Browsing by Author "Ejiogu, Emeka Kingsley"

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    Statistical models for predicting compressive strength and density of sandstone concrete
    (Federal University of Technology, Owerri, 2015-12) Ejiogu, Emeka Kingsley
    This thesis developed statistical models for predicting compressive strength and density of sandstone concrete. The materials used in the laboratory experiment are water, Ordinary Portland Cement, river sand and sandstone. Forty-four (44) mix ratios were developed from eleven (11) solid mix ratios and four water-cement ratios (0.45, 0.5, 0.55, and 0.6). A total of one hundred and thirty-two (132) concrete cubes were cast for compressive strength test, comprising of three cubes per mix ratio. The Saturated surface dry (SSD) densities of the concrete cubes were determined. Using Ibearugbulem‟s regression method, the first twenty-two (22) mix ratios were used to determine the coefficients of the regression while the remaining twenty-two (22) mix ratios were used to validate the models. The results show that, optimum compressive strength of the sandstone concrete obtained from experiment at 28 days is 24.18 N/mm2 and corresponds to 0.6:1:2.5:3.5 mix ratio (for water, cement, sand and sandstone respectively) while the optimum value predicted by the statistical model is 22.55 N/mm2 corresponding to the same mix ratio. The maximum SSD density of the sandstone concrete obtained from experiment at 28 days is 2592.59Kg/m3 and corresponds to 0.6:1:1.75:4.25 mix ratio while the statistical model predicted maximum value of 2567.7Kg/m3 for the same mix ratio. The compressive strength was found to increase as the water/cement ratio increases and fine/coarse aggregate ratio decreases. Also, high water/cement ratio and low fine/coarse aggregate ratio resulted in high densities. Theresults from the models compared favorably with the corresponding experimental results. Predictions from the models were tested with the statistical Fisher test and found to be reliable at 95% confidence level. With the models developed, any desired compressive strength or SSD density of sandstone concrete can be predicted if the mix proportions are known and vice versa.