Phenomenological model for quantifying concrete chloride diffusion coefficient

• Published in: Construction & building materials 2019-11, Vol.224, p.773-784
• Main Authors: Chidiac, S.E., Shafikhani, M.
• Format: Article
• Language: English
• Subjects: Aggregate volume fraction; Chloride diffusion coefficient; Compressive strength; Concrete; Phenomenological model; Porosity; Tortuosity
• ISSN: 0950-0618; 1879-0526
• DOI: 10.1016/j.conbuildmat.2019.07.006

•Model employs tortuosity, aggregate, porosity, strength, cement content & composition.•Effective chloride diffusion coefficient of concrete can be quantified using the following relationships:•DCl=31-Va23-Va10-CCEMCEM100+CGGBFSGGBFS100+CSFSF100fc'+B•CCEM=0.037C3S-0.18•CGGBFS=1.39CaO+MgO+SiO2GGBFSC3SCCEM•CSF=3.60SiO2SFC3SCCEM•B=3.293.56-C3SC3S>0.3410.59C3S≤0.34•The chloride diffusion model is applicable to a wide range of concrete mixture compositions and age.•The proposed model accurately captures the effects of w/cm ratio, inter-connectivity of pores, chemical composition of cement and SCMs, and different replacement levels of SCMs.•Model quantification values are found to be statistically equal to those quantified experimentally to the 95% confidence level. Compressive strength and the chloride diffusion coefficient are the common and principal performance measures for designing durable concrete structures. Hence, a comprehensive and consistent phenomenological model is put forward that quantifies a concrete’s effective chloride diffusion coefficient based on the tortuosity factor, aggregate volume fraction, porosity, cement paste chloride diffusion coefficient, compressive strength, cement and cementitious material content and composition. The phenomenological model results are first shown to depict chloride ion diffusion phenomena in porous media. Experimental data reported in the scientific literature are then used to evaluate visually and statistically the accuracy and precision of the phenomenological model. The model results are found statistically equal to experimentally quantified chloride diffusion coefficient values to a 95% confidence level and possess high accuracy and precision for a wide range of concrete mixture proportions, composition and age. The model average percent error is less than the standard test method repeatability and the COV of the reproducibility. The standard error, samples mean, and R2 are found to be 0.41 × 10−12 m2/s, 1.46 × 10−12 m2/s and 0.94, respectively.