Assessing the reliability of using concrete cubes in predicting the compressive strength of concrete columns

Abstract

The compressive strength of concrete, which is crucial for the safety and durability of structures, is typically evaluated through cube testing. Despite their widespread use, doubts have been raised about the accuracy of concrete cubes in reflecting the strength of actual on-site structures. This research assesses the reliability of using concrete cubes to predict the compressive strength of concrete columns. The experiment study involved casting and testing concrete cubes, alongside assessing the comprehensive strength of columns on construction sites using a rebound hammer. Through quantitative analysis using software like SPSS and Microsoft Excel, the research revealed significant variances between the strengths of the concrete cubes and that of the columns, with the columns having higher strengths. Variances were influenced by factors such as concrete mix designs, curing methods, mixing methods and compaction methods. On average, the difference between the cubes and columns was 2.15N/mm2 at 7 days from casting and 3.525 N/mm2 at 28 days from casting. The variance between the cube and column strength was 14.917 for a mix of 1:1.37:2.78 and 8.170 for a mix of 1:1.5:3. For a mix of 1:3:4, the variance between the cube and column strength was 6.067 and that for a mix of 1:2:4 was 3.714. The variance between the cube and column strength for a mix of 1:2:3 was 7.103. The study concluded that while concrete cubes provide insights into compressive strength, they may not fully capture the behaviour of concrete in columns due to several factors like the environmental and size differences. The study emphasizes the importance of complementing cube testing with non-destructive testing methods for accurate strength predictions. Applying correction factors while using cube testing, statistical techniques and further research on different concrete mix designs and using concrete cylinders during testing to enhance the predictive accuracy is recommended.