Traditional chemical resistance testing of tank linings is a time-consuming and largely empirical process. Both of these characteristics define limits to the formulator's efforts, first by severely constraining the number of experimental formulations that can be tested, and second by reducing the information gained in the developmental process. This paper reports work done comparing the results of long-term testing of several tank linings' resistance to various solvents, to short-term testing based on a new experimental methodology. This methodology draws on ideas from modem thermodynamics and glass transition theory. Inferences are drawn both from those cases where the correlation between the two methods is high, and from those where the correlation is low. The value of the short-term method is discussed both from a developmental and an informational perspective.
KEY WORDS Epoxy, tank linings, chemical resistance, test methods, logistic regression