Statistical modeling of enamel rater value data

Abstract

Enamel rater value (shortly, ERV) of a quick stress test is usually used to evaluate the integrity of an organic coating for the inside of an aluminum (denoted by Al shortly) can. A large positive value of ERV is supposed to indicate the degree of imperfect coating coverage, i.e. the size of an exposed Al area. An Al can filled with some drink, if there is an exposed Al area due to imperfect coating coverage, has Al dissolution brought by corrosion. Thus a smaller value of ERV is desirable to prevent Al dissolution. However, quantitative evaluations of ERV data as well as an accumulated quantity of Al dissolution have never been published, because ERV is involved in complicated anode dissolution of an exposed Al area. Recently our experimental study has found out a relationship between ERV and sizes of exposed Al areas. This relationship enables us to construct a descriptive statistical model for ERV data as well as to evaluate coating effects for Al cans. Furthermore, empirical implications suggest that an instantaneous quantity of Al dissolution is proportional to ERV. Using this fact, we can derive a predictive statistical model for an accumulated quantity of Al dissolution in an Al can.

Keywords:

• enamel rater value
• aluminum cans
• corrosion
• aluminum dissolution
• generalized gamma distribution
• new power normal family