In the 1980s, damage to automotive coatings from acid rain became a problem.
Both dealerships and buyers complained about the ring type or ?water spot?
etches that were caused by acid catalyzed hydrolysis in areas where acid rain
occurred and temperatures were high.

Etches are formed when material is lost from the coating
surface. This material is lost as a result of acid-catalyzed hydrolysis of the
chemical bonds within the coating. When enough bonds are broken, polymer
molecules or fragments become detached from the rest of the coating and are
washed away. The lost material is greatest at the edge of water droplets where
the acid concentrates during evaporation. This leads to the familiar ring or
?water spot? pattern.

Most successful work on improving the acid etch resistance
of automotive coatings has focused on reformulation to reduce or eliminate the
acid-sensitive sites within the polymer network. Most automotive coatings are
comprised of a highly pigmented color coat (or basecoat) that is covered with
a transparent clearcoat.

Blount Island Exposures



Summertime on Blount Island in Jacksonville, Fla., is, and continues to be,
one of the most severe locations in North America for the conditions that
cause acid etch. As a result, this location is the site for numerous annual
field tests to assess the performance of automotive coatings for acid etch
resistance. Varying numbers of hoods, panels and fascia are placed in
Jacksonville each year by automotive OEMs and their suppliers. The hoods,
panels and fascia vary in size and shape, but are mostly black to create the
worst-case scenario for etch testing. The items are exposed in a variety of
ways, but typically tested between 0 and 5 degrees horizontal. The typical
testing period is approximately 14 weeks, from June to September of each year.
The test specimens are rated for etch damage at varying times throughout the
14-week period using a visual method of evaluation with standard panels used
as a guide. While the Blount Island exposures have the benefit of real-world
testing, there are significant limitations to the current procedure in terms
of product development. In addition to the inconvenience of a single location
and narrow time frame for testing, the results of this annual test vary like
the weather.

Accelerated Acid Etch



Many attempts have been made at developing laboratory techniques to predict
etch resistance of automotive clearcoats. There are a number of methods
currently used in the industry, like the Gradient Bar Test, the Acid Spot
Test, etc. None of these tests have included all of the field components that
contribute to the etching of automotive clearcoats ? humidity, UV light,
specimen orientation, acidic solution.

Development of Current Test

BASF recognized the need for a realistic
laboratory-accelerated acid etch test procedure and began to quantify the
critical elements. Many of the critical test parameters could be reproduced in
existing xenon arc test chambers.

Since the 1950s, xenon arc testers have been used to test
the weatherability of coatings. These devices have attained significant
popularity because they utilize a light source which, when properly filtered,
provides an excellent simulation of the full spectrum of sunlight.
Traditionally, these devices have a xenon light placed in the center of the
chamber and the test specimens mounted vertically on a framework that revolves
around the light source, like a carousel. This mechanism is often called a
?rotating drum? style tester (Figure 1), and is available in many models
from several manufacturers. Unfortunately, the vertical specimen mounting
system means any liquid sprayed onto the specimen tends to rapidly run off.

Recently, Q-Panel Lab Products developed and introduced the
Q-Sun Xenon Test Chamber, which has the xenon lamps positioned at the top of
the exposure chamber and positions the test parts and panels underneath the
lamps in a near-horizontal orientation (Figure 2).1 This has several design
advantages. Specifically, any liquid sprayed onto the specimens tends to
remain for an extended time. Instead of quickly running off, as it does on the
older rotating drum-style xenon tester, it slowly dries in place.

It was recognized that this flat-array xenon would be
particularly useful for reproducing the acid rain effects seen on horizontal
specimens in Jacksonville. Because of Q-Panel?s experience in weathering
science, correlation studies and tester design, BASF partnered with Q-Panel to
embark on a joint effort to develop a realistic accelerated acid etch test
procedure.

Quantifying the Exposure Environment



BASF has been monitoring the exposure conditions at the Jacksonville exposure
site for a number of years. Based on that data, the following are the critical
environmental conditions that were considered in the development of a new BASF
Accelerated Acid Test procedure.

Temperature

Early on, we recognized the importance of the effect of
temperature on the Jacksonville field exposure results. Consequently,
temperature parameters were quantified by taking real-time measurements of the
actual specimens that were exposed in Jacksonville (Table 1). In 1993,
thermocouple measurements produced specimen temperatures as high as 80 ?C. In
2002, Jacksonville pyrometer measurements of actual parts and test panels,
under field conditions, show the maximum specimen temperatures to be
approximately 72?C. As a result of this data, 80 ?C was chosen as the target
Uninsulated Black Panel2 temperature for the light exposure step of the lab
test.

Rainfall



We identified the need to use a simulated rain solution with a specific pH and
chemical composition known to produce etch in Jacksonville. Field observations
indicated that trace rainfall of less than 0.25 cm (< 0.1 inch) and low
cloud cover are the conditions most responsible for producing acid etch in
Jacksonville. Studies determined that lower pH rainfalls are the most
responsible for producing etch (e.g., 3.49 pH collected in Jacksonville in
1989). Weather data from Jacksonville indicates that, between June and August,
there are an average of 10-15 days of this type every year (Table 2).

The pH and chemical composition of the simulated acid rain
solution used for the accelerated laboratory method was based on an analysis
of actual Jacksonville rain samples.

Humidity and Wet Time



Research indicated a need to maintain a relative humidity that is consistent
with Jacksonville?s natural exposure environment. This would best mimic the
prolonged dry off seen in the field. Weather data shows that the relative
humidity averages approximately 80% during the summer months (June-August).

Time of wetness (TOW) research from Florida and other
locations indicates that test specimens are wet more than 50% of the time and
that the source of this wetness is dew (Grossman, 1978). Field observations
confirmed this by determining that, on most summer evenings, dew forms on the
parts and panels. Typically, the dew is still in place the following morning.

Therefore, in the lab simulation, a series of pure water
sprays during the dark step are used to simulate the evening dew. At the same
time, a high humidity of 80% is maintained throughout the test to simulate the
summer conditions in Jacksonville.

Specimen Orientation

Panels and parts exposed outdoors in Jacksonville are
positioned at the horizontal or near-horizontal orientation angles that give
severe etch effect. Typically, the most severe etching is observed at 0 to 5
degrees exposure.

To reproduce the most severe field exposure condition for
the accelerated test, the Q-Sun test chamber was modified to orient the test
specimens at 0 degrees. (This is a modification from the normal Q-Sun exposure
angle of 10 degrees.)

UV Light

Q-Panel?s experience dictated that, for the best
correlation to outdoor results, the lab specimens should be exposed to UV
light with a similar Spectral Power Distribution (SPD) and intensity to that
which is seen in the field (Table 3). Q-Panel research on sunlight spectrum
shows that, although the spectrum of daylight changes minute by minute
throughout the day (Figure 3), the peak solar noon summer sunlight maximum is
approximately 0.68 W/m2/nm @340 nm. Q-Panel?s measurements are in essential
agreement with CIE 85, Table 4 and with the new SMART2 spectra currently
proposed by ASTM Committee G 03.

Some older automotive test methods like SAE J1960 use an
?extended UV? xenon spectrum to accelerate the coating degradation,
(Figure 4). This spectrum has the disadvantage of producing short-wavelength
UV below the solar cut-on point of 295 nm. Experience has taught us that this
spectrum can cause unnatural results for some coatings. Consequently, more
recent test protocols, like SAE J2527, allow for a more realistic spectrum by
specifying the Daylight Filter described in ASTM G 155 (Figure 5). (This is
the same spectrum specified in ISO 4892-2 and ISO 11341). The Daylight Filter
spectrum was chosen for the new BASF Accelerated Acid Test procedure because
of its close match with natural sunlight.

Experimental Development

BASF developed a simulated acid rain solution that was based
on the observed acid rain chemistry in Jacksonville and provided Q-Panel an
initial set of test specimens coated with four types of clearcoat systems.

The Q-Sun Xenon Test Chamber was modified into a new model
to incorporate the features dictated by field observations. A 0 degree
specimen mounting plane and a dual spray system were added. The dual spray
system can be programmed to automatically spray either pure DI water or
simulated acid rain solution.

Using the modified Q-Sun test chamber, Q-Panel experimented
with various test cycles, including 100% light, with intermittent acid spray.
Q-Panel determined that a cycle with both a light and a dark time exposure
gave better correlation to the outdoor etch results from Jacksonville. The
volume and frequency of acid spray was adjusted to best simulate the outdoor
results.

After some Edisonian Research, an optimized exposure cycle
was developed. After the cycle was determined, the test specimens were exposed
in the Q-Sun and evaluated at intervals of 200, 300, 400, 500, 600 and 700
hours.

Acid Etch Evaluation Procedure

The evaluation of acid damage is performed visually and the
test specimen is rated on a scale from 0 (best) to 10 (worst). A summary of
the rating scale is described in Table 5.

For certain tests, the scale was extended to allow for finer
discrimination.

Exposure Results Compared

Jacksonville Data



Data from two years? Jacksonville natural exposures were compared to
establish a baseline. As expected, there were differences from year to year in
the absolute values. However, there was perfect agreement in the rank order of
the various systems from year to year (Table 6 and Figure 6).

BASF Accelerated Acid Test Results

The test specimens were rated at 100-hour intervals,
beginning at 200 hours. The results are shown in Table 7 and Figure 7. After
200 hours in the Q-Sun, the relative rank order was well established and
remained unchanged throughout the exposure period.

The accelerated acid test results were compared to actual
Jacksonville natural exposure data. As seen in the figures, after only 200
hours in the Q-Sun, the BASF accelerated acid test procedure gave the same
ranking as the Jacksonville exposures. After 400 hours, it produced both the
correct Spearman rank order (rho = 1.0) and approximately the same level of
etching as seen after 14 weeks of the 2001 Jacksonville exposure (Figures 8
and 9). Test results for 700 hours were essentially identical to 2002
Jacksonville data (Figure 10).

Expanded Testing

Because of the excellent results, the test was expanded to
include more clearcoat systems for which Jacksonville data was already
available. To establish a benchmark, the Jacksonville 2001 and 2002 data were
compared using both Pearson and Spearman correlation methods. For data of this
type, the authors believe that Pearson?s method is the more useful. The data
sets exhibited a Pearson?s correlation coefficient of 0.88 and a Spearman
rank order coefficient of 0.72.

The same systems were exposed for 420 hours to the BASF
Accelerated Acid Test procedure.

The Q-Sun results agreed with the 2001 Jacksonville results
with a Pearson?s correlation of 0.90 and a Spearman?s rank order
coefficient of 0.80 (Table 8).

When the BASF Accelerated Acid Test ratings were compared to
the average of the 2001 and 2002 Jacksonville results, the correlation was
even better (Table 9). Pearson R2 = 0.93 and Spearman rho = 0.80. In short,
the BASF Accelerated Acid Test results agreed with Jacksonville as well as, or
better than, Jacksonville agreed with itself.

Summary and Conclusions



A new BASF Accelerated Acid Test procedure was developed jointly by BASF and
Q-Panel Lab Products. The procedure identified and incorporated all of the
known critical test parameters. In order to accomplish this, BASF developed a
simulated acid rain solution and Q-Panel modified the Q-Sun Xenon Test
Chamber. Correlation between the new procedure and the Jacksonville natural
exposures are better than, or equal to, the correlation between Jacksonville
year to year results.

The development of the new procedure has a number of
significant benefits to industry:

1. It allows for faster development of etch-resistant
coatings. As many as 20 iterations per year of formulate/test/reformulate
vs. current one iteration per year, as dictated by natural Jacksonville
exposures.


2. The relatively pristine condition of specimens tested
in the lab allows for the use of digital evaluation of etches. The
Jacksonville panels cannot practically use this technique due to scratches,
dirt, etc.


3. It is expected that this BASF Accelerated Acid Etch
Test can be used to simulate other acid rain environments, where the rain
chemistry differs from Jacksonville.


4. The new procedure allows the possibility of consistent
monitoring of assembly plant systems for etch as an ?early warning
system.?


5. Ultimately, because of the new test procedure, there
should be fewer acid etch failures in service.