Component Adaptive Life Management

• US 20110054806A1
• Filed: 06/07/2010
• Published: 03/03/2011
• Est. Priority Date: 06/05/2009
• Status: Abandoned Application

• Current Assignee: Jentek Sensors, Inc.
• Original Assignee: Jentek Sensors, Inc.
• Inventors: Ziberstein, Vladimir A., Spencer, Floyd W., Goldfine, Neil J., Grundy, David C., Jablonski, David A., Washabaugh, Andrew P., Lyons, Robert J., Sheiretov, Yanko K.
• Application Number:
  US12/795,561
• Publication Number:
  US 20110054806A1
• Time in Patent Office: Days
• US Class Current: 702/34
• CPC Class Codes: G01N 2203/0212 Theories, calculations
• CPC Class Codes: G07C 3/00 Registering or indicating the condition or the working of machines or other apparatus, other than vehicles

Abstract

A framework for adaptively managing the life of components. A sensor provides non-destructive test data obtained from inspecting a component. The inspection data may be filtered using reference signatures and by subtracting a baseline. The filtered inspection data and other inspection data for the component is analyzed to locate flaws and estimate the current condition of the component. The current condition may then be used to predict the component'"'"'s condition at a future time or to predict a future time at which the component'"'"'s condition will have deteriorated to a certain level. A current condition may be input to a precomputed database to look up the future condition or time. The future condition or time is described by a probability distribution which may be used to assess the risk of component failure. The assessed risk may be used to determine whether the part should continue in service, be replaced or repaired. A hyperlattice database is used with a rapid searching method to estimate at least one material condition and one usage parameter, such as stress level for the component. The hyperlattice is also used to rapidly predict future condition, associated uncertainty and risk of failure.

Claims

1. A computer-readable storage medium comprising computer-executable instructions that, when executed by at least one processor, perform a method comprising acts of:

receiving at least two sets of sensor data, each of the at least two sets of sensor data comprising spatial data for a measured material condition of a component;

spatially registering the at least two sets of sensor data with respect to each other and the component;

computing a change in the material condition of the component from the spatially registered at least two sets of sensor data;

estimating the current condition based at least in part on the change in the material condition; and

predicting a future condition of the component at a future time based at least in part on the estimated current condition, the future condition of the component being predicted using a database comprising a plurality of precomputed material conditions of the component, each precomputed material condition computed for a respective operating condition and time;

the method further comprising generating the precomputed material conditions using a flaw growth model;

storing the precomputed material conditions in the database;

filtering at least one of the at least two sets of sensor data with at least one flaw signature; and

estimating a current condition of the component from the filtered sensor data, and a stored database of historical sensor data from a simple element selected to represent the component material and flaw growth behavior.

2. The computer-readable storage medium of claim 1, wherein each of a plurality of data points in the database are generated for a respective combination of an equivalent number of fatigue cycles and stress level.

3. The computer-readable storage medium of claim 1, wherein the stored database comprises at least one coupon test representative of the component.

4. The computer-readable storage medium of claim 1, wherein predicting the future condition of the component comprises estimating a service loading experienced by the component based at least in part on the change.

5. The computer-readable storage medium of claim 1, wherein the act of estimating the current condition comprises using the change in material properties and archived data, the archived data relating flaw size to sensor signal data.

6. A method of predicting a risk of failure before a future time, the method comprising:

inspecting a feature of a component using a non-destructive testing (NDT) method, wherein the NDT method is performed at a plurality of inspection times at a plurality of locations on the component, the NDT method producing inspection data for the plurality of locations at each of the plurality of inspection times;

storing the inspection data in a computer-readable storage medium;

operating at least one processor to determine, based at least in part on the inspection data, if a damage feature is growing within the component and, when insufficient information exists to reliably detect the damage feature using the inspection data at one of the inspection times, generating an enhanced response from the inspection data at two or more of the inspection times and using a precomputed database with two or more dimensions as a function of sensed condition and usage that is searched to determine the risk of failure before the future time.

7. The method of claim 6, wherein the enhanced response is a function derived from sensor data at a same location on the component for each of the two or more inspection times.

8. The method of claim 7, wherein data at other locations is also used to formulate the function to generate enhanced response.

9. The method of claim 8, wherein a signature library is used to derive the enhanced response.

10. The method of claim 6, wherein the NDT method is an eddy current array.

11. The method of claim 6, wherein the plurality of locations comprise a fatigue critical location on the component, and usage is measured in equivalent fatigue cycles, and the database is generated using a damage evolution model.

12. The method of claim 6, wherein:

the plurality of inspection times comprises a first and second time, the first time being a time before the component is put into service and the second time being a time after the component is put in service, and

the inspection data comprises first inspection data taken at the first time and second inspection data taken at the second time.

13. The method of claim 6, wherein the flaw is a crack in the component and it is determined that insufficient information exists to reliably detect the crack using the inspection data at one of the plurality of inspection times if a probability of detecting the crack is below a threshold.

14. The method of claim 13, wherein the threshold is about 90 percent probability of detection, and failure comprises a size of the crack reaching a critical crack size.

15. A method of predicting a future time at which a critical damage level will be reached, the method comprising:

inspecting a feature of a component using a non-destructive testing (NDT) method, wherein the NDT method is performed at a plurality of inspection times at a plurality of locations of the component, the NDT method producing inspection data for the plurality of locations at each of the plurality of inspection times;

storing the inspection data in a computer-readable storage medium;

operating at least one processor to determine, based at least in part on the inspection data, if a damage feature is growing within the component and, when insufficient information exists to reliably detect the damage feature using the inspection data at one of the plurality of inspection times, to generate an enhanced response from the inspection data at two or more of the plurality of inspection times and using a database generated using a model of damage evolution to predict the probability distribution of future times at which the critical damage level will be reached.

16. The method of claim 15, wherein the future time is measured in equivalent fatigue cycles, the damage level is crack size, the critical damage level is a critical crack size, and the location of interest on the component is a fatigue critical location.

17. A method for tracking process progression comprising:

operating a processor to;

process a first inspection image of a component to rank an original plurality of locations of the component, wherein the original plurality of locations are ranked based on a quantitative measure that correlates with predicted crack growth rate at the respective location;

filter a second inspection image of the component using filters devised from signatures to suppress indications that are not representative of a condition of interest and enhance the response of conditions that are more likely to represent the condition of interest; and

re-rank the original plurality of locations using the filtered second inspection image including additional highly ranked locations,

wherein the second inspection image is acquired at a later time than the first inspection image.

18. The method of claim 17, further comprising:

extracting the signatures from a representative fatigue test article; and

storing the signatures in a signature library on a computer-readable storage medium.

19. The method of claim 17, wherein the quantitative measure is a peak conductivity at the respective location.

20. The method of claim 17, wherein the quantitative measure is a half-height width of conductivity dip at the respective location.

21. The method of claim 17, wherein the quantitative measure is a peak filter response at the respective location.

22. The method of claim 17, wherein a fatigue progression track is identified based on a statistically significant trend in non-destructive test (NDT) data from two or more inspection times, where the top ranked location is maintained for each inspection until the end of the component'"'"'s useful life.

23. The method of claim 22, wherein multiple NDT passes are recorded at each inspection time.

24. A computer-readable storage medium comprising computer-executable instructions that, when executed by at least one processor, perform a method comprising acts of:

spatially registering a baseline response of a component and a current response of the component;

after the spatially registering, subtracting the baseline response from the current response; and

estimating a probability distribution of a current condition of the component by applying a statistical analysis using a set of responses from one or more simplified elements.

25. The computer-readable storage medium of claim 24, wherein:

the baseline response is a digital non-destructive test (NDT) image from an inspection performed prior to deployment of the component,

the current response is a digital NDT image from an inspection performed after deployment and during a service life of the component.

26. The computer-readable storage medium of claim 24, wherein estimating the probability distribution of the current condition of the component comprises estimating a probability density function.