Risk-based material selection process supported on information theory: A case study on industrial gas turbine

Abstract

Gas turbine blades are produced using high-temperature high-strength materials because they are often exposed to severe environments. Material selection for such cases should be performed with sensitivity and using a systematic method. Moreover, various risks may exist about the values of materials properties. For example, the difference between the designed and unexpected conditions may pose some risks to materials properties. Although considering risks for gas turbine blade material selection problem or similar high-tech practical cases is important, a research gap exists in such fields. This paper presents a risk-based material selection algorithm using the principles of Suh and Shannon entropies supported on information theory. In this regard, we develop the risk-based fuzzy axiomatic design approach with the integrated Shannon significance coefficients of attributes. A real-world example about material selection of industrial gas turbine blade is examined using four techniques including the fuzzy axiomatic design, weighted fuzzy axiomatic design, risk-based fuzzy axiomatic design, and weighted risk-based fuzzy axiomatic design approaches. In the example, risk factors are determined to consider the effect of temperature variation on materials properties. Finally, the resultant rankings are compared by calculating Spearman rank correlation coefficients. The comparisons show that considering risk factors in the problem affects the resultant ranking. We validate the results of the proposed methods by the unweighted and weighted fuzzy MULTIMOORA approaches.