Analysis of alignment data

• US 5,980,094 A
• Filed: 03/28/1997
• Issued: 11/09/1999
• Est. Priority Date: 03/28/1997
• Status: Expired due to Term

• Current Assignee: Computational Systems Incorporated (Emerson Electric Company)
• Original Assignee: CSI Technology Incorporated (Emerson Electric Company)
• Inventors: Nower, Daniel L.
• Primary Examiner(s): Trammell, James P.
• Assistant Examiner(s): DAM, TUAN QUANG
• Application Number:
  US08/827,576
• Time in Patent Office: 956 Days
• Field of Search: 364/528.14, 364/528.1, 364/474.34, 364/474.35-474.36, 702/56, 702/33, 33/412, 33/645, 33/661, 33/613, 33/655, 33/533, 33/286, 345/442, 356/153, 356/138, 356/399, 356/373, 356/400, 356/139.04, 356/141.2, 356/141.3
• US Class Current: 700/279
• CPC Class Codes: G01B 21/24 for testing alignment of axes

Abstract

Alignment data is analyzed to determine machine condition information that is not necessarily related to alignment. Alignment fixtures provide alignment data to an analyzer that finds a curve of a predetermined type, such as a sine curve, that best fits the data. A confidence factor is calculated indicating the probability that the alignment data corresponds to misalignment. The alignment data is further analyzed to determine a condition, other than misalignment, that contributed to the magnitudes of the alignment data. Preferably, the analyzer is used to check for machine conditions such as looseness in the machine, rubs, and spall defects in a bearing. Preferably, the condition analysis is based on both the confidence factors and curve parameters, such as phase, offset and amplitude of a sine curve. In the condition analysis mode, the analyzer provides both manual and automatic routine for removing selected segments of data to produce trial data sets that are analyzed to determine machine conditions.

Claims

1. An apparatus for determining the condition of a machine having a first shaft coupled to a second shaft, comprising:

alignment fixtures for producing alignment data composed of a plurality of data points having purportedly corresponding to the misalignment between the first and second shafts measured at a plurality of rotational positions relative to said shafts;

an alignment analyzer having an input for receiving the alignment data, said analyzer for finding a curve of a predetermined type that best fits the alignment data, and for analyzing the alignment data and the curve to produce a confidence factor indicating the probability that the alignment data actually corresponds to the misalignment of the first and second shafts; and

an output associated with said analyzer for outputting information corresponding to the alignment data including information based on said confidence factor.

2. The apparatus of claim 1 wherein said analyzer further comprises means for analyzing the alignment data and the curve for determining a condition other than misalignment of the first and second shafts that contributed to the magnitudes of the alignment data and for selectively producing a condition signal for being transmitted to the output.

3. The apparatus of claim 1 wherein said analyzer further comprises means for analyzing the alignment data and the curve for determining whether a bearing defect contributed to the magnitudes of the alignment data and for selectively producing a bearing defect signal for being transmitted to said output.

4. The apparatus of claim 1 wherein said analyzer further comprises means for analyzing the alignment data and the curve for determining whether a machine rub contributed to the magnitudes of the alignment data and for selectively producing a machine defect signal for being transmitted to the output.

5. The apparatus of claim 1 wherein said analyzer further comprises means for analyzing the number of data points in the alignment data, the rotational spacing between the data points, and the correlation between the alignment data and the curve, and for determining the confidence factor based on the analysis of the number, spacing and correlation of the data points.

6. The apparatus of claim 1 wherein said analyzer further comprises means for deleting selected ones of said data points from the alignment data to produce a trial data set of alignment data and for determining a confidence factor based in part on said trial data set of alignment data.

7. The apparatus of claim 1 wherein said analyzer further comprises means for deleting selected segments of data points from the alignment data to produce a trial data set of alignment data and for determining a confidence factor based in part on said trial data set of alignment data.

8. The apparatus of claim 1 further comprising:

an input user interface for producing input commands in response to a user;

said analyzer in response to the input commands being operable to remove selected data points from the alignment data to produce a trial data set of alignment data, said analyzer being further operable to produce a confidence factor based in part on said trial set of alignment data.

9. The apparatus of claim 1 wherein said analyzer further comprises:

means for producing a plurality of trial data sets, each trial data set different from the others, and each trial data set corresponding to said alignment data with a unique selected segment of data points removed from said alignment data;

means for finding a curve of a predetermined type for each trial data set that best fits the trial data set to thereby produce a plurality of best fit curves, one for each trial data set;

means for analyzing the plurality of trial data sets and their respective best fit curves for determining a plurality of trial data characteristics, one trial data characteristic for each trial data set; and

means for making a determination of a condition of the machine based on at least one of the trial data characteristics.

10. The apparatus of claim 9 wherein said means for analyzing further comprises means for determining a plurality of trial curve parameters, one set of curve parameters for each trial data set and wherein the determination of a condition of the machine is based in part on the trial curve parameters.

11. The apparatus of claim 10 wherein said trial curve parameters comprise phase, offset and amplitude of the trial best fit curves.

12. The apparatus of claim 9 wherein said means for analyzing further comprises means for determining a plurality of trial confidence factors, one trial confidence factor for each trial data set and wherein the determination of a condition of the machine is based in part on at least one of the trial confidence factors.

13. The apparatus of claim 12 wherein said means for analyzing further comprises means for determining a plurality of trial confidence factors and trial curve parameters, one trial confidence factor and one set of trial curve parameters for each trial data set and wherein the determination of a condition of the machine is based in part on at least one of the trial curve parameters and at least one of the trial confidence factors.

14. The apparatus of claim 12 wherein said means for making a determination further comprises means for determining whether all of the trial confidence factors meet predetermined criteria indicating that either the brackets or the machine is loose and for selectively producing a loose signal for being transmitted to the said output.

15. The apparatus of claim 12 wherein said means for determining further comprises means for identifying the largest of the trial confidence factors and for comparing the largest trial confidence factor to predetermined criteria indicating whether a rotationally isolated machine fault exists in the machine and for selectively producing a machine fault signal for being transmitted to said output.

16. The apparatus of claim 12 wherein said means for making a determination further comprises means for identifying the largest of the trial confidence factors and for comparing the largest trial confidence factor and the trial data set corresponding to the largest trial confidence factor to predetermined criteria indicating whether a rub condition exists in the machine and for selectively producing a rub signal for being transmitted to said output.

17. The apparatus of claim 9 wherein said means for making a determination further comprises means for identifying the largest of the trial confidence factors and for comparing the largest trial confidence factor and the trial data set corresponding to the largest trial confidence factor to predetermined criteria indicating the presence of a bearing spall in the machine and for selectively producing a spall signal for being transmitted to said output.

18. The apparatus of claim 1 further comprising:

said output including a visual display;

said analyzer for transmitting signals to said display and displaying a Cartesian coordinate graph showing the best fit curve and the data points;

input means for selecting a segment composed of a plurality of data points; and

said analyzer for responding to a selection of a segment to produce a trial data set corresponding to the alignment data with data removed based on the selected segment.

19. The apparatus of claim 1 further comprising:

means for determining the correlation between the curve and the alignment data and producing a temporary confidence factor; and

means for determining an adjustment factor that is inversely proportional to the number of data points in the alignment data and for reducing the magnitude of the temporary confidence factor by the magnitude of the adjustment factor to produce the confidence factor.

20. The apparatus of claim 1 wherein said analyzer further comprises:

means for determining the correlation between the curve and the alignment data and producing a temporary confidence factor; and

means for determining an adjustment factor that is inversely proportional to the rotational spacing of data points in the alignment data and for reducing the magnitude of the temporary confidence factor by the magnitude of the adjustment factor to produce the confidence factor.

21. The apparatus of claim 1 wherein said analyzer further comprises:

means for determining the correlation between the curve and the alignment data and producing a temporary confidence factor; and

means for determining an adjustment factor that is inversely proportional to the spacing and number of data points in the alignment data and for reducing the magnitude of the temporary confidence factor by the magnitude of the adjustment factor to produce the confidence factor.

22. The apparatus of claim 1 wherein said analyzer finds a sine curve that best fits the alignment data.

23. An apparatus for determining the condition of a machine having a first shaft coupled to a second shaft, comprising:

alignment fixtures for producing alignment data composed of a plurality of data points purportedly corresponding to the misalignment between the first and second shafts measured at a plurality of rotational positions relative to said shafts;

an alignment analyzer having an input for receiving the alignment data, said analyzer for finding a best fit curve of a predetermined type that best fits the alignment data and for analyzing the alignment data to determine whether a rotationally isolated machine fault exists and for selectively producing a machine fault signal; and

an output associated with said analyzer for outputting information corresponding to the alignment data including information based on said machine fault signal.

24. The apparatus of claim 23 wherein said analyzer means further comprises means for deleting selected ones of said data points from the alignment data to produce a trial set of alignment data and for determining whether a rotationally isolated machine fault exists based in part on said trial set of alignment data.

25. The apparatus of claim 23 wherein said analyzer further comprises:

means for deleting selected ones of said data points from the alignment data to produce a trial data set of alignment data;

means for finding a best fit curve of a predetermined type that best fits the trial data set;

means for analyzing the trial data set and the best fit curve for determining at least a correlation between the trial data set and the best fit curve; and

means for making a determination of whether a rotationally isolated machine fault exists based on the correlation.

26. The apparatus of claim 23 wherein said analyzer further comprises:

means for producing a plurality of trial data sets, each trial data set different from the others, and each trial data set corresponding to said alignment data with a unique selected segment of data points removed from said alignment data;

means for finding a curve of a predetermined type for each trial data set that best fits the trial data set to thereby produce a plurality of trial best fit curves, one for each trial data set;

means for analyzing the plurality of trial data sets and their respective trial best fit curves for determining a plurality of trial correlations, one trial correlation for each trial data set; and

means for making a determination of the existence of a rotationally isolated machine fault based in part on the trial correlations.

27. The apparatus of claim 23 further comprising:

said output including a visual display;

said analyzer for transmitting signals to said display and displaying a Cartesian coordinate graph showing the best fit curve and the data points;

input means for selecting a segment composed of a plurality of data points; and

said analyzer for responding to a selection of a segment to produce a trial data set corresponding to the alignment data with the selected segment removed, for finding a trial best fit curve of a predetermined type that best fits the trial data set, for determining a correlation between the trial best fit curve and the trial data set, and for comparing the correlation to predetermined criteria to determine whether a rotationally isolated machine fault exists.

28. The apparatus of claim 23 wherein said analyzer further comprises:

means for deleting a selected segment of said data points of a selected rotational size from the alignment data to produce a trial data set of alignment data;

means for finding a trial best fit curve of a predetermined type that best fits the trial data set;

means for analyzing the trial data set and the trial best fit curve for determining at least a correlation between the trial data set and the trial best fit curve; and

means for making a determination of whether a rotationally isolated machine fault exists based on the correlation and, if a rotationally isolated machine fault exists, determining the type of machine fault based on the rotational size of the segment and the amplitude of the best fit curve.

29. The apparatus of claim 23 wherein said means for analyzing further comprises means for determining a curve fit parameter for the best fit curve and wherein the determination of whether a rotationally isolated machine fault exists is based in part on at least one of the curve parameters.

30. The apparatus of claim 29 wherein said means for determining a trial fit parameter further comprises means for determining phase, offset and amplitude of the best fit curve.

31. The apparatus of claim 23 wherein said means for analyzing further comprises:

means for deleting a selected segment of said data points of a selected rotational size from the alignment data to produce a trial data set of alignment data;

means for finding a trial best fit curve of a predetermined type that best fits the trial data set;

means for analyzing the trial data set and the trial best fit curve for determining at least a confidence factor based on the trial data set and the trial best fit curve; and

means for making a determination of whether a rotationally isolated machine fault exists based on the confidence factor.

32. The apparatus of claim 23 wherein said means for analyzing further comprises:

means for deleting a selected segment of said data points of a selected rotational size from the alignment data to produce a trial data set of alignment data;

means for finding a trial best fit curve of a predetermined type that best fits the trial data set;

means for analyzing the trial data set and the trial best fit curve for determining at least a curve fit parameter; and

means for making a determination of whether a rotationally isolated machine fault exists based on the curve fit parameter.

33. The apparatus of claim 32 wherein said means for analyzing further comprises means for determining phase, offset and amplitude of said trial best fit curve.

34. The apparatus of claim 23 wherein said means for analyzing further comprises:

means for deleting a selected segment of said data points of a selected rotational size from the alignment data to produce a trial data set of alignment data;

means for finding a trial best fit curve of a predetermined type that best fits the trial data set;

means for analyzing the trial data set and the trial best fit curve for determining at least a confidence factor and a curve fit parameter; and

means for making a determination of whether a rotationally isolated machine fault exists based on the confidence factor and the curve fit parameter.

35. The apparatus of claim 23 wherein said means for analyzing further comprises:

means for producing a plurality of trial data sets, each trial data set different from the others, and each trial data set corresponding to said alignment data with a unique selected segment of data points removed from said alignment data;

means for finding a curve of a predetermined type for each trial data set that best fits the trial data set to thereby produce a plurality of trial best fit curves, one for each trial data set;

means for analyzing the plurality of trial data sets and their respective trial best fit curves for determining a plurality of trial correlations and a plurality of trial curve parameters, one trial correlation and one set of curve parameters for each trial data set; and

means for analyzing the correlations and the curve parameters to identify a segment of data points caused in part by a rotationally isolated machine fault.

36. A method for determining the condition of a machine having a first shaft coupled to a second shaft, comprising:

producing alignment data composed of a plurality of data points purportedly corresponding to the misalignment between the first and second shafts measured at a plurality of rotational positions relative to said shafts;

finding a curve of a predetermined type that best fits the alignment data; and

analyzing the alignment data and the curve and producing a confidence factor indicating the probability that the alignment data actually corresponds to the misalignment of the first and second shafts.

37. The method of claim 36 further comprising analyzing the alignment data and the curve for determining whether a bearing defect contributed to the magnitudes of the alignment data.

38. The method of claim 36 further comprising analyzing the alignment data and the curve for determining whether a machine rub contributed to the magnitudes of the alignment data and for selectively producing a machine defect signal for being transmitted to the output.

39. The method of claim 36 wherein said analyzing further comprises analyzing the number of data points in the alignment data, the rotational spacing between the data points, and the correlation between the alignment data and the curve, and determining the confidence factor based on the analysis of the number, spacing and correlation of the data points.

40. A method for determining the condition of a machine having a first shaft coupled to a second shaft, comprising:

producing alignment data composed of a plurality of data points purportedly corresponding to the misalignment between the first and second shafts measured at a plurality of rotational positions relative to said shafts;

finding a best fit curve of a predetermined type that best fits the alignment data, and

analyzing the alignment data to determine whether a rotationally isolated machine fault exists; and

selectively producing a machine fault signal based on the analyzing step.

41. The method of claim 40 wherein said analyzing step further comprises deleting selected ones of said data points from the alignment data to produce a trial set of alignment data and determining whether a rotationally isolated machine fault exists based in part on said trial set of alignment data.

42. The method of claim 40 wherein said analyzing step further comprises:

deleting selected ones of said data points from the alignment data to produce a trial data set of alignment data;

finding a best fit curve of a predetermined type that best fits the trial data set;

analyzing the trial data set and the best fit curve for determining at least a correlation between the trial data set and the best fit curve; and

making a determination of whether a rotationally isolated machine fault exists based on the correlation.

43. The method of claim 40 wherein said analyzing further comprises:

deleting a selected segment of said data points of a selected rotational size from the alignment data to produce a trial data set of alignment data;

finding a trial best fit curve of a predetermined type that best fits the trial data set;

analyzing the trial data set and the trial best fit curve for determining at least a correlation between the trial data set and the trial best fit curve; and

making a determination of whether a rotationally isolated machine fault exists based on the correlation and, if a rotationally isolated machine fault exists, determining the type of machine fault based on the rotational size of the segment and the amplitude of the best fit curve.