Up to Date Overview of Aspects of Steel Reinforcement Corrosion in Concrete – Warren Green, Frank Collins and Maria Forsyth (2017)
Professor Brian Cherry has always been a firm believer in understanding first the fundamentals of any aspects of corrosion science, then the mechanisms, before embarking on the engineering of solutions to the management of materials corrosion. So it has been with steel reinforced concrete. This paper endeavours to walk a reader through the fundamental and mechanistic aspects of the excellent protection afforded to steel reinforcement by concrete (including electrochemistry), corrosion of steel reinforcement (uniform (microcell/mini-cell) and pitting (macrocell) corrosion, corrosion products composition and development), chloride induced corrosion mechanisms (passive film breakdown/pit initiation, metastable pitting, pit growth/propagation,chemical conditions within propagating pits and reinforcing steel quality effects), carbonation induced corrosion mechanisms, leaching induced corrosion of reinforcement and reinforcing steel stray current corrosion and interference.
Professor Brian Cherry in his education of students and practitioners has always ensured that fundamental aspects are conveyed. As a result, it is appropriate to begin this symposium with an overview of key fundamentals and mechanisms when it comes to the protection of reinforcement by concrete. There are several degradative processes which affect some reinforced concrete structures and amongst these, the most common cause of deterioration is the corrosion of steel reinforcement. An up-to-date overview of the key fundamentals and mechanisms of corrosion of steel reinforcement in concrete is also considered warranted.
Much is expected of some of our concrete structures and buildings, some are in extremely aggressive environments, they may be many decades of age, yet still expected to meet service life, they are of critical importance in terms of function or location, it is impossible to replace them, etc such that some require repair and protection during their service lives. Dr John Broomfield will provide an overview of repair and protection aspects relating to corroding steel in concrete in the second paper of this symposium.
The subsequent papers and speakers at the symposium then address other aspects of reinforced concrete corrosion, protection repair and durability that Professor Brian Cherry has taught, researched and provided innovative solutions to over many decades, including:
- Condition assessment of structures;
- Modelling and deterioration prediction;
- Concrete repair and protection;
- Cathodic protection;
- Corrosion inhibitors; and
- Alternate metallic reinforcement.
2. STEEL REINFORCED CONCRETE
Concrete is the most widely used material of construction in the world, with 25 gigatonnes/year consumed globally. Concrete is strong in compression but weak in tension and so much of the concrete is reinforced, usually with steel. The steel reinforcement can take the form of conventional carbon steel (black steel), prestressing steel, post-tensioned steel and steel fibres and its widespread utility is primarily due to the fact that it combines the best features of concrete and steel. The two complement one another and so by combining them together a composite that has good tensile strength, shear strength and compressive strength combined with durability and fire resistance can be formed.
When suitably designed, constructed and maintained, reinforced concrete provides service lives of numerous decades to structures and buildings. Concrete provides reinforcing steel with excellent corrosion protection. The highly alkaline environment in concrete results in the spontaneous formation of a stable, tightly adhering, thin protective oxide film (passive film) on the steel reinforcement surface, which protects it from corrosion. In addition, well proportioned, compacted and cured concrete has a low penetrability, thereby minimising the ingress of corrosion-inducing species via the aqueous phase. It also has a relatively high electrical resistivity, which reduces the corrosion current and hence the rate of corrosion if corrosion is initiated.
There are however, several degradative processes which affect some reinforced concrete structures leading to loss of functionality, unplanned maintenance/remediation/replacement, and in the worst cases, loss of structural integrity and resultant safety risks. Amongst these, the most common cause of deterioration is corrosion of conventional carbon steel (black steel), prestressing steel and post-tensioned steel reinforcement.
3. PROTECTION AFFORDED TO STEEL REINFORCEMENT BY CONCRETE
3.1 Portland Cement and Blended Cement Binders
The hydraulic binder of concrete commonly consists of Portland cement or of mixtures of Portland cement and one or more of fly ash, ground granulated iron blast-furnace slag or silica fume (Standards Australia, 2010). The latter are referred to as Blended cements.
AS 3972 (2010) defines Blended cement as a hydraulic cement containing Portland cement and a quantity comprised of one or both of the following:
- Greater than 7.5% of fly ash or granulated iron blast-furnace slag, or both.
- Up to 10% silica fume.
Furthermore, AS 3972 (2010) designates cements as general purpose and special purpose. By definition, general purpose cements may be portland cements (Type GP) or blended cements (Type GB). Special purpose cements are also defined in AS 3972 (2010) and may be portland or blended cements with restrictions being placed on their composition.
Portland cement consists (nominally) of a mixture of oxides:-
CaO, SiO2, l2O3, Fe2O3, MgO, Na2O and K2O.
These are however combined together as a series of cement compounds.These cement compounds are themselves combinations of the principle oxides (Cherry & Green, 2017).
Fly ash is the solid material extracted from the flue gases of a boiler fired with coal (Standards Australia, 2016). Fly ash for use in cement needs to comply with AS 3582.1 (2016). The fly ash content of a FA blended cement based concrete varies but is commonly in the range 20-35%. For example, some Road Authorities in Australia stipulate in their structural concrete specifications a minimum 25% FA content (Roads & Maritime Services, 2013; VicRoads, 2017) so as to achieve increased durability.
Granulated iron blast-furnace slag is the glassy non-metallic granular material, essentially consisting of silicates, aluminosilicates and calcium, resulting from the rapid chilling of molten iron blast-furnace slag (Standards Australia, 2016). Ground granulated iron blast-furnace slag (BFS) for use in cement needs to comply with AS 3582.2 (2016). The slag content of a BFS blended cement based concrete varies but is commonly in the range 50-70%. For example, some Road authorities in Australia stipulate in their structural concrete specifications a minimum 50-65% BFS content (Roads & Maritime Services, 2013; VicRoads, 2017) so as to achieve increased durability.
Silica fume is a very fine pozzolanic material produced by electric arc furnaces as a by-product of the production of elemental silicon or ferro-silicon alloys. Silica fume is also known as ‘condensed silica fume’ and ‘microsilica’. The amorphous silica content of a silica fume (SF) blended cement based concrete typically varies between 5-15%. For example, some Road Authorities in Australia stipulate in their structural concrete specifications a 10% SF proportion (VicRoads, 2017).
Triple blend cement based concretes are also used for increased durability. For example VicRoads in their “Structural Concrete Specification 610” (2017) stipulate that in a triple blend concrete mix, the Portland cement shall be a minimum of 60% and the individual contribution of Slag, Fly Ash or Amorphous Silica shall be a maximum of 40%, 25% or 10% respectively. For marine durability, triple blend cement based concretes such as 52% shrinkage-limited (SL) cement, 25% FA and 23% BFS have been used (Green et al, 2009).
3.2 Alkaline Environment in Concrete
As mentioned above, the highly alkaline environment in concrete results in the spontaneous formation of a passive iron oxide film on reinforcing steel which protects it from corrosion.
The reaction of the cement compounds of Portland cement or Blended cement with water results in the setting and hardening of the cement paste so that it binds the aggregate (coarse and fine) of the concrete together. A product of the hydration of Portland and Blended cements is Ca(OH)2 together with NaOH and KOH.
As a result, the pH of the pore solution of concrete is normally in the range of 12-14 (Page & Treadaway, 1982; Tuutti, 1982; Tinnea & Young, 2000; Tinnea, 2002; Broomfield, 2007; Abd El Haleem et al, 2010; Ghods, et al, 2011; Cherry & Green, 2017), and maintained due to the CaO-Ca(OH)2 pH buffer until this is overcome by, say, carbonation, refer later, reduced by, say, leaching, refer later, or neutralised by chloride ion induced pitting corrosion reactions, refer later.
For iron (steel) in a pH 12-14 alkaline environment, the Potential – pH diagram, see Figure 1, tells us that the metal will be passive.
3.3 Physical Barrier Provided by Concrete
In addition to providing a high pH passivating environment for reinforcing steel, concrete also provides a physical barrier against the ingress of corrosion-inducing substances.
The quality of concrete as a physical barrier may be assessed from penetrability data (and chemical data, i.e. blended cements have better chloride ion binding/adsorption capacity as well as improving the porosity of the aggregate-paste transition zone). Penetrability data is collected from measurements of a concrete’s water absorption (various tests), water permeability (various tests), chloride diffusion (various tests) and gaseous diffusion/penetration (O2 and CO2, various tests) (Concrete Institute of Australia, 2015).
The factors affecting concrete penetrability include the type and quality of aggregates, type of cement (binder), cement (binder) content, water/cement (binder) ratio and production variables such as mixing uniformity, placement, degree of compaction and adequacy of cure.
3.4 Passivity and the Passive Film on Steel Reinforcement in Concrete
3.4.1 Thermodynamics of Passivity
Thermodynamics, the science of energy changes, is applied to corrosion studies to determine why a particular metal does or does not tend to corrode in a particular environment.
Chemical and electrochemical thermodynamic data [chemical potential (µ0) values, standard free energy change of reaction ( △G0) values, and standard electrode potential (E0) values] provides the means for deciding which of a set of reactions is thermodynamically favoured and for predicting the most stable reaction products under specified conditions of electrode potential and solution composition.
Marcel Pourbaix (1904-1998) recognised that the corrosion state of a metal could be represented by a point on a diagram the two axies of which were the potential and pH. These diagrams are now often termed ‘Pourbaix’ diagrams.
The Pourbaix diagram plots electrochemical stability for different redox states of an element as a function of pH. The Fe-H2O system at 25°C, shown in Figure 1, is of most relevance to reinforcing steel in concrete. Typically the concentrations of dissolved ions, other than H+ and OH–, are taken to be 10-6M and solid phases are assumed to be pure.
If the potential of iron immersed in neutral water (pH about 7) is measured using a standard hydrogen electrode (SHE), it will be seen in Figure 1 to have a value represented by, say – 0.3V. An examination of the Pourbaix diagram at Figure 1 suggests three possible means of achieving a reduction in the corrosion. First, the metal can have its potential so changed in the negative direction that it enters the domain of immunity, i.e. it can be cathodically protected. Secondly, the potential can be changed in the positive direction so that it enters the passive domain. This is anodic protection and is only applicable to some metals such as iron but not for example to zinc. Thirdly, the pH of the electrolyte can be so adjusted that the metal enters the passivity domain, and this process is considered inhibition.
To summarise, Figure 1 shows us that there are three basic zones of behavior representing states of lowest energy for iron (steel):
- Immunity – iron metal is thermodynamically stable and is immune to corrosion.
- Corrosion – Fe2+, Fe3+ and HFeO2– ions are thermodynamically stable and corrosion will occur at a rate which cannot be predicted thermodynamically.
- Passivity – iron oxides are thermodynamically stable. These oxides give rise to a condition termed passivity since significant corrosion may be stifled owing to the formation of a protective oxide layer on the metal surface (see shaded area on Figure 1 in the high pH region).
Having described the information that can be gleaned from Pourbaix diagrams with respect to deciding which of a set of reactions is thermodynamically favoured and which reaction products are most stable, it is necessary to recognise that certain limitations apply to the use of thermodynamics and Pourbaix diagrams. The kinetics (rates) of possible reactions are not considered, and therefore, it is impossible to predict whether a particular reaction, which is thermodynamically favoured, will occur at a significant rate in practice. The aqueous solution composition in the vicinity of the corroding surface also needs to be known. This presents difficulties when considering real corroding systems in which concentration gradients may be developed.
3.4.2 Kinetics of Positivity
Figure 2 shows the anodic polarisation curve for a passive metal, such as reinforcing steel in the alkaline environment of concrete.
At potentials more negative than the equilibrium potential or reversible potential for Fe/Fe2+ (i.e. Eo Fe/Fe2+), iron is immune to corrosion or dissolution. Raising the potential to values more positive than Eo Fe/Fe2+ leads to active corrosion at a rate which initially increases with increasing overpotential (maximum corrosion current, I1). At some potential more positive or noble than the equilibrium potential for the formation of a surface oxide film (i.e. Eo Fe/Fe2O3) active corrosion practically ceases (i.e. I2) owing to surface oxide film formation which renders iron passive and allows the passage of a small ‘leakage current’ (i.e. I2→I3). This ‘leakage current’ results from the necessity to reform the surface oxide passive film at points of local breakdown and to replace oxide lost by dissolution to the environment.
3.4.3 Passive Film Foundation
The corrosion of steel reinforcement in
concrete is, initially at least, a very slow process. As a result of the
hydration reaction of the cement (binder) compounds as previously noted, the
pore water surrounding the steel has a high pH (in the range
12-14) and so the steel is completely covered with a dense gamma ferric oxide (γ-Fe2O3, maghemite) protective passive film. This is formed by the anodic reaction (1):
4Fe + 12OH– → 2γFe2O3 + 6H2O + 12e– (1)
Once the film has formed, the reaction continues (at a very slow rate) at the metal/oxide interface as a result of hydroxyl ions diffusing through the film of iron oxide. The cathodic process is ascribed to the reaction (2):
3O2 + 6H2O + 12e– → 12OH– (2)
This reaction takes place on the surface of the passive film. The cathodic process requires that oxygen molecules diffuse through the concrete to the passive film surface and are there reduced to hydroxyl ions. The electrical circuit is completed by the ionic transport through the passive film of the hydroxyl ions resulting from the cathodic reaction moving towards the metal/film interface and by electronic transport through the protective film to the film/electrolyte interface. Corrosion is proceeding in this state corresponding to the conversion of iron to iron oxide.
Passive (leakage) corrosion current densities (icorr) measured for steel reinforcement in concrete are of the order of 0.1µA/cm2 (Andrade & Gonzalez, 1978; Hansson, 1984). This current density therefore corresponds to an insignificantly slow corrosion (penetration) rate of 1µm/year. This corrosion rate is of the order of 1mm in 1000years and so it can be seen that the reinforced concrete structure or building element may be regarded as unaffected by corrosion. The steel in this state is said to be “passive” or protected by a “passive film” (Cherry & Green, 2017).
Since the passive film maintains a constant thickness (it dissolves from the outside as quickly as it is formed on the inside) no great swelling stresses are applied to the concrete and so no cracking of the concrete takes place (Cherry & Green, 2017).
3.4.4 Passive Film Composition
The precise nature of the passive film which is formed by the reaction of iron with the highly alkaline environment of the pore water in concrete varies. As above, it can be considered to be generally a dense gamma ferric oxide (γ-Fe2O3, maghemite).
The characteristics of oxides formed on steel under alkaline conditions and their electrochemical behavior have been examined by, for example, Nasrazadani (1997), Cornell and Schwertmann (2000) and Freire et al (2009). Iron-oxy-hydroxides like goethite (a-FeOOH), lepidocrocite (γ-FeOOH) and akagonite (β-FeOOH) below which more protective forms of iron oxides including magnetite (Fe3O4), maghemite (γ-Fe2O3) and haematite (a-Fe2O3) are assumed to be the constituents of the passive film formed on steel electrodes under alkaline conditions.
Cohen (1978) indicated that the composition of the passive oxide layer is a spinel a-Fe3O4-γ-Fe2O3 solid solution. Sagoe-Crentsil and Glasser (1989) attributed the passive action of pH>12 concrete pore solution to the formation of a surface layer of Fe2O3-Fe3O4. Ghods et al (2011) have more recently also reported two-layer oxides in passive films, with an inner Fe2+ and Fe3+ layer and an outer pure Fe3+ layer. Al-Negheimish et al (2014) have also proposed a two-layer oxide passive film but the top layer was composed of FeO and FeOOH with Fe2O3 close to the steel surface.
More recently Ghods et al (2013) report of a tri-layer model of oxides for the passive film on the steel when exposed to simulated concrete pore solution namely a layer of FeO at the metal surface, then a layer of Fe3O4 and then an outer layer of Fe2O3.
MnS inclusions are often found in steels and their effect is to change the Fe2O3 in the passive film from γ-Fe2O3 to a-Fe2O3 according to Al-Negheimish et al (2014).
3.4.5 Passive Film Thickness
Singh and Singh (2012) proposed that the passive film formed in simulated concrete pore solution is of “ultra thin thickness” (i.e. <10 nm).
Ghods et al (2013) reported that the thickness
of the tri-layer (FeO/Fe3O4/Fe2O3)
passive oxide film was uniform between
5 and 13 nm.
Al-Negheimish et al (2014) have determined thicknesses of 6 to 7 nm for the passive film on steel in simulated concrete pore solution.
3.4.6 Models and Theories of Passive Film Development on Iron
Veluchamy et al (2017) indicate that the understanding of the passive state of metals started with Faraday (1844). Though several theories, models and experimental works on passivity have been published in the literature, the mechanisms underlying the stability of the passive oxide over the metal still remains a mystery according to Veluchamy et al (2017). They undertook a detailed review of theoretical and experimental results for the iron/electrolyte system invoking the high field model (HFM) (ion-migration mechanism), modified high field model, point defect model (PDM), variants of PDM (VPDM), diffusion Poisson coupled model (DPCM) and the density functional theory based atomistic model. The experimental and model-predicted dependencies on applied voltage, pH, chloride and temperature have also been presented and discussed by Veluchamy etl al (2017).
Suffice to say that the models and theories of passive film development on iron are complex and still not clear, thus discussion of such is beyond the scope of this paper.
4. CORROSION OF STEEL REINFORCEMENT
The passivity provided to steel reinforcement by the alkaline environment of concrete may be lost if the pH of the concrete pore solution falls because of carbonation or if aggressive ions such as chlorides penetrate in sufficient concentration to the steel reinforcement surface. Carbonation of concrete occurs as a result of atmospheric CO2 gas (and atmospheric SOx and NOx gases) neutralising the concrete pore water (lowering its pH to 9) and thereby destroying the passive film. Leaching of Ca(OH)2 (and NaOH and KOH) from concrete also lowers pH to cause corrosion of steel reinforcement. Stray electrical currents, most commonly from electrified traction systems, can also breakdown the passive film and cause of corrosion of steel reinforced and prestressed concrete elements.
The corrosion mechanisms by chlorides, carbonation, leaching and stray electrical currents is provided in subsequent sections but discussion of the different forms of corrosion, composition of reinforcing steel corrosion products (rusts) and the consequence of the corrosion products is provided herewith.
4.2 Uniform (microcell) Corrosion and Pitting (macrocell) Corrosion
According to the different spatial location of anodes and cathodes, corrosion of steel in concrete can occur in different forms (Elsener, 2002):
- As microcells, where anodic and cathodic reactions are immediately adjacent, leading to uniform steel (iron) dissolution over the whole surface. This uniform (or general) corrosion is typically caused by carbonation of the concrete or by very high chloride content at the steel reinforcement.
- As macrocells, where a net distinction between corroding areas of the
steel reinforcement (anodes) and
non-corroding passive surfaces (cathodes) is found. Macrocells occur mainly in the case of chloride induced corrosion (pitting) where the anodes are small with respect to the total (passive) steel reinforcement surface.
Cherry and Green (2017) also suggest that uniform or general corrosion of steel reinforcement in concrete where the cathodic and anodic sites may be separated by millimetres (rather than microns), can be termed mini-cell corrosion.
4.3 Corrosion Products Composition – Chloride Induced Corrosion
The exact nature of reinforcing steel corrosion products (rusts) associated with chloride induced corrosion varies markedly depending on conditions. Corrosion products are of various layers and of various compositions. Generally speaking they will be layers and combinations of ferrous and ferric hydroxides, hydroxyl/oxides and oxides each with possible varying degrees of hydration.
Melchers and Li (2008) and Pape and Melchers (2013) have identified goethite (a-FeOOH), akagenite (β-FeOOH), lepidocrocite (γ-FeOOH) as corrosion products within chloride contaminated reinforcing steel samples from concrete structures. “Green rusts” can also occur. Melchers and Li (2008) and Pape and Melchers (2013) have identified compounds such as iron oxide chloride (FeOCl), hibbingite (a-Fe2(OH)3Cl), iron chloride hydrate (FeCl2.4H2O), “Green Rust I” (carbonate variety) and “Green Rust II” (sulphate variety) within the corrosion products (rusts) of samples taken from steel reinforced and prestressed structures in chloride-rich environments. Discussion of chloride induced corrosion mechanisms is provided at Section 5.
For corroding mortar specimens subject to chloride ingress, Koleva et al (2006) identified corrosion products mainly consisting of highly crystallised goethite (a-FeOOH), lepidocrocite (γ-FeOOH) and akagenite (β-FeOOH). Depending on the ratio of iron and chloride ions, the iron oxychlorides and iron oyhydroxides present different morphologies and exert influences on steel/mortar interface microstructure and on material behavior.
The corrosion products generally detected by Vera et al (2009) from embedded steel in concrete cylinder samples exposed to simulated marine and industrial conditions and a natural marine atmospheric environment were:- lepidocrocite (γ-FeOOH), goethite (a-FeOOH) and magnetite (Fe3O4); but in the chloride contaminated environments the presence of akagenite (β-FeOOH)was detected; and, in the natural marine atmospheric environment the formation of siderite (FeCO3) was also observed.
4.4 Corrosion Products Composition – Carbonation Induced Corrosion
Attack of concrete by carbon dioxide is termed carbonation and the mechanisms of such are provided at Section 6, suffice to say here that in terms of carbonation-induced corrosion products, Kolio et al (2015) on studies performed on existing reinforced concrete facades of 12 buildings that had been in use for 30-43 years identified corrosion products that were mostly hydroxide type of rusts (i.e. goethite/a-FeOOH and lepidocrocite/γ-FeOOH).
Huet et al (2005) identified corrosion products mainly composed of magnetite (Fe3O4) and lepidoocrocite (γ-FeOOH) in laboratory based studies of mild steel in carbonated concrete pore solution.
4.5 Corrosion Products Development – Cracking, Delamination and Spalling
The process of reinforcement corrosion will lead to corrosion products which, generally speaking, will occupy a greater volume than the iron dissolved in its production, refer Figure 4 (Jaffer & Hansson, 2009). Furthermore, when the corrosion products become hydrated the volume increase is even greater (Broomfield, 2007). Refer hydrated haematite (a-Fe2O3.3H2O) (red rust) in Figure 3.
The consequence of higher volume corrosion products is then to develop tensile stresses in the concrete covering the reinforcement. Concrete, being weak in tension, will crack as a consequence of the corrosion. Continuing formation of corrosion product(s) will enhance the expansion which will ultimately lead to cracked pieces of concrete cover detaching, leading to delamination and then spalling. Rust staining of the concrete may or may not occur together with the cracking or as a prelude to delamination and spalling. Section loss, bond loss and anchorage loss of the reinforcement also occurs as a result of the corrosion.
Kolio et al (2015) for example, describing work they had carried out on carbonation initiated corrosion on concrete façade panels, determined that the corrosion products (i.e. as previously discussed, goethite/a-FeOOH and lepidocrocite/γ-FeOOH) had a unit volume increase of roughly 3 times the volume of iron. Furthermore, they identified by taking into account the relative volume of the corrosion products that the required corrosion penetration (metal loss) to initiate visually observable cracks in the studied building façade panels was by average 53.6 µm with a corresponding corrosion (rust) products thickness of 111.7 µm.
Hydrated haematite (a-Fe2O3.3H2O) (red rust) induced cracking, delamination and spalling of cover concrete can be frequently seen in deteriorated reinforced concrete structures and buildings, an example of which is provided at Figure 4 for a marine structure.
4.6 Corrosion Products Development – No Visible Damage
It is most important to note however, that not all corrosion of reinforcement leads to rust staining, cracking, delamination or spalling of cover concrete. Localised pitting, localised corrosion at cracks, localised corrosion at concrete defects, etc can result in marked section loss (loss of bond, loss of anchorage) and ultimately structural failure without the visible consequences of corrosion on the concrete surface, i.e. no rust staining, cracking, delamination or spalling of cover concrete (Green et al, 2013). An example of marked localised section loss of reinforcement due to chloride induced pitting (macrocell) corrosion is provided at Figure 5. The definition of macrocell corrosion being as at Section 4.2 where a net distinction between corroding areas of the steel reinforcement (anodes) and non-corroding passive surfaces (cathodes) is found.
Elsener (2002) pointed out that macrocell (pitting) corrosion is of great concern because the local dissolution rate (reduction in cross-section of the conventional/prestressing/post-tensioned steel reinforcement, loss of bond, loss of anchorage) may greatly be accelerated due to the large cathode/small anode area ratio. Indeed, values of local corrosion (penetration) rates of up to 1 mm/year have been reported for bridge decks, sustaining walls or other chloride contaminated steel reinforced concrete structures according to Elsener (2002).
Broomfield (2007) also notes that ‘black rust’ or ‘green rust’ (due to the colour of the liquid when first exposed to air after breakout) corrosion can also be found under damaged waterproof membranes and in some underwater or water saturated structures. He states that it is potentially dangerous as there is no indication of corrosion by cracking and spalling of the concrete and the reinforcing steel may be severely weakened before corrosion is detected. Reinforcement bars may be “hollowed out” in such deoxygenated conditions.
5. CHLORIDE INDUCED CORROSION MECHANISMS
It is known that chloride ions in sufficient concentration can destroy the passivity of steel reinforcement (conventional, prestressed and post-tensioned) in concrete leading to pitting corrosion. Pitting corrosion is localised accelerated dissolution of metal that occurs as a result of breakdown of the otherwise protective passive film on the metal surface (Frankel, 1998). Various mechanisms of chloride ion induced corrosion of steel reinforcement in concrete are proposed in the literature.
Frankel (1998) comments that many engineering alloys, such as stainless steels and aluminium alloys, are useful only because of passive films, which are thin (nanometer scale), oxide layers. Furthermore that such passive films however, are often susceptible to localised breakdown resulting in accelerated dissolution of the underlying metal and if the attack initiates on an open surface, it is called pitting corrosion and at an occluded site it is called crevice corrosion.
Fundamental studies of pitting corrosion on such engineering alloys appear to have typically focused on the various stages of the pitting process, namely: passive film breakdown, the growth of metastable pits (which grow to about the micron scale and then repassivate) and the growth of larger, stable pits (Frankel, 1998).
Like Frankel (1998), Angst et al (2011) propose from a fundamental viewpoint that chloride induced pitting corrosion of steel reinforcement in concrete can be considered to occur as a sequence of distinct stages:- pit nucleation, metastable pitting and stable pit growth.
The mechanisms of pitting corrosion of steel in chloride contaminated concrete will therefore be considered on a stages basis as well as discussion of some other mechanistic aspects relating to the reinforcing steel quality itself.
5.2 Passive Film Breakdown/Pit Initiation
The American Concrete Institute (ACI) Committee 222 on ‘Corrosion of Metals in Concrete’ (1985) proposes three theories to explain the effects of chloride ions on steel reinforcement:
- The Oxide Film Theory – postulates that chloride ions penetrate the passive film through pores or defects in the film, thereby “colloidally dispersing” the film.
- The Adsorption Theory – postulates that chloride ions are adsorbed preferentially on the metal surface in competition with dissolved oxygen or hydroxyl ions. The chloride ion promotes the hydration of the metal atoms, thus facilitating anodic dissolution.
- The Transitory Complex Theory – postulates that chloride ions form a soluble complex with ferrous ions which then diffuses away from the anodic sites. Thereafter, the complex decomposes, iron hydroxide precipitates, and the chloride ions are released to transport more ferrous ions from anodic sites.
Frankel (1998), in a comprehensive review paper of pitting corrosion of metals, indicates that theories for passive film breakdown and pit initiation of pure metal systems (i.e. not alloys where pitting may also be associated with inclusions or second-phase particles) have been categorised in three main mechanisms that focus on passive film penetration, film breaking, or adsorption, refer Figure 6.
Soltis (2015) on the other hand, after a review of the above three main mechanisms that focus on passive film penetration, film breaking, or adsorption together with mechanisms such as the ‘percolation model’ and ‘voids at the metal-oxide interface’, proposes that despite a number of theories for passive film breakdown/pit initiation, this aspect of localised pitting corrosion remains still the least understood, with a generally accepted theory yet to come. Soltis (2015) then surmises that it is entirely possible that we may fail to develop a single theory and will need to focus on the development of system-specific models, purely because of differences in passive film properties.
Given the Soltis (2015) proposition that there may not be a single theory for passive film breakdown and pit initiation and that development of system-specific models may be necessary, in the remainder of this sub-section, literature has been reviewed with the aim of proposing a system-specific mechanism(s) for chloride ion induced depassivation and pit initiation of reinforcement steel in concrete.
Bird et al (1988) in laboratory based solution studies in the pH range 10-14 and containing NaCl in the range 10-3 – 100 M, supports the idea that the first stages of localised passive film breakdown depend upon the competition for adsorption, at the oxide/liquid interface, between hydroxyl ions and aggressive ions such as chloride ion.
Leek and Poole (1990) for steel in concrete suggest that the breakdown of passivity by chloride ions is achieved by disruption of the passive film to substrate adhesive bond. An initial bond breakdown occurs beyond which internal pressure within the film due to surface tension effects, acts to disband chemically unaltered film, thereby expanding the size of the site of depassivation. Little or no chemical dissolution of the passive film occurs.
Sagoe-Crentsil and Glasser (1990) for steel in chloride contaminated concrete show evidence for the ‘Transitory Complex Theory’ above. X-ray diffraction analysis of steel from chloride contaminated cement paste samples showed the formation of an intermediate, soluble, FeCl2.4H2O complex. Visually, they advise this was evident as a greenish-blue hue.
As noted previously, Pape and Melchers (2013) identified FeOCl, a-Fe2(OH)3Cl, FeCl2.4H2O, “Green Rust I” (carbonate variety) and “Green Rust II” (sulphate variety) compounds within the corrosion products of samples from reinforced and prestressed structures in chloride environments.
It is then proposed that the presence of “green/greenish-blue rusts” may be evidence of the ‘Transitory Complex Theory’ above.
Figures 7 and 8 at Section 5.4.2 show schematically the reactions within a propagating chloride induced pit (Treadaway, 1988; 1991), and as such may also be considered further evidence of the ‘Transitory Complex Theory’.
Ghods et al (2013) have recently stated that the underlying mechanism for the depassivation process by chloride ions for steel in concrete is still not fully understood. They propose that one of the reasons behind this lack of understanding is the unavailability of data on the compositional and nanoscale morphological characteristics of the passive oxide film before and after exposure to chlorides. Furthermore they advise, since traditional electrochemical and most microscopic techniques do not provide such specific data, nano-scale surface chracterisation techniques are required and although such techniques have been widely used to characterise metal and metal alloy oxides, their application to study carbon steel and stainless steel in alkaline environments that are representative of concrete pore solution has been limited.
In the study by Ghods et al (2013), nano-scale transmission electron microscopy (TEM) samples prepared with a focused ion beam were used to investigate the passivity and depassivation of carbon steel reinforcement in simulated concrete pore solutions. It was found that the addition of chlorides at concentrations lower than the depassivation thresholds did not change the physical appearance of the passive oxide films. After exposure to chlorides in concentrations greater than depassivation thresholds, passive oxide films were no longer uniform, with some regions of the surface bare and some pit initiation sites formed on the surface of the steel.
It is proposed that the study by Ghods et al (2013) is evidence for both or either the ‘Adsorption Theory’ or ‘Transitory Complex Theory’ for chloride ion induced passive film breakdown of steel reinforcement in concrete.
However, based on the majority of the above, it is proposed that the ‘Transitory Complex Theory’ is the dominant mechanism for the chloride ion induced depassivation/pit initiation of steel reinforcement in concrete and that the ‘Adsorption Theory’ may also be a contributing mechanism.
5.3 Metastable Pitting
Frankel (1998) advises that metastable pits are pits that initiate and grow for a limited period before repassivating. For metals in solutions, large pits can stop growing for a variety of reasons, but metastable pits are typically considered to be those from micron size at most with a lifetime of the order of seconds or less.
Angst et al (2011) indicate for chloride induced reinforcement corrosion that once a pit has formed, the anodic dissolution reaction has to be sustained, otherwise the pit might cease growing and repassivate within a short time. Nucleation/repassivation events occur during the phase of so-called metastable pitting.
Furthermore, Angst et al (2011) propose that it is generally accepted that many of the nucleated pits never achieve stability and that for a pit to survive the metastable state, it is necessary that the solution within the pit cavity (anolyte) maintains an aggressive chemical composition. This requirement is met by hydrolysis of the dissolving iron cations and the resulting increase in H+ ion concentration as well as migration of chloride and other anions into the pit to maintain charge neutrality. As long as dilution of the aggressive pit chemistry is prevented, pit growth is self-sustaining and likely to achieve stability.
This mechanism of localised acidification due to metal ions hydrolysis is attributed to Galvele (1976). Newman (2010) then highlights that definition of the pit propagation process by the localised acidification model that has since had an enormous influence on the development of corrosion science.
Pit growth and pit propagation together with the chemical conditions within propagating pits including localised acidification is discussed specifically at Section 5.4 and illustrated in Figures 7 and 8.
Lin et al (2010) utilised a scanning-micro reference electrode (SMRE) technique to monitor localised corrosion processes of reinforcing steel in NaCl-containing solution by in-situ imaging of the corrosion potential and corrosion current. Metastable micro-pits occurred randomly and instantaneously on the surface when the reinforcing steel was first immersed in the chloride-containing solution. The number and size of active micro-pits changed with immersion time; some micro-pits lost their activity and some micro-pits gathered into groups to form visible corrosion holes that were covered by yellow-brown corrosion products.
Furthermore, whether metastable pitting nuclei developed into macro-corrosion pits or lost their corrosion activity, was closely associated with the local chloride concentration. Once stable pitting nuclei were formed, further hydrolysis reactions of Fe ions continued to lower the local pH, which accelerated dissolution of Fe at pits (Lin et al, 2010).
A porous cover of corrosion products and remnants of the passive film have also been reported for carbon steel and iron in chloride solutions (Alvarez & Galvele, 1984). Loss of this pit cover it is proposed will result in dilution of the anolyte and might stop anodic dissolution. The pit can survive rupture of the cover and continue to grow in a stable manner only if the pit geometry has already developed in such a way that the pit depth sufficiently limits transport of ionic species to and from the pit bottom, in other words: that the pit geometry (transport path) imposes a resistance against diffusion and migration (Angst et al, 2011).
5.4 Pit Growth/Pit Propagation
Green (1991) has proposed that for chloride induced pit growth to be sustained for reinforcing steel in concrete, the following conditions must be maintained:
- A sufficient concentration of chloride ions at the pit;
- The recycling of chloride ions during the corrosion process (i.e. from the hydrolysis of the intermediate iron chloride/oxy-chloride complexes);
- Diffusion of chloride ions to the pit from the bulk pore solution;
- Development of acidity within the pit;
- Cathodic processes on the steel surface; and
- A continuous electrolytic path between cathodic sites and the pit.
Angst et al (2011) advise that studying the transition from nucleation to stable pit growth for the case of reinforcement steel embedded in concrete is much more difficult owing to experimental reasons. For instance, the reference electrode cannot be placed as close to the pit as in the case of solution experiments, or the IR drop through the concrete might disturb electrochemical measurements. Also, visual examination of an electrode embedded in concrete is impossible during an experiment. In addition, concrete is an inhomogeneous material and thus local chemistry is not as well defined as in solutions. Corrosion kinetics are also clearly different for steel embedded in concrete compared with experiments in solution.
Angst et al (2011) proposed however, that in principle the localised pit growth/propagation mechanisms valid for stainless or carbon steel in solution can be expected to apply also for reinforcement steel in concrete, namely: acidification of the anolyte and migration of chloride ions into the pit are required to maintain the aggressive local pit chemistry, which in turn is required to sustain anodic dissolution. In concrete, convection as a mechanism that promotes dilution of the anolyte can be neglected when compared with the situation of a metal surface directly exposed to bulk solution. However, considering the presence of soluble alkaline cement (binder) hydration products (i.e. NaOH, KOH and Ca(OH)2) providing a pH buffering capacity, local acidification can be considered a critical step (Angst et al, 2011). Particularly lime/portlandite (Ca(OH)2), segregated at the steel/concrete interface, and its ability to retain a fall in pH, as it has been suggested as a major reason for the inhibitive properties of concrete (Page, 1975). As long as OH– ions migrate preferentially into the pit cavity to maintain charge balance, they will inhibit the pH of the anolyte from falling too fast.
Alvarez and Galvele (1984) reported that for iron in alkaline solutions, acidification of the anolyte is the rate controlling step; once a low pH in the pit is reached, the behaviour becomes equal to that found in acid solutions where chloride accumulation in the pit is the rate limiting process.
The chloride ions enter the passive film (via predominantly the ‘Transitory Complex Theory’) and give rise to soluble products such that a small bare area of metal surface is formed. Ferrous ions (Fe2+) diffusing away from the anodic area will react with the high pH environment to form rust product (Fe(OH)2 initially) which will shield the bare metal from oxygen. Further Fe2+ ions will no longer be able to be oxidised by the dissolved oxygen and so may be hydrolysed leading to the production of H+ ions in the incipient pit. The production of H+ ions reduces the pH in the pit. Tinnea (2002) has in fact measured reductions in pH to below 4 within pits for conventional steel reinforcement in chloride contaminated concrete.
A charge imbalance is also then generated by the H+ ions production so that Cl– ions will enter the pit which will now contain iron chloride (including the corrosive FeCl3), iron oxy-chloride and iron chloro-hydroxyl compounds (Treadaway, 1991).
The rate of pitting corrosion is very large (locally) because the portion of the metal covered by the protective passive film is in general very much larger than the portion of the metal where the passive film has been damaged. The area over which the reduction of oxygen takes place is therefore very much larger than the anodic area where metal is dissolved. Consequently as the metal loss that has to balance the oxygen reduction must all take place in the localised area where the passive film has been damaged, the rate of penetration of the metal in that local area is very fast. The result is rapid pitting is observed (Cherry & Green, 2017). As previously noted, indeed, values of pitting (penetration) rates of up to 1 mm/year have been reported for bridge decks, sustaining walls or other chloride contaminated steel reinforced concrete structures according to Elsener (2002).
5.5 Reinforcing Steel Quality
Angst et al (2011) indicate that the metallurgy (physical surface heterogeneities such as inclusions, lattice or mill-scale defects, flaws, and grain boundaries) of the steel reinforcement also impacts the nucleation of chloride induced pitting.
Bhandari et al (2015) note that for engineering alloys, pits almost always initiate due to chemical or physical heterogeneity at the surface such as inclusions, second phase particles, solute-segregated grain boundaries, flaws, mechanical damage, or dislocations.
5.5.2 Reinforcing Steel Defects
Lin et al (2010) have undertaken work to image the chloride concentration on the surface of reinforcing steel, and to further study the effects of interfacial chloride ions on pitting corrosion in reinforcing steel by combining electron microprobe analyser (EMPA) ex-situ imaging of the morphology and chemical composition of reinforcing steel at the same locations. Both the electrochemical inhomogeneities in the reinforcing steel and the non-uniform distribution of chloride at the surface were considered the most important factors leading to various localised attacks. They also developed a scanning-micro reference electrode (SMRE) technique to monitor localised corrosion processes of reinforcing steel in NaCl containing solution by in-situ imaging of the corrosion potential and corrosion current. Combining the data of the SMRE technique with that of the EMPA measurements they confirmed that among the electrochemical inhomogeneities (defects) in reinforcing steel, MnS inclusions play a leading role in the initial corrosion because chloride prefers to adsorb and accumulate at the MnS inclusions, resulting in pitting corrosion.
Accordingly, dissolution of the MnS inclusion is generally accompanied by a local drop in pH around the MnS inclusion, as follows (Lin et al, 2010):
MnS + 3H2O → Mn2+ + HSO3– + 5H+ + 6e– (3)
Once a local reduction in pH and increase in chloride concentration is reached, nucleating pits can initiate and form occluded pits, such that H+, Cl– and HSO3– can accumulate inside pits. As previous, the metastable pitting nuclei either developed into large corrosion pits or lost their corrosion activity, which largely depended on whether the local chloride concentration exceeded the critical chloride concentration. Again as previously, if stable pitting nuclei were formed, further hydrolysis reactions of Fe ions continued to lower local pH, which accelerated dissolution of Fe at pits (Lin et al, 2010).
6. CARBONATION INDUCED CORROSION MECHANISMS
Attack of concrete by carbon dioxide is termed carbonation. The passivity provided to steel reinforcement by the alkaline environment of concrete can be lost due to a fall in the pH of the concrete pore solution because of carbonation and thereby destroying the passive film.
Carbonation may occur when carbon dioxide gas (and atmospheric SOx and NOx gases) from the atmosphere dissolves in concrete pore water and penetrates inwards or when the concrete surface is exposed to water or soil containing dissolved carbon dioxide.
Carbon dioxide dissolves in the pore water to form carbonic acid by the reaction:
CO2(g) + H2O(l) → H2CO3(aq) (4)
Carbonic acid can dissociate into hydrogen and bicarbonate ions. The carbonic acid reacts with Ca(OH)2 (portlandite)in the solution contained within the pores of the hardened cement paste to form neutral insoluble CaCO3. The general reaction is as follows:
Ca(OH)2(aq) + H2CO3(aq) → CaCO3(s) + 2H2O(l) (5)
and the nett effect is to reduce the alkalinity of the pore water which is essential to the maintenance of a passive film on any reinforcing steel that may be present. Whilst there is a pH buffer between CaO and Ca(OH)2 that keeps the pH at approximately 12.5, the CO2 keeps reacting until the buffer is consumed and then the pH will drop to levels no longer protective of steel.
The attack of buried concrete by carbon dioxide dissolved in the groundwater is a two stage process. The calcium hydroxide solution that fills the pores of the concrete, first reacts with dissolved carbon dioxide to form insoluble calcium carbonate. However it then subsequently reacts with further carbon dioxide to form soluble calcium bicarbonate which is leached from the concrete. The extent to which each process takes place is a function of the calcium carbonate/calcium bicarbonate concentration of the ground water (which in turn is a function of the pH and the calcium content) and the amount of dissolved carbon dioxide.
The iron oxide passive film formed on steel in alkaline conditions is stable at pH levels greater than 10 (Broomfield, 2007) or greater than around 9.5 (Savija & Lukovic, 2016). Since the process of the carbonation of concrete lowers the pH of concrete pore water to lower than 9 (as low as 8.3) (Savija & Lukovic, 2016), the passive film is simply dissolved. The steel surface is then exposed to a pH~9 environment and corrosion thereby propagates.
Carbonation of concrete causes general corrosion where anodic and cathodic reactions are immediately adjacent (microcells or minicells), leading to uniform steel (iron) dissolution over the whole surface.
Localised carbonation induced corrosion of steel reinforcement can occur at cracks, concrete defects, low cover areas etc of reinforced concrete elements.
7. LEACHING INDUCED CORROSION OF REINFORCEMENT
The passivity provided to steel reinforcement by the alkaline environment of concrete may also be lost if the pH of the concrete pore solution falls because of leaching of Ca(OH)2 (and NaOH and KOH). Leaching from concrete can lead to a lowering of the pH below 10 to cause corrosion of steel reinforcement.
Natural waters may be classified as “hard” or “soft” usually dependent upon the concentration of calcium bicarbonate that they contain. Hard waters may contain a calcium ion content in excess of 10 mg/l (ppm) whereas a soft water may contain less than 1 mg/l calcium. The capillary pores in a hardened cement paste contain a saturated solution of calcium hydroxide which is in equilibrium with the calcium silicate hydrates that form the cement gel. If soft water can penetrate through the concrete (e.g. joint, crack, defect, etc) then it can leach free calcium hydroxide out of hardened cement gel so that the pore water is diluted and the pH falls. Since the stability of the calcium silicates, aluminates and ferrites that constitute the hardened cement gel requires a certain concentration of calcium hydroxide in the pore water, leaching by soft water can result in decomposition of these hydration products. The removal of the free lime in the capillary solution leads to dissolution of the calcium silicates, aluminates and ferrites and this hydrolytic action can continue until a large proportion of the calcium hydroxide is leached out, leaving the concrete with negligible strength. Prior to this stage, loss of alkalinity due to the leaching process will result in reduced corrosion protection to reinforcement.
Calcium hydroxide that is leached to the concrete surface reacts with atmospheric carbon dioxide to form deposits of white calcium carbonate on the surface of the concrete. These deposits of calcium carbonate may take the form of stalactites or severe efflorescence.
Leaching induced corrosion of steel reinforcement can be localised and consequently more serious where it occurs at concrete joints, cracks, defects, etc. An example of localised corrosion of steel reinforcement in a potable water reinforced concrete water tank is shown at Figure 9.
8. STRAY CURRENT CORROSION AND INTERFERENCE OF REINFORCEMENT
Stray electrical currents, most commonly from electrified traction systems, can also breakdown the passive film and cause corrosion of steel reinforced and prestressed concrete elements.
Stray electrical currents are a potent source of corrosion problems in reinforced concrete. Currents flow through any environment in response to potential differences in that environment. If an extended reinforced concrete structure is immersed in such an environment then these potential differences drive a current through the steel and this gives rise to corrosion. In the regions where the environment has the more positive potential, current will enter the metal by means of the reaction:
O + H2O + 2e– → 2OH– (6)
In the regions where the environment has the more negative potential, current will leave the metal by means of the reaction:
Fe → Fe++ + 2e– (7)
and as the metal is polarised in the positive direction that part of the structure is corroded.
8.2 Ground Currents
Ground currents are particularly effective in causing corrosion in extended reinforced concrete structures such as concrete pipelines, tunnels and retaining walls because the potentials that can build up in the soil between the different portions of the structure are correspondingly larger. Although the driving force for the corrosion is greater, the resistance between the cathode and anode is not proportionately larger and so the corrosion currents are greater.
The commonest cause of stray current corrosion is electrified traction systems. In general the DC power for an electric train or tram is supplied by a sub-station and drawn from a positive overhead cable. The negative return to the sub-station is either through the rails on which the train is running or through a separate conductor rail. Although most of the current driving the trains/trams returns to the sub-station from which it is drawn via the conductor rail of the traction system, it is inevitable that some of the current in the conductor rail leaks from that rail and goes back to the sub-station through the earth. This sets up potential gradients in the earth. If a low resistance path (such as metallic reinforcement to a concrete structure) is present then the current may pass through the reinforcement. It can be seen from Figure 10 that where the current enters the structure, the reinforcement is cathodically protected. Where the current leaves the structure near the sub-station, corrosion is intensified.
The potential loss of reinforcing steel section due to stray current or interference corrosion can be very serious and design measures may need to be developed to mitigate such corrosion. As an example, suitable mitigation systems needed to be designed, implemented and maintained for the Sydney Harbour Tunnel as calculations carried out indicated a potential corrosion problem existed on the steel reinforcing of the immersed tunnel units and up to 150 kg of steel could be corroded per year by stray current (McCaffrey, 1991).
8.3 Interference Currents
Interference corrosion currents are typically associated with cathodic protection schemes. The application of remote anode impressed current cathodic protection to (say) a buried pipeline involves the establishment of an anode(s) (anode groundbed) at some distance from the pipeline with the aim of passing current from the anode(s) (groundbed) to the pipeline. If there is a reinforced concrete structure within this potential field then potential gradients associated with the passage of current through the soil give rise to interference corrosion of steel reinforcement. The effect on the buried reinforced concrete structure is to set up cathodic areas near to the anode groundbed and anodic areas closer to the structure which the cathodic protection system has been designed to protect.
8.4 Local Corrosion Due to Stray Currents
Dependent on the type of reinforced or prestressed concrete structure and elements that are near an electrified traction system or cathodic protection systems, stray current or interference corrosion of steel reinforcement may be localised. As such, structurally significant section loss of reinforcement may occur within short periods of time, since current density can be converted to an equivalent mass loss or corrosion penetration rate by Faraday’s Law.
For example, a section of reinforcing steel receiving say 1 uA/cm2 of stray current equates to a mass loss or corrosion penetration rate of approximately 12 um/yr because of Faraday’s Law (Fontana, 1987). If the stray current is 1 mA/cm2, then the corrosion penetration rate will be approximately 12 mm/yr. Consequently significant section loss of reinforcement may occur within short periods of time for reinforced and prestressed elements suffering stray current corrosion from ground currents or interference currents.
Professor Cherry has always been a firm believer in understanding first the fundamentals of any aspects of corrosion science, then the mechanisms, before embarking on the engineering of solutions to the management of materials corrosion. So it has been with steel reinforced concrete. This paper has endeavoured to walk a reader through the fundamental and mechanistic aspects of the excellent protection afforded to steel reinforcement by concrete (including electrochemistry), corrosion of steel reinforcement (uniform (microcell/mini-cell) and pitting (macrocell) corrosion, corrosion products composition and development), chloride induced corrosion mechanisms (passive film breakdown/pit initiation, metastable pitting, pit growth/propagation, chemical conditions within propagating pits and reinforcing steel quality effects), carbonation induced corrosion mechanisms, leaching induced corrosion of reinforcement and reinforcing steel stray current corrosion and interference.
Concrete is a most wonderful material of construction. When suitably designed, constructed and maintained, reinforced concrete provides service lives of numerous decades to structures and buildings. Concrete provides reinforcing steel with excellent corrosion protection. The highly alkaline environment in concrete results in the spontaneous formation of a stable, tightly adhering, thin protective oxide passive film on the steel reinforcement surface, which protects it from corrosion. In addition, well proportioned, compacted and cured concrete has a low penetrability, thereby minimising the ingress of corrosion-inducing species via the aqueous phase, and a relatively high electrical resistivity, which reduces the corrosion current and hence the rate of corrosion if corrosion is initiated. As Professor Cherry has said many times “good steel in good concrete will not corrode”.
There are however, several degradative processes which affect some reinforced concrete structures leading to loss of functionality, unplanned maintenance/remediation/replacement, and in the worst cases, loss of structural integrity and resultant safety risks and amongst these, the most common cause of deterioration is corrosion of conventional carbon steel (black steel), prestressing steel and post-tensioned steel reinforcement.
The authors would like to acknowledge colleagues, mentors, peers and industry friends for their support.
Author 1 would like to specifically thank Jack Katen and Andrew Haberfield for their help with the paper.
It is noted that the views expressed in this paper are those of the authors and are not necessarily of the organisations that they represent.
Abd El Haleem, S M, Abd El Hal, E E, Abd El Wanees, S and Diab, A (2010), Environmental factors affecting the corrosion behaviour of reinforcing steel: I. The early stage of passive film formation in Ca(OH)2 solutions, Corrosion Science, 52, 3875-3882.
Al-Negheimish, A, Alhozaimy, A, Rizwan Hussain, R, Al-Zaid, R and Singh, J K (2014), Role of manganese sulphide inclusions in steel rebar in the formation and breakdown of passive films in concrete pore solutions, Corrosion, 70, 1, January, 74-86.
Alvarez, M G and Galvele, J R (1984), The mechanism of pitting of high purity iron in NaCl solutions, Corrosion Science, 24, 1, 27-48.
American Concrete Institute (1985), Corrosion of metals in concrete, Committee 222, ACI Journal, 82, 1, 3-32.
Andrade, C and Gonazlez, J A (1978), Quantitative measurements of corrosion rate of reinforcing steels embedded in concrete using polarization resistance measurements, Werkstoffe und Korrosion, 29, 8, August, 515-519.
Angst, U, Elsener, B, Larsen, C K and Vennesland, O (2011), Chloride induced reinforcement corrosion: Rate limiting step of early pitting corrosion, Electrochimica Acta, 56, 5877-5889.
Bhandari, J, Khan, F, Abbassi, R, Garaniya, V and Ojeda, R (2015), Modelling of pitting corrosion in marine and offshore steel structures – A technical review, Journal of Loss of Prevention in the Process Industries, 37, 39-62.
Bird, H E H, Pearson, B R and Brook, P A (1988), The breakdown of passive films on iron, Corrosion Science, 28, 1, 81-86.
Broomfield J P (2007), Corrosion of steel in concrete, Taylor and Francis, 2nd Edition, Oxon.
Cherry, B and Green, W (2017), The Corrosion and Protection of Concrete Structures and Buildings, Australasian Corrosion Association and Australasian Concrete Repair Association, Training Course Notes, Version 1.0, Melbourne.
Cohen, M (1978), The passivity and breakdown of passivity of iron, in: R P Frankeuthal, J Kurager, N J Princietor (Eds.), Passivity of Metals, The Electrochemical Society, 521.
Concrete Institute of Australia, (2015), Performance tests to assess concrete durability, Recommended Practice Z7/07, Concrete Durability Series, Sydney.
Cornell, R M and Schwertmann, U (2000), The Iron Oxides, Wiley-VCH, Weinhem.
Elsener, B (2002), Macrocell corrosion of steel in concrete – implications for corrosion monitoring, Cement & Concrete Composites, 24, 65-72.
Faraday, M (1844), Experimental Researches in Electricity, Vol II, Richard and Edward Taylor, London.
Fontana, M G (1987), Corrosion Engineering, McGraw-Hill, 3rd Edition, Singapore.
Frankel, G S (1998), Pitting corrosion of metals a review of the critical factors, Journal of the Electrochemical Society, 145, 6, 2186-2198.
Freire, L, Novoa, X R, Montemor, M and Camezin, M J (2009), Study of passive films formed on mild steel in alkaline media by the application of anodic potentials, Materials Chemistry and Physics, 114, 962-972.
Galvele, J R (1976), Transport processes and the mechanism of pitting of metals, Journal of the Electrochemical Society, 123, 464-474.
Ghods, P, Isgor, O B, Brown, J R, Bensebaa, F and Kingston, D (2011), XPS depth profiling study on the passive oxide film of carbon steel in saturated calcium hydroxide solution and the effect of chloride ion on the film properties, Applied Surface Science, 257, 10, March, 4669-4677.
Ghods, P, Isgor, O B, Carpenter, G J C, Li J, McRae G A and Gu, G P (2013), Nano-scale study of the passive films and chloride-induced depassivation of carbon steel rebar in simulated concrete pore solutions using FIB/TEM, Cement and Concrete Research, 47, 55-68.
Green, W K (1991), Electrochemical and chemical changes in chloride contaminated reinforced concrete following cathodic polarisation, MSc Dissertation, University of Manchester Institute of Science and Technology, Manchester.
Green, W, Riordan, G, Richardson, G and Atkinson, W (2009), “Durability assessment, design and planning – Port Botany Expansion Project”, Proc 24th Biennial Conf. Concrete Institute of Australia, Paper 65, Sydney.
Green, W, Dockrill, B and Eliasson, B (2013), Concrete repair and protection – overlooked issues, Proc. Corrosion and Prevention 2013 Conf., Australasian Corrosion Association Inc., Brisbane, November, Paper 020.
Hansson, C M (1984), Comments on electrochemical measurements of the rate of corrosion of steel in concrete, Cement and Concrete Research, 14, 4, July, 574-584.
Huet, B, Hostis, V L, Miserque, F and Idrissi, H (2005), Electrochemical behaviour of mild steel in concrete: Influence of pH and carbonate content of concrete pore solution, Electochimica Acta, 51, 172-180.
Jaffer, S J and Hansson, C M (2009), Chloride-induced corrosion products of steel in cracked-concrete subjected to different loading conditions, Cement and Concrete Research, 29, 116-125.
Koleva, D A, Hu, J, Fraaij, A L A, Stroeven, P, Boshkov, N and de Wit, J H W (2006), Quantitative characterisation of steel/cement paste interface microstructure and corrosion phenomena in mortars suffering from chloride attack, Corrosion Science, 48, 4001-4019.
Kolio, A, Honkanen, M, Lahdensivu, J, Vippola, M and Pentti, M (2015), Corrosion products of carbonation induced corrosion in existing reinforced concrete facades, Cement and Concrete Research, 78, 200-207.
Leek, D S and Poole, A S (1990), Corrosion of Steel Reinforcement in Concrete, ed. C L Page, K W J Treadaway and P B Bamforth, Society of Chemical Industry and Elsevier Applied Science, 65-73.
Lin, B, Ronggang, H, Chenqing, Y, Yan, L and Chagjian, L (2010), A study on the initiation of pitting corrosion in carbon steel in chloride-containing media using scanning electrochemical probes, Electrochimica Acta, 55, 6542-6545.
Melchers, R E and Li, C Q (2008), Long-term observations of reinforcement corrosion for concrete elements exposed to the north sea, Proc. Corrosion & Prevention 2008 Conf., Australasian Corrosion Association Inc., Wellington, Paper 079.
McCaffrey, W R (1991), Stray current mitigation systems in Sydney Harbour Tunnel, Proc. Corrosion & Prevention 1991 Conf., Australasian Corrosion Association Inc., Sydney, November, Paper 27.
Nasrazadani, S (1997), The application of infra-red spectroscopy to the study of phosphoric acid and tannic acid interactions with magnetitie (Fe3O4), goethite (a-FeOOH) and lepidocrocite (y-FeOOH), Corrosion Science, 39, 1845-1859.
Page, C L (1976), Mechanism of corrosion protection in reinforced concrete marine structures, Nature, 258, 514-515.
Page, C L and Treadaway, K W J (1982), Aspects of the electrochemistry of steel in concrete, Nature, 297, 109-115.
Pape, T M and Melchers, R E (2013), A study of reinforced concrete piles from the Hornibrook Highway bridge (1935-2011), Proc. Corrosion & Prevention 2013 Conf., Australasian Corrosion Association Inc., Brisbane, November, Paper 086.
Roads & Maritime Services (2013), Specification B80 Concrete Work for Bridges, Edition 6/Revision 5, July, Sydney.
Sagoe-Crentsil, K K and Glasser, F O (1989), Steel in concrete part II: Electron microscopy analysis, Magazine of Concrete Research, 213-220.
Sagoe-Crentsil, K K and Glasser, F P (1990), Corrosion of Steel Reinforcement in Concrete, ed. C L Page, K W J Treadaway and P B Bamforth, Society of Chemical Industry and Elsevier Applied Science, 74-85.
Sajiva, B and Lukovic, M (2016), Carbonation of cement paste: Understanding, challenges, and opportunities, Construction and Building Materials, 117, 285-301.
Singh, J K and Singh, D D N (2012), The nature of rusts and corrosion characteristics of low alloy and plain carbon steels in three kinds of concrete pore solution with salinity and different pH, Corrosion Science, 56, 129-142.
Soltis, J (2015), Passivity breakdown, pit initiation and propagation of pits in metallic materials – Review, Corrosion Science, 90, 5-22.
Standards Australia (2010), AS 3972 Portland and blended cements, Sydney, Australia.
Standards Australia (2016), AS 3582.1 Supplementary cementitious materials for use with portland and blended cement – Fly ash, Sydney, Australia.
Standards Australia (2016), AS 3582.2 Supplementary cementitious materials for use with portland and blended cement – Slag – Ground granulated iron blast-furnace slag, Sydney, Australia.
Tinnea, J and Young, J F (2000), The chemistry and microstructure of concrete: Its effect on corrosion testing, Proc. Corrosion & Prevention 2000 Conf., Australasian Corrosion Association Inc., Auckland, November Paper 026.
Tinnea, J (2002), Localised corrosion failure of steel reinforcement in concrete: Field examples of the problem, Proc. 15th International Corrosion Congress, Granada, Spain, Paper 706.
Treadaway, K (1988), Corrosion of steel in concrete, ed P Schiessl, RILEM Report of Technical Committee 60-CSC, Chapman and Hall, London.
Treadaway, K W J (1991), Corrosion propagation, COMETT Short Course on The Corrosion of Steel in Concrete, University of Oxford.
Tuutti, K (1982), Corrosion of steel in concrete, Swedish Cement and Concrete Association, Report Fo. 4.82.
Veluchamy, A, Sherwood, D, Emmanuel, B and Cole, I S (2017), Critical review on the passive film formation and breakdown on iron electrode and the models for the mechanisms underlying passivity, Journal of Electroanalytical Chemistry, 785, 196-215.
Vera, R, Villarroel, M, Carvajal, A M, Vera, E and Ortiz, C (2009), Corrosion products of reinforcement in concrete in marine and industrial environments, Materials Chemistry and Physics, 114, 467-474.
VicRoads (2017), Specification 610 – Structural Concrete, April, Melbourne.