To promote more construction applications of desert sand (DS), this study provides a method to design and optimize the mix proportion of desert sand concrete (DSC) based on a statistical model. The water to cement (W/C) ratio, sand to aggregate (S/A) ratio, DS to fine aggregate (D/F) ratio, and water reducer (WR) dosage are selected as the experimental factors. In addition, the slump, air content, and compressive strength at 7 days and 28 days of 36 DSC mixtures are investigated experimentally. The statistical models are established based on a central composite design to evaluate the effects of the experimental factors on the DSC properties, and the analysis of variance and the residual for the models indicate that all the models are valid and have predictive capacity. Additionally, the results presented in the response trace or surface plots indicate that an increase in the D/F ratio in the mixtures can decrease the slump and enhance the air content, but the augmentation of the AMC ratio can increase the slump and reduce the air content. An increase in both the DM ratio and W/C ratio cause a reduction in compressive strength. On the other hand, the effects of the S/A ratio and WR dosage on the DSC properties exhibit minor roles relative to the important effects of the W/C ratio and D/F ratio. In addition to the above, optimization criteria are proposed to select the appropriate mix design parameters of DSC for various construction demands. (C) 2019 Elsevier Ltd. All rights reserved.
