Significance and Use
This test method is used to compare the endurance of different materials to the action of corona on the external surfaces. A poor result on this test does not indicate that the material is a poor selection for use at high voltage or at high voltage stress in the absence of surface corona. Surface corona should be distinguished from corona that occurs in internal cavities for which no standardized test has been developed. Evaluation of endurance by comparison of data on specimens of different thickness is not valid.
The processing of the material may affect the results obtained. For instance, residual strains produced by quenching, or high levels of crystallinity caused by slow cooling may affect the result. Also, the type of molding process, injection or compression, may be important especially if the mixing of fillers or the concentration and sizes of gas-filled cavities are controlled in any degree by the process. Indeed, this test method may be used to examine the effects of processing.
The data are generated in the form of a set of values of lifetimes at a voltage. The dispersion of failure times can be analyzed using Weibull or extreme value statistics to yield an estimate of the central value of the distribution and its standard deviation. This is particularly recommended when the dispersion of failure times is large, and a comparison of lifetimes of two materials must be made at a specified level of confidence.
This test is often used to demonstrate the differences between different classes of materials, and to illustrate the importance of eliminating corona in any application of a particular material. When the test is used for such purposes or other similar ones, the need for precision is reduced, and certain time saving techniques, such as truncating a test at the time of the fifth failure of a set of nine, and using that time as the measure of the central tendency, are recommended. Two such techniques are described in 10.2. Both techniques remove the necessity of testing beyond median failure, and reduce the required testing time to approximately half of that required to obtain failures on all specimens.
Insulating materials operating in a gaseous medium are subjected to corona attack at operating voltage on some types of electrical apparatus in those regions where the voltage gradient in the gas exceeds the corona inception level. On other types of equipment, where detectable corona is absent initially, it may appear later due to transient over-voltages or changes in insulation properties attending aging. Certain inorganic materials can tolerate corona for a long time. Many organic materials are damaged quickly by corona, and for these, operation with no detectable corona is imperative. This test method intensifies some of the more commonly met conditions of corona attack so that materials may be evaluated in a time that is relatively short compared to the life of the equipment. As with most accelerated life tests, caution is necessary in extrapolation from the indicated life to actual life under various operating conditions in the field.
The failure produced by corona may be due to one of several possible factors. The corona may erode the insulation until the remaining insulation can no longer withstand the applied voltage. The corona may cause the insulation surface to become conducting. For instance, carbonization may occur, so that failure occurs quickly. On the other hand, compounds such as oxalic acid crystals may be formed, as with polyethylene, in which case the surface conductance will vary with ambient humidity, and at moderate humidities the conductance may be at the proper level to reduce the potential gradient at the electrode edge, and thus cause either a reduction in the amount of corona, or its cessation, thus retarding failure. The corona may cause a “treeing” within the insulation, which may progress to failure. It may release gases within the insulation that change its physical dimensions. It may change the physical properties of an insulating material; for instance, it may cause the material to embrittle or crack, and thus make it useless.
Tests are often made in open air, at 50 % relative humidity. It may be important for some materials to make tests in circulating air at 20 % relative humidity or less (see Appendix X1). If tests are made in an enclosure, the restriction in the flow of air or other gas may influence the results (see Appendix X2).
The shape of the (voltage stress)-(time-to-failure) curve is sometimes useful as an indicator of the useable electric strength of a material in an application involving surface corona and its variation with time of application of voltage, though such comparisons are risky. (Specimen thickness, electrode system, the presence of more than one mechanism of failure, and the details of the ambient, including the nature of the surface corona, all have significant effects.) For instance, on log-log paper, the volt-time curve often obtained by the procedures of this test for void-free materials such as polyethylene sheet generally has a continuous curvature that is slightly concave upward. The low voltage end of the curve tends toward the horizontal and approaches a threshold voltage below which the curve does not go. A similar threshold would be expected for many materials in an application involving surface corona. Moreover, if the material possesses a low electric strength (as measured by Test Method D 149
For materials that possess a basic resistance to corona, such as mica, or, to a smaller degree, silicone rubber, the time required for the curve to reach the threshold produced by corona may be greater by many orders of magnitude than the time required for materials such as polyethylene, polyethylene terephthalate, or polytetrafluoroethylene.
The variability of the time to failure is a function of the constancy of the parameters of the test, such as the test voltages, which should be monitored. It is also a significant material property. The Weibull slope factor, β, is recommended as a measure of variability. β is the slope obtained when percent failure is plotted against failure time on Weibull probability paper. Such a plot is called a “Weibull probability plot” (see Fig. 1).
The shape of the Weibull probability plot can provide additional information. A non-straight-line plot may indicate more than one mechanism of failure. For instance, a few unaccountably short time failures in the set could indicate a small portion of defective specimens with a different failure mechanism from the rest of the lot.
Scope
1.1 This test method differentiates among solid electrical insulating materials for use at commercial power frequencies with respect to their voltage endurance under the action of corona (see Note 1). In general, this test method is more meaningful for rating materials with respect to their resistance to prolonged a-c stress under corona conditions than is dielectric strength.
Note 1—The term “corona” is used almost exclusively in this test method instead of “partial discharge”, because it is a visible glow at the edge of the smaller electrode. This is a difference in location, not in kind. Partial discharges also occur at the edges of electrodes, and in general corona describes an electrical discharge irrespective of its location.
1.2 The values stated in SI units are to be regarded as standard. The values given in parentheses are mathematical conversions to inch-pound units that are provided for information only and are not considered standard.
1.3 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility of the user of this standard to establish appropriate safety and health practices and determine the applicability of regulatory limitations prior to use. For specific hazard statements, see Section 7.
