Diagnosability Modeling and Diagnosability Evaluation of Mechanical System at Conceptual Design Stage

Abstract

Reliability, availability and maintainability (RAM) of mechanical system plays a significant role in decreasing the life cycle cost of product. It is well known fact that reliability and maintainability of mechanical systems depends to a large extent on diagnosability of mechanical system, i.e., systems with higher diagnosability, not only helps in reducing the failure of the system at operational stage, but will also help in identification and isolation of faulty components of the system in quick possible time. However, to reap maximum benefits of it is inevitable to consider the design for diagnosability of mechanical systems right from conceptual design stage. For the implementation of the same, designer should be provided with suitable and efficient tool for carrying out diagnosability design and evaluation at conceptual design stage. In this paper, a methodology is developed for carrying out diagnosability of mechanical systems at conceptual design stage. At conceptual design stage designer is well versed with the functional and structural hierarchies of the product. The relationship between the components and functions are used to determine the relationship between the components/assemblies and their performance monitoring parameters (PMP). These are modeled in terms of fuzzy diagnosability bipartite graph (FDBG). FDBG is transformed into fuzzy diagnosability matrix (FDM). The FDM represents the fuzzy relationship between the components and PMP. Diagnosability of the system design concept is evaluated in terms of diagnosability index (IDN) which is obtained from the normalized matrix by using −plog2p. First of all the system which has lowest diagnosability is evaluated from normalized FDM (derived by considering relationship value of 1 between components and PMP). Then reference diagnosability value is obtained from this matrix by using −plog2p and this is represented by DP, ref. FDM for design concept is obtained from the relationship between the components and the PMP. DPN of the design concept is also obtained. Diagnosability index (IDN) of the design concept is then obtained by the ratio of (DP, ref − DPN)/ DP, ref. The higher value of IDN means higher diagnosability and its lower value means lower diagnosability. Component which has lowest diagnosability is also identified, i.e., which has highest functional relationship. This provides guidelines to designer not only for the improvement of design from diagnosability point of view, but also helps in evaluation of dignosability at system conceptual design stage.