Specification and allocation of reliability and availability requirements

Our model for allocation of requirements is based on a generalized fault tree approach, where the TOP represents the product to be designed. The other parts of the fault tree represent entities, which affect essentially the failure tendency and the repair time of the product. Relations between parts are modeled by two mechanisms. The "gates" determine the partly logical and partly stochastic propagation of faults (primary states). The "strategies" define other relations between TOP and the deepest entities. A consequence of the strategies is that two types of "waiting" (secondary states) can occur. Customer and/or manufacturer data influences the design of product reliability, availability and repair time. The proposed methods can deal with quite different types of requirements. Requirements related to failure tendency can involve number of failures, time between failures, reliability and availability as a function of age, or data concerning first failure. Requirements related to product's repair time again could involve mean time to repair, standard deviation, minimum repair time (0%), and maximum repair time (with corresponding quantile%). The allocation of the failure tendency of a gate (entity) down to its input entities is guided by assessing "importance" and "complexity". Importance takes into account customer's perspective and complexity represent the technical standpoint. The aim is that the more important an entity is, the less it is allowed to fail, and the more complex an entity is, the more it is allowed to fail. The repair time allocation again is based on a direct assessment of repair time ratios between the input entities. The failure tendency and the repair time of an entity can also be locked, whereas the designer can focus only on the unlocked entities. The requirements for TOP are summarized in two "dependability functions" - one for failure tendency and one for repair time. A stepwise allocation process downward in the fault tree leads gate by gate to equivalent dependability functions for other entities. These functions are in every stage tested via simulation and comparison to TOP requirements. The last simulation confirms the final dependability of entities, especially of those to which attention will be paid in a later design process. The simulation produces also a complete list of events, states of entities, their duration, etc. This "logbook" is of course detailed raw material for various supplemental calculations, conclusions, and even further programming