Product and Process Improvement Using Mixture-Process Variable Designs and Robust Optimization Techniques
The quality of an industrial product depends on the raw material proportions and the process variable levels, both of which need to be taken into account in designing a product. This article presents a case study from the food industry in which both kinds of variables were studied by combining a constrained mixture experiment design and a central composite process variable design. Based on the natural structure of the situation, a split-plot experiment was designed and models involving the raw material proportions and process variable levels (separately and combined) were fitted. Combined models were used to study: (i) the robustness of the process to variations in raw material proportions, and (ii) the robustness of the raw material recipes with respect to fluctuations in the process variable levels. Further, the expected variability in the robust settings was studied using the bootstrap.
- Research Organization:
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- AC05-76RL01830
- OSTI ID:
- 960306
- Report Number(s):
- PNNL-SA-53166; TRN: US200923%%312
- Journal Information:
- Journal of Quality Technology, 41(2):181-197, Vol. 41, Issue 2
- Country of Publication:
- United States
- Language:
- English
Similar Records
Stretching the limits of forming processes by robust optimization: A demonstrator
Effect of process variables on the quality attributes of briquettes from wheat, oat, canola and barley