Engineering Papers⌕ Search

Engineering topics

Moore, Leslie M

Publications and source records attributed to Moore, Leslie M.

Batch sequential designs for computer experiments

Computer models simulating a physical process are used in many areas of science. Due to the complex nature of these codes it is often necessary to approximate the code, which is typically done using a Gaussian process. In many situations the number of code runs available to build the Guassian process approximation is limited. When the initial design is small or the underlying response surface is complicated this can lead to poor approximations of the code output. In order to improve the fit of the model, sequential design strategies must be employed. In this paper we introduce two simple distance based metrics that can be used to augment an initial design in a batch sequential manner. In addition we propose a sequential updating strategy to an orthogonal array based Latin hypercube sample. We show via various real and simulated examples that the distance metrics and the extension of the orthogonal array based Latin hypercubes work well in practice.

97 MATHEMATICS AND COMPUTING↗

Statistical design of a uranium corrosion experiment

This work supports an experiment being conducted by Roland Schulze and Mary Ann Hill to study hydride formation, one of the most important forms of corrosion observed in uranium and uranium alloys. The study goals and objectives are described in Schulze and Hill (2008), and the work described here focuses on development of a statistical experiment plan being used for the study. The results of this study will contribute to the development of a uranium hydriding model for use in lifetime prediction models. A parametric study of the effect of hydrogen pressure, gap size and abrasion on hydride initiation and growth is being planned where results can be analyzed statistically to determine individual effects as well as multi-variable interactions. Input to ESC from this experiment will include expected hydride nucleation, size, distribution, and volume on various uranium surface situations (geometry) as a function of age. This study will also address the effect of hydrogen threshold pressure on corrosion nucleation and the effect of oxide abrasion/breach on hydriding processes. Statistical experiment plans provide for efficient collection of data that aids in understanding the impact of specific experiment factors on initiation and growth of corrosion. The experiment planning methods used here also allow for robust data collection accommodating other sources of variation such as the density of inclusions, assumed to vary linearly along the cast rods from which samples are obtained.

36 MATERIALS SCIENCE↗

Experiment planning for an x-ray material study

An experiment is planned in which up to 3 materials will be subjected to conditions determined by 3 factors: dose, dose rate, and x-ray energy. The 3 materials, new compact material, powdered material, and aged compact material, are referred to simply as 'compact', 'powder' , and 'aged' respectively. Initial consideration of experimental settings for this experiment proposed up to 7 levels of dose, 3 levels of dose rate, and 4 levels of x-ray energy. All combinations of these materials and conditions would yield 3 x 7 x 3 x 4 = 252 possible experiments in the proposed study. However, it is expected that at most 23 experiments can realistically be set-up and monitored. This document describes statistical experiment design methods for obtaining a subset of combinations that efficiently provide the desired information needed from the study.

97 MATHEMATICS AND COMPUTING↗