An Integrated Fuzzy MCDM-Based FMEA Approach for Risk Prioritization of Casting Defects in Electro-Pneumatic Brake Units of EMU, MEMU, and DMU Coaches

In railways, electric multiple unit (EMU), mainline electric multiple unit (MEMU), and diesel multiple unit (DMU) coaches are extensively used in the transportation of passengers for short distances. These coaches require frequent stopping and starting, and thus a reliable and robust braking system is essential. The electro-pneumatic (EP) brake systems are installed in these coaches. The brake unit of this EP brake system is fitted under every coach, and upon receiving the signal from the brake controller from motorman’s cabin, this unit plays a major role for the application of brake. Most of the subassemblies of this brake unit is casted by sand-casting process. However, the success rate of the sand-casting process of these subassemblies is around 10–25%, which not only incurs a huge financial burden to the organization, but also responsible for delayed delivery of the brake units. In this work, failure modes and effects analysis (FMEA) is performed to identify the most critical casting defect, their causes, and possible solutions to eliminate the defects. The traditional FMEA approach has been criticized due to its multiple drawbacks. Thus, to overcome those drawbacks, this study proposes an integrated fuzzy multi-criteria decision-making approach (fuzzy MCDM) for the risk ranking of the casting defects. Buckley’s fuzzy analytic hierarchy process (fuzzy AHP) is used for calculating the fuzzy relative importance of the risk factors. Then fuzzy technique for order of preferences by similarity to ideal solution (fuzzy TOPSIS) is used for risk ranking of the casting defects. The reason for incorporating the fuzzy numbers is that the experts linguistically evaluated the risk factors and the failure modes. Fuzzy number is considered as a potential approach to overcome the inherent vagueness and uncertainty associated with the linguistic judgments. Finally, the obtained risk ranking result is validated by performing a sensitivity analysis.