Whirling of Shaft

NAVAL POSTGRADUATE SCHOOL MONTEREY, CALIFORNIA
Sj a
If^
&m iLi JU W &
% APR 2 7 1995
THESIS
ANALYTIC AND EXPERIMENTAL INVESTIGATION OF ROTORDYNAMIC RESPONSE AND BACKWARD WfflRL INDUCED BY SPLIT RESONANCE by Curtis E. Vejvoda December, 1994 Thesis Advisor: Knox T. Millsaps Jr.
Approved for public release; distribution is unlimited.
19950426 117
REPORT DOCUMENTATION PAGE
Form Approved OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instruction, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188) Washington DC 20503.
1.
AGENCY USE ONLY (Leave blank) 2. REPORT DATE December 1994 TITLE AND SUBTITLE ANALYTIC AND EXPERIMENTAL INVESTIGATION OF ROTORDYNAMIC RESPONSE AND BACKWARD WHIRL INDUCED BY SPLIT RESONANCE
REPORT TYPE AND DATES COVERED Master's Thesis FUNDING NUMBERS
6. 7.
9.
AUTHOR(S) Curtis E. Vejvoda PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Naval Postgraduate School Monterey CA 93943-5000 SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)
8.
PERFORMING ORGANIZATION REPORT NUMBER
10. SPONSORING/MONITORING AGENCY REPORT NUMBER
11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government. 12b. DISTRIBUTION CODE 12a. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution is unlimited. 13. ABSTRACT (maximum 200 words) Rotordynamic response phenomena, including backward whirl, were investigated both analytically and experimentally. A two degree-of-freedom rotor model was developed to simulate the steady state, lateral vibration characteristics of a simply supported, single disk rotor. This model includes the effects of direct and cross-coupled, linear damping and stiffness coefficients. The computer model was used to quantify the influence of bearing characteristics on rotordynamic response. The presence of split resonance, which appears to be due to separate and distinct natural frequencies in the two orthogonal lateral directions and the occurence of backward whirl between these two frequencies was studied. The effects of geometric imperfections in the bearing sleeve, gravitational forces and bearing support stiffnesses were isolated using the experimental apparatus. It was determined that the split resonance induced backward whirl and the different natural frequencies were caused by asymmetric stiffness of the bearing support structure as well as gravitational forces. Bearing imperfections did not create the backward whirl phenomena. The bearing support characteristics necessary to create the observed rotor orbits were determined. 14. SUBJECT TERMS 15. NUMBER OF PAGES 135 18. SECURITY CLASSIFICATION OF THIS PAGE Unclassified 16. PRICE CODE 20. UMTTATIONOF 19. SECURITY CLASSIFIABSTRACT CATION OF ABSTRACT UL Unclassified Standard Form 298 (Rev. 2-89)
Prescribedby ANSI Std. 239-18298-102
17. SECURITY CLASSIFICATION OF REPORT Unclassified NSN 7540-01-280-5500
11
Approved for public release; distribution is unlimited
ANALYTIC AND EXPERIMENTAL INVESTIGATION OF ROTORDYNAMIC RESPONSE AND BACKWARD WHIRL INDUCED BY SPLIT RESONANCE
Curtis E. Vejvoda Lieutenant, United States Navy B.S., California Maritime Academy, 1986
Submitted in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE IN MECHANICAL ENGINEERING Accesion For AND MECHANICAL ENGINEER NTiS CRA&I from the NAVAL POSTGRADUATE SCHOOL December 1994
DT!C TAB Unannounced Justificst ion By Distribution /
13 D
Author:
Approved by:
Matthew D Kelleher, Chairman, Department of Mechanical Engineering
ill
IV
ABSTRACT
Rotordynamic response phenomena, including backward whirl, were investigated both analytically and experimentally. A two degree-of-freedom rotor model was
developed to simulate the steady state, lateral vibration characteristics of a simply supported, single disk rotor. This model includes the effects of direct and cross-coupled, linear damping and stiffness coefficients. The computer model was used to quantify the influence of bearing characteristics on rotordynamic response. The presence of split
resonance, which appears to be due to separate and distinct natural frequencies in the two orthogonal lateral directions and the occurence of backward whirl between these two frequencies was studied. The effects of geometric imperfections in the bearing sleeve, gravitational forces and bearing support stiffnesses were isolated using the experimental apparatus. It was determined that the split resonance induced backward whirl and the different natural frequencies were caused by asymmetric stiffness of the bearing support structure as well as gravitational forces. Bearing imperfections did not create the
backward whirl phenomena. The bearing support characteristics necessary to create the observed rotor orbits were determined.
VI
TABLE OF CONTENTS
I.
INTRODUCTION
1 7 9 14 15 15 ^ 16
II. BACKGROUND III. ANALYTIC MODEL A. KXX = 495, KYY = 495, CxX = -025, CYY = -025 B. KXX = 535, KYY = 555, CXX =-025, CYY =-025 C. Kxx = 535, KYY = 535, CXX =-035, CYY =-02 D. Kxx = 535, KYY = 555, CxX =-04, CYY =-02 E. KXX = 535, KYY = 575, CxX = -04, CyY = -025
F. KXX=535, KYY=535, KXY=40, KYX=I0, CXX=-025, CYY=-025... 16 G. Kxx=535, KYY=535, KXY=10, KYX=40, CXX=-025, CYY=-025... 17 H. Kxx=535, KYY=535, CXX=-025, CYY=-025, CXY=--02, CYX=--01 17 I. Kxx=535, KYY=535, CXX=-025, CYY=-025, CXY=-02, CYX=-01 . 18 18 19 22 22 22 23 23 24 24 25
J. KxX=535, KYY=555, KXY=-5, Kyx=3, Cxx=-025, CYY=-025 IV. EXPERIMENTAL INVESTIGATION. A. GOOD BEARING TABLE MOUNT 1. Good Bearing No Imbalance 2. Good Bearing 0.5 Gram Imbalance. 3. GoodBearing 1.0 Gram Imbalance. 4. GoodBearing 1.6 Gram Imbalance B. BAD BEARING TABLE MOUNT 1. Vertical Bad Bearing 1.0 Gram Imbalance 2. Horizontal Bad Bearing 1.0 Gram Imbalance
vil
3. 45 Degrees Left Bad Bearing 1.0 Gram Imbalance 4. 45 Degrees Right Bad Bearing 1.0 Gram Imbalance C. HORIZONTAL SHAFT MOUNT 1.0 GRAM 1MB ALANCE D. VERTICAL SHAFT MOUNT 1.0 GRAM 1MB ALANCE V. DISCUSSION OF RESULTS VI. CONCLUSIONS AND RECOMMENDATIONS A. CONCLUSIONS B. RECOMMENDATIONS APPENDKA: LIST OF DIMENSIONS AND PROPERTIES APPENDKB: ANALYTIC ROTOR DATA APPENDIX C: EXPERIMENTAL ROTOR DATA LIST OF REFERENCES INITIAL DISTRIBUTION LIST
25 25 26 26 29 33 33 34 37 39 59
115
117
Vlll
LIST OF FIGURES
Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6.
Rotor and Disk Assembly in Forward Whirl Rotor and Disk Assembly in Backward Whirl Analytic Lumped Mass and Stiffness Model of Two Degree-ofFreedom Rotating Shaft and Disk Assembly Analytic Pinned-Pinned Model of Shaft and Disk Assembly Configuration of Experimental Rotor System Rotor Orbits and Displacements From Analytic Model For Kxx = 495 lbf/in, Kyy = 495 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 495 lbf/in, Kyy = 495 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .035, Cyy = .02, Cxy = 0, Cyx = 0 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .035, Cyy = .02, Cxy = 0, Cyx = 0 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .04, Cyy = .02, Cxy = 0, Cyx = 0 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .04, Cyy = .02, Cxy = 0, Cyx = 0 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 575 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .04, Cyy = .025, Cxy = 0, Cyx = 0 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 575 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .04, Cyy = .025, Cxy = 0, Cyx = 0 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 40 lbf/in, Kyx = 10 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
3 4 9 13 21
39
Figure 7.
40
Figure 8.
41
Figure 9.
42
Figure 10.
43
Figure 11.
44
Figure 12.
45
Figure 13.
46
Figure 14.
47
Figure 15.
48
Figure 16.
49
IX
Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. Figure 27. Figure 28. Figure 29. Figure 30. Figure 31. Figure 32. Figure 33
Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 40 lbf/in, Kyx = 10 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0,Cyx = 0 50 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 10 lbf/in, Kyx = 40 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 51 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 10 lbf/in, Kyx = 40 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 52 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = -.02, Cyx = -.01 53 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = -.02, Cyx = -.01 54 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = .02, Cyx = .01 55 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = .02, Cyx = .01 56 Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = -5 lbf/in, Kyx = 3 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 57 Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = -5 lbf/in, Kyx = 3 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 58 Rotor Orbits No Imbalance Good Bearing. Speeds from 2760 to 2900 rpm 59 Rotor Orbits No Imbalance Good Bearing. Speeds from 2920 to 3200 rpm "0 Rotor Orbits No Imbalance Good Bearing. Speeds from 5600 to 6200 rpm "I Maximum Displacements and Bode Plot No Imbalance Good Bearing.... 62 Cascade Plot No Imbalance Good Bearing 63 Rotor Orbits 0.5 Gram Imbalance Good Bearing. Speeds from 2760 64 to 2900 rpm Rotor Orbits 0.5 Gram Imbalance Good Bearing. Speeds from 2920 65 to 3200 rpm Rotor Orbits 0.5 Gram Imbalance Good Bearing. Speeds from 5600 66 to 6200 rpm
Figure 34. Figure 35. Figure 36. Figure 37. Figure 38. Figure 39. Figure 40. Figure 41. Figure 42. Figure 43. Figure 44. Figure 45. Figure 46. Figure 47. Figure48. Figure 49. Figure 50. Figure 51. Figure 52. Figure 53. Figure 54.
Maximum Displacements and Bode Plot 0.5 Gram Imbalance Good Bearing Cascade Plot 0.5 Gram Imbalance Good Bearing Rotor Orbits 1.0 Gram Imbalance Good Bearing. Speeds from 2760 to2860rpm Rotor Orbits 1.0 Gram Imbalance Good Bearing. Speeds from 2880 to3200rpm Rotor Orbits 1.0 Gram Imbalance Good Bearing. Speeds from 5600 to5900rpm Rotor Orbits 1.0 Gram Imbalance Good Bearing. Speeds from 6000 to6200rpm Maximum Displacements and Bode Plot 1.0 Gram Imbalance Good Bearing Cascade Plot 1.0 Gram Imbalance Good Bearing Rotor Orbits 1.6 Gram Imbalance Good Bearing. Speeds from 2600 to2830rpm Rotor Orbits 1.6 Gram Imbalance Good Bearing. Speeds from 2860 to2940rpm Rotor Orbits 1.6 Gram Imbalance Good Bearing. Speeds from 3000 to3200rpm Rotor Orbits 1.6 Gram Imbalance Good Bearing. Speeds from 5600 to5900rpm Rotor Orbits 1.6 Gram Imbalance Good Bearing. Speeds from 6000 to6200rpm Maximum Displacements and Bode Plot 1.6 Gram Imbalance Good Bearing Cascade Plot 1.6 Gram Imbalance Good Bearing Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction. Speeds from 2780 to 2860 rpm Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction. Speeds from 2900 to 3200 rpm Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction. Speeds from 5000 to 5800 rpm Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction. Speeds from 5900 to 6000 rpm Maximum Displacements and Bode Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87
XI
Figure 55 Figure 56. Figure 57. Figure 58. Figure 59. Figure 60
Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction. Speeds from 2780 to 2860 rpm 88 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction. Speeds from 2900 to 3200 rpm 89 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction. Speeds from 5000 to 5800 rpm 90 Maximum Displacements and Bode Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction 91 Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation 92 In the Horizontal Direction Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical). Speeds from 2780 to 2860 rpm 93 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical). Speeds from 2900 to 3200 rpm 94 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical). Speeds from 5600 to 5800 rpm 95 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical). Speeds from 5900 to 6200 rpm 96 Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical)
97
Figure 61
Figure 62
Figure 63
Figure 64
Figure 65
Figure 66
Figure 67.
Figure 68.
Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off 9 vertical) Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees off Vertical). Speeds from 2780 to 2860 rpm 99 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees ott Vertical). Speeds from 2900 to 3200 rpm 100 Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees oft 101 Vertical). Speeds from 5600 to 6000 rpm Maximum Displacements and Bode Plot for 1.0 Grain Imbalance Elongated Bearing with Elongation Perpendicular to Direction ot Resonant Displacement (45 Degrees off Vertical) 102 Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees oft Vertical)
Figure 69.
Figure 70
10i
Xll
Figure 71. Figure 72. Figure 73. Figure 74. Figure 75. Figure 76. Figure 77. Figure 78. Figure 79. Figure 80.
Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Horizontal. Speeds from 2760 to 2880 rpm 104 Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Horizontal. Speeds from 2900 to 3200 rpm 105 Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Horizontal. Speeds from 5600 to 6000 rpm 106 Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Shaft Mounted Horizontal 107 Cascade Plot 1.0 Gram Imbalance Shaft Mounted Horizontal 108 Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Vertical. from 2400 to 2800 rpm Speeds 109 110 Ill 112 113
Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Vertical. Speeds from 2900 to 3100 rpm Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Vertical. Speeds from 5400 to 6000 rpm Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Shaft Mounted Vertical Cascade Plot 1.0 Gram Imbalance Shaft Mounted Vertical
Xlll
XIV
LIST OF SYMBOLS
SYMBOL
DEFINITION (units. U.S. Customary)
E I L M Mdisk Mshaft PSD {q} {q} {q} e
Modulus of elasticity of the rotor shaft, (psi) First moment of inertia of the rotor shaft, (in^) Length of the rotor shaft, (in) Mass of the imbalance, (lbm) Mass of the rotor disk, (lbm) Mass of the rotor shaft, (lbm) Power spectral density. (mils/HZ) Linear displacement vector, (in) Linear velocity vector, (in/sec) Linear Acceleration vector, (in/sec^) Distance from shaft centerline to center of mass imbalance (in)
KXx
direction stiffness of bearing plus shaft caused by x
direction displacement, (lbf/in) Kyy y direction stiffness of bearing plus shaft caused by y direction displacement, (lbf/in) KXy x direction stiffness of bearing plus shaft caused by y direction displacement, (lbf/in) KyX y direction stiffness of bearing plus shaft caused by x direction displacement, (lbf/in) Cxx
x
direction damping of bearing plus shaft caused by x
direction velocity, (dimensionless)
xv
Cyy
y direction damping of bearing plus shaft caused by y direction velocity, (dimensionless)
Cxy
x direction damping of bearing plus shaft caused by y direction velocity, (dimensionless)
Cyx
y direction damping of bearing plus shaft caused by x direction velocity, (dimensionless)
F Fx Fy Q, t 0(z) x,y,z
Force caused by mass imbalance, (lbf) Component of imbalance force in x direction, (lbf) Component of imbalance force in y direction, (lbf) Rotational speed of shaft and disk assembly, (rad/sec) Time (sees) Axial displacement shape function (dimensionless) Cartesian coordinates.
XVI
I.
INTRODUCTION
Rotating machinery is widely used throughout the world in various applications. Such devices include gas turbines, steam turbines, pumps, compressors and fans. Rotating
devices are subjected to numerous external and internal forces. These forces can be categorized as either steady or time varying. The time varying, or vibratory motions create alternating stresses, which tend to decrease the operational life of the equipment, due to wear and high cycle fatigue. Proper prediction of the machine dynamics, during the design phase, along with an appropriate condition monitoring program, during the life cycle, is imperative if sufficient component life is to be ensured. Smooth operation of rotating
machines is critical to all users of these machines including the U.S. Navy. Rotating machines can undergo many types of, vibrations which are induced from a variety of sources. Axial compressor and turbine airfoils, excited primarily by upstream wakes, may undergo bending and torsional vibration loads. The shaft and disks can
vibrate in axial, torsional and lateral modes. All of these vibrations may occur due to either external forcing or various self-excitation mechanisms. Lateral vibrations, due primarily to residual imbalance in rotating members, is the most common type of vibration encountered in practice. Rotors tend to be designed to operate at speeds in excess of their first lateral critical speed. This is termed supercritical operation. This type of operation provides low vibration levels at the operating speed and light weight devices leading to less expensive machinery. One major problem with operating a rotor supercritically is that the rotor must proceed through the resonance speed, where high vibratory stresses occur, prior to reaching its operating speed. Transition through the
critical speed should be done as quickly as possible in order to prevent such resonant conditions or to reduce the time at which the equipment experiences the high vibration
levels. To reduce the lateral bending forced response, which is due to residual imbalance, many balancing techniques are employed. However, even when the device is in balance to a specific grade the radial response near the resonant speed can be quite high. The
vibration problems encountered in rotating machinery can lead to significant equipment damage due to the close tolerances encountered in the rotating device. The alternating
stresses generated by these vibrations can also create bearing distress, which can lead to premature failure. The major design considerations relating to lateral shaft vibrations are to determine the critical frequencies and the radial response magnitudes as a function of rotor speed (and possibly rotor angular acceleration) as well as any potential regimes of rotordynamic instability. Rotating equipment design can be improved by having a better understanding of the various rotordynamic phenomena which relate to machinery vibration problems. Understanding the dynamics of the various components in a rotating machine can lead to much improved designs and hence longer operating lives and lower life-cycle costs. An ideal, symmetric rotor with symmetric bearings and supports will typically have circular orbits of the shaft center which get progressively larger until the critical speed is reached After the rotor transitions through the critical speed the orbits remain circular and decrease in size. The orbits remain in the same direction as the rotation of the shaft (forward whirl), as shown in Figure 1. Maximum displacements from the shaft centerline are observed at or near the critical speed (resonance). However, in a recent experimental investigation by Simei [Ref. 1], a much more complex lateral response was observed from a simply supported single disk rotor. The orbits started in a small amplitude, circular forward procession at low speeds. As the rotor speed was increased the orbits became elliptical. At an initial critical speed, the orbit
"collapsed" into a line. At a slightly higher speed the rotor executed a whirl orbit in the
reverse direction of shaft spin. This phenomena is termed "backward whirl" and is shown in Figure 2. There was a small range of rotor speeds where the backward whirl
phenomena occurred. At a slightly higher speed a second resonant peak occurred along with another collapsed orbit. As the speed was further increased the orbits again proceeded in the forward direction and became more circular (less elliptical).
A*
Figure 1. Rotor and Disk Assembly in Forward Whirl
The backward whirl phenomena is believed to be caused by distinctly different natural frequencies of the rotor system in the two orthogonal lateral directions. [Ref. 2] The two
(split resonant) frequencies cause x and y displacement curves to have peak values at separate rotational speeds.
Figure 2. Rotor and Disk Assembly in Backward Whirl
The objectives of this research were to identify the cause of the split resonance and backward whirl phenomena. To determine if, in fact, these two phenomena are caused by different natural frequencies in the two directions created by non-symmetric bearing
rotordynamic coefficients and, if this was determined to be the cause of backward whirl, to identify the specific cause of the asymmetry. In Chapter II a review of the pertinent literature is given to provide sufficient background information concerning this research. In an attempt to identify the causes for the complex behavior noticed in the simple rotor configuration an analytic model is developed in Chapter III and the major parameters varied to influence rotor responses. The analytic solutions are compared to experimental results obtained in a closely controlled environment. In Chapter IV experiments, which were conducted, are described and show the effects of gravity, bearing eccentricity and imbalance. In generating experimental data a variety of rotor bearing configurations was employed, however the shaft remained in the simply supported configuration To determine the effect of imbalance, the rotor was purposely placed out of balance using known weights. The amount of imbalance was gradually increased. To determine the effect of bearings on rotor response, elongated bearings (eccentric) were used and oriented in various relative positions on the bearing mounts. The shaft was then positioned in orthogonal directions with respect to the gravitational field to determine the effects of gravity and bearing support structure on rotor response. A modal test to determine static natural frequencies of the experimental apparatus in the two lateral directions was conducted by installing accelerometers on the rotor assembly and applying a known impulsive force with the use of a precision hammer. The vertical natural frequency was found to be slightly higher than the horizontal natural frequency implying asymmetric stiffness properties in these two directions. A discussion of results from both the experimental and analytic models is contained in Chapter V and recommendations and conclusions are provided in Chapter VI.
II. BACKGROUND
The first documented rotordynamic analysis was submitted by Rankine [Ref. 3] in 1869. He considered a simply supported rotating shaft and proposed that the deflection of the shaft is caused by the centrifugal force due to some nominal shaft deflection. Opposing this force is the shaft bending rigidity (stiffness). In the analysis he came to the erroneous conclusion that supercritical operation was statically unstable in the rotating reference frame and hence not possible for rotating machines. In Rankine's analysis, the Coriolis forces in the rotating reference frame were not considered. If purely radial motion of the shaft is considered this would be the case, however, the orbiting motion creates additional forces which stabilize the shaft. In 1919 Jeffcott [Ref. 4] published an article analyzing the effects of imbalance on rotating machines. This analysis, which is considered the classical reference on the
subject, provides the simplest and most accurate model of the dynamics of a simply supported rotating shaft. Jeffcott corrected the basic mistake made by Rankine. The
Jeffcott rotor is comprised of a long shaft with a concentrated mass disk. This is a two degree-of-freedom model for transverse motion. Jeffcott described the phenomena of
"whirling" and was the first to notice, and correctly assess, the effect of natural frequencies on rotordynamic response. Damping was neglected in his analysis and a pronounced change in phase was observed as the rotor transitions from subcritical speeds through resonance to supercritical speeds. In addition, Jeffcott showed the advantages of
supercritical operation of rotating machinery. Subsequent to Jeffcott's work, many authors have developed analytical models for rotors and compared the predicted results to those obtained experimentally. The influence of bearing dynamics, gyroscopic effects (from the disk), support flexibility and many other phenomena have been investigated In 1986 Muszynska [Ref. 5] published an article
detailing the fundamental response of rotating machines. Muszynska expressed an analytic model of rotor imbalance which included damping forces to predict rotor dynamic response for a wide range of unbalanced rotors. In 1988 Vance [Ref. 2] proposed that the backward whirl phenomena is caused by the rotor imbalance when the rotational speed is between to natural frequency splits caused by bearing stiffness asymmetry. He also surmised that increased damping would cause the backward whirl phenomena to "disappear." He had, in fact, observed the backward whirl phenomena in his experimental results. Simei [Ref. 1], in a Masters Thesis at the Naval Postgraduate School, obtained a rotor analysis kit from the Bentley Nevada Corporation and installed software for data acquisition and signal processing. Under certain shaft loading conditions backward whirl was observed and documented, however an explanation as to the cause of this phenomena was not provided Simei used brass oil impregnated (oilite) bearings for all of his
experiments and configured the rotor assembly as simply supported with a single disk arrangement. The backward whirl phenomena was observed to occur between to distinct maximum displacement peaks in the two orthogonal lateral directions.
III. ANALYTIC MODEL
In order to determine the cause of the split resonance and backward whirl and to predict rotordynamic response, a simple analytic model was developed. The model had to be capable of incorporating non-symmetric bearing characteristics, both stiffness and damping, which was hypothesized to be the cause of the observed split resonance and backward whirl. In this model the rotor shaft was modeled as a two degree-of-freedom (x and y) lumped mass and stiffness shaft and disk assembly as shown in Figure 3.
MASS IMBALANCE (M)
SHAFT
-*
Kyy, Kyx
Figure 3. Analytic Lumped Mass and Stiffness Model of Two Degreeof-Freedom Rotating Shaft and Disk Assembly.
The model includes bearing and shaft forces which are linearly related to the displacements and velocities of the rotor only. The variations in the axial direction have been ignored. All of the properties have effectively been concentrated into the single axial plane. The mass is the algebraic sum of the disk and the shaft masses. The direct stiffnesses, Kxx and Kyy, are a combination of the rotor shaft and the bearings stiffness. The cross stiffness terms, Kxy and Kyx, are due to any self excitation interfacial forces provided by the bearings. These are non-conservative forces. The direct damping terms, Cxx and Cyy, are from both material damping of the shaft as well as bearing and support proportional damping. This model treats those cases when the two bearings are identical, but not necessarily symmetric. This means that the x and y displacements along the shaft follow the functional form of: x(t,z) = x(t:z = zo)0(z) and y(t,z) = y(t: z = z0)tf>(z) (2) (1)
Where, x(t,z) and y(t,z) are the x and y displacements along the shaft, z is the axial direction, Zo is a particular location along the length of the shaft and tp(z) is the whirling mode shape. The equation of motion for this model system, in terms of q, which represents the radial displacement of the shaft centerline, can be written in matrix form as:
[M]|q} + [C]{q} + [KKq} = {F(t)} with,
(3)
10
q.a dt
4
(4)
q = ^F
n
d2q
and the mass, damping and stiffness matrices respectively, are given by:
"M 0 [M] = . 0 Mj [C] = c [K] =

where, Mtot = Mshaft Kxx = Kxx(shaft)
+ +
XX
Cxyl
yy-
(5)
xx yx
Kvxy
yy
Mdisk;
+
KXx(bearing); Kyy = Kyy(shaft)
Kyy(bearing);
KXy and KyX = coupling or cross stiffness ; Cxx = Cxx(shaft)

+
CXx(bearing); Cyy = Cyy(shaft)
Cyy(bearing);
Cxy and Cyx = coupling or CTOSS damping
and, since the force on the shaft, F(t), is created by a mass imbalance and is periodic in time it can be expressed in exponential notation as:
-a*** {F(t)}=Fe
(6)
where j =V^T, Q is the rotational speed of the shaft in rad/sec and t is time. From Figure 3 the force magnitude is:
11
F = F = MeQ2
(7)
where, M is the mass of the imbalance, e is the radial distance from the shaft centerline to the location of the mass imbalance and Q is the rotational speed of the shaft. x and y displacements are known to be sinusoidal and the absolute displacements in the x and y directions are orthogonal, therefore, assume q is of the form: {q} = qeja
(8)
where q is a radial displacement: V(x2+y2) Taking derivatives with respect to time:
q^=qQ)ej2t (9)
qf=q(-Q2)eja
Substituting Equations (7) (8) and (9) into Equation (3) gives [M]q(-fl2)ep + [C]q(jfl)eifll + [K]qept = Fejnt
(10)
but ejnt * o for finite time and solving Equation (10) for q leads to, q = [[K]-Q2[M] + jfl[C]rF dD
Now the displacements x and y can be determined by selectively substituting Equations (4) and (5) into Equation (11) which provides for a direct solution for the displacements:
12
IM
lyl
-Q2
0 + j M,, c
c, c
{FejQt}
(12)
yy
Equation (12) is limited to steady state (constant speed) response, symmetric shaft configurations, and linear stiffness and damping. This equation does not apply to cases where the rotor is accelerating or configured with two different types of bearings. To obtain an approximation for the stiffness terms, the shaft and disk assembly can be modeled as a pinned-pinned beam as shown in Figure 4. Where the force P is the weight of the disk.
14 in P 7 in 7 in
DISPLACEMENT CURVE
Figure 4.
Analytic Pinned-Pinned Model of Shaft and Disk Assembly.
From [Ref. 6] the stiffness of the shaft is: 48EI

T
K. h.ft 'shift
(13)
13
Using the shaft data in Appendix A, the shaft and disk mass can be determined and the shaft stiffness is found to be 495 lbf/in. The bearing stiffness and all of the damping terms must be selectively chosen since no hard data is available and modeling it is non-trivial. The displacements obtained using Equation (12) are for the disk position and a straight line extrapolation can be performed to obtain the displacements anywhere along the length of the shaft. To quantify the effects of the various parameters in Equation (12) values were chosen for stiffness and damping with a 1.0 gram mass imbalance imposed and the equation was solved The model was used to see the effects on rotor response caused by separately imposed analytic conditions. The first solution only considers the stiffness of the shaft and uses symmetric damping and stiffness. A comparison was made to the results obtained in [Ref. 1]. The procedure was repeated many times with varying coefficients in an attempt to isolate the effects of asymmetric stiffness, asymmetric damping, cross stiffness and cross damping. The results of some of these analyses are contained as plots in Appendix B and an explanation of each solution is provided in the following subsections (stiffnesses are in lbf/in and damping coefficients are dimensionless). It may be possible to optimize Equation (12) to match observed data, however, this was not done. The sequential progression through each of the imposed analytic responses lead to an increased understanding of rotordynamic response phenomena. The rotational speed of the shaft is termed (lx), twice rotational speed is (2x) and one half rotational speed is (l/2x). This terminology follows that contained in the reference material discussed in Chapter n.
A. Kxx = 495, KYY = 495, Cxx = -025, CyY = -025

The rotor orbits and displacements are shown in Figure 6. The orbits are circular, as expected, with a resonant speed of approximately 2720 rpm. The displacements in the x
14
and y directions are symmetric with a maximum value of 32 mils. The phase diagram and cascade plot are shown in Figure 7. A smooth transition from subcritical to supercritical speed is observed on the phase diagram at the resonant speed. The cascade plot indicates the energy content of the system is contained at the operating speed of the shaft with no indication of (l/2x) or (2x) vibrations. This is to be expected since sub-harmonic and super-harmonic responses are due to nonlinearities which this model does not contain. The observed resonant speed is lower than that obtained in [Ref. 1] which leads to the conclusion that a bearing stiffness term exists. B. KxX = 535, KYY = 555, CXX = 025, CYY = .025 The bearing stiffness is assumed to be 40 lbf/in where the disk is located along the shaft and 20 lbf/in asymmetric stiffness is imposed The rotor orbits and displacements are shown in Figure 8. The orbits begin in a circular pattern at slow speeds and proceed into elliptical patterns and exhibit the backward whirl condition at 2840 rpm. This closely
matches the resonant speed found in [Ref. 1]. Two distinct resonant speeds are evidenced from the displacement plots which is indicative of the asymmetric stiffness in the x and y directions, however the magnitudes of the displacements are similar. The phase diagram and cascade plot are shown in Figure 9. The phase shift for the x direction remains smooth and the cascade plot is similar to symmetric stiffness.
C. Kxx = 535, KYY = 535, CxX = 035, CYY = 02

The rotor orbits and displacements are shown in Figure 10 for symmetric stiffness and asymmetric damping. The orbits are circular at slow speeds and become elliptical as the resonant speed is reached, however, backward whirl is not observed. The resonant speed is the same in both the x and y directions but the y displacement magnitude is greater due to the lower damping coefficient. The phase diagram and cascade plot are shown in Figure
15
11. The phase transition is smooth while the cascade plot shows the increased amplitude of vibration in the y direction. D. KXX = 535, KYY = 555, Cxx = 04, CyY = 02 For asymmetric stiffness and asymmetric damping the rotor orbits and displacement plots are shown in Figure 12. The orbits are circular at slow speeds and become elliptical near resonance. Backward whirl is evidenced at 2860 rpm which is approximately the midpoint of the maximum x and maximum y displacements. The phase diagram and cascade plot, Figure 13, show a smooth transition through resonance and a lower spectral amplitude due to the interaction of stiffness and damping. E. KXX = 535, KYY = 575, CXX = 04, CYY = -025 To investigate the decreased cascade spectral amplitude the stiffness asymmetry is increased while the damping asymmetry is decreased The rotor orbits and displacement plots are shown in Figure 14. The orbits begin circular and proceed into elliptical patterns as resonance is approached. The backward whirl condition remains through a wider range of speeds, between 2840 and 2920 rpm. The displacement peaks are further separated due to the increase in stiffness asymmetry. The cascade plot, Figure 15, shows no marked change in spectral magnitude and the phase diagram remains smooth throughout resonant transition. F. KXX=535, KYY=535, KXY=40, KYX=10, CXX = 025, CYY = 025 The rotor orbit and displacement plots for symmetric stiffness and symmetric damping with cross stiffness terms are shown in Figure 16. The orbits begin circular but quickly become elliptical as shaft rotational speed is increased. In the region of the y displacement dip, backward whirl is observed with the elliptical orbits perpendicular to those of Figure 14. The magnitude of the x displacement is almost twice that of the y displacement. The 16
cascade plot for the y direction is shown in Figure 17 and the low amplitude of response is depicted. The phase diagram, Figure 17, indicates a 180 degree phase shift at the first resonant peak, another 180 degree phase shift at the base of the first resonant peak and a final 180 degree phase shift at the second resonant peak.
G. KXX=535, KYY=535, KXY=10, Kyx=40, CxX = 025, CyY = .025

The rotor orbits and displacement plots illustrating the effect of reversing the cross stiffness amplitudes are shown in Figure 18. The orbits are similar, however the axis of orbit elongation is perpendicular to those of Figure 16. Backward whirl is evidenced
through the same speed range. The y displacement is now nearly twice the magnitude of the x displacement. The phase diagram and cascade plots are shown in Figure 19. The phase begins to shift as the first resonant speed is approached but is delayed by the large decrease in x displacement amplitude at 2800 rpm. The phase shift then proceeds smoothly through the rest of the transition zone. The cascade plot shows the increased spectral magnitude in the y direction.
H. KXX=535, KYY=535, CXX=025, CYY=025, CXY=-02, CYX=-01

The rotor orbits and displacements for symmetric stiffness and symmetric damping with negative cross damping terms are illustrated in Figure 20. The rotor orbits are slightly elliptical near resonant speeds and circular elsewhere. The large magnitude of
displacements in both the x and y directions stem from the negative cross damping terms. The magnitude of the x displacement is greater than the y displacement. The phase diagram and cascade plot, Figure 21, show a steep phase shift through resonance due to the negative cross damping and the cascade plot shows a large spectral amplitude at the resonant speed.
17
I. KxX=535, KYY=535, CXX=-025, CYY=-025, CXY=02, CYX=01 The rotor orbits and displacement plots for symmetric stiffness and damping with positive cross damping are shown in Figure 22. The rotor orbits become elliptical near resonance but remain in the forward whirl condition and the major axis of the elliptical orbits is perpendicular to those shown in Figure 20. The x displacement has a small area where the curve flattens (near 2800 rpm) and the y displacement magnitude is greater than the x displacement. The phase diagram and cascade plots are shown in Figure 23. The phase diagram shows a small region (at approximately 2800 rpm) where the phase curve flattens. This is the same region with the flattened displacement curve. The cascade plot indicates the spectral energy content. J. KXX=535, KYY=555, KXY=-5, KYX=3, CXX = 025, CYY = 025 The rotor orbits and displacement plots for asymmetric stiffness, symmetric damping and imposed opposite cross stiffnesses are shown in Figure 24. The orbits begin circular and become elliptical as the rotor approaches resonant speed The shaft does not achieve the backward whirl condition. The displacement plot shows a small double peak near the resonant speed. The phase diagram and cascade plot, Figure 25, shows a small overshoot in the phase transition. Spectral peaks are located near the resonant speed and at the forced response speed (operating speed) of the rotor. In conclusion, the analytic model, previously developed, does predict the split resonance and backward whirl phenomena noticed in [Ref. 1]. Distinctly different types of responses are generated depending on the type of stiffness and damping terms employed in the solution. The asymmetric stiffness model matches most closely to the experimental results obtained by Simei. [Ref. 1]
18
IV.
EXPERIMENTAL INVESTIGATION
The analytic results provided in Chapter III showed that asymmetric stiffness (difference in KXx and Kyy terms) created the split resonance and backward whirl phenomena similar to that experienced by Simei. [Ref. 1] To determine the physical cause of the stiffness asymmetry the rotor assembly hardware, used in [Ref. 1], was investigated. Finally as a direct measurement, accelerometers were positioned on the rotor assembly and a precision hammer (load cell) was used to provide a known impulsive force to the rotor. The horizontal (x direction) static natural frequency was found to be 46.00 cycles per second (cps), while that of the vertical (y direction) was 46.75 cps. The disparity in the two frequencies leads to the possibility of asymmetric supports or bearings as a cause of the split resonance. The lubrication regime within the bearing during the split resonance and backward whirl phenomena is unknown. Since the bearing is manufactured with oil impregnated brass, the lubrication may be hydrodynamic (either full or partial oil film), metal to metal rub or the shaft may actually lift off of the surface and be in aerodynamic lubrication. Possible causes were divided into four broad categories: non-circularity of the bearings, gravity effects, degree of imbalance and bearing support asymmetry. 1. Non-Circularity of Bearings
The shaft diameter is 0.375 inches with a design radial bearing clearance of 1 mil. The bearings were removed from the experimental hardware used in [Ref. 1] and one of the bearings was found to be 2 mils out-of-round (non-circular). The physical condition of the bearing and its orientation with respect to the two orthogonal lateral directions may lead to unequal stiffness terms in the x and y directions. This would, in effect, create different KXx and Kyy terms as used in the analytic model.
19
2.
Gravity Effects
The position of the shaft journal in the bearing may constrain the motion of the shaft within the bearing, thus increasing the magnitude of force necessary to cause the rotor to move in that direction. This type of situation can be directly related to stiffness variations in the two lateral directions. 3. Degree of Imbalance
The magnitude of the mass imbalance may generate a forcing function F(t) which is large enough to overcome the constraints of gravity and rotor centrifugal force and produce backward whirl. 4. Support Asymmetry
The bearing support structure is itself asymmetric. The vertical (y) direction is bolted to the assembly base plate, which may make the stiffness greater in the vertical direction. The horizontal (x) direction of the support is not hard mounted and may have more flexibility (less stiffness) as a result. Of the four possible physical causes presented only gravity effects, non-circularity of the bearings or bearing support asymmetry can create asymmetric direct stiffness. In order to determine the actual physical cause of the asymmetric stiffness, the experimental facility developed in [Ref. 1] was used to provide data for a variety of rotor imbalances and bearing and shaft orientations. The ROTOR.VI program developed in [Ref. 1] was used to collect the time trace data which was processed and plotted using a MATLAB computer program. The rotor assembly is shown in Figure 5 and a variety of mounting configurations and bearing arrangements was used to determine the effect of various parameters on rotordynamic responses. The data obtained from the various rotor configurations is The plotted results from the experimental
tabulated at the end of this Chapter. investigations are contained in Appendix C.
20
COUPLING
MOTOR
KEY PHASOR PROBE
Y-PROXIMETER MASS IMBALANCE SHAFT X-PROXIMETER
BEARING SUPPORT DISK SECTION A-A
Figure 5. Configuration of Experimental Rotor System.
21
A. GOOD BEARING TABLE MOUNT The rotor assembly was mounted on a table with a soft foam pad under the base plate. Both the inboard and outboard bearings, with respect to the drive motor, were replaced with circular bearings (good bearings) of 1 mil clearance between the shaft journal and the bearing. The foam pad acts as a low pass filter, thus eliminating any high frequency unwanted inputs to the system. The x direction is horizontal and the y direction is vertical, or in the direction of gravitational force as shown in Figure 5. 1. Good Bearing No Imbalance The rotor orbits for a good bearing (circular with 1 mil radial clearance; tolerance of 0.5 mil) with no shaft imbalance are shown in Figures 26, 27 and 28. The originally circular orbits become elliptical and collapse into a line at a resonant speed of 2840 rpm. Backward whirl is not observed. The orbits at twice critical speed (5600-6200 rpm) remain circular. The radial displacement and phase diagrams are shown in Figure 29. The phase shift is abrupt through the resonant speed while the displacement plot has a wide frequency range in which resonance occurs. The cascade plot shown in Figure 30 indicates the frequency corresponding to the maximum amplitude. A slight amount of spectral energy is evidenced in the (2x) and (3x) regions of the plot. 2. Good Bearing 0.5 Gram Imbalance The rotor orbits for a good bearing with a 0.5 gram imbalance are shown in Figures 31, 32 and 33. The resonant speed is much less abrupt, but a large amplitude response exists between 2800 and 2900 rpm. The orbits begin circular then become elliptical and collapse into a line at 2800 rpm. The orbits then become elliptical again, in the forward whirl direction, and again collapse at 2900 rpm. From this speed onward the orbits revert back into a circular pattern and remain circular at twice the critical speed. The radial displacement shows a slight double peak and the phase diagram indicates higher damping is evident by the gradual transition through resonance, Figure 34. The cascade plot, Figure
22
35, indicates a higher spectral energy content at resonance and (2x) and (3x) vibration levels are observed. 3. Good Bearing 1.0 Gram Imbalance
The rotor orbits for a good bearing with 1.0 gram imbalance are shown in Figures 36, 37, 38 and 39. The orbits are circular at slow speeds and become elliptical as the resonant speed is approached. The orbits collapse into a line at 2800 rpm and then proceed into elliptical orbits in the backward whirl condition. The orbits remain in backward whirl until 2890 rpm when the orbits collapse into a line again. The orbits leave this condition in forward whirl and remain forward throughout the rest of the experimental speeds. At twice the resonant speed two distinct orbits are evidenced. The radial displacement and phase diagrams, Figure 40, show an increase in split resonance range and a higher displacement in the y direction than the x direction. The phase plot continues to show slightly increased damping, however, the transition through critical speed remains abrupt. Figure 41 depicts the cascade or spectral information and indicates increased spectral content at the (l/2x) speed when the rotor is at twice the critical speed corresponding to the two distinct orbits at these speeds. 4. Good Bearing 1.6 Gram Imbalance
The rotor orbits for critical and super-critical speeds for 1.6 gram imbalance are shown in Figures 42, 43,44,45 and 46. The orbits collapse at 2760 rpm and proceed into a much more pronounce backward whirl condition through 2880 rpm. The orbits collapse again at 2890 rpm and proceed into another backward whirl orbit with the major axis shifted to a more vertical direction. The orbits abruptly change to forward whirl at approximately 3125 rpm and remain forward and circular through 7000 rpm. The orbits at twice critical speed again exhibit the dual orbit phenomena. Two very distinct peaks are noticed in Figure 47. The magnitudes of the x and y displacements are almost identical, however, the speed at which they occur is more widely separated. The phase transition, in Figure 47, shows a
23
decrease in the phase slope as the rotor approaches resonance. The peak displacement remains for a longer speed duration which causes the phase to remain at 90 degrees for a short while. Figure 48 shows the cascade plot and indicates spectral energy content at (l/2x) when the rotor is at twice resonant speed but also indicates high energy content at (2x), (3x) and (4x) frequencies when the rotor is at critical speed. B. BAD BEARING TABLE MOUNT To test the effect of a non-circular bearing on rotor response and particularly split resonance and backward whirl, a spare bearing was machined into an elliptical shape. The major axis was measured to be 0.3810 inches while the minor axis was 0.3760 inches. This represents a very non-circular bearing. Only one bad bearing was used in the experimental apparatus and positioned in the inboard bearing support (closest to the proximeters). The use of two bad bearings would have resulted in much higher vibrational displacements than can be measured with the proximeters in the positions illustrated in Figure 5. The outboard bearing was kept as a circular bearing. 1. Vertical Bad Bearing 1.0 Gram Imbalance The bearing was positioned with the major axis of elongation in the vertical (y direction). The rotor orbits for this configuration are shown in Figures 49, 50, 51 and 52. The orbits begin circular at slow rotor speeds and proceed into elliptical patterns as resonant speed is approached. The rotor reaches resonance at 2840 rpm as the orbits collapse into a line. A slight backward whirl is observed until the rotor reaches 2940 rpm at which point the orbits collapse again then proceed in a forward whirl motion and elliptical patterns. The orbits become circular again at 3200 rpm. At twice the resonant speed very large deflection amplitudes are present along with peculiar rotor orbits. The orbits become normal at 6000 rpm. The radial displacement and phase diagrams, Figure 53, show a very wide split peak
24
near the resonant speed and a second double peak near twice the resonant speed. The rotor exhibits a phase shift at each of these speeds. The cascade plot, Figure 54, shows the increased (l/2x) energy content at twice resonance and high vibration levels at (2x) and (3x) frequencies through all rotor speeds. 2. Horizontal Bad Bearing 1.0 Gram Imbalance The rotor orbits with the elongated bearing positioned with the major axis in the horizontal (or x direction) are shown in Figures 55, 56, and 57. The orbits near resonant speed are similar to those for the vertical position, however, the orbits at twice resonant speed are much less erratic. The radial displacement and phase diagrams, Figure 58, again show a split peak near resonance and another at twice resonant speed and the phase again shifts twice. The cascade plot is shown in Figure 59. 3. 45 Degrees Left Bad Bearing 1.0 Gram Imbalance The elongated bearing was positioned with the major axis in the general direction of the resonant displacement (approximately 45 degrees off the vertical). The rotor orbits are shown in Figures 60, 61, 62 and 63. The near resonance orbits are again similar to those for the vertical bearing direction and the twice resonance speed orbits are similar as well except that they occur at slower rotational speeds. The radial displacement and phase diagrams, shown in Figure 64, indicate a split displacement peak near resonant speed and at twice resonant speed, however the phase shift is less pronounced at the higher speeds. The cascade plot, Figure 65, indicates less spectral energy content at the (2x) and (3x) frequencies. 4. 45 Degrees Right Bad Bearing 1.0 Gram Imbalance The elongated bearing was positioned with the major axis perpendicular to the direction of resonant displacement (approximately 45 degrees off the vertical). The rotor orbits are shown in Figures 66, 67 and 68. The amount of backward whirl exhibited in this configuration is slightly reduced near the resonant speed. The orbits at twice resonant
25
speed are much more similar to the good bearing configuration with a 1.0 gram imbalance. The radial displacement and phase diagrams, Figure 69, more closely resembles the good bearing plots depicted in Figure 40. The elongated bearing causes the phase shift to become erratic at speeds above resonant speed. The cascade plot in Figure 70 shows no indication of energy in (l/2x) frequency range when the rotor is at twice resonant speed, however (2x) and (3x) frequency levels are observed. C. HORIZONTAL SHAFT MOUNT 1.0 GRAM IMBALANCE The rotor assembly was wall mounted with the shaft in the horizontal flongitudinal) direction with rubber mounts installed between the base plate and the wall board. The x direction for this configuration remains horizontal and the y direction is in the vertical direction (direction of gravitational forces). Good bearings were installed at both the inboard and outboard positions. The rotor orbits are shown in Figures 71, 72 and 73. The rotor orbits begin circular at slow rotational speeds then proceed into elliptical orbits. The orbits never collapse, however they rotate about the shaft centerline axis and at higher speeds the orbits are tilted off the vertical in a configuration which resembles a normal fluid lubricated bearing position. The displacement plots, Figure 74, show the decreased amplitude in the x direction due to the rubber isolation mounts and the effect of the wall board. The phase shift is very gradual, indicative of increased system damping. The cascade plot shown in Figure 75 shows the decrease in spectral energy at resonant speeds and a decrease in (2x) and (3x) vibration levels. D. VERTICAL SHAFT MOUNT 1.0 GRAM IMBALANCE With the rotor assembly wall mounted with the long axis of the shaft in the direction of gravitational force (vertically) the rotor orbits are depicted in Figures 76, 77 and 78. The orbits are circular at slow speeds and become elliptical near resonance. The collapse of the
26
orbits near 2600 rpm indicate a decrease in the resonant speed and at higher speeds the orbits exhibit characteristics very similar to the analytic orbits obtained using slight asymmetric direct stiffness in Chapter III. Since the rotor is mounted with the y direction in the direction of the wall mounts the displacements are smaller in that direction, Figure 79. The phase diagram shows a more conventional transition through the resonant speed. The cascade plot, Figure 80, shows a very low energy content in the (l/2x), (2x) and (3x) regions indicative of low vibration levels. The results of all of the previously listed experiments are provided in Table 1. The degree of backward whirl was significantly affected by the magnitude of the mass imbalance, but not significantly affected by the relative position of the elongated (noncircular) bearing. The responses obtained from the elongated bearing matched closely to the good bearing response at 1.0 gram imbalance except for the (l/2x) vibration levels and rotor orbits at twice the critical speed. The position of the rotor shaft, disk and mounting assembly relative to the gravitational force greatly affected the rotor response, however backward whirl was not evidenced in either of these configurations.
27
BEARING CONFIGURATION BAD BEARING (TABLE MOUNT) GOOD BEARING (TABLE MOUNT) IMBALANCE 1.0 gram 0.0 0.5 1.0 1.6 gram gram gram gram
1 X RESPONSE BW/SR NO BW/SR NO BW/SR BW/SR BW/SR BW/SR
1/2 X RESPONSE YES NO NO YES VERY MUCH VERY MUCH
VERY BAD BEARING HORIZONTAL (TABLE MOUNT) VERY BAD BEARING VERTICAL (TABLE MOUNT) VERY BAD BEARING 1 RESONANT DISP. (TABLE MOUNT) VERY BAD BEARING II RESONANT DISP. (TABLE MOUNT) GOOD BEARING HORIZONTAL (WALL MOUNT) GOOD BEARING VERTICAL (WALL MOUNT)
1.0 gram
1.0 gram
BW/SR
VERY MUCH
1.0 gram
BW/SR
YES
1.0 gram
BW/SR
VERY MUCH
1.0 gram
NO BW/SR
N3
1.0 gram
NO BW/SR
NO
Table 1. Results From Experimental Investigation
28
V.
DISCUSSION OF RESULTS
The previous analyses will be discussed as well as possible causes which produce the split resonance and backward whirl rotordynamic response phenomena. As expected, the experimental data is not as smooth as the analytic solutions generated by the model but a general insight into the overall characteristics of the rotor shaft and bearing system can be attained and useful information can be gleaned from the results. In all cases the response is greatly influenced by the forcing function F(t) which is created by imbalance. By varying the characteristic parameters (Kxx, KXy, KyX, Kyy, Cxx, CXy, Cyx, and Cyy) in the analytic model the rotor response can be significantly altered. In the first The
analysis, circular orbits were evident throughout the range of speeds of interest.
natural frequency of the system occurred at 2720 rpm which is lower than that obtained in the experimental analyses. The addition of a bearing stiffness term in the analytic model raised the resonant speed to 2840 rpm. The existence of asymmetric direct stiffnesses in the model caused elliptical orbits and the backward whirl phenomena. The addition of asymmetric stiffness terms caused a split peak in the x and y response amplitudes (split resonance). Asymmetric damping also caused elliptical orbits, however the rotor remained in forward whirl throughout the resonant speed range. Bearing cross stiffness terms lead to split peaks and elliptical orbits and the direction of the highest cross stiffness affects the direction of the rotor resonant response. Small
amounts of opposite cross stiffness cause a phase shift overshoot which is also evidenced in the 1.6 gram imbalance experimental results. Cross damping terms provide increased amplitude response if the terms are negative and cause elliptical orbits with forward whirl if the terms are small and positive. The magnitude of the imbalance affects the response of the rotor in a variety of ways. Increased imbalance magnitude causes increased x and y deflections. Increased imbalance
29
magnitude causes increased spectral energy in the (l/2x) region at high speeds and increased spectral content in the (2x) and (3x) range through all operating speeds. The cascade plots show that the dominant energy is contained at the operating speed of the shaft. Even with no imbalance imposed, the shaft will have relatively high vibration levels at the resonant speed of the rotor/bearing system. The 1.6 gram imbalance causes
unexpected responses at the resonant speed and at multiples of the resonant speed, as evidenced by the increased energy level at the (2x), (3x) and (4x) frequencies. The
unorthodox orbits exhibited under this loading are possibly caused by a non-linear response generated by this imbalance. The elongated bearing provided increased response amplitudes in all cases except when the bearing was positioned to oppose the resonant displacement. The orbits at twice the resonant speed show marked differences in appearance and orientation. The
supercritical orbits with the bearing positioned with the elongation in the direction of resonant displacement closely match those with the bearing positioned vertically. The long axes of the orbits are rotated 45 degrees and the speed at which they appear is slightly less but similarities do exist. The response of the rotor with the elongated bearing positioned perpendicular to the direction of resonant displacement is more similar to the good bearing response with 1.0 gram imbalance. The cascade plot for this configuration shows the decrease in energy content at the operating speed and the low energy levels at the (l/2x), (2x) and (3x) regions. Mounting the shaft in the horizontal position, in effect, changed the gravitational effect on the shaft caused by the bearing support structure. The y displacements remained similar to the table mount experiment, however the x displacement amplitude decreased due to rubber isolation mounts being installed between the base plate and wall board. The phase transition was very gradual indicative of high damping in the system. The orbits follow a more conventional, fluid lubricated, bearing orbit as the rotor speed is increased
30
With the shaft mounted vertically the effect of gravity on the bearing and bearing support structure was removed. The displacement in the y direction was damped out by the isolation mounts and wall board, however the orbits showed a pronounced resonance at 2720 rpm. This resonant speed corresponds to that achieved with only the shaft stiffness term in the analytic model.
31
32
VI.
CONCLUSIONS AND RECOMMENDATIONS
A. CONCLUSIONS Using the experimental facility developed in [Ref. 1] and an analytic model to predict rotordynamic phenomena, a deeper understanding of the physical mechanisms that create split resonance and backward whirl was obtained. Analytic variables can be altered to change the rotor output, in effect altering the transfer function which converts the forcing function input to the rotor response output. The analytic model strongly suggests that the KXx and Kyy (direct stiffness) difference is responsible for the split resonance and backward whirl phenomena The modal test confirmed that Kyy is greater than Kxx on the experimental rotor assembly, at least statically. The effect of mass imbalance on the experimental aparatus was shown as well as the effects of non-circular bearings. The rotor system orientation with respect to the gravitational force coupled with the installation of rubber isolation mounts causes rotor responses to change. Backward whirl can be induced by split resonance caused by a mass imbalance or gravitational force in the bearing. The effect of bearing support structure stiffnesses may also cause split resonance. If the bearing is more rigidly mounted in the direction of the bearing support bolts, the y stiffness may be greater than the x stiffness. This asymmetry may cause higher forces in the y direction leading to elliptical orbits and backward whirl. Asymmetric damping within the bearing does not cause the backward whirl phenomena but does produce elliptical orbits. Backward whirl is not caused by bearing non-circularity.
33
The vibrations of the rotor shaft differ depending on the orientation of the elongated bearing. The response amplitudes are higher if the elongation is in the direction of the resonant displacement of the rotor. The orientation of the non-circular bearing does not significantly alter the backward whirl orbits. The bearing stiffness values are a function of the force applied, therefore the bearing stiffnesses change with rotor speed and with rotor orientation with respect to the gravitational field. Vertically mounted rotating assemblies will have a critical speed near the natural frequency of the shaft since bearing stiffness is negligible. At higher speeds,
however the effect of bearing stiffness must be included in the analysis of the machine. B. RECOMMENDATIONS There are many areas yet to be investigated using the rotor kit and data acquisition software. The most prevalent include: 1. Developing an analytic/experimental procedure for rotordynamic parameter identification. This would entail solving for the coefficients in the analytic model using known experimental rotor responses. 2. Developing an optimization technique for determining the analytic model parameters (Kjj and Qj). Given a rotor orbit a mathmatical equation for the response could be written. 3. Study in detail the physical mechanism through which gravity influences the stiffnesses in the lateral directions. This would most likely require a structurally sturdy but configureable mount to position the rotor assembly in any position in three dimensional space. 4. Conduct a detailed analysis of the lubrication regime within the bearing during backward whirl. The forces acting on the shaft journal at the bearing interface would need to be quantified along with lift and drag forces generated by the rotation of the shaft.
34
Additional areas of research include the investigation and determination of the dependence of bearing stiffness on shaft rotational speed, investigation of fluid lubricated bearing dynamics, roller bearing dynamics, and the effect of different isolation mounts on vibrational characteristics. The analytic model developed herein should be modified to allow for the bearing stiffnesses to change as shaft speed increases. This will provide a more realistic rotor response. The rotor kit is provided with two translucent fluid lubricated bearings. The wide use of fluid lubricated bearings in rotating machinery installations warrants research in this area. Use of the translucent bearings will provide both visual and time trace data to evaluate the rotor dynamic characteristics based on hydrodynamic and possibly hydrostatic bearings. The rotor can be configured with one or two fluid lubricated bearings. The use of roller bearings on marine gas turbine installations motivates the research into roller bearing induced shaft vibrations. The amplitudes of response are expected to be higher than the brass oilite bearing since the stiffness of these bearings is higher. The rotor kit is provided with a roller bearing which is mounted between four adjustable springs. This will provide for increased control of the bearing influence on rotor response. The rotor assembly should be hard mounted on a sturdy base which can be adjusted to configure the rotor in a variety of positions with respect to the gravitational field. Additional types of shock mounts should be obtained (with different stiffness values) and sequentially installed to determine their affect on rotor response.
35
36
APPENDIX A.
LIST OF DIMENSIONS AND PROPERTIES
Rotor Shaft: Material: Modulus of Elasticity: Length: Diameter Mass Density 4140 Low Alloy Stee [ 200 GPa 45.7 cm 0.9525 cm 0.253 kg 7766 kg/m3 29.2(106)psi 18.0 in 0.375 in 0.558 lbm 0.281 lbm/in3
Rotor Disk: Material: Width: Inner Diameter Outer Diameter Mass 316 Stainless Steel 2.54 cm 0.9525 cm 1.18 cm 0.816 kg 1.0 in 0.375 in 3.0 in 1.8 lbm
37
38
APPENDIX B. ANALYTIC ROTOR DATA
40
=2500rpm - =2740rpm x =2800rpm
30
20
T =2940rpm
- =2900rpm
I
I 0
j2
=30C0rpm
a. <n
>-
5-10 20
it
-30-
-30
-20
-10 0 10 X displacement - mils
20
30
40
3000
4000
7000
speed rpm
Figure 6. Rotor Orbits and Displacements From Analytic Model For Kxx = 495 lbf/in, Kyy = 495 lbf/in, Kxy = 0 lbf/in, Kyx = 0 Ibf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
39
2(K
15-
ROTOR SPEED - (200X rpm)
FREQUENCY - (4X Hz)
3000 4000 ROTOfl SPEED (rpm) Kxx=<t95.Kyy=495.Kxy=0.Kyx=0 */in Cxx= 02S.Cyy= 025.Cxy=O.Cyx=0
7000
1000
2000
3000 4000 ROTOR SPEED (rpm)
5000
3000
7000
Figure 7. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 495 lbf/in, Kyy = 495 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/m, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
40
<xx=535.Kyy=555 Kxy=0.Kyx=0 *'in Gxx= 025 Cyy= G25.Cxy=O.Cyx=0 -10 r =2~-iCrpm - =2300rpm < =2840rpm 20 - =2920rpm - =2960rpm c 10 - =3000rpm
30
.2
Q.
o -10
>&*** *^:
-20
30
-30
-20
-10 0 10 X displacement- mils
20
30
-iO
"0
1000
2000
3000 4000 spaed -rpci
5000
6000
7000
Figure 8. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
41
20-,
FREQUENCY - (4X Hz)
Kxx=535.Kyy=555.:<xy=O.Kyx=0 */in Cxx= C2S.Cyy= 025.Cxy=O.Cyx=0
1000
2000
5000
S000
7000
Kxx=535.Kyy=555.Kxy=0.Kyx=0 #in Cxx= 025.Cyy= 025Oxy=0.Cyx=0 50 0 -50 100 2< 150

LU 1/1
<; 200 a-250

1000
2000
ROTOR SPEED (rpm)
3000
4000
5000
5000
7000
Figure 9. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 Ibf/in, Kyy = 555 Ibf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/m, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0 42
<xx=535.:<yy=535 <xy=0.!<yx=0 *'in C<x= 035 Cwy= 32 Cxy=O.C/x=0

40r
=2740rpm =2800rom =294Crpm

xxXJOOcc<Xx
30
20
=2920rpm =2960rpm
5 10t-
=30C0rpm
i2
a.
a-10
-20
-30
-30
-20
20
30
40
"0
1000
2000
3000
4000
5000
3000
7000
spead -rpm
Figure 10. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 Ibf/in, Cxx = .035, Cyy = .02, Cxy = 0, Cyx = 0
43
20>
ROTOR SPEEO - (200X rpm)
FREQUENCY - (4X Hz)
Kxx=535.Kyy=535.Kxy=0.Kyx=0 *in Cxx=035.Cyy= Q2.Cxy=0.Cyx=0
-Q u
70
2000
30 ^ ROTOR SPEED (rpm)
OO
"000
Kxx=535.Kyy=535.Xxy=0.Kyx=0 *in Cxx= Q35.Cyy=.02,Cxy=0.Cyx=0
1000
2000
3000 ^000 ROTOR SPEED (rpm)
5000
000
7000
Figure 11. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 Ibf/in, Cxx = .035, Cyy = .02, Cxy = 0, Cyx = 0 44
Kxx=53S Kyy=555 xy=0 :<'/*=0 -'in Cxx= 04 Cyy= 02 Cxy=O.C,x=0
40r-
=2~40rpm =2900rpm x =2840rpm
30h
20
=2920rpm
- =2S60rpm - =3000rpm
o -10
-20
-30
-40. 40
-30
-20
10 0 10 X displacement - mils
20
30
40
-0
1000
200
30 4 speed - rpm
5000
7000
Figure 12. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/m, Cxx = .04, Cyy = .02, Cxy = 0, Cyx = 0
45
FREQUENCY - (4X Hz)
Kxx=535.Xyy=555 Kxy=0.Kyx=0 #/in Cxx= 04.Cyy= Q2.Cxy=0.Cyx=0
^.50 50
Z40'
S 30)
2 20|a
in
1000
2000
5000
6000
7000
Kxx=53S.Kyy=555.Kxy=0.Xyx=0 */ln Cxx= Q4.Cyy-02.Cxy=O.Cyx=0

0
']>
SO 0 -50 10 150 200 250
3>
T!
s <
J-
LU I
tu
Tool
20
5000
6000
7000
Figure 13. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .04, Cyy = .02, Cxy = 0, Cyx = 0
46
:<xx=535.Kyy=575,.<xy=0 <yx=0 -'"m Cxx= 04 C/y= J25 Cxy=O.C/x=0

40r
=27-l0rpm - =29C0rpm x =2940rpm + =2920rpm -- =2960rpm
30
!= 10
- =3000rpm
u -10
-20
-30
-40, 40
-30
-20
20
30
40
40
!L
a-30
>X
20
10
n "0
1000
2000
3000
4000
5000
6000
7000
speed-rpm
Figure 14. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 575 Ibf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .04, Cyy = .025, Cxy = 0, Cyx = 0 47
FREQUENCY - (4X Hz)
Kxx=535.Kyy=575.Kxy=0.Kyx=0 */in Cxx= 04Cyy= 025.Cxy=O.Cyx=0
'0
1000
2000
5000
5000
7000
Kxx=535.Kyy=575.Kxy=0.Kyx=0 */in Cxx= 04.Cyy= 025.Cxy=0.Cyx=0
LU
">-200
Q--
1000
2000
5000
5000
7000
Figure 15. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 575 lbf/in, Kxy = 0 Ibf/in, Kyx = 0 Ibf/in, Cxx = .04, Cyy = .025, Cxy = 0, Cyx = 0
48
Kxx=535 Kyy=535 Kxy=40 Kyx=10 */in C<x= 025C/y= 02S Cxy=0 C/x=0 40 r =2740rCm =29C0rpm x =2S40rpm 20 i- =2920rpm
30
- =2960rpm
10
i>
- =3000rpm
1 o 'J
^-10
-20
-30
-30
-20
-10 0 '0 X aisplacament - mils
20
30
40
40
u
o 5X
-30
y 20
10
J
0 1C
--
speed - rpm
Figure 16. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 Ibf/in, Kxy = 40 lbf/in, Kyx = 10 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
49
ROTOR SPEED - (20QX rpm)
FREQUENCY - (4X Hz)
Kxx=53S.Kyy=535.;<xy=40.Kyx=10 *"n Cxx= 025.Cyy= 025.Cxy=O.Cyx=0 & -40 I
UJ
c 3 20 a
510 x _
<f> -
ui30
1000
2000
5000
5000
7000
Kxx=535.Kyy=535.Kxy=40.Kyx=l0 #/in Cxx=.025.Cyy= 025.Cxy=0,Cyx=0 50
*
LU
0 50
UJ in
2 -100 < -150

-200 1000 2000 3000 4000 ROTOR SPEED (rpm) 5000 6000 7000
a-250
Figure 17. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 Ibf/in, Kxy = 40 Ibf/in, Kyx = 10 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
50
40 r
Kxx=535.Xyy=535 Kxy=!0.:<yx=40 *1n Cxx= 02S Cyy= 325 Cxy=0 Cyx=0
=2740rpm - =2900rpm x =2840rpm
30
20 - * =2920rpm =2960rpm 5 10 =3000rpm
I 0
2
<n
a.
'S -10-
-20
-30
-30
-20
20
30
40
1000
2000
3000 4000 speed - rpm
5000
6000
7000
Figure 18. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 10 lbf/in, Kyx = 40 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
51
20-,
15>
10-
FREQUENCY - (4X Hz)
Kxx=535.Xyy=535.Kxy=10.:<yx=40 *fin Cxx=.025Oyy=.025.Cxy=0 Cyx=0
70
2000
3 ROTOR SPEED (rpm)
5000
5000
7000
Kxx=535.Kyy=535.Xxy=10.Kyx=40 #"n Cxx= 02SOyy= 025.Cxy=0Oyx=0
1000
2000
3000 4000 ROTOR SPEEO (rpm)
5000
S0OO
7000
Figure 19. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 10 lbf/in, Kyx = 40 lbf/m, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
52
40
Kxx=535.Xyy=535.Kxy=0 Xyx=0 */in C<x= :2S.Cyy= 02S.Cxy=- 02.C/x=- 01 =2740rpm =290Cwrn x ^X:2940rpm
30
20h - =2920rpm -- =2S60rpm 10 - =30C0rpm
\1
a T3-10 20
-30
*.
-40. -40 -30 -20
T*-*-
.* .** *
*'
20
30
40
3000
4000
7000
speed - rpm
Figure 20. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 Ibf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = -.02, Cyx = -.01
53
20-,
FREQUENCY - (4X Hz)
Kxx=535.Kyy=535.!<xy=O.Kyx=0 */in Cxx= 02S.Cyy= 02S.Cxy=- 02.Cyx=- 01 50
!50 \
LU
^40
20 10 1000 2000 3000 4000 ROTOR SPEED (rpm) 5000 6000 7000
Kxx=535.Kyy=535.Kxy=O.Kyx=0 #/in Cxx=.025Oyy=025.Cxy=- 02.Cyx=-01 * 9> 303' 0 50
m m 200 a -250
1000 2000 3000 4000 ROTOR SPEED (rpm) 5000 5000 7000
3- 100 < 150

im.
Figure 21. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = -.02, Cyx = -.01
54
Kxx=535.Kyy=535.;<xy=0 Xyx=0 *fm Cxx= 025 Cyy= 325 Cxy= 02.C/x= 01 40r =2740rpm - =2800rom x =2840rpm 20 * =2920rpm - =2960rDm i= 10 =3000rpm
30
O-10
-20
-30
-$r
-30
-20
-10 0 ' 10 X displacement - mils
20
30
40
3000 4000 speed-rpm
000
Figure 22. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 lbf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = .02, Cyx = .01
55
FREQUENCY - (4X Hz)
Kxx=535.Kyy=535.<<xy=0.Kyx=0 */in Cxx= 025.Cyy=.025.Cxy=.02.Cyx= 01
1000
2000
3000 4000 ROTOR SPE50 (rpm)
5000
5000
7000
Kxx=535.Kyy=535.Kxy=0.Kyx=0 *"n Cxx=.025.Cyy=025.Cxy-02.Cyx= 01
50 ' 0
-50 -100 -150 -200 -250 1000 2000 3000 4000 ROTOR SPEED (rpm) 5000 5000 7000
Figure 23. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 535 lbf/in, Kxy = 0 Ibf/in, Kyx = 0 lbf/in, Cxx = .025, Cyy = .025, Cxy = .02, Cyx = .01
56
:<xx=535 Kyy=555 Kxy=-5.Kyx=3 */in Cxx= 02S.Cyy= 025 Cxy=O.Cyx=0
-40
-30
-20
20
30
40
"0
1000
2000
3000 4000 speed - rpm
5000
6000
7000
Figure 24. Rotor Orbits and Displacements From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = -5 lbf/in, Kyx = 3 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
57
FREQUENCY - (4X Hz)
Kxx=535.Kyy=555.Kxy=-5.Kyx=3 #in Cxx= 025.C/y= 025.Cxy=O.Cyx=0
1000
2000
5000
5000
7000
Kxx=S35.Kyy=555.i<xy=-5.Kyx=3 #in Cxx=.02S.Cyy= 02S.Cxy=0,Cyx=0
*
9 T>
50 0
ill
4m
-so
150
a- 1
<
LU
2-200 a-250 1000 2000 3000 4000 ROTOR SPEED (rpm) 5000 5000 7000
Figure 25. Cascade Plot and Phase Diagram From Analytic Model For Kxx = 535 lbf/in, Kyy = 555 lbf/in, Kxy = -5 lbf/in, Kyx = 3 lbf/in, Cxx = .025, Cyy = .025, Cxy = 0, Cyx = 0
58
APPENDIX C. EXPERIMENTAL ROTOR DATA
C"3 "981" ^c>^ r.^lTCAL iP'rzO C :q WBAt^NC3
SC^CP CP8IT3 *<S*fl C.^lT'OL J?==-3 0 2q .MB/*L-NC3
!5r27<iCrom :oi93/T T1KW1 . :0J- ' -jn<j 5r niCtS gOOrom
~ :0t--
I
i'r
SlOI151-
5
3
I
31-
5; >>
,-'
I
tSi-
*"->0l-
i aoi!
t :IS?*_AC2>HSNT miiai
I
1
20
IS
-i0
i ) < "ISP'-^CSMGNr mils)
'S
:o
"Ts
TOTCR CflSlTS N6>H caiTlCAC 3PK0 :0.0q IMBJM^NCS!
ISi-
:oi'33TT
"[ .-MO,
TllOOl
/
- si3
1 OF I /
101-
IT
151-
15p
Z0K
i
) i
Jflj. is
f
10 -5 ) : '-3
-:o
-'S
;Q
-i
:0
25
Figure 26. Rotor Orbits No Imbalance Good Bearing. 2760 to 2900 rpm.
Speeds from
59
3C"p 0F8IT" ê.*a dirc-AL :s^-
JPSEO
:f
) oq
MSAL^NC":
*C~?CP CR8i" vg.a -ntrCAL 3P"==0 '0 Cq ;990rcm
U8AC-NC31
:-> - aart
rued ~ 'Of-' *fla
:oh
jun
~ "Oh -*TKJ
n
/
-is -ia -i > . ' :Q -s JO JS
^
<->
3h
i
I *[ 5
-iSr -201~5 - I < 2iS?'w*C3M6NT mit*) -0
-<
- oi-
isi-zoii
:s :o :s
zy-. is
-JO
5C"fl C39ITS .^SAP C3ITCAL 3P553 >0.3q iMBiLANCTI 2Sr OQOrom !0h Sr
2S, 1 32QCram
20I -Mfl
TOTCS OBSfTS N6>B CSmOL 3??=3 <0 Oq
IM8A(_4NCS1
airt
1
<
a
'-.a.
= =12 1
! I
1
01-
r\
x
.
'0
$ 4 3 i
i
-151-201-
it
1
>".iOh tSh
1
I
rah
TS
3 i > ^ x oisp'.^cakieir -ma
>s5
Is
'15
JO
iS
10
3 J i X aSPVCEMS-Nr :mui
'S
JO
Speeds from
60
3CTOR CA9IT3 A7 UP=3 C^rPCAC i?*-zO :0 Oq iUaAl_ANC3)
=C"Cf CB9IT3 *r XP==1 -lirCAL i??z-3 0 Oq BALANCE
5500rcm - aan
fSOOrom - sa/l .-nice *ia
i
= :l-
1KG
5 i
a it3 31-
,^1
3 I
!'!-
I i
5 I
Si
] f
Jt-
""'1"i
31-ll< GlSP'-^CS-'iSNT mrtsi X îSP'-^CSMENr musi
aQTCfl CP3IT3 Af 3UP53 ~3iT!CAC 5?==D O-Cq
IM8AU*NCS1
=IOTCB CPB17S A7 3UP?3 CSIT1CH. 3PS50 'O.Oq iMB*C*NC31
Al-
SOOOrom -son ntod ' md
I
31rodda
if
'J < ! 1 a*'!'
311
5 ! ah
\ ) v. .
3 I
3!
n
-2 ) X 3i&-ACSME.NT ;mji
*"-*h
4
3 -2 -i ) X aiSP'-XSMENT mat 2 3 3
Speeds from
61
MAX <Y ÎSP'-ACSMSNT /S. 5PEEC 0 'c. MSALAMCc
1000
2000
3000 -1000 ROTOR SPEED {rpm)
5000
5000
7000
MAX DISPLACEMENT /s. ROTOR SPEED ( 0.0 gram imbalance ) 50 r|50h
fejola i 3! I
i
J
30 P r <2ok
0 1000 5000 3000 4000 ROTOR SPEED (rpm) 5000 5000 "COO
V2 101-
PHASE ANGLE VS. ROTOR SPEED ( 0.0 gram imbalance )
3> LU
S solOK
-50
'3-100
LU
<-lS0
^-200 a-250 0 1000 2000 3000 -1000 ROTOR SPEED (rpm) 5000 5000 7000
Figure 29. Maximum Displacements and Bode Plot No Imbalance Good Bearing
62
2CU
150
FREQUENCY - (2X Hz)
Figure 30. Cascade Plot No Imbalance Good Bearing
63
3C"C?> CP91T3 ^ClirCAL 3P<==C O.Sg MgA!_^NC3: TSCtzn Id-
30TCR C38IT3 ME C3I"'CJL 5?3 0 5<J U9il^nC~: 25r:900rsm son
si-
I
- roi- "
ntca
- >i-
X'
/
i I 2 Oh
'Oh ISr
*>.JO ->5 -'0 -i ) ; X OlS^'-ACSMeNT must -a
3 I
a
>
"-;oiisiI -JOl.?!_
wi-55
:v
:s
-JQ
-'S
-'Q
-5
'5
:a
^5
aarcfl snarrs
M*A CRITICAL 3?=O IO.SJ IM8AU*NCS
'OTCR C3BITS N4fl CPlT'CAL 3?=C 0 fg
.MSAI>NCK
201-
30 V
j -art
,s
j- TO
-'IK
= 'i'5 i i > i
< j
y
3
mioela Ol-- *w
t Oh
y
2010 S 20 25
TTs
ioisao
25
Figure 31. Rotor Orbits 0.5 Gram Imbalance Good Bearing, from 2760 to 2900 rpm.
Speeds
64
^C'CF" -39I7S *6.+ CîTCAL 2??=:3 'O.Sq
:M9AL>NC31
3QTCP CfSi^S NgAfl C^rrCAL 3?==3 <Q 5q !9SOrorti
'MBALANCII
.*
i SJ1
ritaatm *od
/i
:t-
Oh
< I i -at"'01iSr [
-iSh -201-
201-
:a
:s
=OTCP "S8tn .is.- CîT'CAL i?K3 '0.5q ua.'.>NCS!

JOOCrom
WOB CH3IT5 Ml came* 3PSSO lO.Sj :M3*L>NCC

!200rom
zo>
>5r -Oh id
= ji
<
J
ih
|
,n
-is -io -5 ' i -a * OlSP'JCSMgNT -IM -5 aa
20t-'
i 5K . lOt--
art UK
51-
3 li5
w
1 I
2
-rOK -1SK
i*
-.,o(.
n
a is :a :s
-zoiJO
,sr -aoiI
$-
Speeds
65
aCTCP Cfl8i*S
A*
SUP?1'
C^T'CAL
SP==3 '0 5g
'M9AI_ANCS
COrcm - aart
.7TKJC3
' **1
-\
5
5
'fr
0h
< ClS?'-4CSM6NT ,,-niisi
X DlSP'-ACSMSNT imiisj
30TC0 C09ITS
A7
3UP=* CPmCAL 3P==D (0 Sq .BALANCE
ROTQP CPSiTS
AT
3UPS3 CamCAL 3PS50 f0 Sg IMBALANCE
if
SOOOrom -n fflxX
13h
3200rwn
3K
1 ' nd
- 1f"
> v
<
- II-
^"\
5 !
ffi 01-
-li. 71
a i *" -2H -3t-
5 <
S-'i'2!-
i 01I -I3
-*(X OlSP'-ACSMENT iflw)
X OlSP*JCcMNT imn;
Figure 33. Rotor Orbits 0.5 Gram Imbalance Good Bearing. from 5600 to 6200 rpm.
Speeds
66
MAX
XY
TJiSP'-ACSMSV 'S. 3PEE7; Ci? M8A..ANCE'.
50h
-tO
:3;3o
201-
1000
2000
3000 -1000 flCTCR SPEED (rpm)
5000
5000
7000
MAX DISPLACEMENT vs. RCTCR SPEED ( 0.5 gram imoaianca -
30 2 50CJOLU
| 30 o
a 52 10-
<20h -4
0
1000
2000
3000 iOOO ROTOR SPEED (rpm)
5000
000
"000
PHASE ANGLE VS. ROTOR SPEED ( 0.5 gram imoaianca )
-250 1000 2000 3000 -i000 ROTOR SPEED (rpm) 5000 3000 7000
Figure 34. Maximum Displacements and Bode Plot 0.5 Gram Imbalance Good Bearing
67
20-,
15-
FREQUENCY - (2X Hz)
Figure 35. Cascade Plot 0.5 Gram Imbalance Good Bearing.
68
aC'CP CPS'TS NEAfl CSrTiCAL 3P==D <* Oq
IMBALANCC!
aC"Cfl C08r3 N6*f CRITICAL 3P==D;i Og'MBALANCEI
- am maa*
j
!5h
- start niCClt *na
. ; *
2 Oh
iah
1 5
o
3r
i
3>Sr
-<
5
-1 I
"10
-1 sfI
**
<0h
*>
I5r
-2oi13 TTs Ho 1 3 5 X OISPI^CSMSNT ,m> io is 20
-i
201-
I5 x ;is?'-cSMeN7 im
20
25
SCTCP CUBITS
2S
CSITICAL $PE=O
3g
IMBI>NCS!
aOTOB on BITS ! 2S60rorn 20t-
NE.B
cnmCAL SP5SD
09
IMBALANCE
I 2810rom 20115! mtaat

tf
mioffl I
<f
Ok ' ana
toi- - w
r
5h
31-
I ' a oi3 i
a ''li
"lOh <5r 2012ft -75 -20 -15 -'0 -5 : X OBPLACEMgNT mi !0 15 20 25
5 i -.loi-
15-
201.3* ii I, it .in
X OISPV^CSMENT iiM
Speeds
69
noron onarrs
NEAR
cnntCAL SPEED M Og IMBALANCE)
SOTOR
CRsrrs
NEAR
carncAi.
3?E=3
' :g
IMBALANCE:
25r920rom
turn
'
15r
/^\
i
_ !0- and
5 -si'I0|15r
>
-20 15 -10 -5 0 S X DISPLACEMENT |m| 10 15 20 25
20h 20
-IS
-10
-5 ) 5 X DISP'-ACEMENT ;mf
-0
15
20
25
OTCH
CBsrrs
NEAR CRITICAL SPEED
t Oo IMBALANCE!
201-
ROTOR ORBITS NEAR CRITICAL SPEED " 3g
IMBALANCE:
201-
I-
'T
|
l
r\
3
M
I5J-
!
Oh
01-
s >
10 r 15 20 '-I0I-
5 20 15 -10
\ _.
-5 1 5 X DISPLACEMENT imul 10 15 20
|
23 :o -: : : X DISPLACEMENT mw io
Speeds
70
*C"B CASH'S AT 3U5=. CSfTiCAL 3P?ED Jg .MBALANCC)
3C"Q CP3'"3 *" 3U== CPfT'CAL S?==0
' :<; 'M8AL>NC=1
sun
.modi
<>
\
3 Oh
"V
3 I
213h
-31-
X OISP'-ACCMENT fmtisi
X DISPLACEMENT imnsi
^CTQfl ORStTS AT SUPS-* CHlTiCAC 3PSS0 " Og IMBALANCE! Si
SQTQfl ORBITS AT SUPS? C3rr;CAl 3P5SO '1 3g IMBALANCE}
it-
31,."" >
I"
5'l
a oi-
V\
2 01-
<
ii I
3 I
5 -31i
1 l ' 3 X OSPUkCEMCMT imM 2
54
*-*!
3h
X 3ISP1.ACSMENT IIM)
Speeds
71
^CTCO CPBITS AT SUPS1' Cfn-ICAL 3P*=3 M 0<j -M8AL>NC1
5r
4 i> OOOnxn
3r 21^ 'LU
- sarr Tiioa na
f 3i.
%.^
-3 -2 J X DlSP'-ACcMeNT tiros)
3CT0P. CBBlTS AT 3UP== CPrr'C-u. SP?53 !l 3q IMBALANCE
itI 31-
5200rom
IT
3 I
3 Oh
< !
""s.^,/
2\I
-2
> X DISPLACEMENT imltl
Speeds
72
MAX
:<y DISPLACEMENT /s. $?HC ' ;g MSAUA.NCC
-0
1000
2000
3000 A000 ROTOR SPEED (rpm)
5000
5000
7000
MAX DISPLACEMENT vs. ROTOR SPEED ( L0 gram imOalancs )
1000
2000
3000 -1000 ROTOR SPEED (rpm)
5000
5000
7000
PHASE ANGLE VS. RCTCR SPEED ( 1.0 gram imtialanca )
1000
2000
5000
5000
7000
Figure 40. Maximum Displacements and Bode Plot 1.0 Gram Imbalance Good Bearing.
73
FREQUENCY - (2X Hz)
Figure 41. Cascade Plot 1.0 Gram Imbalance Good Bearing
74
*OrCfl CHirs Ng*fl CPiftOi. SPSE'J ii 'iy
.M>H.-*NO)
ROTCrt 0P6irs .'JE.-'^'îr.CAL iPSSO I *j
.M,.L.+ NC
I '0
jl-
f
.3$ .0 -i5 10 -5 3 5 10 A OiSP'.ÔtMtNr (mas)
1 u
*4
25
.ô
IS
-10
-5
io
;s
20
25
SO rCH OflS* T5 Mc.*A
CÎTICAC
SPS=Q I t g IMBALANCE)
i
23 J790<tan Jt
...
25 20 IS 4*
sorofl oflsirs e^fl cairoi.
SPSSO
o dg
IMBALANCE)
283Uipni
_
3
5
ITUkM
*
'3
|,0
' #n<l 4
-7 *
*
1
O
1
*
r
X t
iu
<
* / *
a
>- 10 h 7 -20 25 25 20
*. lO
-2
<L
25 20
as
IS '0 5 3 S 10 OttS?l/tCEyENr |mu| IS 20 25 -15 -10 -5 3 5 )Q X OcJPVACSM6fr \tnt*t 15 20 25
Speeds
75
.jorcHCfiBirs.Ne.iflcairtCAL speso (i *j eAt^NCs
aorcrt ORairs Ne.->fl cai HCAL iP?=o (i ^ ,UdAi 25iI
.NCE
2
C
iSH
i
5 5-
23im
mdam
J iol-
7
IS 10 1 i 10 15 20 25
Z II-
<
,- I0(15-
X
20 25
2S
JO
25
-20
-IS
-10 *
DISPLACEMENT
i 10 (mitii
la
20
2S
ROTOfl OA81TS NI=Afi CRITICAL SPEcO O q IMBALANCE)
AOTOR ORBITS NI=-*H CRITICAL SP=0 t1 Sq IMBALANCE! 25
25 20IS 290ft om
>94Q<wn
//
2C
,
z
>on
ntOUM
|
3
5
</l Q
?' s
' *n
*
CJ VI -i -
<
,- -toIS 20 25 25 20
tO
15 20
y
-IS -10 i 0 S 10 X OlSi^L-ACEMfiN r [tndti iS 20 25
25 -25 20 -15 10 5 3 S 10 X DISPLACEMENT titusi 15 20 25
Speeds
76
sorcfl ofleits
N.*P. CRITICAL
aPesoo 3giM8AL*NC)
25 20
*OrOfl ORBITS
N6*
CRITICAL SP0
dy
,M8ALNCE)
JUtWfWn
i5-
15
-nitJOM
3" iu-
| 10
I
<
a
-S
>- 10 IS20-
zs
J5 20 .15 10 5 0 5 '0 A OidPt-ACcMfcNt unusi 15 20 25 -25 -20 -IS -10 -5 0 5 10 < OiSPt-ACcMtiNr (mJiJ IS 20 2S
flran
CHSITS
**
CRITICAL SPSSO
i i j
IMALANCS
ROTOR
oftfltrs
NEAR CRITICAL
Sfleso (i iq
IMBALANCE)
3200* on
i2
.25
-20
15
-10
-5 0 S 10 X DISPLACEMENT (mMi
tS
20
25
Speeds
77
scroflcpeiri r ;ur>s^ cmrcv. i?s=Jn TOTOofisirs"t upsacmrc.

SP5DII ÎMBM-WCSI
*JM8*HIW."=I
i .
ioOOfwn
-<
*'**....'
-5
'2
* OlS?l.*sMl:N r niMal
<ISPL>CMCNT ;.nusi
ores osairs r
SUPSScninc*
P=o 11 j iMa.-NCci
aoTOR oflirs.*r
SUPÂca.riCM. SPEED
11 3 iMbAtAwcsi
iOOfOin
Jnuada ,**
.'"\
1 .'" -N
i
<
v> r
' - , . . <"
"3
1 5 i x oisPt/**;MeNr unu
'
"3
< aSPl/tCEMEN r um
"
Speeds
78
soroRcsmrs r iuPtscairicî. i?e=on iruiwL.MCS;
'4.
Rorcfl caairs *r oPss Cairo,
SPEED H
IMÔMCCJ
3-2-1
Speeds
79
VIAX <v 3lS?'-iC3MEN- .s. SPSS'
5g USALANC:
7Q
20
5000
000
000
MAX DISPLACEMENT vs. ROTOR SP"EED ' 1 5 .jram :moaianc3 )
0i
1000
2000
3000 -1000 ROTOR SPEED (rpm)
5000
5000
70C0
PHASE ANGLE VS. ROTCR SPEED ( 1.3 gram imoalanca )
1000
'000
5000
5000
Figure 47. Maximum Displacements and Bode Plot 1.6 Gram Imbalance Good Bearing.
80
FREQUENCY - (2X Hz)
Figure 48. Cascade Plot 1.6 Gram Imbalance Good Bearing. 81
*OrCH GrtBlTS ::AH CHir:CAc ?60 ,1 Jt) liHT vSAO 3BG)
ôrcBCRBirsNe-rt wBircJi. j?=Hn 39 vear ib 2U

iian
VBAOBBI
15
imudU ' IKl
t "
**
i . o
<
5
*" iQ
'
>-'"
ib
IQ-
10
-15
-10
5 0 5 X OiSr'l.ACcMbNr (mSi
10
15
20
25
iS
-i0
A.
-5 3 5 0 iSr". AC M N r imuaj
10
ROTOR
oflairs
NEAR CRITICAL SPSEO
o og
VEST VSAO SRGI
ROTO
oflatrs NCAR CRITICAL
SP0
11 oq vear
VSAO SAGI
25 2S40<pm 20 tun miaow tS
> an
.-^
10
z I Q u < a .5. 5
*-"
>-.io15
/-
-20.
JO -IS -10 5 J S x DISPLACEMENT (mnl 10 IS 20 25 20 -15 -i0 5 3 5 10 15 20 25
Figure 49. Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction. Speeds from 2780 to 2860 rpm.
82
florcfl opeirs *e** cRtroi.

29O0(pri
SPS=D
n og
VEST VAAO
awG>
RCTCR ORBITS NeRCRir.C.i. 5?60:i Oq vesr'/d aBGi
i5(-
.moulu
/f
3 j
r
75 -20 -'5 -iQ '5 ^ _ ,.
3
/
10 IS 20 25
l5
J
20
-15
-10
-5 ) ) X OlSPL^CcMCNf (mast
10
'S
20
ROTOR ORBITS .MEAfl CRITIC*. SPEED (I Og vER r vBAO 8BGI
ROTOR ORBITS
-<e>R
CRITIC, SPEED
11 oq
VERT
vBAORGi
* sun nxuaM
V
'2 Vs 3 Hs "i 1 5 5 i x oispi^cSMEnr ,n is 20 25
io 5 a X OlSPl^CcMgNT (mal
83
acrcflOflirs *r sons*
OPTICAL SPE=(I
ogveHr'/a*o ana
aorcROH8irs*r3uPSfli;Birii;AL SPfrOu ug vc.ar /a*oafli
- suit rnoau
x 0iSPt_C2Me*r (mum
>
.]
-2
OISPLACSMCNT
-I
1 jnusi
sorofl oflars r
SUPSR CBHIC.AL
speso n <year-BAD awoi
floroR CRars Ar
SUPR CRITICAL SPE=D
(i *j veer
V8A
ano
?
Q.
1 X aiSPt^CEMgNf lm>
X aiSPl-ACSMeNT (IM
84
aorGBonairsarsuPsacaiTOL
SPSSDH *J
êarvaAoasGi
*OTOR
5
OflatTs *r supsa carncAL
SPSSO I
og vear va>o aflG)
OOOfwn
3
- 2
- am mud
|( I0 <
5 3
"3
-3
-2
1 0 I XOiSPl^CaMgNT Haut
<
85
MAX X'
3I3?L.*CSM=N~ .s. 3??C :q /EP~ /9AC =FGi
1000
2000
3000 4000 aCTOR SPEED (rpm)
5000
5000
7000
MAX OISPWCSMENT /s. ROTOR SPEED ( i 0 gram Vertical 1
1000
2000
3000 4000 3CTCR SPEED (rpm)
5000
5000
7000
PÂSE ANGLE VS. PCTOR SPEED ( 1.0 gram Vertical)
aCTOR SPEED (rpm)
Figure 53. Maximum Displacements and Bode Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction.
86
150
FREQUENCY - (2X Hz)
Figure 54. Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Vertical Direction.
87
florGflonairsNe-rt caincAL i^sscn ^ *cftiz
VBA SBO
RQfCH CflaiT N6.-R
C3U"OL
iP? = 0 n jQr<:fli2
'/SAU
dfli
ZQ\-
nioaia
n>uui<i
2 a
IS 20iS 20 2S 20 -IS -10 -5 ) 5 tmMS) '0 'S 30 as
20
-la
-10
SO 5 A. OlSr't.AlJSMCNr ;<TUIi|
30
QISPLACCMENT
aorofl Gflars NfiAfl
CRITICAL
SPSSD (1 Oq nCHl /B*0 SflGt
aoroa 25 260rom 20 IS nmooi
ORBITS NAA CRITICAL SPEED
o og
ORIZ VSAO BGJ
./"
,f
_ 10
oi
X"'
/
TTs "io 5 3 5 * OISH^CEMNT (ml 10 IS 20 25 20 -iS -10 5 0 5 I iS 20 25
Figure 55. Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction. Speeds from 2780 to 2860 rpm.
88
-oroflGflarsNE.**CRITICAL
SPSEO
ii jqHORiz-veAoaRGi
AOrOROf&TS
NAH
CRJTlOl. aP-=0(i OqHCHi^ /AQ dP-Gl
s
It
4r
20 -IS i0 -5 '3 5 0 IS 20 25 IS 10 3 J 5 x OiSPt-ACSMeNT ^Mi)
flOTO Cfi&TS .HfiAH CRITICAL SPSSO (t 0q wOfliZ V8*0 9PG)
ROTOR ORBITS
NAR
CRITICAL aPS=D (l Oq rRtZ V8*0 3ftG)
V
s 10 Hi î 1 5 10 15 20 25 .10 5 3 5 10
89
flOTOfl flBifS *r SUPER
CRITCAI.
i?0 (i *} tOHiZ vdAO 3RG1
rÔrOH OriirS AT
SKSR
CRITICAL SP = Q n Jg r-tCfliZ 78* dfl)
/*nm *"
DISPLACEMENT
im
-1
ôrGfl Oflars *r
SUPER GHITCAA. SPSSQ U
og HORIZ VSAO
SPO
aoroAORBirsAr
SUPERCRITICAL
Pfison qnCRizvSAoanG)
(
4
.)
n3
"
90
MAX <Y ZJISPLACc^SN- /s 3?=5D ' 3q .-CP'Z V5AC 3PG';
1000
2000
300C -1000 ROTOR SPEED (rpm)
5000
5000
7000
MAX DISPLACEMENT vs. 3QTCR SPEED ( 1.0 gram Horizontal)
"0
1000
2000
3000 4000 ROTOR SPEED (rom)
5000
5000
7000
PHASE ANGLE VS. ROTOR SPEED ( 1.0 gram horizontal)
1000
2000
3000 A000 ROTOR SPEED (rpm)
5000
5000
7000
Figure 58. Maximum Displacements and Bode Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction.
91
150
FREQUENCY - (2X Hz)
Figure 59. Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Horizontal Direction.
92
qorCH oflsus *.** cm roc spe=Oii aq-soeG icrf VBAO HI 27aO(n
ROrOflCflBirSNtrAHCfliriCAl. SPSsOil Qg o C6G-crT <*0 Pl 25 20
- 4411 15 THUOlK
' <*nu
~ 10
- s
o
'l5r
jot-
M,t- ~3TTs
Ho
1 5 I x otSPLACjiMeN r imusi
10
IS
20
25
.10
'0
aorofl
2Sr 20li
ORBITS
NgAftCflinc-y.
SPEEDn
og
JSOGG LEFT
veAoaRCi
HO'Cfl ORBITS Ne.Art CRtriCA. SPS20 (1 Og -*5 OEG LCFT VAO 3HG)
TbOOM
' na
?,0-
X LU
3 o
-I0
r"
5 3 7s H 1 3 510 X IS?(-.CMeNr (null IS 20 25
IS-
IS
20-
-20
4-
IS
10
S 0 S < 01SPl_*CsM6NT (rnul
10
IS
20
25
Figure 60. Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical). Speeds from 2780 to 2860 rpm.
93
len juairs NEfi
CRITICAL SPSSQ
n oq4S oec L=rr /ao asci iut-
aorcHCflairsNe^rtcamcu.
SPSSOU
*j4sc60LcrTvaoaf!Gp
^
TIMJUI.
' no
51-
id
-IS2025,'
f
-in % Tio
XGISM-ACSMSNT IOM1I
r
1 5 3 i i 2025
X GlSPLAOcMeNf (miisj 0 IS 20 25
*3TOB ORBITS NS/WI CRITICAL
speeo 1> og as o0
ISFT VBAO
asc
ROTOR
oseirs
NEAR CRITICAL
as=o 11 oq
JS CEG
urT
VBAO
asoi
2520HMO! ' 141
~ 10-
?"
s
3 JA
J -
Q *"I0 152025, 20 IS -10 i 0 i X OlSPLACSMtSNT tnu 10 IS 20 2S -15 20 -10
o
5 3 S A QlSPt-ACEMtN T im*si 10 IS 20
94
florcflOFiairs r iuH=Bcfnrioi.
SPÊOn
oq-isOE tcfr e*oISHI
sQOtQfli
- ii+fi nudl ' *nu
'5 4 -3 -2 -i ) i 2 1 * S
flu TOR QHB1TS Af SUPfifl CBirO.L iP6=0 11 *) 4S 0G LEFT 'J&AO *Gt
jrofl ORsirs
AT SUPER
cm neu.
SPEED I I
*j 4S oea
LSFT VUAO
wai
SOOfgm
* sun
rcwwM ..'*>
U,'-'
a -a -i a i X CiSPUCEUIsNr imttl 2 i
^
"3 3 3
x OlSPl>CEMENr tmui
"
95
AoroHQBirs ^r
SUPERCRITICAL SPEEDo
.x^soeÛfT ,a*odr*Gi
X iSPlACcMfiNT (m)
aorOP,flars *T 3UP6*CHiTtCAL SPEED(i dg -50G
LFT
vBAOaflG
AOTO OBBITS r SuPEfl CRITICS. SPEEO [l O) 45 06G iSFT va>0 flGi
Mfpm
/"
I2
\
5
\ ...v-
i. 3
-3
'5
~5
'
-1
96
MAX XY DISPLACcMS^- 'S. SPEED < 3g -15 3S-3 -E" 73AC 3RG'i
"0
1000
2QCO
3000 -tOOO ROTCfl SPEED (rgm)
5000
S000
7000
MAX DISPLACEMENT vs. ROTCfl SPEED ( 1.0 gram 45 Deg L9rt )
"0
1000
2000
3000 4000 ROTCfl SPEED irpm)
5000
3000
7000
PHASE ANGLE VS. ROTCB SPEED ( 1.0 gram 45 Oeg '.aft )
1000
2000
3000 4000 ROTOfl SPEED (rpm)
S000
5000
"000
Figure 64. Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical).
97
150
FREQUENCY - (2X Hz)
Figure 65. Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation In the Direction of the Resonant Displacement (45 Degrees off vertical).
98
.rOH OH6.rS ,-16) Cainou. iWiJ 11 *)-S 06G *T vd*Q jP, flOrOHOB8lTSNeCRIICAl. PSSOII Oq-'SOEGST vAOflGi
Mli mods
,^>
10 25 .20 -IS -10 -i DISPLACEMENT ktnusi
-5 0 . 5 <0ISPl->*;6MtNr ,musj
ID
13
20
25
ores onairs .ie.*fl Cfliriou.
SPSSD
(i jg oea ar va*o afiGi 20:
flOTCfl onairs
;360fOJn
NEAR CRITICAL SPS=O
<i Og -s oeo *r vaô awsi
2UH > *an lj>moa< ' no
.^
' no
2 u
5
* 10-
^
~ui 3 3 i io is :o 25
=25
2Q
/
HS ^Q
20
'S
5 3 3 DISPLACEMENT iiiwi
15
20
25
Figure 66. Rotor Orbits 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees off Vertical). Speeds from 2780 to 2860 rpm.
99
ôrOfl OH&rS NgAfl CRITICAL iPcO (1 Og -5 0G *r '8*0 aG(
sorcB o*airs
NS**cfliriCAL
3P=o -1 oq ^Q6GarvB*oaBGi
/
1
4IMJ
/
^5 75 Ho 5 3 ~5 To 1 I 25
Vs 3 1 731 s x 'JISPI./>I:SMEN r .-man lo
aero csaiTS w.*n CSITCAL

300Qfom
SPEED
n og 49 PEG
BT VBAO
asm
BOTCH
osairs
N&H CSITICM. SPESO
n *j
JS
oea ar
VSAO
asci
* an moo*
\
Hi J
V
2
& SOn?n 43
x DiSPljCMNr Inuai
20
as
X QlSPtACSMENr (mm
100
ôronoseif5*r iuP5flCfliroi.
SPEED
it oq.soeasTvBAOsi
*}-
AOrCflCnairSAf SUPER CrtirLCALiPScQ il Jy-S^HT /d-u iPG-
S00<i*n - u&n
57Q0fwn
ma
3 j
\ %.*
Nui*
'
'. ii X OlSPL.ACSMENr imitj
'
-3
Borcs cseirs *r
SUPSS
cfure.
SPSEO
11 Og
JSoea
flr va*o asGi
riOTCHOflairs *r SPSflCflincAL speson oq-*scea ar vaA0 3flG>
*t"
OOOfom
./"*
*. V
w*w
*^^
4-3-210 I X OiSPLX:SMcNr ;mui
>
2 <
OCSP1.>CSWCN
l r {mui
"
101
MAX XY DISPLACEMENT /s. SPEED
Og -^ -EG ^T 7SAC 3FG)
1000
2000
3000 J-000 HOTCH SP==D (rom)
5000
5000
"000
MAX OlSP'-ACSMENT vs. ?C7CR 3P==0 ( ! 0 gram ^5 Deg flignt:
7 ,uuu
JC <u^
3 ^ ROTOR SPS=D (rpm)
50
COO
7000
PHASS ANGL= VS. PCTOB SP55D ( 1.0 gram -IS Ceg Signt)
Tool ,uuu
20~
^<JU
3000
noTCR
SPS=D
-1000 (rpm)
5000
5000
7000
Figure 69. Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees off Vertical).
102
150
FREQUENCY - (2X Hz)
Figure 70. Cascade Plot 1.0 Gram Imbalance Elongated Bearing with Elongation Perpendicular to Direction of Resonant Displacement (45 Degrees off Vertical).
103
aOTCH ORBITS NEAP. CPaTICAL SP6E0 (1 Og HCPJiCNTAL SHAFT)
florcn CRsiTs 2900fom 20 - stan 15 ' id 1G 5

LU
HAR
cflincAL
3P=EO
T i 1 Oq CfllZCNMt. iHAF )
'
\ '
sun nioctti
I 0
o
<
<
'
5
5
""-10 15
*" -10 IS 20 5 "20 Hi HO 1 51015 A DISPLACEMENT (mul 20 25 20 20
I,'
-15 -10 -5 0 5 X DISPLACEMENT (mils! '0 15 20
aoron cflBiTS
2360xn
NEAH CRITIC*, SPEEO II
og HOBIZONTAL SHAFT)
25
BOTOB OflSlTS NEAP, CP.ITICAL SPEED () Og HORIZONTAL SHAFT) 2880fpm 20 aart IS 10 Jf s (mod ' and
juri 15}nudol * *nd
5 r 5
a 0 o
3
ft -5
<
a! .5
5
"10 -15 -20
1520.10 -5 0 5 X CXSAJCEMENT (MM 10
lo
ti
Ho
'&
3 3 510 X OlSPl^CfiMENT (Mi)
^5
2025
Figure 71. Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Horizontal. Speeds from 2760 to 2880 rpm.
104
ROTOR ORBITS NEAR CRITICAL SPEEO I I
og
HORIZONTAL SHAFT)
ROTOR ORBITS NEAR CRITICAL iPCSO 11 Qg HORIZONTAL SHAFT)
Z90arom sun
Tiiaal
3 o 5
.JO
15
-10
-5 0 i X DISPLACEMENT JIM
10
15
20
25
ROTOR ORBITS NEAP. CRITICAL SPEED (1 Og HORIZONTAL SHAFT)
ROTOR ORBITS NEAR CHITICAL SPEEO (1 Oj HORIZONTAL SHAFT)
-ft
mKHM * id
=
B 5
10 15-
2010 5 10 2! 20 15 -10 -5 1 5 X DISPLACEMENT (iMsl 10 15 20 25
x asPucsuEMT (IM
105
SOTCR CHBITS AT 3UPEP. CP.ITCAI. SPEED (t Oq HOfllZONTAL jHtfTl
POTOH ORfllTS Af SUPEfl CaiTCAL 3PEEO (1 0<J MCRIJCNfAl. HAF"
%
o
< S -I ;/*
5
*"!
a-
"3
X DISPLACEMENT (ml
"
'5
-i
.J
-I 0 1 X OlSPt-ACEMENT imu.l
>
aorcfl oneTS
T SUPSP. CBITCAL SPEED
(t og MOAIZCNTAI.
SHAFTI
HOTOP.
onarrs
AT SIJPEP. CSITCAL SPEED
(i og
HOBCCWTAI. SHAPTI
3*
nuaala ' id
2 ? E
'
ft-i 23-
v...
"5 3~5
X DISPLACEMENT (mis)
a o
Q-
5 ""2
3-
23
.J
.2
I 0 I X DISPLACEMENT (mi
106
VtAX X DISPLACEMENT JS. 3F5ED ! Og "CRIZCNTAL SHAr"
1000
2000
30C0 4000 HCTOR SPEED (rpm)
5000
5000
7000
MAX Y DISPLACEMENT vs. SPEED (I.Og HORIZONTAL SHAFT)
1000
2000
5000
5000
7000
MAX DISPLACEMENT vs. ROTOR SPEED ( 1 .Og horizontal mount)

30
f 50
z4r2 30(o <20-_J
a.
2 10
"0
1000
2000
5000
5000
7000
PHASE ANGLE VS. ROTOR SPEED ( 1.0g horizontal mount)
1000
2000
5000
5000
7000
Figure 74. Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Shaft Mounted Horizontal.
107
20>
FREQUENCY - (2X Hz)
Figure 75. Cascade Plot 1.0 Gram Imbalance Shaft Mounted Horizontal.
108
ROTOfl CBBirS 25
.'400ftXTl
NEAP.
CRITICAL SPEEO 11 Oq VERTICAL 5HW=T)
flOTCHOfl&rSNeAfl CRITICS. SPOn Jg VERTICAL
SHAF*-)
20
sun
rmod*
' iOd
IS '
miofe
~ 10 5 5
r
J
-<
Q " ''0 15 20
X
\
x.
'
20 IS -10 o 0 S X DISPLACEMENT (mtfsj 10 15 20 25 25 20 -15 -10 -5 0 5 X DISPLACEMENT im| 10 IS 20
ROTOR CRBJTS .NEAR CRITICAL SPEED (1 Oq VERTICAL SHAFT) 25 2770fpm 20
flotcfl Oflarrs
NEAA CRITICAL SPEED M
og VERTICAL
SHAFT)
* sun
IS moatt ' and
J ,0 5 3 o
<
^S^*!*-..
a. .g *"-io
-15 -20
20
-IS
-10
-5 0 S X DISPLACEMENT (mui
10
IS
20
23
*h
10
-5 0 5 X DISPLACEMENT {mut
10
Figure 76. Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Vertical. Speeds from 2400 to 2800 rpm.
109
flOTOH opars NEAR
CRITICAL SPEEO
11 oq VERTICAL
SHAFT)
flOrOHORSlTS NEAH CfllTtCAL3P=0l' Oq 25 20
V6P,TOL SMAFT)
sun
15 10 mrcwa ' and
I
a o
<
j
-5 J
^ t/i
'"IO
-IS -20-2: 20
-10
-5 "0 S X DISPLACEMENT [mi)
10
-IS
-10
-5 0 5 X (DISPLACEMENT !mus(
10
15
30
25
flOTOR oflerrs NEAR 3000fpm
CRITICAL SPED
OQ VERTICAL SHAFT)
ROTOR ORBITS NEAfl CRITICAL SPESO (1 0q VERTICAL SHAFT) JlGOfom san
mod ' nd
i(-
ITUdOl
' nd
^N
******
5..
2%S "
^n iQ
^S 5
HO
1 5 S X DISPLACEMENT (m>
lO
13
20
25
15
-10
-5 0 S X OlSPUCEMENT imui
10
Figure 77. Rotor Orbits 1.0 Gram Imbalance Shaft Mounted Vertical. Speeds from 2900 to 3100 rpm.
110
BOTCH 0B8ITS Ar SUPER CP.ITICAL SPScO {1 O) VERTICAL SHAFT)
ore
OBSITS AT SUPER CRITICAL SPEED H
og vEarcAL
SHAFTI
,'"'
-2
I .1 12 X DISPLACEMENT (mmi
ROTOR ORBITS AT SUPER CRITICAL SPEED (1 Oq VERTICAL SHAFT)
BOTOB 0B8ITS AT SUPER CRITICAL SPEED (! Og VERTICAL SHAFT)
Figure 78. Rotor Orbits 1.0 Gram Imbalance Shaft Mounted VertkaJ. Speeds from 5400 to 6000 rpm.
Ill
MAX X DISPLACEMENT >s. SPEED '1 Cg 75":CAL 3HAh-
Tonn
20
3000 4000 HOTOR SPEED (rcm)
5000
60C0
7000
MAX Y DISPLACEMENT /s. SPEED (VOg VERTICAL SHAF")

SO 50
130
Q
40
2oi-
1000
2000
30 40 HOTOR SPEED (rpm)
57000
MAX DISPLACEMENT vs. ROTCR SPEED ( 1 .Og vertical mount) 50 = 5040 uj | 30 o <20
1000
2000
5000
5000
7000
PHASE ANGLE VS. ROTOR SPEED ( 1.0g vertical mount) 50 0 -50
s <
i
100 150
I-200
I-250
1000 2000 3000 4000 ROTOR SPEED (rpm) 5000
5000
7000
Figure 79. Maximum Displacements and Bode Plot for 1.0 Gram Imbalance Shaft Mounted Vertical.
112
2Q>
15IM
I10a tn a
150
FREQUENCY - (2X Hz)
Figure 80. Cascade Plot 1.0 Gram Imbalance Shaft Mounted Vertical.
113
114
LIST OF REFERENCES
[Ref. 1]
Simei, F.A., "Development of an Experimental Facility for Analysis of Rotordynamic Phenomena," Thesis, March 1994. Vance, J.M., Rotordynamics of Turbomachinery, pp. 130131, John Wiley & Sons, 1988. Rankine, W.A., "On the Centrifugal Force of Rotating Shafts," London Engineer, v.27, p.249, 1869. Jeffcott, H. H., "The Lateral Vibration of Loaded Shafts in the Neighborhood of a Whirling Speed - The Effect of Want of Balance," London Philosophical Magazine, v. 37, pp. 304-314, 1919. Muszynska, A., "Fundamental Response of a Rotor," report No.l, 1986. James, M. L., and others, Vibration of Mechanical and Structural Systems, pp. 89-94, Harper & Row, Publishers, Inc., 1989.
[Ref. 2] [Ref. 3] [Ref. 4]
[Ref. 5] [Ref. 6]
115
116
INITIAL DISTRIBUTION LIST
No. Copies 1. Defense Technical Information Center Cameron Station Alexandria, Virginia 22304-6145 2. Library, Code 052 Naval Postgraduate School Monterey, California 93943-5101 3. Department Chairman, Code ME Department of Mechanical Engineering Naval Postgraduate School Monterey, California 93943-5000 4. Professor Knox T. Millsaps Jr., Code ME/MI Department of Mechanical Engineering Naval Postgraduate School Monterey, California 93943-5000 5. LT Curtis E. Vejvoda 549 Bonita Dr. Tulare, California 93274 6. Curricular Officer, Code 34 Department of Mechanical Engineering Naval Postgraduate School Monterey, California 93943-5100 Naval Sea Systems Command Director, Engines Division (SEA S6X3) Washington D.C. 20362 Bentley Nevada Corporation Mr. Don Bentley P.O. Box 157 Minden, Nevada 89423 2
7.
8.
117
Dr. John Vance Department of Mechanical Engineering Texas A & M University College Station, Texas 77843-3123
118

Whirling of Shaft

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Whirling of Shaft

Uploaded by

Copyright:

Available Formats

NAVAL POSTGRADUATE SCHOOL MONTEREY, CALIFORNIA

&m iLi JU W &

Approved for public release; distribution is unlimited.

REPORT DOCUMENTATION PAGE

Form Approved OMB No. 0704-0188

REPORT TYPE AND DATES COVERED Master's Thesis FUNDING NUMBERS

PERFORMING ORGANIZATION REPORT NUMBER

10. SPONSORING/MONITORING AGENCY REPORT NUMBER

17. SECURITY CLASSIFICATION OF REPORT Unclassified NSN 7540-01-280-5500

Approved for public release; distribution is unlimited

Submitted in partial fulfillment of the requirements for the degree of

Matthew D Kelleher, Chairman, Department of Mechanical Engineering

Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6.

DEFINITION (units. U.S. Customary)

E I L M Mdisk Mshaft PSD {q} {q} {q} e

direction stiffness of bearing plus shaft caused by x

direction damping of bearing plus shaft caused by x

direction velocity, (dimensionless)

y direction damping of bearing plus shaft caused by y direction velocity, (dimensionless)

x direction damping of bearing plus shaft caused by y direction velocity, (dimensionless)

y direction damping of bearing plus shaft caused by x direction velocity, (dimensionless)

Figure 1. Rotor and Disk Assembly in Forward Whirl

Figure 2. Rotor and Disk Assembly in Backward Whirl

III. ANALYTIC MODEL

MASS IMBALANCE (M)

[M]|q} + [C]{q} + [KKq} = {F(t)} with,

"M 0 [M] = . 0 Mj [C] = c [K] =

KXx(bearing); Kyy = Kyy(shaft)

KXy and KyX = coupling or cross stiffness ; Cxx = Cxx(shaft)

CXx(bearing); Cyy = Cyy(shaft)

Cxy and Cyx = coupling or CTOSS damping

where q is a radial displacement: V(x2+y2) Taking derivatives with respect to time:

Analytic Pinned-Pinned Model of Shaft and Disk Assembly.

From [Ref. 6] the stiffness of the shaft is: 48EI

A. Kxx = 495, KYY = 495, Cxx = -025, CyY = -025

C. Kxx = 535, KYY = 535, CxX = 035, CYY = 02

G. KXX=535, KYY=535, KXY=10, Kyx=40, CxX = 025, CyY = .025

H. KXX=535, KYY=535, CXX=025, CYY=025, CXY=-02, CYX=-01

tabulated at the end of this Chapter. investigations are contained in Appendix C.

KEY PHASOR PROBE

Y-PROXIMETER MASS IMBALANCE SHAFT X-PROXIMETER

BEARING SUPPORT DISK SECTION A-A

Figure 5. Configuration of Experimental Rotor System.

1 X RESPONSE BW/SR NO BW/SR NO BW/SR BW/SR BW/SR BW/SR

1/2 X RESPONSE YES NO NO YES VERY MUCH VERY MUCH

Table 1. Results From Experimental Investigation

CONCLUSIONS AND RECOMMENDATIONS

LIST OF DIMENSIONS AND PROPERTIES

APPENDIX B. ANALYTIC ROTOR DATA

=2500rpm - =2740rpm x =2800rpm

-10 0 10 X displacement - mils

ROTOR SPEED - (200X rpm)

FREQUENCY - (4X Hz)

3000 4000 ROTOR SPEED (rpm)

-10 0 10 X displacement- mils

3000 4000 spaed -rpci

ROTOR SPEED - (200X rpm)

FREQUENCY - (4X Hz)

Kxx=535.Kyy=555.:<xy=O.Kyx=0 */in Cxx= C2S.Cyy= 025.Cxy=O.Cyx=0

3000 4000 ROTOR SPEED (rpm)

Kxx=535.Kyy=555.Kxy=0.Kyx=0 #in Cxx= 025.Cyy= 025Oxy=0.Cyx=0 50 0 -50 100 2< 150

<; 200 a-250

ROTOR SPEED (rpm)

SC^CP CP8IT3 <Sfl C.^lT'OL J?==-3 0 2q .MB/*L-NC3