Spot Painting to Extend Highway Bridge Coating Life: Volume 2: Research Overview (2018)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

NCHRP Project 14-30 6 CHAPTER 2 Laboratory Accelerated Performance Testing of Spot Coating Systems 2.1 Background A major project requirement was to, “develop guidelines for the selection and use of a broad spectrum of protective materials for spot treatment of failed coatings on steel highway bridges,” more specifically, “testing protocols to determine the anticipated performance and service life of protective material types for use on rusted, soluble salt-contaminated substrates prepared to SSPC-SP 3 standard.” This requirement was addressed by applying accelerated laboratory testing of candidate spot coatings using modifications of ASTM standard accelerated coating performance tests, in particular ASTM D5894, “Standard Practice for Cyclic Salt Fog/UV Exposure of Painted Metal, (Alternating Exposures in a Fog/Dry Cabinet and a UV/Condensation Cabinet),” and ASTM B117, “Standard Practice for Operating Salt Spray (Fog) Apparatus.” There were numerous reasons for doing this: • Those tests are internationally recognized standards for evaluating protective coatings • Coating manufacturers typically evaluate their coatings prior to manufacture using one or both of those tests providing some basis for anticipating testing performed by KTC • Equipment and test procedures for those tests are widely used by other firms in coatings industry including consultants, end users, industry associations (e.g. AASHTO NTPEP) and universities so any methods developed under this study can be readily adopted by others • Those methods have proven effective for screening coatings by the coatings industry in particular and specified by many industries and technical sectors • There is a large body of information relative to their use • Those methods are adaptable to wide variety of coating materials including standard liquid-applied and non-traditional materials (greases, tapes) • They provide usable test data in a relatively short period (typically less than one year) which is advantageous in the rapidly changing coating industry where coating formulations can be affected by availability, cost, and environmental regulations. • The tests can be used co-jointly to approximate conditions that exist at various locations on bridges • There is some experience, at least from KTC’s perspective, that specific coating performance levels in those tests can be equated to historic service performance on bridges Test protocols to evaluate the performance of candidate spot coatings/systems should: • Provide data in a short timeframe • Be relatable to actual service conditions • Yield reasonable results In the past, field test patches were used to evaluate candidate overcoating coatings/systems for both performance and compatibility with existing coatings. That approach has been mostly abandoned for performance testing due to the effort and timeframe needed to produce viable results. Accelerated laboratory performance testing has largely supplanted field trials. Standard coating test intervals produce satisfactory results in about nine months. Prior to

NCHRP Project 14-30 7 ASTM promulgating its standard for cyclic weather/corrosion testing ASTM D5894 in the mid- 1990s, KTC used that method to test overcoating systems. Prior to this research, KTC employed that method to test approximately 100 different coatings/systems-primarily barrier systems used for overcoating. KTC has also employed ASTM B117 testing to evaluate coatings, but rarely performed it as ASTM D5894 had generally supplanted it for testing coatings to be used for atmospheric service. Those tests included coatings placed over a variety of steel test panel substrates including uncontaminated rust, power-tool cleaned rust, mill scale and abrasively blasted steel (both uncontaminated and contaminated with soluble salts). Many of the coatings/systems that performed well on those tests were subsequently used by the Kentucky Transportation Cabinet (KYTC) on experimental bridge maintenance painting projects. In most cases, they provided good service performance as determined by follow-up periodic inspections of the experimental projects and were adopted by KYTC for routine use as maintenance coatings. To perform those tests, KTC developed standard operating procedures (SOPs) used by laboratory technicians to ensure consistency. Accurate evaluation of candidate spot coatings requires the use of representative service exposures and test panel substrates in the test protocol. Two different types of exposures reflect common service conditions of failed bridge coatings monitored by KTC researchers: 1) bold exposure using ASTM D 5894 and 2) sheltered exposure with extended time of wetness (TOW) using ASTM B117. Spot painting is performed over 1) an area where the existing coating has failed, 2) a boundary between the failed area and existing coating and 3) the intact existing coating. Typically, the failed area contains a surface consisting of rust, mill scale, cohesive/adhesive failed existing coatings or a combination of all three. In many cases, it is contaminated with soluble salts. Prior to spot painting, where rust is present, the failed areas are mechanically prepared to eliminate loose material (SSPC-SP 2, “Hand Tool Cleaning,” or SSPC- SP 3, “Power Tool Cleaning”) or to provide a profiled steel substrate (SSPC-SP 11, “Power Tool Cleaning to Bare Metal,” SSPC-SP 15, “Commercial Power Tool Cleaning,” or various abrasive blasting standards). If a failed area consists of inter-/intra-coat disbonding and it is entirely covered by a portion of the existing coating, it can be lightly abraded by hand using sand paper prior to spot painting. The intact existing coating is usually weathered causing it to deteriorate and form a white powder (chalking). That is usually more severe on bridge surfaces which are exposed to direct sunlight (UV exposure). The surfaces may be contaminated with soils and hydrocarbons, commonly from diesel fumes. They may also be contaminated to a certain extent with soluble salts from atmospheric exposure and deicing salts. Those issues can be addressed by solvent cleaning, washing, dry wiping, and soluble salt remediation using pressure washing, sometimes using chemical treatments. The boundary (or border) between the failed area and intact existing coating must be defined by hand- or power-tool cleaning. An intact border is typically assessed by probing the edge of the existing coating with a dull scraper to ensure that it is firmly adherent. Commonly, the edge of the existing coating is feathered by hand or power tool abrading using sand paper or non-woven pads. The existing coating around the repair area needs to be cleaned where the spot coating will over lap it in coating the repair area and feathered boundary. Where a repair coating is placed over an existing one on bridges, there is a possibility that they will be composed of different generic coating types/chemistries and may experience incompatibility problems. In developing the coatings testing protocol, certain assumptions were to be made about level of surface preparation to be replicated, the degree and types of soluble salt contamination; the coatings (coating systems) employed both as existing and repair test systems and the method of application (for liquid-applied coatings). To replicate large area spot painting methods used by

NCHRP Project 14-30 8 some state highway agencies that employed abrasive blasting to SSPC-SP 6/ NACE 3, “Commercial Blast Cleaning,” or SSPC-SP 10/NACE 2, “Near White Metal Blast Cleaning,” (used with a coating system employing an organic zinc primer) standard ASTM D 5894 testing would probably suffice over properly prepared panels or state highway agencies could use the AASHTO NTPEP steel coatings test results. This protocol, as mandated by the project RFP, was to evaluate cleaning to the SSPC-SP 3, “Power Tool Cleaning” standard. The test protocol reflected not only that requirement, but the intent of the review panel was to address methods that could be readily applied by state highway agency in-house forces. Survey responses from some state highway agencies noted significant regulatory issues with pressure washing bridges related to the capture and disposal of the resulting wastewater. State highway agency requirements varied widely based upon the regulations of Departments of Natural Resources (DNRs) in each state. KTC researchers believed that could be a major stumbling block for some state highway agencies in adopting spot painting, especially when using in-house crews. Additionally, KTC researchers felt that state highway agencies using less vigorous surface preparation methods wished to limit spot painting costs and potential motorist disruptions. Pressure washing, while effective in reducing soluble salt levels, typically requires an extra day on-site for the washed surface to dry properly prior to additional cleaning/painting operations. As part of the KTC protocol, the decision was made to eliminate the use of pressure washing to limit soluble salts (with the exception of several low water consumption tests described below). In the past, KTC had successfully used dry wiping with burlap to clean off surface soils on coating test patches and felt that a dry method to remove chalk, loose materials and soils in a wipe down prior to painting would eliminate state highway agency concerns about capturing and treating wastewater and the need for a hold period to allow for drying after pressure washing. In situations where soluble salt concentrations warranted concern, a state highway agency could elect to employ other salt remediation treatments including pressure washing. Based upon the literature review KTC performed as part of the interim report and other KTC field tests on bridges for chlorides, KTC researchers determined that any test coatings should be placed over prepared substrates with about 20-30 µg/cm2 of chloride contamination in mild/moderate atmospheric exposures. Soluble salt effects from other sulfates and nitrates were not considered as commonplace as chlorides in those exposures. As SSPC-SP3 cleaning was to be incorporated into the protocol, coatings systems evaluated would not benefit from the use of zinc primers. Rather, inhibitive and barrier coating systems would need to be considered. It was anticipated that most of the spot painting work addressed by state highway agencies would be on relatively small areas that could be best painted by brushes or rollers. That coupled with the small size of the test panels, and the irregular shape of some test panels incorporated into the KTC laboratory test program, led KTC to incorporate brush application in the testing protocol. KTC researchers decided to use ASTM A572 high-strength low alloy columbium- vanadium steel (50 grade) for the test panels in the laboratory tests. Although that steel had a higher corrosion resistance than some other grades, it typically corroded to provide a pitted substrate that posed a difficult surface for SSPC-SP 3 cleaning with the potential to provide “hot spots” in the pits that trapped small amounts of soluble salts in high concentrations. The trapped soluble salts would have a negative effect on coating performance and enable the test protocol to be able to deteriorate coatings in a manner consistent with that occurring on bridges. The steel test panels were ordered from a steel distributor pre-cut to the final size for the laboratory tests - 6 in. (15.24 cm) x 4 in. (10.16 cm) x 1/4 in.(0.64 cm). The mill analysis of the ASTM A572 steel is provided in Appendix A. KTC employed two Q-Lab Q-Fog Cyclic Corrosion (fog/dry) Testers capable of performing both ASTM D 5894 and B117 tests (Figure 1). KTC also had four QUV/se Accelerated

NCHRP Project 14-30 9 Weathering Testers capable of performing ASTM D 4587, “Practice for Fluorescent UV- Condensation Exposures of Paint and Related Coatings” with SOLAR EYE irradiance control (Figure 2). That provided KTC a large test capacity to test multiple combinations of spot coatings, existing coatings, test panel substrates, and specimen types. The proposed test program was modified as requested by the project review panel. The final program incorporated three “existing” coating systems consisting of: 1) two coats of acrylic coatings, 2) two coats of alkyd coatings and 3) an epoxy primer with a two-component polyurethane finish coat. Those were chosen to replicate existing coatings to be repaired on bridges. Six liquid-applied coating systems were to be tested under the KTC test protocol: 1) a calcium sulfonate alkyd (applied in two-coats), 2) a two-coat alkyd system (including a phenolic alkyd primer and silicone alkyd topcoat), 3) a two-coat acrylic system (including a direct-to-metal primer and a UV-resisting topcoat), 4) a two-coat moisture cure urethane (MCU) system including a MIO-aluminum pigmented primer with a high-gloss UV-resisting topcoat, 5) an epoxy penetrating sealer with a two-component polyurethane topcoat and 6) a micaceous iron oxide (MIO) pigmented epoxy primer with a two-component polyurethane topcoat. Two non-traditional coatings were also selected for testing: 1) an aluminum-blend, heavy duty multi-purpose grease and 2) a non-curing visco-elastic polymer supplied as a self-adhering tape in rolls 9-in. (22.9 cm) wide x 34 mils (859 microns) thick. Those were selected serve as “repair” coatings for spot coating failures of the replicate “existing” coatings. 2.2 KTC Test Protocol The primary methods used to evaluate candidate spot coatings were accelerated laboratory weathering/corrosion testing (ASTM D5894) and corrosion testing (ASTM B117). Two types of tests were performed utilizing flat panel specimens (Type I) and shaped test panels consisting of lapped/bolted flat panels (Type II). The Type I tests were used to evaluate coating performance over local distress (spot failures). The Type II test was used to evaluate coating performance over irregular substrates with spot failures at edges, faying surfaces and around the bolts. Type I Tests - ASTM D5894 testing was used to replicate spot coating performance under conditions of bold exposure (exposure to direct sunlight-UV deterioration with cyclic condensation and evaporation). ASTM B117 testing was used to replicate sheltered performance with exposure to deicing salt runoff and extended time of wetness (TOW) with inconsequential UV exposure (no UV exposure occurring in the B117 testing). Figure 1. KTC Cyclic Corrosion (Dry/Fog) Test Chambers. Figure 2. KTC QUV Accelerated Weathering Testers.

NCHRP Project 14-30 10 To evaluate the potential steel substrates, half of the Type I test panels had initial test surfaces with intact mill scale and the other had the mill scale removed by blast cleaning to an SSPC–SP 5 White Metal Blast Cleaning condition. ”Existing” coatings were applied to half of the test surface of each panel with the remainder of the surface consisting of steel or mill scale left uncoated (Figure 3). All test panels were to be conditioned by exposing the steel/mill scale surfaces to a corrosive spray in a test chamber to replicate rusted surfaces found on bridges. Initially, KTC planned to weather the “existing” coatings. Due to the duration of the proposed ASTM D5894 testing, concern arose about excessive UV exposure causing premature failure of the “existing” coatings and that step was eliminated. The conditioning was also intended to create rust contaminated with chlorides at a nominal level from 20-30 µg/cm2 (~166 to 250 µS/cm) after SSPC-SP 3 surface preparation (prior to painting). Sulfate contamination was considered less significant than chlorides and nitrates. A series of exposure tests were performed using the test panels in mill scale and blasted substrate conditions using ASTM G85, “Standard Practice for Modified Salt Spraying - Annex A5 Dilute Electrolyte Cyclic Fog/Dry Test.” It was determined that 250 hours of exposure provided sufficient time to properly charge both types of rusted surfaces with chlorides. That procedure was used to condition (corrode) the exposed steel surfaces on the Type I and Type II test panels. The rusted area and a narrow strip of the existing coating were to be subjected to mechanical surface preparation to the SSPC-SP 3 standard. The depth of overlap of the surface preparation into the existing coatings was dependent upon when firmly adherent coating was encountered. Candidate repair coatings were to be applied over the panel surfaces, cured, and subjected to the aforementioned laboratory accelerated performance tests to evaluate their performance as “repair” coatings for spot painting. Exposed Test Surface of Existing 3 in. Weathered 3 in. Rusted Steel/ Mill Scale Cleaned Per SSPC SP 3 1 in. SSPC SP 3 Cleaning Overlap/Feathering over Existing Coating (max.) 5 in. Spot Coating 6 in. 4 in. Spot Coating over E i i Spot Coating over Rusted Steel/Mill Scale Figure 3. Accelerated Performance Type I Test Panel (0.25 in.-Thick ASTM A572 Steel)

NCHRP Project 14-30 11 The Type I test protocol was as follows: A. A sufficient number of appropriate steel test panels for a total of 336 Type I tests were prepared (Figure 4). B. Half of the test panels were blast cleaned to a NACE 1/SSPC-SP 5, “White Metal Blast Cleaning,” condition. The other half remained unprepared with the mill scale intact. C. Prior to further preparation, a moisture cure urethane coating was sprayed on the back faces of the panels to prevent them from corroding during the testing. The “existing” coatings were applied to one-half of the face of test panels in each group providing 3 in. (7.62 cm) x 4 in. (10.16 cm) coated areas. Those coatings were spray applied, dry film thicknesses (DFTs) measured and cured at minimum for 28 days prior to follow- on surface preparation work. Each panel/set was given a unique alphanumeric designation placed on the backside of the panels in indelible ink. That was done to prevent mix-ups in follow-up operations with the panels that, in some cases, spanned nearly two years. D. The panels were subsequently conditioned per ASTM G85 Annex 5 for 250 hours of exposure in accordance with ASTM D5894. The fog/dry chamber ran a 1-hour cycle of fog spray at ambient temperature ~ 72o F (22.2o C) and 1-hour dry off at 95o F (35o C). The fog electrolyte was an aqueous solution of 0.05 % sodium chloride and 0.35 % ammonium sulfate. Deionized water was used for the conditioning and the entire test program (0-5 µS/cm conductivity). During conditioning, the exposed steel/mill scale portion of the panels were positioned above the painted portion in the test chambers which were affixed to the chamber mounting brackets. To prevent rust bleed onto the “existing coatings,” the painted portions of the test panels were masked with tape during conditioning. E. Type I panels previously conditioned and cleaned per SSPC-SP 3 were tested for chloride contamination in accordance with SSPC: TECHNOLOGY GUIDE 15 “Field Methods for Retrieval and Analysis of Soluble Salts on Steel and Other Nonporous Substrates” using the sleeve method 5.2.5.1, 5.2.5.3, and 5.2.5.4 for chlorides, nitrates and sulfates respectively. The panels were tested for chlorides, sulfates and nitrates. The results were: Mill Scale Panels – chlorides 27 µg/cm2 (~227 µS/cm); sulfates 14 µg/cm2 (~116 µS/cm); nitrates 0 µg/cm2 (0 µS/cm) Blast Cleaned Panels – chlorides 24 µg/cm2 (~200 µS/cm); sulfates 7 µg/cm2 (~58 µS/cm); nitrates 0 µg/cm2 (0 µS/cm)

NCHRP Project 14-30 12 F. Mechanical surface preparation to an SSPC-SP 3 “Power Tool Cleaning” condition was performed using a 4-inch (10.16 cm) grinder with a knotted wire brush on the rusted portions of the panels (Figure 5). As no hazardous materials were involved, the grinders were not shrouded. That helped in cleaning at the boundary between the paint and metal substrate and minimized the unnecessary removal of adherent “existing” coatings. After surface preparation, the panels were visually inspected using SSPC- VIS 3 “Guide and Reference Photographs for Steel Surface Prepared by Power and Hand Tool Cleaning” to ensure proper cleaning. The panels were also rag wiped to ensure all loose material had been removed. KTC researchers completed power tool cleaning work within a few days of the “repair” coating application to preclude the possibility of the prepared surfaces flash rusting in the interim. At the boundary of the existing coatings and SP 3-cleaned area, the “existing” coatings were feathered manually using 100 grit sand paper (Figure 6). The cured “existing” coatings were hard and it was difficult to obtain a smooth transition by hand sanding (Figure 7). Rather than risk damaging those coatings by feathering with a power tool, the decision was made to continue hand sanding and accept boundaries where the abrupt transition between the “existing” coating and prepared surface had been blended without a seamless transition. Consequently, the boundaries were visible once the panels received the “repair” coatings. However, after the tests were completed, the hand feathering was found to be acceptable. Surface preparation evaluations included chloride remediation by chemical treatment/washing on 48 Type I panels. The intent of this work was to determine efficacy of soluble salt remediation methods that did not require pressure washing and the generation of large quantities of wastewater. Two soluble salt remediation treatments were investigated. Both consisted of liquid chemical treatments that were applied on the SP 3 prepared surfaces using small hand sprayers. The surfaces were thoroughly wetted with the treatments. Then, they were hand scrubbed with a stiff 2- inch (5.08 cm) wide nylon bristle brush using four strokes to cover the exposed metal substrates. After a short waiting period, the surfaces were flushed with water from a wash bottle, applying sufficient water to rinse the treated surfaces. Those tests include panels with both initial blast cleaned and mill scale substrates along with epoxy/urethane “existing” coatings. They were coated with two liquid-applied “repair” coatings, a grease, and a tape. Figure 4. Test Matrix for Type I Panels

NCHRP Project 14-30 13 G. Candidate liquid-applied “repair” coatings/systems were applied by brushing (Figure 8). The KTC researchers reviewed the product data sheets and safety data sheets prior to painting. They observed mixing/agitation instructions, application surface preparation requirements, acceptable ambient conditions for application, wet film thicknesses, curing times and recoat times/windows. Figure 6. Feathering Exposed Edge of Existing Coating on Test Panel by Hand Sanding. Figure 5. Surface Preparation of Conditioned Panels to SSPC-SP 3, “Power Tool Cleaning,” Standard. Figure 7. Type I Flat Panel Specimen Surface Prepared to SSPC-SP 3 with Feathered Existing Coating.

NCHRP Project 14-30 14 The painting work was performed in the KTC paint shop. To perform the painting, 8 ft. (2.44 m) x 4 ft. (1.22 m) sheets of plywood were laid upon sets of saw horses and the panels. The floor of the shop was washed down to minimize the possibility of any airborne debris settling on the wet panels. Each group of panels being painted with the same system were grouped on the plywood sheets. H. Temperature gages were placed on several panels and atmospheric temperature and relative humidity readings were taken using a sling psychrometer to determine the dew point. When the surface temperature of the test panels was 5o F (2.8o C) above the dew point, the painting would commence. Ambient conditions were monitored several times throughout the painting process. The test panels were wiped down with lacquer thinner to prevent any problems resulting from handling the panels. The coatings were properly mixed and where needed for the epoxy, the proper induction time was observed. KTC technicians applied the coatings on the panels painting down-hand with the panels lying flat on the plywood sheets. They used 3-in. (7.62 cm) wide brushes applying the coatings by brushing across the 4-inch (10.16 cm) width of the panels. As the panels required a 1 in. (2.54 cm) strip of the existing coatings coating to remain uncoated, guide marks were placed on the existing coatings to prevent excess application of the spot coatings using indelible ink. Immediately after application of each coat, every panel in a group was inspected and tested with a tooth gage to measure the wet film thickness (WFT). If bristles were lodged in the wet paint or some other problem including improper WFT arose, the test panels were immediately repaired and re-brushed as necessary. The tooth gage marks were brushed out as well. After the first coat was applied, the panels were cured about 18 to 24 hours prior to receiving the finish coat. The only issue encountered was rust bleeding through the acrylic primer which was addressed by a second primer application. The grease and tape were applied shortly before the onset of the performance testing as they did not require curing. Prior to their application, the test panels were wiped with lacquer thinner. Electrician’s tape was placed around the edges of the test panels prior to the grease application to prevent rusting and rust bleed onto the specimen test surfaces (Figure 9). Application conditions were monitored as with liquid-applied coatings with applications being performed when the surface temperature of the specimens was 5o F (2.8o C) above the dew point to ensure that these coatings were not applied over moist substrates. The grease was applied with a spatula and the thickness measured with a tooth gage. The grease was applied at thicknesses between 20-40 mils (500-1,000 microns) with the wide range related to the difficulty in closely controlling the thickness with the spatula. In the field, grease applications have typically been less sophisticated with the grease slathered on bearings by hand resulting in variable coating thicknesses. The 34-mil (850 micron) thick tape was applied from a roll and cut-to-fit each test panel. The tape was self- Figure 8. KTC Technician Brushing a Spot Coating on a Test Panel.

NCHRP Project 14-30 15 bonding and readily attached to the existing coatings and the SSPC-SP 3 substrate. A small wallpaper roller was run over the applied tape to ensure that it adhered and that no air pockets were present at the tape/substrate interface. Edges around those test panels were covered with electrician’s tape to prevent rust bleed. The only problem with the tape application was that the test face of the tape smudged easily during application (Figure 10). However, that did not pose a performance problem during testing. I. After the liquid-applied coatings had cured sufficiently (dry to handle), dry film thicknesses (DFTs) of the “repair” coatings were measured on each panel per ASTM D7091, “Standard Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals.” For each panel, five readings were taken over the SSPC-SP 3 cleaned portion of the test panels - one near each corner Figure 9. Test Specimens with Grease “Repair” Coatings Ready for Testing. Figure 10. Tape Specimens Prepared for Testing (Note the Smudges on the Tape).

NCHRP Project 14-30 16 of the spot coating and one in the center. As was anticipated, the calcium sulfonate alkyd coating did not harden sufficiently to permit testing. KTC had to rely on previous WFT readings taken with the tooth gages during painting to ensure that coating was properly applied. After the panels were approved for testing, they were cured under ambient/room temperature conditions - 72o + 2o F (22.2o+1oC) for 28 days. The edges of the panels were covered with electrical tape and the panels were photographed to track their condition (Figure 11). The panels were not scribed as KTC researchers did not feel scribe creep was a valid test especially for barrier coatings. Experience showed that accelerated laboratory tests of barrier coatings yield poor scribe performance, yet those coatings worked well in the field on overcoating projects. J. On 144 panels, accelerated performance testing was conducted to replicate direct UV exposure with condensation/evaporation. Three panels of each combination of “existing” coating type/“repair” coating type were tested for 15 two-week-long cycles (5,040 hours) using ASTM D5894-10, “Standard Practice for Cyclic Salt Fog/UV Exposure of Painted Metal (Alternating Exposures in a Fog/Dry Cabinet and a UV/Condensation Cabinet).” The panels were placed in QUV chambers and subject to one-week (168 hours) of exposure in accordance with ASTM D 4587-11, “Practice for Fluorescent UV-Condensation Exposures of Paint and Related Coatings.” The QUV chambers were equipped with UVA 340 bulbs that replicated sunlight. Those were calibrated to operate at a normal irradiance (0.89 µm). Standard test panel mounting brackets were modified with the exposure windows removed. The panels were placed with “repair” coatings closest to the bulbs for maximum UV exposure. The QUVs were set to operate on a repetitive cycle of UV exposure at 140o F (60o C) for four hours alternated with a four-hour condensation cycle at 122o F (50o C) using potable water. Thereafter, the test panels were rotated to fog/dry chambers for a one week exposure per ASTM G 85. The test panels were mounted with the existing coatings end of the panels inserted in the chamber mounting brackets allowing the ”repair” coatings to receive the greatest exposure to the cyclic salt fog during the condensation cycle. The cyclic exposure consisted of one-hour of condensation of fog sprayed from a single nozzle located in the center of the chamber. That was followed Figure 11. Polyurethane over Epoxy Sealer “Repair” Coating System Tested over Test Panels with an Alkyd/Alkyd Existing Coating on a Test Panel.

NCHRP Project 14-30 17 by one hour of evaporation (drying) at 95o F (35o C) with moisture being exhausted from the chamber. Once the cyclic salt fog testing was completed, the panels were returned to the QUV chambers for further testing or temporarily removed for panel evaluation before testing continued. Each two-weeks of QUV-fog/dry exposures constituted a block of testing. K. At 3-cycle (1,008-hour) intervals, the panels were temporarily removed from the test chambers for nondestructive evaluation of “repair” coating performance and photo documentation. Coating performance was evaluated according to ASTM D610-08, “Standard Practice for Evaluating Degree of Rusting on Painted Steel Surfaces,” and ASTM D-714-02, “Standard Test Method for Evaluating Degree of Blistering of Paints.” Failure was set at a rust grade of 5 or less (1-3% without classifying type) and blistering at a “dense” rating. Those ratings were based upon visual references in the ASTM standards. Those levels were based upon the assumption that spot coatings would experience some level of deterioration during their service lives which would be acceptable as long as it did not expose the repair area to significant corrosion due to a large-scale coating breakdown. One technician performed all evaluations throughout the project to provide consistent ratings. After 5,040 hours of total exposure, the testing was complete and the final evaluations were performed. L. After the ASTM D5894 testing was completed, accelerated performance testing was conducted on the other 192 Type I panels to replicate high TOW/deicing salt exposure in sheltered locations. The panels were subjected to ASTM B117-11, “Standard Practice for Operating Salt Spray (Fog) Apparatus,” for 5,040 hours. That included exposure to a 5% aqueous salt solution continually fogged in the same chambers used for the ASTM D5894 testing. The testing was performed at 92o - 98oF (33.3o – 36.7o C)) during the test program measured by data loggers placed in the test chambers. The salt solution was replenished in the test chamber spray tanks on a weekly basis and periodic measurements were made of pH and spray volumes. M. The test panels were temporarily removed from the B117 test chambers at 6-week intervals (1,008-hours) for nondestructive evaluation of coating performance and photo documentation. As with the ASTM D5894 tests, coating performance was evaluated according to ASTM D 610-08 and ASTM D 714-02. After 5,040 hours of total exposure, the testing was complete and the final evaluations were performed (Figures 12 and 13). Unlike the ASTM D5894 testing, KTC researchers elected to remove test panel groups from the B117 testing once the panels in the group had failed in accordance with the KTC criteria. This was done to limit problems encountered in the test chambers due to rust deposits on some components in the chambers that were proving problematic.

NCHRP Project 14-30 18 Type II-Shaped Panel Tests – At the direction of the project review panel, KTC developed shaped panels to be used to evaluate spot coatings placed over irregular surfaces. The Type II shaped panels were made from two-6 in. (15.24 cm) x 4 in. (10.16 cm) x 1/4 in.(0.64 cm) flat plates also used for the Type I test panels. The plates were blast cleaned to SSPC-SP 5 (See Figure 14). An “existing” coating was applied to 2 in. (5.08 cm) of the test surface of each panel with the remainder of the surface left exposed. The panels were drilled to accept 1-in. (2.54 cm) diameter bolts/nuts meeting ASTM A325. The panels were lapped and bolted together. That created a faying surface and exposed edge on the top plate which were also part of the test features of the test panels. The test panels were conditioned with the same procedure used for the Type I panels Figure 12. Polyurethane over MIO-Epoxy “Repair” Coating System Tested over Test Panels with an Epoxy/Urethane Existing Coating on a Test Panel after 5,040 Hours of ASTM B117 Testing. Figure 13. Moisture Cure Polyurethane (2 Coats) “Repair” Coating System Tested over Test Panels with an Epoxy/Urethane Existing Coating on a Test Panel after 5,040 Hours of ASTM B117 Testing.

NCHRP Project 14-30 19 to rust the heads of the bolts and unpainted test surfaces. The rusted areas were subjected to mechanical surface preparation to the SSPC-SP 3 standard using a needle gun. Candidate “repair” coatings would be applied over the panel surface and subjected to accelerated performance testing per ASTM B117 to replicate spot coating performance on typical steel details under conditions of sheltered performance with high TOW/deicing salt exposures. The test matrix is shown in Figure 15. The actions for the Type II test protocol are as follows: A. Ninety-six - 6 in. (15.24 cm) x 4 in. (10.16 cm) x 1/4 in.(0.64 cm) steel plates were used to create 48 test panels for the Type II Tests. B. The test panels were blast cleaned to the NACE 1/SSPC-SP 5 “White Metal Blast Cleaning” standard. C. The epoxy/urethane coating was applied to 48 Type II panels each to 2 in. (5.08 cm) strips as shown in Figure 14 as “existing” coatings. D. The panels were subsequently conditioned per ASTM G85 Annex 5 for 250 hours of exposure to replicate spot coating failures, salt contamination and degradation of existing coatings on bridges. E. The panels were subsequently conditioned per ASTM G85 Annex 5 for 250 hours of exposure in accordance with ASTM D5894 to replicate spot coating failures, salt Figure 14. Accelerated Performance Type II Test Panel (0.25 in.-Thick Steel). Figure 15. Test Matrix for Type II Panels.

NCHRP Project 14-30 20 contamination and degradation of existing coatings on bridges. F. Mechanical surface preparation to an SSPC-SP 3 “Power Tool Cleaning” condition was performed on the rust exposed areas of the Type II panels using a needle gun (Figure 16). The surfaces were wiped with a cloth after cleaning to ensure that no loose material remained on the cleaned surfaces. G. Two-candidate conventional liquid- applied “repair” coatings were applied to the test panels by brushing using the procedure for the Type I test panels. The grease and tape were also used on those panels (Figure 17). The coatings were applied in accordance with Type I procedure G. After a 28-day of curing, spot coating thicknesses for each panel using liquid applied coatings were evaluated per ASTM D7091 “Standard Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals.” The panels were photo-documented for condition tracking. H. To replicate high TOW exposure in sheltered locations the Type II panels were subjected to ASTM B117-11 “Standard Practice for Operating Salt Spray (Fog) Apparatus” for 5,040 hours per Type I panel test protocol K. The initial intent had been to mount the panels in the test chambers with the edge of the outer panel facing upward. However, when condensate was observed ponding on the edge early in the testing, KTC researchers became concerned that this condition was more severe than anticipated and that it would result in early failures of the Type II panels. They elected to turn the panels with that edge facing downward to prevent the ponding. I. The panels were temporarily removed from the B117 test chamber at 6-week intervals (1,008-hours) for nondestructive evaluation of coating performance and photo documentation. As with the Type I B117 tests being performed concurrently with the Type II panels, coating performance was evaluated according to ASTM D 610-08 and ASTM D-714-02. The testing program ran 5,040 hours. Figure 16. Surface Preparation of Complex Shape (Type II) Test Specimen to SSPC-SP 3 Standard Using a Needle Gun.

NCHRP Project 14-30 21 2.3 Results of Accelerated Laboratory Performance Testing The results of the accelerated testing for the Type I flat panel specimens are provided in Tables 1-4 below. If no panels in a group failed after a 1,008- hour evaluation by rusting or blistering, that result was noted as a “0” indicating no failures up to that point. If one panel failed out of the three panels tested for each category, it would be noted as 1/3 meaning one of the three panels failed. After the 1/3 there will be a –R or a –B noting that the panel failed by rusting or blistering respectively. The Type I specimen results for ASTM D 5894 testing are provided in Tables 1 and 2. The Type I specimen results for ASTM B117testing are provided in Tables 3 and 4. The Type II specimen results for ASTM B117 testing are provided in Table 5. The test durations are shown rounded to the nearest 1,000 hours. For the Type I panels that used steel originally with a mill scale finish (Table 1), two of the liquid-applied coating “repair” systems passed the 5,000-hour (nominal) ASTM D5894 tests regardless of the “existing” coating system - the moisture cure urethane (2 coats) and the MIO- epoxy/polyurethane systems. Both the grease and the tape also passed the tests over all “existing” coating systems. The next best performing “repair” system was the epoxy sealer/polyurethane system that passed 3,000 (nominal) exposure hours for the acrylic and epoxy-polyurethane “existing” coatings and 4,000 hours for the alkyd “existing” coatings. In order of performance from best to worst the “repair” coating performances were: moisture cure urethane (2 coats) = MIO-epoxy/polyurethane = tape = grease > epoxy sealer/polyurethane > calcium sulfonate alkyd (2 coats) = alkyd (2 coats) > acrylic (2 coats). For the Type I panels that used steel originally with a blast cleaned finish (Table 2), one liquid-applied “repair” coating system, MIO-epoxy/polyurethane passed the 5,000-hour ASTM D5894 tests over all “existing” coatings systems. Both the grease and the tape also passed the tests over all “existing” coating systems. There was a near tie for the next best performing system - the epoxy sealer/polyurethane passed 4,000 exposure hours over the “existing” coatings and Figure 17. Grease Coated Type II Specimen Prepared for Accelerated Corrosion Testing per ASTM B117.

NCHRP Project 14-30 22 moisture cure urethane (2 coats) passed 5,000 hours over the acrylic “existing” coating; 4,000 hours over the alkyd “existing” coating; and 3,000 hours over the epoxy-polyurethane “existing” coating. In order of performance from best to worst the spot coating performances were: MIO- epoxy/polyurethane = tape = grease > moisture cure urethane (2 coats) = epoxy sealer/polyurethane > calcium sulfonate alkyd (2 coats) = alkyd (2 coats) > acrylic (2 coats). For the Type I panels that used steel originally with a mill scale finish (Table 3), one of the liquid-applied “repair” coating systems, the MIO-epoxy/polyurethane, passed the 5,000-hour ASTM B117 tests regardless of “existing” coating system. Both the grease and the tape also passed the tests for all “existing” coating systems (with and without chemical treatments). The next best performing “repair” system was the epoxy sealer/polyurethane which passed 5,000 exposure hours over the alkyd and epoxy-polyurethane “existing” coatings and 3,000 hours over the alkyd “existing” coating. The chemical soluble salt treatments did not show significant performance improvement in 3 of the 4 coatings tested. One of the treatments improved the performance of the moisture cure urethane “repair” system from 2,000 hours to 4,000 hours. In order of performance from best to worst the “repair” coating system performances were: MIO- epoxy/polyurethane = tape = grease > epoxy/polyurethane sealer > moisture cure urethane (2 coats) > calcium sulfonate alkyd (2 coats) = alkyd (2 coats) > acrylic (2 coats). For the Type I panels that used steel originally with a blast cleaned finish (Table 4), none of the liquid-applied “repair” coating systems passed the 5,000-hour ASTM B117 tests for all “existing” coating systems however, the MIO- epoxy/polyurethane “repair” system did when chemically treated to reduce soluble salts. Both the grease and the tape also passed the tests over all “existing” coating systems (with and without chemical treatments). The epoxy sealer/polyurethane passed 5,000 exposure hours over the epoxy-polyurethane “existing” coating system, but only 1,000 hours over the acrylic and alkyd “existing” coatings. The moisture cure urethane (2 coats) “repair” coating system lasted only 2,000 hours over the acrylic “existing” coatings but 1,000 hours over the alkyd and epoxy-polyurethane “existing” coatings. The chemical soluble salt treatments did show significant performance improvement in the moisture cure urethane and over MIO-epoxy polyurethane “repair” systems which were tested over the epoxy- polyurethane “existing” coating. The chemical treatments provided increased durability from 1,000 to 2,000 hours compared to tests over untreated surfaces. In order of performance from best to worst the “repair” coating performances were: MIO-epoxy/ polyurethane (treated) = tape (all) = grease (all) > MIO-epoxy/polyurethane (untreated) > epoxy sealer/polyurethane > moisture cure urethane (2 coats-treated) > calcium sulfonate alkyd (2 coats) = moisture cure urethane (2 coats- untreated) > alkyd (2 coats) > acrylic (2 coats). For the Type II panels that used steel originally with a blast cleaned finish (Table 5), none of the liquid-applied “repair” coating systems passed the 5,000-hour ASTM B117 tests. Both the grease and the tape passed the tests for all “existing” coating systems. The next best performing liquid-applied “repair” coating system was the MIO-epoxy/polyurethane that passed 3,000 exposure hours for the alkyd and epoxy-polyurethane “existing” coatings. None of the other liquid- applied “repair” coatings lasted more than 2,000 hours. In order of performance from best to worst the spot coating performances were: tape = grease > MIO-epoxy/polyurethane > calcium sulfonate alkyd (2 coats) > epoxy sealer/polyurethane > moisture cure urethane (2 coats) = alkyd (2 coats) > acrylic (2 coats).

NCHRP Project 14-30 23 Table 1. Type I Test Results for Mill Scale Plates Tested for 5000 Hours per ASTM D5894 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs “Existing” Coating System||“Repair” Coating System Acrylic||Calcium Sulfonate Alkyd (2 - Coats) 0 1/3-R 3/3 - R -------- -------- Acrylic||Moisture Cure Urethane (2 - Coats) 0 0 0 0 0 Acrylic|| Epoxy Sealer Polyurethane 0 0 0 1/3 - R 2/3 - R Acrylic|| MIO-Epoxy Polyurethane 0 0 0 0 0 Acrylic||Acrylic (2 - Coats) 1/3 - R 3/3 - B, 1/3 - R -------- -------- -------- Acrylic||Alkyd (2 - Coats) 0 0 1/3 - R 2/3 - R 2/3 - R Acrylic||Tape 0 0 0 0 0 Acrylic||Grease 0 0 0 0 0 Alkyd||Calcium Sulfonate Alkyd (2 - Coats) 0 0 3/3 - R -------- -------- Alkyd||Moisture Cure Urethane (2 - Coats) 0 0 0 0 0 Alkyd|| Epoxy Sealer/Polyurethane 0 0 0 0 1/3 - R Alkyd|| MIO-Epoxy /Polyurethane 0 0 0 0 0 Alkyd||Acrylic (2 - Coats) 3/3 - R 3/3 - B, 3/3 - R -------- -------- -------- Alkyd||Alkyd (2 - Coats) 0 0 2/3 - R 3/3 - R 3/3 - R Alkyd||Tape 0 0 0 0 0 Alkyd||Grease 0 0 0 0 0 Epoxy Polyurethane||Calcium Sulfonate Alkyd (2 - Coats) 0 3/3-R -------- -------- -------- Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats) 0 0 0 0 0 Epoxy Polyurethane|| Epoxy Sealer/Polyurethane 0 0 0 1/3 - R 1/3 - R Epoxy Polyurethane|| MIO-Epoxy/ Polyurethane 0 0 0 0 0 Epoxy Polyurethane||Acrylic (2 - Coats) 1/3 - R 3/3 - B, 1/3 - R -------- -------- -------- Epoxy Polyurethane||Alkyd (2 - Coats) 1/3 - R 1/3 - R 2/3 - R 3/3 - R -------- Epoxy Polyurethane||Tape 0 0 0 0 0 Epoxy Polyurethane||Grease 0 0 0 0 0 *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering.

NCHRP Project 14-30 24 Table 2. Type I Test Results for Blast Cleaned Plates Tested for 5000 Hours per ASTM D5894 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs “Existing” Coating System||“Repair” Coating System Acrylic||Calcium Sulfonate Alkyd (2 - Coats) 0 0 3||3 - R -------- -------- Acrylic||Moisture Cure Polyurethane (2 - Coats) 0 0 0 0 0 Acrylic|| Epoxy Sealer/Polyurethane 0 0 0 0 2||3 - R Acrylic|| MIO-Epoxy/Polyurethane 0 0 0 0 0 Acrylic||Acrylic (2 - Coats) 2||3 - R 3/3 - B, 2/3 - R -------- -------- -------- Acrylic||Alkyd (2 - Coats) 0 0 1/3 - R 1/3 - R 2/3 - R Acrylic||Tape 0 0 0 0 0 Acrylic||Grease 0 0 0 0 0 Alkyd||Calcium Sulfonate Alkyd (2 - Coats) 0 0 3/3 - R -------- -------- Alkyd||Moisture Cure Urethane (2 - Coats) 0 0 0 0 1/3 - R Alkyd|| Epoxy Sealer/Polyurethane 0 0 0 0 1/3 - R Alkyd|| MIO-Epoxy /Polyurethane 0 0 0 0 0 Alkyd||Acrylic (2 - Coats) 3/3 - R 1/3 - B, 3/3 - R -------- -------- -------- Alkyd||Alkyd (2 - Coats) 0 0 0 1/3 - R 3/3 - R Alkyd||Tape 0 0 0 0 0 Alkyd||Grease 0 0 0 0 0 Epoxy Polyurethane||Calcium Sulfonate Alkyd (2 - Coats) 0 0 0 0 3/3 - R Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats) 0 0 0 1/3 - R 2/3 - R Epoxy Polyurethane|| Epoxy Sealer/Polyurethane 0 0 0 0 1/3 - R Epoxy Polyurethane|| MIO-Epoxy /Polyurethane 0 0 0 0 0 Epoxy Polyurethane||Acrylic (2 - Coats) 2/3 - R 1/3 - B, 2/3 - R 1/3 - B, 2/3 - R 2/3 - B, 2/3 - R 3/3 - B, 2/3 - R Epoxy Polyurethane||Alkyd (2 - Coats) 0 0 0 2/3 - R 3/3 - R Epoxy Polyurethane||Tape 0 0 0 0 0 Epoxy Polyurethane||Grease 0 0 0 0 0 *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering.

NCHRP Project 14-30 25 *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering. Table 3. Type I Test Results for Mill Scale Plates Tested for 5000 Hours per ASTM B117 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs “Existing” Coating System||“Repair” Coating System Acrylic||Calcium Sulfonate Alkyd (2 - Coats) 0 1/3-R 3/3 - R -------- -------- Acrylic||Moisture Cure Urethane (2 - Coats) 0 1/3 - B, 2/3 - R 1/3 - B, 3/3 - R -------- -------- Acrylic|| Epoxy Sealer/Polyurethane 0 0 0 2/3 - R 3/3 - B, 3/3 - R Acrylic|| MIO-Epoxy /Polyurethane 0 0 0 0 0 Acrylic||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- -------- -------- Acrylic||Alkyd (2 - Coats) 0 1/3 - R 3/3 - R -------- -------- Acrylic||Tape 0 0 0 0 0 Acrylic||Grease 0 0 0 0 0 Alkyd||Calcium Sulfonate Alkyd (2 - Coats) 0 0 3/3 - R -------- -------- Alkyd||Moisture Cure Urethane (2 - Coats) 0 0 3/3 - R -------- -------- Alkyd|| Epoxy Sealer/Polyurethane 0 0 0 0 0 Alkyd|| MIO-Epoxy/ Polyurethane 0 0 0 0 0 Alkyd||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- -------- -------- Alkyd||Alkyd (2 - Coats) 0 0 3/3 R -------- -------- Alkyd||Tape 0 0 0 0 0 Alkyd||Grease 0 0 0 0 0 Epoxy Polyurethane||Tape 0 0 0 0 0 Epoxy Polyurethane||Tape (Treatment 1) 0 0 0 0 0 Epoxy Polyurethane||Tape (Treatment 2) 0 0 0 0 0 Epoxy Polyurethane||Grease 0 0 0 0 0 Epoxy Polyurethane||Grease (Treatment 1) 0 0 0 0 0 Epoxy Polyurethane||Grease (Treatment 2) 0 0 0 0 0

NCHRP Project 14-30 26 Table 3 Cont. Type I Test Results for Mill Scale Plates Tested for 5000 Hours per ASTM B117 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs “Existing” Coating System||“Repair” Coating System Epoxy Polyurethane||Calcium Sulfonate Alkyd (2 - Coats) 0 2/3-R 3/3 - R -------- -------- Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats) 0 0 1/3 - B 1/3 - R 3/3 - B, 2/3 - R Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats, Treatment 1) 0 0 1/3 - R 1/3 - R 3/3 - B, 2/3 - R Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats, Treatment 2) 0 0 0 0 3/3 - B, 3/3 - R Epoxy Polyurethane|| Epoxy Sealer/Polyurethane 0 0 0 0 0 Epoxy Polyurethane|| MIO-Epoxy /Polyurethane 0 0 0 0 0 Epoxy Polyurethane|| MIO- Epoxy/Polyurethane (Treatment 1) 0 0 0 0 0 Epoxy Polyurethane|| MIO- Epoxy/Polyurethane (Treatment 2) 0 0 0 0 0 Epoxy Polyurethane||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- ------- -------- Epoxy Polyurethane||Alkyd (2 - Coats) 0 1/3 - R 3/3 - R -------- -------- *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering.

NCHRP Project 14-30 27 Table 4. Type I Test Results for Blast-Cleaned Plates Tested for 5000 Hours per ASTM B117 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs Existing Coating System||“Repair” Coating System Acrylic||Calcium Sulfonate Alkyd (2 - Coats) 0 2/3 - R 3/3 - R -------- -------- Acrylic||Moisture Cure Urethane (2 - Coats) 0 0 3/3 - R -------- -------- Acrylic|| Epoxy Sealer/Polyurethane 0 1/3 - R 2/3 - B, 2/3 - R 2/3 - B, 3/3 - R -------- Acrylic|| MIO-Epoxy/Polyurethane 0 0 0 0 0 Acrylic||Acrylic (2 - Coats) Water Based 3/3 - B, 3/3 - R -------- -------- -------- -------- Acrylic||Alkyd (2 - Coats) 2/3 - B, 2/3 - R 3/3 - B, 3/3 - R -------- -------- -------- Acrylic||Tape 0 0 0 0 0 Acrylic||Grease 0 0 0 0 0 Alkyd||Calcium Sulfonate Alkyd (2 - Coats) 0 2/3 - R 3/3 - R -------- -------- Alkyd||Moisture Cure Urethane (2 - Coats) 0 1/3 - R 3/3 - R -------- -------- Alkyd|| Epoxy Sealer/Polyurethane 0 1/3 - B 1/3 - B, 1/3 - R 2/3 - B, 1/3 - R 2/3 - B, 2/3 R Alkyd|| MIO-Epoxy/Polyurethane 0 0 0 0 0 Alkyd||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- -------- -------- Alkyd||Alkyd (2 - Coats) 0 0 3||3 - R -------- -------- Alkyd||Tape 0 0 0 0 0 Alkyd||Grease 0 0 0 0 0 *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering.

NCHRP Project 14-30 28 Table 4 Cont. Type I Test Results for Blast-Cleaned Plates Tested for 5000 Hours per ASTM B117 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs Existing Coating System||“Repair” Coating System Epoxy Polyurethane||Calcium Sulfonate Alkyd (2 - Coats) 0 0 3/3 - R -------- -------- Epoxy Polyurethane||Moisture Cure Polyurethane (2 - Coats) 0 1/3 - B 1/3 - B 2/3 - B, 2/3 - R 3/3 - B, 2/3 - R Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats, Treatment 1) 0 0 0 2/3 - R 2/3 - R Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats, Treatment 2) 0 0 1/3 - R 3/3 - R -------- Epoxy Polyurethane || Epoxy Sealer/Polyurethane 0 0 0 0 0 Epoxy Polyurethane|| MIO- Epoxy/Polyurethane 0 0 0 1/3 - R 1/3 - R Epoxy Polyurethane|| MIO- Epoxy/Polyurethane (Treatment 1) 0 0 0 0 0 Epoxy Polyurethane|| MIO- Epoxy/Polyurethane (Treatment 2) 0 0 0 0 0 Epoxy Polyurethane||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- -------- -------- Epoxy Polyurethane||Alkyd (2 - Coats) 0 3/3 - B -------- -------- -------- Epoxy Polyurethane||Tape 0 0 0 0 0 Epoxy Urethane||Tape (Treatment 1) 0 0 0 0 0 Epoxy Urethane||Tape (Treatment 2) 0 0 0 0 0 Epoxy Urethane||Grease 0 0 0 0 0 Epoxy Urethane||Grease (Treatment 1) 0 0 0 0 0 Epoxy Urethane||Grease (Treatment 2) 0 0 0 0 0 *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering.

NCHRP Project 14-30 29 *0 denotes no failures at specific test durations. 1/3 through 3/3 denote one to three of three panels in a category have failed based upon KTC criteria. – R denotes failure due to rusting. –B denotes failure due to blistering. 2.4 Laboratory Coating Testing Discussion Tests using the Type I flat test panels were conducted without any problems related to the basic specimen design or test protocol. The aforementioned problem with the Type II test specimens ponding of water on the edge created by lapping the two panels and bolting them together was detected early in the testing. It was considered problematic as it constituted an immersion situation which SSPC-SP3 surface preparation was not designed to address. As noted, those panels were inverted to avoid premature failures. In retrospect, the situation might have been also addressed by tilting the specimens in the mounts, but that might introduce another a test variable that would be difficult to account for. If the Type II tests are to be used in the future, the top lapping plate should have its upper edge beveled or cut at an angle to prevent ponding on the edge. The accelerated laboratory performance testing for spot coatings has produced results Table 5. Type II Test Results for Blast-Cleaned Plates Tested for 5000 Hours per ASTM B117 Test Durations in Hours vs. Performance* Test Duration 1000 Hrs 2000 Hrs 3000 Hrs 4000 Hrs 5000 Hrs Existing Coating System||||“Repair” Coating System Alkyd||Calcium Sulfonate Alkyd (2 - Coats) 0 0 3/3 - R -------- -------- Alkyd||Moisture Cure Urethane (2 - Coats) 0 3/3 - R -------- -------- -------- Alkyd|| Epoxy Sealer/Polyurethane 0 3/3 - B -------- -------- -------- Alkyd|| MIO-Epoxy /Polyurethane 0 0 0 1/3 - R 3/3 - R Alkyd||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- -------- -------- Alkyd||Alkyd (2 - Coats) 0 2/3 - B, 2/3 - R 3/3 - B, 3/3 - R -------- -------- Alkyd||Tape 0 0 0 0 0 Alkyd||Grease 0 0 0 0 0 Epoxy Polyurethane||Calcium Sulfonate Alkyd (2 - Coats) 0 1/3 - R 3/3 - R -------- -------- Epoxy Polyurethane||Moisture Cure Urethane (2 - Coats) 0 3/3 - R -------- -------- -------- Epoxy Polyurethane|| Epoxy Sealer/Polyurethane 0 0 3/3 - R -------- -------- Epoxy Polyurethane|| MIO-Epoxy /Polyurethane 0 0 0 2/3 - R 3/3 - R Epoxy Polyurethane||Acrylic (2 - Coats) 3/3 - B, 3/3 - R -------- -------- -------- -------- Epoxy Polyurethane||Alkyd (2 - Coats) 0 3/3 - B, 3/3 - R -------- -------- -------- Epoxy Polyurethane||Tape 0 0 0 0 0 Epoxy Polyurethane||Grease 0 0 0 0 0

NCHRP Project 14-30 30 that were anticipated by KTC researchers based upon published field and laboratory performance tests of those generic types of coatings and KTC experience with experimental overcoating projects on over 200 bridges and extensive in-house laboratory testing. The liquid-applied coating rankings were especially instructive. The moisture cure polyurethane (2 coat) “repair” system performed well in the ASTM D5894 accelerated weathering/corrosion testing replicating bold exposure service. It performed worse in the ASTM B117 test that replicated sheltered bridge exposures with high TOW and soluble salt contamination. That reflects past KTC findings on many overcoating projects using aluminum-pigmented moisture cure primers and intermediate coatings. The best overall performing liquid-applied “repair” coating was the over MIO- epoxy/polyurethane. For spot painting (touch-up), the recently developed MnDOT Maintenance Manual (2017) specifies the use of a coat of polyamide epoxy as a one-coat spot repair or a polyurethane over a polyamide epoxy for a two-coat spot repair, both applied over hand- or power tool-prepared substrates with feathering of the existing coatings. Epoxy coatings are commonly used by state highway agencies as repair coatings over marginally prepared steel substrates. The epoxy sealer/polyurethane “repair” coating performed well despite being thin 1-2 mil (25-50 microns) and not providing significant barrier protection. The polyurethane top coat was expected to provide weathering resistance, but contribute little to the corrosion protection. The epoxy sealer contained inhibitors which may have been beneficial to the system performance The surface preparation of the Type II specimens was problematic due to the poor performance of the liquid-applied coatings. That was probably due to the needle gun surface preparation causing salt contaminated residue to be embedded in the steel surface prior to painting resulting in premature failures. In retrospect, it would have been desirable to use power wire brush cleaning to remove as much salt-contaminated rust as possible on the Type II substrates prior to using the needle gun. That issue should be noted when using a needle gun to clean complex rusted details on bridges. The use of rotary power tools for cleaning prior to using impact tools to create profiles is noted in SSPC-SP 11 and SSPC-SP 15. As anticipated, most coatings that relied on inhibitive properties (alkyds and acrylics) did not perform as well overall as the barrier-type coatings (moisture cure urethanes and epoxy/polyurethanes). That was due to the soluble salt contamination that remained on the test panels prior to painting. When calcium-sulfonate alkyds are subjected to accelerated laboratory testing over salt-free steel substrates, they typically perform well - often better than three-coat systems employing zinc primers. The conventional alkyd did not perform as well as the calcium sulfonate alkyd for most KTC tests. The poorest performer was the acrylic coating system. During application in both the laboratory and the field, the primer showed rust bleed through. In every test, the acrylic failed in less than 1,000 hours. The tape and grease performed excellently providing protection throughout all of the Type I and Type II tests. When several samples of each were stripped of coating after the tests were completed, the existing SSPC-SP 3 surfaces were present. The substrates of the tape-covered test panels were found to be black indicating the presence of Fe3O4 rust, which is a stable corrosion layer produced by corrosion in a low oxygen environment. The performance of those non-traditional coatings reflects previous findings of KTC researchers on experimental field projects. Several observations were made during the tests. There were no incompatibilities between the “existing” coatings and the experimental “repair” coating systems during the tests. Indeed, where the “repair” coatings lapped the “existing” coatings, they provided protection in cases where the exposed portions of the “existing” coatings were beginning to fail towards the end of some tests. There did not appear to be discernable differences between the test panels conditioned with either the mill scale or blast-cleaned substrates in terms of soluble salt contamination or follow-on performance of the coated panels. While the conditioning step is

NCHRP Project 14-30 31 considered important in obtaining good correlations between candidate “repair” coating systems, it is likely that one of those substrate options can be eliminated – preferably the mill scale surface. A problem encountered in the laboratory tests was rust bleed from the taped edges of the test panels. Apparently, the edges of the test panels should have been better protected by coating them prior to applying the experimental spot coatings. The liquid-applied “repair” coatings all performed poorly on the Type II panels. The best performing of those, the MIO-epoxy/polyurethane lasted 3,000 hours. Those had surface preparation in the rusted areas by use of needle guns. KTC did not measure soluble salt contamination on needle gun-treated surfaces and it may have been considerably higher than that obtained by power wire brushing on the Type I panels. That might have been the cause of the poor performance of those coatings. The good performance of the grease and tape on the Type II specimens probably indicates that they are more tolerant of soluble salt contamination than any of the six liquid-applied “repair” coatings. Several of the accelerated corrosion tests involving liquid-applied coatings revealed that the low-water-use soluble salt remediation treatments employed by KTC could provide tangible improvements in coating performance when used over mildly contaminated SSPC-SP 3 substrates. The non-traditional coatings, the grease and tape performed well in the laboratory tests, passing the 5,000-hour protocols for both the Type I and Type II test panels for both the ASTM D 5894 and B117 tests.