 16 28 27 17 32 15 20.5 7 17.69 22.0 0 17.5 4 24.0 3 14.8 2 24.42 3.6. Determination of appropriate blended Portland cement concrete proportions Due to the variability of concrete in production caused by the differences in material properties and workmanship, it is necessary to design a concrete mix such that the expected mean strength is greater than the specified design characteristic strength by a specified margin. The British Research Establishment through the design of concrete mixes specifies that the margin should be calculated as shown in Equation 1 where the terms are as illustrated in Table 7. ksff cm += …………………………………………………………………………………………….(1) Table 7: Illustration of terms used in Equation 1 as given in BS 532828 Terms Meaning Terms Value fm The target mean strength k for 10% defective 1.28 fc The specified characteristic strength k for 5% defective 1.64 s The standard deviation k for 2.5 % defective 1.96 k A constant K for 1% defective 2.33 The standard deviation s for the 28 days compressive strength results was 6.841 as illustrated in Figure 11. The British standards, BS 5328 specifies a k of 1.64 for 5% defective. The 28 days compressive strength margin was thus calculated as shown in Equation 2 giving a compressive strength margin of 11.22MPa for all the strength classes of the blended Portland cement concrete. MPaxks 22.11841.664.1 == ……………………………………………………………………. (2) American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 283 Figure 11: Statistical analysis of the 28 days compressive strength Y4 results Considering the results in Tables 5 and 6, only three out of the six blended Portland cement brands (CEM C, CEM D1 and CEM E) achieved the target compressive strength of 30MPa. The other three brands did not achieve the target 30MPa. Based on the calculated compressive strength margin of 11.22MPa, the target 28 days compressive strength for C15 concrete is 26.22MPa, for C20 is 31.22MPa, for C25 is 36.22MPa and for C30 is 41.22MPa. It was therefore observed for a workable concrete with a slump of ≥ 30mm that the most appropriate mix proportions for the blended Portland cement concrete were: : 1:2.2:3.4 (w/c 0.6) for strength class C15 and 1:1.3:2.2 (w/c 0.5) for strength class C20. It was further noted that none of the blended cement brands achieved the target design strength for strength class C25 and above. It was concluded that the blended Portland cements may not be suitable for producing structural concrete strength class C 25 and above. 4. Conclusion Based on the experiments carried out and the results obtained, the following conclusions can be arrived at: a) The different brands of blended Portland cements from the six different manufacturers have varying physical and mechanical properties which in turn affect the concrete produced when the different brands of cements are used. Other than one brand (CEM F), all the other five brands met the minimum physical and mechanical properties as stated in the Kenyan standards KS EAS 18-1:2001. b) The appropriate concrete mix ratios for the Kenyan blended Portland cement concrete are as follows: Class C15 is 1:2.2:3.4 at a water/ cement ratio of 0.6, and C20 is 1:1.3:2.2 at a water cement ratio of 1st Quartile 23.382 Median 29.025 3rd Quartile 31.959 Maximum 41.892 26.573 30.382 26.832 30.594 5.733 8.484 A-Squared 0.66 P-Value 0.082 Mean 28.477 StDev 6.841 Variance 46.804 Skewness -0.015283 Kurtosis -0.344077 N 52 Minimum 13.874 Anderson-Darling Normality Test 95% Confidence Interval for Mean 95% Confidence Interval for Median 95% Confidence Interval for StDev 4236302418 Median Mean 313029282726 95% Confidence Intervals Summary Report for Y4 American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 284 0.5. Since some cement brands do not meet the minimum characteristic strength of 25MPa required for class C 25, Kenyan Blended Portland cements should not be used to produce concrete of strength class C25 and above since the mix does not achieve the target compressive strength of 36.22MPa at 28 days when no additive is used. c) The construction industry in Kenya should come up with policies to ensure that un qualified personnel do not design and supervise reinforced concrete structures to ensure that quality control measures are adhered to on site. 5. Limitations of the study The main limitation of the study is the use of fine aggregates from the same river bank and coarse aggregates from the same quarry thus the influence of the difference in the properties of the aggregates on the quality of concrete was not investigated. 6. Recommendations From the results of the experiments, the authors recommend that; 1. The blended Portland cements may not be suitable for the production of concrete class C25 and above. 2. Further research should be done to establish the influence of the difference in aggregates properties on the quality of blended Portland cement concrete in Kenya. Acknowledgment The authors would like to thank the Pan African University, Institute of Basic Sciences , Technology and Innovations for funding the research, the Kenya Bureau of Standards Civil Engineering and Testing department and Jomo Kenyatta University of Agriculture and Technology for the laboratory equipment’s and technical staff used during the research. References [1] Government of Kenya, Kenya Vision 2030; A Globally Competitive and Prosperous Kenya, Government printer, Nairobi: Ministry of Planning and National Economic and Social Council (NESC), 2007. [2] K. K. Adewole, W. O. Ajagbe, & I. A. Arasi, “Determination of appropriate mix ratios for concrete grades using Nigerian cements”, Leonardo Electronic Journal of Practices and Technologies, Vol. 26, pp. 79-88, 2015. [3] V. A. Okumu, W. O. Oyawa & S. M. Shitote, “The Effect of the Properties of Constituent Materials on the Quality of Concrete in Kenya”, Proceedings of the 2016 Annual Conference on Sustainable Research and Innovation (SRI), 4th- 6th May, 2016. Nairobi, Kenya. pp. 225-230, 2016. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 285 [4] P. D. Kumbhar & P. B. Murnal, “Assesment of suitability of existing mix design methods of normal concrete for designing high performance concrete”, International Journal of Civil and Structural Engineering , Vol. 3 (1), pp. 158-167, 2012. [5] H. N. Ngugi, R.N. Mutuku, & Z.A. Gariy, “Effects of Sand Quality on Compressive Strength of Concrete: A Case of Nairobi County and Its Environs, Kenya”, Open Journal of Civil Engineering, Vol. 4, pp. 255-273, 2014. [6] R. Maciulaitis, M. Vaiciene, & R. Zurauskiene, “The effects of concrete composition and aggregates properties on performance of concrete”, Journal of civil engineering and management, Vol. 15 (3), pp. 317-324, 2009. [7] S. Ahmad, “Optimum concrete mixture design using locally available ingredients”, The Arabian Journal of Science and Engineering , vol. 31 (1B), pp. 27-33, 2007. [8] M. J. Mohamed, “Effect of Curing Method and Insoluble Residue in Cement On The Compressive Strength of Portland Cement Mortar”, Al-Qadisiya Journal for Engineering Sciences, Vol. 6, No. 1, pp. 74-80, 2013. [9] R. K. Dhir, M. C. Limbachiya, M. J. Mc Carthy & A. Chaipanic, “Evaluation of Portland Limestone Cements for Use in Concrete Construction”, Materials and Structures, Vol 40, pp 459-473, 2007. [10] K. Marar and Ö. Eren, “Effect of cement content and water/cement ratio on fresh concrete properties without admixtures”, International Journal of the Physical Sciences, Vol. 6(24), pp. 5752-5765, 16 October, 2011. [11] I. B. Muhit, S. Haque, and M. R Alam, “Influence of Crushed Coarse Aggregates on Properties of Concrete”, American Journal of Civil Engineering and Architecture, Vol 1 (5), pp. 103-106, 2013. [12] C. GONG, J. ZHANG, S.WANG & L. LU, “Effect of Aggregate Gradation with Fuller Distribution on Properties of Sulpho aluminate Cement Concrete”, Journal of Wuhan University of Technology-Mater. Sci. Ed. Vol 30, pp. 1029- 1035, 2015. [13] J. M Chi, R Huang , C. C Yang and J. J Chang, “Effect of Aggregate Properties on the Strength and Stiffness of Lightweight Concrete”, Cement and Concrete Composites, Vol 25 (2) pp. 197–205, 2003. [14] H. Donza, O. Cabrera & E. F. Irassar, “High Strength Concrete With Different Fine Aggregates”, Cement and Concrete Research, Vol 32(11), pp. 1755-1761, 2002. [15] W. Ke-Ru, B. Chen, W. Yao & D. Zhang, “Effect Of Coarse Aggregates Type on Mechanical Properties of High Performance Concrete”, Cement and Concrete Research. Vol 31(10), pp. 1421- 1425, 2001. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2017) Volume 31, No 1, pp 265-286 286 [16] M. S. Meddah, S. Zitouni & S. Belaabes, “Effect of Content and Particle Size Distribution of Coarse Aggregate on the Compressive Strength of Concrete”, Construction and Building Materials. Vol 24 (4) pp. 505-512, 2010. [17] A. Katz and H. Baum, “Effect of High Levels of Fines Content on Concrete Properties”, ACI Materials Journal. Technical Paper. Vol 103 (6) pp 476-482, 2006. [18] W. O. Oyawa, N. K. Githimba, & G. N. Mang'urio. “Structural response of composite concrete Filled plastic tubes in compression”, Steel and Composite Structures, Vol. 21, No. 3, pp. 589-604, 2016. [19] J. Dils, D. Schutter , & V. Boel, “Influence of Mixing Procedure and Mixture type on Fresh and Hardened Properties of Concrete A Review”, Materials and Structures Journal. 2012. [20] N. B. Singh, K. N. Bhattacharjee, & S. Rai, “Hydraulic Behaviour of Ternry and Quaternary Blended Cements”, Transactions of the Indian Ceramic Society, 57(3), pp. 81-84, 2014. [21] F. M. Wegian, “Effects of seawater for mixing and curing on structural concrete”, The IES J ournal Part A: Civil & Structural Engineering, vol. 3 (4), pp. 235-243, 2010. [22] C. N. Aginam, S. N. Umenwaliri, & C. Nwakire, “Influence of mix design methods on the compressive strength of concrete”. ARPN Journal of Engineering and Applied Sciences, vol. 8(6), pp. 438-444, 2013. [23] Kenya Bureau of Standards. KS EAS 18-1:2001- Cement Part 1: Composition, Specification and Conformity Criteria for Common Cements. Nairobi, Kenya Bureau of Standards, 2005. [24] D. O. Koteng', “Concrete use for Sustainable Development”, Presented at the 20th Engineers International Conference., The Institution of Engineers of Kenya., Kisumu, 2013. [25] D. C. Montgomery, Design and Analysis of Experiments (8th Ed.), John Willey and Sons. 2013. [26] S. Coruh, & S. Elevli, “Optimization Study of Dye Removal by Cement Kiln Dust Using The Central Composite Design of Experiments”, Global NEST Journal, vol. 17 (1), pp. 93-102, 2015. [27] M. Barbuta, E. Marin, S. M. Cimpeanu, G. Paraschir, D. Lepadatu, & R. D. Bucur, “Statistical Analysis of the Tensile Strength of Coal Flyash Concrete with Fibres using Central Composite Design”, Journal of Advances in Material Science and Engineering , pp. 1-7, 2015.