Concrete calcium leaching at variable temperature: Experimental data and numerical model inverse identification

• Published in: Computational materials science 2010-06, Vol.49 (1), p.35-45
• Main Authors: de Larrard, T., Benboudjema, F., Colliat, J.B., Torrenti, J.M., Deleruyelle, F.
• Format: Article
• Language: English
• Subjects: Applied sciences; Artificial Neural Network; Buildings. Public works; Calcium; Computer Science; Concretes; Concretes. Mortars. Grouts; Diffusion; Diffusivity; Engineering Sciences; Exact sciences and technology; Flotation. Enrichment. Beneficiation; Fundamental areas of phenomenology (including applications); Indexing in process; Inverse; Inverse identification; Leaching; Materials; Mathematical analysis; Mathematical models; Measurement and testing methods; Mechanics; Mechanics of materials; Metals. Metallurgy; Modeling and Simulation; Ores; Physics; Porosity; Production of metals; Properties and performance tests of mortar and hardened concrete; Solid mechanics; Structural and continuum mechanics; Tortuosity; Variability
• ISSN: 0927-0256; 1879-0801
• DOI: 10.1016/j.commatsci.2010.04.017

A simplified model for calcium leaching in concrete is presented. It is based on the mass balance equation for calcium in the porous material. This model is implemented in a Finite Volume code and validated by comparison between numerical simulations and experimental results found in the literature, for cement pastes and mortars as well as for concretes, with a satisfactory agreement. Then, a parametric survey has been performed. It enlightens the large influence of porosity and diffusivity on the leaching kinetic. In complement, a large experimental campaign, which aims at acquiring data on the material characteristics variability (within several batches for a same concrete mix design) has been undertaken. This campaign investigates porosity and the degradation depth at different times considering accelerated leaching under variable temperature. Nevertheless, the coefficient of tortuosity (which partially controls diffusivity in concrete) cannot be directly measured, although it is an important parameter to model the calcium diffusion process. Therefore, an inverse identification tool is developed and validated, based upon the Artificial Neural Network theory, using the available experimental data as input data and the numerical model for the network training.