COMPONENT QUALITY FACTOR AND RELIABILITY MEASUREMENT

Quality factor and reliability measurement can be characterized into three areas - detection, diagnosis and prognosis. Detection uses the most basic form of quality and reliability measurement, where overall reliability or quality is measured on a broadband basis, within a range of 0 - 100% or 0 to 1.0. In machines with low quality or reliability of components, the signal produced by low quality factor (Qi) or reliability coefficient (Ri) may imply incipient defects/failure, which is also proportional to high energy level/age indicating severe defects. This type of measurement information can be useful when used for trending, where an increasing level of quality or reliability is an indicator of the growth of the machine condition and the decrease in quality and reliability is an indication of degradation of machine condition, which eventually leads to failure. Trend analysis involves plotting the quality factor or reliability coefficient values as a function of time, frequency and technological inheritance coefficient, using these to predict when the machine must be taken out of service for repairs. Another way of using the measurement is to compare the optimal condition levels with the real-time data event conditions for different type of equipment. Although broadband measurements provide a good starting point for fault/failure detection in terms of time, it has limited diagnostic capability and although a fault/failure can be identified, it may not give a reliable indication of where the fault/failure is (e.g deterioration, misalignment and unbalance). Where an improved diagnostic capability is required, frequency and technological inheritance analysis can be used. This gives a much earlier indication of the development of fault/failure and also the source or root cause of the fault/failure. Under this condition, the component quality condition parameters (Yi) as well as its corresponding technological inheritance coefficients (ai)/(bi) must be known and analyzed to find out how each quality parameter degrades with time from its optimal value. The real-time data of each quality parameters helps to diagnose the root-cause of faults/failure. The reliability coefficient of the different component can be calculated with the help of technological inheritance coefficients of the different quality parameters and the quality factor of components. Technological inheritance coefficients are used to set threshold set points of the time for failures, apart from for determining the maximum and minimum quality/reliability, which makes measurements to be trended over time and in terms frequency and technological inheritance coefficients. Frequency spectrum analysis plays an important role in detecting and diagnosing machine faults/failures.
When a component starts to deteriorates, the resulting time signal often exhibits characteristics features, which can be used to detect a fault/failure. Also, component condition can rapidly progress from a very small defect to complete failure in a relatively short period of time; so early detection during manufacturing processes requires sensitivity to very small changes in the quality/reliability signatures. The knowledge of quality/reliability growth during manufacturing processes will help to obtain maximum achievable quality/reliability, with multivariate regression model. The results are used as the initial inputs for technological inheritance model, used to monitor and maintain components/equipment. With Technological Inheritance Model, it is possible to carry out quality factor and reliability measurements, for a cost effective component/system reliability monitoring and maintenance under a single platform.