Kalman filters are widely used in the turbine engine community for health monitoring purpose. This algorithm has proven its capability to track gradual deterioration with a good accuracy. On the other hand, its response to rapid deterioration is either a long delay in recognising the fault, and/or a spread of the estimated fault on several components. The main reason of this deficiency lies in the transition model of the parameters that is blended in the Kalman filter and assumes a smooth evolution of the engine condition. This contribution reports the development of an adaptive diagnosis tool that combines a Kalman filter and a secondary system that monitors the residuals. This auxiliary component implements a Generalised Likelihood Ratio Test in order to detect and estimate an abrupt fault. The enhancement in terms of accuracy and reactivity brought by this adaptive Kalman filter is highlighted for a variety of simulated fault cases that may be encountered on a commercial aircraft engine.

This paper extends previous work on model order reduction based on singular value decomposition. It is shown how the decrease in estimator variance must be balanced against the bias on the estimate inevitably introduced by solving the inverse problem in a reduced order space. A proof for the decrease in estimator variance by means of multi-point analysis is provided. The proof relies on comparing the Cramer-Rao lower bound of the single point and the multi-point estimators. Model order selection is discussed in the presence of a varying degree of a priori parameter information, through the use of a regularization parameter. Simulation results on the SR-30 turbojet engine indicate that the theoretically attainable multi-point improvements are difficult to realize in practical jet engine applications.

Least-squares-based methods are very popular in the jet engine community for health monitoring purpose. In most practical situations, the number of health parameters exceeds the number of measurements, making the estimation problem underdetermined. To address this issue, regularisation adds a penalty term on the deviations of the health parameters. Generally, this term imposes a quadratic penalisation on these deviations. A side-effect of this technique is a relatively poor isolation capability. The latter feature can be improved by recognizing that abrupt faults impact at most one or two component(s) simultaneously. This translates mathematically into the search for a sparse solution. The present contribution reports the development of a fault isolation tool favouring sparse solutions. It is very efficiently implemented in the form of a quadratic program. As a validation procedure, the resulting algorithm is applied to a variety of fault conditions simulated with a generic commercial turbofan model.

The computation time and the extraction of useful information remain severe drawbacks to systematic use of modern three-dimensional Navier-Stokes codes in a design procedure of multi-stage turbomachines. That explains why throughflow simulation is still widely used at industrial scale. The main limitation of throughflow is however the need for empirical models to reproduce blade-flow interactions and major 3D flow features. The purpose of this work is to investigate the degree to which empiricism could be reduced by using the averaged-passage equations of Adamczyk, combined with a harmonic closure strategy. To that aim, results of a computation performed with a steady three-dimensional Navier-Stokes code are used to calculate some of the additional terms of the circumferentially-averaged equations, the so-called circumferential stresses. The importance of the latter to bring back the mean effect of circumferential non-uniformities, linked to 3D phenomena, is illustrated by injecting them into a throughfow simulation. Then the ability of truncated Fourier series to reproduce the level of non-uniformity in the core flow and near the walls is detailed. It is finally shown that the harmonic approximated stresses can lead to a good reproduction of local 3D flow features in throughflow simulation and to a better accuracy.

In this contribution a quasi-one-dimensional tool for stationary and transient simulations of post-stall flows in multistage axial compressors is presented. An adapted version of the 1D Euler equations with additional source terms is discretized with a finite volume method and solved in time by a fourth-order Runge-Kutta scheme. The equations are discretized at mid-span both inside the blade rows and the non-bladed regions. The source terms express the blade-flow interactions and are estimated by calculating the velocity triangles for each blade row. Several loss and deviation correlations are implemented and compared to experimental data in normal flow, stalled flow and reversed flow regions. Transient simulations are carried out and a parameter study is presented to analyze the shape of the surge cycles and the frequency of the surge oscillations. At last, a bleeding control strategy is implemented to study the recoverability of the instabilities in a compression system.