The invention relates to a method for training machine learning models, having the steps of: detecting data in the form of time series data using one or more computers, said data being obtained by means of one or more measuring devices (60-62), in each case in the form of a sensor for measuring a physical variable; receiving multiple classification data units relating to the data using the one or more computers; receiving a selected part of the data using the one or more computers for each of the classification data units; and training multiple machine learning models using the one or more computers, in each case on the basis of at least one of the classification data units and the at least one corresponding selected part of the data, wherein the multiple machine learning models represent multiple instances of the same machine learning model.

An aircraft engine, has: a shaft rotatable about a central axis and engaged at an end thereof to a rotatable load via splines; a reference tube extending around the shaft and having a first end secured to the shaft and a second end free relative to the shaft for measuring a deformation of the shaft, the reference tube defining at least one tube aperture; an oil nozzle defining an exit flow axis intersecting the at least one tube aperture, the shaft defining at least one shaft aperture through the shaft, an oil flow path extending from the oil nozzle to the splines; and a drain outlet located radially outwardly of an inlet of the at least one shaft aperture for outputting excess oil out of an annular gap defined between the shaft and the reference tube.

Methods, systems, and assemblies for operating an aircraft having a turboprop engine are described. The method comprises obtaining a measured torque of the turboprop engine from a mechanical torque piston measurement system and displaying the measured torque in a cockpit of the aircraft. During operation of the turboprop engine, a torque measurement saturation condition of the torque piston measurement system is monitored to detect when the torque piston measurement system is outside of an operating range. A synthesized torque of the turboprop engine is determined based on one or more actual engine operating parameters of the turboprop engine. In response to detecting that the torque piston measurement system is outside of the operating range, the synthesized torque is displayed in the cockpit of the aircraft.

A measurement system for an aircraft gas turbine engine includes a probe and a heated-gas source in fluid communication with the pressure probe. The probe includes a probe body defining an internal cavity of the probe. The probe further includes a plurality of sensor inlet ports extending through the probe body and configured to receive a sensed fluid flow. The probe further includes a plurality of probe conduits. Each probe conduit of the plurality of probe conduits is coupled to a respective sensor inlet port of the plurality of sensor inlet ports and extending from the respective sensor inlet port to an exterior of the probe body. The heated-gas source is configured to supply a heated gas flow to one or both of: the plurality of sensor inlet ports via the plurality of probe conduits and an interior of the probe body outside of the plurality of probe conduits.

Systems, program products, and methods for detecting thermal stability within gas turbine systems are disclosed. The systems may include a computing device(s) in communication with a gas turbine system, and a plurality of sensors positioned within or adjacent the gas turbine system. The sensor(s) may measure operational characteristics of the gas turbine system. The computing device(s) may be configured to detect thermal stability within the gas turbine system by performing processes including calculating a lag output for each of the plurality of measured operational characteristics. The calculated lag output may be based on a difference between a calculated lag for the measured operational characteristics and the measured operational characteristic itself. The calculated lag output may be also be based on a time constant for the measured operational characteristics. The computing device(s) may also determine when each of the calculated lag outputs are below a predetermined threshold.

There is provided a system and method for detecting excess flow in an engine fluid system, the method comprising sensing a temperature of a fluid flowing in a fluid line of the fluid system, the fluid line located downstream of a fluid flow restrictor configured to receive the fluid from a source upstream thereof and to flow the fluid from the source into the fluid line downstream thereof, comparing the temperature to a temperature threshold, and when the temperature is beyond the temperature threshold, detecting excess flow of the fluid in the fluid line and outputting an excess flow indication accordingly.

A method of identifying a fuel contained in a fuel tank of an aircraft and arranged to power a gas turbine engine of the aircraft is performed by processing circuitry of the aircraft and includes: obtaining at least one fuel characteristic of any fuel already present in the fuel tank prior to refuelling; determining at least one fuel characteristic of a fuel added to the fuel tank on refuelling; and calculating at least one fuel characteristic of the resultant fuel in the fuel tank after refuelling. The method may further controlling the propulsion system of the aircraft based on the calculated at least one fuel characteristic of the resultant fuel in the fuel tank after refuelling.

Provided is a corrosive environment monitoring device capable of monitoring the condition of a turbine for a long period of time without corrosion damage to a sensor caused by turbine steam. A corrosive environment monitoring device 10 includes: a steam extraction part 11 that extracts steam from inside of a casing 21 of a steam turbine 22 to outside thereof; a condensed water storage part 12 that stores therein condensed water produced by condensation of steam passing through the steam extraction part; and a corrosion factor sensor part 13 that detects properties of the condensed water. The condensed water storage part 12 includes a gap simulation part that simulates a gap inside the turbine and has a predetermined gap capable of storing the condensed water therein, and an annular channel formed on an outer periphery side of the gap simulation part. The corrosion factor sensor part 13 includes one or more sensors capable of measuring the properties of the condensed water, and a switching part configured to be capable of switching between contact and non-contact between the one or more sensors and the condensed water stored in the gap simulation part.

The present application provides a hazardous gas detection system to determine hazardous gas concentrations and/or temperatures within a flow of exhaust air in an exhaust duct of a gas turbine compartment. The hazardous gas detection system may include one or more sensors positioned within or in communication with the exhaust duct and a static mixer positioned upstream of the one or more sensors to promote mixing of the flow of exhaust air.

A fuel flow measuring system includes an ultrasonic fuel flow sensor. The fuel flow sensor includes a first transducer and a second transducer. The first transducer is excited at multiple different excitation frequencies and a voltage, an electric current, and a phase difference between the voltage and the electric current is sensed at the first transducer during excitation. Data points are generated based on the sensed readings and a model is fit to the data points to determine a complex impedance spectrum. The complex impedance spectrum indicates a range of excitation frequencies within a range of a peak resonance frequency of the first transducer. One or more characteristics of excitation signals directed to the second transducer are set based on the determined complex impedance spectrum. In this manner, the signal to noise ratio of ultrasonic signals emitted by the second transducer and received by the first transducer can be maximized.