Time-dependence of chloride diffusion for concrete containing metakaolin Hossam S. Al-alaily, Assem A.A. Hassan n Faculty of Engineering and Applied Science, Memorial University of Newfoundland, St. John's, Newfoundland and Labrador, Canada A1B 3X5 article info Article history: Received 7 February 2016 Received in revised form 6 June 2016 Accepted 6 June 2016 Available online 14 June 2016 Keywords: Chloride diffusion RCPT Lifetime prediction Chloride diffusion reduction coefficient Metakaolin Statistical analysis RSM abstract Chloride diffusion coefficient depends on many variables including concrete quality, environmental conditions, and time. In this investigation, the concrete quality and time were studied while maintaining the environmental conditions constant. Fifty-three concrete mixtures were tested based on a refined statistical analysis. Enhanced response surface method (RSM) was used to present the most significant factors affecting the chloride diffusion at different ages. The tested mixtures contained various water-to- binder (W/B) (ratios 0.3–0.5), metakaolin (MK) replacement (0–25%), and total binder content (350– 600 kg/m 3 ). Bulk diffusion test was adopted for two years to determine the time-dependent coefficient m of chloride diffusion for all mixtures based on the error function solution to Fick's law. This coefficient was calculated based on two different bulk diffusion test methods: total and average methods. Design charts were developed to facilitate the optimization of mixture proportions for designers/engineers. The investigation also included some experimental relationships between the rapid chloride permeability test (RCPT), chloride diffusion coefficient, and compressive strength results. The results showed that the values of the chloride diffusion indicated a general reduction from 28 days to 760 days of testing. As the percentage of MK or binder content increased or as the W/B ratio decrease, the chloride diffusion re- duction coefficients, m total and m avr , were found to increase. Based on the analysis of variance (ANOVA) from the statistical model, MK was found to be the most significant factor affecting the chloride diffusion at late ages (360 and 720 days), while the W/B ratio was the most significant factor affecting early ages of chloride diffusion (28 and 90 days). The developed models and design charts in this paper can be of special interest to designers/engineers by aiding prediction of service life of concrete containing MK. & 2016 Elsevier Ltd. All rights reserved. 1. Introduction Enhancing protection and prediction of service life for reinfor- cing concrete structures strongly impacts economic and ecological sustainability. Rebar corrosion is one of the main problems af- fecting the durability of reinforced concrete structures. Service life time of reinforced concrete structures can be drastically reduced if the chloride ions near the rebar surface exceed the chloride threshold level in a short time [1]. Concrete cover is considered the first barrier for chloride ion penetration to reach the rebar surface. Chloride ions diffuse slowly through deeper cover thickness, thus improving the structure's service life. Once the chloride threshold is exceeded at the rebar surface, corrosion may start. The concrete cover protecting the rebars can be cracked due to the expansive nature of rust products. The cracking of concrete covers sig- nificantly increases the ingress of chlorides and other deleterious substances, causing strength reduction of the structural element and acceleration of the corrosion process [2,3]. Predicting the time of corrosion initiation is determined by the rate of chloride ingress (diffusion coefficient), environmental loading, and chloride threshold level [4–6]. The rate of chloride diffusion is affected by concrete quality due to the size and volume of pore structure, and by time due to the blockage and reduction of the capillary pore structure from the hydration process. Concrete is not a static material, as the concrete continues to hydrate with time, the vo- lume of capillary pores decreases, which reduces the diffusion rate with time. The concrete quality depends mainly on binder content, water-to-binder (W/B) ratio, and any supplementary cementing material (SCM), where they are the primary controlling factors for concrete quality. On the other hand, aggregates in concrete gen- erally have negligible permeability and consequently have an in- significant effect on the transport mechanism of chloride through concrete. Metakaolin (MK) as a SCM in concrete is receiving a lot of at- tention recently. The use of MK as a SCM can considerably improve Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/jobe Journal of Building Engineering http://dx.doi.org/10.1016/j.jobe.2016.06.003 2352-7102/& 2016 Elsevier Ltd. All rights reserved. n Corresponding author. E-mail address: ahassan@mun.ca (A.A.A. Hassan). Journal of Building Engineering 7 (2016) 159–169