PSI - Issue 72

Ruhit Bardhan et al. / Procedia Structural Integrity 72 (2025) 507–519

518

Figure 2. Temperature distribution across the thickness for the top three FGM alternatives

Figure 3. Thermal stress distribution across the thickness for the top three FGM alternatives

4.5. Sensitivity Analysis

In order to evaluate the robustness of the neutrosophic TOPSIS results. We varied the criterion weights in a sensitivity study. While keeping the proportional connections between the other weights, Error! Reference source not found. shows how the closeness coefficients of the alternatives vary when the weight of the most crucial criteria ( 3 : Maximum service temperature) is changed. The sensitivity analysis reveals that the ranking of the top two alternatives ( 3 and 2 ) remains stable across a wide range of weight variations, confirming the robustness of the results. However, the relative positions of 1 and 5 could change if significantly different weights were assigned to certain criteria.

Figure 4. Sensitivity analysis with varying weight of criterion

5. Conclusion In this paper, we represented a comprehensive neutrosophic TOPSIS framework for the multi-criteria selection of functionally graded materials. The framework addresses the unique challenges in FGM selection by incorporating neutrosophic set theory to handle uncertainty, imprecision, and indeterminacy in the evaluation process. The main conclusions derived from this study are: 1. The neutrosophic TOPSIS framework provides a robust methodology for FGM selection problems, capturing the complex nature of gradient specific properties and uncertainties in their evaluation. 2. The three-way representation (truth, indeterminacy, falsity) of neutrosophic sets offers enhanced discrimination power compared to conventional TOPSIS and fuzzy TOPSIS methods, as demonstrated in the comparative analysis. 3. For the high-temperature aerospace application case study, the ZrO2-NiCoCrAlY FGM emerged as the optimal choice, followed by Ti-TiB2 FGM. This ranking remained stable under sensitivity analysis, confirming the reliability of the results.