A statistical degradation model for the service life prediction of aircraft coatings: With a comparison to an existing methodology

An advance on the model used by Guseva et al. [1] for estimating the service life of organic coatings under service conditions from accelerated test results has been developed. Instead of modelling just the times to failure, this new approach uses the complete degradation curve and allows for a more general specification of the failure time distribution. This modified model was then applied to the estimation of the service life (defined as gloss loss) for aircraft coatings and the results were compared to those obtained by Guseva et al. It was found that when this model was applied to the naturally weathered data, gloss loss (and thus failure times at a given level of gloss loss) followed a generalised gamma distribution, rather than the Weibull distribution identified by Guseva et al. Further, the new approach suggested a reduction in the warranty time of about one month. When the model was applied to the accelerated test data, it produced more accurate extrapolations of the median failure time associated with the naturally weathered data – (49.3 months compared to a measured 52 months obtained at the naturally weathered site). In fact, the extrapolated distribution obtained by this new approach was much closer to the distribution for the naturally weathered data than the extrapolated distribution obtained by Guseva et al.

 

M. Evans

Polymer Testing, 2012, Vol. 31(1), Pages 46-55 doi: 10.1016/j.polymertesting.2011.08.017