One of the primary causes of the low mechanical properties of rubberized concrete is the weak bond between crumb rubber (CR) and hardened cement paste. Many CR pretreatment techniques have been researched in an attempt to mitigate this problem. The NaOH pretreatment method is one of the most widely used, although the reported results are inconsistent due to the absence of standardized NaOH pretreatment concentrations and CR replacement levels. This study aims to develop models for predicting the mechanical and shrinkage properties of NaOH-pretreated CR concrete (NaOH-CRC) and conduct multi-objective optimization using response surface methodology (RSM). The RSM generated experimental runs using three levels (0, 5, and 10%) of both NaOH pretreatment concentration and the CR replacement level of fine aggregate by volume as the input factors. At 28 days, the concrete's compressive, flexural, and tensile strengths (CS, FS, and TS), as well as its drying shrinkage (S), were evaluated as the responses. The results revealed that higher CR replacements led to lower mechanical strengths and higher shrinkage. However, the strength loss and the shrinkage significantly reduced by 22%, 44%, 43%, and 60% for CS, FS, TS, and S, respectively, after the pretreatment. Using field-emission scanning electron microscopy (FESEM), the microstructural investigation indicated a significantly reduced interfacial transition zone (ITZ) with increasing NaOH pretreatment. The developed RSM models were evaluated using ANOVA and found to have high R2 values ranging from 78.7% to 98%. The optimization produced NaOH and CR levels of 10% and 2%, respectively, with high desirability of 71.4%.
Keywords: CR-concrete; crumb rubber (CR); optimization; pretreatment; response surface methodology (RSM).