U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
This magazine is an archived publication and may contain dated technical, contact, and link information.
|Publication Number: FHWA-HRT-07-006 Date: Sept/Oct 2007|
Publication Number: FHWA-HRT-07-006
Issue No: Vol. 71 No. 2
Date: Sept/Oct 2007
FHWA researchers test the corrosion resistance of various paint systems for steel structures.
|(Above) A worker on a steel bridge over Salmon Creek in northern California is power-tool cleaning the surface in preparation for overcoating, which is a less costly alternative to the traditional practice of enclosing sections of bridge for abrasive blasting, disposing of the removed paint, and repainting. Photo: Caltrans.|
"Of the nearly 200,000 steel bridges in the United States, about 10 percent require rehabilitation to prevent corrosion," says Chief Scientist Steven B. Chase of the Federal Highway Administration (FHWA). Removing rust and repainting a steel bridge is no small job. In fact, the California Department of Transportation (Caltrans) estimates that full removal of paint can cost as much as $35 per square foot. Because old paint typically includes lead, which is hazardous to human health and the environment, full paint removal requires abrasive blasting, dust containment, environmental monitoring, and waste removal and disposal. Worker health and safety must be protected as well.
To keep costs down while maintaining the service conditions of the Nation's steel bridges, Caltrans and many other State departments of transportation are turning to overcoats — applying a new coating on top of the existing one(s) — as an alternative to removing old paint.
"We're overcoating the majority of our steel bridges," says Senior Chemical Testing Engineer Andy Rogerson with Caltrans. The department maintains nearly 800 steel bridges statewide. Most have a red, lead-based primer coat, which for the most part is performing well, Rogerson says. When the topcoats start to fail, Caltrans applies waterborne primers and acrylic latex topcoats or, for harsher coastal climates, three-coat, moisture-cured urethane (MCU) overcoat systems.
Cost is the main advantage. Overcoat applications cost the agency $6 to $10 per square foot — nearly two-thirds less than the cost of full removal. "If rust covers less than 20 percent of a bridge, then we'll keep the lead primer and do an overcoating," Rogerson says.
But how well do these overcoat materials work in the long term? And under what circumstances do they perform most effectively? As new coatings emerge, such as new products with low levels of volatile organic compounds (VOCs), those will need to be evaluated for performance.
Between 2004 and 2006, to fill the need for data on the latest overcoat products, researchers at the FHWA Turner-Fairbank Highway Research Center (TFHRC) conducted an inhouse study to evaluate how various overcoat materials perform when they are applied to different types of aged steel substrates.
The FHWA researchers and their contractor staff selected six lead-free and low-VOC materials to apply over coated, aged, and rusted surfaces. Using a cyclic, accelerated testing method, they studied the overcoat systems in the laboratory and through field exposures, evaluating performance by assessing surface failures and rust creepage developed at scribes (scratches made through the overcoat surface down to the steel substrate). Comparing the results yielded a number of insights into overcoat performance when applied to the three types of substrates.
Most aging steel bridges are covered with one of two types of coating systems: (1) a two-coat system with an alkyd primer and topcoat (both coats could contain lead or just the prime coat) and (2) a two-coat system with an inorganic zinc (IOZ)-rich primer and a vinyl topcoat. Therefore, the researchers chose to study the performance of overcoats when applied to both types of coating systems and to a rusted steel base, cleaned using power tools according to the Society for Protective Coatings' (SSPC) specification SSPC-SP3.
Description of Coating Systems
|Base Systems||Lead Alkyd/Lead Alkyd (125/125 µm)a, (5/5 mil)||IOZ/Vinyl (100/115 µm), (4/4.6 mil)||SSPCSP3 Steel|
|Artificial Weathering||ASTM D5894, 3,360 hours||ASTM D5894, 3,360 hours||None|
|System Number||1||12||Alkyd/Silicone Alkyd
(50/50 µm) (2/2 mil)
(125/50 µ;m) (5/2 mil)
|4||9||15||HB Acrylic/HB Acrylic
(175/175 µm) (7/7 mil)
(175/175 µm) (7/7 mil)
|a Dry film thickness.
b VOC content.
A: High-pressure washed with 21 MPa (3,000 psi) potable water and air-dried.
B: Power-tool cleaned to SSPC-SP3 using a needle gun.
To set up the study, the researchers created rectangular steel test panels measuring 10.0 by 15.0 by 0.48 centimeters (4.0 by 6.0 by 0.19 inches). To prepare the substrates, they applied two coats of lead-based alkyd paint to some of the test panels, which had been cleaned using solvent according to specification SSPC-SP1.
For the samples that would have the IOZ/vinyl coating system as a base coat, they applied the IOZ/ vinyl paint to steel panels that were prepared according to specification SSPC-SP5, involving white metal blast cleaning (removing rust or foreign matter by blowing abrasives against the steel surface). The dry film thickness for the IOZ was 100 microns (4 mils), while the vinyl thickness was 115 microns (4.6 mils). Next the researchers subjected the test panels to 3,360 hours of cyclic, artificial weathering according to ASTM International testing standard ASTM D5894. After the weathering was complete, the researchers removed the panels from the test chamber, pressure-washed them with 21 megapascals, MPa (3,000 pounds per square inch, psi), of potable water, and air-dried them.
Laboratory Testing Conditions
Every 500-hour cycle included the following tests:
1. Freeze: 68 hours
2. UV/Condensation: 216 hours (9 days)
3. Cyclic salt-fog: 216 hours (9 days)
To create the third type of substrate (hereafter referred to as SSPC-SP3 surfaces), the researchers cleaned the remaining test panels according to SSPC-SP5 specifications and placed them outdoors for 6 months to produce a natural layer of corrosion. Then the researchers cleaned the surfaces to the SSPC-SP3 standard using a power tool (a needle gun). All three sets of panels then were painted with the six candidate overcoat materials. A 5.0-centimeter (2.0-inch) diagonal scribe was made on each set of panels.
For each test, the researchers used three scribed replicate panels of each coating system. For the laboratory test, they cycled the panels through freezing, ultraviolet light (UV)/condensation, and salt-fog/dry-air conditions a total of eight times over a period of 4,000 hours. The researchers alternated the hot salt fog, generated with a 5 percent by weight solution of sodium chloride (NaCl), with ambient air at 1-hour intervals during the third phase of each cycle. After each 500-hour test, the researchers examined and/or measured both the surface conditions and the rust creepage at the scribe.
Next, the researchers traced, scanned, area integrated, and calculated the creepage area of rust at the scribe for each panel. To find the average rust creepage, the researchers used the following equation: average rust creepage = A/2L, where A is the integrated area surrounding the scribe line (both sides) and L is the length of scribe line. (They used the areas on both sides of the scribe line to make the calculation more statistically meaningful.)
For each triplicate of panels, they obtained two values using ASTM D7087-05a (a standard test method for an imaging technique to measure rust creepage at the scribe on coated test panels subjected to corrosive environments). Then the researchers plotted the average rust creepage from the scribe, namely scribe creepage, which is defined as the average rust creepage measured from the scribe line. The researchers obtained the scribe creepage values by taking the average of six rust creepage values representing three scribed panels with two test values for each coating material.
In addition, using standard methods, the researchers measured the gloss, adhesion strength, and pencil hardness of the samples before and after the laboratory test.
All the overcoat materials tested during the study had high percentages of solid content, ranging from 66 to 89 percent by weight, which means the materials have low VOC content. The pigment content, as a part of the solid, ranged from 41 to 80 percent by weight of the dry film, with the exception of zero or a low amount found in the epoxy and MCU sealers, respectively. The researchers found that most of the pigment fractions contained a significant amount of aluminum, phosphorus, iron, and zinc, which are effective elements for protecting steel from corrosion. Further, they found that both the elastomeric high-build (HB) acrylic and calcium sulfonate (CSA) contain zinc phosphate, a proven anticorrosive pigment.
The researchers measured the pencil hardness of the six overcoat materials according to the ASTM D3363 method. In the pencil hardness test, the "H" stands for hardness, the "B" stands for blackness, and "HB" refers to hard and black pencils. The scale ranges as follows from softest to hardest, with "F" falling at the middle of the hardness scale: 6B, 5B, 4B, 3B, 2B, B, HB, F, H, 2H, 3H, 4H, 5H, and 6H. The alkyd showed hardness of HB on the pencil hardness test; the two-coat MCU (MCU2), three-coat MCU (MCU3), and epoxy all showed the hardness of 2H. CSA and HB acrylic, however, were very soft, with a hardness of 6B.
Chemical Composition of Overcoat Materials
|Overcoat Material||Alkyd||MCU||MCU Sealer||HB Acrylic||CSA||Epoxy||Epoxy Sealer|
|Solid Content, Percent by Weight of Wet Film||67.5||87.0||76.1||66.2||74.02||83.1||89.0|
|Pigment Content, Percent by Weight of Dry Film||71.0||79.9||19.7||40.6||48.0||68.9||0|
|Elemental Content, Percent of Total Metals||Na||16.2||11.7||12.1|
|Film Thickness, m (mil)||100 (4)||175 (7)||190 (7.6)||350 (14)||350 (14)||185 (7.4)|
As a result of the laboratory testing, the researchers found that the alkyd, MCU2, and MCU3 overcoat materials were more resistant to the UV lamp, retaining 40 to 60 percent of their original gloss (or shine), while the epoxy retained only 20 percent of its gloss. Even though HB acrylic and CSA retained 70 to 80 percent of their gloss, both had very low gloss before the test. High gloss retention is important for ensuring and maintaining coating appearance over time.
Nearly all the overcoat systems retained their adhesion strength values after the test, meaning a reduced chance for moisture, oxygen, or salts to penetrate through the coating film to cause corrosion. The measured adhesion strength is the strength needed for the weakest point to fail in a coating system. The researchers measured the adhesion strength using ASTM D4541 (a standard test for the pull-off strength of coatings) both in the laboratory and in the field. Most of the coating systems had high adhesion strength, except two elastomeric coatings (HB acrylic and CSA) that had adhesion strength of only 2.8 to 3.5 MPa (400 to 500 psi). The researchers found that the majority of the failures during the adhesion test were cohesive failures (which means breaks within the layers, rather than breaks between layers) within the lead-based alkyd and IOZ primer layers themselves. The observed cohesive failures suggest strong adhesion of most of the overcoat materials to the basecoat surfaces. Conversely, they found that the CSA and HB acrylic overcoats showed cohesive failures indicating that they have low mechanical strength themselves. For overcoats on the SSPC-SP3 surfaces, the researchers observed adhesive failures (breaks between layers) between the steel and overcoat primer for all cases, indicating weak adhesion between steel and the overcoats.
Surface failure. Only one of the panel surfaces, system 1 (panel numbers 160 and 167), showed a surface failure, in this case some rust-through. These panels are SSPC-SP3 surfaces with overcoats of two coats of alkyd. The researchers concluded that this rust-through failure may have been caused by the thinness of the overcoat film — only 100 microns (4 mils), as recommended by the manufacturer for overcoating. All the other overcoat materials were applied at a thickness of at least 175 microns (7 mils) over SSPC-SP3 surfaces.
|This bar chart shows the adhesion strength of overcoated systems before and after the 4,000-hour laboratory test. The substrates were aged lead-alkyd, aged IOZ/vinyl, and rusty steel surfaces that were power-tool cleaned. Source: FHWA.|
Based on these results, the researchers concluded that coating thickness plays a critical role in development of surface rust-through. But at the same time, system 1 exhibited only a small amount of rust creepage at the scribe. Normally, a poorly performing coating would show both severe surface failure and large rust creepage at the scribe. These unexpected results suggest that there is no direct correlation between surface performance and scribe performance for overcoat systems. In theory, surface performance depends on the thickness of the coating film, but rust creepage developed at the scribe is affected by adhesion and the corrosion resistance of the primer.
Scribe failure. The scribe rust creepage developed by all the overcoat systems grew linearly with time. The performance results of the various coating materials over lead-based alkyd, IOZ/vinyl, and SSPC-SP3 surfaces after the 4,000-hour laboratory test were as follows.
For the samples with coatings applied over the aged lead-based alkyd surface, the performance of the overcoat materials at the scribe was best for CSA and decreased respectively for MCU2, epoxy, MCU3, alkyd, and HB acrylic. With the exception of CSA as the standout performer, the remaining overcoat materials showed similar defects at the scribe.
For the overcoat materials applied to the aged IOZ/vinyl surfaces, CSA performed the best at the scribe, followed in decreasing order by epoxy and HB acrylic, then MCU2 and MCU3. In fact, the researchers determined that MCU2 and MCU3 were somewhat unsuitable for overcoating IOZ/vinyl surfaces.
And for the SSPC-SP3 surfaces, the steel panels that were allowed to rust and then power-tool cleaned, the researchers ranked the coating performance in decreasing order of effectiveness: CSA; then alkyd; followed by MCU3, epoxy, and MCU2; and, finally, HB acrylic. The researchers concluded that these results show CSA has a unique high affinity to less perfect steel surfaces and provides high resistance to corrosion.
|Panels 160 and 167 are lead-based alkyd surfaces that used alkyd paint as an overcoat. Following the 4,000-hour laboratory testing, these samples showed rust-through spots on the surface, which may be due to the thinness of the overcoat.|
Comparing the results for the rust creepage at the scribe obtained from the three overcoated substrates, the researchers found that CSA performed the best, even though it has low gloss and a tendency to pick up dirt. These characteristics indicate that CSA could be a practical choice for use in areas where corrosion prevention is more important than appearance. The test data further revealed that the performance of the other overcoat materials varies by type of substrate.
|Shown here is the rust creepage at the scribe developed by various coating materials over aged alkyd surfaces after the laboratory test. All the rust creepage at the scribe increased linearly over time. The HB acrylic overcoated system developed the most rust creepage at the scribe, while the CSA system developed zero rust creepage after the 4,000-hour laboratory test. Source: FHWA.|
Concerning performance of the overcoats under real-world conditions, the researchers found that most of the paints lost 80 to 90 percent of their gloss after the 24-month outdoor exposure at Sea Isle City, NJ. HB acrylic was the best performer, losing only 70 percent of its gloss. The high-intensity UV light at Sea Isle City might be the cause of these significant gloss reductions.
Surface failure. System 1 (alkyd, panel numbers 164 and 165) showed severe rust-through, performing much worse than it did in the laboratory test, possibly due to the harsh conditions at the outdoor site. In addition, system 2 (MCU2, panel numbers 174 and 175) developed 8VD (very dense, size 8 blisters as defined in ASTM D714) blistering, a failure mode the researchers did not observe following the laboratory test. The researchers expect that this blistering could lead to coating delamination or rusting over time. The panels with MCU3 overcoats, however, did not show blistering — a difference in performance that the researchers say indicates the effectiveness of using the sealer prior to applying the two-coat MCU over lead-based alkyd.
Scribe failure. After the 24-month exposure, all the coating systems started to show linear growth in rust creepage at the scribe with exposure time. For the lead-based alkyd substrates, the researchers found that MCU2 and MCU3 did not perform as well at the scribe as other overcoat materials in the salt-rich, high-UV environment. HB acrylic, CSA, alkyd, and epoxy, however, all performed about the same — better than MCU2 and MCU3. The panels with CSA overcoats developed rust creepage in similar amounts to the panels coated with the other materials; apparently the CSA did not soften the alkyd substrate to increase the adhesion of the lead-based alkyd to the steel surface at the outdoor exposure site. The researchers concluded that CSA might have lost its softening power because its solvent evaporated faster under the hot temperatures at the exposure site.
|This figure shows the rust creepage at the scribe developed on the three different substrates using overcoats of various materials. The CSA overcoat system showed the least amount of rust creepage at the scribe for each type of substrate. Source: FHWA.|
No rust creepage occurred at the scribe for any of the panels with the IOZ/vinyl basecoat. IOZ takes a long time, usually weeks or even months, to cure and cures more completely at the scribe in an outdoor environment due to the availability of moisture and air ventilation. Recognizing this characteristic, the researchers expect that extended exposure to outdoor conditions will cure the IOZ primer further and thereby increase the protection it offers to steel.
When placed over the SSPC-SP3 steel surface, the epoxy performed three times worse than the other overcoat materials at the scribe and exhibited 17 millimeters (0.67 inch) of rust creepage after 24 months of outdoor exposure. This large creepage may be caused by the fact that epoxy chalks badly under the high-intensity UV light at Sea Isle City. At the scribe, the edge of the entire epoxy layer over the SSPC-SP3 steel surface was exposed to UV light, so the chemical composition of the chalked epoxy changed, thereby losing its ability to protect the steel surface. Again, CSA was the best performer.
|Panels 164 and 165 show the coating condition of lead-based alkyd covered with an alkyd overcoat after 24 months of exposure outdoors in New Jersey. The panels show severe rust-through spots on the surfaces.|
Comparing the performances of the various coating materials when applied to the three substrates and exposed to both laboratory and outdoor testing led to a number of key findings. Although rust creepage developed at the scribe and grew linearly over time for each sample, the overcoat materials performed differently though applied to the same primer substrates. The difference in performance depended upon the wetting or penetrating properties of the individual overcoat material, which contains different types and amounts of solvent and resins with varying penetrating power. The longer the overcoat material takes to cure, the more solvent is available to soften the coating substrate and thereby increase the primer's adhesion to steel. As a result, the researchers concluded that the difference in the amount of rust creepage at the scribe was due to the variability in primer adhesion.
Further, through these tests, the researchers summarized the following findings for overcoat strategies.
Overall, CSA performed the best on all three substrates. However, it is a soft material that picks up dirt easily. Given these strengths and weaknesses, the researchers advise bridge owners to use their best judgment in deciding whether to use CSA as an overcoat material.
|Panels 174 and 175 show the coating condition of SSPC-SP3 surfaces covered with an MCU2 overcoat after exposure outdoors for 24 months. The panels show small blisters all over the surfaces, which are a precursor to eventual corrosion or rust.|
Additional insight: "Stay on top of the surface preparations or the overcoat could fail prematurely," says Rogerson from Caltrans. "Soluble salts, such as sulfates, nitrates, and chlorides, within the base coatings can lead to blistering in the overcoats applied over them. With some acrylic latex coatings, we've had to clean and test for the amount of soluble salt on the steel, and sometimes do an additional cleaning step to make sure the overcoat sticks."
Rogerson adds, "Continued research is important to prove the longevity of overcoats and ensure compliance with the latest VOC requirements. In California, we're continuing to see the maximum allowable VOC content level drop. The limit is 250 grams per liter (2.08 pounds/gallon) VOC content in northern California, but just 100 grams per liter (0.83 pounds/gallon) in the Los Angeles area. And the numbers keep changing. With 35 air pollution control districts in the State, from an owner's perspective it becomes difficult to keep on top of the different rules and definitions. For that reason we're trying to meet the most stringent requirements and use those statewide."
The degree of surface failure for each system assessed during this study will be measured by electrochemical impedance spectroscopy, and the test data will be published at a later date.
|This figure shows the rust creepage at the scribe on the three different substrates with various overcoat materials when the panels were exposed to real-world conditions outdoors in New Jersey for 24 months. The overcoated lead-alkyd system developed 5-6 millimeters (0.19-0.24 inch) of rust creepage at the scribe. The IOZ/vinyl system did not develop any rust. For the SSPC-SP3 steel surface, epoxy was the poorest performer in terms of rust creepage. Source: FHWA.|
Gloss Retention by Overcoat Material on Three Substrates
|The bar graph summarizes the gloss retention of various overcoat materials for three substrates (lead/alkyd, IOZ/vinyl, and SSPC-SP3) after a 4,000-hour laboratory test.|
Rust Creepage at the Scribe on IOZ/Vinyl Surfaces
|Shown here is the rust creepage at the scribe on various coating materials applied over aged IOZ/vinyl surfaces. In all cases, the rust creepage at the scribe increased linearly over time. The CSA, epoxy, and HB acrylic systems developed less rust creepage than MCU2 and MCU3.|
Rust Creepage at the Scribe on SSPC-SP3 Surfaces
|Shown here is the rust creepage at the scribe on various coating materials applied over SSPC-SP3 steel surfaces. In all cases, the rust creepage at the scribe increased linearly over time. The CSA system developed zero rust creepage, while the HB acrylic system developed the highest amount of rust.|
Rust Creepage at the scribe Versus Exposure Time
|This figure shows the rust creepage at the scribe developed by various coating materials over aged alkyd surfaces after the 24-month outdoor exposure. In all cases, the rust creepage increased linearly with exposure time. The MCU2 overcoat system developed the highest amount of rust creepage, while HB acrylic showed the least amount of rust during the test period.|
Rust Creepage at the Scribe Over SSPC-SP3 Steel Surfaces Versus Exposure Time
|This figure shows the rust creepage at the scribe developed by various coating materials over SSPC-SP3 steel surfaces after 24-month outdoor exposure. In all cases, the rust creepage increased linearly with outdoor exposure time. The CSA system developed the least amount of rust while epoxy showed the highest amount of rust creepage at the scribe.|
Dr. Shuang-Ling Chong, recently retired from FHWA, is a senior chemist with more than 30 years' experience in various fields of chemical research. She joined FHWA in 1989 and studied generic types of bridge coatings, including low-VOC coatings, MCUs, waterborne coatings, two-coat systems, and one-coat bridge materials. She also conducted research on leaching of blasted paint residues, chloride testing methods, and failure analyses. Chong served as manager of the Coatings and Corrosion Laboratory at TFHRC until her retirement in April 2007. She was a member of numerous industry and standards organizations, including SSPC, ASTM, the Transportation Research Board, and the American Chemical Society.
Yuan Yao is a senior chemist with SaLUT, Inc. Since 1994 she has been an onsite contractor in the TFHRC Coatings and Corrosion Laboratory. She has supported many chemistry and coating projects, applying a range of testing and analytical techniques to performance evaluations of various coating types.
For more information, please visit www.fhwa.dot.gov/research/tfhrc/ or contact Yuan Yao at 202-493-3092, email@example.com. To view additional figures summarizing the results from this study, please visit the Public Roads Web site at www.fhwa.dot.gov/publications/publicroads/index.cfm.