Corrosion of steel reinforcement bars in concrete
Corrosion of steel reinforcement bars in concrete
Reinforced concrete (RC) is a composite material comprising steel reinforcement bars (rebars) embedded in a concrete mass. Rebars carries the bulk of the tensile load and imparts a degree of cracking resistance to the concrete which itself is compressively loaded. Steel in concrete is normally in a non-corroding, passive condition. However, non-corroding, passive conditions are not always achieved in practice due to which corrosion of rebars takes place. Corrosion of rebars has physical consequences such as decreased ultimate strength and serviceability of the concrete structures. The steel rebar corrosion in concrete is a big universal problem. The damage which happens from corrosion can to a large extent reduce the serviceability and structural integrity of the RC. Corrosion induced damage to the concrete structure necessitates an early repair or in some cases complete replacement of the concrete structure.
Whenever the reinforcement bar embedded in steel corrodes, the corrosion products increase its volume. All forms of iron oxide and hydro-oxide have specific volumes which is greater than that of the steel. The expansive forces generated by the steel corrosion leads to tensile cracking and rust staining of the concrete. This, in turn, causes reduction in the serviceability and structural integrity of concrete besides affecting its aesthetics. Once the corrosion starts, it is only a matter of time before a cumulative amount of damage occurs to the concrete structure and it fails well before its design life.
The two main causes of the corrosion of the rebars in concrete are (i) localized failure of the passive film on the steel by the chloride ions and (ii) general failure of the passivity by neutralization of the concrete alkalinity due to the reaction with atmospheric CO2 (carbon dioxide). The main factors affecting the corrosion of rebars in RC are (i) loss of alkalinity due to carbonation, (ii) loss of alkalinity due to chlorides, (iii) cracks in the concrete because of mechanical loading, (iv) stray currents, (v) atmospheric pollution, (vi) moisture pathways, (vii) water-cement ratio, (viii) low tensile strength of the concrete, (ix) electrical contact with dissimilar metals, and (x) corrosion due to difference in environments.
Electro-chemical nature of corrosion of steel in concrete
Corrosion can be defined as the deterioration or destruction of a material by reaction with its environment. It is an electro-chemical process (Fig 1), which requires a flow of electric current and many chemical reactions. An example of the electro-chemical process is a galvanic cell. For an electro-chemical cell to function, three basic elements are necessary namely (i) anode, (ii) cathode, and (iii) electrolyte. An anode is an electron producing unit, while the cathode is the electron consuming unit. The electrolyte is a medium through which ionic flow can occur. Typical reactions at the anode and cathode for the corrosion of iron are (i) anodic reaction Fe = Fe (++) + 2e(-) (oxidation), cathodic reaction 2H (+) + 2e(-) = H2 (reduction), and depolarization reaction 2H(+) +2e(-) + 0.5 O2 = H2O.
Fig 1 Concept of electro-chemical process of corrosion of rebars in concrete
At the anode, metallic iron (Fe) is oxidized and electrons are generated. Since the metal is to remain at a state of electron equilibrium, an equal amount of electrons are consumed at the cathode to form hydrogen (H2) gas. The H2 gas tends to remain near the surface of the rebar and the reaction becomes self-inhibiting. The cathode is then said to be polarized and no further reaction is possible unless the protective H2 film is removed (depolarized). H2 can be evolved as a gas, but this process is normally quite slow. More important is the breakdown of the H2 film by the depolarizing action of oxygen (O2). In this case, O2 acts to prevent the buildup of H2 gas by consuming the free electrons. Once the H2 layer is broken, the corrosion reactions are free to continue
Since sodium and chloride ions do not participate in the reaction, the total reaction can be expressed as the sum of anodic and the depolarization reactions. Making use of the reaction H2O = H(+) + OH(-) gives the principal corrosion reaction as Fe + H2O + 0.5 O2 = Fe(OH)2. The compound precipitating is ferrous hydroxide, a form of rust with whitish colour. However, in oxygenated solutions, ferrous hydroxide is further oxidized to ferric hydroxide. The product finally formed is the familiar reddish-brown rust
Type of rebar corrosion
There are two types of corrosion which are observed in the rebars embedded in concrete. These are (i) crevice corrosion, and (ii) pitting corrosion. Crevice corrosion is a localized form of corrosion usually associated with a stagnant solution on the micro-environmental level. Such stagnant micro-environments tend to occur in crevices (shielded area). O2 in the liquid which is deep in the crevice is consumed by reaction with the metal. O2 content of the liquid at the mouth of the crevice which is exposed to air is greater. Hence a local cell is formed in which the anode (area under attack) is the surface in contact with the O2 depleted liquid. In case of pitting corrosion, theories of passivity fall into two general categories. The first one is based on adsorption while the second one is based on presence of a thin oxide film. Pitting corrosion in the first case arises as detrimental or activator species, such as chloride ion, compete with O2 or hydroxyl ion at specific surface sites. By the oxide film theory, detrimental species become incorporated into the passive film, leading to its local dissolution or to development of conductive paths. Once initiated, pits propagate auto-catalytically, resulting in acidification of the active region and corrosion at an accelerated rate.
In a medium of perfect uniformity, corrosion is very unlikely to occur. However, reinforced concrete is by no means a homogeneous material and corrosion cells are set up when certain conditions exist. There are numerous reasons for the corrosion enhancing non-uniformity of concrete. Concrete can be honeycombed, porous, and unevenly wet and dry. Cracking causes differences in steel stress, differential aeration, and depositions of salt. There are always inherent non-uniformities in the rebar itself due to initial locked-in residual stresses and the manufacturing processes. As a result, regions of lower potential become anodic and regions of higher potential become cathodic. Moist concrete acts as the electrolyte, the action of which is further accelerated if salt ions exist. Corrosion in reinforced concrete normally falls under two general groups namely (i) cracked concrete, and (ii) uncracked concrete.
In case of fresh uncracked concrete normally there is ample resistance to corrosive attack. The concrete cover over the rebar is very effective in inhibiting the penetration of corrosive agents to the level of the steel. It is obvious that the thicker and denser the concrete cover the more effective it becomes in resisting corrosion. Also, fresh concrete has a very high pH value which usually inhibits corrosion reactions. The pH number is an index of the acidity or alkalinity of a medium. Numbers from 0 to 7 indicate acidity of a solution (in which corrosion is promoted), and numbers from 7 to 14 indicate solution alkalinity (in which corrosion is retarded). Fresh concrete has a high Ca(OH)2 (calcium hydroxide) content which gives it a pH of around 13. The last defense against corrosion is offered by the blue oxide film (mill scale) around the surface of the rebar. This oxide film prevents corrosive agents from coming into direct contact with the bare metal. Thus mill scale provides localized corrosion protection.
However, as time passes the above conditions tend to alter. Water, salt, O2, CO2, and industrial gases (if present) slowly begin penetrating the concrete, the rate of which depends on the permeability of the concrete cover. CO2, which penetrates into concrete through pores and cracks, reacts with Ca(OH)2 and produces calcium carbonate. Thus, both the pH value and the protective quality of concrete are reduced. The general mechanism by which corrosion occurs in concrete is shown in Fig 2.
Fig 2 General mechanism for the corrosion of rebar
The probability of corrosion is high when the pH of concrete falls as low as 8. Crystallizing salt and freeze-thaw effects set up internal forces which adversely affect the durability of the concrete cover. As a corrosive medium reaches the steel, it concentrates its attack at the flaws in the oxide film. More importantly, if salt is present, it destroys the passivity of the oxide film on the steel and corrosion is thus promoted.
It is obvious that at large cracks in the concrete, the penetration phase of the above sequence is considerably shorter and corrosion rapidly begins on the steel below the cracks. In uncracked regions of the concrete, the same sequence takes place as outlined, but at a much reduced rate. That is, corrosion initiates as soon as the corrosion promoting medium penetrates through the concrete to the level of the steel.
It is to be remembered that the presence of salt is an important factor in the corrosion process. Salt ions destroy the passivity of steel, set up corrosion cells, and increase the conductivity of the electrolyte. Without salt ions, corrosion of rebar in concrete can be inhibited for a long period of time. In that case, the corrosion rate is generally controlled by the processes of carbonation. If the concrete cover is relatively impermeable and thick, corrosion cannot occur at all in uncracked areas. However, cracks do not lose their importance in this case because localized corrosion can occur under them.
The corrosion products formed tend to have an inhibiting effect upon continued corrosive reactions. These products can seal off the base metal from the diffusion of O2 and H2 and thus terminate the corrosion reactions. This process is known as self-limiting corrosion. A self-limiting corrosion can take place at high C/D (cover thickness / rebar diameter) ratios which seem to determine the occurrence and extent of longitudinal splitting along the rebars. The longitudinal splitting is mainly due to the tensile forces created by the corrosion products which occupy around three times greater volume than the steel from which they are formed. If the concrete cover is not sufficient to resist such forces, longitudinal cracks develop through which O2 and other external agents gain access to the steel. At this point, it is only a matter of time until the structure reaches a hazardous state of corrosion and is to be repaired or replaced. Repeated loadings can also play a role in breaking the protective effect of the rust scale, but more study is needed to establish its importance.
Steel is thermodynamically unstable in the earth’s atmosphere and hence, always tend to revert to a lower energy state such as an oxide or hydroxide by reaction with O2 and water. These processes continuously occur. The question of interest in the use of steel is to control these processes to occur in practice. Fortunately, only the surface atoms of the steel are exposed to the atmosphere and, hence, are available to react. In the case of a 15 mm diameter bar, this amounts to only around 1 in every 40 million atoms. Any coating on the steel reduces this number even further.
For steel embedded in concrete, the concrete itself provides a coating limiting the access of water and O2 to the steel surface. A second beneficial aspect of concrete is that the solution in the pores of the cement paste has a very high alkalinity and, as indicated in the Pourbaix diagram in Fig 3, at the pH levels typical of concrete, the corrosion products which do form are insoluble. They produce a very thin (around few nano metres) protective coating on the steel (a passive film) which limits the metal loss from the steel surface due to corrosion to around 0.1 micrometers to 1.0 micrometers per year. It is generally considered that, at these passive corrosion rates, the steel embedded in concrete is not normally noticeably degraded within a 75 year lifetime and the volume of corrosion products are not sufficient to cause any damaging stresses within the concrete. The passive film does not form immediately but starts as soon as the pH of the mixing water rises in the concrete when the cement begins to hydrate and stabilizes over the first week to protect the steel from active corrosion.
Fig 3 Pourbaix diagram of iron
Corrosion, whether at the negligible passive rate or the damaging active rate, is an electro-chemical process, involving the establishment of anodic and cathodic half cell reactions on the micro-scopic and /or macro-scopic levels. In high pH solutions and in the absence of chloride ions, the anodic dissolution reaction of iron is balanced by the cathodic reaction the Fe2+ ions combine with the OH- ions to produce the stable passive film.
Both the anodic and cathodic reactions are necessary for the corrosion process to take place concurrently. The anode can be located next to each other or can be separated. When they are located immediately next to each other, that is, on a microscopic scale, the resultant corrosion cell is referred to as micro-cell corrosion. When they are separated by some finite distance, the resulting corrosion cell is referred as macro-cell corrosion. The corrosion of rebars in concrete can be due to combination of micro-cells and macro-cell corrosions. Fig 4 shows both the micro-cell and macro-cell corrosions of rebars in concrete.
Fig 4 Micro-cell and macro-cell corrosions of rebars in concrete
Chloride induced corrosion
The mechanism by which chloride ions break down the passive film is not fully understood, largely because the film is too thin to be examined and because the events occur inside the concrete. One hypothesis is that the chloride ions become incorporated into the passive film and reduce its resistance. This incorporation is not uniform and, where it occurs, it allows a more rapid reaction and the establishment of an anodic area where corrosion continues while the remaining steel remains passive (Fig 5b).
Fig 5 Schematic representation of passive and chloride-induced active corrosion
A second hypothesis is that the Cl- ions ‘compete’ with the OH- anions for combining with Fe2+ cations and, since the Cl- ions form soluble complexes with the Fe2+ ions, a passive film is not formed and the process stimulates further metal dissolution. The soluble iron-chloride complexes diffuse away from the steel and subsequently break down, resulting in the formation of expansive corrosion products and, simultaneously freeing the Cl- ions, which are then able to migrate back to the anode and react further with the steel. In this overall process, hydroxyl ions are continuously consumed, locally decreasing the pH (i.e. making the solution acidic in that localized region) and, thus, enhancing further metal dissolution. The Cl- ions, on the other hand, are not consumed and the attack then becomes ‘auto-catalytic’. Ultimately, the reinforcement cross-section and its structural resistance are seriously compromised.
Either of these hypothesized mechanisms explains the local nature of the attack often observed. The local actively corroding areas behave as anodes while the remaining passive areas become cathodes where reduction of dissolved O2 takes place. The galvanic cells can be macro or micro in scale depending on a number of factors. Thus, the anode and cathode can be widely separated or they can be adjacent on an atomic scale.
Time dependence of rebar corrosion
Corrosion process of rebars has three distinct stages (i) initiation, (ii) de-passivation, and (iii) propagation. Initiation precedes de-passivation which is then followed by propagation to reach the final state (Fig 6). After the initiation, a crack appears on the external concrete surface which propagates and does further damage and develop into. The service life is determined when the rebar reaches the final state which is the time when the spalling of the concrete starts.
Fig 6 Stages of corrosion of rebars
The most detrimental consequence of chloride-induced corrosion of rebars is the build-up of voluminous, insoluble corrosion products in the concrete which leads to internal stresses and, finally, to cracking and spalling of the concrete cover. Clearly, once such damage is visually apparent, the rebars are prone to very rapid further corrosive attack because access to O2 and moisture is no longer limited by diffusion through the concrete cover. All forms of iron oxides and hydroxides have high specific volumes which are greater than the volume of steel of the rebar (Fig 7). Hence, the degree of damage to the concrete produced by a certain amount of corrosion depends on the specific corrosion products formed and their distribution within the concrete cover as well as on the porosity and strength of the concrete itself.
Fig 7 Specific volume of the corrosion products of iron
Sometimes it is assumed that the corrosion products are rust, i.e. Fe2O3.3H2O and because this, the orange coloured product is observed on damaged concrete. Hence, it is also assumed that the corrosion products are more than six times as voluminous as the steel from which they are formed and the predicted stresses in the concrete are based on this conclusion. In fact, analysis of the products formed indicates that there are other products as shown in Fig 7 which have a specific volume ranging between 2.2 times and 3.3 times that of the steel. It is only after cracking and spalling and, thus, exposure to the atmosphere, these products convert to the familiar rust.
Parameters influencing corrosion of steel in concrete
The parameters influencing corrosion of steel in concrete are shown in Fig 8. The steel related parameters are metallurgical properties, prior rusting, size of the bar, and steel arrangement.
Fig 8 Parameters influencing corrosion of steel in concrete
Metallurgical properties – It is perhaps common knowledge that many elements alloyed with steel produce increased corrosion resistance. The major corrosion inhibiting elements include copper (Cu), nickel (Ni), and chromium (Cr), most of which are present in negligible proportions in the steel for rebars. Specific combinations of these and other elements have been found to improve corrosion resistance of steels, but from a practical standpoint have had little impact in the application rebar steels in concrete.
Localized metallurgical differences in the atomic structure of the steel cause differential energy fields within the steel and promote the formation of the anodic and cathodic regions necessary for electro-chemical corrosion. These regions are, in effect, different materials in contact with one another. Energy fields are usually associated with dislocations, mismatched grain boundaries, inclusions, impurities, metallurgical phase boundaries, etc. For example, it has been determined that the ferrite phase of steel is readily attacked, while cementite is resistant to corrosion. Where both phases exist adjacent to one another, the cementite becomes the cathode and the ferrite becomes the anode and a corrosion cell is developed.
It is to be recognized that differential energy field sources for corrosion cells are present in all commercial steels and hence a means of inhibiting corrosion is to be found other than attempting to homogenize the steels, which is impractical and of questionable effectiveness. For this reason it is fortunate that the effect of these various energy fields upon the corrosion of rebar steel is minimal as long as the pH of the surrounding concrete remains relatively high (in the range of 10 to 13).
In addition to the corrosion cell sources associated with the basic atomic structure of the steel, the surface of the rebar offers additional opportunities for cell formation. Such factors as surface roughness, scratches, cuts, and particularly mill scale are frequently responsible for the initiation of corrosion. If the mill scale formed during the hot rolling of the steel does not result in a continuous scale coating then the surface areas coated with mill scale are cathodic to the uncoated adjacent areas.
In certain applications, metallic coatings offer corrosion protection to surfaces of steel. However, such cathodic coatings as Ni and Cu are not effective in rebar steels since they are relatively expensive and are likely to be damaged during construction, hence creating serious localized corrosion conditions. Cadmium (Cd) and zinc (Zn) are anodic to steel and can be used as sacrificial coatings. Galvanized coatings on rebars are perhaps practical, but to be effective, the coating is to be of adequate thickness.
Prior rusting of rebars – The condition of the rebars prior to embedment has considerable influence. In some standards, it is required that loose, ‘flaky’ rust is to be removed from rebar steel prior to use and that normal rough handling generally removes injurious rust. On the other hand, some other standards are less restrictive with respect to prior rusting of rebars in that use of prior rusted rebar is allowed so long as the requirements on deformation height, dimensions and brushed bar weight are met.
Furthermore, it has been reported that normal rust actually increases bond. It has also been found that for 14-day-old concrete the use of prior rusted welded wire fabric resulted in less bond slip in comparison to clean wire. However, the long-term effects of the use of prior rusted rebars are not well-defined. This is especially critical for exposed structures. In fact, it has been suggested that prior rusting of pre-stress tendons can cause serious corrosion after encasement in concrete grout. The same concern can be expressed for prior rusted reinforcing bars in exposed structural elements.
Size of the bar and steel arrangement – There are relatively few corrosion studies which have included variables related to bar size and steel arrangement. In one study, it has been determined that a welded grid of rebars is susceptible no more corrosion than the individually insulated rebars. In another study, it has been observed that a relationship exists between bar spacing and corrosion induced cracking. In this study rebars spaced 300 mm apart generally developed trench like spalls, while those rebars spaced 150 mm apart tended to develop weakened planes.
To reduce and prevent the corrosion of reinforcement steel bars in concrete several methods are employed. Some are related to the making of concrete while the others are related to the quality, composition and coating of steel used in the making of reinforcement of bars. The choice is normally made based on the cost. Coatings employed on the rebars are (i) hot dip galvanizing, (ii) fusion bonded epoxy coating, and (iii) stainless steel cladding. Reinforcement bars of stainless steels are also being used. Coatings suffer from the disadvantage since coatings can be physically damaged or electro-chemically penetrated so that the base steel is again vulnerable to the usual corrosion process. Steel rebars of special composition to resist corrosion have also been tried. Several steel plants have experimented with various compositions of the weathering steels. However after extensive testing, it has been found that there is consistently poor performance of weathering steels when buried. Hence, the production of steel rebars has been abandoned by most of the producers.