Predicting Corrosion Behavior of A36 Plain Carbon Steel and A588 Weathering Steel in Bridge Applications

Abstract

The annual direct cost of corrosion on bridge structures is over $8.3 billion. Much of this cost is due to broadly applied corrosion mitigation and prevention methods for steel components. On-going efforts by many state and local agencies focus on reducing maintenance costs through improved corrosion detection and prevention methods. In our research, we focus on linking corrosion mechanisms to corrosion predictions for materials commonly used for steel beams on bridges. Plain carbon steel remains a common material for bridge construction despite its poor corrosion performance. Porous and fragile corrosion products such as Hematite (α-Fe2O3), Goethite (α-FeOOH), Akagneite (β-FeOOH), and dense and highly adhesive Lepidocrocite (γ -FeOOH) form. Due to the porous corrosion layer, a high concentration of chlorine ions are trapped, further accelerating corrosion. To prevent rapid corrosion, typical protection measures on bridge steels in Connecticut include the application of a three-layer paint system containing primer, intermediate urethane coat, and an epoxy topcoat. Breakdown of the coating can result in pitting corrosion below the paint layers, which is difficult to detect visually. Periodic removal and replacement of paint is essential to long-term maintenance. In fact, the paint layer protection is typically designed to last 15-20 years, and removal and recoating costs range from $5-$20 per square foot, creating a high driving force to identify lower cost options for maintenance. Weathering steel is a potentially lower-cost alternative. A primary benefit of weathering steel (WS) over plain carbon steel is its ability to form dense, protective, and highly adhesive Lepidocrocite (γ-FeOOH) layer on the surface. With continued exposure, corrosion resistance of WS increases. Although the corrosion behavior of plain carbon steel and WS have been heavily researched, prediction of the corrosion performance of these steels coupled in coated and uncoated bridge applications remains difficult. In this study, we conducted a series of accelerated corrosion tests and evaluated the electrochemical corrosion behavior of A36 plain carbon steel and A588 weathering steel to predict corrosion rates. Accelerated corrosion testing of both bare and scratched painted A36 and A588 steel coupons of 50 x 30 x 5 mm was conducted through 200 cycles of wet/dry immersion at room temperature in 3.5% NaCl and simulated seawater (ASTM D1141). Bare steel samples were polished with 800 grit SiC paper and cleaned with ethanol and distilled water. Painted samples were polished with 240 grit SiC paper and cleaned with ethanol and distilled water prior to coating with zinc primer Carbozinc® 11 HS, epoxy coating Carboguard® 893, and polyurethane top-coat Carbothane® 133 LV. After the coating, the samples are scratched with diamond wafering blade on an IsoMetTM saw. Samples were photographed using a Nikon D5000 DSLR camera and weighed using microscale (FisherScientific XA-200DS) after every 20 cycles. A36 and A588 coupons were characterized by Potentiodynamic polarization (PDP) and Electrochemical Impedance Spectroscopy (EIS) testing using a Bio-Logic VSP-300 potentiodynamic tester. Sample surfaces were polished with 800 grit SiC paper and cleaned with ethanol and distilled water for PDP testing to establish baseline corrosion rates prior to oxide formation. EIS testing was applied to coupons in the original (clean surface) and corroded surface state to characterize the electrical nature of the corrosion layers formed on the surface. In addition, all corroded sample surfaces were analyzed using XRD (Rigaku SmartLab) and GIXRD, and microstructural features were examined using optical and electron microscopies. All bare surface samples demonstrated continuous weight gain during wet/dry cycling. Scratched samples gained weight at a slower rate. Both corrosion potential (Ecorr) and corrosion current (i corr) of the A36 samples was found to be higher than that of A588, and the corrosion rate of A36 is calculated at nearly 2,600 mmpy compared to 570 mmpy for A588. The EIS results indicate that A36 surface layers have a slightly higher electrical resistance than A588 surface layers, but the capacitance of A588 corrosion products is higher than for A36. Overall, the corrosion resistance of A588 is better than the corrosion resistance of the A36, as expected. The results from electrochemical testing correspond to the Scanning Electron Microscopy (SEM). The cross-sectional microstructural analysis of sample revealed A588 contains denser and thicker corrosion layer, and A36 corrosion products are highly porous and detached from the substrate. Furthermore, severe pitting was observed on the A36 surfaces, and mostly uniform corrosion was observed on the A588 surfaces. The pitting observed scratched samples for both A36 and A588 was similar and not severe.