Computerized Systematic Approach to Fault Tree Analysis Based on Quantity Dimension Indexing
Fault tree analysis (FTA) is an effective method of ensuring the security and safety of the product by identifying all the possible causes of the problem and fixing them. However, it is not easy for a designer to construct a complete fault tree about various physical phenomena without any misunderstanding or oversight, and some computerized method of managing (i.e., storing, searching and utilizing) knowledge about FTA is needed. To solve the problem, the authors have proposed and studied a method and software tool for knowledge management of FTA based on quantity dimension indexing as a design knowledge management method to avoid ambiguity of literal expression about physical phenomena. In the previous method and software, however, fault values of quantities were limited as just above- and below-normal, and dynamic phenomena such as oscillation could not be described. In this paper, the authors introduce a systematically classified definition of fault values as above/below normal, one-side/both-sides, constant/varying, monotonic/non-monotonic and sudden/gradual, and expand the computerized systematic approach to FTA. Feasibility of the method was examined by applying it to fault tree examples made in a company.