Abstract
The quality of concrete was assessed based on measurements of ultrasonic pulse velocity (UPV) and compressive strength. Portland cement (PC) was substituted with silica fume (SF), fly ash (FA) and blast furnace slag (BFS) as mineral admixtures. The compressive strength and UPV of concrete mixtures were estimated with classifiers alongside the tree model M5P in the data mining field. The compressive strength and UPV were modeled as a function of five input variables: day, cement, SF, FA, and BFS. It was found that UPV and compressive strength had an exponential correlation with SF, FA, BFS, SF + FA, SF + BFS and FA + BFS. However, additive-free admixture and PC-free substitution levels had a different constant in empirical work. According to the results obtained, both compressive strength and UPV of concrete mixtures could be reliably anticipated with the tree model M5P in a proportion of 97%. Hence, this model could be a potential new method for modeling mineral admixture concrete.