https://www.mdu.se/

When performing gas turbine diagnostics using Gas Path Analysis (GPA), a convenient way of extracting the degradations is by feeding the measured data from a gas turbine to a well-tuned gas turbine performance code, which in turn calculates the deltas on the chosen health parameters matching the measured inputs. For this, a set of measured parameters must be matched with suitable health parameters, such as deltas on compressor and turbine efficiency and flow capacity.

In aero engines, the number of sensors are in general limited due to cost and weight constraints and only the necessary sensors for safe engine operation are available. Some important sensors may have redundancy in case of a sensor loss but it is far from certain that this applies to all sensors available.

If a sensor malfunctions by giving false or no values, the functions using the sensor will be negatively affected in some way causing them to either synthesize a fictive measurement, changing operating scheme, going into a degraded operating mode or shutting down parts or the whole process. If an onboard diagnostic algorithm fails due to sensor faults it will lead to a decrease in flight safety, thus there is a need for a robust system.

This paper presents a strategy for automatic modifications of the gas turbine diagnostic matching scheme when sensors malfunction to ensure a robust function. When a sensor fault is detected and classified as malfunctioning, the gas turbine matching scheme is modified according to predefined rules. If possible, a redundant measurement replaces the faulty measurement. If not, the matching scheme will be modified by determining if any health parameters cannot be derived by the functional set of measurements and remove the least valuable health parameter while maintaining a working matching scheme for the remaining health parameters.

A mixer may be damaged either by cracks or mechanical deformation causing a change in geometry. Only the latter case is, in some cases, possible to detect by shifts in physical measurements such as pressures and temperatures. In general, deformations of the geometry of a mixer due to damage is very hard to identify and quantify with the onboard measurements available, especially for turbofans with high bypass ratio (BPR) where a damaged mixer may cause a slight loss in thrust rather than shifts in measurable quantities.

A special case of mixer damage that may be detected is burn-through of the afterburner liner in low bypass afterburning turbofans. The liner is used to protect the outer casing of the gas turbine to the high temperatures during afterburner operation. For this, the liner need to be continuously cooled by bypass air to withstand the temperatures. A burn-through is generally caused by a local blockage of the cooling path, leading to temperatures the liner cannot withstand. In severe cases it may cause a burn-through of the gas turbine outer casing as well where it may cause a fire in the engine bay.

In this paper, two diagnostic routines are developed to identify a burn-through of an afterburner liner. The diagnostics is intended to be performed as a part of the startup check of the gas turbine to increase the confidence that no burn-through has occurred during the last operation. For these methods a mixer model of high enough fidelity is required, which is described in the paper. The main conclusion is that with enough data it is possible to detect a burn-through but the data collection time is so long that the methods need to be further enhanced to be of any practical use.

In gas turbines used for airplane propulsion, the number of sensors are kept at a minimum for accurate control and safe operation. Additionally, when data are communicated between the airplane main computer and the various subsystems, different systems may have different constraints and requirements regarding what data transmit. Early in the design process, these parameters are relatively easy to change, compared to a mature product. If the gas turbine diagnostic system is not considered early in the design process, it may lead to diagnostic functions having to operate with reduced amount of data. In this paper, a scenario where the diagnostic function cannot obtain airplane installation effects is considered. The installation effects in question is air intake pressure loss (pressure recovery), bleed flow and shaft power extraction. A framework is presented where the unknown installation effects are estimated based on available data through surrogate models, which is incorporated into the diagnostic framework. The method has been evaluated for a low-bypass turbofan with two different sensor suites. It has also been evaluated for two different diagnostic schemes, both determined and underdetermined. Results show that, compared to assuming a best-guess constant-bleed and shaft power, the proposed method reduce the RMS in health parameter estimation from 26% up to 80% for the selected health parameters. At the same time, the proposed method show the same degradation pattern as if the installation effects were known.