Statistical approaches for predicting weathering degradation and service life

S. G. Croll, B. R. Hinderliter, S. Liu

Research output: Contribution to journalArticlepeer-review

21 Scopus citations

Abstract

Predicting the service lifetime of any material is very important. One of the major difficulties is relating individual, nanoscale, degradation events to the eventual deterioration in macroscopic properties. Monte Carlo simulations and the Central Limit theorem provide approaches to the kinetics of how general coating topology, or bulk morphology, changes during long term weathering. Results on structural changes, from these approaches, may then be translated, via well-known models, into the degradation of macroscopic, "use" properties. Thus service lifetime may be related to composition and degradation mechanisms. Results on pigmented, acrylic coatings demonstrate the applicability and utility of this statistical approach. Changes in gloss and contact angle were related, via these models and provide insights into the processes of degradation in pigmented coatings. Values of parameters for the models that relate surface roughening with exposure are consistent with results from atomic force microscopy. The models provide justification for using simple expressions for trends in non-destructive monitoring, e.g. gloss, or contact angle or for comparing different systems and estimating long-term changes in properties.

Original languageEnglish (US)
Pages (from-to)75-87
Number of pages13
JournalProgress in Organic Coatings
Volume55
Issue number2
DOIs
StatePublished - Feb 1 2006

Bibliographical note

Funding Information:
The authors would like to thank the US Air Force Office of Scientific Research for support of this research, contract number FA9599-04-1-0368.

Keywords

  • Contact angle
  • Durability
  • Gaussian
  • Gloss
  • Service life
  • Statistics

Fingerprint Dive into the research topics of 'Statistical approaches for predicting weathering degradation and service life'. Together they form a unique fingerprint.

Cite this