Engineering Papers⌕ Search

Engineering topics

Anderson-cook, Christine M

Publications and source records attributed to Anderson-cook, Christine M.

A designed screening study with prespecified combinations of factor settings

In many applications, the experimenter has limited options about what factor combinations can be chosen for a designed study. Consider a screening study for a production process involving five input factors whose levels have been previously established. The goal of the study is to understand the effect of each factor on the response, a variable that is expensive to measure and results in destruction of the part. From an inventory of available parts with known factor values, we wish to identify a best collection of factor combinations with which to estimate the factor effects. Though the observational nature of the study cannot establish a causal relationship involving the response and the factors, the study can increase understanding of the underlying process. The study can also help determine where investment should be made to control input factors during production that will maximally influence the response. Because the factor combinations are observational, the chosen model matrix will be nonorthogonal and will not allow independent estimation of factor effects. In this manuscript we borrow principles from design of experiments to suggest an 'optimal' selection of factor combinations. Specifically, we consider precision of model parameter estimates, the issue of replication, and abilities to detect lack of fit and to estimate two-factor interactions. Through an example, we present strategies for selecting a subset of factor combinations that simultaneously balance multiple objectives, conduct a limited sensitivity analysis, and provide practical guidance for implementing our techniques across a variety of quality engineering disciplines.

42 ENGINEERING↗

Features added in new release of SRFYDO - February 2009

Version 2.0 of SRFYDO, released in February 2009, has a number of new features. This document describes these additions and is designed to complement the document 'Calculating System Reliability with SRFYDO: A User Guide' (Los Alamos National Laboratory Technical Report LA-UR 08-06161). The new features include: (1) Reliability Estimates for Combinations of Components - Users can now estimate the reliability of collections of components, such as a sections or sub-systems. This optional feature is activated via a new table, SectionDef, which is used to define names and component associations for the section or sub-system aggregate. If the SectionDef table is present, section plots and tabulated reliabilities will be created during the SRY3 phase. For more details see section 1. (2) Plots of System or Component Reliability Priors - SRFYDO now creates reliability plots for each system (if system priors were given) or each component (if component priors were given) during the SRY1 phase. These plots, written in the directory 'priorPDFs', are provided to help users visualize the impact of their estimates on the final reliability calculation. For more details see section 2. (3) Convenient Exclusion of Lifecycle Covariates - Both GUI and command-line interfaces to SRFYDO now allow the user to exclude one or both of the lifecycle covariates without requiring modification to the spreadsheet. With this capability users can readily explore the effect of adding or removing lifecycle measures on the resulting statistical analysis. Note that excluding a lifecycle covariate is not the same as setting its rate (-b or -c) to zero. For more details see section 3. (4) MCMC Diagnostic Plots - Diagnostic or 'trace' plots may optionally be created for each of the statistical parameters generated during the Monte Carlo phase (SRY2). These plots can help the user determine if (1) appropriate lifecycle variable choices have been made, (2) more sampling iterations, '-i', are required, and (3) suggest the number of samples to discard, '-x', for optimal results. For more details see section 4. (5) Sample Thinning - A sample thinning feature (-e) allows advanced users to address parameter correlation issues diagnosed by trace plots (section 4). For more details see section 5. (6) Version Compatibility - An incompatibility with the previous version of SRFYDO was introduced. If you use different SRFYDO versions for different phases of an analysis see section 6.

97 MATHEMATICS AND COMPUTING↗