RESPONSE SURFACE METHODOLOGY BASED MODEL FOR PREDICTING COMPRESSIVE STRENGTH OF KUTA GRAVEL CONCRETE

Abstract

Compressive strength is undoubtably the most important property of concrete and measures the performance of a concrete mix. Hence, concrete constituent materials are mostly proportioned in terms of compressive strength. Several Design of Experiment tools have been developed that can be used to improve accuracy, optimize and model concrete properties. In this study, a model to predict compressive strength of Kuta gravel concrete was developed. Central Composite Design in Minitab was used to generate 20 mixes with varied water to cement, coarse aggregate to total aggregate, and total aggregate to cement ratios as design variables. Concrete cubes were tested for compressive strength at 28 days of curing. The model was developed and analyzed using response surface methodology. Results obtained showed that a maximum compressive strength of 27.47N/mm2 can be achieved with mix proportion of W/C of 0.4, CA/TA of 0.55 and TA/C of 3. The model has overall P-value of 0.001, R2 value of 90.2% and Adjusted R2 value of 81.38%. It was concluded among others, that Kuta gravel can be used in producing C15, C20 and C25 grades of structural concrete, and that the developed model is adequate in predicting 28-day compressive strength of Kuta gravel concrete.

Keywords

• compressive strength
• concrete
• kuta gravel
• predictive model
• response surface methodology