Modelling of reinforcement corrosion – Influence of concrete technology on corrosion development

Reinforcement corrosion is influenced by different parameters like resistivity of concrete, setting conditions and also by concrete technology. Moreover the presence of cathodic areas and the possibility of unhampered cathodic reaction influences the reinforcement corrosion. In this paper the development of corrosion without large cathodic areas, called self‐corrosion, considering different concrete parameters, is studied.     

DFG‐research‐group: Modelling reinforcement corrosion – overview of the project

A strong interest for the durability of reinforced concrete structures currently exists in industry and research [1]. Against the background of immense costs for maintaining reinforced concrete structures and repairing damage caused by corroding reinforcement steel, this interest lead to a German joint research project. The aim of this network‐based (www.bam.de/dfg537.htm) research group is, to deliver the basic knowledge of the corrosion propagation and to make a probabilistic tool available for engineers so that a complete design for durability, concerning reinforcement corrosion, will be possible.     

Modelling of reinforcement corrosion – Corrosion with extensive cathodes

Reinforcement corrosion is the most common reason for the premature deterioration of a concrete structure. In case of a partial depassivation of the reinforcement macrocells are formed with considerable metal removal rates. Thus an assessment of the macrocell current becomes of great concern. To find out, whether this can be achieved by numerical calculations, specimens have been fabricated and simulated by the use of a boundary element program. In this paper the results of the calculations are presented and compared to electrochemical measurements on the real specimens.     

Modelling reinforcement corrosion – usability of a factorial approach for modelling resistivity of concrete

As part of the macro cell corrosion process of reinforcement steel, the resistivity of concrete plays a crucial role. In order to investigate the influence parameters on the resistivity of concrete, the results of a discrete quantification were presented and implemented into a factorial approach for modelling corrosion propagation. First results delivered significant deviations from values obtained by measuring. Using Gauss' method of least mean squares provided a decrease in deviations. The hereby obtained deviations were lower than the scatter of the measuring results. A usability of the proposed factorial approach could therefore be approved.     

Modelling of reinforcement corrosion – Investigations on the influence of shrinkage and creep on the development of concrete cracking in the early propagation stage of reinforcement corrosion

Since the initiation stage of the damage process due to reinforcement corrosion had been successfully investigated in the recent past, the damage progress in the propagation stage is currently in the focus of research. This work deals with the mechanisms of cracking and spalling due to corrosion of reinforcement and aims for the development of an analytic prediction model of the damage process. On this occasion the influence of shrinkage and creep on the stress condition within the concrete cover is of major importance to the subsequent analysis of the crack formation due to reinforcement corrosion in the early propagation stage.     

Models for the propagation phase of reinforcement corrosion – an overview

The deterioration of a concrete structure by reinforcement corrosion proceeds in two phases: the initiation stage and the propagation stage. The first stage describes the time to onset of corrosion due to carbonation of the concrete or chloride ingress. The second stage is the actual deterioration stage. Most methods for life time assessments refer only to the first stage, what is on the safe side with respect to design of structures, but also a model for the second stage can be of interest, e.g. if the remaining life time of an existing structure has to be estimated. This paper presents and discusses the state‐of‐the‐art of models for the propagation stage with regard to their different approaches.     

Damage process due to corrosion of Reinforcement bars – Current and future activities –

Against the background of huge costs for maintenance and repair it would be helpful to have a tool to assess the remaining life time of concrete structures. Deterioration is often caused by reinforcement corrosion and research projects have been carried out to develop models for the time‐dependent progress of the degradation. Although these projects have resulted in several steps forward, further work is still needed. This paper presents two actual research activities which deal with modeling of reinforcement corrosion: the first one is the RILEM Technical Committee MAI, the second is a German joint research project.     

Inspection strategies for reinforcement corrosion surveys

This paper describes the procedure of the assessment of the condition of concrete structures regarding reinforcement corrosion and selection criteria for the methods to be used for inspection. It is based on the European Standard EN 1504 on Repair of Concrete Structures and the experience of the members of the task group.     

Theoretical considerations on the supposed linear relationship between concrete resistivity and corrosion rate of steel reinforcement

Traditionally, the assessment of service life of steel reinforced concrete structures has been focused on the prediction of the time required to achieve a transition from passive to active corrosion rather than to accurately estimate the subsequent corrosion rates. However, the propagation period, i.e. the time during which the reinforcing steel is actively corroding, may add significantly to the service life. Consequently, ignoring the propagation period may prove to be a conservative approach. On the other hand the prediction of the corrosion rate may result in a very complex task in view of the electrochemical nature of corrosion and the numerous parameters involved. In order to account for the various influences an essentially empirical model has been introduced in which the electrolytic resistivity of the concrete environment serves as the major parameter. This model will be discussed for carbonation‐induced corrosion based on the commonly accepted theory of aqueous corrosion. An alternative model for microcell corrosion is proposed which is based on the commonly accepted view that anodic and cathodic sites are microscopic and their locations change randomly with time. In line with this view electrolytic resistivity can be incorporated and thus may play a significant role in the kinetics of the corrosion process. For a wide range of corrosion current densities the relationship between corrosion current density, log(i_corr), and concrete resistance, log(R_con), can then be approximated by an almost ideal linear relationship. Assuming a fixed geometrical arrangement of anodic and cathodic sites on the steel surface, this linear relationship is also valid for concrete resistivity, ρ_con. However, from the theoretical treatment of the electrochemical processes underlying reinforcement corrosion it becomes evident that a linear relationship between corrosion current density and concrete resistivity does not necessarily imply that concrete resistance is dominating the overall corrosion cell resistance. In most cases a significant portion of the driving voltage of the corrosion cell will be consumed by the transfer of electrical charge involved in cathodic reactions, i.e. cathodic activation control will dominate.     

Concrete cover cracking owing to reinforcement corrosion – theoretical considerations and practical experience

Reinforcement corrosion might lead to cracking and spalling of the concrete cover owing to the volume expansion associated with the deposition of some of the possible corrosion products. This is not only aesthetically unpleasing, it might also accelerate deterioration processes or become a safety issue for passing traffic. The present paper discusses first the mechanisms of carbonation‐ and chloride‐induced reinforcement corrosion and considers the chemistry of aqueous iron in order to identify the type of corrosion products as well as their location of formation. Furthermore, practical examples are summarised in order to compare the documented behaviour of a number of real structures with the theoretical considerations made. It is shown that for the case of purely chloride‐induced (pitting) corrosion, precipitation of corrosion products is strongly delayed or may even not occur. Implications are discussed with respect to time‐to‐corrosion prediction models and visual inspection of reinforced concrete structures. Both the theoretical considerations and the practical experience illustrate that relying on outwardly visible signs to detect internally on‐going corrosion must be done with caution if localised reinforcement corrosion cannot be excluded.     

Numerical models for the propagation period of reinforcement corrosion ‐ Comparison of a case study calculated by different researchers

This paper describes the results of the calculations on Case Study 1 which have been performed within the activities of the EFC Task Group on modelling of corrosion of steel in concrete. These calculations have been performed by six researchers from different institutes, each of them using his favourite tools and assumptions. Case Study 1 was intended to obtain a preliminary impression of the corrosion models used by researchers as well as to gain information on the underlying scientific backgrounds. It has to be noted that these calculations already have been carried out in 2000 and 2001 and some of the models might have been more or less improved in the meantime. Also only selected details are described in this paper, because the discussion on the details within the EFC task group has shown, that a full scientific discussion of the models including all relevant assumptions is not possible within the framework of this paper. However, it also should be noted that it has never been intended to find the “best” model or to rank the models, but to get an overview and learn from each other. Finally, the participants of the EFC task group came to the conclusion, that all types of models have their advantages, depending on the complexity of the corrosion system to be simulated.     

FEM‐models for the propagation period of chloride induced reinforcement corrosion

The paper reports the results of numerical simulations carried out with FEM and aimed at evaluating the corrosion conditions of steel bars in concrete elements subjected to chlorides. Two case studies were analysed: a reinforced concrete element subjected to de‐icing salt in the presence of a crack and a concrete tunnel in a chloride‐contaminated, water saturated soil. Attention was focused on the selection of proper values of concrete resistivity and of the parameters suitable to describe the electrochemical behaviour of steel in the different conditions of exposure. The results allowed to quantify the effects of the galvanic coupling between active and passive areas on the corrosion rate of steel.     

Measuring the corrosion rate of steel in concrete – effect of measurement technique,polarisation time and current

Both on‐site investigations and laboratory studies have shown that different corrosion rates are obtained when different commercially available corrosion rate instruments are used. The different electrochemical techniques and the measurement parameters used, i.e. polarisation current and time, are in some studies considered the main reasons for the variations. This paper presents an experimental study on the quantitative effect of polarisation time and current on the measured polarisation resistance – and thus the corrosion current density – of passively and actively corroding steel. Two electrochemical techniques often used in instruments for on‐site corrosion rate measurements are investigated. On passively corroding reinforcement the measured polarisation resistance was for both techniques found to be highly affected by the polarisation time and current and no plateaus at either short or long polarisation times, or low or high polarisation currents were identified. On actively corroding reinforcement a large effect of the polarisation time was also found, but only a minor effect of the polarisation current. The effect of the polarisation time was, however, practically independent of the corrosion rate for actively corroding steel. For both techniques guidelines for polarisation times and currents are given for (on‐site) non‐destructive corrosion rate measurements on reinforcement steel in concrete.     

Problems encountered in the detection of reinforcement corrosion in concrete tunnel linings – theoretical considerations

For concrete structures in which oxygen availability is limited, the potentials recorded for passive steel reinforcement may be considerably more negative than ?350 mV versus CSE. This situation may occur for concrete linings of bored tunnels located in water saturated soil. In view of the great practical importance of correct interpretation of results from potential mapping for durability assessment, the influence of oxygen availability on the steady state potential of passive steel has been investigated on a theoretical basis. For values of the ratio between thickness, d, to the oxygen diffusion coefficient D_Ox of the concrete cover approaching a critical value of 3.36 · 10⁶/(A_st · i_pas), the potential will demonstrate a dramatic drop amounting to several hundreds of millivolts. Consequently, in this critical region of d/D_Ox, the corrosion potential of steel in its passive state may vary over a wide range, and commonly used potential criteria to identify the actual electrochemical condition are not applicable. A numerical analysis using a simplified approach has been performed on the typical reinforcement arrangement of a tunnel lining. The analysis clearly demonstrates that the presence of 2 layers of passive steel located at different depths will result in a pronounced effect on the potential recorded at the inner concrete surface, thus hindering correct interpretation.     

Stainless steel reinforcing bars – reason for their high pitting corrosion resistance

In harsh chloride bearing environments stainless steel reinforcing bars offer excellent corrosion resistance and very long service life for concrete structures, but the high costs limit a more widespread use. Manganese bearing nickel‐free stainless steels could be a cost‐effective alternative. Whereas the corrosion behavior of stainless steels in alkaline solutions, mortar and concrete is quite well established, only little information on the reasons for the high pitting resistance are available. This work reports the results of pitting potential measurements in solutions simulating alkaline and carbonated concrete on black steel, stainless steel DIN 1.4301, duplex steel DIN 1.4462, and nickel‐free stainless steel DIN 1.4456. Duplex and nickel‐free stainless steels are fully resistant even in 4 M NaCl solutions with pH 13 or higher, the lower grade DIN 1.4301 shows a wide scatter between fully resistant and pitting potentials as low as +0.2 V SCE. In carbonated solutions with pH 9 the nickel‐free DIN 1.4456 shows pitting corrosion at chloride concentrations ≥3 M. This ranking of the pitting resistance can be rationalized based on XPS surface analysis results: both the increase of the Cr(III)oxy‐hydroxide and Mo(VI) contents in the passive film and a marked nickel enrichment beneath the film improve the pitting resistance. The duplex DIN 1.4462 shows the highest pitting resistance, which can be attributed to the very high Cr(III)oxy‐hydroxide, to a medium Mo(VI) content in the film and to a nickel enrichment beneath the film. Upon time, the protective properties of the surface film improve. This beneficial effect of ageing (transformation of the passive film to a less Fe²⁺ containing, more hydrated film) will lead to higher pitting potentials. It can be concluded that short‐term solution experiments give conservative results in terms of resistance to chloride‐induced corrosion in reinforced concrete structures.     

Reinforced concrete structures – shall concrete remain the dominating means of corrosion prevention?

The acknowledged serious deterioration of reinforced concrete structures due to chloride induced corrosion has been the main fuel for research and development of very dense and impermeable concrete, so‐called high performance concrete (HPC). This development has dominated concrete research up through the 80'ies and 90'ies. The results have technically been successful. However, the practical use of such concretes on site have often posed serious difficulties, resulting in at times very low performance concrete structures although HPC was specified. The discrepancy between concrete quality reached in the laboratory, what is being specified in the design and what can realistically be achieved on site is seldom in balance. Alternative means of more or less reliable means of corrosion prevention, often based on organic materials, have during recent years been developed to protect our inorganic concrete and reinforcement. However, a highly reliable means of corrosion prevention has been the introduction of stainless steel reinforcement, which is available with dimensions and strengths directly interchangeable with ordinary carbon steel reinforcement. It has been proven that stainless steel and carbon steel can be in metallic contact when cast into concrete, without causing galvanic corrosion. This seems, for the present, to be like an unexpectedly simple and highly reliable solution to the corrosion problems. As exemplified, this technology is rapidly gaining momentum in highly corrosive environments – and concretes being much more robust to execution can now take over from HPC.     

Early stage beneficial effects of cathodic protection in concrete structures

Over the last 25 years, cathodic protection (CP) of reinforced concrete structures suffering from chloride induced reinforcement corrosion has shown to be successful and durable. CP current causes steel polarisation, electrochemical reactions and ion transport in the concrete. CP systems are designed based on experience, which results in conservative designs and their performance is a matter of wait‐and‐see. CP systems can be designed for critical aspects and made more economical using numerical models for current and polarisation distribution. Previously, principles of numerical calculations for design of CP systems were reported. The results were satisfactory, except in terms of current density for active corroding systems. This was suggested to be due to neglecting beneficial effects of CP current flow. One of the beneficial effects is pH increase at the steel surface due to oxygen reduction. As the pH increases, the corrosion rate decreases and the current demand decreases. A simple model was set up for this transient process, suggesting that such effects should occur on the time scale of hours to days. This model was validated from start up data of a CP field trial system on part of a bridge. Field results confirmed the modelling proposed here.