Quantifying Paint System Performance

How do you know the paint you just bought will actually meet the manufacturer's claims? Because there is no industry standard for ranking or evaluating paint systems, you don't.

The U.S. Army Tank Automotive and Armaments Command is in the process of leveraging commercial technology to improve military coating systems. Innovative approaches with respect to system design and evaluation are necessary to provide the military with a vehicle that lasts longer and requires less maintenance.

We have found that the automotive industry has done an exceptional job developing highly durable coating systems and the test methods to qualify them. In the automotive industry, corrosion and chip resistance are often taken for granted because of the use of galvanized steel, plastic, aluminum and anti-chip coatings. While automotive manufacturers have recognized the value of multi-coating paint systems in conjunction with more corrosion resistant substrates, for the most part, general industry and the military have not made this quantum change. For these industries, total system performance is driven by the pretreatment, or lack thereof, primer and topcoat.

Although some companies have accepted accelerated corrosion testing as a means to establish the durability of their paint/pretreatment system, the durability of an epoxy-based system, based upon accelerated corrosion testing alone, can be extremely optimistic if other safeguards are not exercised.

In the last few years, this has become a more serious issue with the drive to reduce VOCs. Lower VOC primers have higher viscosities and are less efficient as corrosion inhibitors. Consequently, they tend to require greater film thickness to get the same performance level as their higher VOC versions. Although a thicker epoxy provides improved corrosion resistance, it is more brittle. Considering the industry's move toward lower VOC coatings, a thorough evaluation of multi-coating systems, substrates and pretreatments is necessary.

Therefore, we took the automotive industry's proven methods and made some slight "improvements" to create a quantitative method for evaluating paint performance. Many of our new system contracts have paint and substrate corrosion performance defined in terms of GM 9540P cycles coupled with a gravelometer test for chip resistance and impact test for process control.

Cleaning & Pretreatment - The Variability Begins Cleaning and pretreatment are the foundation for the coating system. Therefore, some general insight into cleaning and pretreatment and the pretreatment methods we used in this study need to be discussed.

Although a cleaning or pretreatment process might perform well in laboratory conditions on flat test panels, the ability to properly clean and pretreat complex structures with multiple metals and surface conditions, corners, edges, cavities, welds and flash/heavy rust is another matter. Temperature control, concentration and contamination levels in the various recycled chemicals need constant monitoring to maintain minimum levels of performance.

Manufacturing processes, such as laser cutting, forming lubricants and corrosion inhibitors to retard flash rusting are sources of constant grief to the finisher. Laser cut edges resist most pretreatments, and some drawing compounds and corrosion inhibitive oils only come off with specific cleaners. Their removal becomes even more difficult if they have been on the surface for considerable periods of time. Welding processes, which deposit a glassy flux on the weld deposit, provide an unreactive surface for both the pretreatment and the primer.

Problems are often compounded when the finisher, due to a lack of manufacturing floor space, chooses to store parts outside with little or no corrosion inhibitors or water-soluble inhibitors. The general and galvanic corrosion cells created by this practice often result in tenacious corrosion products that require both mechanical and chemical intervention for complete removal.

The typical approach is to use an alkaline cleaner in conjunction with an abrasive blast, iron phosphate, zinc phosphate, chromate conversion, vinyl wash primer or organo-ceramic derivative pretreatment. But, chemical cleaning becomes increasingly more complicated when inorganic contaminants and mixed metals are involved. In addition, some cleaners have wetting agents, which augment the removal of certain contaminants but may hide organic contaminant detection by giving the visual impression of a clean water break. Fish eyes can be prevented with this practice, but overall long-term performance is compromised. There is no "easy" way to determine that the surface is free of all inorganic contaminants, since water break criteria do not apply. Adhesion failures in an accelerated corrosion test of adequate duration are often the only reliable means of contaminant detection.

Abrasive blasting is highly regarded by some as providing the optimal "pretreatment" for large platforms. There is no question that for quickly removing old coatings, corrosion and mill scale, various media and blast pressures can be employed to accomplish the task with minimal degradation to the substrate. Complications, however, do arise when the spent media, old coatings and metallic fines need to be contained or removed from the remainder of the structure to preclude primer contamination.

Unless the process is contained in an environmental booth, this process cannot be easily carried out in a manufacturing environment. Although water delivery systems help to contain the residue, a relatively expensive reverse osmosis water system is necessary to preclude the development of almost instantaneous flash rusting on ferrous surfaces using water sources with high mineral salt content.

However, the greatest deterrent to abrasive blasting is its inherently limited ability to develop chemical bonding between the substrate and primer compared to chemical pretreatments. Good mechanical bonding between the primer and substrate can be obtained if the primer is applied shortly after blasting to the proper surface profile.

Product data sheets, which identify a particular level of coating performance in conjunction with abrasive blasting, can be very misleading. Both surface profile and dwell time from blasting to primer application are critical variables. The difference in performance between a 2-hr dwell time (laboratory test coupon) and a 2-day dwell time (typical production) can be dramatic.

Iron phosphating does not meet Army acceptance standards as a pretreatment but is generally an effective cleaner. Numerous attempts have been made in the past to get various manufacturers' products to pass the military's 336-hr salt spray test on scribed panels with an epoxy primer. Not one has succeeded.

The polyvinyl butyral wash primers are recipe driven pretreatments dictated by DOD-P-15328. They contain hexavalent chromium and typically have a VOC of 6 lb/gal with 10% solids. These factors limit its use in some states. These are two-component products containing a phosphoric acid component. They are touted as adhesion promoters. They have no other performance benefits expressed or implied. They have been used on a variety of substrates but require that the film remain wet a minimum of 2 min to allow the acid component to have maximum effectiveness. Furthermore, the wash primer must cure a minimum of 1 hr before primer application (MIL-C-8507). There have been reports that the chromium component leaches through the paint film and has been detected in the drip line of vehicles. This product has been used by numerous manufacturers and is allowable on many Army systems.

Aside from its significant environmental problems, wash primers suffer from two major weaknesses associated with their application:

1. Coating systems that rely on a "paint" for a pretreatment typically have very poor to minimal cleaning practices associated with them. Chemical pretreatment systems have a controlled cleaning process directly preceding them to assure a reactive surface in conjunction with sufficient rinsing to eliminate chemical cross contamination.
2. Applicators have a tendency to "dry spray" the wash primer, apply wet-on-wet with the primer or apply the incorrect film thickness. All degrade the usefulness of the product.

Zinc phosphating and organo-ceramic derivative pretreatments are the preferred method for obtaining maximum corrosion resistance of the coating system on ferrous and galvanized surfaces. Immersion systems provide coverage to internal as well as external surfaces and are the preferred method for large or complex components.

There have been significant advances in the zinc phosphate industry in the last 8 years. Application processes can be run at lower temperature (110F) with less sludging (disposal), which reduces the overall operating cost. Crystal size, morphology and coverage have been optimized to give consistent performance in conjunction with high performance coatings. Functional capabilities have been enhanced.

Low alloy steels previously required a preliminary abrasive blasting operation to remove excessive surface carbon and oxidation products. Failure to perform this operation resulted in surfaces with 20-50% coverage of zinc phosphate crystals and unsatisfactory coating performance. Alkaline cleaning/zinc phosphating technology that can provide the same alloy with 100% crystal coverage at less than 0.5 micron size without the need for abrasive blasting is commercially available.

Improving the chemical robustness of the system has its drawbacks however. Previously both organic and inorganic contaminants often could be detected by the lack of crystal coverage over the contaminated surface on the scanning electron microscope. Now the contaminated surfaces appear to have the same quality and coverage of crystals as the clean surface. It is the bond strength of the zinc phosphate crystal to the substrate that has been compromised. Failure analysis now requires more stealthy analytical methods to determine the type and degree of contamination.

Chromate conversions (MIL-C-5541) have been the benchmark for pretreating aluminum alloys for the military. Due to the hexavalent chromium content in the chemistry, alternatives are being sought.

Corrosion Resistance Testing

To establish benchmarks that could have high statistical correlation to actual exposure conditions, GM 9540P- Accelerated Corrosion Test (ACT) was selected as the protocol for establishing corrosion resistance.¹ The merit of this protocol is derived from its thermal and moisture cycling, which induces mechanical stresses in the coating system. In other words, this test mimics the real-time fatigue of a coating system in a corrosive environment; it is not just a corrosion test. ACT has become an accepted industry standard and is a readily available protocol at independent test laboratories. Test panel substrates were cold rolled steel (CRS), galvanized steel (G 90), AA 5083-H231 and AA 6061-T6.

Both electrocoat epoxies as well as spray epoxies were evaluated. Products were military approved or commercially advertised as having exceptional levels of corrosion resistance. Seven products from five manufacturers are provided in the listing. It should be noted that comparative data from two additional products was purposely omitted. These two products displayed a significant variation in ACT results at two different test facilities. It is not without coincidence that these same two products have demonstrated significant variations in 336 hr salt spray process control test results over time. The source of the variability cannot be established at this time. It is believed that the variation is more likely attributed to variations in the wet paint (lack of process control, solvent manipulation or additive change or omission by the paint manufacturer) than variations in the test laboratory.

All substrates were cleaned (water break free) and pretreated as noted. Surface profiles were established with replica tape. Some received a combination of both abrasive blasting and chemical pretreatments. One product tested recommended a specific surface profile prior to primer application. Some manufacturers are required to abrasive blast the surface prior to chemical processing due to poor storage/handling conditions that create unacceptable levels of corrosion or purchase hot rolled steel with unacceptable levels of mill scale.

Some test coupons were prepared over knowingly inadequate zinc phosphate pretreatments to assess the impact on performance. An unacceptable pretreatment is considered one that has less than 70% crystal coverage by scanning electron microscope evaluation at 500x. High performance coatings have demonstrated that they are capable of providing adequate protection even with marginal pretreatments. Marginal coatings, however, require almost perfect pretreatment conditions to achieve even a minimal level of performance. Unless otherwise noted, the zinc phosphate pretreatment had 100% crystal coverage. All test panels were cured a minimum of 7 days prior to scribing to assure a fully cured condition.

Testing was conducted at four facilities. In many cases duplicate panels of a particular system were tested at two facilities or at the same facility at different times. Test results were averaged on a minimum of six replicates. Coating dry film thickness was reported as the mean value of the range (six readings per test coupon). The range of coating dry film thickness (dft) was a function of the application process. Electrocoated test coupons had the same dft (relative to dft readings recorded to the nearest 0.1mil) over the entire test surface. On sprayed test panels the range varied from 0.2-1.7 mils on a single panel. Typically, the higher the solids content of the coating the greater the applied range.

A significant limitation of the GM 9540P protocol is the method of evaluation after exposure. The test panels are typically rinsed and subjected to high-pressure air to remove any loose primer; occasionally the air nozzle tip may be used to loosen the corrosion scab. Many coating systems fail by cathodic disbondment. This current automotive method was not considered adequate to address this failure mechanism. Consequently, the evaluation protocol was modified.

All test coupons were scraped at a 30-degree contact angle to the test surface with a blunt-edged, 1.5-inch-wide metal putty knife prior to evaluation. To demonstrate the difference in evaluation techniques, Figure 1 shows a test coupon after an "automotive" evaluation. The opposing image is the same test coupon after being scraped. The scribe creep goes from an intermittent condition to an almost continuous response along the scribe line. The maximum scribe creep measured from one side of the scribe increased from 1.6 mm to 1.7 mm.

The failure criteria for steel and galvanized surfaces was the same criteria used since 1985 by the Tank Automotive and Armaments Command (TACOM) for the system of pretreatment and primer contained in TT-C-490, "Cleaning Methods for Ferrous Surfaces and Pretreatments for Organic Coatings." If any of the following four conditions was met, the test coupon was considered to have reached "failure:"

1. Primer creep, blistering or loss of adhesion relative to a scribed line applied prior to corrosion testing was equal to or exceeded 0.125 inch at any point at the scribe (maximum creep measured from one side of the scribe).
2. Corrosion in the field exceeded rust grade No. 9, ASTM D610 on ferrous surfaces.
3. There were more than five blisters in the field in any 24 in² area.
4. Any single blister exceeded 1 mm.

These criteria are considerably more liberal than most current automotive standards.

The benchmark for aluminum systems was established from a prior internal study using two aluminum alloys, two epoxy primers and two thickness criteria. 160 cycles of GM 9540P were required on AA 5083 and AA 6061 test coupons with a chromate conversion before any primer could be scraped from the scribe. There was no degradation in the field.

The test coupons were generally kept in the ACT cabinet until failure criteria had been met (test to failure). When the corrosion scab or paint eruption met the 0.125 inch criteria, the coupon was scraped. Some coupons were removed at a predetermined stop point such as 40 or 80 cycles.

The epoxy coating on the galvanized and aluminum coupons normally does not erupt around the scribe line as on CRS. These materials must be scraped to determine coating failure.

TABLE I - Primer System Longevity
Substrate AA 5083 AA 6061 AA 5053 AA 6061 AA 6061 AA 5083 G90 G90 G90 CRS CRS CRS CRS CRS CRS CRS CRS CRS CRS CRS CRS CRS CRS	MIL-C-5541 X X X X X	Abrasive Blast (MIL) 1.6 0.9 1.2 0.9 1.2 0.9 0.9 1.2 0.9 1.2	Wash Primer X X X	Zinc Phosphate X X X X X X X X X X X X X	Primer ID H1 H1 R1 R1 P1 D2 P1 H1 S3 H1 H2 P1 D2 H2 P2 P2 H2 H2 D2 H2 H2 D2 H2	VOC (lb/gal) 4.3 4.3 4.3 4.3 0.3 0.6 0.3 4.3 2.8 4.3 3.2 0.3 0.6 3.2 0.6 0.6 3.2 3.2 0.6 3.2 3.2 0.6 3.2	dft (mils) 1.3 1.2 1.3 1.2 0.8 3.9 0.6 1.1 5.9 2.1 3.4 0.8 4.4 1.5 1.2 1.2 0.9 3.6 3.2 3.5 2.0 3.7 1.7	Cycles to Failure 160+ 160+ 160+ 160 160+ 48 80+ 40+ 26* 63 59* 53 52 44 44 35 28* 24* 23 23* 18 15 12	Comments 0.0 scribe creep 0.0 scribe creep 0.0 scribe creep 0.03-in scribe creep 0.0 scribe creep 96-hr delay after blast before prime 0.07-in scribe creep 0.03-in scribe creep field failure field failure poor quality treatment
Notes: Dft is direct reading. There was no correction for surface profile. The reported value is the mean of the range. * Indicates extensive coating disbond from scraping as compared to the visual observation for blistering and corrosion at the scribe line. Cycles to failure was reduced based upon this observation. Scribe creep is rated from one side of the scribe line at the maximum value. 40 cycles (40 days) of GM 9540P is approximately equal to 5 year performance with respect to corrosion. The alpha designator in the primer ID identifies the paint company.

Once the coupon was scraped, the test panel was retired from further corrosion testing. If the failure criteria had not been met at this interval, a plus (+) designation was added to the duration. Noting the scraping procedure, the amount of scribe creep could be considerably greater than that visually observed in the cabinet at the point of removal. Those coupons whose scribe creep increased by scraping more than 0.031(1/32) inch than that recorded prior to scraping were so noted by an asterisk following the number of cycles. The number noted represents a deduction of 4 cycles for every 0.031 inch increment greater than 0.156 inch of scribe creep (for example, a maximum scribe creep of 0.236 inch after scraping would have 10 cycles deducted from the original rating).

Beyond Corrosion Resistance

An appropriate engineering decision could be made with respect to a given coating system from this data alone only if the following two conditions are met: 1) the coated product was in a static application not subject to flex or chipping conditions and 2) the product was not subject to direct sunlight (UV exposure).

TABLE II - Flexibility
Primer ID* P1 P1 P1 H1 H1 H3 H3 N2 N2 S2 S2 S2 S2	dft 0.8 1.0 1.2 1.2 1.7 1.0 1.3 1.5 3.1 1.6 2.1 2.6 3.4	Direct Impact (in-lb) 160 160 120 50 40 50 35 45 40 40 40 16 8	Reverse Impact (in-lb)** 160 160 160 6 < 4 4 < 4 6 < 4 8 < 4 < 4 < 4
* VOC of H3, N2 and S2 primers is 3.5 lb/gal ** A coating with an impact value less than 6 in-lb is categorized as brittle (ASTM D 2794). All test coupons were a minmum of 30 days.

Although epoxies are known for their exceptional corrosion and fluid resistance they suffer from two weaknesses. Epoxies show an inverse relationship between dft and flexibility. Consequently, if you need to increase the dft to get corrosion resistance, you are compromising chip resistance in the process. Various tests are used by the industry to assess flexibility of the coating. Those that best accommodate high strain rates reflective of stone pecking damage are the impact tests per ASTM D2764, STM for Resistance of Organic Coatings to the Effects of Rapid Deformation (Impact) and the various automotive gravelometer tests based upon SAE J400, Test for Chip Resistance of Surface Coatings. A comparison of five epoxies with respect to direct and reverse impact is provided in Table II. The criteria for rating were no cracking by unaided visual observation or disbond (tape) of the coating.

A more insidious problem with epoxies is their poor UV resistance, which degrades the performance of the coating (commonly referred to as chalking) via direct degradation or photolytic oxidation.^2,3 Although it is general knowledge that you don't expose epoxies to UV, there is the belief that applying a topcoat will always shield the epoxy. This assumption can have devastating consequences. There is no accepted method for accelerating UV exposure with a real-time correlation. The best protocols are developed around xenon arc or the use of mirrors to track and intensify light exposure in a desert climate coupled with moisture. Many use the UV test protocols to measure changes in color or gloss over time. The serious damage occurs when the topcoat is transparent to UV transmission allowing the epoxy to degrade and generate a disbond between the primer and topcoat.⁴

Many manufacturers have consequently added proprietary UV absorbers or reflective pigments in their topcoats to protect the epoxy. As the user can't readily verify this addition there must be considerable trust in both the topcoat suppliers chemistry and manufacturing controls to provide a consistent, compliant product. Adding excessive pigment to a topcoat formulation (flat gloss) in the belief that it will block the UV photon is not a valid remedy.

Two problems are associated with flat topcoats:

1. It has been theorized that when the pigment volume concentration (PVC) exceeds approximately 50%, close packing of the pigment particles generates scattering within the coating rather than reflection of the UV energy. The generation of more free radicals increases the probability of degrading the binder.⁵ The low binder content of low gloss coatings inherently provides for lower resistance to UV degradation.
2. Heavily pigmented topcoats may exceed the critical pigment volume concentration (CPVC) due to variations in product batching or improper mixing ratios by the applicator in 2K systems. High PVC topcoats have been found to have microcracking in field systems. The microcracking not only makes cleaning difficult but also generates crevice corrosion problems. The topcoat deterioration also provides a pathway for the UV to contact the epoxy primer.

 

Results Based on this research, we can draw some definite conclusions:

 

1. The use of abrasive blasting as the sole surface pretreatment for epoxy primers does not provide adequate protection to the substrate-coating system when the surface is subject to flex and chipping. An aggressive surface profile dictates the use of even greater dry film thicknesses to cover the substrate peaks, which further degrades epoxy flexibility.
2. Wash primer is the worst possible option for a surface pretreatment on steel. It not only has serious environmental issues but also provides no significant performance improvement to the coating system. If the only two options are abrasive blast and wash primer, the abrasive blast option is the preferred approach.
3. The more passive the substrate, the longer the coating system is capable of lasting.
4. Due to the limited film thickness capabilities of electrocoat primer (without special modification of the process parameters), the use of abrasive blasting in conjunction with electrocoating will reduce the performance capabilities of the system. It is still preferred to the alternative approach of trying to coat directly over corrosion.
5. A poor quality zinc phosphate pretreatment can decrease the longevity of an epoxy primer by more than 50%.
6. Environmental constraints have forced paint manufacturers to reduce the VOC of their primers. Unfortunately, all sprayed primers tested to date show a significant drop in the performance level of their low-VOC epoxies compared to the higher VOC versions. The typical response has been to often increase the dft by 50% or more to obtain comparable corrosion performance. The solvent/surfactant package in the wet paint has a major impact on film adhesion properties. Surface wetting facilitates chemical bonding to the pretreatment. Too much wetting, however, can be just as detrimental as not enough.⁶ Thin film, high-solids coatings appear to suffer from poor surface wetting. When high-solids epoxies are applied at thicker levels, the tack time tends to increase. This longer tack time indicates a higher solvent concentration per unit surface area of substrate, allowing for film leveling and surface wetting. The over zealous use of tail solvents alone is not the solution but rather the intelligent blending of various solvents/diluents to obtain a fully cured, homogeneous film. The solvent blends, however, must be compatible with the curing facilities of the applicator. Numerous field coating failures have been noted where the excessive primer film build in conjunction with inadequate curing have created adhesion failures due to solvent entrapment.
7. The lowering of VOC levels with the resultant increase in dry film thicknesses have resulted in finished products with continually receding levels of chip resistance and flexibility. This may be a direct response to formulators lowering the molecular weight of the resin. The response to this situation in conjunction with unquantifiable levels of UV resistance provided by a topcoat warrants a change in coating system design. The addition of a flexible UV barrier or the replacement of sprayed epoxies with thermoset polyurethane primers are possible options. A third option is to use curing agents that improve flexibility by decreasing the cross link density.⁷
8. There are no two epoxy primers tested thus far with equivalent levels of performance at the same film thickness.

 

The reliance on the salt spray test (which has no correlation to real time exposure) for product development and qualification may be responsible for many misconceptions in the non-automotive coatings industry. Coating formulation is highly complex chemistry, and the degradation mechanisms are no less complex. The formulator at a minimum must have knowledge and control of the molecular weight of the resin, crosslink density of the cured polymer, plasticizer and surfactant additives and pigment type/concentration. These factors have a major impact on the adhesion, corrosion resistance, fluid resistance (for example, water, gasoline, hydraulic fluid or decontamination solvent) and flexibility of the final film. The coating chemist must rely on meaningful physical testing to confirm the actual performance of his engineered product. The variables involved and possible synergisms, either positive or negative, are too complex to intuitively predict any outcome.

The writer would like to acknowledge the following individuals for their assistance in developing the test data for this project: Shawn Dolan, Henkel Surface Technologies; John Shaffer, Cathy Trusik and Varina Andriaschko, PPG Industries; and Tom Braswell, UDLP.