Title: Analysis of mixture experiments with process variables using strip-plot design

Authors: Bushra Husain; Fariha Aslam

Addresses: Women's College, Department of Statistics and OR, Aligarh Muslim University, Aligarh, UP, India ' Women's College, Department of Statistics and OR, Aligarh Muslim University, Aligarh, UP, India

Abstract: The characteristics of a product are determined by the formulation of the mixture and the conditions under which it is processed. To optimise product quality (or characteristics), the proportions of the components in the combination as well as the process variables are evaluated and compared. The inclusion of difficult-to-change elements might complicate testing by imposing restrictions on randomisation. A suitable design for restricted randomisation may be a strip-plot design. This paper discusses the use of strip-plot design in analysing mixture-process variable experiments by using the three mixture component with two process variables example of vinyl thickness experiment, when there is only one constraint and when more than one restriction is applied to the mixture components. Mixture experiments with time and cost constraints and hard to change process variables, or in the case of additional restriction or where the binary or ternary blends of components and their interaction with process variables are of interest, it will be recommended to use a strip plot design with replications.

Keywords: mixture experiments; process variables; restricted maximum likelihood; REML; strip-plot design; split-plot design.

DOI: 10.1504/IJEDPO.2024.140469

International Journal of Experimental Design and Process Optimisation, 2024 Vol.7 No.2, pp.115 - 134

Received: 17 Jan 2024
Accepted: 30 Mar 2024
Published online: 19 Aug 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article