Abstract:
Material selection is a challenging issue in manufacturing processes while the inappropriate selected material may lead to fail the manufacturing process or end user experience especially in high-tech industries such as aircraft and shipping. Every material has different quantitative and qualitative criteria which should be considered simultaneously when assessing and selecting the right material. A weighted linear optimization method (WLOM) in the class of data envelopment analysis which exists in literature is adopted to address material selection problem while accounting for both qualitative and quantitative criteria. However, it is demonstrated the adopted WLOM method is not able to produce a full ranking vector for the material selection problems borrowed from the literature. Thus, an augmented common weight data envelopment analysis model (ACWDEA) is developed in this paper with the aim of eliminating deficiencies of WLOM model. The proposed ACWDEA is able to produce full ranking vector in decision making problems with less computational complexities in superior to the WLOM. Two material selection problems are solved and results are compared with WLOM and previous methods. Finally, the robustness and effectiveness of the proposed ACWDEA method are evaluated through Spearman’s correlation tests.
Machine summary:
"A weighted linear optimization method (WLOM) in the class of data envelopment analysis is adopted to address material selection problem which deals with both qualitative and quantitative criteria, effectively.
Multi-criteria decision making (MCDM) methods are widely-used techniques which can be applied as part of engineering design processes especially in material selection problems (Jahan and Edwards, 2015).
Rao and Davim (2008) utilized TOPSIS and AHP methods to present a procedure which is able to consider infinite number of qualitative and quantitative criteria in material selection problems.
(2014) proposed a weighted linear optimization method (WLOM) in the class of DEA-like models for evaluating efficiency of DMUs which is able to cope with qualitative criteria more precisely than other methods.
Therefore, proposed ACWDEA has two main advantages: (i) improving the discriminating power among DMUs with unity efficiency in order to produce a full ranking vector and (ii) decreasing the number of times which the model should be solved.
The following algorithm is presented to solve the ACWDEA model in order to evaluate the efficiency of materials and even any other type DMUs in MCDM problems: Normalize the performance measures using Equations (19)-(20).
(5) The proposed ACWDEA method can be applied on any other decision making problems as well as material selection problems where it is essential to consider qualitative criteria precisely.
To demonstrate the applicability of the proposed ACWDEA, two material selection problems are borrowed from the literature and represented that the WLOM is not able to produce full ranking vectors."