Condition evolution of machinery and its assessment from passive diagnostic experiment

Abstract This paper deals with symptom value evolution and distribution when running machinery is observed diagnostically. It is shown that the Pareto distribution can be used. The fit of this distribution depends on symptom choice and maintenance quality. It was also shown that the Paretian symptom reliability curve can be transformed to the average symptom life-curve of a given machine set. The exponent of the Pareto distribution can be easily assessed from diagnostic observation, and from it one can construct a simple graph for condition assessment and forecasting the residual time to breakdown. The findings of the paper were confirmed by data taken from vibration condition monitoring practice.