Durability modeling of reinforced concrete structures

RE Melchers Infrastructure Performance and Reliability Centre, Newcastle University, Australia
Introduction: The management of reinforced concrete assets increasingly relies on analytical models to predict the time of steel corrosion, especially the time of severe destructive corrosion. Increasingly, it has been discovered that the existing models’ predictions of the occurrence of corrosion of steel bars caused by chlorides cannot be compared well with the observation results and actual experience of actual concrete structures. Examples show that for high-quality concrete structures, even after long-term exposure, there may be a very high chloride content without signs or signs of corrosion. In this case, the latest research that separates corrosion initiation from active corrosion is reviewed. Corrosion occurs mainly due to the existence of voids adjacent to the reinforcement. It is mainly a short-term transient effect and does not indicate the beginning of severe corrosion. Active corrosion is a direct result of the decrease in concrete alkalinity, which is the mechanism of chloride acceleration. This is the cause of severe steel corrosion, not direct chloride corrosion. In addition, the effects of stress-induced cracking and other concrete cracking must be reconsidered. These new interpretations opened the way for improved modeling and prediction of “chloride-induced” corrosion.
More and more asset managers and engineers are more and more interested in the durability of reinforced concrete (RC) assets, especially the durability of steel corrosion. Many structures and infrastructure projects are related to prediction-the probability that the structure will remain safe during the recommended lifetime, or the estimated useful life for a given probability of failure. Such predictions are based on analytical (mathematical) models and are calibrated based on field experience and observations. This applies to various types of loads that can be applied to the structure. It is also suitable for the resistance or load capacity of various types of materials and structural systems. Obviously, if you want to include deterioration in this type of analysis, you need a model to predict the level of possible deterioration at the structural engineering level. Ideally, the model should reach the corresponding complexity and be carried out based on actual conditions and relevant past experience. calibration. Predicting the expected behavior of the structure and its degradation is very important because (1) there are actually many reinforced concrete structures, and (2) the production of reinforced concrete structures with long-term durability has economic and environmental benefits.
The models used for prediction can be at different levels of complexity and different levels of detail. In all cases, they must be based on a good understanding of the basic processes involved, although this may depend on the level of abstraction appropriate for the intended use. For example, the designers of the Sydney Harbour Bridge do not need to know much about the atomic structure of steel, but they do need to know how this steel behaves at a level suitable for engineering design. This means that regardless of the level of abstraction, all models should be consistent with sound and coherent data. They should also be calibrated based on such data. Ideally, the model should be analytical in order to make reliable, valid and defensible inferences. In principle, these requirements also apply to knowledge in so-called “heuristic” models and so-called “expert systems”. Once these are used for forecasting, there will be assumptions about trends and the validity and methods of inferences, regardless of whether these assumptions are clear. In all cases, prediction implies a model inferred from the existing knowledge base. If the model is not understood or the model is imperfect, then any prediction is problematic at best, and “trajectory” at worst.
As far as the deterioration of reinforced concrete (RC) structures is concerned, perhaps the most concern is the structure exposed to the marine environment or chloride conditions. At least for many structures, the concrete matrix itself is not an important factor, but it may be an important factor for infrastructure such as sewers (Wells and Melchers 2015). Here, the durability of the concrete matrix itself is not considered. In addition, by focusing on high-quality, low-permeability concrete, the carbonization of concrete can be ignored as a mechanism to enhance corrosion. Practical experience shows that even under long-term exposure, the depth of carbonization is only a few millimeters (Parrott 1987). Similarly, this article ignores issues such as alkali aggregation reactions.
It is believed that the marine environment can cause erosive corrosion relatively quickly, causing structural damage, especially for unprotected structures. There are many examples in practice to support this view (eg Wig and Ferguson, 1917; Wakeman et al., 1958; Lewis and Copenhagen, 1959; Lewis, 1962; Gjorv, 1971, 1994). Protection can take the form of physical barriers, such as ceramic or other low-permeability tiles, protective coatings (such as durable coatings), bituminous materials or thick membranes, waterproof concrete coverings, or shells inside the weatherproof enclosure of buildings. For new structures, it can also take the form of cathodic protection, and for structures that have shown signs of distress, it can also take the form of cathodic protection. If you use these systems, and provide good design, good execution, and, importantly, well-maintained systems, long-term durability will not be an issue. Alternative systems that avoid the use of steel bars often lack the strength, ductility, or economy that conventional reinforced concrete usually has. However, a question must be raised as to whether these protection or alternative systems are absolutely necessary.
What is sometimes forgotten is that there is a large number of documents documenting the unprotected history of successful, durable reinforced concrete structures long-term exposure to the marine environment (Wakeman et al., 1958; Lukas, 1985; Ozaki & Sugata, 1998; Broomfield, 1999). ; Gjorv, 2009; Melchers et al. 2009; Angst et al. 2012; Melchers et al. 2017). There is also evidence that historically, very strong concrete structures were made of seawater (alkaline materials such as coral in some cases) (Wig & Ferguson, 1917; Narver, 1954; Wakeman et al., 1958; Dewar, 1963; Mather, 1964; Boqi et al., 1983; Burnside & Pomerening, 1984). Despite this experience, due to laboratory observations on reinforced concrete samples (Shalon & Raphael, 1959), and on the basis of some countries, apparently in the 1960s, most countries banned the use of seawater in concrete (Richardson 2002; Gjorv , 2009). Electrochemical testing (see Escalante and Ito, 1990).
It should be clear from this brief introduction that building a model for predicting the enhanced corrosion that may occur in the marine environment raises some interesting questions. There is a lot of practical experience, that is, anecdotal “evidence”, but despite decades of research, it is still believed that people still do not fully understand the mechanism of steel corrosion in concrete structures exposed to the marine environment. Therefore, the corrosion of steel bars in the marine environment is currently referred to as “chloride-induced” corrosion. In order to help solve this problem, the next section will summarize the current conventional knowledge and use theoretical and conventional methods to predict the chloride-induced corrosion of steel bars. Then it briefly summarizes the field experience extracted from the actual structure. The following is a description of some of the latest research results and an overview of the logical model. These developments provide new ideas for the meaning of “chloride-induced” corrosion. They also pointed out ways to develop better analytical models to predict expected corrosion.
For a long time, the accumulation of chloride in the concrete matrix of reinforced concrete structures is related to the occurrence and subsequent development of severe steel corrosion. Although the correlation does not prove any cause and effect, for a long time, the traditional view has held that when the surface chloride concentration on the steel bar reaches a high enough level (called the “threshold concentration”), steel corrosion will occur. Richardson (Richardson), 2002; Hunkeler (Hunkeler), 2005; Gjorv (Gjorv), 2009). Since seawater is no longer allowed to be used as mixed water, the main source of chloride is usually the external environment, usually the marine environment, and there are also sources of deicing salt. Therefore, the chloride must diffuse (or otherwise spread) to the steel reinforcement through the concrete protective layer. Coupled with the critical concentration of chloride will cause the critical value of corrosion, which has led to special emphasis in modern practice on the use of concrete with low permeability and (absorption) absorption characteristics and a thicker concrete cover. Similarly, great importance is attached to using Fick’s second law to simulate the diffusion of chloride ions, taking into account surface effects such as rain leaching, wet and dry cycles, and the fact that actual transportation phenomena are not pure diffusion (e.g. Hunkeler, 2005). These two ideas-chloride diffusion and chloride threshold-are the basis of many current methods for predicting the life of concrete structures (eg Richardson, 2002; Hunkeler, 2005; Gjorv, 2009; Angst et al., 2009).
In practice, there are other effects that may cause problems with conventional methods. In seawater immersion and tidal environments, the inward diffusion may be very low due to the accumulation of calcium carbonate and magnesium carbonate on the surface of the concrete, which tend to act as a diffusion barrier. Although many researchers have conducted extensive studies for more than 20 years, focusing on determining the “chloride threshold” at which corrosion begins, the concept is still as elusive as ever, with very large changes between different environments and high uncertainty (Repeatability), even in nominally similar environments and conditions (Bentur et al., 1997, p. 31; Angst et al., 2009). By citing other standards, the scheme will not change much, such as the ratio of chloride ion concentration to hydroxide ion concentration in concrete originally proposed by Hausmann (1967) (ie (Cl –)/(OH–)). ).
Although there are many electrochemical methods (Bentur et al. 1997) and laboratory studies (Angst et al. 2017), the inability to closely link chloride concentration with steel corrosion suggests that the understanding of the precise processes involved is still imperfect. It is becoming more and more obvious that, compared with steel corrosion, the mechanism involved in corrosion is more complicated than the direct action of simple chloride. As a result, the term “chloride-induced” corrosion is now up to date. After previous work (Melchers & Li, 2006; 2009; Melchers 2010; 2015), we believe that the mechanism involved will be clarified by using observations of actual concrete structures (especially those that have shown excellent overall resistance to concrete) Progress can be made. Long-term steel corrosion.
Gjorv (2009) gives a useful summary of the long-term behavior of reinforced concrete structures, including long-term behavior in a chloride environment. In addition, a summary of many literature reports is provided, dating back to the early 1900s (Melchers & Li 2009). Both data indicate that although there are many cases where reinforced concrete structures have poor durability and serious corrosion problems, there are also many cases where they have good long-term performance. Both reviews indicate that a great deal depends on obtaining high-quality, low-permeability concrete made with “good craftsmanship”. Compared with concrete made from igneous rock and similar aggregates, concrete made from limestone and (non-reactive) dolomite appears to exhibit greater enhanced durability (Melchers and Li, 2009).
Not everything attributed to the chloride effect should be attributed this way. The physical or chemical damage to the concrete matrix may be the cause of the subsequent “chloride-induced” corrosion of steel bars. For example, road bridges without asphalt wear or cycling routes have been found to suffer severely enhanced corrosion, which is usually attributed to deicing salts applied during frost periods, such as calcium and sodium chloride. However, closer observations show that the physical damage to the concrete cover of the bridge deck caused by motor vehicles and trucks is the main cause of corrosion of the steel bars, rather than the inward diffusion of chlorides in the deicing salt (Beaton & Stratfull, 1963). ; Lukas, 1985; Volkswein & Dorner, 1986). Similarly, the change in volume caused by the alkali metal carbonate (ACR) and alkali silica reaction (ASR) may also cause damage to the concrete matrix that protects the steel. These causal mechanisms should be separated from corrosion directly attributable to chlorides (Jensen, 1996; Richardson 2002; Gjorv, 2009). It should be clear that once the concrete protective layer is damaged, the steel bar is the steel exposed to the environment and will corrode under any circumstances, regardless of the performance of the concrete. Another example is cracks transverse to the reinforcement direction. Note that such cracks are usually the result of structural bending or shearing. Conventionally, concrete with acceptable cracks below 0.3 mm is acceptable, although the standard is mainly based on testing laboratory concrete specimens in a relatively short exposure time, including in a simulated marine environment (Beeby, 1978). However, as shown below, for long-term exposure, the standard needs urgent revision.
Although short-term tests and electrochemical tests are used in practice to estimate the life expectancy of structures (for example for design and evaluation purposes), any such tests must be verified against actual conditions. It is in this area that the problem arises. Short-term accelerated tests (such as the use of salt spray, higher temperature, current applied to the steel, etc.) cannot measure the same physical and (chemical) chemical characteristics that drive the long-term corrosion process. Not only for steel corrosion (Poursaee & Hansen 2009), but also for the more general sense (Lee et al., 2010), the relationship between the results obtained in this way and actual usage behavior is still weak. In short, all accelerated test results need to be verified and can only be done by comparing with actual field observation results.
Since corrosion is usually a slow process, it takes time to learn directly from on-site corrosion behavior. This is the case with experimental plans. For example, 16 years of marine corrosion research in the Panama Canal area (Southwell & Alexander, 1970) and NBS soil corrosion work (Romanoff, 1957) were also carried out in similar periods. However, these have not solved the problem of steel corrosion. For any physical infrastructure that requires a service life of 50-100 years, it is too short in any case. In this study, an additional method (Chitty et al., 2005) that is also used to evaluate the corrosion resistance of nuclear waste containers is an “archaeological” method in which the behavior of old structures and systems is examined , Where reconstruction of operations and environmental programs and conditions are necessary.
The shortcomings learned from the practical experience of real structures are: (1) There are usually few structures available for research; (2) They are often one-off, so it is difficult to study causal parameters; (3) They do not have Sufficient documentation (or documentation no longer exists or cannot be found). However, some studies on reinforced concrete structures along this line have been described (eg Melchers et al., 2009; Angst et al., 2012). Figure 1 shows an example where there is a lot of background information (Wood, 1948). This case and other cases (Melchers, 2010, 2011, 2015) show some similarities, including:
1. The chloride concentration on the steel bar is very high, except for very slight corrosion, it is not necessarily the “cause”, and for many years or even decades, there is usually no visible signs of corrosion at all.
2. For high-quality concrete, the occurrence of severe destructive corrosion takes place much later than any form of corrosion initiation measures.
3. The mechanism of causing severe steel corrosion is not clear, but it seems to be different from the mechanism of corrosion.
4. Very serious local steel corrosion has indeed occurred, and there are no obvious signs of external corrosion, such as rust, concrete cracking or spalling, but the clear explanation for such corrosion is still excellent.
Figure 1a. The side view and side view of the Phoenix caisson at low tide in 2009, similar to the caisson used in the Normandy invasion, ran aground in the harbour of Lanston, England. The caisson was built on the adjacent coastline (Haying Island) and had its back broken (in 2 locations) when it was launched in 1943. Please note that the tides are high and there is little corrosion of steel bars. The only aggregate available in this area is calcium.
Figure 1b. Close-up of calcareous aggregates and evidence of shells in concrete matrix. This indicates that sea water is used as mixed water.
Figure 1c. Close-up view of a 16mm diameter. Steel bars extracted from seemingly sound concrete. The steel bars did not crack along the concrete, and there was no normal reddish brown rust.
The conventional model for the initiation and development of reinforcement corrosion is usually attributed to Tuutti (1982), but was later proposed by Clear (1976) (Figure 2a). There was little or no corrosion before the corrosion start time ti after the corrosion started (Richardson 2002; Hunkeler 2005; Gjorv 2009). It has been suggested (Weyers et al., 1997; Bentur et al., 1997) that severe corrosion may occur at some time after ti, and this delay is caused by the need for chlorides to diffuse from the external environment into the steel. , Such as the slight cracking of concrete caused by the bending of beams (Francois & Arliguie, 1999).
Based on field studies (for example, Melchers et al., 2009) and studies of many reports, Melchers and Li (2006) proposed that certain limited amounts of corrosion may start at the beginning time ti, but long-term corrosion is severe. It usually takes a while to start (Figure 2b). In this model, the period (0-ti) corrosion is basically zero or negligible. Subsequently, a relatively short period of corrosion begins at the start time ti, after which there is no or only a very moderate increase in corrosion, until severe damage begins at the activation time tact. corrosion. Melchers and Li (2006) proposed that the corrosion process in the 0-ti period is different from the process that causes severe long-term corrosion. They also proposed that the active corrosion stage is the result of concrete alkali loss, especially because the presence of chloride accelerates the loss of Ca(OH)2. Based on recent research findings, this aspect will be considered in more detail below.
Figure 2 (a) shows the classic model of corrosion (damage) until ti, and then severe corrosion (damage) appears immediately, (b) an improved model with separation between the initiation of ti and the alert initiation of corrosion (Melchers and Li, 2006 ).
Using the separation of ti and tact, Melchers and Li (2009) detailed and analyzed about 300 individual reinforced concrete structures from literature and various reports. The analysis originally included cases of alkali-aggregate reaction (AAR), but it was later deleted due to lack of representation. The analysis reinforces the notion that in practice, the separation between ti and tact is very common, at least for high-quality reinforced concrete structures. The analyzed data is summarized in Figure 3. The graph shows the estimated values ​​of ti and tact as a function of the estimated chloride content of concrete made with two different types of aggregates. Although the data is scattered, it is clear that there is a considerable difference between ti and tact, which applies to both types of aggregation. It is also clear that concrete made with (non-reactive) limestone and dolomite lasts longer both times.
Figure 3. Initiation time (ti) or active corrosion (intermittent) as a function of chloride concentration. For (a) igneous rocks and similar aggregates, and (b) carbonate aggregates, the chloride concentration is close to increase. Please note that due to differences in data sources, the degree of data dispersion is very high. A typical range of acceptable chloride concentration thresholds is shown.
More recently, independent support for the separation of ti and tact came from observations of long-term exposure to laboratory pre-split beams (Yu et al., 2015). From regular detailed observations, it can be observed that some corrosion occurred shortly after the first exposure, but subsequent observations indicated that the corrosion actually stopped shortly afterwards, remained in this state, and then increased again.
In addition to the obvious difference between ti and tact, the most obvious aspect of Figure 3 is that the effect of chloride concentration on ti and tact is present in low concentrations of chloride, while it is much weaker at high concentrations. Collectively used. For real reinforced concrete samples exposed to high humidity for more than 10 years, the chloride content has also been observed to have little effect (Melchers & Chaves, 2017). These results are contrary to the results of many earlier laboratory tests, which seem to indicate that chloride has a considerable influence. However, the important difference is that most of these early observations are for samples exposed to accelerated test conditions or test durations shorter than 10 years (or both), and very few real concrete is used in the test procedure . This suggests that these tests may not adequately represent conditions related to corrosion of steel bars inside actual concrete structures with actual service life.
In fact, the most critical aspect of steel corrosion is the damage to the structure. As defined in Figure 2b, this is the activity, destructive corrosion related to the beat at the beginning. It can be clearly seen from Figure 3 that it has nothing to do with the type of aggregates and the chloride concentration. Usually the start time tact is much later than the start time t1. This supports the view that these two times are driven by different corrosion mechanisms. Consider these possibilities next.
Before continuing, it will be helpful to briefly modify some basic concepts. In order for corrosion to occur, the thermodynamic requirements of the chemical reaction involved must be met, which is represented by the Gibbs free energy or equivalent Pourbaix diagram (Jones, 1996). No electric potential is applied (for example, electric potential is often applied in electrochemical tests). In pure water, ferrous iron may corrode only when the pH is        Fresh concrete has a relatively high pH, ​​usually about 13.5 to 14, which is given by a small amount of alkali metals NaOH and KOH. These are relatively easily leached, and then the main alkali, calcium hydroxide, Ca(OH)2 tends to maintain a pH of about 12 or lower because the concentration of Ca(OH)2 decreases. Although the steel interface maintains such a high pH value, corrosion still cannot start-the pH value needs to be lowered. This is usually the case with “high-quality” concrete-”high-quality” concrete is made with the appropriate cement content (and type) and low permeability, usually reflected in higher concrete strength and density. Generally, this type of concrete also has a higher alkalinity (sometimes easier to visualize as the amount of “acid buffer capacity”), so it can absorb a large amount of “acid” erosion and still maintain a high pH value. It can be seen that the rate of pH decrease over time is controlled by the following factors: (a) the initial “alkalinity” of the concrete, (b) the rate of alkali leaching from the concrete, and (c) the rate of “carbonation”. Although pitting corrosion may occur at higher pH values ​​in a closed environment, general (uniform) corrosion only occurs when the pH value drops below about 9. Therefore, for general corrosion, it is necessary to lose alkalinity at the steel-concrete interface. Regarding Figure 3b, it should be noted that the solubility of calcium carbonate (CaCO3) in water is very low, with a pH of about 10-this may be the reason for the larger ti and tact values.
After careful examination of the actual case, the above-mentioned rather theoretical concept is obvious. In one case (Melchers et al. 2009), it was found that the pH value of concrete remained above 10 for more than 60 years, and it only existed in local areas where the pH value was lower than 9, such as in the fine line cracks of concrete (Probably allowing air, sea water and rain to enter), corrosion of steel reinforcement is observed. Similar observations were made for 80-year-old reinforced concrete piles in tidal seawater (Melchers et al., 2017) and other conditions earlier (such as Chitty, 2005).
Now turning to active corrosion, the conditions under which corrosion must occur in high-quality concrete can be called “wet and stagnant”. Obviously, moisture is necessary for corrosion to occur, but moisture must be present at the correct location at the steel-concrete interface. Stagnant conditions are of great significance to corrosion, as evidenced by the classic work reported by Heyn & Bauer (1908). They observed little difference in corrosion loss for low carbon steel strips in weakly immersed, surrounding but stagnant solutions of many different salts (including NaCl). These observations were initially controversial, but were ultimately explained by Dora Brasher (1967). She believes that under stagnant exposure conditions, the weak passivation film usually formed on mild steel, as first observed in the 1930s (see Evans, 1960), can remain in place, and under such conditions The availability of oxygen determines the rate of corrosion. Mercer & Lumbard (1959) carefully controlled laboratory exposure experiments support this view. They observed that higher concentrations of chloride in the water increased corrosion, but only in a non-stagnation state. Under stagnant or near stagnant conditions, chloride has little effect on the amount of corrosion. One of the unfortunate aspects of many experimental laboratory testing procedures is that stagnation is rarely used. These common solutions are not to use model concrete, but often stir instead of stagnation to accelerate the corrosion process (Poursaee & Hansen, 2009).
As mentioned above, great attention has been given to the so-called onset time (ti) in the corrosion literature, which is generally considered to correspond to the onset of severe enhanced corrosion (ie in the notation of Figure 1). In Figure 2b, assume that tact = ti). However, if ti and tact are different and independent entities, based on the notion that they are governed by different mechanisms (Melchers & Li, 2006), as shown in Figure 2b, the problem lies in what mechanism causes the “initialization”. This is being studied in a long-term exposure experiment started in 2004.
The experiment involved exposing a large number of concrete specimens each 40 x 40 x 160 mm long to various water/cement and aggregate/cement ratios, all of which were made of the same material. Each sample contains a smooth low carbon steel rod with a diameter of 6 mm placed longitudinally. Except for a test series specifically described below, all samples were made using standard commercial (hybrid) Portland (GP type) cement and were purchased from a manufacturer. All samples (and some continued) were continuously exposed to a laboratory fog chamber environment with high humidity (>95%RH) at 25°C. This humid environment ensures that carbonation is reduced. Pacific sea water was used as the mixed water for most of the samples, and the rest was made with fresh water (drinking water). Whenever checked in the fog chamber, it was observed that the external water saturation of the sample was observed. One or two samples are taken from each sample set approximately every 12 months. These specimens are inspected from the outside and then opened to reveal the state of the steel bars and the state of the concrete inside. Overall, the patterns observed over the years are consistent, although certain features have increased over time. As mentioned earlier (Melchers & Chaves, 2017), frequent sampling with one or two samplings is more informative than sampling at a lower frequency with more samplings at any one time.
After 2 to 3 years of exposure, it has been noted that many samples show no corrosion, while others show only limited corrosion. Usually, this kind of corrosion is pitting corrosion or localized corrosion along the lower part of the steel bar away from the casting direction (Figure 4a). This is also the side closest to the steel bars of the shaking table used for concrete compaction.
Figure 4. (a) Typical early localized corrosion (pitting) along the lower part of the rebar, with little or no corrosion elsewhere; and (b) a typical wet air gap in the concrete matrix adjacent to the location of the rebar.
This pattern continued for the next few years, only a modest increase in corrosion was observed. In general, concrete with the greatest apparent permeability and lower density exhibits relatively more corrosion. After 10 years of exposure, the concrete with the highest aggregate cement ratio (ie, the lowest total cement content) and therefore the highest permeability showed moderate corrosion around the steel bars, but not enough to cause the steel bars to crack longitudinally. On the contrary, even after 10 years, the denser concrete (concrete with low aggregate/cement ratio) still retains the pattern shown in Figure 4a.
Especially for high-density, low-permeability concrete, after inspecting the internal concrete surface adjacent to the steel bars, it was found that after the sample broke, there were multiple voids in the thin silicate layer on the concrete surface. Usually around the rebar. The voids contain air and, like the surrounding concrete, are wet, although not necessarily completely filled with water (Figure 4b). The voids mainly exist on one side of the steel bar, and in all cases are found along the bottom of the steel bar with respect to the casting direction of the concrete in the steel mold (Figure 4a). It is important that the concrete vibrates in the same direction on the laboratory shaker. Therefore, the bottom of the strip will be where any air bubbles in the concrete may be trapped. It has been observed that the lower density concrete contains more air voids, including at the steel-concrete interface. Generally, the corrosion of the steel bar along its underside is in the form of pitting corrosion, usually there are multiple pitting corrosion nearby, and it is considered based on a small amount of observations on both the concrete and the adjacent steel surface to reflect the pattern of the void . It was observed that for the less dense concrete, the voids at the steel-concrete interface are more frequent and larger. For these concretes, the density of pits per unit area is also greater.
Multiple measurements of the concrete pH of cracked concrete (next to and far from the steel bars) show that even after 10 years of exposure, the pH ranges from about 12 for low-permeability concrete to about 12 for low-density concrete. 10. , Apparently porous or permeable concrete. Since the chloride concentration in the pore water at the beginning of the experiment was similar to the chloride concentration in the seawater used for concrete (about 35,000 ppm), the only corrosion that occurred at that time (Pourbaix, 1970) was pitted. Indeed, this is what was observed. In addition, since the high chloride content has existed since the beginning of the experiment, it has been so from the beginning, and may last through the entire 10 years of experimental exposure, although the chloride concentration may decrease over time. Leaching. However, the precise change in pH (>9) and the precise chloride content during this period are not important-conditions suitable for pitting corrosion (only) are common at the beginning and are likely to be present throughout the process.
Although there is no mention of whether vibration or other compaction techniques are used (Horne et al., 2007; Bouteiller et al., 2012), laboratory research projects on model concrete have earlier reported similarities to the casting direction. Gap. ). In those experiments, local differences in pore water and chloride content of concrete were observed in areas containing voids. In addition, many other factors that may be potentially important for corrosion in the concrete-steel interface area have been identified, but no conclusions have been drawn regarding their potential effects (Angst et al., 2017). Moreover, the mechanism of corrosion initiation is not proposed.
For real concrete that is different from model reinforced concrete, the voids and separation phenomena observed at the concrete-steel interface have a long history. They are attributed to the degree of compaction of concrete during construction (eg Nawy, 2008). However, at least at the macro level, voids and separations are not always observed. For example, the above two field studies (Melchers et al.: 2009, 2017) show that there are no visible voids beside the steel bars, except for rare and very special locations (see section 4.5 below), and there is even no sign of the beginning of corrosion. After 60 to 80 years of exposure to the marine environment and very high concrete chloride concentrations. In both cases, only when the local concrete has a pH value of less than 9, can the steel corrosion be visually observed. Both structures were built before extensive use of mechanical vibration equipment. This means that concrete compaction could have been manually tamped and tamped. It is easy to prove that if done well, these manual techniques can pour concrete with negligible voids next to the steel bars even on the side away from the casting direction. Another factor is that although it is known that limiting the water-cement ratio is important for strength, there is no specific limit on the amount of water that can be added to the mixture when making these concretes, and the focus is on achieving viable mixing in practice (Wood, 1948) . This will help air bubbles leave the steel-fluid concrete interface area during compaction, rod strikes or vibration, including on the underside of the rod.
Combining these observations together leads to the conclusion that the onset of corrosion (at time ti) is mainly the result of voids in the concrete beside the steel bars, which is the result of unsatisfactory concrete compaction. Specifically, it has been proposed that the mechanism that causes enhanced corrosion (ie, the ti mechanism defined in Figure 2b) is the classic differential aeration (Evans, 1960; Jones, 1996), which is formed by adjacent moisture pores. To the steel surface (Figure 5). Under this mechanism, the void provides air (and oxygen) and enough water (usually oxygen-containing water) to maintain the usual cathodic corrosion reaction in the void. In the presence of higher chlorides, even under high pH concrete, differential aeration will appear as pitting corrosion, which forms at the edges of the moisture pores (Figure 5). Indeed, there are reports (Verbeek, 1975; Angst et al., 2017) that the corrosion pits are not completely aligned with the pore centers, which can be interpreted as consistent with the situation in Figure 5.
For most concrete in marine environments, water in the form of pore water is likely to be continuously supplied inside the concrete matrix. Under these conditions, the pitting corrosion process may stop only after the oxygen in the cavity or the oxygen dissolved in the pore water is exhausted. If the permeability of the (wet) concrete allows oxygen to diffuse to a certain extent from other voids in the concrete matrix or even from the external environment, pitting corrosion may continue. For wet concrete, this may occur at a very low rate, so any corresponding corrosion will also occur at a very low rate. Obviously, for small voids, the pitting effect is small, and for large voids, the pitting effect is greater, which explains why for very dense concrete, traditional aerated corrosion products are usually not observed. More generally, if there is not enough pore water to sustain the corrosion reaction, pitting corrosion may only stop temporarily.
For a long time, it has been believed that oxygen and water, especially humidity in pores or voids, are responsible for corrosion-enhancing materials, and such pores (or voids) provide the necessary cathodic reaction (for example, Andrade et al., 1990). However, the differential aeration mechanism shown in Figure 5 and the fact that oxygen is mainly limited by oxygen in the voids rather than transportation through the concrete cover have not been previously proposed.
In general, the mechanism proposed above is consistent with various actual observations. For example, it has been observed that, as shown in Figure 2b, after the initial stage of exposure, after corrosion, there is a static period between ti and tact (for example, Francois & Arliguie, 1999; Yu et al., 2015). The proposed mechanism is also consistent with the observation of iron rust, such as “green rust” known to form under low oxygen conditions (Gilberg & Seeley, 1981), and is usually observed when the steel bars are inspected before the production cycle. (Melchers et al., 2009, 2017). These observations can be interpreted as reflecting the gradual consumption of available oxygen, which is initially exhausted from the voids and finally diffused by oxygen. Also in this case, the role of chloride is only to allow pitting corrosion to occur at a higher pH than in the absence of chloride. In this sense, chloride induced the initiation to occur earlier. This is consistent with field observations (Figure 3).
Finally, regarding the time when the corrosion starts, it should be pointed out that for the RC structure made of seawater, for example, in the above experiment, ti→0, because the pitting corrosion conditions can be obtained in a high chloride environment from the beginning. For RC structures that have no significant chloride content but are exposed to a chloride environment, it is expected that it will take some time before these conditions are reached by the transport of chloride through the concrete cover. For lower external chloride exposure conditions, this time can also be expected to be longer. Indeed, this is consistent with practical experience.
The time for active corrosion is very important. As mentioned earlier, it was not always separated from “initialization” in the past. In the case of poor-quality, permeable concrete, they may be closely connected, but most of the field evidence of the actual structure shows very significant differences (Figure 3). Estimation of the beat is usually based on the observation of the time of cracking or peeling of the concrete cover due to a large accumulation of corrosion products or obvious signs of corrosion. These are actually hindsight observations. They did not inform the mechanism that may cause active reinforcement corrosion, and the time interval. Let’s explore this now.
As part of the experimental plan outlined in Section 4.3, a series of test specimens (C series) were made with sulfate-resistant cement, rather than the commercial mixed Portland (GP type) cement used in all other series. This has brought unexpected benefits. Sulfate-resistant cement has unique properties, that is, fresh cement or concrete made with it has a dark gray-black color. It is well known that the dark color due to oxidation of iron and manganese sulfide will become lighter in the atmosphere (within a few days) (Hanson, 2015). However, for all concrete mixtures made with this cement in the experimental procedure, the edges of the light-colored concrete were observed immediately after the specimen was opened (Figure 6). Similar observations have been made when the specimen ruptured after 3 years of exposure and the following years. At any time, the light-colored rim or outer area is wider for the more permeable concrete. For a longer exposure time of the corresponding concrete mixture, the width is also greater. The samples made with commercially available GP cement did not show sharp contrast as shown in Figure 6. However, similar to the pattern, subtle differences in color can also be observed between the darker inner core and the lighter outer area. You can see concrete made of sulfate-resistant cement.
Figure 6. Sample C4 made with seawater, showing the edge of light-colored concrete that opened immediately after opening the sample after 10 years of continuous exposure to the fog chamber. The dark interior is a typical feature of concrete made with sulfate-resistant cement.
The analysis of dark and light concrete (shown in Figure 6) reveals interesting differences. When inspected by scanning electron microscopy and energy dispersive spectroscopy (SEM/EDS) and X-ray diffraction (XRD), it was found that the light-colored concrete contained essentially no Ca(OH)2, and almost no C (and therefore no There is calcium carbonate caused by carbonation). In contrast, dark concrete was found to contain high levels of Ca(OH)2 (Melchers & Chaves, 2016). Regardless of the ratio of water/cement and aggregate/cement, similar observations were made for the light and dark concrete of other concrete samples.
Light colored concrete all show relatively low (alkaline) pH readings, usually in the range of 8.5 – 9.5, while dark concrete shows higher pH readings (pH 10-12). The latter is consistent with the pH range of various concentrations of calcium hydroxide in the solution.
When inspected under a microscope, the light-colored concrete showed a more open and presumably more permeable structure than the dark-colored concrete in the inner area, indicating that the material has been lost from the light-colored area (Melchers & Chaves, 2016). It is speculated that due to this more open structure, atmospheric oxygen can enter the Ca(OH)2 depleted concrete. As a result, the iron and manganese sulfides in the sulfate-resistant cement are oxidized and become dark gray. black. It becomes the observed light color. Therefore, the color change is an inevitable result of the outward leaching of hydrolyzed calcium hydroxide. The overall results are summarized in Figure 7.
In summary, these observations indicate that, in accordance with the 2006 recommendations (Melchers & Li, 2006), calcium hydroxide gradually loses from the outside of the concrete sample to the inside, presumably due to leaching caused by dissolution in water. Importantly, although the dissolution rate of calcium hydroxide in pure water is very low (Sagues et al., 1997; Marinoni et al., 2008), the presence of NaCl is directly proportional to its concentration, thus speeding up the process (Johnson & Grove, 1931) . ). It is important in this regard that concrete with better permeability has a larger internal surface area available for the dissolution process. The pattern in Figure 7 is the same for all concrete mixtures made of sulfate-resistant cement – ​​the only difference is that the width of light-colored concrete is greater for more permeable concrete (such as rebound (Schmidt hammer) Surface) estimated) hardness reading) (Melchers & Chaves, 2016).
Figure 7. The appearance of internal concrete and the condition of steel reinforcement after ten years of exposure in the laboratory fog room. There is a significant difference between the internal pH value and the concrete composition. There is calcium hydroxide in dark concrete, while there is no calcium hydroxide in light concrete.
In order to evaluate the influence of chloride on the above mechanism, two other series of experiments (B and K) were conducted simultaneously with the above C series. They were treated as described above-in addition, the Ca(OH) 2 loss depth (d) (ie the depth of light-colored concrete) of each sample was also measured. The results are shown in Figure 8a as a function of rebound hammer surface hardness readings, and in Figure 8b as a function of water/cement (w/c) and aggregate/cement ratio (a/c).
Figure 8b shows that for any of the three concrete series, there is almost no effect on the a/c ratios of 2:1 and 4:1. It also shows that concrete with a larger a/c ratio of 6:1 in terms of d has greater calcium hydroxide loss. As mentioned above, this may be because the concrete with such a high air-fuel ratio has a larger open porosity (porosity), thereby allowing a larger dissolution area. In all cases, the d of the mixture made with sulfate-resistant cement (C series) is much greater than the corresponding B series. This is consistent with the lower strength and therefore higher permeability of the sulfate-resistant cement. For similar A/C ratios of 4/1 and 6:1, the d value of fresh water (K ​​series) is approximately half of that of seawater concrete (B series). This is consistent with chloride (in the B series) increasing the dissolution rate of calcium hydroxide.
Figure 8. The relationship between the depth of calcium hydroxide loss (d) and (a) rebound hammer readings as a (reverse) substitute for concrete permeability; and (b) water/cement and aggregate/cement ratios.
Although the scatter in each set of data points is high, the data does show a trend when the best fit plotting routine is used for fitting. Concrete made with sulfate-resistant cement (C series) has a greater depth of calcium hydroxide loss than concrete made with GP cement (B series). Both figures show that compared with concrete made with fresh water (series B) (given w/c and a/c ratios, or similar surface hardness values), the depth of calcium hydroxide loss is approximately 100% higher. (Series K).
The important actual observation in these results is the effect of concrete permeability (Figure 8a). The loss of calcium hydroxide is much higher, so the loss of permeable concrete is much higher than that of dense and impermeable concrete. This is consistent with actual observations, that is, a concrete structure made of low-permeability concrete with good quality and depth of coverage has a longer service life without significant corrosion of steel bars (Gjorv, 2009). This can also be observed for the situations mentioned in Section 3, except where the concrete cracks. In this case, it should be noted that the results shown in Figure 8 apply to uncracked concrete. It can be expected that rupture will affect the rate of alkali loss.
Under the action of stress, although high-level concrete shear stress is involved, its mode may be more complicated, and RC members tend to crack in the direction perpendicular to the steel line. For a long time, it has been thought that these cracks may cause protection that the concrete protective layer usually cannot provide (see Francois & Arliguie 1999). On the other hand, longitudinal cracks are usually the result of corrosion, rather than a direct cause of enhanced corrosion.
Early studies on the possible effects of concrete cracking on steel corrosion have concluded that the importance of cracking is not great, and self-healing usually reduces the long-term effects. This view is observed by the high concrete pH adjacent to the cracking Support, and in some cases, use fresh water. Similarly, for cracks penetrating steel bars, the severity of corrosion seems to be proportional to the crack width (Beeby, 1978; Bentur et al., 1997, p. 50). Others believe that the existence of cracks is more important than their width (Mohammed et al., 2001). Importantly, most of these conclusions are based on relatively short-term tests and observations of structures built from limestone or similar aggregates. As mentioned above, these aggregates tend to impart greater alkalinity (or buffer capacity to neutralize acids) to concrete. This will help self-heal.
Recent large-scale laboratory tests have shown that regardless of the width, cracks transverse to the reinforcement direction are allowed to start corrosion at the base of the crack. Then, this corrosion may extend along the damaged steel-concrete interface (Francois et al., 2012). Similarly, the observation of the actual reinforced concrete structure also provides strong evidence that the almost unobservable cracks transverse to the steel bars, and even fine-line cracks, may cause severe and localized corrosion of the steel bars (especially under long-term exposure conditions). (Figure 9) (Melchers & Li, 2009; Melchers et al., 2017). This is to be expected because such cracks will expose more of the concrete matrix to the possibility of alkali dissolution. As mentioned above, the key corrosion rate control process is the rate of alkali dissolution. In extreme cases, extensive cracking will expose more of the concrete matrix to the possibility of alkali dissolution, thereby reducing the duration of the buffer protection that concrete can provide sufficient, alkaline, and thus neutralizing acid. In this article, this is related to the effect of chloride, but a similar effect will apply to carbonation. In both cases, this will directly affect the effective life of the steel bars in the concrete structure.
The above observations and the model shown in Figure 2b directly lead to how to model the life expectancy tL of RC structures subject to “chloride-induced” corrosion:
Where ti is the time to start corrosion as before, tpass is the time after the corrosion starts and before the active corrosion starts. As shown in Figure 2b, the life expectancy tL will include a period of time when corrosion has begun. Without compromising structural performance or structural usability, it is estimated that the time period during which this active corrosion can be tolerated is the subject of a large number of recent studies, especially in the field of structural engineering. During this period, this time period is considered to be the extra time beyond the time interval after the onset of severe corrosion.
Although the model of initiation mechanism is proposed above, quantifying the initiation time is still a research problem, and now it is necessary to consider parameters such as compaction and void size. The time period tpass represents a fairly similar difficulty because it is not the mainstream of research activities, and there is no joint effort to conduct research with experience or other means. It seems that the best way is to directly estimate the speed of the aircraft. Again, there are currently only limited public studies focused on directly estimating the interval, but information such as those used to obtain FIG. 3 and information such as those presented in FIG. 8 can be used for the first estimation. In addition, a conservative estimate of the intermittent can be obtained by assuming that the chloride is present from the beginning, so there is no need to estimate the time for the chloride to migrate into the concrete. In this regard, the results in Figure 8 can be considered conservative because of the presence of chloride from the beginning.
For example, consider a (unbroken) concrete with a high water-cement ratio of 0.6 and a total cement ratio of 4:1. If it is made with fresh water, Figure 8 shows that the depth of calcium hydroxide loss in 10 years will be d = 1.1mm or 0.11 / y. If the concrete is made of seawater, d = 3.2 mm or 0.32/y in 10 years. This means that concrete structures with a cover layer of 50mm will begin to actively corrode fresh water and seawater in 450 and 160 years, respectively, provided that the concrete is not broken. Although these estimates appear to be high, estimates of concrete containing high concentrations of chloride are not completely impractical. It can be compared with the situation of the Honeybrook Bridge pile, which is exposed to submerged, tidal and splash zones in the Pacific Ocean. These results show that after 80 years, despite the very high chloride concentration, there is almost no loss of calcium hydroxide in the concrete matrix, and the pH reading is about 12, which is about ten times the normal acceptable threshold (Melchers et al., 2017). However, as described in section 4.4 above, where the concrete cracked to the steel reinforcement, very severe corrosion occurred. This is also the case for 60-year-old reinforced concrete structures. As shown in Figure 9, the corrosion is quite limited but very serious. As suggested in Section 4.4, the key aspect of predicting the life of reinforced concrete structures now appears to be (transversely of the concrete matrix) cracking deep enough to accelerate the loss of calcium hydroxide or other alkalis.
From a practical point of view, the most important parameter derived from the current work is wit. It describes when severe corrosion can be expected to begin, regardless of any previous short-term corrosion indicated by the parameter ti. In most actual structures, once severe corrosion begins, the life of the structure will be severely damaged. Of course, simulation can be performed to determine the extra life (tL) after the beat is reached, but the purpose of any design or rehabilitation exercise should be to extend the time range of the beat, as long as it is economically feasible.
The second actual result of the current work clearly shows that concrete cracking, especially transverse cracking through or near the steel bar, such as cracks that may be caused by high bending stress, will have a very negative impact on the expected life of the steel bar. . Anti-corrosion. The immediate implication of the current work is that in addition to such cracks that may have any effect on the inward transmission of chloride and/or oxygen, they also provide a way for the outward leaching of alkali, thereby providing a way to reduce the pH near the steel bar. This way, finally allowing active corrosion to begin, initially at the crack opening. If this type of cracking is extensive, such as usually cracking in beams subjected to bending or temperature changes, the effect is similar to the overall loss of permeability of the concrete surface. Early research results allow the size of concrete cracks to be limited in the design code, but it seems to ignore the effect of crack depth, so it must now be considered insufficient. As mentioned earlier, this research is based on short-term tests and does not allow the development of the mechanisms revealed by the current long-term experimental work, which can be seen in various field observations and other recent experimental work (see Francois et al. ( 2012). Therefore, the design rules for cracks need to be modified, especially under ocean exposure conditions.
The third aspect drawn from the current work is that it enhances the key role of concrete alkalinity. For a long time, the importance of this has been regarded as a key factor in maintaining the corrosion resistance of steel bars. The latest experimental results of chloride-induced corrosion reinforce this effect, as it has now been found that alkali dissolution is critical to the development of conditions, including in a chloride environment, allowing active and severe corrosion of steel to begin. Since high alkalinity is conventionally related to high cement content, it is directly inferred that high cement content can achieve longer durability. However, in contrast to this, increasing environmental and economic pressures require the reduction of cement content in concrete. In this case, the results shown in Figure 3b are important because they show that there are other ways to improve the acid neutralization capacity of concrete, namely using (relatively high) low alkalinity aggregates (such as limestone and non-reactive Sex) Dolomite. Obviously, when they are used as aggregates, they increase the time for active corrosion. This observation is consistent with (anecdotal) experience. For example, in Europe (Lukas, 1985) and Florida (Lau et al., 2007) the main aggregate is calcium. This is also consistent with the difference in RC durability between different regions of the United States, which has longer durability, which corresponds to those whose main source of concrete aggregate is calcium (Melchers & Li, 2009). For these various situations, it has not been noticed that the potentially lower strength of limestone (but not dolomite, such as marble) is an important (negative) factor. These considerations also indicate that there may be other ways to increase the alkalinity of concrete, such as adding other types of alkali to the mixture. This is an obvious area of ​​research. Another area that needs to be studied is the greater use of cement to increase the balance between the useful life of structures affected by durability issues and the resulting life. This may be due to the initial need not to use cement for renewal or maintenance, and the overall use of cement was reduced during the entire service life of the RC structure, which offset the negative impact of the initial large-scale use of cement on the environment.
An important aspect of current work is the model on which Figures 2b and Eqn are based. 1 Mainly from the direct or indirect observation of the actual RC structure or the behavior of a sufficiently realistic substitute. The method used relies only indirectly on observations that are usually manual laboratory studies, such as those that use alternative solutions of pore water instead of concrete. Since the results of electrochemical tests are also carried out under quite artificial conditions, and are usually carried out under accelerated conditions, current work rarely relies on the results of electrochemical tests. These techniques usually have limitations (Lee et al., 2010; Poursaee & Hansen, 2009), but they are not always clear and may have a serious impact on the transformation of experimental results into realistic structures. Although there are some well-known difficulties in interpretation and data support, especially for older cases, this work does not attempt to solve the problems related to these various technologies, but focuses on observation and interpretation from the actual structure.
In addition to the above content, the proposition of this work also points out many areas that need further research. Intermittent (and passing) estimates still represent a major research challenge. The variables that may be involved have been mentioned in the above expo, but the interaction between them and the direct impact on alertness need to be further explored, including:
2. The internal surface area available for alkali dissolution (related to the density and permeability of the concrete matrix),
3. (Chloride-driven) alkali dissolution rate (a function of chloride concentration (Johnston & Grove 1931),
4. The effect of concrete cracking on concrete permeability and severe (perhaps only local) steel corrosion.
In addition, in practice, it can be expected that there will be a certain degree of interaction between the inward diffusion of chlorides, resulting in a slow increase in chloride concentration (in Figure 1) and the effect of this concentration on the rate of alkali dissolution (In Figure 3). This aspect also needs further study.
Finally, the overall mechanism described in this article differs from traditional views in some important respects. They make chlorides work, which is very different from traditional knowledge, and can distinguish between the beginning of enhanced corrosion and the final, serious, active beginning. In general, the mechanism is consistent with the basic principles of corrosion science, and importantly, consistent with the practical experience and observations of high-quality reinforced concrete structures. The current results also make the corrosion caused by carbonization and the corrosion caused by chloride reach a certain degree of unity. In both cases, the loss of protective alkali will lead to eventual active corrosion.
Recent studies have shown that the main role of chloride in the so-called “chloride-induced” corrosion of steel bars is to accelerate the solubility of alkaline calcium hydroxide in the concrete matrix (including around the steel bars), thereby ultimately strengthening steel passivation. Usually, this takes a long time and is strongly affected by concrete permeability and concrete alkalinity, such as the alkalinity provided by cement or alkali aggregate. On the other hand, the mechanism of corrosion is mainly due to the poor compaction around the steel bar, which results in the local aeration corrosion pool. The separation of the initiation of corrosion and the initiation of active steel corrosion, the dissolution of alkalinity and the effects of the main concrete cracking all provide new ideas for modeling the time when severe steel corrosion occurs under simulated chloride conditions. The framework of this type of modeling and the research challenges are outlined.
The author thanks the Australian Research Council for financial support for some of the work reported in this article. With the support of the Civil Engineering Laboratory of Newcastle University, especially the support of Ian Janes and Goran Simonditch, (the late) Dr. Dick van der Molen tried to reduce Sulfate cement proposal, early research support by Dr. Toril Pape, etc. Recently, Dr. Igor Chaves (Igor Chaves) in the sample analysis and the University of Newcastle (University of Newcastle) Central Science Service Center The support has been recognized by everyone. In addition, thank the reviewers for their useful comments and suggestions.
Andrade, Alonso and Garcia, AM (1990), Oxygen utilization rate in steel corrosion, “Cement Research Progress” (3) (11) 127-132.
Angst, UM, Elsener, B., Jamali, A. & Adey, B. (2012), Concrete cover cracking due to corrosion of steel bars-theoretical considerations and practical experience, “Materials and Corrosion”, 63 (12) 1069- 1077.
Angst, Michigan, Elsener, Larsen, and Venezuela, Ohio (2009), Critical Chlorine Content in Reinforced Concrete – Overview, “Cement and Concrete Research 39, 1122-1138.
Angst, Michigan (Angst, UM), Venezuela, Venezuela, Larsen and B. Elsener (2017), Corrosion of steel bars caused by chloride: from the Norwegian COIN project Proc . Third International Conference on Durability of Concrete in Adelaide from October 22nd to 25th.
Angst UM and 15 others (2017), Steel-concrete interface, materials and structures, 50 (143) (24 pp.)
Beaton, JL and Stratfull, RF (1963), The influence of environment on the corrosion of steel bars in the substructure of concrete bridges, Records of Highway Research, Vol. 14, pp. 60-78.
Beeby, AW (1978), Corrosion of Rebar in Concrete and Its Relationship with Cracking, Structural Engineer, 56A (3) 77-80.
Bentur, A., Diamond, S. And Berke, N. (1997), “Steel Corrosion in Concrete: Foundation and Civil Engineering Practice”, E&FN Spon, London.
Boqi, C., Dinghai, H., Hengquan, G. & Yinghao, Z. (1983), Ten-year Field Exposure Test of Reinforced Concrete Durability in Port Engineering, Cement and Concrete Research, 13, 603-610 .
Bouteiller, V., Cremona, C., Baroghel-Bouny, V. and Maloula, A. (2012), Corrosion initiation of reinforced concrete based on Portland or GGBS cement: chloride content and electrochemical properties change with time, Cement and Concrete Research, 42, 1456-1467.
Brasher, Dora M. (1967), The stability of oxide films on metals related to corrosion inhibition. I. Low carbon steel with corrosive ions, “British Corrosion Journal”. 2, 95-103.
Burnside, Ohio, and DJ Pomeranian (1984), “Experience Survey of the Use of Reinforced Concrete in Floating Marine Structures”, report SSC-321, Ship Structure Committee, U.S. Coast Guard Headquarters, Washington, DC.
Chitty, W.-J., Dillmann, P., L’Hostis V. and Lombard, C. (2005), Long-term corrosion resistance of metal reinforcements in concrete-based on the corrosion mechanism of archaeological artifacts, Corrosion Science 47, 1555-1581.
Clear, KC (1976), Corrosion time of steel bars in concrete slabs. V.3: Performance after 830 salt applications per day, FHWA-RD-76-70, Federal Highway Administration, Washington, DC, April, p. 64.
Dewar, JD (1963), Workability and Compressive Strength of Concrete Made with Seawater, Technical Representative, TRA/374, C&CA, London.
Escalante, E. and Ito, S. (1990), “Measuring the Corrosion Rate of Steel in Concrete”, (In) NS Berke, V Chaker and D Whiting (Eds), “Corrosion Rate of Steel in Concrete”, ASTM STP 1065 , Am. Soc. Test materials, Philadelphia, pages 86-102.
Evans, UR (1960), “Corrosion and Oxidation of Metals: Scientific Principles and Practical Applications”, London, E. Arnold.
Francois, R. & Arliguie, G. (1999), The influence of microcracks and cracks on corrosion development in reinforced concrete members, Journal of Concrete Research, 51 (2) 143-150.
Francois, R. Khan, I., Vu, NA, Mercado, H. and Castel, A. (2012), “Research on the Effect of Local Cracks on Corrosion Mechanism”, “European Journal of Environment and Civil Engineering”, 16 (3) -4) 392-401.
Gilberg, MR & Seeley, NJ (1981), Identification of compounds containing chloride ions in marine iron corrosion products: an important review. Conservation Research, 26 (2) 50-56.
Gjorv, OE (1994), Steel corrosion in concrete structures exposed to the Norwegian marine environment, Concrete International, 16(4) 35-39.
Gjorv, OE (2009), Durability design of concrete structures in harsh environments, London, Taylor and Francis.
Hanson (2015) Sulfate Resistant Cement: Technical Data Sheet, Version 3.3, November 2015, Hanson Heidelberg Cement, Keaton, UK.
Heyn, E. and Bauer, O. (1908), “Uerber den Angriff des Eisens durch Wasser undwässerige Losungen”, “Stahl und Eisen”, 28 (44) 1564-1573.
Horne, AT, Richardson, IG and Brydson, RMD (2007), Quantitative Analysis of Microstructure of Reinforced Concrete Interface, Cement and Concrete Research, 37, 1613-1623.
Huet, B., L’Hostis, V., Miserique, F. and Idrissi, H. (2005), Electrochemical behavior of low carbon steel in concrete: influence of pH and carbonate content in pore solution, Electrochimica Acta , 51 (1) 172-180.
Hunkeler, F. (2005), “Corrosion in Reinforced Concrete: Process and Mechanism”, (In) Corrosion in Reinforced Concrete Structures, (Ed) H Bohni, Woodhead Publishing Ltd., Cambridge, 1-45.
Jensen, V. (1996), Current experience of aggregate testing in Norway, (Proceedings of the 10th International Conference on Alkali-Aggregate Reaction in Concrete, Melbourne, Australia, 133-142).
Johnston, J. And Grove, C. (1931), the solubility of calcium hydroxide in saline solution. Journal of the American Chemical Society, 53(11) 3976–3991.
Jones, DA (1996), “Principles and Prevention of Corrosion,” Second Edition, Upper Saddle River, New Jersey, Prentice Hall.
Lau, K., Sagues, AA, Yao, L. and Powers, RG (2007), Corrosion performance of concrete cylindrical piles, Corrosion, 63 (4) 366-378.
Lee, JS, Ray, RI and Little, BJ (2010), The effect of experimental conditions on the results of laboratory studies using natural coastal seawater, Corrosion 66 (1) 105001-1-6.
Lewis, DA (1962), “Some Aspects of Corrosion of Steel in Concrete”, Proceedings of the First International Conference on Metal Corrosion, London, 1962, pp. 547-555.
Lewis, DA and Copenhagen, WJ (1959), Corrosion of steel reinforcement in concrete in the ocean atmosphere, Corrosion, 15 (7) 382t-388t.
Lukas, W. (1985), The relationship between chloride content and corrosion in concrete in tensile steel bars of Austrian bridges and concrete pavements, Betonwerk and Fertigteil-Technik, 51 (11) 730-734.
Marinoni, N., Pavese, A., Voltolini, M. and Merline, M. (2008), Long-term leaching test in concrete: X-ray powder diffraction study, Cement & Concrete Composites, 30, 700-705.
Mather, B. (1964), “The Influence of Seawater on Concrete”, Miscellaneous Paper 6-690, U.S. Army Engineering Waterway Experimental Station, Vicksburg, MI4.
Melchers, RE (1999), Structural Reliability Analysis and Prediction (Second Edition), John Wiley, Chichester.
Melchers RE (2004), Pitting corrosion of mild steel in seawater immersion environment-1: Maximum pit depth, corrosion, 60(9): 824-836.
Melchers, RE (2010), Carbonate, carbonation and durability of reinforced concrete marine structures, Australia J Structural Engineering 10 (3): 215-226.
Melchers, RE (2011), “Observations on Rebar Corrosion in the Marine Environment”, Proc.Natl.Acad.Sci.USA. The 12th DBMC, International Conf. Regarding the durability of building materials and components, Porto, Portugal, April 12-15, CDRom, paper 6.59.
Melchers, RE (2015), Field experience and long-term durability of reinforced concrete structures, Proc. The second international airport Conf. Research on Performance and Life Cycle Structural Engineering (PLSE 2015), Brisbane, December 9-12, University of Queensland, Brisbane.
Melchers, RE and Chaves, IA (2016), Long-term durability of reinforced concrete structures in marine environments, Proc. September 11-15, Montpellier Eurocor2016.
Melchers, RE & Chaves, IA (2017), Comparative Study of Chloride and Long-term Rebar Corrosion, “Materials and Corrosion”, in press.
Melchers, RE & Li, CQ (2006), Phenomenological modeling of steel corrosion loss in marine environment, ACI Journal of Materials, 103(1): 25-32.
Melchers, RE and Li, CQ (2009), Initiation and activation time of reinforcement corrosion in concrete structures exposed to harsh marine environments, Cement and Concrete Research 39, 1068-1076.
Melchers, RE, Li, CQ & Davison, MA (2009), “Observation and Analysis of the 63-year-old Reinforced Concrete Corridor Railing Exposed to the North Sea”, “Concrete Research” Journal 61 (4) 233-243.
Melchers, RE, Pape, TM, Chaves, IA, and Heywood, R. (2017), “Long-term Durability of Reinforced Concrete Piles for Honeybrook Highway Bridges”, “Australian Journal of Structural Engineering”, in press.
Mercer, AD and Lumbard, EA (1995), Corrosion of low carbon steel in water, British Corrosion J., 30(1) 43-55.
Mohammed, TU, Otsuki, N., Hisada, M. & Shibata, T. (2001) The effect of crack width and rebar type on steel corrosion in concrete, J. Civil Engineering Materials (ASCE), 13(3) 194-201.
Narver (Darver, DL) (1954), High-quality concrete made of coral and sea water, Civil Engineering, 24 (11) 49-52.
Nawy, EG (2008), Handbook of Concrete Construction Engineering, CRC Press, Boca Raton, pp. 30-57.
Ozaki S. & Suguata, N. (1998) Long-term durability of reinforced concrete submerged in the sea, (In) Concrete under severe conditions 2: Environment and load, Vol. 1. 1. Procedures for the second international airport Conf. On the issue of concrete under severe conditions, CONSEC’98, Tromso, Norway, June 21-24, (Editor) OE Gjorv, K. Sakai and N. Banthia, E and FN Spon, London, 448-457.
Parrott, LJ (1987), “Research on Carbonization of Reinforced Concrete in the UK”, Architectural Research Institute of the British Ministry of Environment.
Pourbaix, M. (1970), Importance of pitting corrosion and intergranular corrosion protection potential, Corrosion, 26(10) 431-438.
Poursaee, A. and Hansen, CM (2009), “Assessing Potential Pitfalls of Steel Corrosion Caused by Chloride in Concrete”, Cement and Concrete Research 39, 391-400.
Romanoff, M. (1957), “Underground Corrosion”, National Bureau of Standards Circular No. 579, US Government Printing Office (Publisher), Washington, DC.
Sagues, AA, Moreno, EI and Andrade, C. (1997), Changes in pH during in-situ leaching in small concrete cavities, Cement and Concrete Research, 27 (11) 1747-1759.
Shalon, R. & Raphael, M. (1959), The effect of seawater on corrosion of steel bars, J. ACI, 30 (12) 1251-1268.
Sharland SM & Tasker PW (1988) Mathematical model of crevice and pitting corrosion-I. Physical model, Corrosion Sci. 28(6), 603-620.
Southwell, CR and Alexander, AL (1970), Corrosion of metals in tropical water-ferrous metal structures, material protection 9 (1) 14-23.
Tuutti, K. (1982), Corrosion of steel in concrete, Swedish Cement and Concrete Research Institute, Stockholm, Research Report No. 4. (See also (1984) on the structural service life of embedded steel corrosion, (English) Performance of Concrete in Marine Environment, ACI SP-65, American Institute of Concrete, Detroit, 223-236.)
Verbeek, GJ (1975), Corrosion Mechanism of Steel in Concrete, Publication SP-49, American Concrete Association, Farringham, Mississippi, 21-38.
Volkswein, A. and Dorner, H. (1986), “From highway to highway”, “Forschung Strassenbau and Strassenverkehrstechnik”, “Heft 460″, “German Bundestag”, “German Bundestag”, Abteilung.
Wakeman, CM, Dockweiler, EV, Stover, HE & Whiteneck, LL (1958), “Use of Concrete in the Marine Environment”, Proc. ACI 54(4) 841-856.
Wells, T. and Melchers, RE (2015), “Using theory and field observations to model concrete degradation in sewers”, “Cement and Concrete Research”, 77, 82-96.
Weyers, RE (1997), Service Life Model of Concrete Structures, (In) Repair and Repair of Reinforced Concrete Structures: The Latest Technology, (Ed) WF Silva-Araya, OT De Rincon and LP O’Neill, ASCE, Ray, Virginia Ston, 105-119.
Wig, RJ and Ferguson, LR (1917), Reinforced concrete in seawater fails due to corroded steel, Engineering News Record, Volume 79, pp. 689-693.
Wood, CRJ (1948), Civil Engineers in the “Phoenix” (Yu) War, Symposium on Essays on Wartime Engineering Issues, Volume 2-Dock and Port, London: Institution of Civil Engineers, 336-368.
Yu, Linwen, Francois, R., Vu Hiep Dang & Gagne R. (2015), The development of chloride-induced corrosion in pre-cracked RC beams under continuous load: load-induced cracks, concrete cover and exposure conditions, cement and Concrete Research, 67, 246-258.
Robert E. Melchers is a professor of civil engineering at Newcastle University, Australia. He holds a bachelor’s and master’s degree from Monash University, and a doctorate from Cambridge University in the United Kingdom. He is a Fellow of the Australian Institute of Technology, Science and Engineering, and an Honorary Fellow of the Australian Institution of Engineers. His most recent awards are the 2009 ACA Corrosion Medal, the 2012 Jin S Chung Award (International Society of Marine and Polar Engineers) and the 2013 John Connell Gold Award (Australian Society of Engineers). He used to be the Institution of Engineers and the 2014 Outstanding Lecturer of the Australian Structural Institute.
ACA is a non-profit membership association that disseminates information about corrosion and its prevention or control by providing training, seminars, conferences, publications and other events.

Post time: Sep-30-2020