COMPONENT REMAINING LIFE ESTIMATE VIA DYNAMIC RELIABILITY

摘要

Condition Based Maintenance (CBM) effective Fault Diagnosis and Prognosis is premised on successfully characterizing the operational condition of the component under analysis. In order to achieve this, the component's internal kinematical excitation signal must be extracted from the sensor response signal. This response signal is contaminated with extraneous signals/noise originating at either the internal excitation source point or the transmission path adjacent component contaminants, from which the desired signal characterizing the component's operational condition is extracted. Once the desired signal is extracted then it can be analyzed to determine if an anomaly exists. If an anomaly exists, then the desired signal is utilized to diagnose, isolate and identify, the underlying cause of faulty sub-component operational conditions, such as misalignment, deviant eccentricity, loose coupling, modulations or internal rubbing, which cause variants in the desired signal from nominal to anomalous to faulty. CBM Diagnostic analyses of the desired signal should result in Condition Indicators that are then transformed into dimensionless statistical scores (DSSs). DSSs derived from faulty operational conditions are statistically separability scored against DSS's derived from ideal nominal signals, resulting in a Fault Classification statistical metric. This Fault Classification is accomplished by synthesizing these DSS discriminants into an Operational Health Statistical Score (OHSS). OHSS effects fault diagnosis by tracking machinery operational condition from nominal, to nascent, to degraded, then critical. Therefore, fault detection and diagnosis are sufficient to achieve CBM Diagnostic Health Management (CBM/DHM). However, dynamic failure state tracking from incipient to functional failure is required to achieve Prognosis, which entails tracking the effect of faults and utilization stressors effect on "Remaining Life" estimates throughout the components useful life. CBM Prognosis is the ability to track the "Remaining Life" estimate throughout a dynamic component's useful life, e.g. from current utilization time to projected utilization of next mission pulse cycle time. The Prognostic challenge is how to weight the OHSS with stressors and fault severity effects in order to adjust a component's "Remaining Life" estimate. Consequently, "Remaining Life" estimation is sufficient to achieve CBM Prognostic Health Management (CBM/PHM).