Acoustic apparatus and inspection methods

• US 6,386,038 B1
• Filed: 11/24/1999
• Issued: 05/14/2002
• Est. Priority Date: 11/24/1999
• Status: Expired due to Fees

• Current Assignee: Accurate Automation Corp.
• Original Assignee: Accurate Automation Corp.
• Inventors: Escher, Daric William, Lewis, Carl Edwin III
• Primary Examiner(s): MOLLER, RICHARD ALAN
• Application Number:
  US09/449,228
• Time in Patent Office: 902 Days
• Field of Search: 735/70, 735/79, 735/82, 735/84, 735/87, 735/88, 735/96, 735/97, 735/98, 735/99, 736/00, 736/02, 736/06, 737/60, 737/88, 702/35, 702/36, 702/39, 340/665, 340/666, 340/669
• US Class Current: 73/587
• CPC Class Codes: G01N 2291/0231 Composite or layered materials
• CPC Class Codes: G01N 29/14 using acoustic emission techniques echo of particles G01N29/046; measuring mechanical vibrations or acoustic waves in solids in general G01H1/00
• CPC Class Codes: G01N 29/227 related to high pressure, tension or stress conditions
• CPC Class Codes: G01N 29/4445 Classification of defects
• CPC Class Codes: G01N 29/4481 Neural networks
• CPC Class Codes: G01N 29/46 by spectral analysis, e.g. Fourier analysis or wavelet analysis spectral signal processing per se G06F17/14

Abstract

The invented apparatus includes at least one acoustic sensor, a mapping unit, a controller, and a transducer. The sensor is coupled to an object to be inspected for damage, and generates a sensed signal based on acoustic signal occurring in the object under an applied load, which signal may or may not be the sound of damage occurring in the object. The mapping unit is coupled to receive the sensed signal from the sensor, and generates damage data based thereon. The controller is coupled to receive the damage data from the mapping unit. Initially, the apparatus is in a passive mode in which the apparatus senses an acoustic signal from the object. If the damage data indicates that the object has incurred possible damage in the passive mode, the controller switches to active mode and outputs induced signal data. The transducer is coupled to the object as well as the controller, and induces acoustic signal in the object, based on the induced signal data. The induced acoustic signal is detected by the sensor, mapped by the mapping unit, and supplied as damage data to the controller, after which the controller returns the apparatus to passive mode. If the damage data resulting from the induced acoustic signal indicates that the object has incurred actual damage, the controller can generate an alarm, a display of the damage, and/or store the damage data in a memory to provide a record of damage incurred by the object. The invention also includes related methods.

Claims

1. A method comprising the steps of:

a) sensing at least one acoustic signal occurring in an object under an applied load;

b) determining whether the object has incurred possible damage under the applied load, based on the acoustic signal sensed in said step (a);

c) inducing at least one acoustic signal in the object if said step (b) indicates that the object has incurred possible damage, the induced acoustic signal encountering at least the portion of the object that has incurred possible damage;

d) sensing the induced acoustic signal from the object; and

e) determining whether the possible damage is actual damage of the object, based on the induced acoustic signal sensed in said step (d).

2. A method as claimed in claim 1, wherein said step (b) includes the substep of:

b1) classifying the acoustic signal sensed in said step (a) as pertaining to one of a plurality of categories including at least possible damage to the object and no damage to the object, based on the acoustic signal sensed in said step (a).

3. A method as claimed in claim 2, wherein said step (b) includes the substep of:

b2) extracting at least one feature from the acoustic signal sensed in said step (a) for use in the performance of the classifying in the substep (b1).

4. A method as claimed in claim 2, wherein said step (b) includes the substep of:

b2) before performing the substep (b1), training a learning system to classify a predetermined set of sensed acoustic signals into the categories including at least possible damage and no damage,

the substep (b1) being performed by the learning system.

5. A method as claimed in claim 4, wherein the learning system includes a neural network.

6. A method as claimed in claim 2, wherein a sensed signal is generated in said step (a), based on the acoustic signal sensed in said step (a), the method further comprising the steps of:

f) sampling the sensed signal to generate time and amplitude data; and

g) converting the time and amplitude into frequency and amplitude data, the substep (b1) performed using the frequency and amplitude data.

7. A method as claimed in claim 2, wherein the object is a composite material including fiber and matrix layers, and wherein the classifying of the substep (b1) is performed to further classify the possible damage to the object into possible fiber damage, possible fiber-matrix interface damage, possible fiber-matrix debonding, and possible matrix damage, said step (b) determining that the object has incurred possible damage if the damage is classified in said step (b1) as possible fiber damage, possible fiber-matrix interface damage, possible fiber-matrix debonding, and possible matrix damage.

8. A method as claimed in claim 7, wherein the substep (b1) further classifies the possible fiber damage into possible damage for respective fiber layers.

9. A method as claimed in claim 2, wherein the object is a monolithic material, and wherein the classifying of the substep (b1) is performed to further classify the possible damage into possible surface cracking, possible necking, possible fatigue, and possible cracking.

10. A method as claimed in claim 1, wherein said step (e) includes the substeps of:

e1) classifying the induced acoustic signal sensed in said step (d) as pertaining to one of a plurality of categories including at least actual damage and no damage.

11. A method as claimed in claim 10, wherein said step (e) includes the substep of:

e2) extracting at least one feature from the acoustic signal sensed in said step (d) for use in the performance of the classifying in the substep (e1).

12. A method as claimed in claim 10, wherein said step (e) further includes the substep of:

e2) before performing the substep (e1), training a learning system to classify a predetermined set of induced acoustic signals into categories including at least actual damage and no damage,

the substep (e1) being performed by the learning system.

13. A method as claimed in claim 12, wherein the learning system includes a neural network.

14. A method as claimed in claim 10, wherein a sensed signal generated in said step (c) based on the induced acoustic signal sensed in said step (d), the method further comprising the steps of:

f) sampling the sensed signal to generate time and amplitude data; and

g) converting the time and amplitude into frequency and amplitude data, the substep (e1) performed using the frequency and amplitude data.

15. A method as claimed in claim 14, further comprising the step of:

h) storing at least one of the amplitude data, time data, and applied load data, in a memory in correspondence with the category of actual damage for the acoustic signal as classified in the substep (e1).

16. A method as claimed in claim 10, wherein the object is a composite material including fiber and matrix layers, and wherein the classifying of the substep (e1) is performed to further classify the possible object damage into actual fiber damage, actual fiber-matrix interface damage, actual fiber-matrix debonding, and actual matrix damage, said step (e) determining that the object has incurred actual damage if the possible damage is classified in said step (e1) as actual fiber damage, actual fiber-matrix interface damage, actual fiber-matrix debonding, and actual matrix damage.

17. A method as claimed in claim 16, wherein the substep (e1) further classifies the actual fiber damage into actual fiber damage for respective fiber layers.

18. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual matrix damage count if said step (e1) classifies the possible damage as actual matrix damage;

g) comparing the actual matrix damage count with a predetermined matrix damage count value; and

h) generating an alarm, based on said step (g).

19. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual fiber damage count if said step (e1) classifies the damage actual fiber damage;

g) comparing the actual fiber damage count with a predetermined fiber damage count value; and

h) generating an alarm, based on said step (g).

20. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual fiber-matrix interface damage count if said step (e1) classifies the possible damage as actual fiber-matric interface damage;

g) comparing the actual fiber-matrix interface damage count with a predetermined fiber-matrix interface damage value; and

h) generating an alarm, based on said step (g).

21. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual fiber-matrix debonding damage count if said step (e1) classifies the possible damage as actual fiber-matrix debonding damage;

g) comparing the actual matrix damage count with a predetermined fiber-matrix debonding damage count value; and

h) generating an alarm, based on said step (g).

22. A method as claimed in claim 10, wherein the object is a monolithic material, and wherein the classifying the actual damage in the substep (e1) into actual surface cracking, actual necking, actual fatigue, and actual cracking.

23. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual surface cracking damage count if said step (e1) classifies the possible damage as actual matrix damage;

g) comparing the actual surface cracking damage count with a predetermined actual surface cracking damage count value; and

h) generating an alarm, based on said step (g).

24. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual necking damage count if said step (e1) classifies the damage actual necking damage;

g) comparing the actual necking damage count with a predetermined actual necking damage count value; and

h) generating an alarm, based on said step (g).

25. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual fatigue damage count if said step (e1) classifies the possible damage as actual fatigue damage;

g) comparing the actual fatigue damage count with a predetermined actual fatigue damage value; and

h) generating an alarm, based on said step (g).

26. A method as claimed in claim 16, further comprising the step of:

f) incrementing an actual cracking damage count if said step (e1) classifies the possible damage as actual cracking damage;

g) comparing the actual cracking damage count with a predetermined actual cracking damage count value; and

h) generating an alarm, based on said step (g).

27. A method as claimed in claim 1, further comprising the step of:

f) generating a display of the actual damage, if any, determined in said step (e).

28. A method as claimed in claim 27, wherein said step (b) determines the location of the actual damage in the object, and wherein the display is generated in said step (f) to include the location of the actual damage.

29. A method as claimed in claim 27, wherein said step (b) determines the extent of the possible damage in the object, and wherein the display is generated in said step (f) to include the extent of the actual damage.

30. A method as claimed in claim 1, further comprising the step of:

f) generating an alarm if the object is determined to have actual damage in said step (e).

31. A method as claimed in claim 1, wherein said steps (a) through (e) are repeatedly performed.

32. A method as claimed in claim 31, further comprising the step of:

f) tracking the amount of actual damage incurred by the object;

g) determining whether the amount of actual damage has exceeded a predetermined amount of damage; and

h) generating an alarm, based on the determination of said step (g).

33. A method as claimed in claim 31, further comprising the step of:

f) incrementing the load applied to the object after each performance of said step (e) and before subsequent performance of said step (a).

34. A method comprising the steps of:

a) sensing at least one acoustic signal occurring in an object under an applied load, and generating a signal indicative of the sensed acoustic signal;

b) sampling the signal indicative of the acoustic signal sensed in said step (a) to generate time and amplitude data;

c) converting the time and amplitude data of said step (b) into frequency and amplitude data;

d) classifying the frequency and amplitude data converted in said step (c), as pertaining to one of a plurality of categories including at least possible damage and no damage;

e) inducing at least one acoustic signal into the object if said step (d) indicates that the object has incurred possible damage, the induced acoustic signal encountering at least the portion of the object that has incurred possible damage;

f) sensing the acoustic signal induced in said step (e), and generating a signal indicative of the sensed acoustic signal induced in said step (e);

g) sampling the signal indicative of the acoustic signal sensed in said step (f) to generate time and amplitude data;

h) converting the time and amplitude data of said step (g) into frequency and amplitude data; and

i) classifying the frequency and amplitude data converted in said step (h), as pertaining to one of a plurality of categories including at least actual damage and no damage.

35. A method as claimed in claim 34, further comprising the step of:

j) extracting at least one feature from the frequency and amplitude data of said step (c) for use in the performance of said step (d).

36. A method as claimed in claim 34, further comprising the step of:

j) extracting at least one feature from the frequency and amplitude data of said step (h) for use in the performance of said step (i).

37. A method as claimed in claim 34, further comprising the step of:

j) training a learning system to perform the classifying of said step (d) with a predetermined set of frequency and amplitude data derived from acoustic signals and corresponding categories.

38. A method as claimed in claim 34, further comprising the step of:

j) training a learning system to perform the classifying of said step (i) with a predetermined set of frequency and amplitude data derived from acoustic signals and corresponding categories.

39. A method as claimed in claim 34, wherein said steps (a) through (i) are repeatedly performed.

40. A method as claimed in claim 39, further comprising the steps of:

j) tracking the amount of actual damage incurred by the object;

k) determining whether the amount of actual damage has exceeded a predetermined amount of damage; and

l) generating an alarm, based on the determination of said step (k).

41. A method as claimed in claim 34, further comprising the steps of:

j) recording the category of actual damage classified in said step (i) in correspondence with at least one of the time data, amplitude data, and applied load data indicative of the amount and direction of the applied load.

42. A method as claimed in claim 34, further comprising the step of:

j) displaying the object and any actual damage thereof.

43. A method as claimed in claim 34, further comprising the step of:

j) amplifying the signal indicative of the acoustic signal sensed in said step (a), before performing said step (b).

44. A method as claimed in claim 34, further comprising the step of:

j) analog-to-digital converting the signal indicative of the acoustic signal sensed in said step (a), before performing said step (b).

45. A method as claimed in claim 34, further comprising the step of:

j) generating a signal for the acoustic signal induced in said step (e), said step (e) including the substep of transducing the signal generated in said step (j) to induce the acoustic signal in the object.

46. A method as claimed in claim 45, wherein the signal generated in said step (j) is in digital form, the method further comprising the step of:

k) digital-to-analog converting the signal generated in said step (j) into an analog signal before transducing the analog signal in the performance of said step (e).

47. A method as claimed in claim 46, further comprising the step of:

k) amplifying the signal generated in said step (j) before transducing the amplified signal in the performance of said step (e).

48. A method as claimed in claim 34, further comprising the step of:

j) amplifying the signal indicative of the acoustic signal sensed in said step (f), before the performance of said step (g).

49. A method as claimed in claim 34, further comprising the step of:

j) analog-to-digital converting the signal indicative of the acoustic signal sensed in said step (f), before performing said step (g).

50. An apparatus for detecting damage to an object under an applied load, the apparatus comprising:

acoustic inspection means operatively coupled to the object, the acoustic inspection means having passive and active modes of operation,

in the passive mode, the acoustic inspection means sensing at least one acoustic signal occurring in the object under the applied load to determine whether the object has incurred possible damage, the acoustic inspection means switching from passive mode to active mode if the object has incurred possible damage, and remaining in passive mode if the object has incurred no damage; and

in the active mode, the acoustic inspection means inducing at least one acoustic signal into the object to probe the possible damage in the object, the acoustic inspection means sensing the induced acoustic signal from at least a portion of the object having possible damage, and determining whether the possible damage is actual damage of the object, based on the sensed acoustic signal, the acoustic inspection means generating a signal indicative of the actual damage of the object, based on the sensed acoustic signal, and returning to the passive mode after generating the signal indicative of the actual damage of the object.

51. An apparatus as claimed in claim 50, wherein the acoustic inspection means extracts at least one feature from the acoustic signal sensed in the passive mode and classifies the acoustic signal as possible damage or no damage, based on the extracted feature.

52. An apparatus as claimed in claim 50, wherein the acoustic inspection means transduces the acoustic signal sensed in the passive mode into a signal indicative thereof, the acoustic inspection means sampling the signal indicative of the acoustic signal sensed in the passive mode to generate time and amplitude data, the acoustic inspection means converting the time and amplitude data into frequency and amplitude data from which the acoustic inspection means extracts the feature.

53. An apparatus as claimed in claim 50, wherein the acoustic inspection means extracts at least one feature from the acoustic signal sensed in the active mode and classifies the acoustic signal sensed in the active mode as actual damage or no damage, based on the extracted feature.

54. An apparatus as claimed in claim 50, wherein the acoustic inspection means transduces the acoustic signal sensed in the active mode into a signal indicative of the acoustic signal, the acoustic inspection means sampling the signal indicative of the acoustic signal sensed in the active mode to generate time and amplitude data, the acoustic inspection means converting the time and amplitude data into frequency and amplitude data from which the acoustic inspection means extracts the feature.

55. An apparatus as claimed in claim 50, wherein the acoustic inspection means switches from passive mode to active mode, and from active mode back to passive mode, to evaluate each acoustic signal sensed by the acoustic inspection means.

56. An apparatus as claimed in claim 50, wherein the acoustic inspection means tracks the amount of actual damage incurred by the object, compares the amount of actual damage to a predetermined damage amount, and generates an alarm if the actual damage amount exceeds the predetermined damage amount.

57. An apparatus as claimed in claim 50, wherein the acoustic inspection means generates a visual display to indicate the location and extent of object damage, based on the signal indicative of the actual damage.

58. An apparatus as claimed in claim 50, wherein the acoustic inspection means generates and transduces a signal to induce the acoustic signal in the object in the active mode.

59. An apparatus for detecting damage to an object under an applied load, the apparatus comprising:

at least one sensor mounted to sense an acoustic signal from the object, and generating a sensed signal, based on the sensed acoustic signal;

a mapping unit coupled to the sensor, for generating damage data indicative of possible damage in a passive mode of operation of the apparatus, based on the sensed signal, and for generating damage data indicative of actual damage incurred by the object in an active mode of operation of the apparatus, based on the sensed signal;

a controller coupled to the mapping unit, the controller restricting the apparatus to the passive mode if the damage data indicates that an object has incurred no damage a processor for switching the apparatus from the passive mode to the active mode to generate induced signal data to induce at least one acoustic signal in the object if the damage data indicates that the object has incurred possible damage, the processor determining whether the possible damage is actual damage to the object or no damage, based on the damage data corresponding to the induced acoustic signal, the processor switching an apparatus from the active mode to passive mode after making the determining of whether the possible damage is actual damage or no damage; and

a transducer coupled to the controller and mounted to the object, the transducer inducing the acoustic signal in the object based on the induced signal data.

60. An apparatus as claimed in claim 59, further comprising;

a data acquisition unit coupled to the sensor, the data acquisition unit sampling the sensed signal to generate time and amplitude data, and converting the time and amplitude data in time domain, to frequency and amplitude data in frequency domain, the data acquisition unit supplying the frequency and amplitude data to the mapping unit.

61. An apparatus as claimed in claim 60, wherein the data acquisition unit includes

an amplifier coupled to the sensor, for amplifying the sensed signal to generate an amplified signal;

an analog-to-digital converter coupled to the amplifier, for converting the amplified signal into a digital signal; and

a digital signal processor coupled to the analog-to-digital converter, the digital signal processor generating a sampling signal supplied to the analog-to-digital converter to cause the analog-to-digital converter to sample amplitude of the amplified signal at regular time increments, the digital signal processor receiving the amplitudes in association with respective time increments to generate time and amplitude data, the digital signal processor converting the time and amplitude data in time domain, to frequency and amplitude data in frequency domain, the digital signal processor supplying the frequency and amplitude data to the feature extraction unit.

62. An apparatus as claimed in claim 60 wherein the mapping unit includes

a feature extraction unit coupled to the data acquisition unit, the feature extraction unit extracting at least one feature from the frequency and amplitude data;

and

a damage classification unit coupled to receive the feature from the feature extraction unit, and generating the damage data to indicate whether the object has incurred possible damage in the apparatus'"'"' passive mode of operation, based on the extracted feature, and for generating the damage data to indicate whether the object has incurred actual damage in the apparatus active mode of operation, based on the extracted feature.

63. An apparatus as claimed in claim 59, further comprising:

a signal generation unit coupled between the processor and a transducer, the signal generation unit generating an induced signal, based on the induced signal data, that is transduced into the induced acoustic signal by the transducer.

64. An apparatus as claimed in claim 59, further comprising:

a memory coupled to the controller, for storing at least one count for the amount of actual damage incurred by the object.

65. An apparatus as claimed in claim 64 further comprising;

an alarm unit coupled to the controller, the controller comparing at least one count stored in the memory with a respective predetermined damage limit value, the controller generating an alarm signal supplied to the alarm unit if the count exceeds the predetermined damage limit value.

66. An apparatus as claimed in claim 59, wherein the object includes a composite material having matrix and fiber materials, and the damage data includes a plurality of categories including possible fiber damage, possible fiber-matrix interface damage, possible fiber-matrix debonding, possible matrix damage, and no damage in the passive mode, and the damage classification data includes a plurality of categories including actual fiber damage, actual fiber-matrix interface damage, actual fiber-matrix debonding, actual matrix damage, actual fiber damage, and no damage in the active mode.

67. An apparatus as claimed in claim 66, wherein the fiber damage class for the passive mode and the active mode is further divided into fiber layer damage for actual damage to respective fiber layers of the composite material.

68. An apparatus as claimed in claim 59, wherein the object includes a monolithic material, and the damage data includes a plurality of categories including possible surface cracking, possible necking, possible fatigue, possible cracking, and no damage in the passive mode, and the damage classification data includes a plurality of categories including actual surface cracking, actual necking, actual fatigue, actual cracking, and no damage in the active mode.

69. An apparatus as claimed in claim 60 has an acoustic inspection unit which includes:

a display unit coupled to the controller, the controller coupled to receive the time and amplitude data from a pulsating of sensors mounted at spaced positions to the object from the data acquisition unit, the controller generating a display signal supplied to the display unit to generate a visual display, based on the time and amplitude data from the plurality of sensors.

70. An apparatus as claimed in claim 59, wherein the controller includes a microprocessor.