Development of a Rubber-Based Product Using a Mixture Experiment: A Challenging Case Study

Abstract

Many products used in daily life are made by blending two or more components. The properties of such products typically depend on the relative proportions of the components. Experimental design, modeling, and data analysis methods for mixture experiments provide for efficiently determining the component proportions that will yield a product with desired properties. This article presents a case study of the work performed to develop a new rubber formulation for an o-ring (a circular gasket) with requirements specified on 10 product properties. Each step of the study is discussed, including: 1) identifying the objective of the study and requirements for properties of the o-ring, 2) selecting the components to vary and specifying the component constraints, 3) constructing a mixture experiment design, 4) measuring the responses and assessing the data, 5) developing property-composition models, 6) selecting the new product formulation, and 7) confirming the selected formulation in manufacturing. The case study includes some challenging and new aspects, which are discussed in the article.

Keywords

Rubber formulation Product development Experimental design Mixture experiment Mixture models Optimization

Get full access to this article

View all access options for this article.

References

Piepel

G.F.

, Mixture experiments. Encycl. Stat. Qual. Reliabil., John Wiley & Sons, Ltd., Chichester, UK (2007) 1099–1112.

Piepel

G.F.

, and Cornell

J.A.

, Mixture experiment approaches: examples, discussion, and recommendations. J. Qual. Technol., 26 (1994) 177–196.

Cornell

J.A.

, Experiments With Mixtures: Designs, Models, and The Analysis of Mixture Data. 3rd Edition, John Wiley & Sons, Inc., New York (2002).

Smith

W.F.

, Experimental Design for Formulation, ASA-SIAM Series on Statistics and Applied Probability, American Statistical Association, Alexandria, VA and Society for Industrial and Applied Mathematics, Philadelphia (2005).

Stat-Ease, Design-Expert Version 8, Stat-Ease Inc., Minneapolis, MN (2010).

Minitab, Minitab Version 16, Minitab Inc., State College, PA (2010).

JMP, JMP 9, SAS Institute, Inc., Cary, NC (2011).

Piepel

G.F.

, Szychowski

J.M.

, and Loeppky

J.L.

, Augmenting Scheffé linear mixture models with squared and/or crossproduct terms. J. Qual. Technol., 34 (2002) 297–314.

Draper

N.R.

, and Smith

, Applied Regression Analysis, Third Edition, John Wiley & Sons, New York (1998).

10.

Piepel

G.F.

, 50 years of mixture experiment research: 1955-2004. Chapter 12 in Response Surface Methodology and Related Topics, Khuri

A. I.

, Ed., World Scientific Press, Singapore (2006) 283–327.

11.

Borosy

A.P.

, Quantitative composition–property modelling of rubber mixtures by utilising artificial neural networks. Chemomet. Intell. Lab. Systems, 47 (1999) 227–238.

12.

Piepel

G.F.

, and Landmesser

S.M.

, Mixture experiment alternatives to the slack variable approach. Qual. Engineering, 21, (2009) 262–276.

13.

Anderson

M.J.

, and Whitcomb

P.J.

, Mixture DOE uncovers formulations quicker. Rubber Plastics News, (October 21, 2002) 16–18.

14.

Myers

R.H.

, Montgomery

D.C.

, and Anderson-Cook

, Response Surface Methodology: Process and Product Optimization Using Designed Experiments, 3rd Edition, John Wiley & Sons, New York (2009).

15.

JMP, JMP 7, SAS Institute, Inc., Cary, NC (2007).

16.

Piepel

G.F.

, and Cornell

J.A.

, A catalog of mixture experiment examples, BN-SA-3298, Rev. 16.0, Battelle-Pacific Northwest Division, Richland, WA (2008).

17.

Piepel

G.F.

, Anderson

C.M.

, and Redgate

P.E.

, Response surface designs for irregularly-shaped regions (Parts 1, 2, and 3). 1993 Proc. Section Physical Engin. Sciences, American Statistical Association, Alexandria, VA (1993) 205–227.