Published: Jan 1981
| ||Format||Pages||Price|| |
|PDF Version (228K)||21||$25||  ADD TO CART|
|Complete Source PDF (5.2M)||21||$55||  ADD TO CART|
Two classical methods for estimation of Weibull parameters, namely, the moment estimation and the maximum likelihood estimation are reviewed. It is demonstrated that, assuming the location parameter to be zero, the errors made in estimating both the scale and the shape parameters by these methods increase linearly with the true value of the location parameter.
An alternative method for Weibull parameter estimation is proposed. This method provides estimates for all three parameters when the shape parameter is close to one, and estimates the scale and the shape parameters more accurately than the other two methods for higher values of the shape parameter. Furthermore, unlike the other methods, this method is capable of separating the constituent components in a multi-component sample and estimating the parameters of each component. This property makes the method suitable for statistical analysis of composite material strength and fatigue life data.
composite materials, Weibull distribution, statistical analysis, strength data, fatigue life data, reliability assessment, fatigue (materials)
Associate professorresearch scientist, Technical University of DenmarkRisø National Laboratory, LyngbyRoskilde,
Paper ID: STP27626S