Steel--Corrosion

The penetration of chloride ions through concrete can compromise the integrity of
a structure. The chloride concentration, [Cl-], at which the corrosion process initiates is
termed the critical chloride concentration or chloride threshold, [Cl-]th. One of the
purposes of this research was to determine the [Cl-]th for various reinforcing alloys.
Furthermore, the time-to-corrosion (TIC) was measured to determine the time at which
bars become active. Both parameters, [Cl-]th and TTC, were found to be distributed;
therefore, statistical analyses were performed to forecast the probability of activity. A
new experimental procedure was introduced to increase the TTC data set by electrically
isolating the top bars as they became active.
The research also compared the [Cl-] for core samples with those values obtained
from along the top rebar trace. In general, this analysis demonstrated that core sample
[Cr] was lower than at the rebar-trace.

Qualification criteria for cathodic protection of pre-tensioned tendon wires in concrete were studied with regard to the risks of embrittlement due to chromium microalloying and existing corrosion damage. The selected materials included two microalloyed (with 0.23 and 0.24 w% chromium) and one non-microalloyed, high carbon prestressing steels. The slow strain rate testing technique was used to evaluate the effects of polarization to -0.90 and -1.30 VSCE upon strength and ductility of the steels. Fractographic analysis was performed using scanning electron microscopy. Based upon statistical analysis, an attempt was made to relate the remaining strength of the corroded wire to the extent of corrosion damage for different corrosion morphologies.

Experiments were conducted to evaluate occurrence of any deterioration of prestressing steel tendon to concrete bond as a consequence of cathodic polarization. Pretensionned concrete specimens were cathodically polarized with current densities ranging from 50 to 5000 m^2 of steel while exposed to a constant flow of natural sea water. The concrete and steel dimensional changes were monitored by strain gages mounted on the tendons and embedded in the concrete. Contractions of the steel of 25 to 50 percent of the initial tensioning were recorded after 17 to 36 MC/m^2 were transferred to the tendons on specimens polarized at the highest currents. This corresponds to 54 to 114 years of polarization at 10mA/m^2 if bond loss was solely dependent on the charge transfer. It was noticed that the smaller the current, the more the charge that was transferred before steel contraction began. These results imply that cathodic polarization should impose no threat to the prestressing steel-to-concrete bond on typical structures over their expected lifetime.

Dimensional changes of thin cement paste specimens were monitored during exposure to distilled water, hydrochloric acid, alkaline and FeCl$\sb2$ solutions; and comparison was made with the experimental results obtained by other authors. Post experimental examination of reaction products formed on the specimen surface exposed to FeCl$\sb2$ was conducted using X-Ray diffraction. The data resulting from the measured dimensional changes were input to a finite element model to project the stresses that should occur in association with localized corrosion of embedded steel and wetting of adjacent cement by aqueous FeCl$\sb2$. Implications of the findings with regard to cement and concrete cracking in association with embedded metal corrosion are discussed.

Experiments were developed and performed to determine if glass electrodes can be employed to accurately measure pH of porewater in concrete. An attempt was made to determine the reason for the general underestimation of the alkalinity of porewater by one or more pH units when glass electrodes are used as sensors compared to the OH- concentration obtained from titration of pore solutions expressed from hydrated cement samples. The linear response and alkaline error of glass electrodes was investigated. pH was measured for mortar samples cast in air as well as under CO2-free conditions. A flat-surface combination electrode was mostly employed for this purpose, but studies also involved the use of a microelectrode. The possible role played by Ca(OH)2 liberated during cement hydration in decarbonation of porewater is discussed based on experiments.

Deterioration of reinforcing steel in concrete structures due to rebar corrosion is of national and international concern. Cathodic protection has evolved as the appropriate mitigating technique for existing structures. Another method which is being investigated is the electrochemical removal of chlorides from concrete. The current density used for this process leads to steel potentials at which hydrogen evolves. This raises concern regarding hydrogen embrittlement of reinforcing steel in concrete. For the purpose of analyzing this embrittlement tendency, a series of constant extension rate experiments was conducted in saturated calcium hydroxide solutions using notched and smooth specimens. While notched specimens did not show any effect of hydrogen embrittlement, the smooth specimens have revealed a reduction in ductility at high current densities. Experiments have shown that this loss in ductility is recoverable. Tests were also conducted on mortared smooth specimens which essentially exhibited the same trend observed as in the calcium hydroxide solution. The loss in ductility has been attributed to the enhanced void development in the presence of hydrogen during plastic deformation.

The objective of this study was to determine the influences of chlorides, pH and surface films on the corrosion of reinforcing steel in alkaline solutions. Anodic and linear polarization experiments were conducted on 1018 steel specimens exposed to these environments. The results indicate that there is a threshold amount of [Cl-] beyond which passivity is compromised. The i-passive in the electrolyte of pH 12.24 was approximately 0.01 of that in the electrolyte of pH 10.13. At least a one day period was necessary for the effect of the excess crystals of Ca(OH)2 to be felt. Corrosion rates of specimen with a Ca(OH)2 surface film were lower than without this film. [Cl-] /[OH-] in the range 2.5 to 3.5 appeared to be the threshold for breakdown of passivity.

The objective of this study was to determine whether the presence of undissolved calcium hydroxide at the steel interface helped maintain or delay the breakdown of passivity under adverse conditions, such as the presence of chlorides and carbonation. Saturated calcium hydroxide solution was used as an electrolyte in the test cell, and steel specimens were exposed to a range of chloride ion concentration and carbonation. The results indicate that undissolved calcium hydroxide has an important role in the passivation of reinforcing steel. Also, the critical chloride to hydroxide ratio may be more relevant in characterizing the breakdown of passivity than the threshold value of total or soluble chloride present in the electrolyte. It was noted that in the presence of undissolved calcium hydroxide, steel remained passive for as high a chloride ion concentration as 0.54% (by weight of electrolyte).

A series of experiments were performed on a single
tendon in a prestressed concrete slab for the purpose of
studying the susceptibility of the high strength steel to
brittle fracture as a result of cathodic protection. The
procedure first involved impressing an anodic current until
active potentials were recorded along the tendon length.
Subsequently, cathodic polarization of the midsection of
the tendon to -1.3 V(SCE) was effected. After a prescribed
time, the tendon was removed and sectioned into 15 ern
lengths. Each segment strand was then notched and strained
to failure, in three-point bending, in order to determine
any effect of cathodic protection on the mechanical properties
of the material. Differences in average failure loads
were compared to determine if a loss of load-bearing
capacity and brittle fracture occurred due to hydrogen
absorption.

A model for calcareous deposit induced fatigue crack
closure has been previously reported based upon the
criterion that the deposit thickness within the crack
equals or exceeds one-half the minimum crack opening
displacement. In the present paper an expanded and refined
model is proposed by considering (1) compaction of
calcareous deposits during the closure period of the stress
cycle and (2) the relationship between R ratio and
threshold stress intensity range. Compression tests upon
calcareous films grown on steel surfaces have been
performed, and these reveal the change in deposit thickness
as a function of stress. The implications of these models
to calcareous deposit induced fatigue crack closure are
discussed in terms of mechanical and electrochemical
parameters.