Optimal mechanical performance prediction of a sediment-based geopolymer road sub-layer

Abstract

One of the major obstacles for the implementation of dredged sediment in the manufacture of concrete is its variable mineralogical properties. However, its use in the manufacture of geopolymer binder as proven to be an innovative solution for their recovery and could allow the emergence of a new ecological binder as a replacement of Portland cement. Therefore, the present study aims to assess the variability of the physico-chemical properties of sediments in several ports in the Gironde department in France (La Rochelle, Bayonne, Bordeaux and the Arcachon Basins) in order to assess the impact on the mechanical properties of the final developed material. The study includes two parts. On the one hand, a data base that included sixteen fine sediments coming from the different ports has been made. The sediments were then used as a precursor, without any preliminary treatment, in order to develop an ecological sediment-based geopolymer binder with an alkali reagent composed of sodium hydroxide (NaOH) and sodium silicate (Na2SiO3). Compressive and bending strength linked to elastic modulus, porosity and permeability were assessed at 28 days for different binder samples. Using both multivariate linear regression and response surface methodology, a performance prediction of the mechanical properties of the binder samples were quantitatively estimated and linked to their initial properties. This statistical approach highlights the effects of various parameters influencing the life performance of the binder after 28 days. On the other hand, the optimum formulation was then used in order to evaluate the space variability trough X-ray tomography and ultrasound speed. The results were then transposed to a mechanical finite element model of a road sub-layer coupled to a probabilistic model to analyse the effect of the variability of different mechanical parameters on the prediction of the material mechanical performance.