Models For Predicting Compressive Strength Of Concrete

ABSTRACT

Several approaches using regression functions have been proposed over the years for predicting either the desired compressive strength of concrete or the mixture proportions of its constituent materials. This study as part of the continuing effort, presents models developed using Ibearugbulam's regression principle for predicting the desired compressive strength of granosand concrete when the mix proportions are known or vice-versa. Laboratory experiments were carried out with different mix proportions of four (4) component materials (granite, sand, cement and water). Sixty (60) different mixes were used for experimental test by casting four (4) cubes for each of the mixes. Results revealed that the developed models gave an optimum compressive strength of 32.64Nmm with a mix ratio of 0.525 :1: 2.14 : 2.86 (Water : Cement: Fine Aggregate : Coarse Aggregate) which generated a performance index value of 2.19. From the statistical analysis, the calculated f value of 1.74, using the Fisher's test was less than the allowable f value of 1.89 from statistical table at 95% confidence level. Also, the calculated t value of 2.10 from the Student's test was less than 2.26 obtained from the statistical table. These showed that the difference between laboratory compressive strength results and model compressive strength results were not significant. Thus, the models formulated were adequate and reliable for predicting the compressive strength of granosand concrete.