Elsevier

Corrosion Science

Volume 52, Issue 7, July 2010, Pages 2469-2480
Corrosion Science

On the penetration of corrosion products from reinforcing steel into concrete due to chloride-induced corrosion

https://doi.org/10.1016/j.corsci.2010.03.025Get rights and content

Abstract

Reinforced concretes were corroded to varying degrees by exposing to cyclic NaCl spray and 40 °C drying. The amount of corrosion products and induced damage were measured using image analysis. We found that corrosion products can accumulate at steel–concrete interface as well as penetrate cement paste and deposit within hydration products, relicts of reacted slag, and air voids. As corrosion increases, the products tend to accumulate at the steel–concrete interface, while the amount penetrating cement paste remains relatively constant. Only a small amount of corrosion is needed to induce visible cover cracking. Implications on modelling time to cover cracking are discussed.

Introduction

Corrosion of steel in reinforced concrete is an electrochemical process that causes the dissolution of iron to form a range of solid products. Corrosion products is a complex mixture of iron oxides, hydroxides and hydrated oxides that evolves according to the prevailing local environment [1], [2], [3]. Depending on their level of oxidation and the availability of moisture, corrosion products have specific volumes ranging from about two to six times that of the iron consumed [4], [5]. As such, the main damage caused by corrosion of reinforcement in concrete is not the loss of steel cross-section, but cracking of the concrete cover due to expansive stresses exerted by the continued deposition of corrosion products near the steel–concrete interface. This leads to progressive deterioration and even spalling of the cover.

However, not all corrosion products contribute to the build-up of stresses and initial cracking of the cover. Despite the apparent insolubility of the final corrosion products, the process of corrosion involves soluble species that can dissolve in the concrete pore solution and subsequently migrate or diffuse through the cement paste matrix away from the corroding steel [6], [7]. In the presence of chloride ions, the dissolution of iron is increased by formation of intermediary chlorocomplexes, i.e. ‘green-rusts’ [9]. These soluble complexes migrate away from the anodic sites, and subsequently break down in oxygen-rich areas, releasing the chlorides to transport more ferrous ions from the corroding steel [4], [8], [9]. The mobility of these soluble species and subsequent precipitation in oxygen-rich areas results in the corrosion products being distributed within the porous cement paste.

The ability of the pore structure (capillaries, microcracks and air voids) to accommodate corrosion products, at least in the initial stages of corrosion, may have an effect of prolonging the time to cracking of the concrete cover. The pores near the corroding steel act as repositories, hence extending the period between onset of corrosion and pressure development that leads to damage of the concrete cover. This effect has been shown experimentally for example, by using hydraulic pressurisation to simulate the corrosion and damage processes in reinforced concrete [10]. Alonso et al. [11] observed that an increase in porosity (w/c ratio) produces a delay in crack initiation and this was attributed to the higher amount of void space available to accommodate corrosion products without stress generation. Indeed, many recent theoretical models for predicting time to initial cover cracking have incorporated a ‘free-expansion’ step to account for the time needed for corrosion products to fill the porous cement paste around the rebar. For example, Liu and Weyers [12] assumed the existence of a 12.5 μm thick ‘porous zone’ around the steel that must be filled with corrosion products prior to any development of expansive pressure. This has been adopted in many subsequent works by others (e.g. [13], [14]) although without any experimental verification. In another study [15], a 40 μm thick ‘porous zone’ with the porosity estimated using Powers’ model was proposed instead.

Very little previous research has examined the amount and distribution of corrosion products and their effect on the microstructure of the steel–concrete interface and contribution to cover cracking. Previous studies, using various analytical and microscopy techniques, have shown that corrosion products forms in the cement paste adjacent to a corroding rebar (e.g. [7], [16], [17], [18], [19], [20], [21], [22]). Various studies of corrosion in archaeological artefacts containing metallic reinforcements found that the steel–binder interface contains a dense inner corrosion product layer and a transformed medium that contains corrosion products and elements from the binder [23], [24]. However, there have been no rigorous attempts to quantify the extent and distribution of the corrosion products accumulation at the steel–concrete interface and penetration into the adjacent cement paste, and their damaging effects. Presumably, the amount of corrosion products that can be accommodated during free expansion depends on the volume and connectivity of the pore structure, which is in turn related to factors such as mix composition, degree of hydration and compaction. Thus, assuming a constant value for all cases in modelling appears to be an oversimplification, but a more refined approach is not currently possible since fundamental understanding of this phenomenon is lacking. Clearly, research is needed in this area in order to advance service-life prediction models based on estimating time to cover cracking.

The objective of this work is to characterise the amount and distribution of corrosion products, particularly during the transition period between corrosion initiation and the development of significant damage. Our study is based on reinforced concrete samples that have been subjected to salt spray on one exposed surface to induce corrosion and cover cracking. It was decided that corrosion should not be accelerated by admixed chlorides, imposed electrical current or manufactured cracks, because such treatments would artificially influence the formation and distribution of the corrosion products. The samples examined should span a range of degrees of corrosion; however, severely deteriorated samples are excluded because their microstructure would have been significantly altered since the time of crack initiation. To obtain objective and quantitative information, measurements should be made on many flat, random cross-sections of the corroded rebar–concrete interface, but sample preparation can be challenging and problematic. Irregular, fractured surfaces, although easier to study, are not representative because these tend to expose more of the damaged and corroded areas. Fractured surfaces also do not allow accurate measurements to be made.

Section snippets

Materials

The samples examined were taken from reinforced concrete panels that contained three 16 mm nominal diameter deformed (corrugated) carbon steel bars at 20 mm cover spacing (Fig. 1). The steel had a composition of C: 0.14–0.22%, Si: 0.12–0.30%, Mn: 0.30–0.65%, P < 0.045% and S < 0.050%. The rebars were used as-received and no efforts were made to remove existing millscale. More than one thousand panels were prepared as part of a long-term research project on chloride-induced corrosion in a cyclic

Visual inspection

Visual inspection of the slices immediately after cutting using a stereomicroscope found that the colour of the corrosion products is mainly dark brownish-red, although in some areas the corrosion products tend to be of a brighter red hue. The slices exhibited various amounts of corrosion and cracking, but it was clear that those taken from near the side face of the panels were more affected than farther away (Fig. 3). In every slice, the corrosion products at the steel–concrete interface are

Discussion

When steel in concrete corrodes, the oxidation of iron in the anodic region produces soluble ferrous ions (Fe2+), balanced by reduction of oxygen in the cathodic region that forms hydroxyl (OH) ions. The presence of chloride ions greatly influences the mechanism and kinetics of the corrosion process. The exact mechanism is believed to involve soluble intermediary ‘green complexes’ [9]. These ions dissolve in the concrete pore solution, which enables them to move through the pore structure of

Conclusions

Corrosion of steel in reinforced concrete produces soluble corrosion products that can migrate or diffuse through the cement paste. This study observed that the corrosion products can be deposited in air voids, in the outer and inner hydration products and in rims and relicts of reacted slag. A distinct boundary between the affected and unaffected paste areas can be seen in BSE images, indicating the extent of the penetration front. The corrosion products penetrates the aggregate-paste ITZ as

Acknowledgements

H.S.W. acknowledges the financial support provided by the Engineering and Physical Sciences Research Council, UK. Y.X.Z. would like to thank the National Natural Science Foundation of China (Grant No. 50538070 and 50808157) for supporting her academic visit to Imperial College. We wish to express our thanks to Mr. Andrew Morris for his help in preparing the samples for microscopy work.

References (49)

  • H.S. Wong et al.

    Monte Carlo simulation of electron–solid interactions in cement-based materials

    Cem. Concr. Res.

    (2006)
  • K.O. Kjellsen et al.

    X-ray microanalysis of hydrated cement: is the analysis total related to porosity?

    Cem. Concr. Res.

    (1998)
  • C. Famy et al.

    What causes differences of C–S–H gel grey levels in backscattered electron images?

    Cem. Concr. Res.

    (2002)
  • W.-J. Chitty et al.

    Long-term corrosion of rebars embedded in aerial and hydraulic binders—mechanisms and crucial physico-chemical parameters

    Corros. Sci.

    (2008)
  • I.G. Richardson

    The nature of C–S–H in hardened cements

    Cem. Concr. Res.

    (1999)
  • C. Famy et al.

    The C–S–H gel of Portland cement mortars: part 1. The interpretation of energy-dispersive X-ray microanalysis from scanning electron microscopy, with some observations on C–S–H, AFm and Aft phase compositions

    Cem. Concr. Res.

    (2003)
  • H.F.W. Taylor et al.

    An electron microprobe study of a mature cement paste

    Cem. Concr. Res.

    (1984)
  • J.I. Escalante-Garcia et al.

    The chemical composition and microstructure of hydration products in blended cements

    Cem. Concr. Compos.

    (2004)
  • G.K. Glass et al.

    Corrosion inhibition in concrete arising from its acid neutralisation capacity

    Corros. Sci.

    (2000)
  • G.K. Glass et al.

    The participation of bound chloride in passive film breakdown on steel in concrete

    Corros. Sci.

    (2000)
  • P. Gégout et al.

    Action of chloride ions on hydrated cement pastes: influence of the cement type and long time effect of the concentration of chlorides

    Cem. Concr. Res.

    (1992)
  • A. Delagrave et al.

    Influence of chloride ions and pH level on the durability of high performance cement pastes (part II)

    Cem. Concr. Res.

    (1996)
  • E. Revertégat et al.

    Effect of pH on the durability of cement pastes

    Cem. Concr. Res.

    (1992)
  • D. Talbot et al.

    Corrosion Science and Technology

    (1998)
  • Cited by (243)

    • A Dissolvable Micromechanics Model for Composites

      2024, International Journal of Mechanical Sciences
    View all citing articles on Scopus
    View full text