Service life of reinforced concrete exposed to chloride environment highlighting the effect of metakaolin in the mixture

Alalaily, Hossam S. (2017) Service life of reinforced concrete exposed to chloride environment highlighting the effect of metakaolin in the mixture. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (6MB)

Abstract

Experimental and numerical analysis were conducted to investigate the service life of reinforced concrete exposed to chloride attack highlighting the effect of using metakaolin (MK) in concrete mixtures. The research also aims to study and optimize the mixture proportions/composition of MK mixtures in order to maximize their service lifetime. Design tools/charts were also developed in this research to facilitate and simplify the service life prediction of the mixtures. Three main service life periods of reinforced concrete exposed to chloride environment were studied: corrosion initiation, propagation, and damage periods. The research was divided into three stages; the first stage experimentally studied and investigated the three corrosion periods (initiation, propagation, and damage periods) in concrete with MK and concrete with different curing conditions. The second stage developed mechanical and chloride permeability prediction equations for concrete containing MK. Finally, the third stage utilized probabilistic and extended finite element (XFEM) in predicting the corrosion periods for concrete containing MK. In the first stage, two corrosion testing methods were used to study the three corrosion periods: impressed current corrosion and wet/dry cycles testing methods. Seventy-eight concrete samples were tested under the impressed current corrosion testing. In this investigation, three variables were studied: percentage of MK% (0% to 20%), cover thicknesses (20, 30, and 40 mm), and curing conditions. On the other hand, thirty concrete samples were tested under the wet/dry corrosion testing for two years to study the corrosion period under natural corrosion process and to compare the results with those of the impressed current corrosion testing. The results of the first stage indicated that the incorporation of MK significantly increased both the initiation and propagation periods with little increase in the damage period. Also, the different curing conditions had a clear effect on the initiation and propagation periods while length of the damage period was not significantly affected. The results from the wet/dry corrosion testing were compared with the impressed current corrosion technique and proved that the impressed current corrosion test can be used effectively to evaluate and compare the corrosion activities in different concrete qualities but cannot be used after the initiation of the first crack. The second stage tested fifty-three concrete mixtures containing MK to develop prediction equations using statistical analysis. Three factors were considered in this stage: total binder content (350kg/m³-600Kg/m³), the percentage of MK (0%-25%), and the water-to-binder ratio (0.3-0.5). The mixtures were examined based on the rapid chloride permeability test, chloride diffusion test, compressive strength, modulus of elasticity, splitting tensile strength, flexural strength, and cost of mixture per cubic meter. More over, bulk diffusion test was adopted for two years to determine the time-dependent coefficient of chloride diffusion for all mixtures based on the error function solution to Fick’s law. Finally this stage also included some experimental relationships between the rapid chloride permeability test, chloride diffusion coefficient, and compressive strength results.moreover The derived models and design charts based on the experimental results in this stage were useful for optimizing and predicting the concrete mechanical and chloride permeability properties. Finally, the third stage is divided into two parts: the first part was using probabilistic method to predict the probability of corrosion initiation and the second part was developing iv XFEM to enhance predicting the propagation period (corrosion induced cracking). Three variables were investigated in this stage: total binder content (350kg/m³-600Kg/m³), MK% (0%-25%), concrete cover (20 mm- 60 mm) and W/B (0.3-0.5). In the first part, Monte Carlo simulation technique was used to predict probability of corrosion initiation in a concrete structure containing MK with different mixture proportions and concrete covers. Statistical modeling was then utilized to develop prediction models/charts to predict the probability of corrosion initiation for concrete containing MK, and identify the most significant factors affecting this stage. The results showed that the probability of corrosion decreased as the percentage of MK increased. The results also showed that the most significant factor affecting the corrosion probability was found to be MK replacement, W/B ratio, and binder content respectively, in order of significance. On the other hand, the second part utilized XFEM to investigate the crack propagation and expansive behavior of corroded steel bar on the concrete cover for a concrete structure containing MK. statistical analysis was incorporated with XFEM to develop prediction model/charts for the expansive pressure resulted from corrosion products. Finally validation samples were tested under accelerated corrosion to verify the enhanced model. The results indicated that the time required for corrosion-induced cracking obtained from the developed prediction model showed a good agreement with the experimental results of the accelerated corrosion samples. Also, the cracks predicted by the XFEM showed a similar trend of variation with that found in the accelerated corrosion samples.

Actions (login required)

View Item