Probabilistic modeling and simulation of metal fatigue life prediction

Loading...
Thumbnail Image
Authors
Heffern, Thomas V.
Subjects
Metal fatigue
Variable strain-life
Coffin-Manson
Monte Carlo simulation
Service life extension
Evolutionary algorithms
Genetic algorithms
Aluminum 7050-T7451
Probabilistic investigation
Advisors
Kolar, Ramesh
Wood, E. Roberts
Date of Issue
2002-09
Date
September 2002
Publisher
Monterey, California. Naval Postgraduate School
Language
Abstract
The work of this thesis was to investigate the probability distributions of test data taken for aluminum 7050-T7451, and to attempt to develop a probability based model from the variation of the 4 fatigue life constants ( , f ̤, ' f ,Ì b,c). The goal was to create strain-life-probability curves that would more accurately describe the likelihood of failure at a given strain amplitude. The investigator determined that the test data did not demonstrate any consistent known probability density function. The investigator cautioned against assuming a normal distribution before it could be completely established as the predominate probability density function. Possible consequences of invalid assumptions were presented. Attempts were made to explain the disparity of sample data between two different laboratories testing of the same material. Assuming random behavior within an established range, probability based models were developed using the 4 strain- life constants. It was determined that in order to create a complete probability based model, an accurate regression of the test data must fit all strain levels to include the intermediate strain level's "knee". In an attempt to solve that problem, 8 parameter equations were explored. Methods to predict the 8 parameters included random number simulation combined with non- least squares curve fits, evolutionary algorithms and genetic algorithms.
Type
Thesis
Description
Series/Report No
Department
Aeronautics and Astronautics
Organization
Identifiers
NPS Report Number
Sponsors
Funder
Format
xviii, 116 p. : ill. (some col.)
Citation
Distribution Statement
Approved for public release; distribution is unlimited.
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.
Collections