About the project
This PhD project has the potential of significantly improving the fatigue design and maintenance of an aircraft. It is funded by Airbus Operations Ltd through an EPSRC Industrial CASE Award. Fatigue design and fatigue damage estimates are based on many factors such as design and performance assumptions, manufacturing, component material, stress cycle calculations and service conditions.
Of these, the stress cycles and service conditions experienced by the aircraft are related to the loads acting on the aircraft during flight, such as those resulting from the number (and type) of gusts and manoeuvres experienced by the aircraft.
Thus, refined information on the number and severity of gusts could improve the fatigue damage estimates of the aircraft. The hypothesis is that machine learning could be applied to historic datasets of in-service and flight test data to find specific performance clusters and automatically identify or separate the aircraft dynamics due to aircraft control from the dynamics due to gust or environmental factors or phenomenon.
More accurate, in-service and flight-test based, statistical data could then be used for fatigue design to improve damage estimates and ultimately reduce the amount of time an aircraft spends on the ground for inspection of component failures.
Such research is of utmost importance for the aerospace industry and Airbus has a strong interest in developing a maintenance system which makes the aircraft available when needed by the airline at minimum maintenance cost and which opens opportunities for new maintenance business models and services.