Prediction of service life for naturally exposed coatings remains problematic. A number of factors can have an effect on coating performance, but none more so than surface preparation, the primary factor that influences how long a coating will last in service. A method of surface preparation must be selected based on coating type, the intended exposure and the desired life cycle. Poor, inadequate or improper surface preparation can lead to premature failure of a coating, sometimes catastrophically. The U.S. Naval Research Laboratory (NRL) conducted a study of various surface preparation methods to determine their effect on the common topside epoxy primer barrier coatings used by the U.S. Navy. The independent variables that were changed included type of profile (both the extent of cleanliness and the type of tool used), profile height, coating thickness and the amount of chloride contamination present. The dependent variable in each case was coating performance, which was evaluated either visually by rust-through, scribe cutback (i.e. scribe creep, the extent of corrosion that undercuts the coating at an intentional holiday), blistering or adhesion of the coating.
As seen in the July 2016 Edition of the Journal of Protective Coatings & Linings (JPCL) and on Paintsquare.com
TEST APPROACH
The Effect of Profile Type Three mechanical preparation methods were tested side-by-side on a 12-inch-square steel panel. The panel was sectioned into three equal portions and each portion was prepared by the method indicated in Table 1.
TABLE 1: SURFACE PROFILE TYPE
| Suface Profile No. | Suface Prep Type | Target Profile Height |
| 1 | SSPC-SP 10 Near White BlastCleaning | 2–3 mils |
| 2 | SSPC-SP 11 Power Tool Cleaningto Bare Metal (Needle Gun) | 1–1.5 mils |
| 3 | Power Tool Cleaning to Bare Metal(Power Wire Brush) | 0.5–1 mil |
The three different surface preparation methods were intended to simulate field conditions in a laboratory environment. One aspect of field maintenance conditions that the laboratory test panels did not simulate was the presence of aged coatings, corrosion products and soluble salts. The laboratory steel panels used were virgin (SSPC-VIS 1 Condition A) and not pre-conditioned or coated, and the specified SSPC-SP 10/NACE No. 2, Near White Blast achieved a cleanliness level that approached or attained a SSPC-SP 5/NACE No. 1, White Metal Blast level. 
Fig. 1: This images shows the final surface profile types.
All figures courtesy of the authors unless otherwise noted.The power wire brush surface preparation sample was cleaned to bare metal but not identified as SSPC-SP 11, Power Tool Cleaning to Bare Metal, since SP 11 requires a 1 mil minimum profile. Not only were the profiles for each surface preparation method expected to be different, but it was expected that each method would also produce a different type of profile in terms of roughness and angularity, as seen in Figure 1. Profile measurements were taken on each section of the panel, recorded in accordance with ASTM D4417, Method B with a digital profilometer and were within the specification given. Nine different coating systems were applied to the panels — five of primer only (two coats of epoxy each) and four of primer and topcoat (two coats of epoxy and one coat of silicone alkyd/polysiloxane). These coating systems were intended to be a sampling of legacy epoxy, high-solids (HS) and ultra-high-solids (UHS) epoxies, the most common topside systems used by the Navy. 
Fig. 2: The final panel for exposure with scribes. Note the touched-up areas.All systems were spray-applied in accordance with the manufacturer’s instructions and checked for proper dry film thickness (DFT) with a digital coating thickness gage. Each coating system was applied to four replicate panels. Each panel was scribed twice across the face of the panel (Fig. 2). Baseline adhesion testing was performed on each section of each panel in accordance with ASTM D4541. Two panels were exposed to 1,000 hours of ASTM B117 salt fog and two panels were exposed to one year of atmospheric exposure in Key West, Fla. Upon completion of the exposures, the panels were rated for degree of rust-through, blistering, scribe cutback and adhesion using ASTM D610, ASTM D714, ASTM D1654 and ASTM D4541, respectively. 
Fig. 3: This image shows the excessive profile height panel.The Effect of Profile HeightSteel panels were procured and abrasive blasted in accordance with SSPC-SP 10. The target profile was 10 mils as measured via ASTM D4417, Method B. This excessive profile and the subsequent coating performance could then be compared to that of the normal profile height from the panels described in the previous section of testing. A sample panel is shown in Figure 3. To mitigate the effect of the excessive profile height, four coating systems were applied at four different wet film thicknesses (WFTs). Due to the roughness of the profile, the WFT was measured using a witness (control) panel. The four coating systems were all epoxy primers commonly used by the U.S. Navy. The target WFTs over the 10-mil profile were 10 mils, 11 mils, 12 mils and 14 mils. All coatings were spray-applied in accordance with the manufacturer’s instructions. 
Fig. 4: The final excessive profile height panel ready for exposure.The panels were scribed and exposed in a salt fog cabinet in accordance with ASTM B117 for 1,000 hours (Fig. 4). Upon completion of the exposures, the panels were rated for degree of rust- through, blistering and scribe cutback using ASTM D610, ASTM D714, and ASTM D1654, respectively. 
Fig. 5: Hand tool prepared panel ready for exposure in Ft. Lauderdale, Fla.The Effect of ContaminationBare steel panels were exposed to 10 cycles using the GMW14872 Cyclic Corrosion Laboratory Test. The panels were then prepared to SSPC-SP 11 using a power sanding disc with 100-grit sandpaper with a target profile height of 1 mil. The residual contamination was characterized by measuring conductivity on each panel. All panels were coated with the same epoxy primer in accordance with the manufacturer’s instructions. Panels were then subjected to either 100 cycles of GMW14872 or six months of natural exposure in Fort Lauderdale, Fla. (Fig. 5). Upon completion of the exposures, the panels were rated for degree of rust-through and blistering using ASTM D610 and ASTM D714, respectively.
RESULTS
The Effect of Profile TypeAdhesion data was taken on each panel before and after exposure over each different type of surface preparation. The results are shown in Figures 6, 7 and 8. All of the systems showed adequate adhesion after exposure. The adhesion tests performed provided no meaningful results pertaining to coating performance other than showing which coatings were the weakest to begin with. For example, System 6 was a silicone alkyd topcoat and its mode of failure was disbondment between the topcoat and primer. There was no correlation between adhesion and type of surface preparation/profile before or after exposure. 
Fig. 6: Pre-exposure adhesion test results.
Fig. 7: Post-exposure adhesion results after 1,000 hours as per ASTM B 117.
Fig. 8: Post-exposure adhesion results after one year in Key West, Fla.Ratings taken on each panel for blistering and rust-through also provided no results pertaining to the difference in performance. However, cutback from the scribe was measured for each coating and surface preparation type and provided meaningful results. Figures 9 and Figure 10 illustrate accelerated exposure and Figures 11 and 12 illustrate atmospheric exposure. Figure 13 shows three sample panels after one year of natural atmospheric exposure on various coating types.
Fig. 9: Cutback data by coating system after 1,000 hours as per ASTM B117.
Fig. 10: Cutback data by surface preparation, all coating systems after 1,000 hours as per ASTM B117.
Fig. 11: Cutback data by coating system after one year of natural exposure.
Fig. 12: Cutback data by surface preparation method, all coating systems after one year of natural atmospheric exposure.
Fig. 13: Three sample panels after removal from one year of natural atmospheric exposure. Surface preparation methods were (left to right) power wire brush, needle gun and abrasive blast. Some coating systems performed better than others over an SSPC-SP 11 surface. Note the massive delamination of the coating over the left two-thirds of the center panel. No trend was recognized as to why this occurred. 
Fig. 14: Excessive profile panels after 1,000 hours as per ASTM B117. Film thickness increases (left to right) at 10 mils WFT, then 11 mils, then 12 mils and on the far right, 14 mils WFT.The Effect of Profile HeightThe meaningful data collected for the excessive profile panels included rust-through and blistering. Figure 14 shows one coating system set after accelerated exposure. Note that as coating film thickness increased, rust-through and blistering decreased. The results of the rust-through ratings separated by primer thickness are shown in Figure 15. A rating of 10 is no rust-through. The profile differential is the WFT converted to DFT with the average surface profile height subtracted (-1 mil = 10 mil WFT*90% solids – 10 mil profile). The data and visual inspection both show that as primer film thickness increases above the profile height, rust-through decreases. Blister rating data is shown in Figure 16. This rating number is a composite numerical representation of the ASTM D714 density and size rating that allows the rating to be shown graphically. (A rating of 10 is no blistering.)
Fig. 15: Average rust-through rating by primer thickness.
Figure 16: Average blister rating by primer thickness. The scribe cutback data showed no undercutting of the coating at any film thickness, and thus the data is not shown. Note that in the first part of testing presented, cutback was the primary method of coating failure, with little to no rust-through or blistering, while in this portion of testing, rust-through and blistering were the primary modes of failure and there was little to no cutback observed. The Effect of ContaminationPanels were rated after exposure for performance by rust-through and blistering. Figure 17 shows the average rating of all panels (both accelerated and atmospheric exposure) by conductivity. Conductivity was grouped as either high salts (>70 µS/cm) or low salts (<70 µS/cm) based on the U.S. Navy standard for topside conductivity allowance in Standard Item 009-32. Again, the blister rating number is a composite numerical representation of the ASTM D714 density and size rating that allows the rating to be shown graphically.
Figure 17: Final results of coating performance for all exposures (100 cycles GMW14872/six months natural atmospheric exposure, Ft. Lauderdale, Fla.).
CONCLUSIONS
Differences in surface prep type and the extent of cleanliness did not seem to impact adhesion of the coating or the rust-through/blistering but did affect the cutback performance across all coating systems. Cutback over an abrasive blasted SSPC-SP 10 surface was lower on average than that of a needle gunned, SSPC-SP 11 surface, which in turn was lower than that of a power wire brushed SSPC-SP 11 surface. It appears that both cleanliness and the profile type (including height and angularity) provide for better cutback prevention. Adhesion performance seems to be primarily influenced by the coating type (and its cohesive strength) and not by the surface preparation variables. Additionally, some individual coating systems tended to perform better than others over different types of surface preparation. The second portion of testing of excessive profiles showed that a higher profile will even further limit cutback. The abrasive blasted panels with a 10-mil profile height in the second portion of testing outperformed abrasive blasted panels with a 2-to-3-mil profile height from the first portion of testing. However, the risk associated with going to a profile too high is rust-through and blistering of the coating system if not applied at a high enough film thickness. The coating thickness must be applied several mils higher than the height of the surface profile to prevent premature breakdown of the coating. However, this practice, if done incorrectly, may lead to other problems such as solvent entrapment. An effective balance between abrasive blast profile height and coating thickness must be achieved. Finally, coating performance is affected by chloride contamination. On average, panels with a conductivity above the U.S. Navy acceptable level for topside coatings showed worse overall performance than those with a conductivity level below the acceptable level. These factors should be taken into account when selecting the proper surface preparation method for a job and should be weighed against the applicable cost of the surface preparation and intended lifecycle requirements.
ABOUT THE AUTHORS
Patrick Cassidy has been working in the corrosion and coatings industry for over eight years and is currently a senior engineer with Elzly Technology Corporation. He has been involved in a diverse number of programs including coatings research, field investigation and application of corrosion control products. He holds a Bachelor of Science degree in mechanical engineering from the University of Virginia. Cassidy is an SSPC-certified NAVSEA Coatings Inspector and has completed additional training in Navy Ship Corrosion Assessment and Cathodic Protection Design. In 2015, he was profiled in the JPCL annual bonus issue, Coatings Professionals: The Next Generation.
Paul Slebodnick is employed by the Naval Research Laboratory in the Washington D.C., Center for Corrosion Science & Engineering, under the marine engineering Section. He currently leads research programs in developing technologies for the United States Navy that produce maintenance reductions and reduce Ships Force workload. Slebodnick is responsible for demonstrating new technologies aboard Fleet combatants to determine readiness with in-service evaluation of the technologies prior to transitioning to the Fleet. He also leads Engineering for Research and Development of Tank Coatings under Naval Sea Systems Command, Technical Warrant Holder for Coatings and Corrosion Control — Ships, SEA-05P in Washington D.C.
James Tagert is a materials research engineer working at the Naval Research Laboratory and has over 10 years of experience working in the coatings industry. He graduated from the University of Maryland in 2004 earning a Bachelor of Science degree in mechanical engineering and is a member of both the American Society of Naval Engineers and National Association of Corrosion Engineers. Tagert has worked at NRL since 2008 supporting U.S. Navy research and engineering programs related to materials science with an emphasis on the development and transition of advanced coating systems.
James Martin has been with the Naval Research Laboratory for over 16 years. He is the head of the Marine Coatings Technology and Systems section Code 6138. Martin is responsible for introducing coatings technology to the Fleet through applied research and development, testing and demonstrations. He has been active in addressing Fleet concerns from both maintenance and new construction with respect to coatings. Martin continues to introduce new technology that will help to reduce the life cycle and ownership costs of today’s Fleet.
ACKNOWLEDGMENTS
This material is based upon work supported by Naval Sea Systems Command (NAVSEA) SEA 05 Painting Center of Excellence (PCoE). The work described was performed by the Center for Corrosion Science and Engineering (CCSE), Code 6130 of the Chemistry Division at the Naval Research Laboratory (NRL), Washington D.C. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of NAVSEA.