M. Castellote

Chloride threshold values to depassivate reinforcing bars embedded in a standardized OPC mortar

The chloride threshold to develop active corrosion of the reinforcing steel does not seem to be a unique value and it depends on several factors, such as concrete mix proportions, cement type, C3A content of cement, blended materials, water/cement ratio, temperature, relative humidity, steel surface conditions and source of chloride penetration among others. Numerous studies have been already devoted to the study of the chloride threshold value for depassivation of the steel embedded in concrete. One of the reasons found for the scatter is the large number of variables that influence the chloride amount for depassivation. The other reason is the lack of accordance for the definition of the chloride threshold itself, either on the determining parameters (visual observation, corrosion potential or corrosion current) or on the expression of the threshold (as Cl−/OH− ratio or by weight of cement or concrete). The present paper presents chloride thresholds that were studied in mortar based on corrosion current measurements and expressed as total, free and Cl−/OH− ratio. For this study, mainly smoothed bars were used, but also some ribbed bars were tested. Chloride thresholds in the range of 1.24–3.08% and 0.39–1.16%, by weight of cement, for total and free chlorides, respectively, and in the range of 1.17–3.98 for Cl−/OH− ratio were found for chlorides admixed in the mixing water. Active corrosion is considered when, in a small exposed area, the corrosion rate of the rebar is higher than 0.1 μA/cm2. The threshold in the case of Cl−/OH− results a bit higher than that found in a previous work for synthetic pore solution, although the two types of data can be fitted together finding a good correlation.

Chloride threshold dependence of pitting potential of reinforcements

Abstract Chloride threshold is one of the key parameters needed for service life prediction of concrete structures. The main difficulty that the determination of this parameter encounters is that the chloride threshold inducing active corrosion is not a unique value and depends on several variables. Among them the electrical potential exhibited by the rebar surface is the one less experimentally studied, even though it was identified some decades ago as the most comprehensive controlling parameter for the disruption of passivity. In the present paper the influence of the potential of reinforcements embedded in mortar on the chloride threshold has been studied for the first time in a region from +250 to −650 mV with respect to SCE. Identical samples have been tested in order to perform a statistical treatment and so mean values are given. The chloride thresholds found are expressed as total and free chloride as well as chloride/hydroxide (Cl − /OH − ) ratio. The results show that chloride threshold is independent of the potential for values more anodic than −200±50 mV (SCE) being this region the typical potentials exhibited by passive reinforcements in atmospheric conditions. For potentials more cathodic than −200±50 mV (SCE), the threshold progressively increases and a linear fitting is possible. The dependence of chloride threshold with the potential approaches to the understanding why in previous works researchers have found significant differences.

Measurement of the steady and non-steady-state chloride diffusion coefficients in a migration test by means of monitoring the conductivity in the anolyte chamber. Comparison with natural diffusion tests

Abstract The precise calculation of steady-state chloride diffusion coefficients is laborious, expensive, and time consuming due to the fact that it is necessary to take liquid samples for the chlorides to be analysed. In this paper, the concentration of chlorides is monitored by measuring the conductivity of the anolyte which initially is distilled water. Additionally, the calculation of Ds and Dns is made from the same test using the concept of “time-lag” and of the “equivalent time” between diffusion and migration experiments. The correlation between Ds and the conductivity has been done by testing 10 different cementitious matrixes (concrete and mortar) with different binders, and considering the influence of the type of solution that is the source of chloride ions as well as the variation of the electrodes and the effect of the kind of matrix. The results obtained have been compared with natural diffusion results in order to validate them. As a result, for the possibility of making it automatic, the calculation of Ds and Dns diffusion coefficients has been verified.

COMPOSITION AND MICROSTRUCTURAL CHANGES OF CEMENT PASTES UPON HEATING, AS STUDIED BY NEUTRON DIFFRACTION

Composition and microstructure changes of cement pastes when heating until 620 °C and cooling afterwards have been monitored on site by neutron diffraction. The parameters involved in the study have been the heating ramp, the state of the sample (in block or ground) and the type of cement. The residual state of the samples has also been studied by mercury intrusion porosimetry (MIP) and thermogravimetry analysis (TGA). As a result, it has been possible to monitor the major features of the experiments, i.e., the phases existence domains and their growing and decaying for portlandite, ettringite, calcite, lime, larnite, and hydrated calcium silicate (CSH gel), as well as the loss of free and bounded water during the experiments. In addition, the different residual phases in the matrix depending on the experimental set up have been established, which might have a significant influence in the durability of a concrete after a fire.

Chloride-binding isotherms in concrete submitted to non-steady-state migration experiments

Abstract Chloride binding is an essential step in the determination of the chloride transport across concrete. The binding ability has been previously modelled in different ways, among which is that of assuming an adsorption process and fitting to the mathematical expression of an isotherm from data obtained from natural diffusion tests. In the present paper, chloride concentration profiles from non-steady-state migration tests have been fitted to a Brunauer, Emmett, Teller (BET) isotherm general equation to model chloride binding under these conditions and to compare it with natural diffusion results. The results indicate that chloride ions remains unbound until the concentration in the pore solution exceeds about 28 g/L and that interactions in the chloride/cement matrix could be similar to those of natural diffusion when the chloride external concentration is higher than 1 M in Cl − . In addition, it has been found that the “chloride isotherm” is of type V with a BET characteristic constant C

Accelerated simultaneous determination of the chloride depassivation threshold and of the non-stationary diffusion coefficient values

Numerous studies have been already carried out to study the chloride threshold value for depassivation of the steel embedded in concrete. Most of them were based on natural penetration of chlorides, whose main disadvantage is that it is very time consuming and makes necessary a long time to obtain reliable data. Other methods used were potentiostatic tests that, depending on the potential applied, may also be very long. In present paper, a new accelerated method to determine the threshold value of corrosion of steel is tried on a mortar specimen with an embedded rebar. The procedure is based on the use of the mechanism of migration to force the penetration of chlorides into the specimen, but switching off the external voltage and measuring the resulting threshold value in quasi-natural conditions. In spite of some considerations regarding the fact of the acceleration, the thresholds found (total, free Cl− and Cl−/OH− ratio) are in good enough accordance with those found for a twin specimen in natural conditions and within the range reported for mortars in the literature. In addition, this method has the advantage of enabling the simultaneous determination of the non-stationary chloride diffusion coefficient, Dns from the migration data, which has also been validated in natural conditions.

Electrochemical removal of chlorides: Modelling of the extraction, resulting profiles and determination of the efficient time of treatment

Due to the increase in the hydroxyl ion concentration generated as a consequence of the cathodic reaction of water electrolysis, the diffusion coefficient of chloride ions decreases during chloride extraction experiments. In a previous work carried out by the authors, an empirical equation to model this decrease in the diffusion coefficient as the treatment proceeds was given. Based on this equation, mathematical modelling of the extraction is carried out in this paper. This modelling allows the calculation of the time needed to achieve a particular percentage of chloride extraction as a function of the electrical charge passed, as well as the remaining chloride profile in the concrete. Numerical examples applied to the case in which chlorides were homogeneously distributed in the specimens are given to illustrate the modelling. These are also applied to experimental results previously published by other authors, finding a good accordance.

Oxygen and chloride diffusion in cement pastes as a validation of chloride diffusion coefficients obtained by steady-state migration tests

Abstract When chloride ions diffuse through concrete, it has been found that they interact with the surface charge and electrical double layer developed at the cementitious matrix/pore solution interface. As a consequence of this interaction, the diffusion of chloride ions is retarded in comparison with that of dissolved oxygen molecules, although the two species have very similar diffusion coefficients in infinitely aqueous dilute solutions. This is taken to imply that electrostatic, rather than purely steric, factors limit the ability of chloride ions to diffuse through pores below a critical radius. If an electrical field is applied to accelerate chloride transport through concrete, it is uncertain what effect this may have on the size range of pores that are accessible to the migrating ions. At high applied voltages, it might even be that the normal electrostatic interactions that influence ionic diffusion are nullified so that chloride ions can traverse pores of similar radii to those that are accessible to dissolved oxygen. This would be expected to give rise to an apparent increase in the value of ‘diffusion coefficient’ obtained for chloride. To investigate whether, during a steady-state migration test, chloride ions diffuse through the essentially the same range of pores as during a natural steady-state diffusion test, experimental studies (oxygen diffusion, natural diffusion and migration tests) have been carried out with a limited range of cement pastes. The results indicate that, at the recommended applied voltage (12 V) and for cementitious materials of the type studied, steady-state migration tests can be used to determine ‘effective diffusion coefficients’ that are of similar magnitude to those obtainable from natural steady-state diffusion measurements.

Relation between colourimetric chloride penetration depth and charge passed in migration tests of the type of standard ASTM C1202-91

Abstract The great interest in the knowledge of the amount of chlorides that have penetrated into concrete is due to the fact that corrosion of reinforcement in concrete initiates when chloride content in contact with rebars exceeds a threshold value. So, a number of methods to measure the diffusion coefficient of Cl− in concrete have been developed; one of them, from the use of a colourimetric method, derived from its ease in practice. Some questions have been raised about this method, concerning the amount of chlorides at which the colour change takes place. In this article, comparison between total chloride profiles in migration experiments of the AASHTO type (ASTM C1202-91), and results from colourimetric test, as well as a relationship between colourimetric penetration depth and charge passed, are given.

Alkaline leaching method for the determination of the chloride content in the aqueous phase of hardened cementitious materials

Abstract The need of determining “free” chloride content in the aqueous phase of hardened cementitious materials arises from the fact that chloride is the responsible agent of the depassivation of the rebars in reinforced concrete. This paper presents a method of obtaining free chlorides by leaching in an alkaline solution during 24 h. The method is calibrated by comparison with the values obtained by means of the pore-pressing technique (PPT). The agreement found is reasonably good in spite of the risk of excessive leaching that aims to solve part of bound chlorides. The use of decarbonated water as extraction agent has resulted appropriately, perhaps due to the fact that the samples are not powdered, instead, pieces of around 2 mm in size are used.