System and method for evaluating components using thermography
AssigneePratt & Whitney Canada Corp.
InventorsGuy Joel Rocher, Jing Fang
Life of this patent
Estimating the term…
Abstract
Systems and methods for non-destructive evaluation of components using infrared thermography are provided. A method includes mechanically exciting the component, and using an infrared sensor to acquire a thermographic image of part of the component containing a defect. Using a machine learning algorithm, a health condition of the component is determined based on the new thermographic image. The machine learning algorithm is trained using machine learning and historical data associating previous thermographic images with previous remaining service lives for the component.
Claims20
A method for quantifying a remaining service life of a component containing a defect using non-destructive infrared thermography, the method comprising:
- mechanically exciting the component to induce a thermal response in the component;
- acquiring at a computer from an infrared sensor a thermographic image of a part of the component containing the defect and taken while the component exhibits the thermal response;
- executing a machine learning algorithm to process the thermographic image to determine the remaining service life of the component based on the thermographic image;
- and
- generating an output indicative of the determined remaining service life of the component,
- wherein the machine learning algorithm has been trained using machine learning on historical data associating previous thermographic images with previous remaining service lives for the component.
A method of manufacturing an aircraft power plant, the method comprising:
- manufacturing a component of the aircraft power plant;
- while the component is uninstalled from the aircraft power plant:
- mechanically exciting the component to induce heating in a part of the component containing a defect;
- acquiring a thermographic image of the part of the component;
- and
- using a machine learning algorithm, relating the thermographic image to a health condition of the component;
- and
- when the health condition is indicative of the component being suitable for service, installing the component in the aircraft power plant.
A system for non-destructive evaluation of a component using infrared thermography, the system comprising:
- a transducer operable to induce a vibration in the component;
- an infrared sensor operable to acquire a new thermographic image of the component;
- one or more data processors;
- and
- non-transitory machine-readable memory storing:
- a trained model trained using machine learning and historical data associating previous thermographic images with previous health conditions of the component;
- and
- instructions executable by the one or more data processors and configured to cause the one or more data processors to:
- using the trained model and the new thermographic image, assign a health condition to the component based on the new thermographic image;
- and
- generate an output indicative of the health condition of the component.
Dependencies are inferred from the claim text.