Predictive Modelling of Compressive Strength in POFA-Modified Concrete using Power Series Regression

This study dealt with the predictive function of the compressive strength of concrete partially replaced with Palm Oil Fuel Ash (POFA). A total of 360 cubes, covering 3 per POFA replacement percentage per nominal mixes of M1, M2, M3, M4, and M5 were cast and tested for their compressive strength at 7, 14, 21, and 28 days curing ages. Using power-series function of the form where a, c2, c3 …cn are function coefficients, x1 = compressive strength, x2, x3 ……… xn are water and aggregates to binder ratios respectively, the natural logarithmic values of compressive strength, water, and aggregates to binder ratios were inputted into excel regression module and analytical runs done to obtain parameters for the power-series predictive functions. Taking K, R, and S as the coefficients for the function, the output from analyses showed that the adjusted coefficient of determination had a minimum value of 89.77%, the highest K value was at 21 days curing age of 6.4597 with the least regressed value of 5.9144 at 7 days. Likewise, R values had highest output of 2.9645 at 21 days of curing and the least being 2.3612 at 7 days. S values ranged from 7.092 to 0.9296 from 28 down towards 7 days, all negative. From the results obtained, power-series function can adequately predict compressive strength of concrete mixes of which the cement contents are partially replaced by POFA up to 25% by weight.