Perhaps the most difficult task confronting the paint test engineer is the determination of the “durability” of an organic protective coating, that is, the protection it affords under conditions of actual service. The majority of painted articles are subjected to a wide variety of external influences which may vary over a relatively wide range so that it becomes very difficult to study the effect produced by each individual factor, or by any combination or sequence of factors, and to correlate the results into a reliable picture of the value of the paint as a protective medium.
One solution of this problem consists of actual service tests on an, extended scale and the application of statistical methods to the exposure results. In this case, no particular effort need be made at controlling or studying the effect produced by all variables individually, but a sufficiently large number of tests must be conducted...
