Component Adaptive Life Management

• US 20110060568A1
• Filed: 06/07/2010
• Published: 03/10/2011
• Est. Priority Date: 06/05/2009
• Status: Active Grant

• Current Assignee: Jentek Sensors, Inc.
• Original Assignee: Jentek Sensors, Inc.
• Inventors: Spencer, Floyd W., Grundy, David C., Jablonski, David A., Washabaugh, Andrew P., Sheiretov, Yanko K., Goldfine, Neil J., Zilberstein, Vladimir A., Lyons, Robert J.
• Granted Patent: US 8,494,810 B2
• Time in Patent Office: Days
• US Class Current: 703/6
• 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. An inspection system comprising:

a sensor for collecting first spatial data from a component, the first spatial data comprising a first sensor response for at least one location on the component; and

a computing system comprising;

an inspection archive comprising second spatial data having a second sensor response for the component;

a database comprising a plurality of data points, each data point storing a material condition of the component and a time;

a filtering module to spatially register the first spatial data and the second spatial data;

an estimation module to estimate a current condition of the component based at least in part on the spatially registered first and second spatial data; and

a prediction module to predict a future condition of the component, based at least in part on the current condition, using the database.

2. The inspection system of claim 1, wherein said database is generated offline using a phenomenological model.

3. The inspection system of claim 1, wherein:

the data points span a range of interest for each of the material condition and the time, and

each database data point also includes values for at least two properties to be estimated, the values being generated using a model.

4. The inspection system of claim 3, wherein the material condition is crack size, the time is measured in equivalent cycles, and one of the two values estimated at each data point is remaining life.

5. The inspection system of claim 1, wherein the current condition is described by a probability density function, and wherein the prediction module generates a distribution function describing, probabilistically, the future condition of the component.

6. The inspection system of claim 5, wherein the prediction module predicts the future condition of the component at a next scheduled inspection time, and the computing system further comprises:

a decision module that determines whether the future condition predicted at the next scheduled inspection time has a risk of failure that exceeds a damage tolerance limit.

7. The inspection system of claim 6, wherein if the decision module determines that the risk of failure at the next scheduled inspection exceeds the damage tolerance threshold limit, and in response provides an instruction to replace or repair the component.

8. The inspection system of claim 6, wherein if the decision module determines that the risk of failure at the next scheduled inspection exceeds the damage tolerance limit, then in response the decision module reschedules the next inspection time to an earlier time.

9. 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.

10. The computer-readable storage medium of claim 9, wherein the at least two sets of sensor data comprises three sets of sensor data, and the method further comprises:

computing a second change, of the material condition using the three sets of sensor data,

wherein predicting the future condition is further based on the second change.

11. The computer-readable storage medium of claim 9, wherein the future condition of the component is 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.

12. The computer-readable storage medium of claim 11, wherein predicting the future condition comprises interpolating the future condition at the future time from the precomputed material conditions in the database.

13. The computer-readable storage medium of claim 9, wherein predicting the future condition of the component comprises determining a distribution function describing, probabilistically, the future condition of the component at the future time.

14. The computer-readable storage medium of claim 9, wherein the future time at which the future condition of the component is predicted is a next scheduled inspection time.

15. The computer-readable storage medium of claim 14, wherein the method further comprises determining whether the future condition predicted at the next scheduled inspection time has a probability above a threshold that said future condition may exceed a damage tolerance limit and, if so, providing an instruction to replace or repair the component or to reschedule the next inspection time to an earlier time.

16. The computer-readable storage medium of claim 9, wherein predicting the future condition comprises determining the future time as a time at which the future condition meets a replacement condition for the component.

17. The computer-readable storage medium of claim 16, wherein the future time is measured in equivalent fatigue cycles.

18. The computer-readable storage medium of claim 16, wherein the future condition is a predefined crack size limit.

19. The computer-readable storage medium of claim 9, wherein the method further comprises an act of predicting a probability distribution for a future condition of the component using a probability distribution function for the current condition and a hyperlattice generated using a progression model for a process.

20. The computer-readable storage medium of claim 19, wherein the process is crack initiation and growth.

21. The computer-readable storage medium of claim 19, wherein the process is the evolution of a medical condition.

22. The computer-readable storage medium of claim 19, wherein the process is a manufacturing process.

23. The computer-readable storage medium of claim 19, wherein the process is a motion of a device.

24. The computer-readable storage medium of claim 19, wherein the process is a machining process.

25. A method for generating a hyperlattice, the method comprising acts of:

obtaining calibration information, the calibration information comprising data obtained from sensor measurements; and

operating a processor to perform acts of;

selecting parameters for a model such that the calibration information is predicted by the model; and

generating the hyperlattice using the model as configured with the selected parameters.

26. The method of claim 25, wherein the parameters of the model are selected to minimize least squared error between the calibration information and results predicted by the model.

27. The method of claim 25, wherein the reference part calibration information comprises maximum likelihood information.

28. The method of claim 25, wherein the reference part calibration information comprises first uncertainty distributions.

29. The method of claim 28, wherein selecting parameters comprises selecting second uncertainty distributions for the model parameters so that a predicted uncertainty distribution for a future condition matches the first uncertainty distribution.

30. The method of claim 28, wherein the uncertainty distribution is selected with constructive and destructive cumulative uncertainty from uncertainty sources.

31. The method of claim 30, wherein the uncertainty sources comprise at least one of model input, operating condition, sensor error, ground truth errors in calibration data, ground truth error in recalibration data, and ground truth errors in population data.

32. The method of claim 25, wherein the model is a fracture mechanics model.

33. A computer-readable storage medium comprising computer-executable instructions that, when executed by at least one processor, perform a method for estimating uncertainty of a predicted condition of a component at a future time, the method comprising:

estimating a probability distribution function for a current condition of the component; and

estimating uncertainty of a future condition of the component at the future time by performing a plurality of look-up iterations on a hyperlattice to construct a probability distribution function for the future condition of the component.

34. The computer-readable storage medium of claim 33, wherein the method further comprises:

executing a model to produce a result for each of a plurality of input conditions, each input condition specified by a value for each of at least one input variable, wherein the at least one input variables are each varied over a respective range of values to produce the plurality of input conditions;

forming the hyperlattice, the hyperlattice comprising each of the results in association with the respective input condition.

35. The computer-readable storage medium of claim 33, wherein:

the model is a facture mechanics model for fatigue crack initiation and growth on a metal component, and

the inputs that are varied are initial crack size and material properties, and where the database is configured with two dimensions including remaining fatigue life and a sum of applied and residual stress at a critical location, and where the measured quantities are equivalent cycles and a digital NDT response at two or more times, and where the digital NDT data is first used to estimate the likely crack distribution at a first inspection time during service and then used to predict the a future condition and uncertainty related to crack growth.