Projects that have been performing .....

SPONSORED PROJET LIST

Effective October, 1999

PROPOSALS TITLE PROJECT MANAGER

QRE99-001               Product Testing and Process Control                      Susan L. Albin
                                   In a Multistage Production System

QRE99-002               Reliability Prediction Based on Degradation         David W. Coit
                                   With Multiple Changing Stresses

QRE99-003               Optimal Allocation of Stress Levels and                 Elsayed A. Elsayed
                                   Sample Sizes in Accelerated Life Testing

QRE99-007               Process Capability Indices for Normal and             Douglas C. Montgomery and J. Bert Keats
                                  Nonnormal

QRE99-008               Hybrid Acceptance Sampling Plans                        Doug Montgomery and  Connie M. Borror

DESCRIPTION OF PROJECTS

PROJECT NO.: QRE 99-001

PROJECT NAME: Product Testing and Process Control in a Multistage Production System
PROJECT MANAGER: Susan L. Albin

DESCRIPTION: When a product is the result of a multistage production process, it is desirable to identify problems in the early stages and safely evaluate the product with a minimum of highly costly final product testing. The purpose of this proposal is to develop a feasible and economic strategy for evolving from a policy of heavy dependence on final product testing to one based more on process control in each stage and intermediate product testing. The central concept of the model is that the yield of the final product depends on (1) the product variance and yield in earlier stages; (2) the transmission of variance from earlier to later stages and (3) the interaction of process conditions among various stages. The model we propose would result in less costly product testing, improved product yield, and insights into quality improvements regarding the interaction of production process stages.

PROJECT NO.: QRE 99-002

PROJECT NAME: Reliability Prediction Based on Degradation with Multiple Changing Stresses
PROJECT MANAGER: David W. Coit

DESCRIPTION: For highly reliable systems and components, failures can be rare; however, the repercussions of failure may be quite severe. Degradation modeling represents an opportunity to observe and analyze performance deterioration and to predict reliability prior to the occurrences of failure. Parametric drift or deterioration is modeled as a dynamic (time-variant) probability distribution. Reliability predictions are made by considering shifts in the distribution compared to a failure threshold. This can be a particularly useful and efficient for systems that are exposed to prolonged storage, or are highly reliable. In this project, currently available methods will be extended to consider changing stress levels and uncertain stress conditions.

PROJECT NO.: QRE 99-003

PROJECT NAME: Optimal Allocation of Stress Levels and Sample Sizes in Accelerated Life Testing
PROJECT MANAGER: Elsayed A. Elsayed

DESCRIPTION: Accelerated life testing (ALT) is a widely used approach that estimates reliability of components or systems at normal operating conditions using data obtained at accelerated condition. The accuracy of reliability estimates depends on the stress levels and the corresponding sample sizes. In this proposal, we intend to investigate efficient and practical approaches for determining the appropriate stress levels and sample sizes of test units so that the reliability estimates meet a specified significance level. The approaches will determine the optimal stress levels and sample size at each level sugject to a predetermined test time and a significance level of reliability estimate. We will utilize the models for reliability estimates obtained under earlier funding by the Center.

PROJECT NO.: QRE 99-007

PROJECT NAME: Process Capability Indices for Normal and Nonnormal
PROJECT MANAGER: Douglas C. Montgomery and J. Bert Keats

DESCRIPTION: The objective of this research project is to develop a statistically valid procedure and an associated software module for assessing process capability. We will explore both normal and non-normal distributions. The software package will (1) assess the stability of the process data, (2) check the assumption of normality, (3) compute point estimates of the appropriate capability (or potential) indices, (4) construct confidence intervals on the indices, and (5) present appropriate sypothesis tests on the mean and variance. The original research component of this project is in determining appropriate confidence intervals for capability indices in the nonnormal case and in selecting appropriate statistical tests for the process parameters for nonnormal data.

PROJECT NO.: QRE 99-008

PROJECT NAME: Hybrid Acceptance Sampling Plans
PROJECT MANAGER: Douglas C. Montgomery and Connie M. Borror

DESCRIPTION: Many companies employ acceptance-sampling techniques as a part of the total quality program. The standard acceptance sampling plans address either attributes or variables, but not both. There is a need for a sampling scheme that can incorporate components of both attribute data and variables data that will not require 100 percent inspection. Those companies currently using attribute criteria can use variable criteria to gain more valuable information with the software to be developed.

Top