Model Development for Prediction of Concrete Compressive Strength: Advancing Construction Industry Practices and Quality Control Standards

Abstract

In construction, the selection of concrete mix grades centres on the specified strength outlined in the design. However, achieving the desired strength necessitates laborious and costly experimental investigations. This inspires the current study, which seeks to establish an equation to predict concrete compressive strength (CS) based on water to cement ratio (WCR), reducing the need for costly experimental investigations. 90 concrete cube samples were made using Portland limestone cement of 42.5 N grades, with WCR ranging from 0.45 to 0.65 and two different mix ratios. Strength was tested at 7, 14, and 28 days, with statistical analysis focusing on the 28-day CS. Models developed using Design Expert software exhibited over 94% accuracy in predicting 28-day compressive strength, indicating strong alignment with experimental data. Fit Statistics indicated a satisfactory fit with adjusted R² of 0.9932 and predicted R² of 0.9715. Adequacy precision, signalling the signal-to-noise ratio, exceeded 4, indicating a robust model. P value was significant (<0.05), and the F-value (583.28) suggested the model's significance in predicting CS. The findings imply the model's potential for guiding design decisions effectively.

Keywords — Model, Water, Cement, Strength, Concrete, Statistical