FAULT TREE ANALYSIS
Reliability engineering, in a nutshell is concerned with failures. How failures occur, how often they occur, and how to minimize them are the basic problems. It can be seen very easily that each and every failure has a cause behind it. It is an effect of some change in the parameter of a component or system. It may be a result of poor material, wrong manufacturing technique, improper design or the presence of some other unwanted signal in the system. Finding the real or root cause of a failure is one of the most important aspects of reliability engineering. This is a real challenge to the engineers associated with reliability.
The fault - tree technique as analyse complex systems. It is based upon deductive reasoning. This provides a concise and orderly description of the various combinations of possible occurrence within the system, which can result in a predetermined critical output event.
A fault tree is a diagrammatic representation of all possible fault events, their logical combinations, and their relationship to the system failure. The system fault is normally represented at the top of the tree and faults at the lowest, basic level at the bottom. The events at the lowest level are known as "basic events". Events resulting from the combination of basic events are known as "intermediate events". The failure probabilities of basic events are combined to obtain the failure probabilities of intermediate events and finding the top event, the system failure. Logical "AND" and "OR" gates are used to combine two or more events
Fault tree can be used for both qualitative as well as quantitative reliability analysis. In qualitative analysis, fault - tree can help in tracing the cause of a specific failure and can effectively locate the root of the basic problem. In quantitative analysis, fault tree can help in evaluating the probability of occurrence of a particular event from all the possible failure causes. Probability of occurrence of basic events are evaluated from previous reliability data and are combined using AND and OR gates to get the higher events probability.
Fault tree methods can be applied in the early design phase, and then progressively refined and updated as the design evolves to track the probability of an undesired event. Initial fault tree diagrams might represent functional blocks, becoming more definitive at lower levels as the design materializes in the form of specific parts and materials.
FAULT TREE CONSTRUCTION
The fault tree itself is a graphical model of the combinations of various parallel and sequential fault events that will result in the occurrence of the undesired event. The faults can be events associated with component hardware failures or human errors, any other pertinent events which can lead to the undesired event.
Before constructing the fault tree we must have a detailed functional block diagram of the system. The functional block diagram of the system and equipment must clearly define the critical failure mode in terms of the system level malfunction or symptom to be avoided.
The fault-tree logic diagram is constructed relating all possible sequences of events whose occurrence would produce the undesired events identified in the functional block diagram. The fault tree should depict the paths that lead to each higher level in the functional configuration.
In some cases it may be necessary to construct fault trees for each possible operating mode, function or mission event in the mission sequence. The accuracy of a fault tree is highly dependent on the analyst. Unlike the parts list available in FMECA there is no means of checking back to make sure that a significant potential failure contribution has not been overlooked.