Failure prediction model, degradation

In the present scenario, we assume that a CMS is installed on the following items gearbox, pitch system and yaw system (the main downtime drivers) and that their degradation is continuously measured. Coupled with an assumed prediction model, they enable to predict the Time-To-Failure... 

The degradation of the composite laminates can be modelled as simple strength criteria for fibres or matrix, before delamination occurs. Eigure 9.19 shows the test force history for a 200 x 200 mm plate and the various numerical EE (77) predictions (Davies et al. [42]), The hnear elastic case does get the time of the event but underestimates the peak force by a factor of 3. The damaged EE prediction overestimates the force but does get the departure from the linear solution correctly. The fuUy degraded solution (with much fibre failure) does match the experimental history. Delamination was confined to one interface near the mid-plane as the C-scan image indicates. 

The time-to-failure of a structure or its components is an important issue for structural safety considerations in fire. On the basis of the strength degradation models for FRP materials under elevated and high temperatures developed in Chapter 5, the time-to-failure is predicted for GFRP tubes and laminates under both thermal and mechanical loading in compression. Temperature responses were again calculated using the thermal response model presented in Chapter 6. 

As an input for this model, one has to understand the fundamentals of performance losses and degradation. Specific experimental data is needed, to correlate degradation and deterioration phenomena to operating conditions for stationary applications, and identify the paths leading to failure phenomena. This experimental part is an effort to obtain data on degradation quantitatively and reproducibly, since the understanding of kinetics of various degradation processes are the key of final performance and lifetime prediction. 

In addition to the operating conditions parameters, it may be considered to include also the indicators, except indicator i, in the model input of the indicator i. Obviously, model prediction will be better but it may be hazardous for the detection process. Indeed, failures detection is based on residual value as explained in equation (2). Therefore, the model has to be as independent as possible of degradation effect in order to maximize the residual in case of engine degradation. If other indicators are used to assess indicator i, the model estimation will take into account the degradation effect, indirectly introduced by other indicators. Indeed, indicators have been chosen to be as significant as possible of failures. 

By using thresholds and a decision procedure, fault diagnosis can provide an indication of the onset of an incipient fault that may lead to a failure. If a model of the anticipated degradation of parameters is available, the time evolution of monitored parameters or features can be extrapolated. The time from the starting point of a fault until the projected parameter value intersects with a user defined alarm threshold is then an estimation of the remaining useful life. The prediction of the RUL is affected by uncertainties in the monitored parameters, by the choice of the degradation model, uncertainties in its parameters and uncertainties in the failure alarm thresholds. As... 

This way, the identified degradation model is directly used to predict the RUL of the faulty component. If multiple faults have been isolated, if for each of them a degradation model has been identified, and if for each fault a failure alarm threshold has been set, then the latter ones can be substituted into the solutions of the degradation models to obtain a RUL for each faulty component. 

Huang et al. (2006) proposed that membrane mechanical failure can be a result of reduction in membrane ductility combined with a mechanical strain excursion caused by constrained drying in the fuel cell. RH-induced strain is not uniform in membranes. Their model prediction indicates that strain around the edge is higher than in the center region, which means that the membrane edge is more vulnerable to mechanical degradation. 

A number of studies have been conducted to investigate how well SLP accomplishes these goals for selected polymer-based materials [1 ]. From these, reasonable estimates of lifetime were obtained as compared to actual values of time to failure observed for outdoor exposures. Some of the predictions covered the entire degradation pathway, for which changes in material properties generally followed a trend similar to that measured during outdoor weathering. The SLP methods employed varied somewhat from study to study and all used measured climate data to test their models. 

With the kinetic parameters known, it is now possible to calculate the degradation in each parsed time interval relative to some reference condition. This leads to a correlation factor between laboratory test conditions and the modeled conditions. Alternatively, one could calculate the absolute amount of degradation in that interval leading to a prediction of the number of years before failure. The advantage of the... 

---