Predictive Modeling and Multi-response Optimization of Physical and Mechanical Properties of SCC Based on Sand’s Particle Size Distribution

This paper focuses on the modeling and optimization of the physical and mechanical properties of self-compacting concrete (SCC), prepared using the modified packing model design, taking into consideration sand’s particle size distribution (PSD) and fineness modulus (FM). The optimization is predicated using the response surface methodology. The analysis of variance is exploited to determine the statistical significance of the PSD on the studied properties of SCC. For that, we studied twenty-four SCC mixtures, using three kinds of sands with different FM and particle shapes: dune sand, river sand, and crushed sand. The results show that PSD properties of used sand, despite its shape, are good predictors for SCC fresh properties, but less predictive for compressive strength. The multi-response optimization allowed the estimation of sand’s PSD parameters that give the optimal physical and mechanical properties of SCC, with an overall desirability of 0.923.