DOI
https://doi.org/10.25772/X1PG-MA04
Defense Date
2012
Document Type
Thesis
Degree Name
Master of Science
Department
Mechanical and Nuclear Engineering
First Advisor
Brian Hinderliter
Abstract
Natural weathering and wear of structural materials in service nearly always generate surface roughness, and follow the Central Limit Theorem prediction for surface topology. This study couples experimental and statistical theory, and FEM to extend knowledge of life of materials from initial service surface conditions through surface damage accumulation. Statistical moments and other parameters were correlated with fracture locations probability (H/N), versus auto correlation length, and depth. As the surface grows to a full Gaussian, H/N increases its dependence on profile’s Average and RMS Roughness, and derived parameters. This dependence shows an asymptotic limit behavior that approaches agreeably Griffith’s crack criterion, though with multiple pit locations. Importantly, a Transitional Region was observed, below which the probable location of fracture is uncorrelated to the parameters studied. This is because introduced roughness is insufficient to compete with impurities, internal and external manufacturing flaws, and scratches, due to handling and machining, on the samples.
Rights
© The Author
Is Part Of
VCU University Archives
Is Part Of
VCU Theses and Dissertations
Date of Submission
May 2012