Steel--Corrosion

In the present research butt welded ABS DH-32 steel
specimens were fatigue tested under freely corroding and
cathodic protection conditions by a time-series simulated
wide-band spectra of a North Sea wave climate. Previous
research pertaining to the probabilistic features of wave
dynamics, wave force models, Markov-Chain sea state
evolution and time-series simulation of wide-band spectra
are addressed as a background review. The resulting freely
corroding data are discussed on the basis of the SN plot
which developed from the previous investigations at FAU.
Calculation of the damage of failed specimens by employing
the "Equivalent Narrow-Band Approach" resulted in excellent
agreement with the linear damage summation assumption.

Initiation of corrosion of reinforcing steel in concrete is
often caused by chlorides. Using a pressurized method for
the purpose of accelerating penetration of sea water into
concrete, the threshold chloride ion concentration for
corrosion of reinforcing steel in Type I portland cement
concrete has been studied in detail. The variables that
have been investigated include water-cement ratio and steel
surface preparation. When corrosion was detected by
electrochemical potential measurement, the test was
terminated; and chloride ion concentration was evaluated. No
correlation between threshold concentration and water-cement
ratio was found. With regard to surface condition, the pre-rusted
steel specimen showed a tendency to corrode at a
lesser chloride ion concentration than for the other preparation
techniques, which were sand blasting and
pretreatment in a saturated calcium hydroxide solution.
The results are presented and discussed within a perspective
of established concrete and corrosion technologies.

It is recognized that reinforced concrete structures are
often susceptible to stray current corrosion damage. Experiments
have been performed upon a reinforced concrete bridge
model in sea water, and these indicate that 0.01 percent of
stray direct current in this electrolyte may enter the structure.
Alternating current may also contribute to corrosion
of reinforcing steel, though experimental results reveal detection
to be more difficult than for direct current. Two
other sources of current in reinforcing steel are from cathodic
protection systems of utility attachments and from galvanic
coupling of the covered steel and bare steel. Mechanisms
by which the above types of current contribute to deterioration
of reinforced concrete structures are discussed.

A summary of some pertinent previous research by
other authors is presented separately from the actual
laboratory report. The report presents the results of air
and corrosion fatigue tests which are. interpreted in terms
of time to failure, crack growth history, and fracture
mechanics techniques. In addition, the results of cathodic
protection of corrosion fatigue tests are interpreted in
terms of simple-design criteria, as well as relative to
Critical Corrosion Rate Theory. One of the more significant
results presented is the greater longevity of corrosion
fatigue, as compared to air fatigue, notched specimens
stressed within 10-15 KSI above the air endurance limit.
Also, the potential required to cathodically protect such
notched specimens is shown to be 50 mV lower than for
comparable smooth specimens. Both phenomena are attributed
in part to the lowering of crack pH to acidic levels in
notched specimens.

Experiments were conducted to investigate frequency dependence of sea water corrosion fatigue of notched 1018 steel under constant deflection, reverse bend test conditions. Frequencies of 1850, 1200, and 600 cycles per minute (CPM) were studied. Results indicate that frequency effects are not observed at high stresses, apparently due to the mechanical damage component of the fatigue process dominating over corrosion considerations. At low stresses reduced frequency resulted in decreased cycles to failure. Acidification of the crack tip solution is proposed as possibly responsible for this behavior. For specimens subjected to equal stress levels in the frequency dependent region, greater cathodic protection is required to mitigate corrosion fatigue for 600 CPM specimens (-0.86 volts standard calomel electrode) (V SCE) than for 1850 CPM specimens (-079V SCE).

This study employed half-cell potential (Eoc), polarization resistance (PR) and galvanic current tests to explore effects of surface condition and steel reinforcement type on corrosion initiation in mortar. A low impurity steel plate and #10 grade 60 rebar were used. Four surface conditions (as-received, atmospherically rusted, #240 SiC paper polished, and #500 SiC paper polished) were employed. Cyclic wet and dry exposure of specimens in 15% NaCl solution was conducted. Half-cell potential measurements showed that the corrosion probability of specimens with as-received and pre-rusted surfaces was lower than that of specimens with a polished surface. However, the PR and galvanic current tests indicated higher corrosion rate for specimens prepared with atmospherically rusted and as-received steels. Good agreement was found between results of the polarization resistance test and galvanic current measurements. There is a lack of correlation between half-cell potential and corrosion rate.

A study was performed of variables that affect the chloride threshold for corrosion of reinforcement, including pore water alkalinity, water-to-cement ratio, the use of Florida aggregate, and the addition of fly ash. To accomplish this, specimens were subjected to a periodic wet-dry ponding cycle using 15% by weight NaCl solution. Half cell potential and macrocell current measurements were taken to determine the time to corrosion. Upon active corrosion, the specimens were removed from the ponding cycle and dissected. Powder samples were collected from the concrete at the steel depth to ascertain the total chloride concentration. The determination of the pore water pH was attempted using a leaching method. To date only a portion of the specimens have displayed corrosion. The role of cement alkalinity and water-to-cement ratio in affecting corrosion resistance is reported and the results are discussed within the context of designing concrete structures for corrosion resistance.

Field and laboratory ambient and simulated deep seawater sacrificial anode cathodic protection experiments were performed by coupling steel specimens to Al-Zn-Hg anode through an appropriately sized external resistor and thereby permitting a particular level of cathodic protection from freely corroding to overprotection to be simulated. The effects of sea current velocity, surface preparation, initial current density, temperature and hydrostatic pressure upon cathodic polarization and characteristics of calcareous deposits were investigated in the context of slope parameter and steady-state potential and current density trend. The results revealed that a linear relationship between cathode potential and current density is applicable for design of sacrificial anode cathodic protection systems and analysis of cathodic protection survey data from existing structures both in shallow and deep waters. It was also found that for cathodically polarized steel specimens in ambient (shallow) seawater steady-state cathode potential and current density varied according to a sigmoidal trend that indicates the importance of calcareous deposits in such exposures and demonstrated the utility of rapid polarization. On the other hand, no sigmoidal trend was apparent for field and simulated deep water tests; but instead steady-state current density was constant for potential range between -0.80 and -1.05 v (Ag/AgCl). This disclosed that no particular benefit could be derived from employing rapid polarization in cold water exposures. SEM, EDX and X-ray diffraction analysis revealed that the calcareous deposits formed upon specimens exposed at 5C and ambient pressure or 8.96 MPa laboratory experiments exhibited two layer structures--an outer layer of CaCO3 as aragonite and inner layer of Mg(OH)2 as brucite. The morphology and coverage of the deposits depended on the design slope parameter (initial current density). Field testing results indicated that deposits were also composed of CaCO3 and Mg(OH)2 where the former was calcite. Because of the poor coverage of the deposits formed in the deep water condition, limited current density reduction was noted. These results suggest that a different cp design approach and strategy should be considered for deep, cold seawater compared to that commonly used in shallow water environments.

A study of the corrosion protection afforded by epoxy coatings on reinforcing steel was performed. To accomplish this bars were acquired from ten sources and coatings were characterized in terms of defects, thickness, solvent extraction weight loss and hardness. Testing involved exposure in various aqueous solutions at both ambient temperature and 80$\sp\circ$C and in chloride-contaminated concrete slabs under outdoor exposure. Direct pull-off adhesion testing was performed on tested and virgin epoxy coated reinforcing steels (ECRs). Electrochemical impedance spectroscopy (EIS) scans were made periodically, and a curve fitting technique was employed to analyze coating parameters. Conventional electrochemical measurements were also made, and corrosion morphology of ECR specimens was examined. It was found that the density and size of coating defects was the primary factor affecting ECR performance. The circuit analysis indicated that poorly performing defect-free coatings absorbed water and oxygen; and these species reached the coating/substrate interface and electrochemical reactions at the interface caused coating degradation. By way of contrast, the impedance response for well-performing ECR specimens showed no signs of active degradation at the interface although diffusional processes similar to those noted for poorly performing bars occurred here also. Experimental results indicated a relationship between corrosion behavior and bar source. Weight loss upon solvent extraction correlated with impedance reduction from hot water exposure. Coating defects developed during most of the tests, especially in high pH solutions containing chloride ions. ECRs with excessive coating defects, either initially present or ones which developed in service, performed poorly in every test category regardless of source. Forms of coating failure were extensive rusting at defects, blistering, wet adhesion loss, cathodic delamination, underfilm corrosion and coating cracks. These occurred sequentially or concurrently, depending on the condition of the ECR and nature of the environment.

Effects of (1) cement alkalinity (low, normal and high), (2) exposure conditions (RH and temperature), (3) rebar surface condition (as-received versus cleaned) and (4) density and distribution of air voids at the steel-concrete interface on the chloride threshold and time-to-corrosion for reinforcing steel in concrete have been studied. Also, experiments were performed to evaluate effects of RH and temperature on the diffusion of chloride in concrete and develop a method for ex-situ pH measurement of concrete pore water. Once specimens were fabricated and exposed to a corrosive chloride solution, various experimental techniques were employed to determine time-to-corrosion, chloride threshold, diffusion coefficient and void density along the rebar trace as well as pore water pH. Based upon the resultant data, several findings related to the above parameters have been obtained as summarized below. First, time for the corrosion initiation was longest for G109 concrete specimens with high alkalinity cement (HA). Also, chloride threshold increased with increasing time-to-corrosion and cement alkalinity. Consequently, the HA specimens exhibited the highest chloride threshold compared to low and normal alkalinity ones. Second, high temperature and temperature variations reduced time-to-corrosion of reinforcing steel in concrete since chloride diffusion was accelerated at higher temperature and possibly by temperature variations. The lowest chloride threshold values were found for outdoor exposed specimens suggesting that variation of RH or temperature (or both) facilitated rapid chloride diffusion. Third, an elevated time-to-corrosion and chloride threshold values were found for the wire brushed steel specimens compared to as-received ones. The higher ratio of [OH-]/[Fe n+] on the wire brushed steel surface compared to that of as-received case can be the possible cause because the higher ratio of this parameter enables the formation of a more protective passive film on the rebar. Fourth, voids at the steel-concrete interface facilitated passive film breakdown and onset of localized corrosion. This tendency for corrosion initiation increased in proportion to void size irrespective of specimen type. Also, [Cl -]th decreased with increasing void diameter. In addition, new ex-situ leaching method for determining concrete pore water alkalinity was developed.