Corrosion of Steel in Cracked Concrete: Experimental Investigation using External Polarization Mingdong Bi and Kolluru Subramaniam* Department of Civil Engineering, City College of New York, USA
Abstract Corrosion of steel reinforcement is one of the main causes of damage in concrete structures. Often the corrosion of steel embedded in concrete is influenced by cracks in concrete. Cracks in the concrete are often produced due to loads, shrinkage or thermal gradients. Once a crack is formed, the crack provides an easy access for ingress of water, chloride ions and oxygen to the steel surface. This paper presents the results of an experimental program which aims to investigate (a) the influence of a crack on the corrosion mechanism and (b) external polarization of steel in cracked concrete. A series of electrochemical measurements (half cell potential and linear polarization resistance) were applied on cracked and uncracked concrete specimens. The results indicated that the crack is a dominating factor in corrosion of reinforcing bars in concrete. Experimental results from external polarization, which show a spatial variation relative to the crack, are presented. It is shown that the applied current generated by the polarization is confined to the active area of steel in the vicinity of the crack, irrespective of the size of the counter electrode. Keywords: corrosion; reinforced concrete; steel; crack; nondestructive tests; electrochemical techniques. Department of Civil Engineering, City College of New York, Convent Ave at 140th St New York, NY, 10031, USA ______________________________________ Email:
[email protected], Tel: (212) 650-6569
1. 0 Introduction Reinforced concrete is the most widely used construction material, especially in transportation infrastructure. Many of these concrete structures, especially highway bridges, exhibit early deterioration caused by corrosion. Corrosion damage in reinforced concrete is the leading factor contributing to the deterioration of the nation’s highway infrastructure. In 1997, it was estimated that the cost of corrosion damage in US highway bridges exceeded $150 billion. The annual expenditure for repair and rehabilitation of concrete structures in the 90s has exceeded fifty percent of the total construction costs [1] and it is expected that this trend will continue in the future. Detection and characterization of corrosion of steel in concrete, therefore, is very important for the condition assessment and initiation of timely repairs in reinforced concrete structures. Previous research on corrosion of steel reinforcement in concrete has primarily focused on corrosion in pristine (uncracked) concrete. The factors influencing corrosion and the methods for determining the rate of corrosion of steel embedded in concrete have been extensively researched [2-9]. Several electrochemical methods which provide for assessing the rate of corrosion, which is considered uniform along the entire length of the steel bar have also been developed. In these studies, it is implicitly assumed that crack in concrete is the result of expansion stresses due to corrosion product. The alkalinity of concrete provides good chemical protection to steel against corrosion. Therefore, in the absence of a crack, the steel reinforcement in reinforced concrete structures usually exhibits a prolonged corrosion initiation period. However, cracks are often found in concrete structures. These cracks are introduced due to the action of loads, restrained shrinkage or thermal gradients. Cracks result in a disruption in spatial continuity of the medium material. The main transport mechanism for chloride ingress in cracked concrete is convection due to capillary suction of deicing water, which is faster than the diffusion based-mechanism in uncracked concrete. Once a crack is formed in reinforced concrete, it provides an easy and fast access for ingress of ions such as chlorides, oxygen and water to the steel surface. The steel in the crack zone is exposed to the environment and is de-passivated, which results in a faster initiation of corrosion. Relatively little work has been done to study the corrosion of steel embedded in cracked concrete. A crack has been shown to produce spatial variations in the potential and the rate of corrosion of steel [10]. The corrosion rates have been shown to increase significantly in the vicinity of bending cracks, which intersect the main reinforcement [11]. Furthermore, it has been observed that the corrosion process continues through the formation of macrocell, where a small anode located at
the crack is supported by a large cathode comprising of passive steel in the uncracked concrete around the crack [10]. The macrocell mechanism has been shown to contribute significantly to the overall corrosion rate; at reinforcing bars intersecting the cracks, the macrocell rate was 70 times higher than the micro-cell rate [11]. The formation of a macrocell has been shown to result in localized corrosion. This localized corrosion in the presence of a crack has been shown to be significantly higher than that determined using standard electro-chemical methods from the concrete surface. Using a model system comprised of stainless steel cathodes immersed in various solutions, the average corrosion rate determined using conventional electrochemical measurements have been shown to under-estimate the local corrosion rate by a factor of five to ten, when the macro-cell mechanism is formed [12]. Some of the geometric and material parameters which could influence the rate of corrosion in cracked concrete have also been investigated. The cover depth and the quality of concrete have been shown to significantly influence the corrosion rate obtained from the macrocell mechanism. Berke reported that the corrosion rate in cracked concrete was reduced on increasing the concrete quality, increasing the cover depth or addition of a corrosion inhibitor [13, 14]. Arya studied the relationship between crack frequency and reinforcement corrosion and found that decreasing the frequency of cracking resulted in a decrease in corrosion [15]. Most methods for determining the rate of corrosion rely upon applying external polarization to the system undergoing corrosion. This process is complicated in the case of steel embedded in cracked concrete, which forms a macrocell due to the spatial variations in the resistance of the medium and the rate of corrosion along the length of the bar. Understanding the response of a steel bar undergoing corrosion, which results from the formation of a macro-cell subjected to external polarization is essential for developing procedures which allow estimating the local corrosion rate in such a situation. Sagues et al. developed numerical procedures for studying the response of a macrocell subjected to external polarization [16]. Through numerical simulations it has been shown that the excitation current generated by external polarization enters predominantly the active areas. This observation has been verified through laboratory experiments on model-macro-cells [4]. However, the response of steel bar embedded inside cracked concrete, which is subjected to external polarization have not been reported so far. Currently, the polarization studies on cracked concrete have not been reported. In this paper, results of an experimental program investigating the corrosion of concrete in cracked concrete are reported. The results pertaining to the influence of the crack on the external polarization response of steel embedded inside cracked concrete are presented. It is shown that
steel embedded in cracked concrete forms a macro-cell, which results in a spatial variation in the potential of the steel bar. Polarization measurements indicate that the steel close to the crack undergoes active corrosion, while the steel away from the crack is passive. Finally, it is shown that the low polarization resistance of the active area close to the crack, the applied current is confined to the active area, irrespective of the size of the counter electrode. Therefore, the polarization resistance estimated using a large counter electrode would under-estimate the local corrosion rate.
2.0 Objectives The objectives of the work presented in this paper are 1.
To study the influence of a crack on the corrosion of steel embedded in concrete.
2.
To determine the spatial variation in the potential produced by the crack
3.
To study the influence of external polarization (potential) on the response of the cracked
specimen.
3.0 Materials and Methods Two series of concrete specimen labeled A and B were cast; specimens A without a crack and B with a single crack in the middle of the specimen. Specimens B were intended to study the response of steel embedded in cracked concrete. Two specimens were cast for each series. The specimen geometry and dimensions are shown in Fig 1. Type 1 Portland cement (ASTM C 150) was used for all specimens. The cement: sand: aggregate ratio of concrete by weight was 1:1.70:2.42 and the water/cement ratio was 0.45. Ordinary tap water was used for mixing the concrete. An air entraining agent, MB-VR by Master builders (ASTM C 494), was used. The concrete cover depth for all specimens was equal to 30 mm. Plain carbon steel bars with diameter equal to 12.7 mm were used for all specimens. The steel bars were cleaned with acetone and polished using 600-grit Sic paper. Electrical connections were made by soldering a copper wire to the steel bar close to one of its ends. In specimens A, the length of the steel bar was equal to 250 mm, which is smaller than the overall length of the concrete specimen. The steel bar was completely encased in concrete after casting. In specimens A, the surface area of steel exposed to the concrete was determined to be equal to 100 cm2. In specimen B, both ends of rebar were sealed using an epoxy resin. Only the middle section of rebar with length equal to 1118 mm was left exposed to the concrete, which provided a rebar area approximately equal to 446 cm2. A crack was
introduced at the time of casting using a thin plastic sheet placed in the middle of the specimen. All specimens were cured in a 100% relative humidity (RH) chamber for 90 days following which they were subjected to periodic wetting and drying cycles. Each wetting-drying cycle involved three days of wetting and a four days drying period. During wetting, the specimens were subjected to 100%RH at 23ºC. The drying comprised of exposing the specimens to the laboratory environment, which was maintained at 50% RH and 23ºC. At 90 days age, specimens B were subjected to 3 points flexural to initiate a crack along the plastic sheet. During the loading procedure, as the crack faces moved apart, the plastic sheet was pulled out from the inside of the concrete. After the crack was created, the specimens were exposed to periodic wetting-drying cycles.
Fig 1: A schematic diagram of Specimen A and B
4.0 Measurements Half-cell potential and linear polarization measurements were performed on all specimens immediately after the wetting period. Before each electrochemical measurement, the concrete specimens were covered with a thin sponge soaked in 3% NaCl solution. In cracked specimens, NaCl solution was poured into the cracks prior to initiating readings. The sponge was used to keep the surface of the concrete wet and to provide an electrical connection between the counter electrode and concrete surface. 4.1 Half-cell Potential A saturated calomel electrode (SCE) was used to measure the potentials at the surface of concrete. The half-cell potential is an indication of relative probability of corrosion activity, and the
guideline described in ASTM C 876 provides general principles for evaluation of the reinforcing steel corrosion in concrete using the potential measurement with respect to SCE as shown in Table 3-4. In specimen B, the half cell readings were taken at the location of the crack and at different locations away from the crack. Table 1: Guideline for half-cell potential data interpretation Calomel (SCE) E>-126mV
Interpretation Greater than 90% probability that no corrosion is occurring
-126mV < E < -276mV
E
