Method-It is possible that electrical insulation in service will fail as a result of tracking, erosion, or a combination of both, if exposed to high relative humidity and contamination environments. This is particularly true of organic insulations in outdoor applications where the surface of the insulation becomes contaminated by deposits of moisture and dirt, for example, coal dust or salt spray. This test method is an accelerated test that simulates extremely severe outdoor contamination. It is believed that the most severe conditions likely to be encountered in outdoor service in the United States will be relatively mild compared to the conditions specified in this test method.

Test Results-Materials can be classified by this test method as tracking-resistant, tracking-affected, or tracking-susceptible. The exact test values for these categories as they apply to specific uses will be specified in the appropriate material specifications, but guideline figures are suggested in Note 4. Tracking-resistant materials, unless erosion failure occurs first, have the potential to last many hundreds of hours (Note 5). Erosion, though it is possible that it will progress laterally, generally results in a failure perpendicular to the specimen surface. Therefore, compare only specimens of the same nominal thickness for resistance to tracking-induced erosion. Estimate the extent of erosion from measurements of the depth of penetration of the erosion. Place materials that are not tracking-susceptible in three broad categorieserosion-resistant, erosion-affected, and erosion-susceptible. When the standard thickness specimen is tested, the following times to failure typify the categories (Note 6):

Note 4-Tracking-susceptible materials usually fail within 5 h. Tracking-affected materials usually fail before about 100 h.

Note 5-This information is derived from the individual experiences of eight laboratories using this test method since its publication as a suggested test method in June 1957, and from the results of an organized test program among these laboratories.

Note 6-In a normal distribution approximately 68 % of all test values are included within ±1 standard deviation of the mean.

Interpretation of Test Results-This test method provides information that allows classification as described in 6.2. The comparison of materials within the same group is likely to be ambiguous unless three or more replicate specimens are tested. When the test method is used for specification purposes, do not establish simple minimum values without consideration of the large variance to be expected in test results. It is recommended that quality levels and specification minima be determined by statistical techniques.