|
Activity Number:
|
428
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
| Abstract - #306665 |
|
Title:
|
Comparison of Resampling Methods to Least Squares and Maximum Likelihood Methods for Estimating Fracture Strength Parameters of Glass Test Specimens
|
|
Author(s):
|
Gerald Shaughnessy*+ and Peter Hovey
|
|
Companies:
|
University of Dayton
|
|
Address:
|
5834 Daffodil Circle, Dayton, OH, 45449,
|
|
Keywords:
|
Weibull ; resampling methods
|
|
Abstract:
|
The use of the Weibull model to characterize the failure of brittle materials has enjoyed widespread acceptance. The Weibull probabilistic model is founded on a "weakest link" mode of failure which allows the probability of failure to be related to specimen size. This paper will utilize resampling methods to estimate the strength characteristics of glass specimens. The results of the resampling estimation will then be compared to the traditional estimation procedures using least squares and maximum likelihood.
|
