This paper studies the effect of deviating from the normal distribution assumption when considering the power of two many-sample location test procedures: ANOVA (parametric) and Kruskal-Wallis (non-parametric). Power functions for these tests under various conditions are produced using simulation, where the simulated data are produced using MacGillivray and Cannon's [10] recently suggested g-and-k distribution. This distribution can provide data with selected amounts of skewness and kurtosis by varying two nearly independent parameters.

Abstract. This paper studies the effect of deviating from the normal distribution assumption when considering the power of two many-sample location test procedures: ANOVA (parametric) and Kruskal-Wallis (non-parametric). Power functions for these tests under various conditions are produced using simulation, where the simulated data are produced using MacGillivray and Cannon's [10] recently suggested g-and-k distribution. This distribution can provide data with selected amounts of skewness and kurtosis by varying two nearly independent parameters.

No abstract

No abstract

The effect of weld properties on the steady-state creep rupture stresses, failure lives, and failure positions within multi-material finite element models of three typical branched steam pipe configurations is investigated using service-aged CrMoV creep properties. A general formulation, multi-material interpolation technique for steady-state creep is shown to be applicable to the parametric study of a three-material welded branched pipe with respect to the steady-state creep constants. An approximate interpolation procedure for power-law creep is implemented to reduce the number of analyses needed to span the range of material parameters. The method is used to predict creep stresses at several critical regions within the heat-affected zones, parent material, and weld metal regions of the branched pipe for three typical CrMoV weldment material sets. The advantages of the method are related to the small number of analyses required and the simple and compact way of presenting results for weld assessment and design. The results can be used to improve life-assessment procedures and to provide guidance to component designers.