Tsuneo Someya (Editor)

Journal-Bearing
Databook
With Contributions by
T. Someya, J. Mitsui, J. Esaki, S. Saito,
Y Kanemitsu, T. Iwatsubo, M. Tanaka,
S. Hisa, T. Fujikawa, H. Kanki

With 460 Figures and 115 Tables

Springer-Verlag Berlin Heidelberg GmbH

Professor Dr. Tsuneo Someya
Department of Mechanical Engineering,
Faculty of Engineering, The University of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113jJapan

Title of the original Japanese edition:

Collected Data on the Static and Dynamic Characteristics of Plain Bearings
© by The Japan Society o f Mechanical Engineers

ISBN 978-3-642-525 11-7

Library of Congress Cataloging-in-Publication Da ta

Journal-bearing databook I Tsuneo Someya, editor ;
with contributions by T. Someya ... (et al.]. p. em.
ISBN 978-3-642-52511-7 ISBN 978-3-642-52509-4 (eBook)
DOI 10.1007/978-3-642-52509-4
l. Journal bearings--Handbooks. manuals, etc. I. Someya. Tsuneo. 1931 -
T J I 063.168 1988 621.8'22--dc 19 88-30820

This work is subject to copyright. All rights are reserved. whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting. reuse of illustrations, recitation. broadcast-
ing, reproduction on microfilms or in other ways. and storage in data banks. Duplication o f this
publication or parts thereof is only permitted under the provisions o f the German Copyright Law of
September 9. 1965, in its version of June 24, 1985, and a copyright fee must always be paid. Violations
fall under the prosecution act of the German Copyright Law.
© Springer-Verlag Berlin Heidelberg 1989
Originally published by Springer-Verlag Berlin Heidelberg New York in 1989
The use of registered names. trademarks. etc. in this publication docs not imply. even in the absence
of a specific statement. that such names are exempt from the relevant protective laws and regula tions
and therefore free for general use.
Typesetting : Daten- und Lichtsatz-Servicc, Wurzburg

2161 /3020-543210 - Printed on acid-free paper

Members of the JSME Research Subcommittee
on Dynamic Characteristics of Journal Bearings
and Their Applications (PSC 28)

Chairman Tsuneo SOMEYA University of Tokyo

Secretary Shinobu SAITO Ishikawajima-Harima Heavy
Industries Co., Ltd.

Members Hiroshi AOKI Nihon University

(in alphabetic order) Seiichi AOKI Komatsu Ltd.
Kyoji CHIKIRA Hitachi Zosen Corporation
Jinro ESAKI Mitsubishi Heavy
Industries, Ltd.
Takes hi FUJIKAWA Kobe Steel, Ltd.
Hirotsugu HAYASHI Waseda University
Shoichi HISA Toshiba Corporation
Takuzo IWATSUBO Kobe University
Yoichi KANEMITSU Ebara Corporation
Hiroshi KANKI Mitsubishi Heavy
Industries, Ltd.
Hiroshi KINOSE Nippon Kokan K. K.
Harutaka KOIKE Kawasaki Heavy
Industries, Ltd.
Takahiro MASUDA Mitsubishi Electric
Corporation
Jun'ichi MITSUI Tao Nenryo Kogyo K. K.
Takafumi NAKAI Aoyamagakuin University
Takuya SUGANAMI Mitsubishi Electric
Corporation
Kazuhiko SUGITA Toyoda Machine
Works, Ltd.
Katsuhiko TANAKA Nippon Seiko K. K.
Katsuyuki TANAKA Hitachi, Ltd.
Masato TANAKA University of Tokyo
Nobuo TSUMAKI Hitachi, Ltd.
Ryuji WADA Toyoda Machine Works, Ltd.
Hiromasa YAMAGUCHI Komatsu, Ltd.
Shigeo YANABE Technological University
ofNagaoka

v
Companies Supplying Data (in alphabetic order)

Calculated Data (Chapter 2)

Ebara Corporation
Hitachi, Ltd.
Hitachi Zosen Corporation
Ishikawajima-Harima Heavy Industries Co., Ltd.
Kawasaki Heavy Industries, Ltd.
Kobe Steel, Ltd.
Mitsubishi Electric Corporation
Mitsubishi Heavy Industries, Ltd.
Mitsui Engineering & Shipbuilding Co., Ltd.
Nippon Kokan K. K.
Nippon Seiko K. K.
Toshiba Corporation
Toyoda Machine Works, Ltd.

Experimental Data (Chapter 3)

Ebara Corporation
Hitachi, Ltd.
Hitachi Zosen Corporation
lshikawajima-Harima Heavy Industries Co., Ltd.
Mitsubishi Electric Corporation
Mitsubishi Heavy Industries, Ltd.
Mitsui Engineering & Shipbuilding Co., Ltd.
Toshiba Corporation

VI
Preface

Journal bearings, which are used in all kinds of rotating machinery, do not only
support static loads, such as the weight of rotors and load caused by transmitted
torque of reduction gears, but are, in addition almost the only machine element that
is able to suppress various exciting forces acting on the rotating shaft.
As rotating machines have become large and multi-staged, while compactness,
high speed, and high output have also been realized in recent years, not only has
the bearing load increased, but also the magnitude and variety of exciting forces.
Therefore, the role and importance of journal bearings have increased tremendous-
ly.
In particular, for the design of rotating machines with low vibration levels and
high reliability, knowledge of the exact characteristic data of bearings, and especial-
ly of the stiffness or spring coefficients and the damping coefficients of oil films in
bearings, is essential.
However, the amount of reliable data now applicable to practical design is
limited. Through the activity of the Research Subcommittee on Dynamic Charac-
teristics of Journal Bearings and Their Applications (designated as PSC 28), estab-
lished and organized in June 1979 through May 1982 within the Japan Society of
Mechanical Engineers (JSME), these coefficients, together with static characteris-
tics, have been calculated and also measured on a number of new test rigs. These
data are the principal contents of the present databook, in which almost all types
of bearing actually in use are included. The book can be said to be the fullest
collection of information currently available. In addition it contains explanations
on the characteristics of oil films, and also examples of practical applications, so
that it will be useful to either newcomers to the field or experts, as well as to research
workers and designers.
As mentioned above, the characteristic data in this databook are based on the
hard work of the JSME Research Subcommittee on Dynamic Characteristics of
Journal Bearings and Their Applications. The names and affiliations of the mem-
bers are listed before, and their help is gratefully acknowledged. Thanks are also
due to the companies listed for allowing us to publish the data.
This volume is an English version, with some revisions, of the Japanese original,
published under a licence agreement between JSME and Springer-Verlag. The
translation was supplied by the authors of this databook, and their selfless dedica-
tion is much appreciated. Special thanks are due to Prof. M. Tanaka and Dr. S.
Saito, who have devotedly acted as secretaries throughout the progress of the
project since 1979. Timely help from Mr. M. Takahashi, the manager of JSME, and
Mr. A. von Hagen, acquisitions editor of Springer-Verlag, in publishing the book
is acknowledged with thanks.

Tokyo, Autumn 1988 Tsuneo Someya

VII
Contents

Nomenclature . . . . . . . . XIII

1. Introduction (By T. Someya) 1

1.1 General Information . . . . . 1
1.1.1 Types of Journal Bearings 1
1.1.2 Main Dimensions of Journal Bearings . 1
1.1.3 Coordinates for Journal Bearings . . . 3
1.1.4 Assumptions in the Databook . . . . 3
1.2 Static Characteristics of Journal Bearings . . 3
1.2.1 Equilibrium Position of Journal Centre 4
1.2.2 Frictional Coefficient of Journal Bearings 4
1.2.3 Oil Flow of Journal Bearings . . . . . 4
1.2.4 Heat Balance in Journal Bearings 6
1.2.5 Procedure for Calculation of Heat Balance in a Journal Bearing 7
1.3 Dynamic Characteristics of Journal Bearings . . . . . . . . 9
1.3.1 Definition of Coefficients of Oil Film . . . . . . . . . 9
1.3.2 Determination of Coefficients of Oil Film by Calculation 10
1.3.3 Determination of Coefficients of Oil Film by Experiment 10
1.3.4 Comparison Between Calculated and Experimental Values for
Coefficients of Oil Film . . . . . . . . . . . 10
1.4 How to Use the Characteristic Data of Bearings in the
Present Databook . . . . 11
1.4.1 Nature of the Data . . . . . . . . . . . . . 11
1.4.2 An Example of Data . . . . . . . . . . . . 11
1.4.3 Influences of L : D Ratio, Turbulence, Load Direction, etc. 11
1.4.4 Some Notes on Application of the Data 12
References . . . . . . . . . . 12

2. Calculated Data 13

Calculation Type of bearing L:D mP Remarks p.

1. . 2 Axial grooved cylindrical 0 0.5. 0. 14

2. . 2 Axial grooved cylindrical 0 1 0. 17
3. . 4 Axial grooved cylindrical 0 0.5. 0. 20
4. . 4 Axial grooved cylindrical 0 1 0. 23
5. . Pressure dam cylindrical 0 1 0. 26
6. . Partial arc 0 0.5. 0 120 ° 0 29
7. . Partial arc 0 0.5. 0 150 ° 0 32
8. . Partial arc 0 1 0 120 ° 0 35
9. . Partial arc 0 1 0 150 ° 0 38
10. . 2 Lobe. 0 0.5. 1/2 0 41
11. . 2 Lobe. 0 0.5. 2/3 0
44
12. . 2 Lobe. 0 0.5 . 3/4 0
47
13. . 2 Lobe. 0 1 1/2 0 50
14. . 2 Lobe. 0 1 2/3 0
53
15. . 2 Lobe. 0 1 3/4 0 56
16. . 3 Lobe . 0 0.5. 1/2 0
59

IX
Contents

Calculation Type of bearing L :D mP Remarks p.

17. . 3 Lobe. . 0.5. 2/3 . 62
18. . 3 Lobe. . 0.5. 3/4 . 65
19. . 3 Lobe. . 0.5. 4/5 . 68
20. . 4 Lobe . . 0.5. 1/2 . 71
21. . 4 Lobe. . 0.5. 2/3 . 74
22. . 4 Lobe. . 0.5 . 3/4 . 77
23. . 4 Lobe . . 0.5. 4/5 . 80
24. . 4 Pad tilting pad. . 0.5 . 0. . LBP . 83
25. . 4 Pad tilting pad. . 0.5 . 1/2 . LBP . 86
26. . 4 Pad tilting pad. . 0.5. 2/3 . LBP . 89
27. . 4 Pad tilting pad. . 0.5. 3/4 . LBP . 92
28. . 4 Pad tilting pad. . 1 0 . . LBP. 95
29. . 4 Pad tilting pad. .1 1/2 . LBP . 98
30. . 4 Pad tilting pad. . 1 2/3 . LBP . 101
31. . 4 Pad tilting pad. . 1 3 /4 .LBP . 104
32. . 4 Pad tilting pad. . 0.5 . 0 . . LOP . 107
33. . 4 Pad tilting pad. . 0.5 . 1/2 . LOP. 110
34. . 4 Pad tilting pad . . 0.5. 2/3 . LOP. 113
35. . 4 Pad tilting pad. . 0.5. 3/4 . LOP. 116
36. . 4 Pad tilting pad. . 1 0 . . LOP . 119
37. . 4 Pad tilting pad. . 1 1 /2 LOP
. . 122
38. . 4 Pad tilting pad. . 1 2/3 . LOP. 125
39. . 4 Pad tilting pad. .1 3/4 . LOP . 128
40. . 5 Pad tilting pad. . 0.5. 0 . . LBP. 131
41. . 5 Pad tilting pad. . 0.5. 1/2 . LBP . 134
42. . 5Pad tilting pad. . 0.5 . 2/3 . LBP . 137
43 . . 5Pad tilting pad. . 0.5. 3/4 . LBP . 140
44. . 5 Pad tilting pad. .1 0. . LBP . 143
45. . 5 Pad tilting pad. . 1 1/2 . LBP . 146
46. . 5 Pad tilting pad. .1 2/3 . LBP . 149
47. . 5 Pad tilting pad . . 1 3/4 . LBP . 152
48 . . 5Pad tilting pad. . 0.5. 0. .LOP . 155
49. . 5Pad tilting pad. . 0.5. 1/2 . LOP . 158
50. . 5 Pad tilting pad. . 0.5. 2/3 . LOP . 161
51. . 5 Pad tilting pad. . 0.5. 3/4 . LOP . 164
52. . 5 Pad tilting pad. . 1 0 . . LOP . 167
53 . . 5Pad tilting pad. .1 1/2 . LOP . 170
54. . 5Pad tilting pad. .1 2/3 . LOP . 173
55. . 5 Pad tilting pad. .1 3/4 . LOP . 176
56. . Hydrostatic 179
57. . Spherical spiral grooved 182
58. . Taper spiral grooved 185
59. . Flat spiral grooved 188

3. Experimental Data . . . . . . . . . . . . . . . . . 191

. . . .
Test Type of bearing Diameter Length Radial Remarks p.
bearing D(mm) L(mm) clearance
no . CP, Cb(mm)
1 2 axial grooved 140 140 0.139 With two oil
circular bearing grooves at hori-
zontal mating
surfaces 192

X
 Contents

Test Type of bearing Diameter Length Radial Remarks p.

bearing D(mm) L(mm) clearance
no. CP, Cb(mm)

2 2 axial grooved 100 100 0.105 With two oil

circular bearing grooves at hori-
zontal mating
surfaces 196
3 2 axial grooved 100 50 0.140 With two oil
circular bearing grooves at hori-
zontal mating
surfaces 200
4 2 axial grooved 100 50 0.130 With two oil
circular bearing grooves at hori-
zontal mating
surfaces 203
5 2 axial grooved 50 25 0.033 With two oil
circular bearing grooves at hori-
zontal mating
surfaces 210
6 2 lobe bearing 508 304.8 cp = o.660 mP = 0.5 213
cb = o.33o
7 2 lobe bearing 100 50 cp = o.3 mP = 0.67 216
cb = 0.1
8 2 lobe bearing 100 40 cp = o.28o mP = 0.75 219
cb = o.o7o
9 4 pad tilting pad 140 95 0.175 p = 73.7°,
bearing mP= 0,
rx/p = 0.5, LBP 222
10 5 pad tilting pad 100 50 cp = o.216 p = 60°,
bearing cb = o.1o5 mP = 0.51 ,
rx/p = 0.5, LOP 224
11 5 pad tilting pad 76.4 31.6 0.086 p = 58°, mP = 0,
bearing rx/p = 0.5, LOP 227

4. Explanations . . . . . . . . . . . . . . . .
. . . . . . 231
4.1 Method of Calculation for Bearing Characteristics (By J. Mitsui) . 231
4.1.1 Generalized Reynolds Equation for Laminar Flow 231
4.1.2 Reynolds Equation for Turbulent Flow . . . 232
4.1.3 Method of Calculation for Reynolds Equation 233
4.1.4 Boundary Condition for Oil Film Pressure 234
4.1.5 Calculation of Static Bearing Characteristics . 236
4.1.6 Method of Calculation for Dimensionless Stiffness and Damping
Coefficients of Oil Film . . . . . . . . . . . . . . . . 237 . .
4.2 Methods of Measurement for Journal Bearing Characteristics
(By J. Esaki) . . . . . . . . . . . . . . . . . . . . . . . . .
240
4.2.1 Outline of Performance Characteristic Tests of Journal Bearing 240
4.2.2 Methods of Measurem'e nt for Oil Film Coefficients . . 242
4.2.3 Example of a Test Rig for Dynamic Characteristics of
Journal Bearings . . . . . . . . . . . . . . . . . 243
4.2.4 Experimental Results Obtained with Circular Bearings with
Two Axial Grooves . . . . . . . . . . . . . . 245
4.3 Estimation of Effective Oil Film Temperature (By S. Saito) 248
4.3.1 Estimation from Calculated Data 248
4.3.2 Estimation from Measured Data . . . . . . . . . 249

XI
Contents

4.4 Boundary Condition for Oil Film Pressure Distribution

(By Y. Kanemitsu) . . . . . . . . . . . . . . . . 249
4.5 Influences of Length: Diameter Ratio on Journal Bearing
Characteristics (By J. Esaki) . . . . . . . . . . . . . 251
4.6 Influences of Turbulent Flow on Journal Bearing Characteristics
(By J. Esaki) . . . . . . . . . . . . . . . . . . . . 266
. .
4. 7 Analysis of Bearing Characteristics with Due Consideration for
Viscosity Variation (By J. Mitsui) . . . . . . . . . . . . . 278
.
4.7.1 Thermohydrodynamic Lubrication Theory . . . . . . . 278
4. 7.2 Examples of Temperature Distribution in the Bearing Bush 279
4.7.3 Bearing Characteristics Calculated by THD Theory . . . 279
4.8 Unbalance Response of Rotors in Journal Bearing (By T. Iwatsubo) 282
4.8.1 Introduction . . . . 282
4.8.2 Equation of Motion . . . . 283
4.8.3 Unbalance Response 283
4.8.4 Unbalance Vibration Mode 283
4.8.5 Numerical Calculation of Unbalance Response 284
4.8.6 Damping Effect of Bearing . . . . . 285
4.8.7 Application to a Complicated System . . . . 287
4.8.8 Summary . . . . . . . . . . . . . . . . 287
4.9 Stability of Rotors Supported by Plain Bearings (By M. Tanaka) 288
4.9.1 Introduction . . . . . 288
4.9.2 Oil Whip Phenomenon . . . . . . . 288
4.9.3 Theory of Oil Whip . . . . . . . . 289
4.9.4 Occurrence and Growth of Oil Whip 290
4.9.5 Stability Criterion for Generalized Rotor-Bearing-System . 291
4.9.6 Stabilization Measures . . . . . . . . . . . 292
4.9.7 Flow-Induced Vibration . . . . . . . . . . 294
4.10 Application to Steam Turbine Generator (By S. Hisa) 295
4.10.1 Introduction . . . . . . . . . . . . 295
4.10.2 Bearing Types and Operating Conditions 295
4.10.3 Unbalance Response . 297
4.10.4 Self-Excited Vibration . . . . . . . . 300
4.10.5 Concluding Remarks . . . . . . . . . 304
4.11 Application to Turbo-Compressors (By T. Fujikawa) 304
4.11.1 Introduction . . . . . . . . . . . . . . 304
4.11.2 Dynamic Design of Rotor Systems According to
API Standard . . . . . . . . . . . . 305
4.11.3 Method of Vibration Analysis . . . . . . . . 306
4.11.4 Application to a Centrifugal Compressor . . . 309
4.11.5 Trouble Shooting in the Event of Centrifugal Compressor
Vibration . . . . . . . . . 311
4.11.6 Concluding Remarks . . . . 312
4.12 Application to Pumps (By H. Kanki) 312
4.12.1 Introduction . . . . . . . 312
4.12.2 Bearings Installed in Pumps . 313
4.12.3 Differences Between Pumps and Other Rotating Machines . 314
4.12.4 Method of Design Study for Pump Bearings . 314
4.12.5 Guidelines for Evalution 320
4.12.6 Concluding Remarks . . . . . . . . . . . 322

XII
N otnenclature

C = R - r Radial clearance
Cb Assembled clearance
CP Machined clearance
C;j = Cpwc;)W Dimensionless damping coefficient
D Bearing diameter
F Oil film force
H Heat generated in the film per second
K;j = C P kij/ W Dimensionless stiffness or spring coefficient
L Bearing width
N Rotational speed (rps)
Qe = qe/ (RwCPL) Dimensionless circumferential oil flow at the trailing edge of
pad or at the boundary of film rupture
Q1 = q1/ (RwCPL) Dimensionless forced oil flow
Q, = q,/ (RwCPL) Dimensionless side flow of oil
R = D/2 Bearing radius
Re = CPU fv Reynolds number
S = 11NLD (R /Cp) 2 / W Sommerfeld number
T Temperature (0 C)
U Journal surface velocity
W Bearing load
cij Damping cofficient of oil film (i,j = x, y)
d Journal diameter
e Journal eccentricity
f Oil film force
h Frictional coefficient on journal
h = C + e cos () Oil film thickness
hmin Minimum oil film thickness
kij Stiffness or spring coefficient of oil film (i,j = x, y)
mP = 1 - Cb/CP Preload factor
p Oil film pressure
p1 Feed pressure of oil
Pm = W / (LD) Mean bearing pressure
r = d/ 2 Journal radius
t Time
x, y, z Vertical, horizontal and-axial coordinates
oc Angle of pivot position from leading edge
jJ Pad angle
e = e/Cb Journal eccentricity ratio
() Circumferential angle
Jl Viscosity of lubricant
v Kinematic viscosity of lubricant
Q Density of lubricant
<jJ Attitude angle
If/= CP/R Clearance ratio
w = 2 nN Angular velocity of journal
Bold letters denote vectors or matrices

XIII
1. Introduction

This databook is concerned with the static and dy- Chapter 1 deals with general and fundamental
namic characteristics of oil films in the journal bear- items relating to the databook as a whole and how
ings most frequently used in high-speed rotating to use it; Chaps. 2 and 3 contain the characteristic
machinery. These characteristics, which exert a fun- data of journal bearings; and Chap. 4 explains the
damental influence upon the vibration and stability procedures applied to obtain the characteristics, the
of rotors, have been obtained both theoretically influences of various parameters on them, and their
and experimentally and are presented in tables and applications to vibration analysis of rotors.
graphs. The book is intended to contribute to the Numerical values are given wherever possible in
proper and vibration-free design of rotating ma- the form of dimensionless numbers, and otherwise
chinery. with SI units in most cases.

1.1 General Information

This section is devoted to a discussion of the types LOP) or between two pads (load between pads, or
of journal bearings that govern their characteristics, LBP). Well-known examples of group (iii) are float-
the coordinates system used to represent the char- ing bush bearings, consisting of an additional bush
acteristics, and the assumptions basic to the whole in the bearing clearance, and foil bearings, with
of the databook. sliding surfaces consisting of flexible lamellae. Bear-
ing group (iii) is not dealt with in the present data-
book. Hydrostatic bearings (2) and hybrid bearings
1.1.1 Types of Journal Bearings (3) are dealt with only insofar as an example of the
former type with four pockets and axial grooves
Bearings that support the rotating shaft by a sliding between the pockets is given.
motion can be classified, according to the direction
of load, as radial type and axial type. With few
exceptions the databook is limited to those of the
former type, which are also called journal bearings. 1.1.2 Main Dimensions of
As shown in Fig. 1.1, journal bearings can be Journal Bearings
divided into (1) hydrodynamic bearings in which
the external load is supported by hydrodynamic
The sliding surface of a journal bearing usually con-
pressure generated through the relative motion of
sists of a c ylinder or a set of cylindrical surfaces. In
sliding surfaces; (2) hydrostatic bearings, in which
the latter case the bearing shells are first put togeth-
the load is supported by the hydrostatic pressure of
er with shims, bored, and then assembled after re-
oil fed in, usually, by an external pump; and (3)
moval of the shims, as shown in Fig. 1.2. Therefore,
hybrid bearings, which utilize both hydrodynamic the following main dimensions are generally used.
and hydrostatic pressure. The databook is con-
cerned mostly with the hydrodynamic type (1), D = 2R: inner bearing diameter (m),
which can be grouped into (i) bearings with fixed R: inner bearing radius or radius
sliding surfaces; (ii) tilting pad bearings or bearings of inner cylindrical surface (m),
with sliding surfaces (called pads) that are pivoted r: radius of journal (m),
and free to tilt; and (iii) others. Many of these, such CP = R- r: machined clearance, usually
as a) to g) in Fig. 1.1, belong to group (i). Group (ii) taken as the representative
can be subdivided according to the number of pads bearing clearance (m), (1)
and the relative direction of external load to posi- Cb: assembled clearance. In case of
tion of pads, which depends upon whether the load 360° cylindrical bearing, cp
goes through one of the pivots (load on a pad, or and Cb coincide and shall be
 1 Introduction

a) 360° cylindrical bearing with or

without axial grooves
00
b) Partial bearing
0 0
c) Two-lobe bearing
without or with offset
00
0
Bearing
( i )with fixed d) Three-lobe or four-lobe
sliding surfaces bearing (Q)
e) Bearing with pressure dam
0
f) Spiral groove bearing

g) Others

(I) Hydrodynamic
bearing
With 3 or 4 pads
( i i ) Tilting pad - {
beanng Foil bearing
LOP or LBP

Floating bu sh bearing

Journal iii)Others .. J
~ Pourous metal bearing
0
bearing

(2) Hydrostatic
bea nng
--1r - - - - - - - - - With 4 or 5 pockets
00
0
,___ _ _ _ _ _ _ _ _ With axial grooves
between pockets

(3) Hybrid bearing

Fig. 1.1. Types of journal bearings

the characteristics of the bearing, can then be de-

fined:
L: D: length-to-diameter ratio,

If/ =
cR: clearance ratio, (2)

c
m = 1 - C b : preload factor. (3)
p
a Machining b Assembly
Geometrical parameters have to be defined fur-
Fig. 1.2a, b. Main dimensions of journal bearings. L , length ther as needed, e.g., number of segments (pads),
angular breadth of segments, offset of segments,
load direction relative to segment, positition, shape
and size of oil grooves.
Depending upon the main dimensions just de-
designated as C:
scribed, limiting space is obtained in which the cen-
C = CP = Cb = R- r, (1 ') tre of the journal can move freely. This space is
L: bearing length (m). called the clearance "circle", and is in fact a circle
with radius C in the case of a 360° cylindrical bear-
From these main dimensions the following di- ing; it has a different form in other cases, however,
mensionless geometrical parameters, which govern as shown in Fig. 1.3.

2
 1.2 Static Characteristics of Journal Bearings

Clearance circle

0 s : centre of segment circle

0 , : centre of journal
a 360° cylindrical bearing b Two-lobe bearing c Three-lobe bearing

Fig. 1.3 a -c. Geometry and clearance circle of journal bearings

The angle cjJ between the x-axis and On OJ is called

attitude angle. The velocity of the journal centre is
represented by x and y or e and <fi.
The journal is assumed to rotate anticlockwise
with a constant angular velocity:
w=2nN (radjs) (5)
where N represents the number of revolutions per
second. The x- and y-components of the oil film

- force acting on the journal are denoted by Fx and F,

y
respectively. The positive directions of variables ~.
y
y, x, y, e, c/J, e, <fi, f'x, and F;, are indicated by arrows
in Fig. 1.4.

1.1.4 Assumptions in the Databook

Fig. 1.4. Coordinates for
journal bearing
In this databook, the following assumptions are
made:
1.1.3 Coordinates for Journal Bearings 1. The bearing operates under fluid film lubrica-
tion.
In the present databook the following coordinates 2. The lubricant is incompressible and newtonian.
system is used to represent the characteristics of 3. No deformation of the bearing or the journal
journal bearings. With the origin at the centre On of takes place. Their axes remain parallel.
the bearing, rectangular coordinates on axes x and 4. If not explicitly stated otherwise, the flow of lu-
y are defined for the centre OJ of the journal, as bricant is isothermal, isoviscous, laminar and
shown in Fig. 1.4. The distance e = On 0 J is called free of inertia.
the eccentricity, and the ratio of e to Cb is called the 5. Vibration of the journal takes place in the vicin-
eccentricity ratio: ity of its static equilibrium, and its amplitude is
sufficiently small compared with the bearing
(4) clearance for the oil film force to be linear.

1.2 Static Characteristics of Journal Bearings

This section explains how to determine the charac- static characteristics are also required for determi-
teristics of oil film in a journal bearing in static nation of the dynamic characteristics.
equilibrium with the aid of this databook. These

3
1 Introduction

1.2.1 Equilibrium Position where J..l is the absolute viscosity of the lubricant
of Journal Centre (Pa · s). For f..l, the appropriate mean value has to be
taken when, because of heat dissipation, variations
in temperature and hence in viscosity are not negli-
A journal supported by journal bearings and rotat- gible. The units of variables W, N, D, and L are N,
ing with a constant rotational speed N is assumed 1/s, m, and m, respectively. For various bearing
to be loaded vertically and downwards by an exter- geometries these functional relationships
nal force of constant magnitude W, as shown in
Fig. 1.5. e = e(S)
(7)
Then, the centre 0 1 of the loaded journal shifts 4J=4J(S)
from the centre OB of the bearing, so that a conver-.
gent clearance space is formed between the journal are presented m tables and graphs in Chaps. 2
and the bearing surface, into which the oil is forced and 3.
by its viscosity and adhesion, so that hydrodynamic After determination of the values e and ¢, the
pressure,p, is generated in the oil film. This pressure minimum film thickness, hmin ' and its circumferen-
is called the wedge film pressure. tial location are obtained for the given geometry of
In the remaining divergent clearance space the the aforementioned clearance circle. For example,
lubricant flow is usually cavitated by aeration (ex- in the case of 360°cylindrical bearings the following
pulsion of air or gas), and the pressure in this space equation is obtained:
is close to the atmospheric pressure under normal hmin = C(1 - e) (m) (8)
operating conditions.
The resultant force P of pressure p is balanced and the location of hmin is given by ¢.
with the external load W. Because of the asymmetri-
cal distribution of p against the line OB 0 Jo the cen-
tre 0 1 of the journal does not move vertically or in
the direction of the applied load, but shifts, as 1.2.2 Frictional Coefficient
shown in Fig. 1.5, along a sort of semicircle when of Journal Bearings
the load or rotational speed is changed. For a given
set of geometrical parameters, including L : D and
mP as defined in Sect. 1.1.2, the eccentricity ratio e !he tangential stress of the oil film acting on the
and attitude angle 4J are functions of a dimension- JOUrn~ results in frictional torque T., (N m). If we set
less parameter called the Sommerfeld number, de- T., = F1 r, the coefficient of friction of the journal
fined as follows: bearing, f1o is defined as follows:

(6)
fJ=~= TJ. (9)
W rW

The ratio ~f f 1 to If/ will be also given in Chap. 2,

together w1th e and ¢, as a function of S:
fJ
p - = func (S) . (10)
If/

1.2.3 Oil Flow of Journal Bearings

The oil flow q through a bearing is usually obtained

as the sum of (i) natural flow qw induced by journal
rotation and (ii) forced flow q1 , caused by feed pres-
sure:
(11)

(a) Natural Flow qw

Imagine a bearing consisting of several segments as
shown in Fig. 1.6. For the oil flow of the i-th seg-
X W ment, there is side flow qsi on both sides of the
Fig. 1.5. Static equilibrium of journal bearing segment, and end flow q ei · A part of q ei is carried

4
 1.2 Static Characteristics of Journal Bearings

and therefore
(13)

The value of the parameter x depends upon the

structure, shape, and type of bearing, operational
conditions, and so on; it can vary between 0 and 1
and is difficult to estimate in general.
As a first approximation in design, if no empirical
value is known, a safe-side estimate of temperature
rise is suggested with
X= 1. (13) '

Now, the oil flows qs and qe can be made dimen-

sionless by

Q - q. dimensionless side flow, (14)

s- RcoCPL
Fig. 1.6. Natural flow by journal rotation q"'
q,,: side flow of segment i
q.,: end flow of segment i
Q qe dimensionless end flow. (15)
e- RcoCPL
qc,: carry-over of segment i

The values for Q. and Qe will be presented in

over to the next segment; let the ratio of this carry- Chap. 2 as functions of S:
over qci to qei be X;:
Qs = Qs(S),
(16)
(12) Qe = Qe(S).

Qs reaches maximum or minimum values for in-

The total outflow ( = inflow) in a bearing is given finitely short or infinitely long bearings, respective-
by: ly. In the case of 360° cylindrical bearings we get
qw = Lqsi + "L.qei- Lqci· Qs = 2e for L : D = 0 ,
When the sums over all segments are defined as (17)
Qs = 0 for L:D = oo.
qs = L.qsi• qe = L.qei• and qc = L.qci
Qe is defined in the databook as the circumferen-
and the overall carryover ratio as tial flow into the cavitated region at the end of
region of positive pressure distribution. In the case
X= qc (12)' of bearings with multiple segments the sum over all
qe the segments is taken. Therefore, Qe does not vanish
we get even for 360° cylindrical bearing. For the method
used in calculation of Qs and Qe, the reader is re-
L.qei- L.qci = qe- qc = (1 - x) qe, ferred to Sect. 4.1.

0.15

tiE-
0.10
/"
/
J~':1::-
0.08
<>:- /
0.06 /
'=>'oo
II
eli' 0.04 /
/
/
0.02 0. 0 1 '---'---'----'----'-----''---'---'----'----'---'
0 0.2 0.4 0.6 0.8 d/L 1.0 0 0.2 0.4 0.6 0.8 a/L 1.0
a Circular hole b Axial oil groove

Fig.1.7a, b. Forced flow q1 [4]

5
 1 Introduction

(b) Forced Flow q1

For the forced flow q1 with feed pressure p 1 (Pa) the
following equation is obtained [4]:
shr
qf=-pf·Qf (m3/s) (18)
f.i.
where h 1 (m) is the film thickness, which is assumed
to be constant around the feed hole or groove, and
Q1 is dimensionless forced flow. Figure 1.7 gives the
calculated values for Q1 in the case of a circular hole

X
and axial groove.
In case of 360° cylindrical bearing with central/
circumferential grooves the following equation ap-
plies:
3
HHJ
nC · ( 1 + 3 e2) · p 1 (m 3 /s)
=~
4
qf 2 (19) Fig. 1.8. Heat balance in journal bearing
3f.i,- H: heat dissipated in oil film
D H 1 : net heat flowing out with oil
H 2 : heat transfer from bearing housing to surroundings
where I.: is the effective length of bearing or L minus H 3 : heat conduction from bearing housing to foundation
the axial width of the groove. H 4 : heat transfer to coolant
Therefore, the total oil flow is obtained from H 5 : heat transfer from journal
Eqs. (11)-(19). H 6 : heat transfer from machine parts to bearing
8: mean temperature of oil film
q = q(J) + qf = q. + (1 - x) qe + qf 8A: surface temperature of bearing housing
80 : temperature of surroundings
or W: external load
q = RwCPL{Q. + (1- x) Qe} + q1 (m 3 /s) (20) 1}: frictional torque

where q1 can be calculated from Eq. (18) or (19).

overall heat balance in the oil film. This procedure,

originally presented by Vogelpohl [5], is explained
1.2.4 Heat Balance in Journal Bearings below. Under thermal equilibrium the equation for
heat balance is as follows (Fig. 1.8).
Formally, the static characteristics of the bearings H=H 1 +H2 +H 3 +H4 + H 5 -H6 (W) (21)
are now determined, using the Sommerfeld number
S. Dynamic characteristics will be also obtained, as where
will be seen later. However, the value for the mean H: heat dissipated in oil film by friction per
viscosity, f.i., still has to be calculated, since f.i. is second (W),
included in the Sommerfeld numberS. The f.i. value, H 1 : net heat flowing out with oil (W),
which depends heavily upon the temperature, can H 2 : heat transfer from bearing housing to the
be estimated quite easily when the speed and the surroundings (W),
load are low and therefore the heat dissipated by H 3 : heat conduction from bearing housing to
friction is negligibly small. If the temperature distri- foundation (W),
bution in the oil film is known empirically or exper- H 4 : heat transfer to coolant, if forced cooling is
imentally, the mean value for f.i. can be also deter- undertaken in bearing housing (W),
mined by averaging. In this case it is desirable to H 5 : heat transfer from journal to the surround-
know the temperature values of the bearing surface ings and rotating parts, such as disc and
at as many locations as possible. But there may also coupling (W),
be a number of cases where the heat dissipation is H 6 : heat transfer from machine parts at high
not negligibly small or no empirically known values temperatures to bearing, if such parts are
are available for temperature. To be exact, the dis- present (W).
tribution of temperature in oil film, bearing metal,
To be on the safe side, H 3 , H 4 , and H 5 should be
and housing should be calculated according to
disregarded. In addition, H 6 represents a special
thermohydrodynamic lubrication theory in such
case and is disregarded in the rest of the text. Equa-
cases (see also Sect. 4.7).
tion (21) therefore yields:
Nevertheless, for design purposes the mean tem-
perature of the oil film should be estimated from the H = H1 + H 2 (W). (21)'

6
 1.2 Static Characteristics of Journal Bearings

Since the heat H dissipated in the oil film per Eq. (24) becomes
second is given by the product of frictional torque
H 2 = &A(.9- 9 0 ) (W). (28)
'F.J and angular speed w, Eq. (9) gives:
Therefore, Eqs. (21)', (22), (23), and (28) yield, for the
H = T1 w =f1 rWw (W). (22)
thermal equilibrium of a bearing, the following
It should be pointed out that there are also cases, equation:
such as thrust bearings of the pad type, in which
heat dissipation through churning is high [6] and f 1 r Ww = C(Jq(92 - 91 ) + &A(9- 90 ) (W). (29)
has to be added to the right-hand side of Eq. (22). No reliable relation is yet established between the
The net heat H 1 flowing out with the oil is given mean temperature 9 and the oil inlet and oulet
by temperatures, 91 and 92 • But, as long as the temper-
H1 = Ceq(.9 2 - .9 1 ) (W) (23) ature difference .9 2 - 9 1 is not too wide, a weighted
mean can be used for 9, so that:
where
9 + (1 -A) .92 ( C) (30)
c: specific heat of oil (J/kg oq, = ).91 0

e: density of oil (kgfm 3 ), where for the weighting factor

q: total oil flow (m 3 fs), 2=13 (31)
.91: temperature of oil at inlet eq,
.9 2 : temperature of oil at outlet (0 C). may tentatively be suggested. Eliminating 9 2 from
Eqs. (29) and (30), we get
The heat transfer H 2 from the bearing housing into
the surroundings is given by .9 = (1 - ).) (f;r Ww + &A9o) + CQq.9l eq (32)
(24) Ceq+ &A(1- ).)
Equation (30) yields
where
&: heat transfer coefficient of the surface of 9 - 9 - A91 (OC)' (33)
bearing housing (Wfm 2 oq,
2 - 1- A
A: surface area of bearing housing (m 2),
,9A: Surface temperature Of bearing housing eq, (34)
.90: temperature of the surroundings (0 C).
For the value of&, Vogelpohl [5] recommends In addition, a relation between temperature and
viscosity is given as a graph or a formula for a given
& = 9.807 X (0.7 + 1.2 fw) (W/m 2 C).
0
(25) oil. For example, the following equation, known as
Walther's formula, applies:
The surface area A of the bearing housing can be
calculated with reference to a drawing, but for ordi- log log (v* + 0.8) = - m log T +k (35)
nary types of housings the following approxima- where
tions are suggested [5]:
m, k are constants for a given oil,
A= nHB(LB + HB/2) (m 2 ) after Palmgren (26) v* is the kinematic viscosity of oil (eSt),
or T= 273.15 + 9: absolute temperature of oil (K).
A = 9D JD (m 2 ) after Reitemeyer (26)' The kinematic and absolute viscosity in SI units are

where v = 10- 6 v* (m 2 /s),

w: mean velocity of the surrounding air (m/s), f.l = ev = 10- 6 Q [exp{exp(- m log[273.15 + 9]
HB: height of bearing housing (m), + k)}- 0.8] (Pa · s). (35)'
LB: length of bearing housing (m), All equations required for calculating the overall
D: diameter of journal (m). heat balance are given above.
For example, for w = 1 (m/s) we get

& = 9.807 X 1.9 ':::; 20 (Wjm 2 C).

0

1.2.5 Procedure for Calculation of Heat

If the temperature distribution in bearing metal Balance in a Journal Bearing
and housing is disregarded and their mean temper-
ature is assumed to be equal to the mean tempera- Figure 1.9 shows a procedure for solution by itera-
ture of oil film, that is, tion of the simultaneous equations for overall heat
(27) balance obtained in Sect. 1.2.4 for a given type and

7
1 Introduction

( Start )

f
CD Bearing type
D,L, r, C,, Main parameters to be
t - - given at the beginning
W, N, pf, o'ioo . . I> of design
c,Q, A, a, x, ..!

i
CD LID, R=~
C"=R-r ... Eq. (1)
Calculation of
S_ · · · Eq. (2) t - - geometrical parameters
1/J= R

m =I-S
Cp · · · Eq. (3)
(J

i
G) Assume{} {}: Mean temperature of oil film

1
<D p=fi ({}) .. For example,
Eq. (35)'
fl. is calculated according
to temperature-viscosity relationship

~
CD S= 1!:!!2!:__
2
ljJW
· · · Eq. (6) Sommerfeld number S, is calculated

i
c= c(S) }
· · · Eq. (7)
¢=¢(S) r - - - According to data in Chapter 2 or 3
for example,
hmin
· · · Eq. (8)

i
CD !J= 1/Jfunc(S) · · · Eq. (10)
1 - - According to data in Chapter 2
Qs= Qs(S)
} Eq. (16)
Q, =Q/S)

~
® q= Eq. (20) Oil flow q is calculated

® {}
oow
•
= Right hand side
of Eq. (32) {}"'w: New value for {}

@)
1{}-.,?>ncw l <0
• (For
example
Yes @ {}={}new @ •
Dynamic characteristics are
calculated by interpolation
1 oq {}2 = {} -,\,?>I of data given in Chapter 2
· · · Eq. (33)
1-..l or 3 using the values of S
tNo determined at @
@ {}=<&new {} <&--<&-
=T=T ··· Eq. (34)
l
2 -{}1

'
@ ~r {}obtained at@, calculations
4 - @are repeated once again
i
( Stop

@ Now, all static characteristics

are obtained:
""ll -&2, {Jo2- iJ.l, f.l, S, £, ¢, hmin' jj,q

@ Change values
of appropriate
parameters
Yes @
excessively high? 'iNa
Is{}, {}2 --&-1 or frictional loss

given at G) @
Yes
Is hmin too small"!

Yes @ .No
Is thermal expansion excessively large?

'No

Fig. 1.9. Procedure for calculation of heat balance in journal bearing.

D, L, r, Cb: dimensions of bearing o: density of oil
W: external load A: surface area of bearing housing
N: rotational speed of journal li: heat transfer coefficient of the surface of bearing
p 1 : feeding pressure housing
90 : temperature of surrounding x: overall carry-over ratio of oil
9 1 : temperature of oil at inlet A.: weighting factor for mean temperature
C: specific heat of oil
8
 1.3 Dynamic Characteristics of Journal Bearings

size of bearing and operational parameters. Start- After passing through Q3) "' @ one can proceed
ing from step CD in Fig. 1.9, the condition of heat to @, where dynamic characteristics are obtained
balance is satisfied at (j)'. from the data given in Chaps. 2 or 3 using the value
If the mean temperature 9, temperature increase for S determined above at @ '.
9 2 - 9 1 or frictional loss is excessively high at Q3), As will be explained later, in Sect. 4.3.1, it is also
cooling oil flow should be increased and/or clear- possible to estimate the temperature increase by
ance and other sizes of bearing corrected, and the assuming the ratio !;_ of the heat H 1 to the heat
calculation should be repeated from step aJ. If at (jJ) dissipation H: fi = H 1 / H, with approximately
the minimum oil film thickness hmin is too small to fi = 0.6 in most cases. Then Eq. (32) has to be re-
maintain full film lubrication, the parameters in CD placed by the following equation
should be changed. Further, if at @ the change of
bearing clearance by thermal expansion is not neg-
ligible, the change of parameters C, If/, m given at aJ 9 = (1- A)f1 rWwl;_ + 91 . (32)'
should be also taken into account. c (!q

1.3 Dynamic Characteristics of Journal Bearings

Since in journal bearings the directions of external Thus, expressing oil film force as a vector:
load and corresponding movement of journal cen-
tre do not generally coincide, as mentioned in
Sect. 1.2.1, the state of equilibrium of a rotating
F= [;~] (N) (37)
shaft supported by journal bearings becomes un-
stable under certain conditions, and then a self-ex- we get the following equations:
cited whirling of the shaft occurs, which must be F = F0 + LJF (N) (38)
prevented. When the journal vibrates, squeeze film

[~~] + [c] [~~]

pressure is also generated, in addition to the wedge
film pressure mentioned in Sect. 1.2.1. These pres- AF = [k] (N) (39)
sures give rise to spring and damping forces of oil
film and therefore influence the onset speed of the where the suffix 0 denotes static equilibrium and L1
above-mentioned whirling, the critical speeds and stands for small increments in the variables. Sym-
amplitude of the rotor system. The coefficients rep- bols [k] and [c] on the right-hand side ofEq. (39) are
resenting these spring and damping effects of oil matrices with two rows and two columns and are
film are defined below, and the methods that can be defined as follows:
used to determine the coefficients are outlined.
oFx OFx
ox oy
(N/m) (40)
oFY oFY
1.3.1 Definition of Coefficients of Oil Film
ox oy o
When the centre of a journal supported by a journal spring coefficient matrix of oil film,
bearing moves in the clearance circle, the resultant
force of oil film pressure acting on the journal 8Fx OFx
varies. The components Fx and FY of oil film force ox oy
[c] = [cxx Cxy] = (Nsjm) (41)
are expressed as functions of coordinates x and y of Cyx Cyy oFY oFY
the journal centre 0 1 and its velocity components x ox oy
and y (Fig. 1.4)
0

damping coefficient matrix of oil film .

Fx = Fx(x, y, X, y),
(N) (36) Here, the first suffix i ( = x or y) of elements kii
FY = Fy(x, y ,x, y).
and cii represents the direction of oil film force,
Functions on the right-hand side of Eq. (36) are whereas the second, j ( = x or y, x or y), represents
not linear in each variable, but can be approximat- the direction of displacement or the velocity of the
ed with sufficient accuracy by the Taylor series of journal centre.
the first order when vibration of the journal occurs Further, kii are spring coefficients of oil film, and
in the vicinity of the static equilibrium position and cii• damping coefficients of oil film. These coeffi-
both amplitude and velocity are negligibly small. cients are dimensional, and are usually made di-

9
1 Introduction

mensionless as follows ship between small movements of the journal centre

around the position of static equilibrium and corre-
K = [Kxx Kxy] = CP [k], (42) sponding small increments in the oil film force. Ac-
Kyx Kyy W cordingly, these coefficients can be determined by
giving the journal a known exciting force or known
C= [Cxx Cxy] = CPw [c]. (43) motion and by measuring the corresponding mo-
Cyx Cyy W tion of the journal centre or variation of oil film
These dimensionless coefficients of oil film force force, respectively [1, 9]. For ease of experimenta-
are functions of the Sommerfeld number S or the tion and measurement it is common to fix the jour-
eccentricity ratio for a given set of geometrical nal centre and mount a test bearing on the journal.
parameters, including L/D and mP as defined in Because it is necessary to test in the range where the
Sect. 1.1.2. They are also called dynamic character- oil film force changes linearly or Eqs. (38) and (39)
istics of oil film in journal bearings or simply dy- hold, the amplitude of the journal centre has to be
namic characteristics of journal bearings. kept sufficiently small relative to the bearing clear-
ance. But, at the same time, it has to be measured
with sufficient accuracy (less than 1 f.!m tolerance).
Since the oil film force increases rapidly with in-
1.3.2 Determination of Coefficients creasing eccentricity e, it is preferable to keep the
of Oil Film by Calculation ratio of amplitude of exciting force to static load
nearly constant for varied e, rather than to keep the
amplitude of journal centre motion constant. For
To calculate the coefficients of oil film defined details of experimentation the reader is referred to
above, it is necessary to differentiate, following the Sect. 4.2. In Chap. 3 the principles of test rigs and
definitions in Eqs. (40) and (41), the oil film force,
the test conditions are outlined, together with char-
which is obtained by integration of pressure p over acteristic data obtained experimentally with rigs.
the journal surface [1 , 8]. Or it is possible to im-
prove the accuracy of this calculation by inverting
the sequence of differentiation and integration [7].
The distribution of p is obtained by solving the 1.3.4 Comparison Between Calculated
Reynolds equation under a boundary condition
that applies for the pressure at the peripheries of the and Experimental Values
oil film. This condition is not fully known or settled for Coefficients of Oil Film
yet, because of the complexity of rupture or cavita-
tion of oil film. But all data in the present databook, There are open points in calculated and experimen-
unless explicitly stated otherwise, have been calcu- tal values for coefficients of oil film as follows. The
lated assuming Reynolds boundary condition, reliability of calculated data depends upon various
which requires that the pressure gradient vanishes assumptions, especially the boundary conditions
at the end of the positive pressure region. This con- for pressure, temperature distribution, and defor-
dition seems to be the most common at present. For mation of the bearing metal.
further details and the calculation procedure, the There is experimental data in which the accuracy
reader is referred to Sect. 4.1. for measuring the small amplitude and the means of
bearing support and excitation require further im-
provement. Therefore, efforts were made for the
1.3.3 Determination of Coefficients purposes of this databook to determine and com-
pare the coefficients of oil film force both by calcu-
of Oil Film by Experiment lation and experimentally, with the same conditions
as far as possible. At present, differences between
As Eq. (39) predicts, the coefficients of oil film force them seem to be unavoidable, and there is a real
defined by Eqs. (40) and (41) represent the relation- need for further research.

10
 1.4 How to Use the Characteristic Data of Bearings

1.4 How to Use the Characteristic Data of Bearings

in the Present Databook

1.4.1 Nature of the Data The Sommerfeld number is

Data collected in the present databook are calculat- s= pN LD

w (Rc )
p
2
= 0.244.
ed data (Chap. 2) and experimental data (Chap. 3).
Chapter 2 presents calculated static characteristics The table gives c =: 0.3 and spring and damping
for 59 kinds of bearing (eccentricity ratio, attitude coefficients as follow
angle, oil flow rate and coefficient of friction), and Kxx '=: 6.82, kxx = 66.9 MN/m
also the dynamic characteristics (spring and damp- (6.82 x 104 kgf/cm);
ing coefficients) in the form of tables and graphs.
KYY '=: 4.56, kyy = 44.7 MN/m
The Sommerfeld number, S, is used as an indepen-
(4.56 x 104 kgf/cm);
dent variable. In Chap. 3 experimental data (eccen-
tricity ratio, attitude angle, spring and damping co- Cxx =: 7.25, cxx = 56.6 kN · sfm (57.7 kgf · sfcm);
efficients) for 11 kinds of bearing are collected and cy y =: 5.78, Cyy = 45.1 kN. s/m (46.0 kgf. sfcm);

are also compared with calculated data. kxy = k y x = Cx y = Cy x = 0 .

For S values other than those presented in tables,
characteristic data should be obtained from the ta-
1.4.2 An Example of Data bles by interpolation of values near the given S or
obtained from graphs.
As the data collected in the present databook are
presented in dimensionless form, it is usually neces-
sary to convert the values to a dimensional form 1.4.3 Influences of L : D Ratio,
before application. For conversion, the following Turbulence, Load Direction, etc.
relations obtained from Eqs. (42) and (43) apply:
w Although data is presented for many bearing types
k.1, ] -=c
- K l., J'. (44) and specifications in the present databook, there
p
are still more variations in practical use. In the case
w of other L : D ratios than those given here, for exam-
C·
!,] c (1) !, J'
·=--C.·
p
(45)
ple, the calculation method described in Sect. 4.5 is
applicable. The data given in the present databook
where the suffixes i and j represent x and y. are concerned with the case of laminar flow. For
An example of conversion is presented for item turbulent flow, a procedure for calculation is given
53 in the calculated data (tilting pad bearing with 5 in Sect. 4.6.
pads, LOP, mP = 0.5, L: D = 1). The following In this databook the external load on a journal is
specifications are assumed: assumed to act vertically downwards. For other
N = 200 1/s, L = 80 mm, D = 2R = 80 mm, load directions the data here is not of much practi-
cal use except for special cases like that of 360 o
cb = 100 Jlm, cp = 200 Jlm, cylindrical bearings. This is because the characteris-
W = 1960 N (200 kgO, T;n = 40 oc tic data of nonisotropic bearings, such as two-seg-
Oil: Turbine oil # 90 ment bearings, depends heavily upon the direction
of external load. Nevertheless, data are presented
To obtain the characteristic data of a bearing for both directions of load, namely for LOP and
(dimensionless or dimensional), the condition of LBP in the case of tilting pad bearing. Therefore,
heat balance and the mean temperature of the oil approximate data for intermediate load directions
film first have to be determined as accurately as may be calculated by interpopulation. The influ-
possible, as explained in Sect. 1.2. For this purpose ences of load direction on bearing characteristics
the procedure in Fig. 1.9 can be used as a guide. are discussed elsewhere [1].
In this example, however, let the mean tempera- In this databook the direction of journal rotation
ture rise and the mean temperature of the oil film be is assumed to be counterclockwise, as defined in
assumed to be 30 oc and 70 ( = 40 + 30) oc, respec- Sect. 1.1.3. When the geometry of sliding surfaces is
tively. The mean viscosity of oil film Jl = 9.37 mPa · s completely symmetrical across the vertical loading
( = 0.955 10- 7 kgf.sjcm 2 ) is obtained from the tem- line through the centre of the bearing the dynamic
perature-viscosity table. characteristics do not change if the direction of

11
1 Introduction

journal rotation changes, but they have to be calcu- Sects. 1.2.4 and 4.3. This databook is intended to
lated afresh if the surfaces are asymmetrical. contribute to the prediction of vibrational be-
It goes without saying that the coordinates x, y of haviour, including that of the stability of a rotating
the journal centre and the positive directions of shaft supported by a journal bearing. How much
components Fx, FY of the oil film force acting on the the dynamic characteristics of oil film influence the
journal (not on the bearing) are those defined in vibrational behaviour depends upon the shape of
Sect. 1.1.3. In cases of other coordinates systems the the shaft system. When this influence seems to be
conversion given in Sect. 4.1 is needed. large, as in the case of an overhanging shaft, the
accuracy of calculated vibrational behaviour
should be checked carefully.
1.4.4 Some Notes on Application In general, calculated and experimental data
have to be compared against each other, checked,
of the Data and then applied. Accurate prediction of other
forces than that of the bearing, such as the forces of
For effective use of the present databook, it is most labyrinths, discs, seals etc., is also important, as is
important to give the appropriate mean tempera- the correct evaluation of the influences of structures
ture of the oil film, as already mentioned. Methods that support the bearings.
of predicting the temperature are described in

References
1. Glienicke J (1970) VDI-Z, Fortschritt- Berichte 1: 22 5. Vogelpohl G (1967) Betriebssichere Gleitlager. Springer,
2. Birembaut MY (1977) Raideur et Amortissement des Paliers Berlin Heidelberg New York
Hydrodynamiques. Mem Techn CETIM (Paris) 32 6. Neal PB (1982) Heat transfer in pad thrust bearings. PIME
3. JSME (1982) Dynamic characteristics of journal bearing and 196: 217 - 228
vibration of rotating machineries - Application to design 7. Schaffrath G (1969) Forschung im lngenieur-Wesen 35 (6)
based on the newest data (in Japanese). JSME, Tokyo (Text- 8. Someya T (1976) Transaction of JSME 42 : 360, 2599 (in
book no. 548 for the JSME lecture course) Japanese)
4. Hirano F (1957) Transactions of JSME 23: 134, 717; Junkat- 9. Someya T (1976) Transactions of JSME 42 : 362, 3303;
su (JSLE) (1962) 7: 5, 225 (in Japanese) Junkatsu (JSLE) (1975) 20 : 3, 191 (in Japanese)

12
2. Calculated Data

In this chapter, calculated data for 59 kinds of bearing geometries is collected. The format of the data
collections, except those for calculations 56-59, is as follows.
First page : Geometry of the bearing whose data is calculated, outline of the calculation method, and
tabulation of the calculated results
Second page: Figures for eccentricity, attitude angle, and oil flow versus Sommerfeld number, and for
friction coefficient versus Sommerfeld number
Third page: Figures for spring coefficients versus Sommerfeld number and for damping coefficients
versus Sommerfeld number
The main symbols used are as follows (see Nomenclature).

S: Sommerfeld number=
J.1NLD
W (R)
Cp
2

e: Eccentricity ratio = ~b
</>: Attitude angle (degrees)
Q: Nondimensional flow= (Rm~PL)
f 1 If/- 1 : Friction coefficient for journal
Kii: Nondimensional spring coefficient = C~ij (i,j = x, y)

C;/ Nondimensional damping coefficient = cp;,cij (i,j = X, y)

13
Calculation No. 1

1. Calculation no.: 1 5. Bearing geometry

2. Bearing type: Cylindrical, with two
axial grooves
3. L: D ratio: 0.5
4. Special items
Oil groove:
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 18 (uni-
form spacing) Circum-
ferential direction: 99
(uniform spacing) for
each of upper and
lower pads

Static Characteristics Spring Coefficients Damping Coefficients

s
€ r/> Qs Q, IJ'V' K xx Kxy K yx Kyy Cx x C xy C yx C yy

5.96 . 0750 82.6 .0663 l . 00 111.5 1.77 l3 . 6 -13. 1 2 .72 27. 2 2.06 2.06 14.9

4.43 .100 80.4 • 0880 .999 83.0 1. 7 5 10.3 -9.66 2.70 20.6 2 ,08 2,08 11.4

2.07 .200 71.6 .170 .999 39.5 1. 8 8 5.63 -4.41 2.56 11.2 2.22 2.22 6.4 3

l. 24 .300 64.1 .243 • 997 24.4 2.07 4.27 -2.56 2.34 8.50 2.32 2.32 4.73

• 798 .400 57.5 .309 • 986 16.4 2.39 3 .75 -1. 57 2.17 7.32 2 .24 2.24 3,50

.517 • 5 00 51.1 .366 . 967 11. 3 2.89 3.57 -. 924 2.03 6.81 2.10 2.10 2.60

.3?3 .600 44.7 .416 • 939 7.75 3. 65 3. 62 -.427 1.92 6.81 2.08 2.08 2.06

.187 .700 38.2 .459 .900 5.13 4. 92 3.88 .0235 l. 83 7.32 2.16 2.16 l. 70

.135 .750 34.9 .478 .875 4.07 5. 90 4.11 • 2 58 1.80 7,65 2 .10 2.10 l. 46

.0926 .800 31.2 . 496 .846 3 . 13 7. 35 4.46 .5 27 l. 7 8 8.1 7 2 .05 2.05 l. 24

.0582 .850 27.2 .510 . 806 2.30 9.56 4.92 .805 1. 73 9.12 2.06 2.06 l. 06

.0315 .900 22.7 .524 .758 l. 56 l3 .8 5.76 l. 24 l. 72 10.6 2.03 2.03 ,846

.00499 .975 12.5 .543 .663 .54 43. 3 9.61 4. 2 6 l. 99 22.4 2.93 2. 93 .637

14
 Calculation No. 1

0.8

0.6
~ cu Qs -e.

0.4 0.4

30°

1s•

0 0
0.005 0.02 0.05 0.1 0. 2 0.5 2 5
s

lOOr----------------------------------------------------------------,

15
Calculation No. 1

s
lOOr---- ----------- ----------- ----------- ----------- ----------- ---,

50

20

·~

0.02 0.05
s

16
 Calculation No. 2

1. Calculation no.: 2 5. Bearing geometry

2. Bearing type: Cylindrical, with two
axial grooves
3. L: D ratio: 1.0
4. Special items
Oil groove: 10° X 2
Calculation method : FDM (over-relaxa tion
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 18
(uniform spacing) Cir-
cumferentia l direction:
99 (uniform spacing)
for each of upper and
lower pads X

Static Characteristi cs
s Spring Coefficients Damping Coefficients
c ¢ Qs Q, !J<P-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
2.95 .0500 83.0 . 0351 .999 55.1 20.1 -19.8 2.90 ~0.7 1.65 1.65 1~.5
2.38

1. 96 . 075 80.8 . 0523 .999 36.6 2.18 13.5 -13 .o 2,88 27.4 1.66 1.66 9.81

1. 45 .100 78.7 .0692 • 998 27.3 2.04 10.2 -9.62 2. 77 20.8 1. 67 1. 67 7.55

.689 .200 69 . 9 .132 • 998 13.2 1. 99 5.45 -4.38 2.56 11.3 1. 78 1. 78 4.39
.424 .300 61.0 .181 .998 8.41 2.20 3.96 -2.52 2.31 8.67 1. 99 1. 99 3.50

.284 .400 54.6 .220 . 991 5.93 2.37 3.34 -1.60 2.06 7.57 2.10 2.09 2 . 94

.194 .500 49.1 .251 • 972 4.37 2.67 3.09 -1. 00 1. 89 6.99 2.00 2.00 2.31

.131 .600 4 3. 8 .274 • 943 3. 26 3.19 3.06 -.553 1. 77 1. 82 1. 82

6.69 1. 74

,0 828 .700 38.4 . 295 . 911 2.39 4,16 3 . 29 -.0851 1. 75 7 , 07 2.00

2.00 1. 59

. 0469 .BOO 32.1 .302 .855 1. 69 6.10 3.82 .420 1. 74 8.08 2.15 2.15 1. 34

.0195 .900 24.0 .300 .770 1. 04 11.9 5.28 1. 32 1. 80 10.1 2.06 2.06 .916

.00866 . 950 18 .0 .297 . 710 .684 24.0 7.81 2.79 2.01 14.2 2 .44 2.44 • 791

17
Calculation No. 2

goo

0.8
75°

0.6 60°

45°
0.4
30°

0.2 0·2
15°

18
 Calculation No. 2

lOOr--------------- --------------------- --------------------- -----,

....
~

20

....
0

19
Calculation No. 3

1. Calculation no.: 3 5. Bearing geometry

2. Bearing type: Cylindrical, with four
axial grooves
3. L: D ratio: 0.5
4. Special items
Oil groove:
Calculation method: FDM (over-relaxation y
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(uniform spacing) Cir-
cumferential direction:
16 (uniform spacing)
for each of upper, low-
er, left, and right pads

Static Characteristics Spring Coefficients Damping Coefficients

s JJ<P- l C yy
€ r/> Qs Qe Kxx Kxy K yx Kyy C xx C xy C yx

16.8 .0500 87.5 .0232 2.00 294 1.90 20 . 0 -19.0 1.53 40.1 1.43 1.50 36.4

8.25 .100 85.3 .0467 1.99 145 1.88 10 . 1 - 9.53 1.52 20.2 1.52 1.51 18.8

3.90 .200 80.7 .0942 1.98 69 . 9 2 .03 5.34 - 4 . 30 1.46 10 . 3 1.46 1.46 8.87

2.35 .300 75.6 .143 1.96 43.3 2.32 3.84 - 2.16 1.31 7.21 1.27 1.26 4.90

1.49 .400 69.6 .196 1.93 28.7 2. 63 3, 21 - 1.25 1.22 5.76 1.19 1.20 3.21

1.16 .450 65.9 .224 1.90 23.4 2.92 3. 02 - • 737 1.10 5 . 33 1.02 1.02 2,27

.904 .500 60.6 .255 1.67 16.5 3.37 2.89 - .200 .939 4.99 ,742 .744 1.33

.658 .550 53.4 .287 1.62 14.0 3.65 2.96 .0660 .940 5.05 . 704 .709 .954

.469 .600 45.6 .315 1. 76 10.5 4.39 3.21 .166 1.04 5 .49 . 844 .850 ,867

.336 .650 39.3 .337 1.71 7.94 5.02 3.51 .267 1.14 6 . 10 1 .01 1 . 02 .832

.239 .700 34.2 .357 1.66 6.05 5.78 3. 86 .319 1.24 6.89 1.22 1.23 .835

.111 .800 26.7 .392 1.57 3.42 6.04 4.66 . 361 1.46 9.28 1.86 1.87 .940

.0353 .900 19.7 .424 1.47 1.61 14.9 5.88 .840 1.39 ll.5 1. 74 1. 76 .627

.0129 .950 15.0 .438 1.40 .907 26.6 7.80 1.44 1.50 15.6 1.99 1.97 . 527

20
 Calculation No. 3

2.0
1.8
1.6
1.4

1.2
~ 1.0 w -e..

0.8

0-6
0.4

0·2

21
Calculation No. 3

50.--------------------------------------------------------------,

·~

::.::
2
Kyy

0.5

50.------------------------------------------------------------,

C xy

22
 Calculation No.4

1. Calculation no.: 4 5. Bearing geometry

2. Bearing type: Cylindrical, with four
axial grooves
3. L: D ratio: 1.0
4. Special items
Oil groove: 10° X 4
Calculation method: FDM (over-relaxation
y
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(uniform spacing) Cir-
cumferential direction:
16 (uniform spacing)
for each of upper, low-
er, left, and right pads

Static Characteristics Spring Coefficients Damping Coefficients

s ¢ Qe JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
e Qs

8.57 .0500 87.9 .0130 2.00 151 2.24 20.0 -17.7 1.26 40.0 1.29 1.25 35.2

4.24 .100 85.9 .0260 2.00 74 .a 2.00 10.1 -8.89 1.26 20.1 1.25 1.24 17.7

2.01 .200 81.8 .0523 1.99 36.2 2.10 5.26 -3.88 1.20 10.3 1.18 1.17 8.04

1.22 .300 77.1 .0797 1.98 22.7 2.26 3.73 -2.28 1.16 7.10 1.14 1.15 5.01

• 794 .400 71.8 .109 1.95 15.4 2.60 3.03 -1.11 1.02 5.56 .916 .920 2.80

.636 .450 68.4 .125 1.93 12.7 2.84 2.82 -.622 .919 5.09 .781 .781 1.97

.498 .500 63.8 .142 1.90 10.3 3.13 2.69 -.290 .856 4.80 .692 .697 1.42

.366 .550 56.3 .161 1.86 7.91 3.58 2. 71 .0346 .824 4. 77 .608 .613 .921

.260 .600 47.6 .178 1.80 5.92 4.11 2.92 .196 .903 5.15 .695 .701 • 767

.186 .650 40.1 .190 1. 74 4.51 4. 71 3.23 .291 1.01 5.79 .853 .860 .723

.134 .700 34.2 .201 1.68 3.50 5.41 3.58 .349 1.11 6.62 1.05 1.06 .719

.0661 .BOO 26.1 .218 1.59 2.12 7.38 4.52 .368 1.34 9.10 1.67 1.68 .824

.0242 .900 19 .a .235 1.49 1.16 13.6 5.37 .855 1.30 10.9 1.55 1.56 .563

.00980 .950 15.2 .243 1.42 .728 24.2 7.37 1.41 1.47 15.3 1.97 1.94 .528

23
Calculation No.4

24
 Calculation No. 4

SOr-------------------------------------------------------1

...,
~

...,
0

25
Calculation No. 5

1. Calculation no.: 5 5. Bearing geometry

2. Bearing type: Cylindrical, with pres-
sure dam
3. L: D ratio: 1.0
4. Special items L
Oil groove: 20° x 2
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 18
(uniform spacing) Cir-
cumferential direction:
60

Static Characteristics Spring Coefficients Damping Coefficients

s
€ ¢ Qs Qe ht/rl Kxx Kxy Kyx K yy Cxx Cxy Cyx Cyy

1.72 0.500 6 2.5 0.5 0 4 0. 9 73 27.7 4.5 9 8.55 -2.7 7 5.3 0 1 4.0 6. 1 5 5.9 0 8.44

1.0 9 0.5 50 57.8 0. 5 3 7 0.9 7 0 1 8.3 4.6 7 6.8 0 - 0.8 3 2 4.1 9 1 0.8 4.3 8 4.3 7 5.18

0.7 4 2 0.6 0 0 5 2.7 0.5 6 4 0.9 6 9 1 2.9 4.7 2 5.7 9 -0.2 7 3 3.3 4 9.2 4 3.4 6 3.4 6 3.60

0.517 0.6 50 4 7. 9 0.5 9 4 0.9 56 9. 52 4.8 5 5.2 2 -0.0993 2.8 5 8.08 2.9 5 2.9 5 2.7 2

0.36 2 0.7 0 0 4 3.2 0.611 0.9 52 7. 0 9 5.4 3 5.1 1 0.231 2.5 0 8.30 2.8 4 2.84 2.11

0.249 0.7 50 38.6 0.6 3 1 0. 9 3 9 5.2 7 6.2 0 4.8 5 0.414 2.21 8.0 1 2.6 3 2.64 1.70

0.165 0.8 0 0 3 4.0 0.6 4 6 0.9 21 3.8 5 7.5 7 5.1 5 0.697 2.1 7 8.58 2.4 2 2.4 3 1.52

0.10 2 0.8 50 2 9.5 0. 6 4 6 0.9 2 0 2.7 1 9. 6 9 5.33 0.985 2.0 5 8.95 2.3 9 2.4 0 1.21

0.05 58 0. 9 0 0 2 4.0 0.6 4 4 0.9 0 1 1. 7 9 14. 1 6.28 1.38 1. 9 9 10.8 2.3 3 2.3 4 1.0 I

0.0219 0.9 50 17.1 0.6 31 0.8 6 7 0.98 5 30.0 7.86 2.91 1.82 1 3.0 1.77 1.78 0.5 55

26
 Calculation No. 5

50r-------------------------------------- --,

27
 Calculation No. 5

50r----------------------------------------,

....,
:<

20~--------------------------------------------~

....,
0

28
 Calculation No.6

1. Calculation no.: 6 5. Bearing geometry

2. Bearing type: Partial arc
3. L: D ratio: 0.5
4. Special items
p: 120°
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points : Axial direction: 12
(uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Q, !J<P-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
5.52 .1 75 . 1 .402 2.89 37.3 2.07 10.5 -2.14 1.14 20.9 .902 1.11 4.18

2.55 .2 63.2 .737 2.59 18.4 2.22 5.94 -1.16 1.20 11.8 1.22 1.23 2.44

1.49 .3 54.6 .997 2.27 11.8 2.42 4.62 - .843 1.27 9.12 1.33 1.34 1. 89

.931 .4 48.1 1. 20 1.97 8.35 2.70 4.12 - .695 1.38 8. 01 1.45 1.46 1.64

.588 .5 43.2 1.35 1. 66 6.11 3.14 3.93 - .564 1.46 7.52 1.57 1.59 1.49

.359 .6 38.5 1.46 1.36 4.94 3.98 2.82 - .285 1.38 7.43 1.65 1. 68 1.34

. 203 .7 33.5 1.50 1.05 3.24 5.24 3.95 - .0348 1.36 7.61 1.61 1.66 1.10

.145 .75 30.9 1.50 .892 2.70 6.23 4.05 .133 1.34 7.87 1.60 1.64 .982

.0984 .8 28.0 1.47 .774 2.19 7.66 4.29 .324 1.34 8.34 1.60 1.65 .879

.0612 .85 24.8 1.41 .574 1.71 9.89 4.73 .560 1.37 9.13 1.62 1.69 .778

.0328 .9 21.0 1.32 .409 1.25 14.0 5.64 .897 1.47 10.6 1. 68 1. 76 .675

.0124 .95 15.9 1.16 .241 . 767 26.6 7.79 1.86 1.60 13.9 1. 79 1.92 .546

29
Calculation No.6

1.0,---------------------------------------- -----------------,

50,-----------------------------------------------------~

30
 Calculation No.6

50.---------------------------------------------------------.

50,----------------------------------------------------.

31
Calculation No. 7

1. Calculation no.: 7 5. Bearing geometry

2. Bearing type: Partial arc
3. L: D ratio: 0.5
4. Special items
P: 150°
Calculation method: FD M
Boundary condition: Reynolds condition
Number of grid points: Axial direction : 12
(uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Q, Qe fJtP-l Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

4 .72 .1 77.6 .500 2.86 39.7 1.65 10.5 -3.97 1.71 20.7 1.00 1.58 7.26

2.20 .2 67.6 .941 2.56 19 . 6 1. 78 5.85 -2.14 1. 75 11.5 1.59 1.71 4.16

1.31 .3 60.1 1.31 2.26 12.7 1.94 4.54 -1.57 1.82 8.70 1. 75 1.81 3.14

.838 .4 53.9 1.62 1.97 9.04 2.31 3.94 -1.04 1. 73 7.54 1.85 1. 89 2.59

• 539 •5 48.1 1.87 1.67 6.68 2.85 3. 71 - .661 1.66 7.04 1.85 1.91 2.21

.335 .6 42.4 2.05 1.34 4.95 3. 68 3.70 - . 324 1.59 6.90 1. 78 1.82 1.66

.192 •7 36.4 2.16 1.06 3.56 4.97 3.83 .0236 1.51 7.16 1. 74 1. 78 1.31

.139 .75 33.3 2.18 .899 2.96 5.93 4.00 .197 1.51 7.48 1. 74 1.77 1.17

.0947 .8 30.1 2.17 • 740 2.39 7.23 4.37 .3 57 1.55 7.98 1. 73 1. 78 1.03

.0594 .85 26.4 2.13 .578 1.85 9.42 4.86 .615 1.57 8.81 1.75 1.80 .895

.0322 .9 22.2 2.04 .413 1.33 13.7 5.65 1.06 1.57 10.3 1.81 1.88 .768

.0123 .95 16.5 1.86 .242 .799 25.9 7.98 1.97 1. 73 13.8 1.90 2.01 .596

32
 Calculation No.7

1.0.---------------------------------------- -----------------.

50.--------------------------------------- ---------------,

33
 Calculation No. 7

50~----------------------------------------------------~

5
s

50~------------------------------------------------------~

·~

5
s

34
 Calculation No. 8

1. Calculation no.: 8 5. Bearing geometry

2. Bearing type: Partial arc
3. L: D ratio: 1.0
4. Special items
{J: 120°
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(uniform spacing) Cir-
cumferential direction :
30 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
€ r/> Q, Qe JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

2.19 .1 71.9 . 271 2.94 14.9 2 . 04 10.7 -1 . 40 .920 21.4 .924 .923 2.82

1.03 •2 59.0 .477 2.65 7.57 2.19 6.07 - .776 .966 12.3 1.08 1.01 1.69

.615 .3 50.4 .623 2.35 5.03 2.38 4.69 - .574 1.04 9.63 1.17 1.13 1.36

.394 .4 44.0 .723 2.05 3.67 2.61 4.17 - .488 1.12 8.51 1.26 1.26 1.21

.256 .5 39.4 . 790 1. 74 2.80 2.92 3. 89 - . 438 1.21 7.98 1.37 1.39 1.14

.165 .6 35.7 .832 1.44 2.18 3.47 3.75 - .342 1.25 7.82 1.51 1. 53 1.10

.100 .7 31.8 .837 1.13 1. 70 4.52 3.68 - .127 1.22 7.94 1.57 1.61 1.00

. 0750 .75 29.7 .824 .964 1.49 5.34 ), 76 .0096 1.22 8.10 1.55 1.60 . 914

.0539 .8 27.4 .798 .802 1. 29 6.49 3.93 ,170 1.24 8.43 1.55 1.61 .833

.0361 .85 24.7 . 755 .636 1.08 8 . 33 4 . 32 .372 1.30 9.08 1.58 1.64 .754

.0213 .9 21.3 .691 .468 .867 12.2 4.97 .829 1.34 10.3 1.65 1. 72 .672

.00922 .95 16.5 .590 • 296 .607 23.8 6.91 1.92 1.49 13 . 5 1.81 1.92 .568

35
Calculation No. 8

QJ w

50,------------------------------------- ---------------,

5
s

36
 Calculation No. 8

50~------------------------------------------------------~

50,---------------------------------------------------------,

Cxx "-._

·~

5
s

37
Calculation No. 9

1. Calculation no.: 9 5. Bearing geometry

2. Bearing type: Partial arc
3. L:D ratio: 1.0
4. Special items
p: tsoo
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
e rp Qs Q. /J¢-1 Kxx Kxy K yx Kyy Cxx Cxy Cyx C yy

1.65 .1 75.2 .373 2.91 14.0 1.67 10.5 - 2 .59 1.36 21.0 1.06 1.31 4.97

.786 .2 63.8 .677 2.63 7.11 1. 72 5.87 -1.43 1.40 11 . 8 1.42 1.44 2.93

.479 .3 55.8 .907 2.34 4.78 1.89 4.52 -1.08 1.46 9 . 11 1.54 1. 56 2.92

.315 .4 50.0 1.09 2.05 3.54 2.06 3.93 - .904 1.51 7.90 1.64 1. 66 1.98

.212 •5 45.4 1.22 1. 76 2.76 2.49 3.55 - .615 1.45 7.32 1. 72 1. 75 1.77

.141 .6 40.7 1.31 1.45 2.20 3.10 3.40 - .374 1.41 7 .09 1.71 1. 75 1.51

.0883 •7 35.8 1.33 1.14 1. 75 4.15 3.42 - .0613 1.36 7.17 1.65 1. 70 1.23

.0674 .75 33.2 1.32 .982 1.54 4.93 3.54 .0980 1.36 7.41 1.66 1. 70 1.11

.0493 .8 30.3 1.29 .819 1.34 6.00 3.78 .2 62 1.39 7.81 1.67 1.71 1.00

.0336 .85 27.0 1.24 .654 1.13 7.91 4.17 .565 1.41 8 .53 1.71 1. 76 .897

.0202 .9 23.0 1.15 .487 .902 11.4 5.02 .952 1.51 9.87 1.77 1.84 . 781

.00898 .95 17.5 1.02 .315 .627 23.1 7.03 2.06 1.63 1).1 1.91 2.03 .640

38
 Calculation No. 9

1.0r----------------------------------------------------,

50r-----------------------------------------------------~

39
Calculation No. 9

50r---------------------------------------------------------,

5
s

50r------------------------------------------------------.

...,
0

40
 Calculation No. 10

1. Calculation no.: 10 5. Bearing geometry

2. Bearing type: 2lobe
3. L:D ratio: 0.5
4. Special items
Oil groove: 10° X 2
mP: 1/2
P: 170°
Calculation method : FDM (matrix-column
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing) Cir-
cumferential direction:
33 (uniform spacing)
for each of upper and
lower pads X

Static Characteristics Spring Coefficients Damping Coefficients

s c: ¢ ! Qe fJtP-J Kxx Kyy Cyy
Qs Kxy Kyx Cxx Cxy Cyx

4.79 .0556 87.2 . 292 .530 104 61.5 33 . 1 -43.3 2. 06 123 -43.3 -42.5 36.6

3.06 .0868 87.2 .293 .524 65.9 39.4 21.7 -28.0 1. 43 80.0 -27.7 -27.7 23.4

1. 93 .137 86 . 9 .196 . 520 41 .6 25.1 13.7 -17.7 1.16 50.2 -17.9 -17.5 15.0

.993 .262 85.9 .312 .502 21.4 13.5 7.50 -9.20 .897 26.5 -8.47 -8.34 8.28

.496 . 438 84 . 4 . 343 .469 10.8 7.88 4.52 -4.60 .831 14.0 -3.42 -3.36 4. 71

.306 .608 80.8 . 370 .441 6.99 6.06 3. 7 3 -2.82 . 940 10.5 -1.13 -1.11 3.45

.209 . 718 77.2 .402 . 404 4. 91 5.66 3.60 -1.67 1. 16 8.74 -.0503 - . 0240 2.72

. 105 . 888 68.3 .433 .364 2. 73 6.51 4. 19 -.115 1. 57 7.86 1.44 1. 45 1. 89

.0507 .970 56.8 .436 • 353 1. 61 9.47 5. 46 . 945 1. 94 9.03 2.29 2.31 1. 38

.0303 .995 48.6 . 424 . 359 1. 12 13.2 6.45 1. 72 2.12 10 . 4 2.58 2.62 1.16

.0207 1.01 42.5 .407 . 370 . 891 17.2 7. 81 2.33 2.24 11.8 2. 70 2.73 1.02

.0100 1.01 34.1 . 372 .400 . 590 29 . 6 9.68 3. 28 2.39 15 . 3 2.88 2.93 .875

41
 0.2

0 0~--~~--~~~~~~~~--~~~~~~~~----~----~~~~ 0°
0.005 0.02 0. 01 0.05 0.1 0.2 0.5 1 2 5 10

100~--------------------------------------------------------,

50

20

10

5
-;-
"S.
~

0.5

0 - 2~--~~~~~~~~~~~--~~~~~~~----~~--~~~~
0.005 0. 01 0.02 0.05 0. 1 0.2 0.5 1 2 5 10
s

42
 Calculation No. 10

lOOr----------------- ----------------------- ----------------------.

....,
~

100r---------------- --------------------- --------------------- ,

50

20

10

5
....,
G

-Cxy • - C y x
1

0.5

0.2~~~~--~~~~~~~~~~--~~~~~----~--~~~~~
0.005 0.01 0.02 0.05 0.1 0.5 1 2 5 10
s

43
Calculation No. 11

1. Calculation no.: 11 5. Bearing geometry

2. Bearing type: 2lobe 170°
3. L: D ratio: 0.5
4. Special items
Oil groove: 10° X 2
mP: 2/3
p: 170°
Calculation method: FDM (matrix-column
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing) Cir-
cumferential direction:
33 (uniform spacing)
for each of upper and
lower pads X

Static Characteristics Spring Coefficients Damping Coeffi ci en ts

s ¢ ljt/J -1
€ Qs Qe Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

5.25 .0313 79.0 .374 . 362 135 224 55.7 -120 23.0 314 101 -101 66 . 2

3.00 . 0547 79,0 • 375 . 362 76.7 129 30.2 -68.6 13.2 176 -57.8 -57.7 38.0

2.02 .0828 79.0 .375 .362 50 .7 87.4 19.3 -46 . 1 8.97 112 -38.7 -38.5 25 .5

1. 01 . 167 78.8 . 377 .360 25.8 43.9 9. 85 -23 . 1 4.61 61.9 -19.0 -18.9 13. 0

.515 .317 78.3 . 381 .354 13.2 24 .0 5.29 -11.8 2.60 30.2 -9.42 -9.38 6.86

• 304 .505 77.3 .394 . 341 8.01 15.5 4.07 -6. 77 1. 89 19. 7 - 5.27 -5.2 4 4.42

.201 .662 76.0 .402 .331 5 . 52 12 .3 3. 74 -4.61 1. 56 15.1 - 2.89 -2 .82 3.13

.0985 . 921 73.0 . 430 .296 3.01 10.3 4.0 1 - 1.86 1. 50 10.8 - .272 -.245 1. 95

.0524 1.06 67.6 .445 . 280 1. 79 11.6 4.79 - .23 1 1. 62 9. 72 .6 19 .632 1.28

.0307 1.13 61.9 .444 .277 1. 22 14.4 5.82 . 820 1. 75 10 . 9 1. 51 1. 55 1. 04

.0206 1. 15 57.0 .440 . 28 1 . 941 17. 7 6. 92 1. 52 1. 88 12 . 2 1. 96 2.02 .957

.0100 1.13 47.1 .416 . 300 . 602 28 .0 9.23 2.89 2. 14 15. 6 2. 71 2.75 . 841

44
 Calculation No. 11
1.2

1.0 1.0
goo
0.8 0. 8
75°

0.6 0. 6 60°
C1 w -e.
45°
0.4 0.4 Qs

0.2 0.2

0 0 oo
0.005 10
s

100~-----------------------------------------------------------,

50

20

10

5
-;
-s.
~

0.5

0.2~~~~--~~~~~~~~~--~~._~~~~~----~~~._~~~.
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10
s

45
Calculation No. 11

100r---------------------------------------------~----------~

50

20

10

0.5

1 2 5 10
s

46
 Calculation No. 12

1. Calculation no.: 12 5. Bearing geometry

2. Bearing type: 2 lobe 170°
3. L: D ratio: 0.5
4. Special items
Oil groove: 10° X 2
mP : 3/4
fJ: 170°
Calculation method: FDM (matrix-column
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing) Cir-
cumferential direction :
33 (uniform spacing)
for each of upper and
lower pads X

Static Characteristics Spring Coefficients Damping Coefficients

s
c ¢ Qs Qe fJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

5.06 .0227 76 . 9 .412 .282 154 474 72.0 -215 49.8 548 157 -157 92.5

3.01 .0383 76 . 9 . 412 .282 91.7 282 40 . 6 -128 29.7 319 -94.3 -94.1 54.4

2.04 .0571 76.9 . 412 .281 62.1 186 26.5 -86.7 18.2 215 -63.8 -63 . 6 36.9

.983 .108 76.8 .412 .281 30 . 0 93.7 12.7 -43.2 9. 79 106 -30.8 -30.7 17.9

.477 .254 76.8 .414 .279 14.7 46.6 6.83 -20.4 4. 79 50.9 -15.0 -15.0 8.77

.286 .395 76.4 . 417 .275 8.94 29.5 4 . 91 -12.3 3.24 33 . 0 -8 . 45 -8.42 5.45

.197 .544 75.8 .422 .270 6.27 22.4 3.88 -8.44 2. 35 24.7 -5.76 - 5.75 3.93
.0999 . 862 74.7 .434 .256 3.41 15.6 3.82 -3.75 1. 61 16.1 -2.13 -2.09 2. 24

.0504 1. 11 71.6 .456 . 231 1. 92 14.7 4.36 -1.11 1.40 12.3 .0850 .0901 1. 32

.0306 1. 22 68.0 .457 .228 1. 29 16.9 5.21 .0379 1. 49 11.2 .582 .586 . 978

.0199 1.28 63.9 .458 .226 . 958 20.1 6.32 1.08 1. 62 12.1 1. 24 1. 25 . 866

.00976 1. 30 57.0 .452 .230 .618 29.8 9.04 2.59 1. 93 15.8 2.63 2.66 .840

47
Calculation No. 12

1.4.-----------------------------------------------------------~

1.2

1.0 1.0

0.8 0.8

0-6
C1 w

0.4

0.2 0.2

0 0
0.005 0.01 0.02 0.05 0.1 0.2
s

100~--------------------------------------------------------~

50

20

10

-' 5
~
~

0.5

0 . 2~~~~--~--~~~~~----~--~~~~~----~--~._~~~
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10
s

48
 Calculation No. 12

20

10

:::;.
~

0.5

50

20

10

0.5

0·2~~~~~~--~~~~~~--~~~~~~~----~~~~~~~
0.005 0.01 0.02 0.1 0.2 0.5 1 2 5 10
s

49
Calculation No. 13

1. Calculation no.: 13 5. Bearing geometry

2. Bearing type: 2lobe 170"

3. L:D ratio: 1.0

4. Special items
Oil groove:
mP:
P:
Calculation method: FDM (matrix-column
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing) Cir-
cumferential direction:
33 (uniform spacing)
for each of upper and
lower pads X

Static Characteristics Spring Coefficients Damping Coefficients

s
€ ~ Qs Qe h<P-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

4.99 .Q272 91.5 .177 .541 99.0 140 77.4 -85.9 -2. 15 282 -106 -107 73.3

2.93 .0459 91.5 .177 .541 58.5 82.7 45.1 -50.0 -1.34 167 -62.6 -62.7 43.2

2.08 .0647 91.5 .177 .540 40.3 58.7 31.2 -35.4 -.823 117 -44.3 -44.5 30.7

• 961 .140 91.1 .180 .537 19.2 27 .3 15. 4 -16.4 -.101 55.2 -20.0 -20.0 14.8

.500 .265 89.9 • 190 .524 10.0 14.6 7.81 -8.61 .424 28.0 -9.58 -9.59 8.00

.305 . 393 87.8 .201 .507 6.42 9.53 5.36 -5.23 . 798 18.1 -5.10 -5.10 5.25

.207 .514 85.8 .218 .476 4.45 7.01 4. 27 -3.59 . 977 13.3 -2.55 -2.51 3.82

.104 • 712 78 . 6 .248 .423 2.39 5.03 3.40 -1.43 1. 22 8. 82 .0753 .0761 2.33

.0496 .885 67.4 .280 .364 1. 35 5.78 3.78 .0881 1. 41 7.60 1.12 1.14 1. 42

.0291 . 968 57.6 .267 .353 . 931 8.13 4.51 .775 1.60 8.44 1.84 1.87 1. 11

.0199 • 990 50.5 .260 . 357 .733 11.0 5.29 1. 52 1. 76 9.48 2.20 2.23 • 978

.00993 1.01 39.0 .229 .383 .509 19.1 7.42 3.20 2.09 11.9 2.63 2.65 .856

50
 Calculation No. 13

1.2

1.0 1.0
goo
0.8 0.8
75°

0. 6 0.6 60°

w -e.
CJ
45°
0.4 0.4
30°

0. 2 0.2
15°

0 0 oo
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 10
s

100~------------------------------------------------------~----~

50

20

10

0 . 2~~~~--~~--~~~~--~~~~~~~~----~_.~~~~~
0.005 0. 01 0.02 0.05 0. 1 0.2 0.5 1 2 5 10
s

51
 Calculation No. 13

....
::<

s
100~----------------------------------------------------------~

10

....
0
2

- Cxy•- Cyx

0.2~~~----~--~~~~~L--L~~~_._L~~~----~~--~~~_u
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10
s

52
 Calculation No. 14

1. Calculation no.: 14 5. Bearing geometry

2. Bearing type: 2lobe 170
3. L :D ratio: 1.0
4. Special items
Oil groove:
mP:
P:
Calculation method: FDM (matrix-column
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing) Cir-
cumferential direction:
33 (uniform spacing)
for each of upper and
lower pads X

Static Characteristics Spring Coefficients Damping Coefficients

s
E: ~ Q, Q, JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx C yy

4.90 .0172 81.2 .218 .379 133 391 117 -198 30.7 609 -197 -197 120

2.92 .0289 81.2 .218 .379 78.8 234 67.2 -118 18.2 361 -118 -117 70.5

1. 97 .0430 81.2 .218 .379 53.2 158 44.6 -79.6 12.3 241 -79.2 -79.2 47.3

• 956 .0898 81.2 . 219 . 378 25.5 77.3 20.9 -39.0 6.07 113 -39.4 -39.2 23.8

.521 . 165 81.0 .219 . 377 14.1 42. 3 12.5 -21.1 3.42 63.5 -20.7 -20.7 12.9

• 305 .277 80.7 .222 .372 8.26 25.6 7.40 -12.3 2.23 38.1 -12.2 -11.8 7.80

.192 .427 80.0 .228 .363 5.33 16.8 5.30 -7.80 1. 71 24.7 -7.00 -6.98 5.15

.100 .655 78.3 .239 .344 2. 93 10.7 3.82 -3 . 97 1. 34 15.0 - 2. 81 -2.79 3.01

.0513 .925 74.0 .256 . 311 1. 66 8.92 3.55 -1.61 1. 24 11. 1 - . 490 -.481 1.80

.0313 1.07 69.1 .263 .292 1.15 9.61 4.02 -.401 1. 36 10.0 . 33C . 345 1. 29

.0202 1.15 64.0 .265 .281 .865 11.7 4.80 .570 1.56 10.3 . 87 . 873 . 995

.0104 1. 15 53.9 .247 . 296 .559 20.2 6.60 2.24 1. 90 11.5 1.55 1.58 • 731

53
Calculation No. 14

1.2 €

1.0 1.0

0.8 0.8

0.6

0.4

0.2

100~------------------------------------------------~------~

10

5
-'
-s.
~

0.5

54
 Calculation No. 14

...,
~

5
...,
0

-Cxy'- Cyx

55
 Calculation No. 15

1. Calculation no.: 15 5. Bearing geometry

2. Bearing type: 2lobe
170°
3. L:D ratio: 1.0
4. Special items
Oil groove: 10° X 2
mP: 3/4
P: 170°
Calculation method: FDM (matrix-column
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing) Cir-
cumferential direction:
33 (uniform spacing)
for each of upper and
lower pads

Static Characteristics Spring Coefficients Damping Coefficients

s
€ ~ Q. Q, !JtP-J Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
5.24 .0128 78.1 . 234 . 296 173 840 144 -371 78.0 1060 -319 -319 178

3.01 .0221 78.1 .234 .2 96 99.5 483 83.3 -213 44.9 610 -184 -184 101

1. 91 .0346 78.1 .234 .296 63 . 5 306 54.9 -135 28.6 391 -117 -116 64.1

. 996 . 0659 78.1 .234 .296 33.2 159 29.5 -70.5 14.9 206 -60.6 60.5 33.4

.503 .128 78.0 .235 .295 16.8 80.9 15 . 5 -35.6 7.61 106 -30.5 -30.4 16.9

. 313 .203 78.0 . 235 .294 10.6 50.9 10.1 -22.1 4.82 65.6 -18.9 -18.9 10.6

.208 . 303 77.8 .236 .292 7. 00 35.1 6. 60 -14.7 3.36 42.5 -12.1 -12.0 7.14

.0973 .584 76.7 .243 . 281 3.41 18.8 4.18 -6.85 1. 95 22.3 -5.30 -5.29 3.63

.0480 • 903 74.7 .252 .263 1.85 13.5 3.68 -3.01 1. 46 14.0 -1.54 -1.52 2.08

.0307 1.11 72.3 .263 .245 1.30 12.8 4.14 -1.53 1.38 12.6 -.299 -.265 1. 52

.0208 1. 22 69.2 .262 .239 . 976 13.8 4.72 -.559 1.45 11.8 .529 .532 1. 21

.00987 1. 34 62.1 .261 .233 .587 21.7 6.88 .209 1. 92 13.2 1.61 1.63 .836

56
 Calculation No. 15
1.4

1.2

1.0 1.0

90"
0.8 0.8

0.6 0.6 60"

a I<) "&.

0.4 0.4

0·2 0· 2

100~----------------------------------------------------------~

50

20

10

0 . 2~~~--~~~~~~~~~--~~~~~~~----~----~~. .~
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10
s

57
Calculation No. 15

....
~

50

20

10

0 5

0.5

58
 Calculation No. 16

1. Calculation no.: 16 5. Bearing geometry

2. Bearing type: 3 lobe L

3. L:D ratio: 0.5

"I
lJ
4. Special items
Oil groove: 20° X 3
mP: 1/2
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 10
(uniform spacing) Cir-
cumferential direction:
60

Static Characteristics Spring Coefficients Damping Coefficients

s
e: rp Qs Qe !JtP-1 Kxx Kxy Kyx K yy Cxx Cxy Cyx Cyy

4.70 .0500 55·7 0.126 0.760 150.5 24.8 31.6 -33.4 22.0 69.8 -0.572 -0.462 65.8

2.33 .100 57·3 0.127 0.760 75.0 13.9 15.8 -17.5 10.1 36.0 -0.844 -1. 36 31.9

1.115 .200 58.5 0.130 0.755 36.5 8.44 7.84 -8.72 4.94 19.2 -0.852 -0.816 15.4

0.695 .300 58.7 0.136 0.750 23.2 6.84 5·70 -5·76 2.30 14.3 -0.220 -0.364 9.26

0.475 .400 59.0 0.143 0.740 16.4 6.28 4.56 -4.03 2.76 11.6 0.109 0.125 6.58

0.33 .500 57·8 0.151 0.7)0 11. 9 6.28 4.24 -3.16 2.04 10.4 0.608 0 .232 4.22

0.233 .600 57.0 0.161 0.715 8.8 6.29 3·92 -2.32 2.10 9.34 0.728 0.740 4.08

0.159 .?00 55.2 0.171 0.705 6.45 7·67 3.88 -1.49 1.88 8.70 0.954 0.982 2.96

0.100 .Boo 52.2 0.185 0.685 4.45 8.76 4.18 -0.604 1.66 8.74 1.24 1. 24 1.99

0.0505 .900 46.6 0.209 0.655 2.65 11.92 5.30 0 .348 1.60 10.I,Q 1.82 1.83 1.23

0.0267 .950 39·7 0.227 o.63o 1.66 17.92 6.10 1.28 1.50 11.20 1.64 1.66 0.690

59
Calculation No. 16

1.0
90°
0.8
75°

0·6
OJ (I)

0.4

200~-------------------------------------------------.

100

60
 Calculation No. 16

50~------------------------------------------------------~

100.---------------------------------------------------~

....
0

61
Calculation No. 17

1. Calculation no.: 17 5. Bearing geometry

2. Bearing type: 3 lobe L
3. L:D ratio: 0.5 I" 'I
4. Special items
Oil groove: 20° x 3
mP: 2/3
Calculation method: FEM
I

ttJ
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 10
(uniform spacing) Cir-
cumferential direction:
60

Static Characteristics Spring Coefficients Damping Coefficients

s
c ¢ Qs Qe !JtP-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

2.16 .0500 40.0 0.155 0.520 96.7 49.5 J6.0 -J9.J 44.7 84.0 -1.J8 -1.25 78.6

1.06 .100 42.0 0.155 0.520 46.7 26.J 17.7 -20.7 21.2 4).8 -1.66 -1.60 J8.7

0.510 .200 44.5 0.156 0.520 2).1 15.J 8.16 -10.5 10.0 22.8 -1.42 -1.40 17.4

0.)22 .)00 45.9 0.158 0.517 14.9 12.4 5.16 -7-59 6.)) 15.9 -1.57 -1.55 11.6

0.224 .400 47.5 0.161 0.513 10.6 10.9 4.59 -5.64 4.89 14.5 -0.9)0 -0.912 7.86

0.161 .500 48.1 0.16) 0.510 7-90 10.8 ).81 -4.74 ).81 12.) -0.921 -0.906 6.84

0.117 .600 48.5 0.167 0.50) 6.0J 10.9 ).60 -).75 J.21 11.J -O.J60 -0.525 4.47

o.o84o -700 48.5 0.171 0.500 4.60 11.7 ).42 -).06 2.84 10.6 -0.265 -0.253 4.05

0.0592 .Boo 48.5 0.175 o.49J 3.50 1) .1 ).54 -2.)0 2.5) 10.2 -0.004 -0.015~ ).27

0.0)79 .900 47.) 0.179 0.48) 2.50 15-7 4.08 -1.)) 2.16 10.5 0.)9) 0.405 2.29

0.0288 -950 4605 0.182 o.4Bo 2.04 17-9 4.59 -0.786 1.97 11.0 0.618 0.6)0 2.00

62
 Calculation No. 17

1.0~------------------------------------------------~

2~--~---L~~~~~----~--~~~~~~----~~--~
0.01 0.2 0.5 2 5
s

63
 50r-------------------------------------------..-----.

5
s

64
 Calculation No. 18

1. Calculation no.: 18 5. Bearing geometry

2. Bearing type: 3 lobe ,. L
3. L: D ratio: 0.5
LJ
4. Special items
Oil groove: 20° x 3
mP: 3/4
Calculation method: FEM I
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 10
(uniform spacing) Cir-
cumferential direction:
-ii"l
60

Static Characteristics Spring Coefficients Damping Coefficients

s
€ rp Q, Q. /Jt/1-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

1.27 .0500 31.0 0.169 o.t.oo 71.3 73-2 )6.1, -lt2.lt 61t.B BB.It -J.It5 -J.Jt 85.6

0.625 .too JJ-7 0.170 o.t.oo J5.J JB.It 18.2 -2J.It )0.9 t.B.o -2.52 -2.ltlt ltlt.O

0.299 .200 )6.6 0.170 0.)98 17.1 21.9 8.52 -12.5 llt.2 26.2 -2.16 -2.12 20.5

0.187 .)00 )8.0 0.171 0.)98 10.8 17.1 5-56 -8.56 8.92 19.2 -1.Bo -1.77 12.2

0.131 .~too t.o.G 0.172 0.)95 7-77 15.) lt.28 -6.88 6.1,8 16.0 -1 . 67 -1.67 9.0it

0.0950 .500 lt2.1 0;173 0.)95 5-85 11t.B ).It? -5.68 5.0lt 1J .Bit -1.52 -1.50 6.92

0.0700 .600 lt).2 0.17lt O.J9J lt.5J 15.8 J.OJ -lt.72 lt.2lt 12.5 -1.17 -1.29 5-20

0.0516 .700 t.t..o 0.176 0.)90 J.5J 16.0 2.9J -lt.Oit J-55 11.8 -0.9ltlt -0.972 lt.28

O.OJ71t .Boo 1,1,.7 0.178 O.J85 2-73 17.8 J.OJ -J.J2 ).10 11.lt -0.688 -0.668 ).56

0.0256 .900 1,1,.8 0.179 O.JBJ 2.06 20.8 J,l,t, -2.1,1 2.70 11.6 -0.298 -0.288 2 .58

o.o2o1t -950 1,1,.7 o.1Bo O.JBo 1.75 2).2 ).86 -1.92 2.1,1,1, 12.0 -0.128 -0.112 2.28

65
Calculation No. 18

l.Or------------------ ----------------------- --,

66
 Calculation No. 18

100~----------------------------------------~

....,
0

2
s

67
Calculation No. 19

1. Calculation no.: 19 5. Bearing geometry

2. Bearing type:
3. L: D ratio:
3 lobe L .,
0.5
II
4. Special items
Oil groove: 20° x 3
mP: 4/5
I
-rt·i
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 10
(uniform spacing) Cir-
cumferential direction:
60

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Q, h<P-l Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

0.856 .0500 27.0 0.178 0.)26 57-2 94.5 )9.6 -46.8 84.0 99.0 -4.16 -4.01 94.0

0 .42 .100 29.1 0.178 0.)26 28.2 50.0 18.4 -24.) 39 .4 50. 5 -).98 - ).90 45-7

0.202 .200 )2.0 0.178 0.)26 13-7 28.4 8.00 -1).9 18.1 27.0 -2. 79 -2-75 22.5

0.128 .)00 )4.4 0.179 0.)26 8.82 21.9 4.75 -9.60 11.7 19-5 -1.91 -1.88 1).2

0.0892 .400 )6.2 0.179 0 .)24 6.)0 19.4 4.05 -7.80 8.20 17-7 -1.96 -1.9) 9-55

0.0652 -500 )8.) 0.180 0.)24 4.76 18. 5 ).64 - 6 .60 6.)5 16.5 -1.72 -1.70 7-35

0.0484 .600 40.0 0.18o O.J22 J.70 18. 7 ).2) -5.65 5-15 15.1 -1. 50 -1.48 5-80

O.OJ57 -700 41.0 0.181 O.J20 2.88 19.8 ).08 -4.76 4.)8 11•• 4 -1.11 -1.09 4.27

0.0262 .Boo 42.1 0.182 0.)18 2.26 21.6 ).58 -4.16 J.76 15.1 -0.790 -0.775 ).64

0.0182 .900 42.6 0.182 0.)18 1.72 25.0 4.14 -J.J9 ).16 15-7 -0.)88 -O.J7J 2.89

0.0151 -950 4).2 0.18) 0.)16 1.50 27.) 4-57 -2.98 2.97 16.2 -0.175 -0.160 2.67

68
 Calculation No. 19

69
 Calculation No. 19

lOOr---------------------------------~~

70
 Calculation No. 20

1. Calculation no.: 20 5. Bearing geometry

2. Bearing type: 4lobe
3. L: D ratio: 0.5
4. Special items
Oil groove: 10° x 4
mP: 1P
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points : Axial direction: 20
(uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)
for each of upper, low-
er, left, and right pads

Static Characteristics Spring Coefficients Damping Coefficients

s Cyx Cyy
e ¢ Qs Q. /J¢- 1 Kxx Kxy Kyx Kyy Cxx Cxy
4.94 .050 60 . 1 . 0833 182. 20. 3 33. 1 -33.7 20.4 69 . 6 . 156 .304 70.5
2.44 .100 59.2 .0848 90.3 10.5 17.5 -17 . 0 10.6 36.7 .522 .600 35.5
1.17 . 200 58.4 .0903 43.9 6 . 16 9.24 -8. 65 6.03 19. 1 1. 10 1. 14 18. 4
.719 .300 56.6 . 100 27.6 5.33 6 . 57 -5.82 4. 63 13.5 1. 28 1. 30 12.5
.478 . 400 54.5 . 114 19 . 0 5 .26 5 . 73 -4.01 4 . 08 11.8 1. 62 1. 64 8.91
.323 .500 52 . 2 . 132 13.4 5 . 68 5 . 12 -3. 2 1 3.74 10.4 1. 80 1. 82 7.43
.215 .600 49.4 . 152 9.48 6. 50 4.87 -2. 18 3. 41 9.78 1. 83 1. 85 5 . 48
. 137 .700 46.4 .173 6 . 55 7. 8 1 4.83 -1 . 26 3. 00 9.65 1. 76 1. 78 3. 78
. 0793 .800 43.0 . 193 4 . 29 9 .86 5.09 -. 666 2 .61 10. 2 1. 94 1. 95 2. 94
.037 6 .900 38 .3 . 2 11 2 .48 14.2 5. 83 .1 56 2. 05 11 .9 2.06 2 .07 1. 78

.01 99 . 950 33. 7 .220 1. 57 20 . 5 6.60 . 994 1. 65 13 .2 ~ 1. 86 1.87 .982

71
Calculation No. 20

72
 Calculation No. 20

Kxx

Kxy

.....
0

73
Calculation No. 21

1. Calculation no.: 21 5. Bearing geometry

2. Bearing type: 4lobe
3. L: D ratio: 0.5
4. Special items
Oil groove: 1oox 4
mP: 2/3
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)
for each of upper, low-
er, left, and right pads

Static Characteristics Spring Coefficients Damping Coefficients

s
c rp Qs Q. fJ'fl-' Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
2. 17 .050 43.6 . 106 112. 43 .8 41.0 -40.9 4 3. 5 87.4 .006 . 189 85.'5
1. 07 . 100 43.3 . 107 56 . 2 22.7 20.9 -20.6 22.0 44.3 . 162 .252 43.0
. 513 . 200 43.0 . 108 27.1 12.8 10.7 -10.1 11. '3 22.8 .297 .345 22.5
. 318 .300 42.5 .111 17. 1 10 .2 7.76 -6.89 7.97 17.0 .540 .573 13.9
.213 .400 41.7 . 114 11.8 9.53 6.23 -5.09 6.33 14. 1 .714 .741 11.0
. 147 . 500 40.8 . 119 8.50 9.79 5.36 -3.91 5.36 12.4 .714 .735 7.69
. 101 . 600 39.8 . 124 6. 19 10.6 4.94 -3. 14 4 . 71 11.5 .906 .924 6.59
.0675 .700 38.7 . 131 4.43 12. 1 4.82 -2.37 4. 16 11.4 .930 .948 5.05
.0418 .800 37.3 . 139 3 . 06 14.7 4.92 -1.80 3.68 11.8 1.06 1. 07 3.97
.0229 .900 35.7 . 148 2.01 19.5 5.57 -1.09 3. 10 13.4 1. 29 1. 31 2.89
.0153 .950 34.5 . 152 1.60 24. 1 6.22 -.444 2. 71 15.0 1. 39 1. 40 1. 97

74
 Calculation No. 21

75
Calculation No. 21

Kxx

Kxy
5
:;::.
::.:::
Kyy
2

1
-Kyx

.....
0

76
 Calculation No. 22

1. Calculation no.: 22 5. Bearing geometry

2. Bearing type: 4lobe
3. L: D ratio: 0.5
4. Special items
Oil groove: 1oo x 4
mP: 3/4
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)
for each of upper, low-
er, left, and right pads

Static Characteristics Spring Coefficients Damping Coefficients

s
e ~ Qs Q, !J<P-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
1. 24 .050 34.4 . 113 81.7 64.6 43.9 -44.2 65.0 98.5 -.0680 . 140 98.4
.614 . 100 34.5 . 113 40.5 33.2 22.4 -22.2 32.6 50.7 .0360 . 140 49.3
.293 .200 34.6 . 114 19.6 18. 7 11.8 -11.3 16.5 26.7 . 136 . 192 24.2
. 182 .300 34.5 . 115 12.3 14 . 7 8. 19 -7.33 11. 1 19. 1 . 0967 . 132 15.3
. 123 .400 34.2 . 116 8.55 13 .5 6. 32 - 5 . 62 8.52 15. 3 . 196 .224 11.3
.085 3 .500 34. 1 . 118 6. 18 13.6 5.36 -4.61 6.92 13 .6 .288 .308 8 . 8q
.0593 .600 34.0 .121 4.51 14. 6 4. 81 -3.53 5.85 12.7 .236 .256 6.38
.0400 .700 33.5 . 123 3.26 16.5 4.53 -3.08 5.04 12.5 . 376 .392 5.32
.0256 .Boo 33.3 .126 2. 31 19.8 4.56 -2.62 4.32 13 .0 .492 .508 4.35
.0147 .900 32.8 .130 1. 57 25.9 5. 14 -1 .74 3 . 61 14.6 . 560 .580 2.70
.0103 .950 32.4 . 131 1. 29 31.4 5.82 -1.3 6 3.07 16 .4 .7 76 .79 2 2.28

77
Calculation No. 22

75°

60°

45°

30°

0.01 0.02 0.05 0.1 0.2 0.5 1 2 5

s

78
 Calculation No. 22

5
:;:.
:.::
Kyy
2

0.005 0.01 0.02 0.05 0.1 0. 2 0.5 2

s

....
c3

79
Calculation No. 23

1. Calculation no.: 23 5. Bearing geometry

2. Bearing type: 4lobe
3. L: D ratio: 0.5
4. Special items
Oil groove: 10° x 4
mP: 4/5
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)
for each of upper, low-
er, left, and lright pads

Static Characteristics Spring Coefficients Damping Coefficients

s h<P-1
€ ~ Qs Q, Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
.819 .050 28.3 . 116 64. 1 68.9 36.5 -3 7 . 9 69 . 0 84 . 1 -. 104 .0760 86.2
.405 . 100 28.8 . 116 32.7 36.3 19.2 -19 . 9 35.4 44 . 5 -.060 .0360 45.0
. 194 .200 29.3 . 117 15.4 19.7 9.86 -10 .0 17.3 23.0 -.004 .0440 21 .8
. 120 .300 29.6 . 117 9.69 15 . 2 6.85 -6 . 57 11 .5 16.8 .004 .0360 13.6
.0813 .400 29.5 .118 6. 73 13.8 5 . 43 -4.94 8.5 7 14.0 -.032 -.004 9.65
.0568 . 500 29.7 . 119 4.87 13.7 4.54 -4. 13 6.82 12.5 .020 .0440 7.54
.0398 .600 30.0 . 120 3.57 14.5 4. 14 -3.50 5.62 12.1 .092 . 112 5.88
.0272 .700 30.3 . 121 2. 61 16.2 3. 91 -2.86 4.75 12. 1 .080 .096 4.24
.0178 .800 30 . 5 .122 1. 87 19.3 3.88 -2.59 4.03 12 . 9 . 196 .212 3.41
.0106 .900 30.8 .124 1. 31 25.2 4. 12 -2. 27 3 . 35 13 . 6 . 304 .324 2.66
.00772 .950 30.9 .124 1. 11 31. 1 4.05 -1 . 92 2.74 13.6 .300 . 320 2.22

80
 Calculation No. 23

81
Calculation No. 23

82
 Calculation No. 24

1. Calculation no.: 24 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 0.5
4. Special items
mP: 0
P: goo
ajp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
X

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Q, JJ.p-t Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
5. 76 .1 0 .459 5. 71 54.3 .901 0 0 .901 10.8 0 0 10.8

2.68 .2 0 .803 5.13 27.4 1.22 0 0 1. 22 6.03 0 0 6.03

1. 59 .3 0 1. 07 4. 59 17.7 1.46 0 0 1.46 4.37 0 0 4.37

1.05 .4 0 1.26 4.09 12.9 1.68 0 0 1.68 3.61 0 0 3.61

• 7 56 .5 0 1.34 3.67 10 . 3 1.94 0 0 1.94 3.33 0 0 3.33

.550 .6 0 1.35 3. 24 8.39 2.26 0 0 2.26 3.21 0 0 3. 21

.381 .7 0 1.39 2.82 6.64 2.65 0 0 2.65 3.06 0 0 3.06

. 259 .a 0 1.39 2.40 5.27 3.15 0 0 3.15 3.01 0 0 3.01

.168 .9 0 1.39 1.99 4.10 3.72 0 0 3.72 2.97 0 0 2.97

.103 1.0 0 1.35 1.59 3.14 4.47 0 0 4.47 2.99 0 0 2.99

.0576 1.1 0 1. 27 1.20 2.34 5.63 0 0 5.63 3.11 0 0 3.11

. 0277 1.2 0 1.14 .817 1. 65 7.83 0 0 7.83 3.05 0 0 3.05

83
 Calculation No. 24

1.4r-----------------------------------------------------,14

Of

50..---------------------------------,

5
s

84
 Calculation No. 24

sor-----------------------------------------------------~

....
~

5
s

5
s

85
Calculation No. 25

1. Calculation no.: 25 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L:D ratio: 0.5
4. Special items
mP: 1/2
P: 80°
rxjp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing) X

Static Characteristics Spring Coefficients Damping Coefficients

s
e ~ Q, Q. ~j•p-t Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
. 976 .1 0 2.78 5.58 34 .5 14.7 0 0 14.7 16 . 4 0 0 16 . 4

.474 .2 0 2. 77 5.59 17.1 8.03 0 0 8.03 8 .49 0 0 8.49

.301 .3 0 2. 75 5.57 11.1 6.13 0 0 6.13 5.96 0 0 5.96

.208 .4 0 2. 75 5. 57 8.02 5. 39 0 0 5.39 4.75 0 0 4 . 75

.150 .5 0 2.72 5. 60 6. 09 5.13 0 0 5.13 4 . 08 0 0 4.08

.109 .6 0 2.68 5.62 4.75 5. 20 0 0 5.20 3. 69 0 0 3.69

.0797 . 7 0 2.62 5.64 3.78 5.48 0 0 5.48 3.48 0 0 3 . 48

. 0571 .8 0 2.52 5.72 3.01 6.01 0 0 6. 01 3 . 40 0 0 3.40

.0398 .9 0 2. 35 5.79 2.40 6 . 85 0 0 6.85 3.47 0 0 3.47

.0264 1.0 0 2.17 5.88 1.89 8.16 0 0 8.16 3.64 0 0 3.64

.0162 1.1 0 1.95 5.99 1.45 10. 3 0 0 10.3 3.95 0 0 3.95

.00868 1.2 0 1.69 6. 10 1.05 14 . 4 0 0 14.4 4.74 0 0 4.74

86
 Calculation No. 25

1.4r---------------------------------------------------~14

5or-----------------------------------------------------~

87
Calculation No. 25

50.------------------------------------------------------.

5
s

50.------------------------------------------------------.

88
 Calculation No. 26

1. Calculation no.: 26 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 0.5
4. Special items
mP: 2/3
P: 80°
ajp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Q, JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
.419 .1 0 2. 7 5 3. 63 22 . 5 25 . 2 0 0 25.2 18.5 0 0 18.5

.204 .2 0 2. 74 3.64 11.2 13 . 4 0 0 13.4 9.51 0 0 9.51

.130 .3 0 2. 71 3.64 7.31 9.87 0 o· 9.87 6.64 0 0 6.64

. 0914 .4 0 2.69 3.65 5.33 8.41 0 0 8 . 41 5.29 0 0 5.29

.0669 . 5 0 2. 65 3.66 4.11 7. 84 0 0 7.84 4.58 0 0 4.58

.0499 .6 0 2.61 3.67 3.27 7.80 0 0 7.80 4.22 0 0 4.22

.0371 . 7 0 2.57 3. 68 2.64 8.14 0 0 8.14 4.03 0 0 4. 03

.0273 .8 0 2.51 3.70 2.15 8.84 0 0 8 . 84 3.99 0 0 3.99

.0195 .9 0 2. 43 3. 72 1. 75 10. 0 0 0 10.0 4. 07 0 0 4.07

.0134 1.0 0 2.35 3.74 1.40 11.9 0 0 11.9 4.26 0 0 4.26

.00848 1.1 0 2.25 3. 77 1.09 15.0 0 0 15.0 4 . 59 0 0 4.59

.004 73 1.2 0 2.16 3. 77 . 805 21.0 0 0 21.0 5.17 0 0 5.17

89
Calculation No. 26

I.4r-----------------------------------------------------~14

0.2

5
s

90
 Calculation No. 26

50~------------------------------------------------------~

5
s

sor-------------------------------------------------------~

"-cxx=Cyy

...,
0

0.02 0.05 0. 1 5
s

91
Calculation No. 27

1. Calculation no.: 27 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 0.5
4. Special items
mP: 3/4
p: 80°
rxfP: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing) X

Static Characteristics Spring Coefficients Damping Coefficients

s
e ~ Qs Q. JJ.p-• Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
.248 .1 0 2.61 2.70 18.0 35 . 5 0 0 35.5 21.0 0 0 21.0

.121 •2 0 2.60 2.70 8.86 18 . 7 0 0 18.7 10.8 0 0 10.8

.0772 .3 0 2.59 2.70 5.82 13 . 6 0 0 13.6 7.56 0 0 7.56

.0544 .4 0 2.56 2. 71 4.25 11.5 0 0 11.5 6.04 0 0 6.04

.0407 .5 0 2.54 2.72 3.29 10.6 0 0 10 . 6 5.24 0 0 5. 24

.0301 .6 0 2. 50 2.73 2.63 10.4 0 0 10.4 4.80 0 0 4.80

.0227 .7 0 2.46 2.73 2.14 10.8 0 0 10.8 4.57 0 0 4. 57

.0168 .8 0 2. 42 2.75 1. 75 11 . 7 0 0 11.7 4.51 0 0 4.51

.0122 .9 0 2.36 2. 76 1.43 13.1 0 0 13.1 4. 57 0 0 4.57

.00850 1.0 0 2.30 2.78 1.15 15.5 0 0 15.5 4.76 0 0 4.76

.00551 1.1 0 2.23 2.80 .905 19.6 0 0 19.6 5.11 0 0 5.11

.00315 1.2 0 2. 14 2.82 . 674 28.2 0 0 28.2 5.75 0 0 5. 75

92
 Calculation No. 27

1.4r-------------------------------------------------------~14

0.8
QJ

0.6

0.4

0.2

5
s

93
Calculation No. 27

sor---------------------------------------------------------,

94
 Calculation No. 28

1. Calculation no.: 28 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 1
4. Special items
mP: 0
p: goo
ajp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing) X

Static Characteristics Spring Coefficients Damping Coefficients

s JJ,p-1 Cyy
€ rfi Q. Q. Kxx Kxy Kyx Kyy Cxx Cxy Cyx

0 .244 5. 78 30.1 .462 0 0 .462 11.7 0 0 11.7

3.19 .1

1.52 •2 0 .413 s. 28 15.6 .832 0 0 .832 6.59 0 0 6.59

.931 .3 0 • 538 4.80 10.4 1.24 0 0 1. 24 4.86 0 0 4.86

.612 .4 0 .638 4.33 7.52 1. 54 0 0 1.54 3.94 0 0 3.94

.442 .5 0 .673 3.89 6.04 1. 79 0 0 1. 79 3.60 0 0 3.60

.321 .6 0 .675 3.46 4.94 2.06 0 0 2.06 3.44 0 0 3.44

.229 •7 0 .689 3.25 4.03 2.41 0 0 2.41 3.31 0 0 3.31

.159 .8 0 .682 2.60 3.26 2.88 0 0 2.88 3.26 0 0 3.26

.105 .9 0 .671 2.17 2.61 3.45 0 0 3.45 3.21 0 0 3.21

.0664 1.0 0 .645 1. 74 2.07 4.21 0 0 4.21 3.22 0 0 3.-22

.0386 1.1 0 .604 1.31 1.61 5,18 0 0 5.18 3. 28 0 0 3.28

.0196 1.2 0 .536 .902 1.21 7.19 0 0 7.19 3.60 0 0 3.60

95
Calculation No. 28

t.4r------------------------------,I4

sor---------------------------~

96
 Calculation No. 28

sor-----------------------------------------------------~

5
s

....
G

5
s

97
Calculation No. 29

1. Calculation no.: 29 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 1.0
4. Special items
mP: 1/2
p: 80°
ajp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing) X

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Q, !JtP-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
.617 .1 0 1.37 5.99 22.1 14.0 0 0 14.0 18.7 0 0 18.7

.300 . 2 0 1.36 6.00 10.9 7.65 0 0 7.65 9.61 0 0 9.61

.192 .3 0 1.36 5.98 7. 17 5.85 0 0 5.85 6.74 0 0 6.74

.134 .4 0 1.34 6.00 5.20 5.14 0 0 5.14 5.32 0 0 5.32

.0973 .5 0 1.32 6.00 4.00 4.93 0 0 4.93 4.54 0 0 4.54

.0718 .6 0 1. 29 6.00 3.16 4.97 0 0 4.97 4.08 0 0 4.08

.0529 . 7 0 1.23 6.01 2.56 5.13 0 0 5.13 3.78 0 0 3.78

.0385 .8 0 1.15 6.05 2.08 5.56 0 0 5.56 3.64 0 0 3. 64

.0275 .9 0 1.06 6.08 1.71 6.32 0 0 6.32 3.64 0 0 3. 64

.0187 1.0 0 .953 6.12 1.39 7.51 0 0 7.51 3.76 0 0 3.76

.0119 1.1 0 .820 6.18 1.11 9.46 0 0 9.46 4.07 0 0 4.07

.OOS69 1.2 0 • 675 6.24 .850 13.2 0 0 13.2 4. 58 0 0 4.58

98
 Calculation No. 29

1.4 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - , 1 4

5
s

99
Calculation No. 29

sor---------------------------------------------------------.

sor-------------------------------------------------------~

100
 Calculation No. 30

1. Calculation no.: 30 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 1.0
4. Special items
mP: 2/3
P: goo
ajp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction : 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing) X

Static Characteristics Spring Coefficients Damping Coefficients

s
€ rp Qs Q. JJ<ft-l Kxx K xy K yx K yy C xx C xy C yx Cyy

.285 .1 0 1.32 3.97 15 . 5 25.2 0 0 25.2 21.3 0 0 21.3

.139 . 2 0 1.32 3.97 7.69 l3 .3 0 0 13.3 10.9 0 0 10 . 9

.0884 .3 0 1.31 3.97 5.05 9. 63 0 0 9. 63 7.51 0 0 7.51

. 0624 .4 0 1.30 3.97 3.7 2 8.07 0 0 8 . 07 5.91 0 0 5.91

.0461 .5 0 1. 28 3.98 2. 90 7. 45 0 0 7.45 5.04 0 0 5.04

.0347 .6 0 1.26 3.98 2.34 7.37 0 0 7.37 4.56 0 0 4.56

.0262 . 7 0 1.24 3. 98 1.92 7 .63 0 0 7.63 4.29 0 0 4. 29

. 0196 .8 0 l. 22 3.99 1.60 8 . 25 0 0 8.25 4 . 20 0 0 4. 20

.0143 .9 0 1.18 4 . 01 1.33 9. 28 0 0 9.28 4. 25 0 0 4 .25

.0100 1.0 0 1.14 4. 02 1.10 10.9 0 0 10.9 4.42 0 0 4.42

.00659 1.1 0 1.10 4.03 .888 13.7 0 0 l3. 7 4 .73 0 0 4.73

. 00383 1.2 0 1.05 4.05 . 684 19.5 0 0 19.5 5.27 0 0 5.27

101
Calculation No. 30

1.4r---------- --------------- --------------- --------------- .14

sor-------------------------------------------------------

5
s

102
 Calculation No. 30

sor-------------------------------------------------------~

5
s

50

....,
G

5
s

103
Calculation No. 31

1. Calculation no.: 31 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L: D ratio: 1.0
4. Special items
mP: 3/4
p: 80°
r:t../P: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing) X

Static Character istics Spring Coefficients Damping Coefficients

s ljr/J-1
€ ¢ Qs Q, Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

.174 .1 0 1.24 2.96 12 .9 34 . 9 0 0 34 . 9 23.7 0 0 23.7

. 0849 .2 0 l. 24 2.96 6.39 18.3 0 0 18.3 12.1 0 0 12.1

.0545 .3 0 1.23 2.96 4.21 13.2 0 0 13.2 8.38 0 0 8.38

.0388 .4 0 1.22 2. 97 3.12 11.1 0 0 11.1 6.64 0 0 6.64

.0288 .5 0 1.21 2.98 2.44 10.2 0 0 10.2 5.69 0 0 5. 69

.0219 .6 0 1.20 2.98 1.98 9.93 0 0 9.93 5 . 16 0 0 5.16

.0167 . 7 0 1.18 2.99 1.63 10 . 2 0 0 10.2 4.88 0 0 4.88

.0126 .8 0 1.15 3.00 1.36 10.9 0 0 10. 9 4. 76 0 0 4. 76

.00933 .9 0 1.13 3.01 1.14 12.1 0 0 12.1 4.79 0 0 4.79

.00664 1.0 0 1.10 3.03 .942 14 .3 0 0 14.3 4.94 0 0 4.94

.00443 1.1 0 1.06 3.05 .763 18.1 0 0 18.1 5.25 0 0 5.25

.00262 1.2 0 1.02 3.07 .589 26.0 0 0 26 . 0 5. 84 0 0 5.84

104
 Calculation No. 31

1.4~------------------------------------------------------~14

Qt

5or-------------------------------------------------------~

-
I
-s. 5
~

0.5
0.005 0.05 0. 1 0.2 0.5 1 2 5
s

105
Calculation No. 31

50~--------------------------------------------------------,

5
s

sor----------------------------------------- ----------------,

5
s

106
 Calculation No. 32

1. Calculation no.: 32 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 0
{J: goo
rxj{J: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points : Axial direction : 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)

Static Charact eristics Spring Coefficients Damping Coefficients

s JJ,p-1
€ ~ Qs Q, Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
5.96 .05 0 .203 9.19 79.8 .524 0 0 .249 20.8 0 0 37.4

3.25 .1 0 • 262 9. 02 44.3 .868 0 0 .136 12 . 7 0 0 20.4

0 . 363 8.67 20.3 1.67 0 0 .0595 7.18 0 0 8.93

1.42 •2

.3 0 .430 8 . 34 12.0 2.25 0 0 .0332 5.39 0 0 4.99

.795

0 .445 8.03 8.18 2. 77 0 0 .0212 4.93 0 0 3 . 18

.507 .4

0 .448 7.73 5.57 3.45 0 0 .0132 4.67 0 0 1.98

.316 .5

. 185 .6 0 .443 7.43 3.69 4 .43 0 0 ,00773 4.57 0 0 1.16

.7 0 .428 7.12 2.36 5.83 0 0 .00413 4.55 0 0 . 619

.0987

.0449 .8 0 .397 6.82 1.44 7.99 0 0 , 00188 4 .73 0 0 .282

,0272 0 .371 6.68 1.09 10.3 0 0 .00114 5.10 0 0 .171

,85

.0143 .9 0 . 335 6,54 .780 14.5 0 0 .000597 5.86 0 0 .896

0 .287 6.41 .484 27.5 0 0 .000221 7,35 0 0 .0332

.0052\ .95

107
Calculation No. 32

1.4.---------------------------------------------------------.14

5
s

108
 Calculation No. 32

5Qr-------------------------------------- ----------------.

:::-
:::<
5
Kxx

5
s

sor---------------------------------------------------------,

.....
c3

5
s

109
Calculation No. 33

1. Calculation no.: 33 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 1/2
P: s~
afP: o.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction : 16
(per pad) (uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s e Qs Q, /Jcp-1
rfi Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

9.73 .0100 0 . 449 343 34 .2 0 0 34.3 40. 5 0 0 40 . 5

4.84 .0200 0 ,449 171 17.1 0 0 17.1 20 .2 0 0 20 .2

3.21 .0300 0 .449 114 11.4 0 0 11.3 13 . 4 0 0 13.4

1.91 .0500 0 .449 67.6 6.82 0 0 6.73 8.03 a a 7. 96

.934 .100 a . 449 33.2 3.52 a 0 3.29 4.04 a 0 3.89

.448 .200 0 .446 16.1 2.13 0 0 1.58 2.18 a a 1.86

.258 .300 0 .441 9.62 1.67 0 0 .910 1.54 0 0 LOB

.169 .400 0 .435 6.57 l. 79 0 0 .597 1.34 a 0 .706

.103 . 500 0 .425 4.28 1.81 a 0 . 363 1.16 a 0 .429

.0673 .600 0 .411 3.05 2.30 0. a .237 1.26 a 0 .280

.0386 .700 0 .397 2.00 2. 87 0 0 .136 1.35 a 0 .161

.0190 .800 0 .370 1.20 4.02 0 0 .0667 1.55 a 0 .0789

.00643 .900 0 .328 . 596 7.19 0 a .0227 2.06 a 0 .0267

.00346 .925 0 .314 .398 6.83 0 0 .0122 1.83 0 0 .0144

110
 Calculation No. 33

1.0.-------------------------------------------------------~

¢= 0

t
g_l:-oo~s:-~-.~.-:!-l::-=-----=~--L---L..,...i,~u..Jo.J..-1___jo.L2---lL.-.J.....J-L....J....J....J.L____:r::::r:::::t::±s::::~:::t::t::.J...ll
s

111
Calculation No. 33

....
0

0.5

0 . 005 0.01 0. 1 0.2

s

112
 Calculation No. 34

1. Calculation no.: 34 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 2/3
p: 80°
rxfP: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 16
(per pad) (uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s e ¢ Qs !J<P - 1
Q. Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

4.27 .0100 0 .443 230 26.1 0 0 26.1 20.6 0 0 20.6

2.14 .0200 0 .443 115 13.1 0 0 13.0 10.3 0 0 10.3

1.42 .0300 0 .443 76.6 8. 77 0 0 8.69 6.89 0 0 6.87

. 854 .0500 0 .443 46.0 5.34 0 0 5.21 4.16 0 0 4.12

.427 .100 0 .443 23.0 2.87 0 0 2.61 2.15 0 0 2.06

.182 .200 0 .441 9.97 1.47 0 0 1.11 1.02 0 0 .876

.112 .300 0 .438 6.30 1.19 0 0 .681 .755 0 0 .538

.0772 .400 0 .434 4.52 1. 24 0 0 .471 .703 0 0 .372

.0496 .500 0 .428 3.08 1.28 0 0 .303 .650 0 0 .239

.0315 . 600 0 .421 2.13 1.41 0 0 .192 . 639 0 0 .152

.0129 . 750 0 . 407 1.08 1.82 0 0 .0787 .665 0 0 .062

.00895 .800 0 .401 .831 2. 08 0 0 .0546 .699 0 0 .043

.00647 .850 0 .394 .676 3.22 0 0 .0395 .922 0 0 .031

.00293 .900 0 .386 .396 3.24 0 0 .0178 .844 0 0 .014

113
 Calculation No. 34

1.0
goo

75°

0.6
cy -e.
(<) Q,

~= 0

114
 Calculation No. 34

10
s

115
Calculation No. 35

1. Calculation no.: 35 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 3/4
P: soo
rxfP: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 16
(per pad) (uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s lj¢-1
€ ~ Q. Q, K xx Kxy K yx Kyy C xx Cxy Cyx C yy

2.30 . 0100 0 .421 167 19. 9 0 0 19.9 12.1 0 0 2. 1

1.14 .0200 0 .421 83. 5 9.97 0 0 9.95 6.05 0 0 6.04

.768 .0300 0 .421 55.7 6. 67 0 0 6.63 4.04 0 0 4.03

.461 .0500 0 .421 33.4 4 .04 0 0 3.98 2.44 0 0 2.42

.230 .100 0 .420 16.7 2.11 0 0 1.99 1.26 0 0 1.21

.115 .200 0 ,419 8.45 1.27 0 0 .995 .720 0 0 .604

. 0686 .300 0 .416 5.17 .978 0 0 .593 .521 0 0 .360

.0466 .400 0 .413 3.63 . 933 0 0 .402 .456 0 0 .244

.0298 .500 0 .409 2.45 .907 0 0 .257 . 404 0 0 .156

.0202 .600 0 .403 1.78 1.09 0 0 .174 .427 0 0 .106

. 0109 . 700 0 .396 1.09 1.10 0 0 . 094( .388 0 0 .0571

.00833 • 750 0 .392 .898 1.28 0 0 .071~ . 418 0 0 .0436

• 00654 .800 0 .388 .761 1.83 0 0 .056J .526 0 0 .0322

116
 Calculation No. 35

1.0
90°

75°

0.6
o- -e.
w 45°
Q.

~= 0

~. LOO...I.5-L...I.~-:----:...l...:--'---..JW-LL.i..OU.-l--OL.2-.1..-~0L.5::i:I:J::I:t::==:t:2=::1--L.J.5...L.L..L.Jlt
s

0.2~~~~~----~--~~~~~-----'----L-~~~~----~--~~~~~
0. 005 0.01 0.02 0.05 0.1 0.2 1 10
s

117
Calculation No. 35

...,
:<

10
s

...,
G

0.005 0.01 0. 05 10
s

118
 Calculation No. 36

1. Calculation no.: 36 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: 0
P: 8~
aJP: o.s
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 12
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)

Static Char act eris tics Spring Coefficients Damping Coeffici ents
s JJ.p-1
€ ¢ Qs Q, K xx K xy K yx K yy Cxx C xy Cyx C yy

.05 0 .364 9.12 146. 1.12 0 0 .655 19.4 0 0 34. 4

10.9

5.70 .1 0 .483 8,92 77.7 1.48 0 0 .343 11,5 0 0 18.0

.698 8.53 34.8 2.01 0 0 .147 6. 48 0 0 7.73

2.44 .2 0

,826 8.18 20,6 2.45 0 0 .0824 4.98 0 0 4.33

1. 37 .3 0

.861 7.87 13.9 3. 04 0 0 ,0521 4,57 0 0 2.74

.865 .4 0

7.58 9.22 3.80 0 0 . 0316 4.31 0 0 1.66

.525 .5 0 .879

0 ,880 7. 29 5. 92 4. 83 0 0 ,0180 4. 22 0 0 .944

.298 .6

7.00 3. 62 6.19 0 0 • 00921 4. 22 0 0 . 484

.153 •7 0 .860

2.07 8. 71 0 0 .00398 4.52 0 0 .209

.0661 .8 0 .BOO 6.73

.750 6.59 1.49 11. 2 0 0 .00232 4.97 0 0 .122

.0386 .85 0

.682 6.46 1.01 16.0 0 0 .00115 5. 72 0 0 .0606

.0192 .9 0

.572 29.8 0 0 .000395 7. 26 0 0 ,00207

.00655 .95 0 .587 6.34

119
 Calculation No. 36

1.4~----------------------------------------------------~14

50

5
s

120
 Calculation No. 36

sor---------------------------------------------------------,

....,
:<

5
s

sor---------------------------------------------------------,

....,
G

5
s

121
Calculation No. 37

1. Calculation no.: 37 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: 1P
p: soo
rxJP: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 16
(per pad) (uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s e t/1 Qs Qe /J'r' Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

6.13 .0100 0 .226 219 33.3 0 0 ~3.3 46.2 0 0 46.2

3.05 .0200 0 .226 109 16.6 0 0 16.6 23.0 0 0 23.0

2.03 .0300 0 .226 72.3 11.0 0 0 ll.O 15.3 0 0 15.3
1.21 .0500 0 .226 43.1 6.63 0 0 6.56 9.15 0 0 9.08
.591 .100 0 .225 21.2 3.41 0 0 3.21 4.60 0 0 4.45
.284 .200 0 .224 10.4 2.03 0 0 1.54 2.46 0 0 2.14

.166 .300 0 .221 6.26 1.61 0 0 .904 1.71 0 0 1.25

.110 .400 0 .218 4.30 1.66 0 0 .599 1.45 0 0 .830

.0687 .500 0 .213 2. 91 1.54 0 0 .373 1.28 0 0 .517

.0457 .600 0 .206 2.12 2.11 0 0 .248 1.33 0 0 .344

.0272 • 700 0 .199 1.45 2.66 0 0 .148 1.39 0 0 .205

.0194 .750 0 .193 1.15 2.95 0 0 .106 1.41 0 0 .146

.0141 .800 0 .185 .929 3.67 0 0 .0764 1.60 0 0 .106

.00515 .900 0 .164 .502 6.69 0 0 .0279 2.09 0 0 .038

122
 Calculation No. 37

1.0
go•
0.8
75°

0.6
a -e.
(<)

0.4

~.~0~05~~~~~~-J~~UU~--~O.L2~~~~~~~==~z==~~±5~~
s

123
Calculation No. 37

10
s

124
 Calculation No. 38

1. Calculation no.: 38 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: 2/3
P: 8~
afP: o.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 16
(per pad) (uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s € ~ Q. Qe fJ'p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
2.87 .0100 0 .217 157 26.1 0 0 26.1 23.3 0 0 ~3.3

1.43 .0200 0 .217 78.6 13.1 0 0 13.1 11.7 0 0 1.7

.956 .0300 0 .217 52.4 8. 76 0 0 8.70 7. 80 0 0 7.77

.573 .0500 0 .217 31.4 5.31 0 0 5. 22 4.70 0 0 4.66

.287 .100 0 .217 15.8 2.80 0 0 2.61 2.41 0 0 2.33

.124 .200 0 .216 6.93 1.34 0 0 1.13 1.16 0 0 1.01

.0765 .300 0 .215 4.42 1.06 0 0 .697 .855 0 0 .622

.0533 .400 0 .214 3.19 1.16 0 0 .485 .761 0 0 .433

.0350 .sao 0 .211 2.24 1.20 0 0 .318 . 694 0 0 .284

.0228 .600 0 .207 1.59 1.34 0 0 .208 .670 0 0 .186

.0145 .700 0 .203 1.14 1.69 0 0 .132 .743 0 0 .118

.00995 .750 0 .200 .864 1.71 0 0 .0906 .705 0 0 .0809

.00706 .800 0 .197 .686 1.96 0 0 .0643 .741 0 0 .0574

.00520 .850 0 .194 .569 3.00 0 0 .0473 • 942 0 0 .0422

125
 Calculation No. 38

l.Or--------------------------------------------
-----------------~

10
0.4

0. 2

126
 Calculation No. 38

10
s

127
 Calculation No. 39

1. Calculation no.: 39 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: 3/4
p: 8~
afp: o.s
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 16
(per pad) (uniform spacing) Cir-
cumferential direction:
36 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s € ¢ Q. Qe JJ.p-• K xx Kxy K yx Kyy Cxx Cxy Cyx C yy

1.62 .0100 0 .205 121 17.8 0 0 17.7 13.9 0 0 13.9

.811 .0200 0 ,205 60.4 8.90 0 0 8.87 6.99 0 0 6.98

.541 .0300 0 .205 40.3 5.95 0 0 5.91 4.67 0 0 4.66

.324 .0500 0 .205 24. 2 3.62 0 0 3.55 2.81 0 0 2.79

.162 .100 0 .205 12.1 1.91 0 0 1. 77 1.44 0 0 1.40

.0811 .200 0 .204 6.12 1.22 0 0 .887 .795 0 0 .698

.0490 .300 0 .203 3.80 .942 0 0 .536 .572 0 0 .422

.0336 .400 0 . 202 2.71 .892 0 0 .367 .489 0 0 .289

.0220 .500 0 .199 1.88 .873 0 0 .240 .430 0 0 .189

.0152 .600 0 .197 1.39 1.01 0 0 .166 .451 0 0 .130

.00855 .700 0 .193 .894 1.04 0 0 .0934 .415 0 0 .073~

.00663 • 750 0 .192 .748 1.21 0 0 .0725 .444 0 0 .05711

.00526 .800 0 .189 .643 1.71 0 0 .0575 .540 0 0 .045

. 00293 .850 0 .187 .426 1.66 0 0 .0320 .486 0 0 .0252

128
 Calculation No. 39

1.0~--------------------------------------------------------------,

0.8

0.6

w
0.4

0.2

129
Calculation No. 39

50

130
 Calculation No. 40

1. Calculation no.: 40 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L: D ratio: 0.5
4. Special items
mP: 0
p: 60°
li../P: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
60 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s f;<P-1
€ rp Qs Q, Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
11.4 .D200 0 .0381 2.46 178 1.47 0 0 .775 16.4 0 0 8.65

6.89 .0500 0 .0590 2.43 103 1.51 0 0 .796 10.8 0 0 5.71

4.45 .100 0 .0812 2.38 65.2 1.56 0 0 .825 7.63 0 0 4.03

2.67 .200 0 .106 2.30 43.9 1.80 0 0 .948 5.70 0 0 3.01

1.80 .300 0 .120 2.20 30.8 1.95 0 0 1.03 4.76 0 0 2.51

1.23 .400 0 .126 2.13 22.4 2.19 0 0 1.16 4.25 0 0 2.24

.837 .500 0 .130 2.05 16.4 2.55 0 0 1.35 3.97 0 0 2.10

.559 .600 0 .131 1.97 11 .9 3.02 0 0 1.60 3.83 0 0 2.02

.352 .700 0 .129 1.89 8.27 3.64 0 0 1.92 3.76 0 0 1.99

.212 .800 0 .125 1.82 5.67 4.45 0 0 2.35 3.70 0 0 1.95

. 114 .900 0 .118 1.74 3.72 5.71 0 0 3.02 3.81 0 0 2.01

.0773 .950 0 .113 1.70 2.88 6.73 0 0 3.55 3.84 0 0 2.03

.0515 1.00 0 .107 1.66 2.19 8.02 0 0 4.23 4.00 0 0 2.11

.0304 1.05 0 .0998 1.63 1.66 10.0 0 0 5.29 4.40 0 0 2.32

131
Calculation No. 40

goo

75°

60°
-e.
45°

':E 0.2
s

0.2 2 5 10
s

132
 Calculation No. 40

0.02
s

100-------------------------------------------------------------,

50

20

133
Calculation No. 41

1. Calculation no.: 41 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L:D ratio: 0.5
4. Special items
mP: 1/2
P: 600
ajp: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
60 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
e </> Qs Qc !Jt/1-' Kxx Kxy Kyx K yy Cx x Cxy Cy x Cyy

11.8 .0100 0 .327 1.13 372 137 0 0 137 167 0 0 167

5.90 .0200 0 .327 1.13 186 68.5 0 0 68.5 83.4 0 0 83 .5

2.42 .0500 0 .326 1.13 76.0 28.3 0 0 28 .1 34.3 0 0 34.2

1.23 .100 0 .326 1.13 38.4 15.0 0 0 14.4 17.7 0 0 17.4

.577 .200 0 .324 1.13 18.4 8.44 0 0 7.38 9.32 0 0 8.61

.349 .300 0 .321 1.13 11 .5 6 .65 0 0 5.17 6.56 0 0 5.61

.234 .400 0 .317 1.14 8.02 6.11 0 0 4.25 5.35 0 0 4.19

.168 .500 0 .311 1.14 5.95 6. 12 0 0 3.90 4.84 0 0 3.48

.109 .600 0 .303 1.15 4.27 6.65 0 0 3.91 4.37 0 0 2.89

.0670 .700 0 .292 1.15 3.07 7.55 0 0 4.22 4.22 0 0 2.58

.0421 .800 0 .279 1.15 2.21 8.98 0 0 4.87 4.33 0 0 2.49

.0268 .900 0 .258 1.16 1.61 10.9 0 0 5.83 4.64 0 0 2.57

.0148 1.00 0 .215 1.18 1.20 13.4 0 0 7.12 5.21 0 0 2.82

134
 Calculation No. 41

2.0 1.0
1.8 go•
1.6 0.8
75°
1.4
1.2 0.6 60°
(::§ 1.0 (<) -e.
45°
0.8 0. 4
0.6 30°
0.4 0.2
15°
0.2
0 0 o·
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10
s

-
I
~ 5
~

0.5

0.2 0.5 1 2 5 10
s

135
Calculation No. 41

50

20

10

...,
~ 5

0.5

0.005 0.01 0. 02 0.05 0.1 0.2 0. 5 1 2 5 10

s

...,
c.> 5

Cyy

1 Cxy= 0
Cyx= 0

0.5

0.005 0.01 0. 02 0.05 0.1 0.2 0.5 1 2 5 10

s

136
 Calculation No. 42

1. Calculation no.: 42 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L:D ratio: 0.5
4. Special items
mP: 2/3
P: 6~
y
afP: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
60 (uniform spacing)

Static Characteristics Spring Coefficients Damping Coefficients

s
€ ¢ Qs Q, !J </J- I Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

10.2 .00500 0 .310 .745 487 500 0 0 500 398 0 0 398

5.10 .0100 0 .310 .745 244 250 0 0 250 199 0 0 199

2.55 .0200 0 .310 .745 122 125 0 0 125 99.5 0 0 99.5

1.21 .0500 0 .310 .745 48.6 50.2 0 0 50.1 39.8 0 0 39.8

.534 .100 0 .310 .745 25.3 26.9 0 0 26.3 21.2 0 0 20.9

.236 .200 0 .308 .743 11.4 13.9 0 0 12.4 10.3 0 0 9.53

.145 .300 0 .306 .741 7.18 10.8 0 0 8.67 7.22 0 0 6.26

.0928 .400 0 .304 .740 4.88 9.62 0 0 6.90 5.76 0 0 4.58

.0670 .500 0 .301 .737 3.64 9.57 0 0 6.28 5.18 0 0 3.80

.0425 .600 0 .296 .738 2.64 10.1 0 0 6.07 4.91 0 0 3.24

.Q305 .700 0 .292 .741 2.06 11.1 0 0 6.35 4.92 0 0 3.04

.0204 .BOO 0 .286 .744 1.60 12.5 0 0 6.88 5. 12 0 0 2.98

.0132 .900 0 .279 .747 1.24 14.7 0 0 7.93 5.50 0 0 3.08

137
 Calculation No. 42

0 . 005 0.01 0.02 0.05 0.1 0. 2 0.5 1 2 5 10

s

138
 Calculation No. 42

100~--------------------------------------------------~-----------,

50

20

10

0.5

0.005 0.01 0.02 0.05 0.1 0.2 1 2 5 10

s

100~--------------------------------------------------~---------,

10

Cxx
::;:.
Cj
5
Cyy

I
Cxy= 0
Cyx= 0

0.005 0.01 0.02 0. 05 0.1 0.2 5 10

s

139
Calculation No. 43

1. Calculation no.: 43 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L: D ratio: 0.5
4. Special items
mP: 3/4
P: 600
rxjp: 0.5 y
Calculation method: FD M
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
60 (uniform spacing)

Stat~c Characteristics Spring Coefficients Damping Coefficients

s
€ ¢ Q. Q, !J<P-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

16. 1 .00200 0 .289 .547 1020 2010 0 0 2010 1230 0 0 1230

6.42 .00500 0 .289 .547 406 803 0 0 803 490 0 0 490

3.21 .0100 0 .289 .547 203 401 0 0 401 245 0 0 245

1.57 .D200 0 .289 .547 99.2 196 0 0 196 120 0 0 120

.515 .0500 0 .289 .547 33.0 66. 1 0 0 65.2 40.2 0 0 39.8

.243 .100 0 .289 .547 15.8 32.8 0 0 31.6 19.7 0 0 19.2

.120 .200 0 .288 .547 7.85 18.3 0 0 16.5 10.5 0 0 9.85

.0763 .300 0 .287 .546 5.15 14.5 0 0 11.7 7.76 0 0 6.72

.0484 .400 0 .285 .548 3.52 12.9 0 0 9.18 6.33 0 0 4.93

.0356 .500 0 .283 .549 2.71 12.7 0 0 8.31 5.82 0 0 4.17

.0264 .600 0 .280 .550 2.12 13.0 0 0 7.95 5.59 0 0 3.72

.0187 .700 0 .277 .553 1.69 13.8 0 0 8.00 5.60 0 0 3.48

.0128 .800 0 .272 .555 1.32 15.7 0 0 8.78 5.80 0 0 3.41

140
 Calculation No. 43

1.0 1.0
go•
0.8 0.8
75°

0.6 0.6 60°

(<)
~
"S-
45°
0.4 0.4
Qs
Jo•
0.2 0.2
15•

0 0 o•
0.005 0.01 0.02 0.05 0. 1 0.2 0.5 1 2 5 10
s

0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10

s

141
Calculation No. 43

500

200

100

50
~

10

0. 005 0. 01 0.02 0.05 0.1 0.2 0.5 1 2 5 10

s

500

200

100

0
·~

50

Cxy= 0
20 Cyx= 0

10

Cxx
5
Cyy

2
0.005 0.01 0.02 0. 05 0.1 0.2 0.5 1 2 5 10
s

142
 Calculation No. 44

1. Calculation no.: 44 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L: D ratio: 1.0
4. Special items
mP: 0
P: 600
ajp: 0.5 y
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction :
30 (uniform spacing)
N
Definition of Q1N: QIN = L {Qin)(NDLCP)} X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
€ QIN !JI/J-i K xx Kyy Cxx Cyy

2.73 . 124 7.78 45.1 l. 31 0.974 11.6 11. 6

l. 52 .247 7.62 25.5 l. 59 0.842 6.90 6.70

.935 . 371 7.47 16 .1 l. 9 7 l. 04 4.99 4.51

.579 .494 7.30 10.5 2.42 l. 28 4.36 3.34

.350 .618 7.13 6.91 3 . 00 l. 58 4.09 2.69

.199 . 742 6.96 4.47 3.79 2.00 4.00 2.35

.102 .865 6.78 2.80 4.99 2.64 4.03 2.22

. 0429 .989 6.60 l. 66 7. 2 2 3.81 4.15 2.22

. 0245 1.051 6.50 1.12 8.97 4.78 4.58 2.45

. 0119 1.112 6.41 .886 13.2 6.95 5 .25 2 . 78

.0039 2 1.174 6.34 .557 2 4.4 12.9 6.02 3.18

143
Calculation No. 44

7.5
1.2

7.0

0.4

6.5
0.2

o1 0
0.2 2 5

so.-------------------------------------------------------~--~

20

10

5
-
I
~
~

0. 5

144
 Calculation No. 44

50~------------------------------------------------------------~

20

10

5
...,
:<

0.5

0.2~~~~~~~----~--~~~~~U-----~--L-~~~~L-----L-~~~
0.002 0.005 0. 01 0.02 0. 05 0. 1 0. 2 0. 5 1 2 5
s

50r-----------------------------------------------------------~

20

Cxx

0.5

145
 Calculation No. 45

1. Calculation no.: 45 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L: D ratio: 1.0
4. Special items
mP: Q5
P: 600
r:t/P: o.5 y
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q1N: QIN = L {Q;n)(NDLCP)}
X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
e QIN !Jtf - I Kxx Kyy Cxx Cyy

.673 .124 4.27 22.5 12.0 11.3 15.5 15.0

.314 .247 4.26 10.8 7.12 5. 92 8.29 7.45

.186 . 3 71 4.26 6.64 6.00 4.34 6.09 4.94

.117 .494 4.25 4.48 5.95 3.81 5.13 3.74

.0735 .618 4.24 3.06 6.44 3.78 4.63 3.05

.0446 • 742 4.22 2.18 7.55 4.19 4.46 2.70

.0249 .865 4.21 l. 38 9.23 4.98 4. 77 2.70

.0120 .989 4.18 l. 00 13.3 7.05 5.31 2.89

.00740 1.051 4.16 .825 18 . 3 9.15 5. 98 3.03

.00393 1.112 4.14 .593 24.4 12.9 6. 04 3.21

146
 Calculation No. 45

1.2.---------------------------------------------~

1.0

4.5 0.8

w 0.6

0.4

0.2

oL OL-~~~L---~~~-LLLLU----L--L-L~_u_u
0.005 0 . 02 0. 05 0. 2 0. 5 1
s

50,_----------------------------------------~

10

5
-I
-s.
~

0.5

0.005 0.01 0.02 0.05 0.1

s

147
Calculation No. 45

20

10

0.5

0.005 0.01 0.02 0.05 0. 1 0.2 0.5 1

s

50~----------------------------------------------~

20

10

....
G 5

0. 5

148
 Calculation No. 46

1. Calculation no.: 46 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L:D ratio: 1.0
4. Special items
mP: 0.667
P: 6~
ajp: 0.5 y
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction : 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q 1N: QIN = L {Q;n)(NDLCP)}
X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
€ Q IN lj<P-1 Kxx Kyy Cxx Cyy

. 288 . 124 2.96 14.4 20.6 19.7 17.7 17.2

.134 . 2 47 2.96 6.90 11 . 7 9.89 9.27 8.27

.0802 . 3 71 2.95 4.31 9.55 7.21 6.69 5.56

.0515 .494 2.95 3.00 9.18 6.13 5.60 4.19

.0333 .618 2.95 2.00 9.48 5 . 76 5.28 3 . 54

.0209 .74 2 2. 94 l. 46 10.8 6.14 5.29 3 . 24

.0123 .865 2 . 94 1.12 13.7 7.44 5.51 3.15

.00620 .989 2.94 . 809 19.4 10.4 5.87 3.22

.00396 1. 051 2.95 .624 24.6 13 . 1 6.10 3.29

149
Calculation No. 46

1.0r-----~----------------------------------------,

3.0 0.8

z
0.6
C1 1<.1

0.4

2.5
0.2

ol 0
0.005 0.01 0. 02 0.05 0. 1 0.2 0. 5 1
s

50r---------------------------------------;

20

10

0.5

150
 Calculation No. 46

50r-----------------------------------------~

0.02 0. 05 0.2 0.5

s

50r-----------------------------------------~

20

10

·~
5
0

0. 005 0. 02 0.05 0.1 0.2 0.5

s

151
Calculation No. 47

1. Calculation no.: 47 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LBP
3. L: D ratio: 1.0
4. Special items
mP: 0.75
P: 600
ajp: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q,N: QIN = L {Q;n)(NDLCP)}
X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
€ QIN ljt/J-1 Kxx Kyy Cxx Cyy

.163 .124 2.29 10.9 28.7 27.6 19.2 18.7

• 0770 .247 2.29 5.36 16.3 14.2 10.1 9.29

• 0465 • 371 2.29 3.23 12.8 9. 92 7.50 6.24

.0302 .494 2.29 2.21 12.0 8.21 6.43 4.81

. 0199 . 618 2.29 1. 65 12.5 7.78 5.99 4.05

. 0128 . 742 2.29 1. 28 14.4 8.29 5.88 3. 64

.00769 .865 2.29 .973 18.1 9.91 5.97 3.46

.00400 .989 2.29 .660 24.9 13.3 6.19 3.42

152
 Calculation No. 47

1.0

2.5 0.8

0.6
!':
cy w
0. 4

2 0.2

0.05 0.1 0.2 0.5

s

50.---------------------------------------

20

5
'7
-s.
~

0.5

153
 Calculation No. 47

50r---------------------------------------- -.

0.5

50

' 20

10

5
::::;-
u

0.5

154
 Calculation No. 48

1. Calculation no.: 48 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 0
P: 60° y
ajp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 14
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
X

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Qe h<P-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

8.69 .050 0 88.3 .943 0 0 .678 11.6 0 0 16.6

5.58 .100 0 58.4 1 .1 9 0 0 .700 8.27 0 0 11.0

3.1 5 ,200 0 35.1 1.67 0 0 .694 5o99 0 0 6.63

1. 97 .300 0 23.5 2.16 0 0 .651 5.08 0 0 4o45

1.24 .400 0 16.2 2.76 0 0 ·577 4.64 0 0 3.04

.764 o500 0 11 .o 3.53 0 0 o475 4o44 0 0 2.03

.439 .600 0 7.18 4o59 0 0 .355 4o41 0 0 1.28

.224 .700 0 4.37 6.24 0 0 .233 4o 51 0 0 • 717

.149 o750 0 3.28 7o44 0 0 .174 4o57 0 0 • 501

.0923 ,800 0 2.37 9.02 0 0 .121 4o57 0 0 .326

.051 5 .850 0 1.63 11 .o 0 0 .0755 4o52 0 0 .191

.0241 .900 0 1.05 14.6 0 0 .0396 4o75 0 0 .0943

.00771 o950 0 o581 25.7 0 0 ,0142 5· 91 0 0 .0319

.00204 .980 0 .305 65.3 0 0 ,00400 9.33 0 0 ,00871

155
Calculation No. 48

1.0 go•

0.8
so•
0.6
w '9..

0.4 Jo•

0.2

0.02 0. 2 0.5
s

0.005 0.05 1 2 5 10

156
 Calculation No. 48

50 ~-------------------------------------------------------------------

20

10

5
...,
:.2

...,
G

0.005 0.01 0.02 0.05 0. 1 0.2 0.5 1 2 5 10

157
Calculation No. 49

1. Calculation no.: 49 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 1/2
P: 60° y
afp: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 14
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
X

Static Characteristics Spring Coefficients Damping Coefficients

s
e tp Qs Q, !J </J- I Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
2.45 .050 0 81 .o 28.4 0 0 28.0 34.2 0 0 34.0

1 • 21 .100 0 40.1 14.7 0 0 14.0 17.3 0 0 16.9

.573 .200 0 19.3 8.42 0 0 6.98 9.10 0 0 8.22

.345 .300 0 12.0 6.98 0 0 4o56 6.58 0 0 5.18

,223 .400 0 8.05 6.90 0 0 3.22 5o 52 0 0 3.54

,144 -500 0 5o 50 7o56 0 0 2.32 s.oo 0 0 2.42

.0883 .600 0 3. 71 8,86 0 0 1.60 4.72 0 0 1 • 61

.0498 .700 0 2.39 1 o. 7 0 0 1.02 4-56 0 0 .982

.0353 .750 0 1. 87 12.2 0 0 .774 4-58 0 0 .726

.0236 .BOO 0 1.42 14.4 0 0 • 554 4.74 0 0 .508

,0145 .850 0 1.04 17.8 0 0 .364 5,14 0 0 .328

.00765 ,goo 0 .705 25.6 0 0 .206 5.88 0 0 • 1 81

,00280 .950 0 .402 49.8 0 0 .0812 9.02 0 0 .0696

158
 Calculation No. 49

1.0 r---------------------------------------------------------------,90°

oo

50r-----------------------------------------~-----------------,

20

10

-
I
5
~
~

0.5

159
Calculation No. 49

50 ~-------------------------------------------------------------,

20

10

0.5

5
:::o
\..)

0.5

160
 Calculation No. 50

1. Calculation no.: 50 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP: 2/3
p: 60° y
afP: 0.5
Calculation method: FDM (over-relaxation
method)
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 14
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
X

Static Characteristics Spring Coefficients Damping Coefficients

s
€ ¢ Q. Q, JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
.993 .oso 0 49.7 48.9 0 0 48.6 38.1 0 0 38.1

o491 .100 0 24.6 25.2 0 0 24.3 19.3 0 0 18.9

.233 .200 0 11.9 1 4· 1 0 0 1 2.1 1o. 1 0 0 9.24

• 142 .300 0 7.43 11.2 0 0 7.83 7.20 0 0 s.8s

.0929 .400 0 5.07 1 o. 5 0 0 5·55 5·85 0 0 4.02

.0613 .sao 0 3.57 1 o. 7 0 0 4.02 5· 13 0 0 2.82

.0391 .600 0 2.49 11.9 0 0 2.84 4.80 0 0 1 .92

.0231 .700 0 1.68 1 4·3 0 0 1.87 4.83 0 0 1.22

.0169 .750 0 1 .35 16.5 0 0 1.45 4.98 0 0 .930

.0118 .800 0 1.06 19.7 0 0 1 .07 5.37 0 0 .675

.00757 .850 0 .793 25.3 0 0 .732 5.88 0 0 .450

.00417 .900 0 .553 37.2 0 0 .432 7.50 0 0 .260

161
Calculation No. 50

l.Or---------------------------------------------------------------,900

60°

sor-----------------------------------------------~------------------~

0.5

162
 Calculation No. 50

50~----------------------------------------------------------------~

.....
0

163
Calculation No. 51

1. Calculation no.: 51 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 0.5
4. Special items
mP : 3/4
P: 60° y
r:xjp: 0.5
Calculation method : FDM (over-relaxation
method)
Boundary condition : Reynolds condition
Number of grid points: Axial direction: 14
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
X

Static Characteristics Spring Coefficients Damping Coefficients

s
e ¢ Qs Q. /Jtf;-1 Kx x Kxy Kyx Kyy C xx Cxy Cyx Cyy

·543 .050 0 36.5 68.4 0 0 67.6 41.2 0 0 41.2

.269 .100 0 1 8.1 35.0 0 0 33.9 20.8 0 0 20.5

.128 .200 0 8.78 19.2 0 0 16. B 10.8 0 0 1 o.o

.0788 .300 0 5·55 14.6 0 0 11 .o 7.48 0 0 6.36

.0524 .400 0 3.85 13.3 0 0 7.80 6.04 0 0 4.40

.0352 .500 0 2.75 13.4 0 0 5.68 5.36 0 0 3.12

.0230 .600 0 1.97 14.8 0 0 4.08 5.oa 0 0 2.16

.0141 .700 0 1.37 17.8 0 0 2.75 5.36 0 0 1 .40

.0105 .750 0 1.11 20.8 0 0 2.16 5.52 0 0 1 .09

.00750 .BOO 0 .883 25.2 0 0 1.62 5.88 0 0 .aoo

.00490 .a so 0 .673 33.0 0 0 1.11 6.64 0 0 ·540

.00276 .goo 0 .473 49.2 0 0 .656 8.96 0 0 .317

164
 Calculation No. 51

l.Or---------------------------------------------------------------~90"

60°

30°

o·
0.005 0. 01 0.02 0.2 1

165
Calculation No. 51

50~------------------------------------------------------------,

·~

166
 Calculation No. 52

1. Calculation no.: 52 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: 0
P: 600 y
~P : ns
Calculation method: FDM
Boundary condition : Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q1N : QIN = L {Q;n)(NDLCP)} X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
€ QI N JJcp- l Kxx Kyy Cxx Cyy

3.30 .100 7.81 54.5 1. 30 1.10 14.0 14.0

1. 94 .200 7.70 32.2 1. 64 . 883 8.64 8.17

1. 26 .300 7.58 21.2 2.11 .764 6.38 5.30

.812 • 400 7.45 14.1 2.66 . 64 3 5.22 3.47

.509 .500 7.32 9.30 3.39 . 518 4. 72 2.24

.300 .600 7.18 5.90 4.39 . 388 4. 71 1. 44

.158 .700 7.05 3.52 5.93 .256 4.84 .828

.0679 .800 6.91 1. 90 8.25 .137 5.06 . 392

.0391 .850 6.83 1. 23 11.0 . 0877 5.65 .237

.0191 .900 6. 77 .876 16.2 .0477 6. 45 .122

.00633 .950 6.70 .504 30.1 .0177 7. 43 • 0428

167
Calculation No. 52

8.0

1.0

7.5
0.8
w

0.6

7.0 0.4

100~----------------------------------------------------------~

50

20

0.5

168
 Calculation No. 52

10~---------------------------------------------------------,

0.5
::;:.
::.::

0.2

0.1

0.05

5
s

..., 0.5
0

0.2

0.02 0.2 2 5

169
 Calculation No. 53

1. Calculation no.: 53 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: Q5
P: 600 y
rxfP: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q1N: QIN = L {Qin)(NDLCP)} X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
€ QIN !J<P-1 Kxx Kyy Cxx Cyy

.835 . 100 4.27 27.9 14.4 13.9 19.0 18.6

. 399 .200 4.27 13.5 8.32 6.97 10.0 9.11

. 243 .300 4.26 8.65 6.82 4.56 7.25 5.78

.158 .400 4.26 5.76 6.68 3.25 6.05 3.97

.103 .500 4.25 3.97 7 . 26 2.35 5. 45 2.76

.0650 . 600 4.24 2.76 8.64 1. 66 5.27 1.86

.0376 .700 4.23 1. 73 10.8 1. 08 5.68 1.17

.0186 .800 4.22 1.13 15 . 9 .610 6.41 .628

. 0117 .850 4.22 .877 20.9 .409 6.87 .412

.00627 .900 4.22 .596 29.8 .235 7.39 .232

170
 Calculation No. 53

1.0

0.8
4.5
0.6
~
~ w

0.4

4.0 0.2

o'L

50~---------------------------------------
-----~

20

171
Calculation No. 53

SOr---------------------------------------------~

50~-----------------------------------------,

·~

172
 Calculation No. 54

1. Calculation no.: 54 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP: 0.667
P: 600 y
rxjp: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q1N: QIN = L {Q;n) (NDLCP)} X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
e QIN JJ.p-l Kxx Kyy Cxx Cyy

.358 .100 2.96 18.0 25.1 24.2 24.3 21.3

.170 .200 2.96 8.69 13.9 12.0 11.2 10.4

.105 .300 2.95 5.49 11.0 7.84 8 . 01 6.61

.0692 .400 2.95 3.81 10.4 5.59 6.63 4.58

.0463 .500 2.95 2.61 10.7 4.08 6.19 3. 24

.0302 .600 2.95 1. 86 12.3 2 . 93 6. 20 2.24

.0183 .700 2.94 1. 34 15.9 1. 98 6.51 1. 46

.00958 .800 2.95 .895 23.1 1.16 7.05 .824

.00620 .850 2.95 .667 29.5 .796 7.38 .555

173
Calculation No. 54

1.0

0.8
3.0
0.6
~
a w

0.4

2.5 0.2

0.02 0. 2
s

.50r---------------------------------------------,

20

174
 Calculation No. 54

50~------------------------------------~

20

0.02 0. 2 0.5
s

50r-------------------------------------,

20

·~

0.02 0.2 0.5

175
 Calculation No. 55

1. Calculation no.: 55 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 1.0
4. Special items
mP : 0.75
P: 60° y
rx/P: 0.5
Calculation method: FDM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
30 (uniform spacing)
N
Definition of Q1N: QIN = L {Q;n,;/(NDLCP)} X
i= 1

Static Characteristics Spring Coefficients Damping Coefficients

s
€ QIN ljt/J-1 Kxx Kyy Cxx Cyy

.203 .100 2.29 13.5 35.0 33.9 23.6 23.1

• 0974 .200 2.29 6. 41 19.3 16.9 12.3 11.4

.0602 .300 2.29 4.15 14.8 11.0 8.93 7.26

• 0403 .400 2.29 2.88 13.6 7.92 7.54 5.08

.0275 .500 2.29 2.12 14.2 5.86 6.98 3.65

.0183 .600 2.29 1. 58 16.4 4.27 6.86 2.58

. 0114 .700 2.29 1.15 20.8 2.93 7.03 1. 73

.00614 .800 2.29 .736 29 . 4 1. 74 7 . 41 1. 01

176
 Calculation No. 55

1.0

2.5 0.8

0.6
~
0 w

0. 4
2.0
0. 2

oL 0
0.02 0· 2 0.5

50~-----------------------------------

0.5

177
 Calculation No. 55

50~------------------------------------~

20

~ 5

SOr------------------- -------------------,

20

178
 Calculation No. 56

1. Calculation no.: 56, EPB-1 (hydrostatic

bearing)
2. Bearing type: Hydrostatic bearing
(orifice throttle, not
rotating)
3. L: D ratio: 1
4. Special items
N umer of pockets: 4
Geometry of pockets: G 1 = 0.6, G2 = 0.889,
G3 = 0.5
Orific design
parameter: cs= 12.5
Calculation method: FEM
Boundary condition: Reynolds condition
Number of grid points: Axial direction: 20
(per pad) (uniform spacing) Cir-
cumferential direction:
20 (uniform spacing)
5. Bearing geometry and definitions

L e: density of oil cs: 3 fincdd~fi./(C~ jQP.),

I d 0 : diameter of orifice orifice design parameter
I! I Ca: flow coefficient of orifice W: W j(DLP,), dimensionless
g: acceleration of gravity load-carrying capacity
P,: supply pressure Q: 12fi.Q/(C: P,),
dimensionless flow
wl: J"iiC~
L, n: number of pockets K,/ CbKii/(P,DL), dimen-
sionless spring coefficient
-t- - - ·- r----0 Gl: LI /L )
y Gz: n8z/(2n) geometry Cii : Cbw 1 cj(P,DL),
G3 : 81182 of pocket dimensionless damping
coefficient

Note: Symbols not defined here are defined in the " Nomen-
HI cl a ture " section.

Pocket

Static Characteristics Spring Coefficients Damping Coefficients

w e ¢ Q /J!f -1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

.0359 .0500 0 31.1 .711 0 0 .715 4.05 0 0 4.04
.0711 .100 0 31.1 .674 0 0 .700 4.06 0 0 4.03
.139 .200 0 30.6 .655 0 0 .670 4.08 0 0 3.97
.202 .300 0 30.0 .562 0 0 .600 4.14 0 0 3.86
.258 .400 0 29.3 .475 0 0 .530 4.32 0 0 3.74
.305 .500 0 28.5 .421 0 0 .459 4.80 0 0 3.63
.346 .600 0 27.8 .352 0 0 .355 6.05 0 0 3.63
.380 .700 0 27.2 .317 0 0 .288 9.43 0 0 3.88
.411 .800 0 26.7 .306 0 0 .217 20.8 0 0 4.97
.447 .900 0 26 .3 .442 0 0 .169 87.3 0 0 10.4
.474 .950 0 26.2 .721 0 0 .149 313 . 0 0 25.0

179
Calculation No. 56

l.Or-------------------------------~----,100

180
 Calculation No. 56

Kxy =c 0
Kyx=c 0

0 .1.____---l~-L--L--L-L..ILWL...l..---..l------'-L__IL..L...J......l....U
0.01 0.02 0.05 0.1 0. 2 0.5 1
w

lOOr---------------------------~----~

Cxy=c 0
Cyx=c 0
·~

G Cxx

181
Calculation No. 57

1. Calculation no.: 57, SGB-1 (spiral

grooved bearing)
2. Bearing type: Spherical spiral
grooved bearing
3. L: D ratio: Sphere
4. Special items
rx = 30°, H 0 = 2, y = 1,¢ 1 = 45°, </1 2 = 90°,
k = 10
Calculation method : FEM
Boundary condition: P =I:, at <P = n/2
Number of grid points: MB = 3, MG = 15,
M M G= 2, M M R = 2

5. Bearing geometry

ez : eccentricity ratio = 1 - ezfAR

ex: groove angle
h0 : groove depth
y: land/groove width ratio
</> 1 : angle
</> 2 : angle
z k: number of grooves
L1 R: radial clearance
ez: eccentricity in z direction
R: radius of sphere

Static Characteristics Spring Coefficients Damping Coefficients

s
Kxx Kxy Cxx Cxy
ez f·lj!-1
Kyy -Kyx Kzz Cyy -Cyx Czz

. 3 4 7 .9 3 . 66 846 2.77 .954 5.87 =. o 83.0

.454 . 5 3 . 3 3 .604 .603 . 7 3 4 1 .33 =;a 12 . 1

.623 .0 3.65 . 3 83 . 351 . 495 .746 =.o 5.94

1 . 06 - 1 4 . 76 .168 . 179 . 2 1 6.388 =. o 2 . 88

2.24 -3 7.44 . 04 91 .0685 .0678 . 136 ='=iO 1 . 33

182
 Calculation No. 57

Ir-----------~--------------------------------------~

-1

-2

1~----~~---L--~-L~-L~~------~~--~--~~~~~w
0.1 0.2 1 2 5 10
s

183
Calculation No. 57

-TI_______-~3~-------------.
lOr-----------~o~.rg_o,.~s--To_____

...,
G

184
 Calculation No. 58

1. Calculation no.: 58, SGB-2 (spiral

grooved bearing)
2. Bearing type: Taper spiral grooved
bearing
3. L: D ratio: Taper, l/R = 1
4. Special items
r:x = 30°, y = 1, 19 /l = 0.5, ljR = 1, p = 20°,
k = 10
Calculation method: FEM
Boundary condition: P = Fa at</> = n/2
Number of grid points: M 8 = 3, M 6 = 18,
MMG = 2, MMR = 2

5. Bearing geometry

e, : eccentricity ratio
= 1 - h,/h0
cc groove angle
h0 : groove depth
y: land/groove width ratio
k: number of grooves
h,: bearing clearance
R: radius at the side of large
diameter
z p: half cone angle
I: axial length of cone
/ 9 : axial length of groove

Static Characteristics Spring Coefficients Damping Coefficients

s
K xx Kxy Cxx C xy
ez J· cp-l Kzz
Cyy - C yx
Czz
Kyy -Kyx

.0739 .9 3.62 1 0 53 l 93 9.12 3l 3 '=iO 324

282 .5 3.06 l 09 6 3.93 2 0 1 5 6.30 '=i O 5.39

1 02 8 0 7.45 .293 .750 .380 1 02 3 '=iO .658

4.02 -0.5 l 6 0l .0676 .263 .0939 .430 '=iO 0l 93

9.92 - 1 3 0 06 .0216 0122 .0305 . 1 9 7 '=i O .0797

185
Calculation No. 58

1~---------------------------------------------------,

0.5

-0.5

10
s

0. 9 0.5 0 -0 . 5 -1
50~~~----------~~------------~--------~r--------ft

186
 Calculation No. 58

....,
1
::<

0.5 Kxx (Kyy)

0.2

sr-~o.rg----------~~--------------~o__________-_o~-~5_______
-~I

....,
G

187
Calculation No. 59

1. Calculation no.: 59, SGB-3 (spiral

grooved bearing)
2. Bearing type: Flat spiral grooved
bearing
3. L: D ratio: Plane
4. Special items
r:t.. = 30°, y = 1, R;/R 0 = 0.5, RG/R 0 = 0.6,
k = 10 MMc
Calculation method: FEM
Boundary condition: P = Fa at R = R 0
Number of grid points: MB = 3, MG = 18,
M M G = 2, M M R = 2

5. Bearing geometry
(J)

t X

z
az
r=-
a,
e.: eccentricity ratio = 1 - h,/ h0 h,: bearing clearance
IX: groove angle Ri: inner radius
h0 : groove depth R.: radius of groove end
y: land/groove width ratio R 0 : outer radius
k: number of grooves

Spring Damping
Static Characteristics Coefficients
Coefficients

s
E:z ft/J-' Kzz Czz

.134 . 9 2. 6 2 1l . 3 340

.505 . 5 2.29 3. 50 5.83

2.27 0 5. 6 2 .622 .562

7 . 05 - 5 12.3 . 1 55 .136

1 7 . 2 -1 23 . l .0515 .0286

188
 Calculation No. 59

lr----------------------------------------------------,

0.5

.: 0

-0.5

0.1 20
s

189
 Calculation No. 59

190
3. Experitnental Data

This chapter contains the following experimentally determined data for eleven sets of test bearings:
Circular bearing with two oil grooves 5 sets
Two-lobe bearing 3 sets
Tilting pad bearing 3 sets
Total 11 sets
The data was determined by means of eight different rigs. Test bearing sets 1 and 9 were tested on the same
rig; bearing sets 2, 7 and 10 were also tested on one rig. A different test rig was used for each of the other
bearing sets.
The data is listed in the order of test bearing no. as follows :
Table 3.o.1 Experimental results for oil film coefficients
Table 3.o.2 Test conditions
Fig. 3.o.1 Experimental results for oil film coefficients
Fig. 3.o.2 Cross section of bearing test rig
where o stands for the test bearing no.
In these figures, the experimentally determined values of the dimensionless oil film coefficients are shown
as functions of the Sommerfeld number, and the curves theoretically determined have been plotted and
drawn in as solid lines for comparison. Each experimental value can be dimensionalized by reference to its
test condition.
All formulations of bearing characteristics, static or dynamic, require evaluation of the pressure distribu-
tion of the oil film. Ideally, characteristics should be formulated in terms of parameters definable at the
design stage and easily measurable during service. Then, for a given pressure profile and, hence, a given
bearing load, the thickness and viscosity distributions of the oil film and the journal speed need to be known.
Details on these points are discussed in Sect. 4.1 with reference to calculation method and in Sect. 4.2 with
reference to measurement method.

191
Test Bearing No. 1

Test Bearing No. 1: Circular Bearing with Two Oil Grooves

D = 140 mm, L = 140 mm, CP = 0.139 mm

Table 3.1.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
ratio Kxx Kxy Kyx Kn cxx cxy cyx c yy
s e ¢ deg. q cm 3 js roc

1 0.025 83 45.7 7.522 1.553 0.755 1.074 - 0.459 -0.470 -0.360 0.291
2 0.035 83 47.0 7.974 2.218 1.334 1.212 0.222 -0.391 -0.479 0.356
3 0.037 83 40.9 8.191 2.260 1.185 1.231 0.322 -0.345 -0.369 0.391
4 0.052 83 40.9 8.639 3.083 1.927 1.530 1.507 -0.0020 -0.220 0.606
5 0.070 133 55.7 6.728 3.512 1.727 1.825 -0.167 -2.749 -0.837 -0.630
6 0.084 117 49.8 5.482 3.456 1.441 1.844 -0.085 -2.474 -0.882 -0.659
7 0.088 100 42.5 5.911 2.876 1.273 1.409 1.066 -0.610 -0.681 0.361
8 0.126 191 60.5 2.792 2.913 0.0064 1.741 3.795 0.288 1.322 1.261
9 0.133 83 42.3 4.050 3.237 0.431 1.733 3.958 1.042 1.081 1.827
10 0.158 167 54.3 2.761 3.160 0.131 1.798 3.314 -0.141 0.979 1.064
11 0.195 183 59.0 2.057 3.373 -0.295 1.611 4.854 0.887 1.466 2.317
12 0.195 167 53.2 2.672 3.274 -0.143 1.870 4.893 0.871 1.766 1.838
13 0.225 100 42.3 2.483 3.188 -0.235 0.755 4.529 0.829 0.374 2.397
14 0.284 167 54.8 2.972 4.056 -0.537 1.669 5.847 -0.885 1.654 2.877
15 0.336 150 58.0 1.753 3.584 -1 .059 1.392 7.020 1.857 1.605 4.015
16 0.442 133 51.3 2.934 4.624 -0.745 1.694 6.783 -1.244 0.951 3.179
17 0.498 167 55.4 7.971 9.659 -0.330 1.872 15.77 -1.377 -0.554 0.132
18 0.715 133 57.1 3.185 8.338 -1.701 2.551 16.90 -1.374 2.179 4.011
19 0.913 100 45.3 2.615 7.425 -2.075 2.339 17.63 2.599 0.794 5.058
20 1.458 133 56.8 2.615 14.60 -7.075 2.952 33.44 4.130 3.053 14.09
21 1.947 117 43.4 4.841 14.24 -7.170 2.711 36.40 4.740 -0.767 15.25

Notes:
• At a given steady load-speed combination, test vig temperatures were stabilized and data were separately obtained, first for the
determination of eccentricity ratio and attitude angle, and secondly for the determination of oil film coefficients. In order to examine
the effects of the thermal drift of proximity sensors and the bearing distortion, data for a complete check on the bearing centre was
also recorded after the dynamic tests.
• Determination of effective film temperature. The weighted average value (7;. + 2 I;,",)/3 is assumed to be the effective temperature,
where 7;. is inlet oil temperature and I;,", is outlet temperature of the lower oil pocket.
• Effective viscosity for Sommerfeld number is calculated from the effective film temperature.

192
 Test Bearing No. 1

Table 3.1.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wDf2n Hz WkN Pm MPa p 1 MPa J;n °C J;,ut oC
1 0.025 8.33 15.0:30.0 35.6 1.81 0.1 40.5 48.2
2 0.035 8.33 15.0:30.0 24.3 1.23 0.1 41.4 49.8
3 0.037 8.33 15.0: 30.0 28.1 1.43 0.1 37.8 42.5
4 0.052 8.33 15.0:30.0 20.2 1.03 0.1 37.6 42.5
5 0.070 33.3 15.0:30.0 35.6 1.81 0.1 43.3 61.9
6 0.084 33.3 15.0:30.0 35.6 1.81 0.1 34.2 57.6
7 0.088 16.7 15.0:30.0 22.6 1.15 0.1 37.7 45.0
8 0.126 66.7 15.0:30.0 32.3 1.67 0.1 43.6 68.9
9 0.133 16.7 15.0:30.0 14.8 0.756 0.1 37.8 45.1
10 0.158 66.7 15.0:30.0 32.8 1.67 0.1 33.7 64.6
11 0.195 66.7 15.0:30.0 22.3 1.14 0.1 43.0 67.0
12 0.195 66.7 15.0:30.0 27.7 1.41 0.1 38.9 60.3
13 0.225 16.7 15.0:30.0 8.95 0.457 0.1 37.9 44.6
14 0.284 66.7 15.0:30.0 18.0 0.916 0.1 40.8 61.8
15 0.336 66.7 15.0: 30.0 13.8 0.702 0.1 42.7 65.6
16 0.442 66.7 15.0:30.0 13.5 0.687 0.1 32.6 60.7
17 0.498 66.7 15.0:30.0 10.0 0.512 0.1 42.1 62.0
18 0.715 66.7 15.0:30.0 6.8 0.347 0.1 42.8 64.3
19 0.913 66.7 15.0: 30.0 6.6 0.336 0.1 31.6 65.8
20 1.453 66.7 15.0:30.0 3.3 0.167 0.1 43.0 63.08
21 1.947 66.7 15.0:30.0 3.3 0.167 0.1 30.6 56.3

Notes:
• Properties of the lubricant
# 90 turbine oil
Specific gravity @ 15/4 °C 0.866
Kinetic viscosity mm 2 /s @ 37.8 °C 28.66
@ 50.0 °C 19.90
@ 98.9 °C 4.906
• Determination of amplitude ratio and dynamic load. The ratio of vibration amplitude to bearing clearance is adjusted within the
range 1/6 ~ 1/7 by changing the dynamic load 2 ~ 4 kN (peak).

193
 Test Bearing No. 1

1.0 100°
0: £ }
Experiment L
~;.:.p

.8 80° I

!
i

{
.6 60°
!
"' ·=t=· ~
~c--.
.4 40° .... II
I
jjl

.2 200 rI

·0 .02 .05 .1 .2 .5
oo
2
Bearing material : White metal WJH
Backing metal S45C
a S=Y..!Y. ( R )'
P• "C; Bearing geometry

r-------------------------- --------0
35 f-.

30 f-. 0: Kxx I> : Kxy 0 : Cxx I> : Cxy 0

} Experiment } Experiment
0 : Kyx o : Kyy 0 : Cyx o :C11
----Theory ----Theory
25-

20 f-.

0
0
0 Cxx
Uis
II
15 ,_

-Sf-
I I I I
.02 .05 .1 .2 .5 2 .02 .2 .5 2
0 0
s=Y..!Y.
Pm Cp
(..!i.)' c
s=Y..IY.(..!i
Pm Cp
. )'
b

Fig. 3.1.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

194
 Test Bearing No. 1

Fig. 3.1.2. Cross section of bearing test rig.

CD Test bearing casing (]) Tension rod
~ Bellows cylinder (for static loading) ® Support ball bearing
CD Displacement sensor ® Flexible diaphragm coupling
@ Accelerometer ® Rotational speed sensor
(3) Load cell @ Test shaft
@ Hydraulic exciter head (for dynamic loading) © Test bearing
@ Flexible bar (for preventing side inclination)
Diameter of test journal: Max. 2RO mm. usual 140 mm Static load: Max. 50 kN, from three directions
Speed range: 100 ~ 10000 rpm Dynamic load: Max. 10 kN, peak sinusoidal
DC~ 500Hz

Outline of Test Rig

a) Measurement method and provision of static and dynamic loads
Three sets of compressed-air bellows used for the static loading are attached to the test bearing casing both
horizontally and vertically. In addition, two hydraulic exciters are located left and right of the test bearing
casing 45° from vertical. Each of the static loads applied by the three bellows and the bias loads by the two
exciters is independently adjustable.
The two exciters allow dynamic sinusoidal forces to be applied together, with any relative phase and
magnitude.
The static load applied by the bellows is measured by Bourdon-type pressure gauges. The static bias load
and dynamic load applied by the hydraulic exciters are measured by the load cell attached to the exciter
heads.
b) Measurement method for static displacement and vibration response
The relative displacements between test bearing and journal are measured by four pairs of contact-less eddy
current proximity sensors, located on each side of the bearing in the horizontal and vertical directions.
The absolute accelerations of the test bearing casing are measured by four pairs of piezoelectric ac-
celerometers located at each position of the proximity sensors.
c) Other details, such as data processing, are described in Sect. 4.2.3.

195
Test Bearing No. 2

Test Bearing No.2: Circular Bearing with Two Oil Grooves

D = 100 mm, L = 100 mm, CP = 0.105 mm

Table 3.2.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
ratio Kxx Kxy Kyx K,, cxx cxy c,x en
s e </> deg. q cm 3/s r oc
1 0.0393 0.95 24 70.0 55.0 12.5 2.61 2.52 1.92 2.23 0.099 0.101 0.290
2 0.0486 0.91 28 75.0 55.0 9.58 2.07 2.16 2.09 1.00 0.187 0.010 0.395
3 0.0588 0.80 29 103.3 64.2 5.86 1.48 1.13 1.48 1.18 0.355 0.085 0.419
4 0.0655 0.85 33 75.0 54.7 9.99 2.54 2.75 2.59 1.08 0.185 0.058 0.572
5 0.0720 0.73 31 130.0 70.3 4.95 1.60 0.843 1.46 1.68 0.544 0.176 0.476
6 0.0752 0.75 34 100.0 63.0 6.33 1.90 1.34 1.66 1.59 0.337 0.208 0.545
7 0.0779 0.64 27 146.7 78.0 4.76 1.81 0.933 1.49 i 2.11 0.740 0.226 0.497
8 0.0785 0.73 30 175.0 86.8 4.11 1.93 0.609 1.43 2.33 0.730 0.285 0.584
9 0.0925 0.68 36 130.0 69.4 5.20 2.03 0.890 1.70 2.31 0.528 0.392 0.663
10 0.1003 0.62 33 148.3 76.8 4.30 2.12 0.513 1.70 2.61 0.633 0.603 0.804
11 0.1023 0.66 37 183.3 85.2 4.13 2.25 0.524 1.60 2.93 0.763 0.585 0.828
12 0.1034 0.76 40 70.0 53.3 11.8 4.34 3.98 3.95 1.94 0.031 0.366 1.04
13 0.1066 0.68 40 96.7 61.5 6.99 2.65 1.65 2.09 2.43 0.267 0.481 0.825
14 0.1260 0.63 46 125.0 68.8 5.55 2.69 0.960 2.03 3.32 0.420 0.730 0.983
15 0.1354 0.53 44 146.7 76.4 5.00 2.86 0.745 2.06 3.77 0.531 0.700 1.01
16 0.1377 0.59 44 175.0 84.6 3.91 2.67 0.305 1.79 4.04 0.434 0.932 1.34
17 0.1611 0.61 58 91.7 60.6 7.42 3.54 1.72 2.54 3.57 0.031 0.880 1.43
18 0.1899 0.56 58 121,7 68.5 5.23 3.26 0.796 2.49 4.81 0.016 1.21 1.62
19 0.2032 0.45 57 141.7 75.9 5.11 3.00 1.01 2.06 4.85 0.319 0.990 1.27
20 0.2096 0.50 56 163.3 84.0 3.58 3.10 0.053 2.00 5.37 0.070 1.33 1.98
21 0.2131 0.57 53 65.0 51.7 13.3 7.07 5.02 6.73 4.02 0.481 1.12 2.28
22 0.2977 0.47 62 80.0 60.8 6.99 4.53 1.37 3.54 6.03 0.610 1.78 2.74
23 0.3552 0.44 76 96.7 70.0 5.32 4.17 0.233 3.19 7.90 0.627 2.12 3.73
24 0.3924 0.31 78 120.0 77.1 3.70 4.47 1.10 3.14 7.73 0.227 2.39 3.98
25 0.3950 0.37 73 130.0 85.0 3.21 3.83 1.79 2.69 7.82 0.606 1.93 5.07
26 0.4229 0.38 70 63.3 51.4 6.99 4.70 0.126 3.73 7.20 0.822 1.46 3.71
27 0.6143 0.33 75 70.0 62.7 8.73 5.86 0.344 4.84 10.5 1.04 2.33 5.70
28 0.6642 0.30 88 80.0 72.8 7.53 5.50 1.40 5.17 13.0 0.139 2.53 8.56
29 0.7358 0.17 90 91.7 79.6 6.07 7.46 4.00 4.77 11.2 0.366 2.09 9.07
30 0.7800 0.19 85 105.0 87.5 7.91 6.29 5.88 2.80 14.2 2.98 0.388 12.8

196
 Test Bearing No. 2

Table 3.2.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps Wn/2n Hz WkN Pm MPa p 1 MPa J";n oC T,ul oC
1 0.0393 50.1 190 47.9 4.79 0.134 39.8 47.4
2 0.0486 49.8 190 38.4 3.84 0.140 41.9 47.8
3 0.0588 99.9 190 47.5 4.76 0.142 40.0 54.3
4 0.0655 50.0 190 28.9 2.89 0.144 43.2 47.8
5 0.0720 150 190 48.0 4.80 0.141 37.2 59.9
6 0.0752 100 190 38.9 3.89 0.140 40.0 52.5
7 0.0779 200 190 48.0 4.80 0.144 39.1 71.7
8 0.0785 250 190 47.9 4.79 0.138 46.3 85.5
9 0.0925 150 190 38.6 3.86 0.143 37.8 58.8
10 0.1003 200 190 38.4 3.84 0.150 38.9 69.9
11 0.1023 250 190 38.2 3.82 0.143 46.0 83.4
12 0.1034 50.1 190 19.4 1.94 0.140 42.8 46.7
13 0.1066 100 190 28.8 2.88 0.138 39.1 51.6
14 0.1260 150 190 28.8 2.88 0.136 38.0 57.8
15 0.1354 200 190 28.8 2.88 0.146 38.6 68.1
16 0.1377 251 190 28.8 2.88 0.145 45.4 82.2
17 0.1611 101 190 19.7 1.97 0.131 38.5 50.9
18 0.1899 150 190 19.4 1.94 0.131 37.7 57.2
19 0.2032 200 190 19.4 1.94 0.141 38.3 67.4
20 0.2096 250 190 19.2 1.92 0.139 44.4 81.8
21 0.2131 50.3 190 9.93 0.993 0.138 42.6 45.9
22 0.2977 100 190 10.6 1.06 0.131 39.2 51.0
23 0.3552 150 190 9.85 0.985 0.135 38.5 57.3
24 0.3924 200 190 9.76 0.976 0.148 39.4 69.8
25 0.3950 249 190 9.91 0.991 0.140 42.8 85.4
26 0.4229 50.2 190 5.05 0.505 0.138 41.7 45.1
27 0.6143 100 190 4.80 0.480 0.137 41.9 51.5
28 0.6642 150 190 4.86 0.486 0.139 38.3 57.8
29 0.7358 200 190 4.90 0.490 0.132 39.4 72.8
30 0.7800 249 190 4.74 0.474 0.140 41.6 91.2

Notes:
• Properties of lubricant
# 90 turbine oil
Specific gravity @ 20/4 °C 0.865
Kinetic viscosity mm 2 fs @ 40 °C 31.64
@ 50 °C 21.00
@ 10o oc 5.267
• Effective oil film temperature and equivalent viscosity for calculation of Sommerfeld number. The average bearing metal temper-
ature over 20 points is assumed to be the effective temperature of oil film, and the equivalent viscosity is calculated on the basis
of the effective temperature.
• Determination of exciting force. The ratio of vibration amplitude to bearing clearance is adjusted within the range from 1/6 to 1/20
by changing the exciting force (1 ~ 3.5 kN, peak).

197
 100°

80°

60°

40° .....

20°

00
2
Bearing material: White metal ASTM B23
S=i!N(_B_)' Backing metal SS41
a Pm Cp
Bearing geometry

0: Kxx IJ. : Kxy 0 Cxx A Cxy

} Experiment } Experiment
c : Kyy o Cyy

---Theory ---Theory

·~

<>
a
... ::;:.. c.ij:S:
II
Cxx
c.ijs ·~
0
II 0
0 G 0
0
:<
·~

0
0 0 10
Kxy cP
0
0
5 0 Cyy

Cxy, Cyx
0 0

0 Kyx
"'
-5 -5

.02 .05 .1 .2 .5 2 .02 .2 .5 2

b
s=lili (_B_ )'
P• Cp c s=iili(li_
pm Cp
r
Fig. 3.2.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

198
 Test Bearing No. 2

Fig. 3.2.2. Cross section of bearing test rig.

CD Test shaft @ Support bearing (j}l Rotational speed pickup
(£) Intermediate shaft (J) Test bearing © Acceleration pickup
(] Reduction gear ® Bearing box QY Spring
@ Gear coupling ® Hydraulic exciter @l Displacement pickup
(3) 110-kW direct current motor (!g) Bellows © Load cell
Diameter of test journal: 100 mm Static load: Max. 68.6 kN
Speed range: 500 ~ 20000 rpm (continuously variable) Dynamic load: Max. 25 kN (peak)

Outline of Test Rig

a) Measurement method and provision of static and dynamic loads
The static force, which is the sum of the following three loads, can be applied to the test bearing: (1) the
test bearing weight; (2) the forces generated by pressure on the three bellows cylinders installed at the left,
right, and lower sides of the bearing housing; and (3) the forces applied by two hydraulic exciters installed
under the bearing housing 45° from vertical.
The dynamic forces are applied to the test bearing by the hydraulic exciters.
The static load applied by the bellows is given by measuring the pressures in the bellows, and the static
and dynamic loads applied by the hydraulic exciters are detected by load cells installed between exciters
and bearing housing.
b) Measurement method for static displacement and vibration response
Eight eddy current gap sensors, which measure the relative displacements between bearing housing and
shaft, are installed at the both ends of the bearing housing in the vertical and horizontal directions.
The accelerations of the bearing housing are measured by accelerometers installed on the bearing
housing.
c) Data acquisition
The high-speed data (vibrations and dynamic loads) and the low-speed data (temperatures, pressures and
static loads) are fed into a minicomputer through an A- D converter and through a DVM, respectively.
The data are automatically analyzed by the computer, and the static and dynamic characteristics are
relayed to a visual display and to a printer.

199
Test Bearing No.3

Test Bearing No.3: Circular Bearing with Two Oil Grooves

D = 100 mm, L =50 mm, CP = 0.140 mm

Table 3.3.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
K xx Kxy Kyx Kyy cxx cxy cyx c yy
ratio
s e 4> deg. q cm 3 /s T °C

1 0.0549 0.835 26.7 56.9 11.3 4.58 2.02 1.98 7.01 1.10 2.01 0.889
2 0.0854 0.793 32.4 55.8 7.80 4.67 1.12 2.11 8.35 2.46 3.08 1.82
3 0.114 0.755 35.6 55.6 5.47 4.10 1.795 2.23 8.18 2.82 3.32 2.24
4 0.123 0.749 29.6 64.4 5.97 4.24 0.493 1.95 7.29 1.72 2.54 1.54
5 0.182 0.685 32.7 62.3 3.98 3.44 0.0932 1.84 6.63 2.05 2.63 1.95
6 0.244 0.627 36.0 60.8 3.12 2.90 0.0989 1.69 5.94 1.87 2.18 1.80
7 0.318 0.572 39.5 60.1 2.53 2.65 0.125 1.48 5.66 1.79 1.71 1.64
8 0.408 0.521 43.7 59.3 1.99 2.62 0.194 1.37 5.77 2.08 1.40 1.61
9 0.501 0.467 49.4 58.5 1.60 2.45 0.203 1.33 5.66 2.08 1.22 1.55
10 0.564 0.464 62.1 62.0 2.43 2.04 -0.354 1.22 5.94 2.08 1.36 2.19
11 0.700 0.418 68.3 61.8 2.39 1.90 -0.534 1.10 6.01 1.66 1.34 2.51
12 0.871 0.372 75.8 61.0 2.48 1.83 -0.777 0.967 5.91 0.953 1.35 2.87
13 1.06 0.328 82.8 60.7 2.54 2.16 -1.01 0.811 6.03 0.640 1.34 3.45
14 1.31 0.287 90.4 60.8 3.12 2.40 -1 .45 0.763 4.81 -1.59 1.38 4.57
15 1.60 0.233 98.5 60.9 3.70 3.52 -1.80 0.671 7.92 - 2.77 0.988 5.76
16 2.21 0.157 108 61.3 4.17 6.03 -2.94 0.162 17.4 -3.38 -1.87 7.83

Table 3.3.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wD/2n Hz WkN Pm MPa p 1 MPa T;n °C 7;,., oc
1 0.0549 26.7 45 4.50 0.901 0.1 50 54
2 0.0854 26.7 45 3.00 0.600 0.1 50 53
3 0.114 26.7 45 2.25 0.450 0.1 50 53
4 0.123 66.7 45 4.02 0.804 0.1 50 58
5 0.182 66.7 45 2.88 0.577 0.1 50 58

6 0.244 66.7 45 2.25 0.450 0.1 50 57.1

7 0.318 66.7 45 1.76 0.352 0.1 50 57
8 0.408 66.7 45 1.41 0.281 0.1 50 57
9 0.501 66.7 45 1.17 0.235 0.1 50 56.9
10 0.564 100 20 1.41 0.281 0.1 50 61

11 0.700 100 20 1.14 0.228 0.1 50 60.5

12 0.871 100 20 0.938 0.188 0.1 50 60.1
13 1.06 100 20 0.782 0.156 0.1 50 60.5
14 1.31 100 20 0.626 0.125 0.1 50 60.8
15 1.60 100 20 0.51 2 0.102 0.1 50 60.9

16 2.21 100 20 0.367 0.075 0.1 50 61.0

Notes:
• Properties of the lubricant
Turbine oil ISO VG32
Specific gravity @ 15/4 oc 0.862
Kinetic viscosity @ 37.8 oc 34.6
@ 50.0 oc 21 .3
@ 98.9 oc 5.50
• Determination of effective film temperature. The effective film temperature is assumed to be the average of four temperatures
measured at the locations shown in Fig. 3.3.2 by means of copper-constantan (CC) thermocouples implanted within 3 mm of the
bearing metal surfaces.
• Effective viscosity for Sommerfield number. Effective viscosity is calculated from the effective film temperature.
• Determination of amplitude and frequency of the dynamic load. Horizontal and vertical dynamic loads are set up so that the
amplitude of the vibration response in each direction reaches approximately one-tenth of the minimum bearing clearance it is or
taken as less than 300 N.

200
 Test Bearing No. 3

1.0 100°
{>

.8
{> w

0 : € l 0
{>
80° 100
I
6 : ¢ J Experiment
.6 60°
IU Theory
"' C>
C> ~
.4 40°..,
"" ~~'*"~~-+ ~
~
~

.2 €
20°
Holes for insertion
.0
ao L=50 of thermocouples
.02 .05 .1 .2 .5 2
Bearing material : White metal WJ2
S=yjj_ (lL)' Backing metal SS41
a P• Cp
Bearing geometry

25

20
0 :
A
A
v
: Kxy
•

o : Kyy
Kxx
. Ky:r
l Experiment
20,------------------------------------.

15 --Theory

..,-
·~

~::: 10
II
·~

·::::=--- Cn
A
<>
A
A
- 5 '-----'---'--'-'--'-'-'-'-..____ __,___.__._._-'-'-'-'-'----'---'---'--' I I I I I I I
.01 .02 .05 .1 .2 .5 2 . 02 .05 .1 .2 .5 2

b
S= yJj_
P• Cp
(.Jl_)' c s"' yJj_ ( .Jl_ )'
pm Cp

Fig. 3.3.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

201
Test Bearing No. 3

Fig. 3.3.2. Cross section of bearing test rig.

CD Housing ®Piano wire
(2) Bellows cylinder (for static load) (/) Shaft
Q) Electric exciter in the horizontal direction ® Displacement sensor
@ Hydraulic exciter in the vertical direction ® Supporting ball-bearing
(3) Load cell ® Accelerometer
Diameter of test journal : 100 mm Dynamic load: Max. 2.2 kN (x-direction)
Speed range: 1300 ~ 6000 rpm Max. 0.29 kN (y-direction)
Static load : Max. 4.5kN

202
 Test Bearing No. 4

Test Bearing No.4: Circular Bearing with Two Oil Grooves

D = 100 mm, L =50 mm, CP = 0.130 mm

Table 3.4.1. Experimental results of oil film coefficients at N = 12.5 rps

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
Kxx Kxy Kyx KYY cxx cxy cyx cyy
ratio
s e if> deg. q cm 3 /s roc
1-1 0.0421 0.959 18.6 - - 6.69 2.39 0.106 1.38 0.624 -0.0927 0.259 0.346
1-2 0.0520 0.953 20.9 - 29.9 7.15 2.60 0.0391 1.40 0.893 -0.0165 0.259 0.41 7
1-3 0.0683 0.898 23.6 - - 7.35 2.82 -0.242 1.33 1.38 0.178 0.325 0.564
1-4 0.102 0.846 28.0 - - 7.20 2.89 -0.0867 1.36 1.73 0.278 0.464 0.678
1-5 0.132 0.800 31.8 - 28.7 6.71 2.82 0.102 1.33 2.10 0.340 0.512 0.712
1-6 0.193 0.732 39.5 - 27.4 5.59 2.90 0.200 1.38 2.54 0.630 0.559 0.935
1-7 0.249 0.683 43.1 - - 3.93 2.10 0.1 70 1.10 2.24 0.530 0.509 0.904
1-8 0.350 0.602 50.0 - 26.8 3.44 1.94 0.249 1.05 2.14 0.442 0.454 0.989
1-9 0.595 0.463 62.2 - - 3.04 1.61 0.0761 0.933 2.24 0.247 0.244 1.31

Table 3.4.2. Test conditions at N = 12.5 rps

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wD/2n Hz WkN Pm MPa p 1 MPa T.n °C 7;,., oc
1-1 0.0421 12.5 30 10.9 2.18 0.098 - -
1-2 0.0520 12.5 30 8.81 1.77 0.098 26.2 27.7
1-3 0.0683 12.5 30 6.71 1.34 0.098 - -
1-4 0.102 12.5 30 4.50 0.901 0.098 26.1 27.4
1-5 0.133 12.5 30 3.46 0.693 0.098 25.4 26.5
1-6 0.193 12.5 30 2.37 0.475 0.098 - -
1-7 0.249 12.5 30 1.84 0.369 0.098 - -

1-8 0.350 12.5 30 1.31 0.263 0.098 25.3 25.9

1-9 0.595 12.5 30 0.771 0.154 0.098 - -

Notes:
• Properties of the lubricant
# 90 turbine oil
Specific gravity @ 15/4 °C 0.866
Kinetic viscosity mm 2 /s @ 37.8 oc 32.5
@ 50.0 oc 20.0
@ 98.9 oc 5.33
• The effective film temperature is assumed to the temperature at 3 mm below bearing metal surface.
• The effective viscosity for the Sommerfeld number is calculated from the effective film temperature.
• The exciting force range is set at 0.981 kN to 1.96 kN, and the ratios of vibration amplitude to the bearing clearance vary from 1/100
to 1/4.

203
Test Bearing No. 4
w
!
1.0 100°
0 0

.8 80°
50
o:e
.6 .:. : ; 60°

--Theory ....
.4 40°

.2 20°

00
0
.02 .05 .1 .2 .5 2 70
Bearing material : White metal WJ2
a
s=u.E. (.li. )'
pm Cp Backing metal SS41
Bearing geometry

0 : Kxx " : Kxy 0 : Cxx " : Cxy }

} Experiment Experiment
o : Kyy 0 : Cyx o : C yy

---Theory ---Theory

....·~
.:,-
a
Js
II Js
·~
II
:.:: ·~

-5 -5

.02 .05 .1 .2 .5 .02 .1 .2 -~ 2

b c
s=u.E. (.li.)
p .. Cp
2

Fig. 3.4.1 a-c. Experimental results for oil film coefficients at N = 12.5 rps. a Eccentricity ratios and attitude angles. b Dimensionless
stiffness coefficients. c Dimensionless damping coefficients

204
 Test Bearing No. 4

Fig. 3.4.2. Cross section of bearing test rig.

CD Test shaft ® Flexible coupling
(Zl Support ball bearing (!) DC motor
Q) Test bearing Qj) Supporting bar
@ Bellows cylinder ©! Outlet oil tube
~ Compressed air

Diameter of test journal: 100 mm Static load : Max. 20 kN

Speed range: 0 ~ 3000 rpm Dynamic load : Max. 6 kN, peak sinusoidal
DC~ 100Hz

Outline of Test Rig

a) Measurement method and provision of static and dynamic loads
Three sets of compressed-air bellows used for static loading are attached to the test bearing casing both
horizontally and vertically. In addition, two hydraulic exciters are located left and right of the test bearing
casing 45° from vertical. Each of the static loads applied by the three belows and the bias loads applied by
the two exciters is independently adjustable.
By means of the two exciters dynamic sinusoidal forces can be applied together, with any relative phase
and magnitude.
The static loading applied by the bellows is measured by means of Bourdon-type pressure gauges. The
static bias load and dynamic load applied by the hydraulic exciters are measured by the load cell attached
to the exciter heads.
b) Measurement method for static displacement and vibration response
The relative displacements between test bearing and journal are measured by four pairs of contact-less eddy
current proximity sensors located on each side of the bearing in the horizontal and vertical directions.
The absolute accelerations of the test bearing casing are measured by means of four pairs of piezoelectric
accelerometers located at each positition where there are proximity sensors.
c) Measurement method for oil film coefficients
Eight oil til coefficients are determined by a method based on single frequency sinusoidal wave excitation.
The excitation forces are applied at 45° from vertical, both left and right of the test bearing.

205
Test Bearing No. 4

Table 3.4.3. Experimental results for oil film coefficients at N = 25 rps

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
K xx Kxy Kyx Kyy cxx cxy cyx c yy
ratio
s e 4> deg. q cm 3 /s roc

2-1 0.0373 0.941 16.8 - 42.5 5.09 1.98 0.150 1.19 0.601 -0.106 0.120 0.235
2-2 0.0519 0.881 19.1 - 40.7 6.71 2.65 0.605 1.38 1.34 0.0313 0.440 0.384
2-3 0.0643 0.857 20.2 - - 6.13 2.26 0.215 1.30 1.63 0.156 0.399 0.475
2-4 0.0844 0.810 23.2 - 39.0 6.12 2.78 0.146 1.29 1.98 0.136 0.404 0.526
2-5 0.126 0.745 27.8 - 38.2 5.73 2.73 0.108 1.19 2.20 0.0846 0.390 0.626
2-6 0.164 0.690 32.0 - - 5.90 3.11 0.0879 1.24 2.74 0.241 0.657 0.865
2-7 0.239 0.603 41.3 - - 5.29 2.98 0.0137 1.16 3.04 0.458 0.796 1.14
2-8 0.308 0.548 45.5 - - 4.31 2.80 -0.0105 1.07 3.34 0.657 0.830 1.30
2-9 0.432 0.452 53.0 - 34.8 3.57 2.73 - 0.331 1.02 3.56 0.805 0.777 1.71
2-10 0.734 0.333 59.5 - - 3.42 3.20 -0.863 0.778 4.41 1.20 0.632 2.81
2-11 1.14 0.130 90.0 - 34.1 4.01 2.72 -1.68 0.859 5.69 -0.612 -0.0517 4.42

Table 3.4.4. Test conditions at N = 25 rps

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps Wn/2rc Hz WkN p,. MPa p 1 MPa T;. oc T;,., oc
2-1 0.0373 25 37.5 15.2 3.03 0.981 32.7 38.3
2-2 0.0519 25 37.5 10.9 2.18 0.981 32.9 37.6
2-3 0.0643 25 37.5 8.81 1.77 0.981 - -
2-4 0.0844 25 37.5 6.71 1.34 0.981 32.6 36.7
2-5 0.126 25 37.5 4.50 0.901 0.981 - -

2-6 0.164 25 37.5 3.64 0.693 0.981 - -

2-7 0.239 25 37.5 2.37 0.475 0.981 - -
2-8 0.308 25 37.5 1.84 0.369 0.981 - -
2-9 0.432 25 37.5 1.31 0.263 0.981 31.5 34.3
2-10 0.734 25 37.5 0.771 0.154 0.981 - -

2-11 1.14 25 37.5 0.496 0.0991 0.981 31.7 34.2

206
 Test Bearing No. 4

1.0 100°
0
tJ.
.8
80°
0
0: €
.6 } Experiment
tJ.:; 60°
...
--Theory
.4 40°.....
0

.2 20°
0

0
.02 .05 .1 .2
oo
.5 2

a S=idf_ ( Jl. )'

P• Cp

0 : Kxx 6
. Kxy } Experiment
: 0 : Cxx A :
. C
Cxy }Experiment
0 : Kyx o . K>Y o·: Cyx 0 . yy

---Theory - - - Theory

.;,-
·~

a
U\s
·~
II
·~

~ G

.02 .2 . 02 2

b
S=Jdf_ (.ll.)'
P• Cp c

Fig. 3.4.3a-c. Experimental results for oil film coefficients at N = 25 rps. a Eccentricity ratios and attitude angles. b Dimensionless
stiffness coefficients. c Dimensionless damping coefficients

207
Test Bearing No. 4

Table 3.4.5. Experimental results for oil film coefficients at N = 50 rps in test bearing no. 4
Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
K xx Kxy K yx KYY cxx cxy cyx cyy
ratio
s e rjJ deg. q cm 3/s r oc

3-1 0.0511 0.822 21.0 - 52.0 4.98 2.18 1.19 1.66 0.714 -0.103 0.0571 0.162
3-2 0.0711 0.786 24.0 - 48.5 5.16 2.54 1.22 1.84 1.31 0.0686 0.351 0.392
3-3 0.0880 0.762 26.0 - - 5.54 2.40 0.289 1.09 1.53 -0.0294 0.296 0.474
3-4 0.116 0.714 29.5 - 44.8 4.63 2.23 0.519 0.972 1.44 -0.0495 0.0628 0.434
3-5 0.172 0.654 33.3 - - 4.46 2.44 0.644 0.897 1.98 0.0795 0.268 0.641
3-6 0.224 0.603 36.8 - 41.1 3.58 2.47 0.0896 0.698 2.58 0.384 0.455 0.921
3-7 0.327 0.532 41.4 - - 2.75 2.16 -0.134 0.435 3.14 0.610 0.639 1.26
3-8 0.421 0.476 49.2 - 39.5 2.33 2.28 -0.388 0.276 3.76 0.880 0.761 1.61
3-9 0.591 0.397 55.6 - - 1.75 2.31 -0.726 -0.173 3.95 1.04 0.583 2.15
3-10 1.01 0.278 69.0 - 39.8 1.67 2.75 -1.48 -0.912 5.28 0.867 0.357 3.33

Table 3.4.6. Test conditions at N = 50 rps

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps w 0 j2n Hz WkN PmMPa p 1 MPa 7;. oc 7;,., oc
3-1 0.0511 50 75 15.2 3.03 0.0981 33.9 46.3
3-2 0.0711 50 75 10.9 2.18 0.0981 33.6 45.1
3-3 0.0880 50 75 8.81 1.77 0.0981 - -
3-4 0.116 50 75 6.71 1.34 0.0981 33.2 43.5
3-5 0.172 50 75 4.50 0.901 0.0981 - -

3-6 0.224 50 75 3.46 0.693 0.0981 32.6 41.2

3-7 0.327 50 75 2.37 0.475 0.0981 - -
3-8 0.421 50 75 1.84 0.369 0.0981 32.0 39.3
3-9 0.591 50 75 1.31 0.263 0.0981 - -
3-10 1.01 50 75 0.771 0.154 0.0981 31.5 40.2

208
 Test Bearing No.4

1.0 ,-----------------------------------------,100°

.8

.6

--Theory
.4

.2

0
.02 .05 .1 .2 .5
S= 0Y_ ( _!i_ )'
a Pm Cp

0: Kxx A : Kxy } 0 : Cxx A : Cxy }

.K Experiment . Experiment
0: Kyx 0 . yy 0 : Cyx 0 . c,.,.
------ Theory ---Theory

.;,-
~~ a
U\s U\s
II
II
u
·~

:.;:
·~

Kxy

-5

.02

Fig. 3.4.4a-c. Experimental results for oil film coefficients at N = 50 rps. a Eccentricity ratios and attitude angles. b Dimensionless
stiffness coefficients. c Dimensionless damping coefficients

209
Test Bearing No. 5

Test Bearing No.5: Circular Bearing with Two Oil Grooves

D = 50 mm, L = 25 mm, CP = 0.033 mm

Table 3.5.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
Kxx Kxy Kyx Kyy cxx cx y cyx c yy
ratio
s e cjJ deg. q cm 3/s r oc
1 0.122 0.750 31.5 0.150 58.6 7.02 3.32 2.32 2.95 10.2 2.70 1.03 2.44
2 0.167 0.682 43.4 0.151 58.6 4.79 2.30 1.46 2.04 6.58 2.81 1.19 2.28
3 0.221 0.628 46.3 0.200 61.4 3.95 1.74 0.952 1.43 4.73 2.78 0.603 2.17
4 0.252 0.600 46.1 0.222 65.8 3.82 1.55 0.552 1.39 4.68 1.54 0.600 2.23
5 0.294 0.591 50.8 0.107 55.5 3.66 1.81 0.401 1.50 5.33 1.53 0.587 2.51
6 0.382 0.528 53.0 0.214 64.8 3.48 1.79 0.280 1.50 4.87 1.82 0.604 2.73
7 0.449 0.483 57.5 0.279 69.3 3.82 2.13 0.301 1.51 5.45 1.89 0.711 2.78
8 0.569 0.452 58.9 0.095 50.3 4.16 2.34 0.099 2.31 5.71 2.77 0.805 3.43
9 0.756 0.380 72.8 0.125 56.0 5.08 3.30 -0.403 3.50 7.06 3.79 0.918 4.71
10 0.879 0.344 68.5 0.212 63.4 5.73 4.18 -0.893 3.63 7.73 4.71 1.33 5.08

Table 3.5.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps Wn/2n Hz WkN PmMPa p 1 MPa T;n °C T,., oc
1 0.122 25.0 30.0:70.0 1.52 1.21 0.1 40.3 55.2
2 0.167 34.5 30.0:70.0 1.52 1.22 0.1 40.7 55.8
3 0.221 50.0 30.0:70.0 1.54 1.24 0.1 41.5 57.8
4 0.252 66.7 30.0:70.0 1.60 1.28 0.1 40.5 61.9
5 0.294 37.5 30.0:70.0 1.04 0.832 0.1 38.8 53.0
6 0.382 62.5 30.0:70.0 1.01 0.808 0.1 41.2 61.8
7 0.449 83.6 30.0 : 70.0 1.02 0.812 0.1 41.1 64.8
8 0.569 29.1 30.0:70.0 0.495 0.396 0.1 37.0 48.6
9 0.756 45.8 30.0:70.0 0.485 0.388 0.1 37.8 53.2
10 0.879 66.6 30.0:70.0 0.485 0.388 0.1 40.6 60.9

Notes:
• Properties of the lubricant
* 90 turbine oil
Specific gravity @ 15/4 °C 0.871
Kinetic viscosity mm 2 js @ 30 °C 34.6
@ 50.0 oc 15.4
@ sooc 6.50
• Measurement method for inlet oil temperature ('I;") and outlet oil temperature (T,.,). 7;" and T0 . , are measured by using CC
thermocouples mounted in the inlet oil hole and the outlet oil hole.
• Effective viscosity for Sommerfeld number. The effective film viscosity is assumed to correspond to the temperature measured within
3 mm of the bearing metal surface in the direction of load.

210
 Test Bearing No. 5

1.0 100°

.8 80°

0 :
e~t
£
} Experi m 60°
.6 "' :~ '/, '//#-

....
-
- - Theory

.4 40°
Hf~a-+~~a- q
Inlet hole for
supply lubricant
.2 20°
1:'1'1I J1IV1/.'n'
I
00
.02 .2 .5 2
Bearing material: White metal WJ2

a
S= y]1_ (-lL)'
P• C p Bearing geometry
Backing metal SS41

0 : Kxx I> : Kxy} Experiment 0 : Cxx t;. : Cxy }

o : Kyy ¢ : Cyx 0 : Cyy Experiment
¢ : Kyx

- - - Theory --- Theory

.:;
-~

a
Uj=::
II
J

-5 -5

.05 .1 .2 .5 1 2 .02 .05 .1 .2 .5 2

b
S=MN(_B__)'
pm C p c

Fig. 3.5.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

211
Test Bearing No. 5

® ® ® ®

Fig. 3.5.2. Cross section of bearing test rig.

<D Test bearing ® Static load rig
(ZJ Accelerometer (/) Flexible coupling
Q) Connecting rod ® Transmission gear
@ Electric exciter ® Alternating current motor
~ Load cell (for static load) ® Speed meter
Q]) Supporting roller bearing
Diameter of test journal: 50 - 90 mm Static load: Mas. 4.8 N
Speed range: 550- 5500 rpm Dynamic load: Max. 295 N
DC -1kHz

212
 Test Bearing No. 6

Test Bearing No.6: Two-lobe Bearing

D = 508 mm, L = 304.8 mm, CP = 0.660 mm, mP = 0.5

Table 3.6.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
Kxx Kxy Kyx Kyy cxx cxy cyx Cy y
ratio
s e ¢ deg. q cm 3js roc

1 0.033 0.973 60.9 430 67.1 5.90 2.04 0.375 1.162 2.72 1.329 0.160 1.159
2 0.037 1.00 68.0 440 64.3 5.30 1.860 0.350 1.208 2.50 1.310 0.192 1.131
3 0.045 0.969 69.3 440 65.2 4.48 1.451 0.267 1.351 2.24 1.313 0.233 1.042
4 0.054 0.939 76.8 458 65.8 3.87 1.259 0.208 1.502 2.06 1.319 0.268 1.882
5 0.068 0.931 83.6 442 67.0 3.60 1.200 0.173 1.672 2.21 1.328 0.431 0.855

6 0.080 0.885 84.3 450 67.0 3.61 1.197 0.814 1.883 2.39 1.323 0.723 0.966
7 0.086 0.858 86.2 455 66.4 3.66 1.248 0.0388 2.06 2.47 1.330 0.851 0.953
8 0.096 0.862 88.2 435 63.0 3.73 1.324 0 2.27 2.65 1.353 1.00 0.971
9 0.110 0.836 90.3 450 59.5 3.84 1.436 - 0.0518 2.47 2.78 1.371 1.308 1.099
10 0.133 0.781 92.6 440 66.0 4.03 1.592 -0.201 2.81 2.85 1.401 1.596 1.124

Table 3.6.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wnf2n Hz WkN Pm MPa p1 MPa T;. oc T,ut oC
1 0.033 60 36.40.45.51.4 388.5 2.51 -0.051 49.0 67.1
2 0.037 60 36.40.45.51.4 383.9 2.48 - 0.051 47.0 64.3
3 0.045 60 36.40.45.51.4 311.1 2.01 -0.054 47.8 65.2
4 0.054 60 36.40.45.51.4 251.9 1.63 -0.053 47.9 65.8
5 0.068 60 36.40.45.51.4 191.3 1.23 -0.065 47.9 67.0
6 0.080 60 36.40.45.51.4 163.1 1.06 -0.069 47.7 67.0
7 0.086 60 36.40.45.51.4 154.8 1.00 -0.071 47.7 66.4
8 0.096 60 36.40.45.51.4 158.3 1.02 -0.069 41.6 63.0
9 0.110 60 36.40.45.51.4 153.3 0.99 -0.065 40.3 59.5
10 0.133 60 36.40.45.51.4 97.1 0.63 -0.071 47.5 66.0

Notes:
• Properties of lubricant
Turbine oil IS032
Specific gravity @ 15/4 oc 0.868
Kinetic viscosity mm 2 js @ 37.8 °C 34.6
@ 50.00C 21.3
@ 98.9 °C 5.50
• Measurement method for inlet oil temperature (7;.) and outlet oil temperature (T,u1). 7;. and T,u, are measured by chromel-alumel
(CA) thermocouples attached to the inside of bearing casing and each pipe.
• Equivalent viscosity to calculate Sommerfeld number. T,u, is assumed as an effective temperature of oil film. Therefore, measured
oil pressure distribution corresponds to the calculated oil pressure distribution at T,u•·
• Condition of amplitude ratio and exciting force. Exciting force is 35.5 - 58.8 kN, so that the ratio of vibration amplitude to bearing
clearance will be 1/15 - 1/6.

213
 Test Bearing No. 6

1.0 100°
0 ; E

A:{> } Experiment
.8 goo

--Theory
... .6 A 60°
....
.4 40°

.2 20°

0
.01 .02 .05 .1 .2 .5
Bearing material: White metal WJH
a
S=y_N (
pm Cp
Ji..)' Backing metal SS41
Bearing geometry

20 20,-----------------------------------~

0: Kxx A: Kxy } 0 : Cxx 6 C }

. K Experiment ·
15 r- () : Kyx 0 0 yy
0 : Cyx D .
: xy Experiment
Cyy
Cxx
--Theory ---Theory

10 f-
::;:-

~K.,
~
·~
"a
~:;:: ~:;::
II 5>-- Kxy I~ 5
~ 0
D D
G ~

Kyy
0
~ Kyx
-5 f-1 I -5
I I I
.01 .02 .05 .1 .2 .5 .08 .2 .5

b S=~.N( Cp
R)'
c
S=y_N
Pm
(.R.)'
Cp

Fig. 3.6.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

214
 Test Bearing No. 6

Fig. 3.6.2. Cross section of bearing test rig

CD Test rotor ~ Air bellows (static load)
<2) Test bearing ® Hydraulic exciter head (dynamic load)
Q) Bearing casing (f) Support dynamic load
@ Support bearing
Diameter of test journal: Mean 508 mm, Max. 812.8 mm Static load : Max. 1000 kN
Speed range: 200 ~ 4000 rpm Dynamic load : Max. 89 kN
DC ~ 200 Hz, hydraulic exciter

Outline of Test Rig

a) Measuring method and provision of static and dynamic loads
Static load is supplied by three sets of compressed-air bellows attached to the bearing casing both horizon-
tally and vertically. For the excitation test the static load which is more than twice the dynamic load
supplied by the hydraulic exciter, is added as the bias load. Dynamic load is supplied by the hydraulic
exciter head located at the bearing casing 45° from horizontal. Static load is calculated from the pressure
of the bellows. Bias load and dynamic load are measured by the load cell attached to the hydraulic exciter
head.
b) Measuring method for static displacement and vibration response
Static displacement and relative vibration between test bearing and rotor are measured by means of eight
eddy current proximity sensors mounted horizontally and vertically on each side of the test bearing.
Vibrations of the journal at all support bearing positions are also measured with eddy current proximity
sensors mounted horizontally and vertically for correction of test rotor vibration.
c) Method of data processing
Sinusoidal excitation is applied in the steady-state bearing condition at predetermined speed and load, and
response displacement is measured. In this case, more than two kinds of exciting frequency are applied in
each exciting test. All data (exciting frequency, displacement, temperature and etc.) is stored online by the
mini-computer. A pair of data sets recorded at the same frequency ratio is processed in order to get eight
oil film coefficients.

215
Test Bearing No. 7

Test Bearing No. 7: Two-lobe Bearing

D = 100 mm, L = 50 mm, CP = 0.3 mm, mP = 0.67

Table 3.7.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
ratio Kxx Kxy Kyx K,, cxx cxy c,x c,,
s e cjJ deg. q cm fs
3 Toe
1 0.0058 1.01 28 230 47.3 23.8 2.50 2.02 1.57 1.32 -0.294 -0.020 0.239
2 0.0073 1.01 33 230 46.6 26.9 3.39 2.58 1.75 1.86 -0.319 0.060 0.287
3 0.0097 1.02 38 230 46.5 30.2 4.59 2.90 1.97 3.04 -0.264 0.208 0.392
4 0.0097 0.98 34 288 52.2 20.1 3.17 1.83 1.60 1.94 -0.494 -0.053 0.365
5 0.0123 1.00 40 292 50.9 21.1 4.13 2.01 1.87 2.91 -0.424 0.181 0.471
6 0.0123 0.92 39 338 57.0 18.4 3.82 1.78 1.77 2.55 -0.588 -0.007 0.483
7 0.0145 0.84 43 405 67.9 16.3 4.71 2.18 2.28 3.23 -0.633 -0.252 0.630
8 0.0149 0.86 40 375 60.2 17.7 4.51 1.94 2.11 3.32 -0.979 -0.032 0.631
9 0.0150 1.06 46 233 45.9 33.4 6.00 2.99 2.35 5.50 0.018 0.622 0.624
10 0.0156 0.95 43 353 56.3 19.1 4.45 1.65 1.95 3.39 -0.426 0.162 0.659
11 0.0175 1.01 46 292 49.3 22.0 4.96 1.95 2.17 4.37 -0.346 0.471 0.692
12 0.0185 0.89 52 405 66.8 16.6 4.93 1.94 2.40 4.12 -0.580 -0.134 0.964
13 0.0189 0.89 46 377 59.8 19.0 5.28 2.29 2.51 4.53 -0.542 0.065 0.704
14 0.0209 0.98 50 358 55.3 19.5 5.19 1.50 2.19 4.82 -0.247 0.286 0.869
15 0.0244 0.94 53 380 60.0 17.7 5.27 1.38 2.35 5.94 -0.326 0.045 1.09
16 0.0251 0.95 59 405 65.6 16.4 5.34 1.63 2.58 5.71 -0.284 -0.250 1.24
17 0.0272 1.01 52 297 48.1 21.7 5.70 1.22 2.44 6.70 -0.074 0.588 1.12
18 0.0323 1.03 55 241 43.7 24.0 6.20 1.01 2.91 9.65 0.859 0.800 1.28
19 0.0334 0.99 57 353 53.6 18.4 5.81 0.719 2.59 7.24 0.027 0.410 1.41
20 0.0383 0.99 62 370 58.0 17.1 5.62 0.613 2.80 7.52 0.065 0.028 1.72
21 0.0384 0.96 65 400 64.9 16.1 5.22 0.654 2.72 7.18 -0.293 -0.542 1.83
22 0.0572 0.99 63 297 47.4 19.0 5.85 0.243 3.32 10.7 0.403 0.545 2.11
23 0.0651 0.99 61 233 43.2 21.6 5.13 1.17 4.36 12.8 0.676 0.752 2.11
24 0.0722 0.98 66 350 51.2 17.2 5.17 - 0.476 3.34 10.7 - 0.063 -0.288 2.58
25 0.0828 0.92 68 367 56.3 14.8 5.38 - 1.65 3.76 9.79 - 0.813 -0.622 3.29
26 0.0831 0.89 73 387 63.8 17.0 4.68 -2.02 3.55 8.45 -1.34 -3.03 3.55
27 0.1201 0.87 66 300 46.2 19.0 5.34 -0.343 5.13 14.3 -0.537 -0.991 3.72
28 0.1398 0.77 67 333 50.7 17.5 4.60 -1.99 5.07 14.4 -1.48 -2.66 4.40
29 0.1593 0.73 72 347 55.6 15.3 8.04 -2.88 4.63 15.7 0.482 -3.43 4.61
30 0.1658 0.71 77 363 62.1 19.5 4.54 -5.74 5.38 13.9 -3.72 -6.13 6.70

216
 Test Bearing No. 7

Table 3.7.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wDf2n Hz WkN PmMPa p 1 MPa T;. oc T.•• oc
1 0.0058 49.9 190 23.8 4.77 0.142 40.7 43.1
2 0.0073 49.7 190 19.3 3.87 0.139 40.4 42.6
3 0.0097 50.0 190 14.7 2.9 4 0.136 40.9 42.8
4 0.0097 100 190 23.9 4.78 0.142 40.1 44.8
5 0.0123 98.6 190 19.2 3.85 0.143 39.3 43.8
6 0.0123 150 190 23.9 4.78 0.141 40.5 48.1
7 0.0145 250 190 23.9 4.78 0.150 40.8 56.5
8 0.0149 200 190 23.9 4.77 0.145 36.7 49.2
9 0.0150 50.0 190 9.74 1.95 0.143 40.9 42.6
10 0.0156 150 190 19.3 3.86 0.143 41.0 48.1
11 0.0175 100 190 14.6 2.92 0.142 37.6 42.4
12 0.0185 250 190 19.4 3.88 0.149 40.9 56.3
13 0.0189 200 190 19.0 3.80 0.145 38.0 49.9
14 0.0209 150 190 14.7 2.95 0.139 41.0 47.9
15 0.0244 200 190 14.6 2.93 0.142 40.0 51.0
16 0.0251 250 190 14.8 2.96 0.147 41.1 55.9
17 0.0272 99.1 190 9.74 1.95 0.140 37.0 41.9
18 0.0323 49.8 190 4.95 0.990 0.140 39.5 41.1
19 0.0334 150 190 9.84 1.97 0.140 40.3 47.1
20 0.0383 199 190 9.90 1.98 0.133 38.4 49.7
21 0.0384 249 190 9.82 1.96 0.140 41.6 55.7
22 0.0572 99.9 190 4.83 0.966 0.136 36.7 41.7
23 0.0651 50.0 190 2.52 0.503 0.136 39.5 41.1
24 0.0722 150 190 4.94 0.987 0.142 37.2 44.7
25 0.0828 200 190 4.84 0.967 0.142 38.4 49.4
26 0.0831 251 190 4.76 0.952 0.145 40.8 54.6
27 0.1201 100 190 2.41 0.483 0.140 36.5 41.1
28 0.1398 150 190 2.60 0.520 0.141 40.4 46.3
29 0.1593 199 190 2.56 0.511 0.141 38.6 49.4
30 0.1658 250 190 2.51 0.503 0.136 39.2 53.7

Notes:
• Properties of the lubricant
# 90 turbine oil
Specific gravity @ 20/4 oC 0.865
Kinetic viscosity mm2 f s @ 40°C 31.64
@ 50°C 21.00
@ 100 °C 5.267
• Effective oil film temperature and equivalent viscosity for the calculation of Sommerfeld number. The average bearing metal
temperature over 20 points is assumed to be the effective temperature of the oil film, and the equivalent viscosity is calculated on
the basis of the effective temperature.
• Determination of exciting force. The ratio of vibration amplitude to bearing clearance is adjusted within the range of 1/6 to 1/20
by changing the exciting force (1 ~ 3.5 kN, peak).

217
Test Bearing No. 7

0
1.0 0 0 0
0 0 0 0 0 ~0 100°
0
00 cfJ
8CID
.8 ~ 80°
0 : <
} Experiment
b. : ~

.6 --Theory 60°

....
.4 40°

I>

.2 20°

0 00
.002 . 005 .02 . 05 .1 .2
Bearing material : White metal ASTM B23
S=y}l ( Jl.)' Backing metal SS41
a pm Cp
Bearing geometry

0 : Kxx b. : Kxy }
Experiment
<> : Kyx o :K,,
0 : Cxx
---Theory b. : Cxy } Experiment
() : Cyx 0 . c,,
---Theory
0

0 0
Cxx
0 ·~

0 a"
Uis
0 0
~~ 0

Uis 0
0
co 0
Kxx II ·~

0 u
:J
0
b. oS 0

fig~ Kxy Cyy

Ky,
0 0

Kyx
<> <>
e
-5
<>
.005 .02 .1 .2 .002 .01 .02 .1 .2

b
s=Mli(
Pm Cp
1l.)' c
S=y}l ( Jl.)'
P• Cp

Fig. 3.7.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

218
 Test Bearing No. 8

Test Bearing No. 8: Two-lobe Bearing

D = 100 mm, L = 40 mm, CP = 0.280 mm, mP = 0.75

Table 3.8.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
ratio Kxx Kxy Kyx Kyy cxx cxy c,x cyy
s e ¢ deg. q cm 3 /s T °C

1 0.005 64.8 22.1 2.60 1.28 0.40

2 0.0081 59.6 19.8 7.40 0.80 1.52 17.0 2.40 1.20 0.40
3 0.0131 55.1 19.2 7.00 0.40 1.20 15.0 1.84 1.20 0.32
4 0.0294 48.6 17.6 6.60 0.0 1.20 11.6 0.80 0.0 0.64
5 0.0463 47.6 16.8 6.48 0.0 1.12 11 .8 1.00 0.0 0.96
6 0.0781 48.3 16.0 6.20 -0.20 1.04 13.6 1.40 -0.80 0.60
7 0.148 48.5 14.6 7.20 -2.00 1.40 17.3 1.60 -0.48 1.00

Table 3.8.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wv/2n Hz WkN Pm MPa p 1 MPa T;. oc 7;,., oc
1 0.005 60 80 18.865 4.911 0.102
2 0.008 59.5 80 13.720 3.571 0.102
3 0.013 58.5 80 10.290 2.679 0.102
4 0.029 58 80 5.145 1.339 0.102
5 0.046 59 80 3.430 0.893 0.102
6 O.Q78 59 80 2.058 0.536 0.102
7 0.148 60.5 80 1.029 0.268 0.102

Notes:
• Properties of the lubricant
Turbine oil ISO 32
Specific gravity @ 20/4 oc 0.87
Kinetic viscosity mm 2 /s @ 30 °C 43.7
@ 50 °C 19.2
@ 100 °C 6.09
• Equivalent viscosity to calculate Sommerfeld number. Viscosity at mean temperature of bearing surface was used.
• Eccentricity ratio and attitude angle are measured independently of oil film dynamic characteristics.

219
Test Bearing No. 8

1.6r-----------------------------------~

1.4

1.2

1.0 100°
0 : £ }
"' A : (> Experiment
80°
.8
-Theory

.6 60°

....
.4 40°

.2 20°

0 0
.002 .2

S=MN(Jl.)'
a P• Cp

25 25r-----------------------------------~

0 Kxz
0

0: Kxx 6: Kxy}
() : Kyz a .. K yy Experiment
- - - Theory ·~ 0 0
u
a
rj:s:
II
0: Cxx 6 : Cxy}Experiment
1:. ·~ () : Cyx D: Cyy
G
--Theory
Kxy Cyy
1:.
Kyy
()
D
0

Kyx -5

.002 .02 .2
S=P-N (JL)'
P. Cp
b c

Fig. 3.8.1a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

220
 Test Bearing No. 8

Fig. 3.8.2. Cross section of bearing test rig.

CD Test rotor ® Bellows cylinder (vertical)
(2) Test bearing (J) Support bearing
Q) Bearing housing ® Oil supply to support bearing
@ Capacitative proximity sensor ® Oil supply to test bearing
<3) Bellows cylinder (horizontal)
Diameter of test journal : 100 ~ 140 mm Dynamic load: 0.05 ~ 0.4 kN
Speed range: 300 ~ 12000 rpm 2 directions (horizontal and vertical)
Hydraulic exciter

221
Test Bearing No. 9

Test Bearing No.9: Tilting Pad Bearing with Four Pads

D = 140 mm, L = 95 mm, CP = 0.175 mm, mP = 0.0, p = 73.7°,
r:t/fJ = 0.5, LBP

Table 3.9.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
Kxx Kxy Kyx Kyy cxx Cxy Cyx Cyy
ratio
s e cjJ deg. q cm 3 js roc

1 0.033 800 59.3 2.598 -0.048 -0.223 2.149 0.306 0.095 -0.117 0.507
2 0.057 775.0 55.3 2.653 -0.0040 -0.182 2.211 0.382 0.168 -0.232 0.766
3 0.087 766.7 52.8 2.675 0.054 -0.211 2.245 0.362 0.078 -0.300 1.009
4 0.214 750.0 47.8 2.800 0.135 -0.202 2.378 1.061 -0.017 -0.403 1.672
5 0.305 716.7 46.3 2.717 0.189 -0.180 2.344 1.457 -0.143 -0.369 1.989
6 0.442 750.0 45.3 2.805 0.198 -0.277 2.391 1.915 -0.230 -0.288 2.442
7 0.620 750.0 46.3 2.933 0.159 -0.544 2.544 2.650 -0.346 -0.563 3.280
8 1.444 733.3 37.5 3.625 0.118 -0.957 3.135 3.980 - 0.423 -0.890 4.248
9 3.652 716.7 32.5 4.849 0.052 -1.383 4.189 5.982 -0.756 -1.288 5.795

Table 3.9.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps wDf2n Hz WkN PmMPa p 1 MPa T;n °C T.•• oc
1 0.033 52.0 15.0 : 30.0 39.2 2.95 0.10 28.0 31.3
2 0.057 56.7 15.0:30.0 29.2 2.2 0.097 25.5 28.9
3 0.087 58.3 15.0:30.0 22.5 1.69 0.097 25.9 29.4
4 0.214 63.6 15.0:30.0 12.4 0.933 0.097 26.1 30.2
5 0.305 65.9 15.0:30.0 9.46 0.711 0.097 24.4 28.2
6 0.442 71.7 15.0 : 30.0 7.44 0.560 0.10 26.3 30.3
7 0.620 75.8 15.0 : 30.0 5.361 0.403 0.10 26.1 30.7
8 1.444 78.8 15.0:30.0 3.42 0.257 0.097 25.7 29.8
9 3.652 61.7 15.0:30.0 1.74 0.130 0.10 26.1 28.8

Notes:
• Properties of the lubricant
# 90 turbine oil
Specific gravity @ 15/4 °C 0.866
Kinetic viscosity mm 2 js @ 37.8 °C 28.66
@ 50.0 °C 19.90
@ 98.9 °C 4.906
• Equivalent viscosity for Sommerfeld number. The effective film temperature is assumed to be the average of two temperatures
measured at points each 25° from vertical (x-axis) by means of CA (chromel-alumel) thermocouples 3 mm from the bearing metal
surface. Effective viscosity is calculated from the effective film temperature.
• Determination of the exciting force. The ratio of vibration amplitude to bearing clearance is adjusted within the range of 1/6 - 1j7
by changing the exciting force.

222
 Test Bearing No. 9

r---{ IT·--

Bearing material: White metal WJ1

Backing metal S45C
Bearing geometry

35 1- 35

30 f- 0: Kxx 6 : Kxy } 30 0 : Cxx 6 : Cxy }

Experiment Experiment
o: Kyy 0 ; Cyx o : Cyy
----Theory Theory
25 1- 25
(Kxy=Kyx= 0) (Cxy=Cyx=O)

20 f- 20
.:;
·~

... ·~ a
~~
~~ 15 f- II
II ·~
15
~
·~
0

10 - 10

51-
El
Kxx= Kyy 0 5
0

8 8 8~ 0 8 8
e e~-~
0 0
~
A
0 ~
0
0 ~ ij ~ ¢ ¢
~
0

-5 -
I I l_ I
.02 .05 .1 .2 .5 2 .02 .05 .1 .2 2

b
s=0:!..(ll.)'
P• Cp c s=0:!.. (.fl.)'
P• Cp

Fig. 3.9.1 b, c. Experimental results for oil film coefficients. b Dimensionless stiffness coefficients. cDimensionless damping coefficients

223
Test Bearing No. 10

Test Bearing No. 10: Tilting Bad Bearing with Five Pads
D = 100 mm, L =50 mm, CP = 0.216 mm, mP = 0.51, p= 60°,
r:t/p = 0.5, LOP

Table 3.10.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
ratio Kxx Kxy Kyx Kn cxx Cxy cyx cyy
s e cP deg. q cm 3 js T °C

1 0.029 0.85 533 53 .6 10.35 -0.17 -0.45 1.23 1.43 0.07 0.17 0.929
2 0.040 0.79 553 57.7 9.44 -0.22 -0.57 1.72 1.54 0.04 0.16 1.26
3 0.046 0.77 517 53 .0 11.16 -0.23 -0.84 1.70 2.25 0.12 0.30 1.34
4 0.048 0.71 558 59.3 9.17 -0.15 -0.59 2.00 1.27 0.03 0.05 1.41
5 0.052 0.68 558 62.4 9.09 -0.34 -0.46 2.12 1.55 -0.07 -0.41 1.83
6 0.052 0.61 553 70.4 9.03 -0.18 -0.54 2.91 1.32 1.06 -0.08 2.10
7 0.054 0.65 558 66.4 9.37 -0.40 -0.64 2.63 1.48 0.20 -0.32 2.05
8 0.065 0.71 517 51.0 9.95 -0.17 -0.88 2.17 2.54 0.16 -0.01 1.72
9 0.065 0.68 538 56.7 10.16 -0.31 -0.64 2.32 2.18 0.24 0.13 1.74
10 0.073 0.63 555 58.9 9.56 -0.39 -0.97 2.94 1.48 0.88 0.08 2.07
11 0.083 0.61 563 60.9 10.54 -0.69 -0.58 3.16 2.54 0.10 -0.58 2.51
12 0.088 0.62 533 54.4 10.00 -0.37 -0.82 2.99 2.64 0.26 0.02 2.48
13 0.089 0.58 557 63.4 10.24 -0.52 -0.58 3.66 2.56 0.74 -0.57 2.99
14 0.090 0.55 547 67.0 10.42 -1.41 -0.47 3.56 3.89 0.59 -0.45 3.09
15 0.098 0.62 517 51.7 10.12 -0.20 -0.79 2.94 3.56 0.40 0.15 2.39
16 0.106 0.58 553 56.9 9.52 -0.54 -0.73 3.28 2.25 0.27 -0.24 2.79
17 0.117 0.54 558 59.1 10.40 -0.40 -0.93 4.27 2.19 1.59 -0.28 3.35
18 0.126 0.52 555 61.5 12.00 -0.89 -1.29 5.20 3.37 1.87 -1.15 3.62
19 0.136 0.47 542 64.2 11.28 -0.95 -0.60 5.52 4.37 2.20 -1.49 4.12
20 0.142 0.53 600 54.2 9.88 -0.04 -0.10 4.20 3.92 0.62 -0.14 3.82
21 0.160 0.48 558 54.3 10.32 -0.87 -0.79 4.84 3.79 1.00 0.0 4.69
22 0.191 0.46 555 56.5 12.16 0.13 -1.23 5.76 2.96 1.39 -0.58 4.20
23 0.198 0.48 533 50.4 9.92 -0.15 -0.78 5.54 5.74 0.77 -0.02 4.27
24 0.207 0.40 525 61.6 13.36 -0.25 -0.72 7.20 6.77 2.43 -2.34 5.97
25 0.210 0.42 555 59.1 16.6 1.91 0.44 7.56 4.56 1.77 -2.81 4.87
26 0.298 0.39 508 52.2 12.00 -0.45 -0.48 7.97 6.50 1.58 -0.05 6.13
27 0.374 0.37 558 53.5 12.56 -0.76 -1.59 9.60 7.09 1.87 0.32 8.46
28 0.422 0.33 558 53.8 15.04 2.11 -2.73 11.92 7.11 2.99 - 0.70 8.71
29 0.463 0.31 558 56.1 18.40 1.68 -2.83 12.72 9.61 6.87 -2.21 9.43
30 0.494 0.29 537 57.6 17.76 5.41 -3.13 14.24 10.55 7.39 -4.09 8.89

224
 Test Bearing No. 10

Table 3.10.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps Wv/2n Hz WkN Pm MPa p 1 MPa T;. oc T.ut oC
1 0.029 83.3 190 14.7 2.94 0.140 40.0 45.4
2 0.040 133 190 14.7 2.94 0.139 39.7 48.3
. 3 0.046 83.3 190 9.80 1.96 0.136 40.5 45.8
4 0.048 167 190 14.7 2.94 0.135 39.0 49.2
5 0.052 200 190 14.7 2.94 0.135 40.0 52.9
6 0.052 267 190 14.7 2.94 0.132 40.3 63.6
7 0.054 233 190 14.7 2.94 0.129 40.3 57.9
8 0.065 83.3 190 7.35 1.47 0.139 40.0 43.3
9 0.065 133 190 9.80 1.96 0.137 39.5 48.8
10 0.073 167 190 9.80 1.96 0.132 40.5 50.5
11 0.083 200 190 9.80 1.96 0.131 39.5 53.9
12 0.088 133 190 7.35 1.47 0.136 39.5 47.1
13 0.089 233 190 9.80 1.96 0.131 40.0 57.1
14 0.090 267 190 9.80 1.96 0.135 40.0 62.9
15 0.098 83.3 190 4.90 0.980 0.140 40.5 45.7
16 0.106 167 190 7.35 1.47 0.133 38.5 49.3
17 0.117 200 190 7.35 1.47 0.133 40.0 52.7
18 0.126 233 190 7.35 1.47 0.133 40.0 56.7
19 0.136 267 190 7.35 1.47 0.137 40.0 62.0
20 0.142 133 190 4.90 0.980 0.137 39.0 47.9
21 0.160 167 190 4.90 0.980 0.133 38.0 48.2
22 0.191 200 190 4.90 0.980 0.135 39.0 52.7
23 0.198 83.3 190 2.45 0.490 0.140 40.5 45.4
24 0.207 267 190 4.90 0.980 0.134 40.0 62.2
25 0.210 233 190 4.90 . 0.980 0.135 40.0 56.1
26 0.298 133 190 2.45 0.490 0.137 39.0 47.7
27 0.374 167 190 2.45 0.490 0.135 39.0 49.1
28 0.422 200 190 2.45 0.490 0.135 39.0 51.3
29 0.463 233 190 2.45 0.490 0.135 40.5 55.7
30 0.494 267 190 2.45 0.490 0.139 39.5 62.2
Notes:
• Properties of the lubricant
# 90 turbine oil
Specific gravity @ 20/4 °C 0.865
Kinetic viscosity mm 2 /s @ 40 °C 31.64
@ 50 °C 21.00
@ 100 °C 5.267
• Effective oil film temperature and equivalent viscosity for calculation of Sommerfeld number. The average temperature over five
pads, measured at the pivot position for each pad, is assumed to be the effective temperature of oil film, and the equivalent viscosity
is calculated on the basis of the effective temperature.
• Determination of exciting force. The ratio of vibration amplitude to bearing clearance is adjusted within the range of 1/6 to 1/20
by changing the exciting force.

225
Test Bearing No. 10
1.0 r - - - - - - - - - - - - - - - - - - - - ,

0 o Experiment
Theory

.4 ·+-
.2

.1 .2 .5 2 Bearing material : White metal W87

Backing metal S15C
S=yj'{(ll.)'
P• Cp Bearing geometry
a

0: Kxx .0. : Kxy} Experiment 0 : Cxx .0. : Cxy

Experiment
<> : Kyx c :Kyy <> : Cyx c: Cyy
15
--Theory Theory
(Kxy= Kyx= 0) ( Cxy= Cyx= 0)

...:::-
0
0 ..,.
·~

Uj:;::
0 a
II 0 Uj:;::
c II
:<
·~

0
oc
·~

0 0 c G
0
10
Kxx

.o..~~.

0
+.o.
<> <>
<>
Ooo <>
-5 <> <>
<>

.02 .05 .1 .2 .5 2 . 02 .2 .5 2

S= E!{(_B._)' S= E!{( _B._)'

b P• Cp c P• Cp

Fig. 3.10.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios and attitude angles. b Dimensionless stiffness
coefficients. c Dimensionless damping coefficients

226
 Test Bearing No. 11

Test Bearing No. 11: Tilting Pad Bearing with Five Pads
D = 76.4 mm, L = 31.6 mm, CP = 0.086 mm, mP = 0, fJ = 58°,
a/fJ = 0.5, LOP

Table 3.11.1. Experimental results for oil film coefficients

Test Sommer- Static characteristics Dimensionless stiffness coefficients Dimensionless damping coefficients
no. feld
number Eccen- Attitude Oil flow Temper-
tricity angle rate ature
ratio Kxx Kxy Kyx Kyy cxx cxy cyx cyy
s e ¢ deg. q cm 3 /s PC
1 0.106 0.88 0.6 168 42.0 3.03 -0.093 -0.353 0.395 0.264 -0.080 -0.173 0.344
2 0.144 0.85 1.0 169 41.5 3.38 -0.075 -0.320 0.414 0.234 -0.059 0.020 0.436
3 0.169 0.81 1.2 166 46.8 2.98 0.056 -0.025 0.300 0.286 -0.011 -0.118 0.592
4 0.224 0.74 0.0 167 47.0 3.34 0.058 0.074 0.284 0.769 -0.021 -0.052 0.716
5 0.306 0.69 -1.4 167 46.0 4.04 0.073 0.033 0.275 1.40 0.039 -0.197 1.14
6 0.496 0.65 1.5 167 45.8 4.28 -0.146 -0.226 0.177 2.89 0.007 0.446 1.61
7 0.708 0.54 2.1 167 46.0 3.72 0.494 0.336 0.553 2.95 0.095 0.048 1.90
8 0.786 0.57 1.6 166 44.5 3.75 -0.324 -0.269 0.242 3.66 -0.008 0.632 2.13
9 1.48 0.44 -2.2 166 45.8 2.79 -0.204 -0.088 0.361 3.73 -0.574 0.133 3.42
10 1.76 0.46 -5.5 164 41.5 2.72 -0.551 0.181 0.833 4.13 -0.772 -0.346 3.77

Table 3.11.2. Test conditions

Test no. Sommerfeld Revolutional Frequency of Static load Bearing mean Supply oil Lubricant inlet Lubricant outlet
number speed dynamic load pressure pressure temperature temperature
s N rps Wn/2n Hz WkN Pm MPa p 1 MPa oc r..
7;,., oc
1 0.106 49.7 44.9 9.48 3.92 0.1 40.3 42.0
2 0.144 49.9 44.9 7.14 2.95 0.1 39.5 41.5
3 0.169 99.2 99.9 9.49 3.92 0.1 41.0 46.8
4 0.224 99.8 99.8 7.09 2.93 0.1 41.0 47.0
5 0.306 102 100 4.74 1.96 0.1 41.3 46.0
6 0.496 104 99.7 3.56 1.47 0.1 41.0 45.8
7 0.708 100 99.6 2.36 0.979 0.1 41.0 45.8
8 0.786 103 99.3 2.36 0.977 0.1 40.3 44.5
9 1.48 101 99.9 1.19 0.490 0.1 40.8 45 .8
10 1.76 102 100 1.20 0.495 0.1 38.5 41.5

Not es:
• Properties of the lubricant
# 140 turbine oil
Specific gravity @ 15/4 °C 0.874
Kinetic viscosity mm 2 /s @ 37.8 °C 61.8
@ 50.0 °C 35.4
@ 98.9 °C 7.78
• Measurement method for inlet oil temperature (T;.) and outlet oil temperature (7;,.,). T;. and 7;,., are measured by means of
thermocouples.
• Effective viscosity for Sommerfeld number. The effective film viscosity is assumed to have the value corresponding to the measured
outlet temperature.
• Determination of amplitude ratio and dynamic load. The dynamic loads are varied from 0.423 kN to 2.05 kN with an increase in
the static load. The ratio of vibration amplitude to bearing clearance is adjusted within the range 1 : 11 to 1 : 13.

227
Test Bearing No. 11

l.Or----------------------------------------

0.8c~ ~w
0 Experiment ' 0
0.61- ~ 0
- Theory o~
00

0.4t- ~

0.2

0 I I I 'I
0.02 0.05 0.1 0. 2 0.5 1 2
Bearing material : White metal WJ2
s=Mli (lL)' Backing metal S20C
a P• Cp
Bearing geometry

0: Kxx A : Kxy } 0: Cxx A : Cxy }

Kyy Experiment Experiment
0 : Kyx 0 : 0: Cyx o: Cyy
------Theory ---Theory
( Kxy= Kyx= 0) ( Cxy= Cyx= 0 )

.:,-
·~

s
dl~
II
·~

Cxx

01-------

- 5 - 5

0.02 0.05 0.2 2 0.02 0.2 2

c
S= Mli
P• Cp
(lL) 2

Fig. 3.11.1 a-c. Experimental results for oil film coefficients. a Eccentricity ratios. b Dimensionless stiffness coefficients. c Dimension-
less damping coefficients

228
 Test Bearing No. 11

Fig. 3.11.2. Cross section of bearing test rig.

CD Housing ® Load cell
~ Shaft (J) Displacement sensor
Q) Supporting bearing ® Hydraulic exciter (for dynamic load)
@ Diaphragm coupling ® Accelerometer
(3) Below cylinder (for static load) @ Connecting link
Diameter of test journal: 76.2 mm Static load: Max. 20 kN
Speed range: 1000 ~ 12000 rpm Dynamic load : Max. 8 kN (peak value)
DC~ 500Hz

Outline of the Test Rig

a) Measurement method and provision of static and dynamic loads
The static load is applied to the test bearing floated on the journal from the horizontal (bilaterally) and
vertical directions by means of the bellows cylinder. The dynamic load is applied from two directions 45°
from vertical by means of the hydraulic exciter. Dynamic and static loads are measured by means of load
cells.
b) Measurement method for static displacement and vibration response
The relative displacement between the bearing and the journal is measured from x and y directions by
means of an eddy current gap sensor. The absolute dynamic displacement is measured by means of an
accelerometer.
c) Method of data recording
Data relating to load, displacement, vibration response, etc. is simultaneously recorded by means of an
electromagnetic oscillograph.

229
4. Explanations

4.1 Method of Calculation for Bearing Characteristics

4.1.1 Generalized Reynolds Equation Variation of the lubricant desnsity is disregarded
for Laminar Flow here. Integrating the continuity over the range 0 to
h with respect to y using the boundary condition
specified below:
Bearing characteristics can be calculated from the
pressure distribution of the oil film. v = 0 at y = 0,
In this section, the generalized Reynolds equa- v = vh at y = h
tion, the basic equation for pressure distribution, we get the following equation:
for laminar flow is derived. The isoviscous case is
also derived from the generalized case. The isovis- a h a h ah
cous Reynolds equation is used in order to calculate -a Judy+ -a Jwdy- uh -a + vh = 0 (4)
X 0 Z 0 X
the data in this databook. The Reynolds equation
for turbulent flow will be shown in Sect. 4.1.2. where v represents the film velocity component of
The equation of motion of the lubricating film the y-direction.
can be derived from the Navier-Stokes equation Substitution of Eq. (3) into Eq. (4) yields the gen-
and is expressed as follows: eralized Reynolds equation for iaminar flow.

(/l
With the following dimensionless variables
ap = ~ au) ap- a (flaw) (1) f) = x jr, 1J = yjh, ( = zjr
ax ay ay ' az - ay ay
u = uj(rw), v = vj(rw), w= wj(rw)
0

where uh = uh j(r (J) ), vh = vh j(r (J))

x: coordinate in the sliding direction 1f1 = cj r
y: coordinate in the film thickness direction A = p lf/3 /(/lin W)
z: coordinate in the axial direction
p: film pressure fl = fl fflin
f1 : viscosity coefficient the dimensionless generalized Reynolds equation is
u: film velocity component of x-direction obtained as shown below:
w: film velocity component of z-direction
The viscosity coefficient fl is a function of x, y and a ( 3 aA) a ( 3 aA)
ae H E ae + ac H E ac (5)
z in Eq. (1). Boundary conditions of the film velocity
are given by Eq. (2): _ a auh _ aH _1 _
= uh ae (H G) + H G · 8if - uh 8if + If/ vh.
u = 0, w = 0 at y = 0 (bearing metal surface),
u = uh, w = 0 at y = h Uournal surface) . (2) Both E and G in Eq. (5) are the functions of film
viscosity and they are given by the following equa-
Film velocity components u and w are expressed by tions:
Eq. (3), below, which is obtained by integrating E(fJ, () = cP 3 (fJ, 1, ()- cP~(fJ, 1, ()/cP 1 (fJ, 1, (),
Eq. (1) twice with respect toy under the boundary (6)
condition specified by Eq. (2) G(fJ, () = 1 - cP 2 (fJ, 1, ()/cP 1 (fJ, 1, (),

ap
u= -
(yJ-Y dy - J0h-flY dy J0 -dyfl j J0h -dy)
y
where

ax 0 fl fl
y dylh dy
+ Uh· J- J -, (3) (7)
ofl ofl

w = ap
az
(s l:
0 fl
dy -1 ydy
0 fl
J dy/1 dy)
0 fl 0 fl
0

231
4 Explanations

The Reynolds equation for the steady state con-

dition is obtained by setting e = 0 and </> = 0 in 1 1

Eq. (10):
0

a ( 3 aA) a ( 3 oA) aH
aB H aB + ac H ac = 6 ao · (11)

Static characteristics of the bearing performance,

such as the bearing load-carrying capacity, the fric-
tional coefficient and the oil flow rate, can be ob-
tained by solving this Reynolds equation.

4.1.2 Reynolds Equation

for Turbulent Flow
Fig. 4.1.1. Velocity components of the point OP on the journal
surface
Variation in the flow condition, and therefore the
change from laminar to turbulent flow, brings
about a perceptible change in bearing performance,
e.g. in load capacity and friction coefficient. There-
For the isoviscous case, values of E and G are con- fore the turbulent flow theory of bearings is indis-
stant at 1/12 and 1/2, respectively. pensable for bearings with higher speed. Many in-
Dimensionless velocity components iih, vh at the vestigations have already been performed on this
journal surface can be determined from the geomet- subject. This section shows how the Reynolds equa-
rical relation indicated in Fig. 4.1.1: tion for the turbulent flow condition is derived, on

iih = 1 - If! eq>' cos() + 1f1 e1 sin()

(8)
the basis of the mixing length theory.
Components of the shearing stress rx, ry for tur-
bulent flow are expressed as follows, the inertia
vh =If!(~:+ e</> sinO+ e cos())
1 1 force of the lubricant film being disregarded:

where rx = f1
au
ay- QU V'
I I

(12)
e = ejC: eccentricity ratio
</>: attitude angle
rz = f1
aw
ay - (! v w'
I I

e </> derivatives of e, </> with respect to di-

1
,
1
:

mensionless time r = w t where, U V W are velocity components caused by

1
,
1
,
1

Substituting Eq. (8) into Eq. (5), with some omis- the turbulent flow and Q is the density of the lubri-
sion we get the following equation (If/ ~ 1 is taken cant. The second term on the right-hand side of
into consideration): Eq. (12) is called the Reynolds stress. Shearing
stress for the laminar flow is obtained by setting

:o (H E ~~) + :, ( H E ~~) - QU V = 0 and - QV W = 0.

1 1 1 1

3 3
Reynolds stress is given by the following equa-
tion based on the mixing length model of Prandtl
= :() (H G) + e</> sin 0 + el cos 0.
1 (9) [4]. Flow in the sliding direction is assumed to be
predominant
The Reynolds equation for the isoviscous case
is obtained by setting E = 1/12 and G = 1/2 in - eu v I I = e[2 . 'au' au
oy . oy,
(13)
Eq. (9):
I
ay . aw
- ev w = e[2 . 'awl ay,
I

:() ( H3 ~~) + :( ( H3 ~~) where 1is the mixing length.

oH . (10)
Substituting Eqs. (12) and (13) into the equation
=6 oO + 12(e</>1 smO- e cosO).
1
for motion of the film

The stiffness and damping coefficients in this data- (14)

book are obtained by using the solution of Eq. (10).

232
 4.1 Method of Calculation for Bearing Characteristics

we get: that rxfr 0 exists within a limited region. r 0 and Th

are the shearing stresses at y = 0 and y = h, respec-
ap = ay
ax j_ {(fl. + Q [2\au\)
ay ayau} ' tively.
The Reynolds equation for the turbulent flow is
(15) derived by substituting the averaged flow velocities
j_ {(fl.+ eF\aw\)
ap = ay
az ay aw}·
of Eq. (16) into the equation of continuity:
oy
This is the equation of motion for the turbulent
aea (H 3 Gx aA)
ae + a(a (H 3 Gz aA)
a(
flow.
As Eq. (15) is unwieldy as it stands, it is simplified =!2 aH
ae + (uf/ sin()- e' cosO). (18)
in various ways [5-9]. Constantinuescu [5] evaluat-
ed velocity distribution for the turbulent flow. This
study gives two flow regions, the quasi-laminar flow
region near the wall, where the theory for laminar 4.1.3 Method of Calculation
flow is applicable, and the perfect turbulent flow for Reynolds Equation
region, where the velocity distribution is approxi-
mated by a logarithmic one. Boundary velocity be- Various numerical methods of solving Reynolds
tween both regions is equalized so as to be smooth- equation have already been developed. They in-
ly connected at the boundary. Ng and Pan [6] clude the finite difference method (FDM), the finite
introduced the turbulent kinematic viscosity e in- element method (FEM) and the methods using in-
stead of the mixing length. e is so determined that finite series function or Green's function. In this
the velocity distribution becomes consistent with section, the numerical procedure used to solve
that obtained with reference to the wall theory. Reynolds equation [10] by FDM is described.
Aoki and Harada [7] calculated the mixing length 1. When the film pressure, A, is calculated numeri-
It is so determined that the velocity distribution is cally, a numerical error increases in the neighbour-
consistent with that obtained with reference to the hood of the minimum film thickness position, be-
wall theory. cause the pressure increases rapidly in that region.
Integrating the velocity distribution obtained in The corrected pressure function, A, is commonly
these manners with respect to y, averaged film ve- introduced to reduce the numerical error. A is given
locities urn and wrn are obtained as follows: by the following form [10]:
h ap u
2

urn=- Gxj; ax+ 2' (16) aea (H 3

Gx
aA)
ae + a(a (H 3
Gz
aA)
a(
h2 ap
wrn = - Gzj; az' =!2 oH
ae + (e¢' sinO+ e' cosO). (19)

where Gx and Gz are called the turbulent flow coef- Applying the transform by Eq. (19) into Eq. (10),
ficients and are the correctional terms of viscosity the following equation is obtained :
caused by the turbulent diffusion.
The turbulent flow coefficient is given by the fol- az A a2 A
-2+ - + C1 -
aA + C -aA + C A- =C .(20)
lowing form by using the local Reynolds number, ae a(
2 ae 2 o( 3 4

Re1:
1 Generally, coefficients C 1 and C 2 are not equal to
zero. cl and c2 are zero only in the isoviscous case.
Gx = 12(1 + rxxRefx}'
(17) C 3 and C4 are given by Eq. (21) for the isoviscous
1 case:
Gz = 12(1 + rxzRef'} ·

Wada and Hashimoto [8] obtained the relation- (21)

ship between the average flow velocity and the
shearing stress at the wall by investigating a large C4 = 6H -~2 {aH . e' cos
ae + 2(e¢' sm(J- e)} .
number of experimental results already published.
They introduced a model of the equivalent pressure Referring to Fig. 4.1.2, we have the dimensionless
flow and gave the equation for Gx and Gz based on pressure, Aii• for each point of the mesh over the
that model. Elrod et al. [9] obtained the relationship () - (plane. The subscripts i,j are grid line numbers
between the shearing stress ratio, rh/r 0, and the in the () and ( directions, respectively. The partial
turbulent flow coefficient under the assumption derivatives of Aii are expressed as follows [11] with

233
4 Explanations

where (n) is the number of repetitions and l/.. repre-

sents the acceleration coefficient.
t=L/r~------------------~ The iterative process is repeated until the numer-
L18 L18 ical error is smaller than the prescribed value Jc.
j+ 1
The error is defined by
No

x.
Azj

"''
(n)
j M-1 N-1
L L
No
1- +'~) < J c • (26)
j- 1
"'' i=2 j= 2
(n

Aii
i- 1 i+l
t= 0 The allowable error, J 0 , must by very small, in the
~--------------------~---8 order of 10- 6 "' 10 -!2, to ensure a good level of
8= 0 8= 2 1r
Fig. 4.1.2. Finite-difference mesh accuracy in the calculation.

finite increments A(} and A(:

a.Aij 1 _ _
80 = 2A(} (A;+t.i- A;-t),
4.1.4 Boundary Condition
for Oil Film Pressure
a.Aij 1 _ _
----a{= 2 A( (A;,i+ 1 - A;,j-t),
(22) The film pressure at the bearing edges must be
a.Aij 1 _ _ _ known before Reynolds equation can be solved. In
802 = A(J2 (A;+t,i- 2A;,i + A;-t), the case of partial arc bearings with a smaller arc
angle or tilting pad bearings, the pressure boundary
u::12Aii
-
1 _ _ _
8(2 = A(2 (A;,i+1- 2A;,i + A;,j- 1). condition is simple. The film pressure is generated
over the whole of the bearing area, and therefore
the boundary pressure is taken as the atomospheric
By applying Eq. (22) to Eq. (20), we get
pressure. In contrast, in the case of full circular
bearings or partial arc bearings with larger arc an-
Aii = D!;~(C4ii- D2ii.Ai+1 , j- D3ii.Ai-t,i
gles, the boundary condition is not so simple, be-
- D4iiAi,i+l- Ds;iAi,j-1) (23) cause of the ruptured region or the cavitation re-
where gion of the film. Moreover, in the case of the
2 2 unsteady state, no accurate boundary conditions
Dtii = C3ii- A(J2- A(2' have been proposed so far.
First, the boundary condition for the steady state
1 Cuj 1 C 1 .. is investigated. Figure 4.1.3 shows the circumferen-
- -A(J2 +--
D2ii- 2AO'
D3ii-
-A(J2
-- - -
2AO'
'1 (24)
tial boundary conditions commonly used in the full
circular bearings. Figure 4.1.3 a shows the Sommer-
D _1_ - c2 ij feld boundary condition. The pressure calculated
sij- A(2 2A(
using this condition is point-symmetrical of(} = n .
C 1 ,;i' C 2 ii and C 3;i are the values of Cu C 2 and C 3 Although a negative pressure in the same order as
in Eq. (21) at each mesh point (i,j), respectively. the positive one is generated in the region of
M and N represent the total numbers of each grid n < (} < 2 n, the pressure in this region is actually
line for the 0 and ( directions, respectively, and the atomospheric pressure, because of the film rupture.
boundary values of film pressure at i = 1, M and This condition is seldom applied except when
j = 1, N are known, the (M - 2) x (N - 2) simulta- higher pressure is supplied. Figure 4.1.3 b shows the
neous equations concerning A are obtained corre- Gumbel boundary condition. The negative pressure
sponding to (M - 2) x (N - 2) mesh points. The in the Sommerfeld boundary condition is replaced
pressure is calculated by using an iteration method. in this case by atmospheric pressure. Although this
The over-relaxation method is commonly used be- condition is physically inappropriate, because the
cause of its rapid convergency. flow continuity is not satisfied at (} = n, it has been
Application of this method to Eq. (23) yields the used so far because of its simplicity. The Reynolds
following equation for determination of A ii: boundary condition shown in Fig. 4.1.3 c mitigates
(n + 1) (n) (n) this defect of the Gumbel condition and is com-
- - -1 -
Aii =A;i + ('/. {D1ii(C4ii- D2iiAi+1,i monly used at present. It is used in the calculations
(n+ 1) (n) in this databook.
- D3ii A;-t ,i - D4iiAi,i+1 The Reynolds boundary condition gives a zero
(n + 1) (n) pressure gradient at the point where the film pres-
-Dsii A;,j - 1 ) - Aii} (25) sure falls to the same as the ambient pressure. This

234
 4.1 Method of Calculation for Bearing Characteristics

p~----------------~h
L! D=O. 7
e=O.S
:
+
:z::
II
I'<:

"....
~
"'0.
....
0
8
e -0 . 1 Circumferential
"'"' angle 8
a Sommerfeld boundary condition "'
"2
0
-0.2
·;:;; -0 .3
p~----------------,h c
8"' -0.4
6
-0.5
-0.6 Isoviscous theory
-0.7

Fig. 4.1.4. Determination of Reynolds boundary condition

conditions for the wedge film pressure and the

squeeze film pressure are used [12, 13].
b Gumbel boundary condition As an example, for a circular journal bearing, the
pr-----------------~h dimensionless film thickness H is expressed as fol-
lows:
H = 1 + e cos 0. (28)
Thus, Reynolds equation [Eq. (10)] is rewritten as
follows :
2 n: h= 0

:o ( H3 ~~) + :( ( H3 ~~)
= - 6e(1 - 2¢') sinO- 12e' cosO. (29)
c Reynolds boundary condition The first term on the right-hand side of Eq. (29)
represents pressure generation by the wedge film
Fig. 4.1.3a-c. Pressure boundary conditions for fully circular
journal bearings effect and the second term, pressure generation by
the squeeze film effect.
For A 1 , which is the solution of Eq. (29) with
yields the following conditions: e' = 0 and 1 - 2¢' = 1, the Reynolds boundary
condition [Eq. (27)] is used. The following boundary
A =0 at 8=0 condition is used for A 2 , the solution of Eq. (29)
oA withe'= 1 and 1 - 2¢' = 0:
A= 0, ao = 0 at 8 = 0* (27)
A =0 at ' = 0, (* oA2 = o at 0 = 0,2n
ao
where (* is the dimensionless bearing width and is
defined as (* = Ljr. (30)
The film pressure obtained using the Reynolds The film pressure for the case that the two terms
condition is positive in the region ofO < 0 < 0* and of the right-hand side in Eq. (29) exist simulta-
zero in 0* < 0 < 2 n. 0* is not the prescribed value, neously can be expressed as follows:
but is determined during the calculation. 0* is a
function of ( . If the axes of the journal and the (31)
bearing are parallel to each other, 0* yields a con-
Some conditions are proposed for Eq. (31):
vex and symmetrical curve in the axial direction.
Figure 4.1.4 shows the process used to determine (i) A 2 = 0 at the location of A 1 = 0 (0 = 0, 0*)
0*. (Fig. 4.1.5 a).
Next, the boundary condition in the unsteady (ii) A 2 exists in the region of 0 < 0 < 0* and
state is investigated. In this case, two boundary A 2 = 0 at 0 = O* (Fig. 4.1.5b).

235
4 Explanations

given by the following equations:

F.= 4 (LD) l JA cos(Jd(Jd(,

Fq, = 4 (LD) l JA sin(Jd(Jd(, (32)
a Condition (i)
A,H~-----------------;
- J~
F= 2 2
+Fq,,
where A is the area of the bearing surface.
In the static equilibrium state, F must be equal to
the bearing load applied. F is called the dimension-
less load-carrying capacity at the static equilibrium
position. The Sommerfeld number S can be ob-
b Condition (ii) tained as follows by using F:
1
S=--· (33)
2nF
Referring to Fig. 4.1.6, the following equation to
calculate the attitude angle <P is obtained:
<P = tan- 1 (-Fq,/F.). (34)
c Condition (iii)

0/J

d Condition (iv)

Fig. 4.1.5a-d. Boundary condition of dynamic film pressure for

fully circular journal bearings

(iii) The same as (ii) but A 2 = 0 where A 1 < 0 X

(Fig. 4.1.5 c).
Fig. 4.1.6. Components of dimensionless film reaction force
(iv) The circumferential boundary locations 0 1 , 02
are determined from A = 0. A 1 and A 2 must be b) Friction Coefficient
calculated separately by applying each condi-
tion (Fig. 4.1.5 d). Friction forces acting on the journal and the bear-
ing surfaces, fJ, j 8 , are given by the following
In this databook, boundary condition (i) is ap- equation:
plied for convenience.
fJ = JJ(f..l aau) rdOdz ,
A Y y=h
(35)
4.1.5 Calculation of Static Bearing
Characteristics fB = JJ(f..l aau)
A Y y=O
rd(Jdz .

a) Bearing Load-Carrying Capacity The friction coefficient is defined as the ratio of the
and Attitude Angle friction force to the bearing load. It is given by the
following equation:
The bearing load-carrying capacity is defined as the
reaction force of the lubricating film. Therefore, it h If/ - 1 = n: SS(_!_ auI ) dO dC ,
can be obtained by integrating the film pressure ' A H a17 ~=1 (36)
calculated from the static Reynolds equation
[Eq. (11 )]. The dimensionless film reaction force, F, JB If/- 1 = n: SS(_!_ au I ) dOd' '
and its components with respect to the eccentricity ' A H a17 ~=0
direction and its perpendicular direction, .F:, Fq, , are where h and h are the friction coefficients on the

236
 4.1 Method of Calculation for Bearing Characteristics

journal and bearing surfaces, respectively. oii/81'/ film reaction force, F, acting on the journal centre,
can be calculated as follows : 0 1 , can be expressed as a linear function of the

au 2 aA (
a11 = H ae 17 - 2 + 1 ·
1) (37)
displacement and its velocity. Figure 4.1.7 shows
the Cartesian coordinate system, OB- xy, with the
origin of the bearing centre, OB. When F is ex-
panded in Taylor's series and the terms of second
The friction coefficient calculated in this data-
and higher powers are disregarded, the following
book is that on the journal surface. It is calculated
equation is obtained:
from the first part of Eq. (36).

c) Oil Flow Rate

In this databook, the dimensionless side leakage (40)
flow rate, Qs, and the dimensionless flow rate, Qe,
are defined. In the multipad-type bearing, such as
the tilting pad bearing, Qs and Qe are defined as the
sum of Qsi and Qei· Qsi and Qei are the flow rates of + iJFY
each pad. Qe of the circular bearing is defined as the
flow rate in the rotational direction across the ay o
Reynolds boundary. where
Fx, FY: components ofF in x andy direc-
Qsi and Qei are calculated from the following tions

II
equations: F0 x, F0 Y: components of F0 in x and y direc-
tions at the static equilibrium
Qs = (L/D) !(Hw){=od1'Jdei, (38) point 0 10
{* 1 ·= dfdt: differentiation by time t
Qe = (L/D) J0 0J(Hii)o=o~d1JdC (39) iJ Fx/ ox lo etc.: partial differential coefficients at

= = =
where (* L fr, ii uf(rw), w w/(rw). (}~ repre-
sents the location of Reynolds boundary at ( = 0.
static equilibrium point
The second and third terms on the right-hand
side of Eq. (40) are called the stiffness and damping
matrices k, c of the oil film, respectively. Then, k and
c can be written as follows:
4.1.6 Method of Calculation for iJFx
Dimensionless Stiffness ay
and Damping Coefficients of Oil Film k=
iJFY iJFY
ax ay o (41)

c
The stiffness and damping coefficients of the lubri-
cating film were defined in Sect. 1.3.1. In this sec- iJFx
-
tion, the theoretical procedures for calculation of ax ay
c=
these coefficients are explained. iJFY iJFY
If the journal vibration about its static equilibri- ox ay o
um point, 0 1 o, is very small, the nonlinear dynamic
The dimensionless stiffness and damping coeffi-
cients are commonly defined by the following equa-

=(K
tion:
K xx Kxy) = CP k
K yx Kyy W ' (42)
C (Cxx
= Cxy) = Cvw c.
Cyx Cyy W
When these coefficients are applied to Eq. (40), the
following dimensionless equation is obtained:

(~=;::)=Fa· K ·(i =~;:)

Fig. 4.1.7. + F0 · C · ((~ - ~Jo):) (43)
Small vibration in the vicinity of the static equilibrium point (X1 - Y10)

237
4 Explanations

where F0 is the dimensionless load-carrying capaci- Fx- Foy

ty, X= X /CP, Y = Y /CP, and ( )' djdr (r = wt). =
As the dimensionless film reaction force, F, is = ar.l sm</>
(& . 7ft aF"'I cos</> ) L1e
0 - 0
commonly represented by two components,£., Fq,, 0 0

+ {(ar.l
o<f> + F_q,) sm</>
. (aF"' o<f> I F_) sm</>
. }L1</>
as described in Sect. 4.1 .5, it is convenient to express
t_!le left-hand side of Eq. (43) by using£., Fq,, F0 , and 0
0 0-
0
- 0, 0
Foq,· Then the left-hand side is rewritten using these
+ (ar.j . . ~. aF"'I
oel cos ..~.) L1e
four terms, which are proportional to the small
increments of displacement from OJo• that is, oel sm 0 '1'0 - 0 '1'0
I

(e - e0 ), (</> - </> 0 ) and their derivatives, (e - e0 )',

(</> - </> 0 )'. These small increments of (e - e0 ), ar.l sm. <l>o - aF"'I
+ (o</>1 0 o</>1 cos <l>o) L1</> ' 0
I

(</> - </> 0 ) etc. are expressed by using (X - XJ 0 ),

(Y- YJ 0 ) etc. By comparing the right-hand side of
the rewritten equation with that of Eq. (43), K and where L1e, L1</> etc. are given as follows:
Care obtained as functions of£., Fq, etc.
Referring to Fig. 4.1.7, we have the following re- L1e = (XJ- XJo) cos</> 0 + (YJ- }j 0 ) sin</> 0 ,
lation between Fx, f'y, ... and £., Fq,, .. .:
L1</> = e0 1 { - (XJ- XJo) sin</> 0
(JS)
FY
= (c~s </>
sm </>
sin<!>)
-cos</>
(~),
Fq, (44)
+ (YJ- YJ 0 ) cos</> 0 }, (48)
L1e1= (XJ- XJ 0 )' cos</> 0 + (YJ- }j 0 )' sin</> 0 ,
(~:) = G:s:: -:~:::) (~:). L1</>1= e0 1 { - (XJ- XJ 0 )' sin</> 0
+ (YJ- }j 0 )'cos</> 0 }.
If e, </>, F. and Fq, are expressed respectively by the
sum of both the static values at the equilibrium
</> 0 is the attitude angle at the static equilibrium
point, l!_o ,</>Q.:. F0 " F0 q, and their small variations, L1e, point. It is expressed as
L1</>, L1F,, L1Fq,, the following equation is obtained
e = e0 + L1e, • .+. -Foq,
s1n 'l'o = - _- . (49)
</> = </>o + L1</>' Fo
(45)
F.= Fo, + L1£.,
Fq, = F0 q, + L1~. When Eqs. (48) and (49) are substituted into
Eq. (47) and the result is compared with the right-
L1£. and L1Fq, are expressed as follows : hand side of Eq. (43), the elements of the matrices
for the dimensionless stiffness and damping coeffi-
- ar.l ar.l
L1F, = & L1e + o</> L1</> + Oe1 L1e + aF"'I ar.l
o<f> 0 11 "'..~.~ '
I
cients of the lubricating film are obtained as fol-
lows:
(46)
- Fg, ar.,_ Fgq, oF"'I- Fo ,Foq,
Kxx-
_
L1Fq, =
aF"'I L1e + aF"'I
7ft aF"'I oL1e + aF"'I
o</> o L1</> + M o</> oL1</>. I I F-30 oe 0 e0 F-30 o</> 0 F- 03
0

Substituting Eqs. (44), (45) and (46) into the left- X (oF"''
oe 0- 1 oF:'
eo o</> 0 ) -eoFFo'
,
0

hand side of Eq. (43) and neglecting the terms of

- Fgq, oF"'I
Kxy - -3
Fg, Fo ,Foq, ar.l
F0 oe 0 + e0 F-30 o</> 0 + F03
second and higher power, we get following equa-
tions:
Fx- Fox X (a£.1
-
oe 0
+-1 -aF"'I ) + -F0-
e0 o<f> 0
"'
e0 F0 '
= (~;~ 0 cos</> 0 + 0~"'1 0 sin</> 0) L1e K
yx
= - Fg, oF"'I- Fgq, ar.,_ Fo,Foq,
-3 a -3 ()..!.
(50)
Fo e o eo Fo 'I' o Fo3
+ {(~21 + Foq,) cos<f>o+ (~11
0 0 - F ,)sin</> }L1</>
0 0
x
(&ar.j o + eo1 aF"'I )
8</> o '

+ (~~~ 0 cos</> + ~~~~ 0 sin</> L1e1

0 0)

ar.l cos . ~. + aF"'I

+ (o</>1 0 o</>1 sm. ..~.) L1'..~.~
'1'0 1' '
0 '1'0
(47)

238
 4.1 Method of Calculation for Bearing Characteristics

Reynolds equation [10], we have small variation

in the film reaction force. The ratio of the small
change of the reaction force to the given pertur-
bation yields an approximate value for the par-
tial derivative used in Eqs. (50) and (51) [15, 16).
Details of these methods can be found in the litera-
ture references cited.
Figure 4.1.8 shows the values for partial deriva-
tives calculated by using these two methods. The
dotted and solid lines show the results obtained by
methods (i) and (ii), respectively. In the case of
(51)
method (ii), the value for each partial derivative
varies with the amount of each perturbation. The
optimum range of perturbation is found to exist for
each variable. In the example shown in Fig. 4.1.8,
when the amount of each perturbation is about
10- 2 , the partial derivative of each variable is con-
stant and agrees with that obtained by method (i)
Both methods are used in the calculations in this
databook. If the amount of perturbation is appro-
priately chosen the results calculated by the two
methods are consistent.
e0 is the eccentricity ratio at the static equilibri- In this databook, directions of the displacement
and the film force as defined in the cartesian coordi-
um point. F0 , Fo e and F0 <P are calculated from
Eq. (32). Thus, K and C can be determined when the nate system in Fig. 4.1.1 are chosen, which is not
partial derivatives at the static equilibrium point always the case elsewhere in the literature. There-
are given. fore, in calculation of the rotor vibration using any
The following two methods are commonly used other coordinate system the coefficient values in
to calculate the partial derivatives at the equilibri- this databook should be converted to those in the
um point. coordinate system to be applied in the calculation.
For example, when the coordinate system used in
(i) Following partial differentiation of Reynolds this data book, 0 B - x y, rotates by 0 in a counter-
equation [1 0] with respect to e, </>, e' and</>', four clockwise direction, we have a new system,
new partial differential equations for 8A /8e, 8Af OB- uv. The relation between the dimensionless
8</>, 8A /8e', and 8A /8</>' are obtained. When their film reaction forces represented in both coordinate
solutions are integrated over the bearing area, systems is expressed as follows:
each partial derivative used in Eqs. (50) and (51)
can be obtained [12-14].
(ii) Given the perturbations from the static equilib-
rium point with respect to e, </>, e' and </>' for
(Fv~) = ( - c~sO
smO
sinO)
cosO J FY
(JS) (52)

50 Full circular journal bearing -a P.,

i:: (lQ0Oil grOOVe, horizontal) ao:' 0
·g_20 ~!Q::~-~-=-~0.:..8__________ .£:.._ - - -
e:I
·.:
@ 10
·~
"
-5" 5
'iii
">"'
-~
2
·.:>
"
"0

1 1

0.5,L-~~~~~~~~~~~~~--~=-~~~~·
Fig. 4.1.8. Partial derivatives of the film reaction
0.05 0.5
force at the static equilibrium point Perturbation Je, J¢. Je,' J¢'

239
4 Explanations

where F, and F, are the components of F in the References for Sect. 4.1
OB- uv system. The film spring and damping coef-
ficients of the OB- uv system can be calculated [1] Dowson, D., Int. J. Mech. Sci., Vol. 4 (1962), 159.
from those of the OB - x y system by applying the [2] Pinkus, 0., Sternlicht, B., Theory of Hydrodynamic Lubri-
following conversion: cation, (1961), McGraw-Hill.
[3] Hinze, J.O., Turbulence, (1975), McGraw-Hill.

[~:::J [~:;1 '

(4] Schlichting, H., Boundary Layer Theory, (1968), McGraw-
Hill.
= C* [5] Constantinescu, V.N., Trans. ASME, Ser. D , Vol. 81, No.1
Cy' x' Cyx
(1962), 139.
Cy' y' Cyy [6] Ng, C.W., Pan, C.H.T., Trans. ASME, Ser. D, Vol. 87,
(53) No. 1 (1965), 675.
[7] Aoki, H . and Harada, M., Lubricant, Vol. 16, No.5 (1971),

r
cos 2 fJ sinfJcosfJ sinfJcosfJ sin 2 fJ ~ 348.
*- -sinfJcosfJ cos 2 B -sin2 B sinBcosB [8] Wada, S. and Hashimoto, H., Journal of JSLE, Vol. 44,
C - - sinBcosB - sin 2 B cos 2 B sinO cosO No. 382, (1978), 2140.
sin 2 8 -sinO cos() - sinBcosB cos 2 8 [9] Elrod, H.G., Ng, C.W., Trans. ASME, Ser. F, Vol. 89,
No. 3, (1967), 346.
(54) [10] Vogelpohl, G., VDI-Forschungsheft, 386, (1937).
[11] Akasaka, T., Numerical Calculation, (1977), Korona-sha.
Theoretical procedures that can be used to ob- [12] Glienicke, J., Dissertation, Karlsruhe (1966).
tain the bearing characteristics have been briefly (13] Someya, T., Dissertation of the University of Tokyo,
described. These procedures or methods can be ap- (1972).
plied for all journal bearings. Concerning the film [14] Nakanishi, Y., Chikira, S., et al., Hitachi Zosen Technical
thickness function H, although only one case of the Review, Vol. 37, No.1, (1976), 49.
[15] Dandage, S.R., Dissertation, Univ. of Wisconsin, (1974).
circular journal bearing is shown here, H takes var- [16] Pafelias, T.A., Dissertation, Rensselaer Polytechnic Institu-
ious function forms according to the bearing types. te, (1974).
In addition, in the case of pivoted pad bearings, the
balance of the moment of each pad about the pivot
point should be also taken into consideration in
calculation of the film reaction force.

4.2 Methods of Measurement for Journal Bearing Characteristics

Performance characteristic tests of journal bearings 4.2.1 Outline of Performance

are carried out to determine either static character- Characteristic Tests
istics, such as the load capacity, frictional loss, tem-
perature rise and oil flow, or dynamic characteris- of Journal Bearings
tics, such as the oil film coefficients and the stability
of the oil whirl, and then to verify the reliability. a) Loading Devices
For many cases, the tests can be performed on a
reduced- or full-scale model bearing set in a test Static loading of the test bearing is attained by
apparatus. In other cases, a bearing assembled in an joining such parts as a hydraulic cylinder, pneumat-
actual machine can be tested directly. This section ic bellows cylinder, link, lever or softening spring
presents the test methods, especially those used for support. Figure 4.2.1 illustrates a concept of the
the dynamic characteristic tests, with actual exam- static loading mechanism in which a hydraulic cyl-
ples. In the case of circular bearings with two axial inder is used for the rigidly mounted test bearing. In
grooves shown in Chap. 3, a comparison then fol- addition, Fig. 4.2.2 shows an actual example where
lows of the experimental data measured by the dif- compressed-air bellows are used for the static load-
ferent methods and apparatuses and of the calculat- ing of the test bearing casing resting freely on its
ed data. shaft.

240
 4.2 Methods of Measurement for Journal Bearing Characteristics

Dynamic loads act on the bearing to greater or centre and the reaction forces of the bearing, i.e., the
lesser degrees under real operating conditions. The dynamic characteristics of the bearing. Devices that
operating conditions can be classified in two broad can be used to apply the dynamic load include elec-
groups: those in which the journal centre moves all trodynamic, electrohydraulic, unbalance weight-
around the bearing clearance, as in the case of bear- type and cam-type exciters. The test apparatus
ings in reciprocating engines; and those in which shown in Fig. 4.2.2 has two electrohydraulic ex-
the journal centre moves with a small displacement citers which act on the test bearing casing with
around its static position of equilibrium. small movements. Details of the apparatus are ex-
The dynamic load is required to measure the plained in Sect. 4.2.3.
relation between those movements of the journal

Test bearing
b) Measurements of Friction Torque
Support bearing
In a case where a test bearing rests freely on its
shaft, the friction torque of the bearing can be ob-
tained by measuring the reaction force of an arm
overhung from the bearing. In other cases, the
torque can be measured directly by a rotary torque
Test shaft
meter or a torsion shaft coupled into the drive train
between the driving and the driven shafts. Other-
Static load
wise, in the case of bearings in actual machines, the
frictional loss is usually estimated from the oil feed
Hydraulic cylinder rate and the temperature rise.
for static loading The frictional coefficient of a well-lubricated
bearing is normally in the order of 0.001, and the
friction force is much smaller than the static load.
/
Consequently, to obtain better accuracy of the fric-
Fig. 4.2.1. An example of a static loading mechanism in a bear- tion torque, losses other than those in the test bear-
ing test rig ing should be measured in advance.

(D Test bearing casing (J) Tension rod Shaft diameter Max. 280 mm
0 Bellows cylinder ® Support ball bearing regular size 140 mm
(for static loading) I00 ~ 10000 rpm
® Flexible diaphragm coupling Speed range
0 Displacement sensor
® Rotational speed sensor Static load Max. 50kN
@ Accelerometer from three directions
® Test shaft
Dynamic load Max. 10 kN, peak sinusoidal
(i) Load cell
@ Test bearing DC ~soo Hz
@ Hydraulic exciter head @ Flexible bar
(for dynamic loading) (for prevention of side inclination)

Fig. 4.2.2. An example of a test rig for dynamic characteristics of journal bearings - cross section of the main parts

241
4 Explanations

c) Measurements Recorded and Data Processing Hagg and Sankey [1] derived the four coefficients
which disregarded the effect of cross-coupling be-
With static characteristics the following variables tween the major and minor axes of the journal lo-
are measured in addition to the friction torque: for cus. However, to eliminate the four cross-coupling
control purposes the rotational speed, steady state terms and to reduce the original eight coefficients to
load, oil feed pressure, oil feed rate and oil inlet four equivalent coefficients with the nature of the
temperature, for example; and for experimental bearing dynamic characteristics, each vibration
data, such variables as the steady-state locus of the phenomenon has to be analysed individually, as
shaft, the oil film pressure and temperature distri- discussed elsewhere [4].
bution in the bearing. In dynamic testing, such dy- In the method adopted in [3] and [7], after the
namic high-frequency magnitudes as the exciting four stiffness coefficients had been determined from
forces and the resultant displacements, velocities the static equilibrium locus of the journal, the resid-
and accelerations must be measured in addition. ual four damping coefficients were obtained from
These numerous data are processed by computer. the dynamic response of the journal to unbalance
The practical example is described in Sect. 4.2.3 g. excitation.

b) Incremental Forcing Methods

The oil film coefficients are measured directly for
4.2.2 Methods of Measurement these methods in accordance with the definition (see
for Oil Film Coefficients Sect. 1.3.1 ). Small changes in static load produce
incremental journal centre displacements Ax and
The relations of oil film coefficients expressed in Ay. If two known, separate, and different incremen-
Sect. 1.3.1 are nonisotropic and nonreciprocal and tal loads are applied, changes in oil film forces give
are thus quite different from the relations which four equations, two of each of the forms:
apply in common elastic systems. Therefore, if there
L1Fx = kxxL1X + kxyL1y ,
is a difference between the directions of the coordi-
nate axes determined in this databook and those AFY = kyxL1x + kyyL1y.
used for vibration calculations and also the direc- The given measurements of the incremental dis-
tions of the measurement axes of the exciting force placements and oil film forces and simultaneous
and the motion response, the oil film coefficients solution of each pair of equations yield the four
must be corrected by transformation of axes. stiffness coefficients.
By definition, oil film coefficients are the coeffi- Similarly, the four damping coefficients are ob-
cients that represent the relations between journal tained by applying forces and measuring the result-
movement of small amplitude and the reaction ing velocities in both x and y directions. However,
force of oil film. In the experimental assessment, this method involves difficulties associated with ap-
these coefficients can therefore be determined by plying and measuring the forces, and also necessi-
measuring the motion response of the journal or the tates four sets of successive measurements to obtain
reaction force change of the oil film to the known the eight coefficients.
input force or motion on the journal. The above two methods are applicable to actual
Since Hagg and Sankey [1, 2] published the first machines under operating conditions. A method of
experimental data relating to oil film coefficients application developed by Morton is that of group
which consisted of four major components for ap- b. The method involves applying transient forcing
plication to unbalance vibration calculations, vari- directly to a rotating shaft. Full details of the prac-
ous investigations have been carried out on the tical considerations and the analysis technique are
experimental determination for a full set of eight given elsewhere [9].
coefficients [3, 5 - 16]. Most of those test methods On the other hand, in laboratory or workshop
fall into one of the following groups. tests, it is more advantageous for application and
measurement of the forces to float a test bearing on
a) Unbalance Excitation Methods its shaft in contrast to the real machine situation.
The following methods employ such a test appara-
These methods [1, 3, 7] are valid for the tilting-pad tus that consists of a test shaft in rigid slave bear-
bearings in which the cross-coupling terms can be ings and carrying a floating test bearing.
virtually disregarded, which means, strictly speak-
ing, only in ideal cases where there is no pad inertia.
In other cases, however, since tests with pure unbal- c) One-Directional Sinusoidal Wave Excitation
ance disturbance can yield only four unknown coef- Methods
ficients, the four remaining coefficients have to be At a steady running condition, if the test bearing is
obtained by some other method. shaken sinusoidally in one direction, the bearing

242
 4.2 Methods of Measurement for Journal Bearing Characteristics

moves in both horizontal and vertical directions.

For the bearing motion of small amplitude, the
frequency is equal to that of the applied force, and
the path of motion is approximately an ellipse with
the steady running position as centre, showing that
the characteristics are approximately linear. Then
the measurements of bearing motions and applied
forces can be used to determine the oil film coeffi-
cients.
These methods [5, 8, 11, 12, 14, 15] require two
sets of measurements with different loading condi-
tions from each other, which may be achieved by
either exciting in two directions at a single frequen-
cy or exciting in one direction at two frequencies.
The former technique was adopted for test bearings
2, 3, 4, 7, 8, 10 and 11 in Chap. 3. The second
method was adopted for test bearing 6.

d) Two-Directional Compound Sinusoidal Wave

Excitation Methods

To obtain the eight oil film coefficients, two or more Fig. 4.2.3. Photograph of test rig for dynamic characteristics of
separate and different loadings must be applied in journal bearings
the above methods, for which it is essential to
ensure that the duty parameter for the bearing
remains constant throughout the test sequence.
The methods in group c can be modified to ob-
tain the eight coefficients at once, by simultaneous test bearing ©, which is set in the test bearing
application of sinusoidal forces in two directions casing Cl), floats on its shaft. Three sets of com-
with different frequencies and measurement of the pressed-air bellows (blare attached to the test bear-
resulting motions in two directions. This is the two- ing casing horizontally and vertically. The bellows
directional compound sinusoidal wave excitation are used for static loading of the test bearing and
method [10, 13, 16], which was adopted for test are so flexible that the test bearing casing can orbit
bearings 1, 5 and 9 in Chap. 3. freely around the shaft. The concentric setting and
Details of the analytic procedure for extraction of operation of the test bearing is secured by means of
the eight coefficients by the methods in groups c an isotropic arrangement of eight flexible bars @.
and d can be found elsewhere [10]. In addition, two hydraulic exciters are located
left and right of the test bearing casing at 45° from
vertical.
4.2.3 Example of a Test Rig for Dynamic Each exciter head @ is connected to the test
bearing casing by means of the tension rod (/), and
Characteristics of Journal Bearings the crevice joints at both attachment points, to the
test bearing casing and the rig bed. These exciter
The test rig developed by Glienicke [5] for the mea- connectors are able to impose negligible constraint
surement of oil film coefficients is frequently quoted on the test bearing casing.
as a typical one, which has been followed by many Since each of the static loads provided by three
other apparatuses, refined mainly in connection bellows and each of the bias loads from two exciters
with the loading mechanism and the processing of is independently adjustable, the static load vector
the measurements. can rotate through 360°. This function is adequate
Eight test rigs are shown in Chap. 3, with the for bearings with a noncircular cylindrical bore,
experimental results obtained with each. In the which the equilibrium curve itself is dependent on
paragraphs below a representative survey is made the angular position of the static load.
of the rig shown in Fig. 3.1 .2, in which the com- The two exciters make it possible to apply dy-
pound sinusoidal wave excitation method can be namic sinusoidal forces together with any relative
applied. Figure 4.2.2 shows the configuration of phase and magnitude. The ratio between the two
the principal mechanical components of the rig. exciting frequencies can also be adjusted to any
Figure 4.2.3 is a photograph of the rig. ratio with an integer up to 20 (e.g. 1 : 2, 3 : 4).
The test shaft (j]) is supported at both ends in Provision is made for the following measure-
comparatively rigid-support ball bearings ®. The ments:

243
4 Explanations

a) Static Displacement and Vibration Response meter of pulse generating type through an FV (fre-
quency to voltage) converter.
The relative displacements between test bearing
and journal are measured by four pairs of contact-
f) Measurement from Inside Rotating Shaft
less eddy current proximity sensors Q), located on
each side of the bearing in the horizontal and verti- The thickness and pressure distribution of the oil
cal directions. film around the journal surface provide useful infor-
Eddy current proximity sensors offer the advan- mation on the operating conditions of the bearing.
tages of noncontacting, compact and lightweight In order to measure these variables by minimum
design and ensure broad ranges of frequency re- transducers, eddy current proximity sensors,
sponse and amplitude linearity. Some of them, how- piezoresistive pressure transducers and CA thermo-
ever, are likely to cause zero drift with temperature couples are incorporated in the rotating shaft.
changes, which must be compensated for static dis- These transducers are very small and light. Their
placement measurements. outputs are directed through a slip ring.
The inertia of the test bearing and its associated
structure bypasses a proportion of the forcing, so
g) Data Processing
that the true force across the oil film differs from the
exciter output force. To compensate this effect by Figure 4.2.4 is a schematic of the data processing
means of mass cancellation, the absolute accelera- system with a minicomputer. For determination of
tions of the test bearing casing are measured by four test conditions or for control purposes, low-speed
pairs of piezoelectric accelerometers @, each locat- data such as the rotational speed, static load, oil
ed at the position of a proximity sensor. feed pressure, oil feed rate, oil inlet temperature,
and also certain significant temperatures in the
b) Static and Dynamic Loads bearing are fed into the minicomputer through an
analog-to-digital converter, digital voltmeter and
The static loading applied by the compressed-air scanner.
bellows is measured by Bourdon-type pressure The exciting forces and the resultant displace-
gauges. ments, velocities and accelerations from which the
The static bias load and dynamic load applied by oil film coefficients are determined are high-speed
the hydraulic exciters are measured by the load cells data. These high-speed data are also fed into the
G) attached to the exciter heads. The total static minicomputer simultaneously with the low-speed
force vector is the sum of the loads applied by three data via a high speed A/D converter and a sample
bellows, two exciters and the gravitational load of and hold amplifier.
test bearing assembly itself. In this mode, sampling periods should be
matched with those of all high-speed data, to allow
c) Rotational Speed of Test Shaft
exact FFT calculations. In this system, the sam-
pling interval is synchronized with the lower of the
The magnetic sensor @ indicates shaft orientation exciting frequencies.
and provides a pulse for an accurate rotational The data processing program provides the fol-
speed indicator. This pulse signal allows the exciters lowing functions:
to be run in synchrony with the test shaft, and the
resulting sinusoidal forces can also be combined at
the same amplitude to provide a rotating force of
constant magnitude, as required. Low speed data 100 ch

0
I YODAC-8
d) Inlet Oil Temperature and Bearing Metal 99
Temperature
Chromel-alumel (CA) thermocouples are used for
all temperature measurements. The white metal
temperature within 3 mm of the surface is measured
at each of the test bearing locations to obtain the
mean oil film temperature.
0
I
AID
e) Supply Oil Pressure and Oil Flow Rate 15
CONVERTER

The supply oil pressure is measured by a Bourdon-

type gauge. The oil flow rate is detected by a flow Fig. 4.2.4. Schematic of data processing system

244
 4.2 Methods of Measurement for Journal Bearing Characteristics

(i) Acquisition and presentation of data to deter-

mine the clearance distribution and the origin
of displacement coordinates of the test bearing.
(ii) Acquisition and calculation of data associated
with three measurement methods of oil film
coefficients, using compound sinusoidal wave
excitation, single-frequency sinusoidal wave
excitation, and random wave-form excitation;
with transforming coordinate axes referring to
the measurement directions of exciting forces
and motion responses.
(iii) Calculation of the means and standard devia-
tions of measured values.
(iv) Plotting or tabulation of measured values. Di-
mensional and dimensionless output of oil film Fig. 4.2.6. Example 2 of test rig for large bearings
coefficients and bearing characteristic parame-
ters. pecially in the high-speed region, and also to check
(v) Comparison of measured values with prepared the capacity in the presence of such severe load
reference data. conditions as misalignment in the low-speed region.
CPU memory size required for these data pro- Further difficulties associated with mathematical
cessing modules is about 32 K words. The mini- modelling include the problems involved in taking
mum sampling time for high-speed data is 160 JlS account of oil recirculation, determination of the
(10 JlS/channel x 16 channels), and the difference correct viscosity distribution, estimation of the
between each channel and the next is kept within bearing distortion, and precise definition of the cav-
2 JlS by means of sample and hold (see Fig. 4.2.4). itation boundaries. Therefore, continuous efforts
In this system, all measurement data are stored are still in progress to improve the accuracy of pre-
on-line and are automatically analysed by the com- diction for the performance characteristics of bear-
puter. The results of analysis are indicated on a mgs.
visual display and fed to a printer or a plotter so
that the validity of the test results can be estimated
immediately.
In addition to a reduced-scale model test in the 4.2.4 Experimental Results Obtained
rig mentioned above, full-scale testing on large with Circular Bearings
bearings more than 500 mm in diameter is per-
formed in a large rig. The photographic examples of with Two Axial Grooves
the practical application for large turbine generator
bearings are shown in Figs. 4.2.5 and 4.2.6. These Experimental static displacement data for circular
photographs correspond to the schematic of the bearings with two axial grooves, which are mea-
principal mechanism illustrated in Figs. 4.2.1 and sured by five different types of rig, as shown in
3.6.2, respectively. Chap. 3, are plotted together in Fig. 4.2.7 for the
Full-scale testing is usually conducted to verify eccentricity ratio and Fig. 4.2.8 for the attitude an-
the performance characteristics and reliability, es- gle. Similarly, the oil film coefficients are compared
in Figs. 4.2.9 and 4.2.1 0. The measured data points
for each test bearing are plotted discriminately by
the following symbols:
EB : Test bearing 1
0 : Test bearing 2
e : Test bearing 3
8 w~: Test bearing 4, with rotational speed
conditions
N = 12.5, 25, 50 rps, respectively
() : Test bearing 5

Each figure also contains the corresponding the-

oretical curve in a solid line for L : D = 0.5. Since
the L: D for test bearings 1 and 2 is 1.0, the Som-
Fig. 4.2.5. Example 1 of test rig for large bearings merfeld number of these bearings is modified and

245
4 Explanations

----- Theory
90 L!D=0.5
"0.8
0 .... 75
-~ 0.6 " 60
01
;:. §
·o
•t::
c: 0.4 "
-o 45
§ --Theory -~
30
~ 0.2 L!D=0.5 ~
15
0~~~~~~~~~~~~~--~~~ 0
0.01 0.02 0.05 0.1 0.2
1 0.5 2 5 0.01 0. 02 0.05 0.1 0.2 0.5 2 5
Sommerfeld number S Sommerfeld number S
Fig. 4.2.7. Eccentricity ratio of circular bearings with two oil Fig. 4.2.8. Attitude angle of circular bearings with two oil
grooves grooves

15 15

:.:' 0

~
c 10 0 10
"
,•
0
·u
e 0

"0
0
0 0 0
9 (J 0
•
(.)
0 0 0
9Q

"'"'
0
~
"
0
:.:
Q Q

@ e
";::J 5 Q e e 5
"'
"'
""'
i3 • •
.9
rg
" 0
.§ ----- Theory 0
(:! L!D=0.5

0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 0.01 0.02 0.05 0.1 0.2 0.5 2 5
a Sommerfeld number S b s
15,---- -- -- -- - - - - - - -- -- -- ------, 15

10 10

0
~
~ 5 0
~
5 ooo
0 0

0 0 e
..
0.01 0.02 0.05 0.1 0.2 0.5 5 0. 01 .02 0.05 0.1 0.2 0.5 2 5
c s d s
Fig. 4.2.9. Dimensionless stiffness coefficients of oil film in circular bearings with two oil grooves

the measured data is plotted on the same figures as where

follows.
log a= 0.75 logp- 1.488
(L /D) }" {1.0}1.531 P= angle of bearing arc = 170°
Serf = { (L/ D)o S = O.S S = 2.89 S
(L: D) 0 : reference L: D ratio of standard bearing

246
 4.2 Methods of Measurement for Journal Bearing Characteristics
15.-----------------------------. 15.-----------------------------.
---Theory 0

LID=0.5 0

.. 10

Q:> •
.,.o
•
5 ..
.. •

• 0
•
0.01 0.02 0.05 0.1 0.2 0.5 2 5 0.01 0.02 0.05 0.1 0. 2 0.5 2 5
a Sommerfeld number S b s
15.-----------------------------. 15r-----------------------------~

10 10
0
0
..
(.) " J
5 5 .. .. 0
•
. .. • .. 8 •
• • •tJ(J;t{"
..a.
0 J> .. (JC> ..
~_.oo

. ..
• e
J ~~el'~i,~ •
0 0 , .....

0 89
.... e
.. 0 ,.a;."\~o'"~
0.01 0.02 0.05 0.1 0.2 0. 5 2
0

0.01 0.02 0.05 0. 1

•
5 0.2 0.5 2 5
c s d s
Fig. 4.2.10a-d. Dimensionless damping coefficients of oil film in circular bearings with two oil grooves

This relationship of modification is obtained by ratios, e > 0.7) and light loadings (i.e., low eccentric-
interpolation of calculated results as discussed in ity ratios, e < 0.3). This certainly cannot be ex-
Sect. 4.5. plained in terms of errors in the nominal clearance
Calculated values are based on constant viscosity assumed or in the measurement of movement in the
assumptions using the finite difference technique. journal or bearing. An allowance for variable vis-
Details of the calculation method are described in cosity certainly seems to improve the agreement
Sect. 4.1. between theory and experiment, but there is still a
Although these results are derived from a combi- significant difference. One possible explanation is
nation of five different experimental data and are nonlinearity of the oil film.
somewhat scattered, it can be said that the experi-
mental data shows fairly good agreement with the
theoretical curves. The experimental data itself b) Theoretical Assumption of Boundary Conditions
should be still examined, mainly in the individual for the Oil Film Extent and Pressures
test condition, as previously mentioned. The pres- at the Film Boundaries
ent data, however, shows the following points of
special interest. Stiffness coefficients show better agreement be-
tween theory and experiment than do damping co-
efficients. It is necessary at this point to examine
a) Nonlinearity of Oil Film boundary conditions for the oil film extent and
pressures at the film boundaries.
The discrepancies between isoviscous theory and Since stiffness coefficients are ratios of the reac-
experimental results become comparatively signifi- tion force change of the oil film to the statical dis-
cant at both high loadings (i.e., high eccentricity placement of the journal centre, only the wedge film

247
4 Explanations

pressure is reflected in these values; damping coeffi- References for Sect. 4.2
cients, in contrast, depend on both wedge film pres-
sure and squeeze film pressure, because the damp- [1] Hagg, A.C., Sankey, G .O., J. Appl. Mech., 23-2 (1 956),
ing force of the oil film arises from both pressures. 302.
The boundary conditions should still be examined [2] Hagg, A.C., Sankey, G.O., J. Appl. Mech., 25-1 (1958),
141.
when the mean bearing pressure and the oil inlet [3] Nakagawa, E., Proc. JSME, 104 (1963), 31.
condition seem to be of particular concern. [4] Kanki, H., Kawakami, T., Esaki, J., Proc. JSME, 810-16
(1981), 57.
[5] Glienicke, J., Proc. IME, 181-3B (1966- 67), 116; Disserta-
tion T.H. Karlsruhe (1966).
c) Reductions in Damping Coefficients [6] Mitchell, J.R., Holmes, R., Ballengooyen, H.V., Proc.
at High Eccentricity Ratios IME, 180-3K (1965-66), 90.
[7] Woodcock, J.S., Holmes, R., Proc. IME, 184-3L (1969- 70),
111.
Among four damping coefficients, Cxx (where x is [8] Morton, P.G., J. Lub. Techno!., Trans. ASME, Ser. F, 93-1
the direction of static load) in particular shows the (1971), 143.
[9] Morton, P.G., IUTAM Symp. Lyngby, Dynamics of Ro-
widest discrepancy between theory and experiment, tors, (1974), 350.
the experimental results being lower. At high eccen- [10] Someya, T., Dissertation University of Tokyo (1972).
tricity ratios the damping coefficients derived ex- [11] Nakanishi, T., Chikira, K., Simoda, H., Hitachi Zosen
perimentally also differ qualitatively from the theo- Technical Review, 37-1 (1976), 49.
[12] Nakanishi, T., Chikira, K., Simoda, H., Hitachi Zosen
retical values. This tendency for a drop in damping Technical Review, 38-4 (1977), 224.
coefficients at high eccentricity ratios is observed [13] Esaki, J., Furukawa, T., Matsumoto, I., Mitsubishi Juko
with almost all test bearings and seems to reflect the Giho, 16-1 (1979), 61.
actual state of the coefficients. [14] Hisa, S., Matsuura, T., Someya, T., International Confer-
The method of measurement for the dynamic ence on Vibrations in Rotating Machinery, Churchill Col-
lege, Cambridge (1980), 223.
characteristics of journal bearings is discussed in [15] Masuda, T., Suganami, T., Oishi, N., Taguma, Y., Proc.
this section. The rigs, transducers and data process- JSME, 813-5 (1981), 91.
ing systems are still subject to improvement. [16] Mitsui, J., Trans. JSME, C, 48-428 (1982), 565.

4.3 Estimation of Effective Oil Film Temperature

The dynamic characteristics of oil film depend on in Sect. 1.2.4, H is newly defined as follows, disre-
the temperature of oil film, because oil viscosity is garding H 2 :
heavily dependent on its temperature. If an error is (1)
made in the estimation of oil temperature, the dy-
namic characteristics calculated may be incorrect. where I(is the ratio of heat used for the temperature
Therefore, it is of the utmost importance that the rise of oil film to the total heat generated; the value
effective oil film temperature be estimated correctly 0.6 is commonly used. This practice means that
if the best possible use is to be made of this data- 60% of the heat generated is consumed to raise the
book. Because the temperature is different in the temperature of the oil film and the remainder is
circumferential and axial directions in an actual oil removed from the bearing by heat transfer.
film, however, analysis must take the change of vis- The next equation is obtained for the heat bal-
cosity into account, as shown in Sect. 4.7, for cor- ance of journal bearing using Eqs. (22) and (23) in
rect estimation. In this section a comparatively sim- Sect. 1.2.4, and Eq. (1):
ple method of estimating the effective oil film
temperature is explained, which allows the best pos- (2)
sible use of this databook; it is calculated under the
assumption of constant temperature (viscosity). e
The effective temperature of oil film is finally
given by the following equation, which is derived
from Eq. (2) above and Eq. (30) in Sect. 1.2.4:

4.3.1 Estimation from Calculated Data e= (1 -A) Jj, · W · wi(/(ceq) + 01 . (3)

The value of e is calculated by the iteration pro-

Though the detailed discussion on the heat balance cedure shown in Fig. 1.9 in Sect. 1.2.4.
of journal bearing is explained in Sect. 1.2.4, a sim- The outline is as follows: First, the effective tem-
pler method is introduced here. In Eq. (22), shown perature of oil film, e, is assumed and the effective

248
 4.4 Boundary Condition for Oil Film Pressure Distribution

4.5
viscosity is estimated. Next, the Sommerfeld num- a Tilting pad
5 pad
ber S is calculated on the basis of the effective vis- 4.0 LID=0.5
cosity. The friction coefficient, jj,, and the oil flow, L.O.P.
Qs and Qe, are obtained by interpolating data 3.5
shown in this databook. The new effective tempera- mp

ture can be calculated from Eq. (3). This iteration is ~ 3.0 0 0

•...
0.3: !':.
~

continued up to the limit within which the error is -"""

-..... 0.5:
}.) 2. 5
small enough. 0.8 :
•
~
.
-.....
2.0

....• 1.5
4.3.2 Estimation from Measured Data "q

1.0

Sometimes there are existing data on bearing tem- 0.5

perature that have been recorded in a condition
similar to the one in which the effective temperature 0 '---~-=--.,........,~-=--::--::-
0. 2 0.4 0.6 0. 8 1.0
is to be estimated. In this case we can obtain the a
effective temperature on the basis of measured data.
As the bearing temperature changes along the cir- Relationship between A Trnean!Pm and e
cumferential and axial positions, however, the loca-
tions at which the temperature should have been
measured are debatable. There are no commonly
accepted opinions on this point. The effective tem-
perature in this databook is so determined that the
scatter of the measured data in the S - e chart scat- 3.5
ter is as small as possible. 3. 0
Cylindrical
In this case, the effective temperature is assumed N!i LID= I
to be equal to the average of many temperatures ;:;:; 2.5
measured along the circumferential position at the -"""
-.....
}.)
centre line. Two examples, in which the relationship 2.0
•
between the average temperature rise A Tmean divid-
ed by the mean bearing pressure Pm and the eccen-
~
.
-..... 1.5

tricity e is depicted, are shown in Fig. 4.3.1, where •

....
"q 1.0
(a) and (b) represent data for a tilting-pad bearing (5
pads, L/D = 0.5, LOP) and for a cylindrical bearing 0.5
with two axial grooves, respectively. It is found that
0
estimation of the effective temperature by the pres- 0.2 1.0
ent method is useful, because temperature rise does b
not scatter especially in the case of tilting pad bear- Relationship between A Trnean/ Pm and e
ing. However, the best way of using measured data
effectively is still the subject of discussion. Fig. 4.3.la, b. Relationship between AT and e

4.4 Boundary Condition for Oil Film Pressure Distribution

For calculation of the pressure distribution of jour- The following are well known to be the condi-
nal bearings, the following must be known: (1) the tions in which circumferential oil film rupture oc-
pressure at the oil feed position, (2) the pressure at curs:
the bearing end, and (3) the circumferential region
within which the oil film exists. In this databook the a) Sommerfeld condition (no rupture)
pressures at the oil feed position and at the bearing b) Gumbel condition
end are set at zero. This means that the ambient c) Reynolds condition
pressure and the oil supply pressure are equal to the d) Prandtl-Hopkins condition
pressure of atmosphere. The third item (3) is dis- e) Coyne-Elrod condition
cussed in this section. f) Floberg condition

249
4 Explanations
100.,....-----------------------,
These conditions can be applied to the wedge - Boundary condition A
action, but there are not yet any established theo- J 50 -- Boundary condition B
ries that would apply to the squeeze action. Though
-;::
conditions (e) and (f) are considered to fit approxi- <!)
'(]
mately the actual situation, these require compli- s <!)

cated iteration. Therefore, Reynolds condition is 8 10

bl)

commonly used. Calculations in this book are t::

worked with reference to Reynolds condition. -~

"0
There are, incidentally, two methods of calculat- ~
ing the spring and damping coefficients of oil film. -a
<!)

The two are different in the way the oil film force is ·~
5
calculated, the static and dynamic force compo.-~ E
nents being separated for one and not the other. Iii Q 0. 5
the latter method, Reynolds condition is commonly
applied to the total pressure. In this section, the
following two conditions are examined fbr the fox:,- Sommerfeld number S
mer method. ·
Fig. 4.4.2. Damping coefficient
A. The pressure distribution by squeeze action is
calculated on the assumption that there is no film both these conditions is shown in Fig. 4.4.2. It can-
rupture independent of that caused by wedge ac- not be judged from the example whether either is
tion. The pressure is then integrated within the re- more realistic than the other, though the values
gion of positive oil film pressure by wedge action, obtained with condition B are smaller than those
which is calculated under the Reynolds condition obtained for condition A within the region where
(Fig. 4.4.1 a). the Sommerfeld number is small.
B. The pressure distribution by squeeze action is Next, the stability of the Jeffcott rotor is dis-
calculated on the assumption that the pressure by cussed with reference to these two values. The sta-
squeeze action also becomes zero at the boundary bility chart is shown in Fig. 4.4.3. Here, A. and v are
of film rupture by wedge action calculated on the the modified Sommerfeld number (S/v) and the on-
assumption that the Reynolds condition obtains set speed (w jfg/C), respectively. Further, S is the
(Fig. 4.4.1 b). Sommerfeld number, w is the rotational angular
speed, g is the acceleration of gravity, and C is the
radial bearing clearance. Though the onset speeds
with A and B are equal with large A., they are very
different from each other with small A., and it should
Squeeze action be noted that the onset speed in A becomes ex-
tremely low. Small A. means, physically, that the
...... bearing mean pressure and the bearing radial clear-
ance are large. It is well known that self-excited
a
vibration does not occur in such a case. Therefore,

- Stability limit
---- ( C,-y+ C yx) !( C xx+ Cn )
I
Wedge action 0.8 8
~ Squeeze action
I
I
l
J -----....... '
+~ 0.6
b
C,..) '\ unstable
'\
....... ~~
Fig. 4.4.1 a, b. Pressure boundary condition
B.C. (B) ''\\
\

Conditions A and B give the same results in cal- '

0.2 2 ~
culation of the spring coefficients of oil film, because \ .. _.: B.C. (A)
these coefficients depend only on the wedge action. Stable
The two conditions yield different results only for 0 0L-L-~-L~-L-~~~~~~~~.

calculation of the damping coefficients, because 0 . 001 0.01 0.05 0.1 0.5 1 5 10
). ~ Siv
these depend on the squeeze action. An example of
the calculation of damping coefficients for a cylin- Fig. 4.4.3. Sta bility chart. B.C.(A), boundary condition A ;
drical journal bearing with two axial grooves with B.C.(B), boundary condition B

250
 4.5 Influences of Length : Diameter Ratio on Journal Bearing Characteristics

it may be concluded that condition B is more real- Here, as the term (kxy- kyx) is the same for con-
istic than condition A. ditions A and B, the difference in stability is caused
Next, the subject is discussed from the viewpoint by the first, second and third terms on the right-
of energy. The infinitely small vibration around the hand side. Now, the assumption ax= aY is made for
equilibrium point of journal is expressed as follows : the sake of simplicity. The energy dissipation, that
is, the stability of the rotor, can then be discussed on
Ax= ax cos(wt + <f>x), the basis of the ratio (cxy + cyx)l(cxx + Cyy), shown
L1y = aY sin (w t + </>y} . in Fig. 4.4.3. With small A values the ratio is smaller
The x andy components of oil film reaction force with condition B than with A. This is consistent
are given by the equations below : with the idea that the energy dissipation is larger
with B than with A and the stability with B is
L1Fx = + kxyL1y + CxxL1x + CxyL1y,
kx xL1x superior to that with A.
L1FY = kyxL1x + kyyL1y + CyxL1x + cYYL1y . From the discussions above it may be concluded
that condition B is more realistic than condition A.
Consequently, the energy dissipated during one Therefore, the data in this book has been calculated
cycle of vibration is obtained as follows: on the basis of condition B whenever the static and
2x dynamic oil film forces are computed separately.
E= J (L1Fx· x + L1Fy· y) dt
0
=nw 2 axcxx+nw
2 2 2
aycyy
- nw axay(cxy + Cyx) sin(<f>x- </>y)
2

- nwaxay(kx y - kyx) cos(<f>x- </>y).

4.5 Influences of Length: Diameter Ratio on Journal

Bearing Characteristics
As a basis for discussion of the influences of length- The method of interpolating bearing data of an
to-diameter ratio on journal bearing characteris- L: D ratio not listed in this databook from the data
tics, the following calculated data is compiled and given is explained in this section.
used in this section (pp. 253 - 265). The similarity rule of bearing characteristics be-
tween the same type of bearings with different L : D
Calculation no. Bearing type L : D ratio ratios is given by the value S (L: Df in the infinites-
imal short bearing theory [1]. Moreover, simple
LBD-1 Cylindrical bearing with 0.35, 0.5, 0.75, 1.0,
two axial grooves 1.5
equations have so far been presented, which express
bearing data of finite length bearings as a function
LBD-2 Two-lobe bearing ; 0.5, 1.0, 1.5
of the eccentricity ratio e by adding correction fac-
mP = 0.67
tors to data calculated for partial arc bearings ac-
LBD-3 Five-pad tilting pad 0.3, 0.5, 0.7
cording to the infinitely long bearing theory [2] or
bearing ; mP = 0.5, LOP
for cylindrical bearings according to the infinitesi-

102 102

10 1
-L!D=0.5
1.0 -·-LID=0.1

0.6 ··--- L/D=0 .3

0.4 -L/D=O.S
---LID=0 .1 ~·
0.2

0 L.......................1o.__,...
, ..............................1.._o--~....._...........~~ 10- ' 1o-l 10-' 10-l 1
a s<ff b s<ff c s<ff
Fig. 4.5.1 a-c. Effective Sommerfeld no. for L/D ratio. 5 - tilting pad bearing, mP = 0.5, P= 60°, rx/ P= 0.5, LOP

251
 4 Explanations

mally short bearing theory (3]. These simple meth- The value of a decreases with decreasing angle of
ods imply that bearing characteristics for different arc p, and is nearly equal to 1.0 for multilobe bear-
L : D ratios can be approximat ed by introducing an mgs.
effective Sommerfeld number defined by The revised data obtained from Eq. (1) is shown
in Fig. 4.5.1 for calculation no. LBD-3, where
(L/D) }a (L/D)0 is 0.5. It is found that for different L : D ratios
Seff = { (L /D)o S. (1)
the eccentricity ratio e, stiffness coefficient K;j, or
Here, the denominato r (L/ D)0 is the L : D value damping coefficient C;j can be represented by a
that has been selected as standard. single curve. Though the correction factor depends
The value for exponent a is obtained as follows on the bearing type, it is useful in practice to express
from calculated data in this databook: data as a function of the effective Sommerfeld num-
ber for the estimation of data for different L: D
a) for partial arc bearings: values.
log a= 0.75 logp- 1.488 (2)
where p = angle of partial arc (degrees), provid-
ed that References for Sect. 4.5
p= 180° if p ~ 180°
[1] Ocvirk, F.W., NACA TN 2808, (1952).
b) for two-lobe bearings and five-pad tilting pad [2] Warner, P.C., J. Basic Eng., Trans. ASME, Ser. D, 85-2
bearings: (1963), 247.
[3] Nakagawa, E., and Aoki, H., Trans. JSME, 31-229 (1965),
a~ 1.0 (3) 1398.

252
 Calculation No. LBD-1

1. Calculation no.: LBD-1 5. Bearing geometry

2. Bearing type: Cylindrical, with two
axial grooves
3. L: D ratio: 0.35, 0.5, 0.75, 1.0, 1.5
4. Special items
Oil groove:
Calculation method: FDM (over-relaxation
method)
Boundary conditions: Reynolds condition
Number of grid points: Axial direction: 10 (uni-
form spacing) Circum-
ferential direction: 99
(uniform spacing) for
each of upper and
lower pads X

L!D=0.35
Static characteristics Stiffness coefficients Damping coefficients
s rp /Jtp-1
€ Q. Q. Kxx Kxy Kyx K.vy Cxx Cxy Cyx Cyy
11.5 . 0750 83.6 .0700 1. 00 215. 1.52 13.6 -13.1 2.68 27.1 2.23 2.23 17.3

8.55 .100 81.2 .0929 1. 00 160. 1. 59 10.3 -9.68 2.69 20.6 2.25 2.25 13.2

3.99 .200 72.4 .180 1. 00 76.0 1. 81 5.69 -4.42 2.56 11.2 2.40 2.40 7.42

2.38 .300 65.0 .260 .995 46 . 5 2.03 4.36 -2.57 2.37 8,43 2.38 2.38 5.09

1. 51 .400 58.2 .333 .983 30.8 2.40 3.86 -1.57 2.21 7.23 2.19 2.19 3.52

.961 ,500 51.5 .400 • 965 20.8 2.96 3.73 -.894 2.08 6.86 2.34 2.24 2.82

.587 ;600 45.0 .462 . 937 13.9 3.83 3.82 -.375 1. 98 6.80 2.10 2.10 2.09

• 327 .700 38.1 .518 . 894 8.82 5.23 4.10 .0633 1.86 7.19 2.02 2.02 1.59

.153 .800 30.7 .569 .834 5.03 7.85 4,67 .513 1. 76 8,22 1. 99 1. 99 1.18
.0466 .900 21.9 .616 .745 2.22 14.7 5.90 1. 07 1. 63 0.4 1. 80 1.80 .724

,0160 .950 16.1 .643 • 701 1.12 30.1 8.66 2.60 l. 84 15.9 2.43 2.43 .693

253
Calculation No. LBD-1

L!D=O.S
Static characteristics Stiffness coefficients Damping coefficients
s /J¢·1
e ¢ Qs Qe Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

5.96 .0750 82.6 .0663 1. 00 111.5 1. 77 13.6 -13.1 2. 72 27.2 2.06 2.06 14.9

4.43 .100 80.4 . 0880 .999 83.0 l. 75 10.3 -9.66 2.70 20.6 2.08 2.08 11.4

2.07 .200 71.6 .170 . 999 39.5 1. 88 5. 63 -4.41 2.56 11.2 2.22 2.22 6.43

1. 24 . 300 64.1 .243 . 997 24.4 2.07 4 .2 7 -2 . 56 2.34 8.50 2.32 2.32 4.73

. 7 98 . 4 00 57.5 .309 .986 16.4 2.39 3.75 -1.57 2.17 7.32 2.24 2.24 3.50

.517 • 500 51.1 .366 . 967 11.3 2.89 3.57 -.924 2.03 6.81 2.10 2.10 2.60

. 323 .600 44.7 .416 . 939 7.75 3.65 3.62 -.427 1.92 6.81 2.08 2.08 2.06

.187 . 700 38.2 .459 • 900 5.13 4. 92 3.88 .023~ 1. 83 7.32 2.16 2.16 1. 70

.135 . 750 34.9 .478 .875 4.07 5.90 4.11 .258 1.80 7.65 2.10 2.10 1. 46

.0926 .800 31.2 . 4 96 . 84 6 3.13 7.35 4. 46 .527 1. 78 8.17 2.05 2.05 1. 24

.0582 .850 27.2 .510 .806 2.30 9.56 4.92 .805 1. 73 9.12 2.06 2.06 1. 06

.0315 .900 22.7 .524 .758 l . 56 13. 8 5.76 1. 24 l. 72 10.6 2.03 2.03 .846

.00499 .975 12.5 .543 .663 .54 43.3 9.61 4.26 1. 99 22.4 2. 93 2. 93 .637

L! D=0.75
Static characteristics Stiffness coefficients Damping coefficients
s ht/J-1
e ¢ Qs Q. Kxx Kxy K yx Kyy Cxx Cxy Cyx Cyy

2.99 .0750 82.3 . 0596 .999 56.0 1. 83 13.5 -13.1 2.71 27.2 1. 82 1. 82 11.8

2.22 .100 80.1 .0791 .999 41.7 1. 79 10.3 -9.66 2.67 20.6 1.84 1.84 9.08

1. 05 .200 70. 8 .152 . 999 20.0 1. 93 5.54 -4.39 2.55 11.2 1. 96 1. 96 5.19

.637 .300 62.2 .210 . 998 12.6 2.17 4.10 -2.54 2.31 8.64 2.16 2.16 4.04

.418 .400 55.9 .261 .988 8.66 2.39 3.52 -1.60 2.09 7.44 2.11 2.11 3.11

.278 .500 50.3 .308 • 972 6.16 2.76 3. 32 -.960 1. 97 6.91 2.11 2.11 2.53

.181 .600 44 . 4 • 341 . 944 4.43 3.39 3 . 32 -. 4 91 1.85 6.76 1. 97 1. 97 1. 93

.110 .700 38.3 .367 .905 3.12 4.46 3.52 -.0564 1. 77 7.14 2.03 2,03 1. 61

.0590 .BOO 31.8 .384 .849 2,08 6.54 4, OS .437 1. 74 8.01 2.04 2.03 1.26

.00963 • 950 17.6 .396 .707 .741 24.9 7,88 2.65 1. 94 14,5 2.41 2.41 .760

254
 Calculation No. LBD-1

LID=!
Static characteristics Stiffness coefficients Damping coefficients
s Qs Q, JJ,p-1
€ ~ K xx Kxy K yx Kyy Cxx Cxy Cyx C yy

2.95 .0500 83.0 • 0351 .999 55.1 2.38 20.1 -19.8 2.90 ~0.7 1.65 1.65 1~.5

1. 96 . 075 80.8 . 0523 .999 36.6 2 . 18 13.5 -13.0 2.88 27.4 1. 66 J.. 66 9.81

1. 45 .100 78.7 . 0692 • 998 27.3 2. 04 10.2 -9.62 2. 77 20.8 1. 67 1. 67 7.55

. 68 9 . 200 69.9 .132 . 998 13.2 1. 99 5.45 -4. 38 2.56 11 . 3 1. 78 1. 78 4. 3 9

. 42 4 .300 61.0 .181 • 998 8.41 2 .20 3.96 -2 . 52 2 . 31 8. 67 1. 99 1. 99 3.50

.284 . 400 54.6 .220 . 991 5.93 2.37 3.34 -1.60 2.06 7.57 2.10 2.09 2.94

.194 .500 49.1 .251 .972 4.37 2 .6 7 3.09 -1.0 0 1. 89 6,99 2. 00 2.00 2.31

.131 .600 43.8 .274 . 943 3.26 3. 19 3.06 -. 55 3 1. 77 6.69 1. 8 2 1. 82 1. 74

.0828 .700 38.4 . 2 95 • 911 2.39 4 .1 6 3.29 -.085 1 1. 75 7. 07 2. 00 2.00 1. 59

.0469 .800 32.1 .302 .855 1. 69 6 .1 0 3.82 .420 1. 74 8.08 2 .15 2 . 15 1. 34

.0195 .900 24.0 .300 .770 1. 04 11.9 5.28 1. 3 2 1. 80 10 . 1 2 .0 6 2.06 . 91 6

.00866 . 950 18.0 .297 .710 .684 24.0 7.81 2. 7 9 2.01 14.2 2 .4 4 2.44 . 7 91

L!D=1.5
Static characteristics Stiffness coefficients Damping coefficients
s Q, Ijt/J-1
€ ~ Qs Kxx Kxy Kyx Kyy C xx C xy Cyx C yy

1. 82 .0500 82.7 .0268 .999 34 .1 2.~ 7 20.0 -19. 7 2.95 40.9 1. 47 1. 47 11.3

1. 21 .0750 80.5 .0399 . 998 22. 7 2 . 25 13.4 -1 3 . 0 2.91 27. 4 1. 47 l . 47 7. 65

.902 .100 78 .4 .0528 .997 16. 9 2.09 10. 1 -9. 62 2 .7 8 20. 8 1. 4 9 1. 4 9 5.9 2

.434 .200 68.3 .0985 .996 8.34 2 . 10 5.28 -4.34 2.59 11.4 l . 59 l . 59 3.53

.272 .300 59.4 .132 . 997 5.45 2 . 21 3.74 -2. 4 9 2.32 8. 73 l. 78 1. 78 2. 8 9

.1 8 7 • 4 00 53.0 . 15 7 . 997 3 .96 2 . 31 3 .0 7 -1. 58 2.04 7. 68 2 .00 2 .00 2.6 3

.13 3 .500 47.9 .175 • 977 3 .0 6 2 . 53 2.8 1 - 1 . 02 1.8 5 7.0 € l. 91 l. 91 2 .11

.0938 .600 43.2 .187 • 948 2. 3 9 2. 97 2.78 -.584 1. 73 6.85 l. 89 1.89 1. 77

.0624 • 700 38.3 .198 • 915 1.86 3.81 3. 01 -.12 1.72 6. 96 1. 91 1. 91 1. 51

.376 .800 32.4 .198 .859 1. 40 5.61 3 . 55 .375 1.71 7.81 2 .03 2 .03 1. 28

.0169 .900 24. 4 . 191 .770 . 92 8 11.0 5. 0 0 1. 24 1. 78 9,8 2. 0 4 2 .04 • 92<

.00786 .950 18.4 .188 .712 .637 22.9 7.64 2.81 2.05 14.1 2.51 2.51 .83

255
Calculation No. LBD-1

L!D=0.35

--- ____ L!D=0.75

--- - - L!D=l.S

---- L! D=0.35

--- - ___ L!D=0.75

20
--· - - - L!D=l.S

2
s

256
 Calculation No. LBD-1

- - - - - - - LID=0.35

------- LID=0.75

- - - - - - LID=1.5
20

,,
2
~
/
/
I
I
I
I
I

s
Ioor--------------------------------------------------------~

- - - - - - - L/ D= 0. 35

------- LID=0.75

----- -------

257
Calculation No. LBD-2

1. Calculation no.: LBD-2 5. Bearing geometry

2. Bearing type: Two-lobe 170°
3. L: D ratio: 0.5, 1.0, 1.5
4. Special items
Oil groove: 10° X 2
mP: 0.67
Calculation method: FEM
Boundary conditions: Reynolds condition
Number of grid points: Axial direction: 20 (uni-
form spacing) Circum-
ferential direction : 36
(uniform spacing) for
each of upper and
lower pads

L/D=0.5
Static characteristics Stiffness coefficients Damping coefficients
s Q, /Jt/J-1
€ ~ Qs Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

3.52 .050 78.8 .351 139. 145. 28 . 9 -75.1 14.6 196. -68.0 -67.7 43 .1
1. 75 .100 78.7 .351 69 .3 72.9 14.4 -37.5 7. 31 97 . 6 -33.7 -33. 6 21.6
.858 .200 78.5 .353 36.7 37.2 7. 71 -18.9 3.92 48.6 -16.3 -16.2 10.7
.554 .300 78.2 .357 22.3 25.1 5.54 -12.3 2.75 32.5 -10.3 -10.2 7.12
.397 .400 77.9 .362 16.2 19.1 4.53 -8.86 2.19 24.5 -7.13 -7.11 5.30
.299 .500 77.4 .368 12.5 15.5 4.02 -6.71 1.89 19.7 -5.04 -5.02 4.19
.230 .600 76.7 .375 9.89 13.3 3.60 -5.11 1.65 16.1 -3.63 -3.61 3.39
.178 .700 75.7 .383 7.97 11.8 3.54 -3.87 1.53 13.9 -2.39 -2.37 2.81
.136 .800 74.5 .392 6.28 10.8 3.51 -2.80 1.43 12.1 -1.43 -1.42 2.31
.101 .900 72.7 .402 5.39 10.3 3.74 -1.82 1.40 11.2 -.441 -.432 1.91
.0841 .950 71.4 .407 4.99 10.3 3. 77 -1.28 1.35 10.6 -.045 -.036 1.68

258
 Calculation No. LBD-2

L!D=l
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Qs Q, lj<P-' Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy
1. 73 .050 81.4 . 196 66.4 132 . 33.7 -65.4 11.6 210. -69.4 -69.1 41.5
.863 .100 81.3 .197 33.1 65.9 16.8 -35.6 5.!)9 104. -34.4 -34.3 20.7
.425 .200 81.1 .199 16.4 33.7 9.07 -16 . 5 3.08 52.1 -16.7 -16.6 10.4
.276 .300 80.7 .201 10.7 22.7 6.33 -10.8 2.16 34.5 -10.6 -10.5 6.92
.200 .400 80.3 .205 7.87 17.1 5.03 -7.87 1. 74 25.8 -7.37 -7.33 5.19
.152 .500 79.7 .209 6.12 13.8 4.28 -6.01 1. 54 20.6 -5.24 -5.21 4.15
.119 .600 78.9 .214 4.92 11.7 3.75 -4.65 1.41 16.8 -3.77 -3.75 3.41
.0942 .700 78.0 .219 4.05 10.2 3.49 -3.59 1.35 13.9 -2.73 -2.71 2.83
.0742 .BOO 76.7 .225 3.40 9.25 3.34 -2.69 1.30 11.9 -1.85 -1.84 2.35
.0575 .900 74.9 .231 2.98 8.73 3.27 -1.83 1.25 10.3 -1.08 -1.07 1.92
.0497 .950 73.8 .234 2.92 8.42 3.26 -1.37 1.23 9. 77 -. 717 -.710 1.72

LID=1.5
Static characteristics Stiffness coefficients Damping coefficients
s ¢ Qs Q, fJcp-1 Kyy Cyx Cyy
€ Kxx Kxy Kyx Cxx Cxy
1.38 .050 82.5 .126 51.9 124. 36.0 -61.0 10.9 218. -69.7 -69.4 40.6
.682 .100 82.4 .126 25.7 62.0 17.2 -30.2 5.33 106. -34.3 -34.2 20.2
.336 .200 82.2 .127 12.8 31.5 9.49 -15.3 2.78 52.9 -16.7 -16.6 10.1
.219 .300 81.8 .129 8.39 21.2 6.69 -10.1 1.96 35.2 -10.6 -10.5 6.78
.159 .400 81.3 .131 6.16 16.0 5.28 -7.39 1.60 26.3 -7.31 -7.30 5.11
.122 .500 80.7 .134 4.81 13.0 4.38 -5.66 1.39 20.7 -5.33 -5.31 4.08
.0959 .600 79.9 .138 3.89 10.9 3.88 -4.42 1.30 16.6 -3.99 -3.98 3.35
.0765 .700 78.9 .141 3.23 9.53 3.47 -3.42 1.25 13.7 -2.93 -2.92 2.80
.0610 .800 77.6 .145 2.74 8.55 3.29 -2.61 1.23 11.5 -2.09 -2 .08 2.33
.0479 .900 75.8 .149 2.45 7.97 3.22 -1.84 1.22 9.99 -1.31 -1.30 1.92
.0419 .950 74.6 .151 2.45 7.69 3.12 -1.40 1.20 9.40 -.924 -.918 1.72

259
Calculation No. LBD-2

L!D=0.5

------ LI D= I
- - - L!D=l.5

1.0~------------------------------------------------------~

L!D= O.S
LID= I
L!D=l.5

260
 Calculation No. LBD-2

Ll D=0 .5
------ LID=!

- - - LID=l.5

- - - L ! D= 0.5

------LID=!

---LID=1.5

...,
G

261
Calculation No. LBD-3

1. Calculation no.: LBD-3 5. Bearing geometry

2. Bearing type: Tilting pad, 5 pads,
LOP
3. L: D ratio: 0.3, 0.5, 0.7
4. Special items
mP: 0.5
P: 600
rxjp: 0.5 y
Calculation method: FDM
Boundary conditions: Reynolds condition
Number of grid points : Axial direction: 10 (uni-
(per pad) form spacing) Circum-
ferential direction: 30
(uniform spacing)
n
Definition of Q,N: X

L/ D=0.3
Static characteristics Stiffness coefficients Damping coefficients
s
e QIN ljl/1-' Kxx K yy Cxx Cyy

2.01 .100 4 . 96 66.2 14.7 14.1 15.7 15 . 4

.953 .200 4.96 31.9 8.59 7 . 07 8.35 7.52

.573 .300 4.95 19.7 7.19 4. 60 6.07 4.73

.367 .400 4.93 13.2 7.15 3.24 5.08 3.20

.233 .500 4.92 8.86 7.97 2.30 4.65 2.17

.140 .600 4.89 5.84 9.63 1. 57 4. 72 1. 41

.0769 .700 4.86 3.65 12.5 . 977 5.14 .843

.0350 .800 4.82 2 . 05 18.0 • 510 6.07 .419

.0102 .900 4 . 76 .910 32.7 .1 7 3 8.26 .135

.00349 .950 4. 71 . 580 65.3 . 0635 11.4 .0486

262
 Calculation No. LBD-3

L!D=O.S

Static characteristics Stiffness coefficients Damping coefficients

s
€ QIN fJ'P- I Kxx Kyy Cxx Cyy

1. 20 .100 4.61 39 . 8 14.4 13.8 17.1 16 . 7

• 572 .200 4.61 19 . 3 8.37 6.94 9.06 8.20

• 346 .300 4.60 12.0 6.95 4.54 6 . 58 5.19

.223 .400 4.59 8.07 6.88 3.21 5 . 52 3.53

.144 .500 4.58 5. 4 7 7.56 2.30 5.00 2.43

.0885 .600 4.56 3 . 72 9.10 1. 60 4.90 1. 61

• 0497 .700 4.54 2.27 ll. 5 1. 02 5. 2 9 .980

.0237 . 8.00 4.51 1. 42 16.8 .553 6.25 .509

. 0144 .850 4.49 . 860 2 0.4 .362 7.03 .326

. 00757 .900 4.47 .697 30.3 .204 8.40 .180

• 00277 .950 4.44 .487 60.5 .0802 11.5 . 0689

- 7
LID-o
Static characteristics Stiffness coefficients Damping coefficients
s
€ QIN ht/J-1 Kxx Kyy Cxx Cyy

.969 .100 4.42 32.2 14.2 13.7 18.0 17.6

.463 . 2 00 4.41 15.6 8.26 6.87 9.5 3 8.63

. 281 .300 4.41 9.73 6.82 4. 5 0 6.90 5 . 47

. 182 • 400 4.40 6.59 6.73 3.20 5. 77 3.74

.118 .500 4.39 4.51 7.37 2.30 5 . 22 2.58

.0735 .600 4.37 3.10 8.82 1. 61 5.10 l . 73

.0418 .700 4.36 1. 92 11.1 1. 04 5 . 44 1. 07

.0203 .800 4. 34 1. 2 3 16.2 . 577 6.3 4 .565

.0126 .850 4.32 . 759 19.6 .383 7.21 .367

.00674 .900 4.31 . 632 29.3 .220 8.47 .206

• 00254 .950 4.29 .454 60.8 .0887 10 . 7 .0812

263
Calculation No. LBD-3

5.0

---------
:; 4.5
01
0.6
---- --- ---
0.4--

0.2 ------ L!D=0.5

4.0 ---L/D=0.7

or
s

- - - L!D=0.3

------ L/D=0 . 5

- - - L!D=0.7

264
 Calculation No. LBD-3

---L!D=0.3

LID=0.5

20 - - - L!D=0.7

....
~

50 ---L!D=0.3

L / D=0.5

20 - - - L/D=0 .7

265
4 Explanations

4.6 Influences of Turbulent Flow on Journal Bearing Characteristics

The transition of bearing oil flow from the laminar 1.0
- - Re=O
to the superlaminar has similar influences on the
bearing characteristics as the increase of oil viscos-
~ ----- Rc =3000

\
0.8
- · - Rc=aooo
ity. In the superlaminar region the load-carrying
capacity and the friction coefficient increase with 0.6
increasing Reynolds number and the discrepancies
between the actual bearing characteristics and the 0.4
data calculated by the laminar theory widen. In this
section the influences of superlaminar flow are dis- ~.
cussed with reference to the data listed below 0.2
~
(pp. 267- 277). ~
0
0. 1
a s,,,
Calculation no. Bearing type Specifications
TBL-1 Two-lobe L:D = 1, mP = a.67, p = 1 7ao,
bearing 2 oil grooves of 10° each
TBL-2 Four-pad L: D = 1, mP = a, p= sao, 10 -Re=O
tilting pad rxfp = a.5, LBP ----- Re=3000
bearing

The calculation method for bearing characteris-

tics in the turbulent region is explained in
Sect. 4.1.2. In this section the method of interpolat-
ing bearing data in the turbulent region is shown.
For this purpose, an effective Sommerfeld num-
ber is defined by Eq. (1). This correction is similar to
0.1
the introduction of turbulent flow coefficients into b S,11
Eq. (18) in Sect. 4.1.2, which corresponds to the in-
crease of viscosity by turbulent diffusion. Fig. 4.6.1 a, b. Effective Sommerfeld no. for turbulent flow.
Seff = (1 + 0.0007768 R~· 9) • S (1) 4 - tilting pad bearing, L /D = t.a, mp =a, p =sao, rxjp = a.5,
LBP
The data calculated from Eq. (1) are shown in
Fig. 4.6.1 for the case of calculation no. TBL-2. The
calculation in the turbulent region is worked using
turbulent flow coefficients Gx and Gz, depending on The transition of oil flow from the laminar to the
the local Reynolds number of Elrod-Ng [1]. The turbulent occurs at Reynolds numbers ranging
coefficient 0.0007768 and the exponent 0.9 of from 1500 to 3000, though it depends on the size of
Reynolds number in Eq. (1) are values obtained by bearing and the operating condition.
trial and error. It is found that the eccentricity e,
stiffness coefficients Kii• or damping coefficients Cii
can be represented by every single curve. Especially,
data calculated for the condition of a nonzero Reference for Sect. 4.6
Reynolds number show good agreement in prac-
tice. The correction equation, however, depends on [1) Elrod, H.G., and Ng, C.W., J. Lub. Technology, Trans.
the type of bearing and the turbulence theory used. ASME, Ser. F, S9-3 (1967), 346.

266
 Calculation No. TBL-1

1. Calculation no.: TBL-1 5. Bearing geometry

2. Bearing type: Two-lobe
170"
3. L: D ratio: 1.0
4. Special items
Oil groove: 10° X 2
mP: 0.67
P: 170°
Calculation method: FDM
Boundary conditions: Reynolds condition
Number of grid points: Axial direction: 22
(uniform spacing)
Circumferential di-
rection: 33 (uniform
spacing) for each of
upper and lower
pads X

Reynolds number: 0, 3000, 5000

Re= 0
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Qs Qe h"'-1 Kxx Kxy K yx Kyy Cxx Cxy Cyx Cyy

4.90 .0172 81.2 .218 .379 133 391 117 -198 30.7 609 -197 -197 120

2.92 . 0289 81.2 .218 .379 78.8 234 67.2 -118 18.2 361 -118 -117 70.5

1.97 .0430 81.2 .218 .379 53.2 158 44.6 -79.6 12.3 241 -79.2 -79.2 47.3

• 956 .0898 81.2 .219 .378 25.5 77.3 20.9 -39.0 6.07 113 -39.4 -39.2 23.8

.521 .165 81.0 . 219 .377 14.1 42.3 12.5 -21 . 1 3.42 63.5 -20.7 -20.7 12.9

• 305 . 277 80 . 7 . 222 .372 8.26 25.6 7.40 -12.3 2.23 38.1 -12.2 -11.8 7. 80

.192 .427 80.0 .228 .363 5.33 16.8 5.30 -7.80 1. 71 24.7 -7 . 00 -6.98 5.15

.100 .655 78.3 .239 .344 2.93 10.7 3.82 -3.97 1. 34 15.0 -2.81 -2.79 3.01

.0513 .925 74 . 0 .256 .311 1. 66 8.92 3.55 -1.61 1. 24 11.1 -.49C -.481 1.80

.0313 1 . 07 69.1 .263 . 292 1.15 9. 61 4.02 -.401 1. 36 10.0 • 33( .345 1.29

.0202 1.15 64.0 .265 .281 • 865 11.7 4.80 . 570 1.56 10.3 • 87 .873 • 995

.0104 1. 15 53.9 .247 . 296 .559 20.2 6.60 2.24 1. 90 11.5 1.55 1.58 • 731

267
Calculation No. TBL-1

Re =3000
Static characteristics Stiffness coefficients Damping coefficients
s
e ¢ Qs Q, ljtV' Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

4. 96 .0119 85.2 .250 .393 366 567 209 -299 24.9 908 -318 -307 203

3.11 .0189 85.2 .250 .393 230 356 131 -188 15.6 570 -199 -193 127

2.09 .0283 85.2 .250 • 393 153 237 86.0 -126 10.4 377 -134 -129 84.9

• 974 .0629 85.2 .250 .393 71.5 111 39.8 -58.6 5 . 05 176 -62.3 -61.2 39.8

.517 .119 85.0 . 251 .392 37.8 59 . 5 21 . 3 -31 . 3 2.80 93.6 -32.8 -31.9 21.1

. 315 .195 84.6 .253 . 387 23.2 36.6 13.7 -19.3 1. 71 58.3 -20.0 -19.2 13.1

. 196 .310 84.3 .258 .380 14 . 4 23.4 8.52 -12.0 1.32 36.3 -12.1 -11.7 8.42

.0992 .535 83.0 .274 .353 7.37 13.0 5.21 -6.19 1.11 20.0 -5.64 -5.44 4.77

.0514 .812 81.1 .303 . 299 3.85 8.58 3.62 -3.01 1.02 11.9 -2.24 -2. 14 2.92

.0306 1. 07 77.5 . 342 . 212 2.43 7. 55 3.91 -1.48 1.04 9.37 -.422 -.340 2.12

.0202 1. 18 73.3 . 353 .169 1.68 8.05 4.48 -.431 1. 38 8.99 .678 .743 1.72

.0105 1. 26 63.8 . 352 .139 .986 12.4 6.13 1. 41 1. 88 10.2 1. 78 1.86 1.28

Re=5000

Static characteristics Stiffness coefficients Damping coefficients

s
e ¢ Qs Q. !Jtf- I Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

4.88 .0095 85.9 .258 . 394 467 682 262 -365 26.5 1110 -394 -377 255

3.02 .0154 85.9 .258 .394 289 422 163 -226 16.3 685 -244 -233 157

2.07 .0225 85.9 .258 . 394 198 290 110 -155 11.2 466 -167 -160 107

1.02 .0459 85.9 .259 .393 96.9 142 53.0 -76.4 5.52 227 -82.4 79.1 52.9

.514 .0927 85.9 . 259 .393 48.5 72.6 26.1 -38.5 2.84 110 -41.4 -39.7 26.8

.301 .158 85 . 8 .260 .390 28.6 42.5 16.2 -22.6 1. 75 68.0 -24.0 -23.2 15.9

.202 .253 85.4 .263 .388 19.2 29.0 11.0 -15.2 1.25 45.8 -15.9 -15.5 10.9

.0997 .458 84.6 .275 .380 9.51 15.2 5. 81 -7 . 60 . 943 22.8 -7.45 -7.12 5.74

.0486 .750 82.8 • 302 • 315 4.69 9.11 3.95 -3.65 .874 12.9 -3.06 -2.71 3.17

.0299 1.01 79.1 • 338 .251 3.00 7.24 3.62 -2.01 • 980 9. 92 -1.09 -.955 2.36

.0199 1.16 75.4 .354 .203 2.08 7.46 4.00 -.904 1.16 8.72 .0796 .165 1.85

.0103 1. 30 67.4 .372 .129 1.14 12.9 5.56 1. 74 1.64 8.42 1.28 1.40 1.27

268
 Calculation No. TBL-1
1.4
€
- - · - - Re= 0
1.2 "",... -------- Re=3000
,•,

~ Re =5000
1.0 1.0
''
\\ '
----- ---- ----- ----- -
\
0.8 0.8 ---------------------~------

0.6 0 .6
c::y w

Qs
45
0.4 0.4

0.2 0.2 - 30

15

0 0
0. 005
s

50

20

10

~
'
~
2
Re= 0

1 Re =3000

R e =5000

0.5

0.2~~~--~-~~~~~~-~-~~~~~---L-~~~
0.005 0.01 0.02 0.05 ~~
0.1 0.2 0.5 2 5 10
s

269
Calculation N o. TBL-1

Re= 0

Re =3000

Re=5000

5 10
s

-- Re= 0

---- ---- Re=3000

I
I
Re=5000

I
0. 05 0.1 0.2 0.5 1 2 10
s

270
 Calculation No. TBL-1

Turbulent flow coefficients:

1. 0
9
8
z5

0.1
9
8
l .....;;
.......... !'-...

""'
5 ........
4
I'-
3

2
I'..
I'-..
I' Gx
0.0 1
9
8
"""'
7
6
Gz
4
3

0.00 1
3 4 5 6 7 8910' 3 4 8
5 6 7 910' 345678910'
10 2

Local Reynolds No. Ro= ~h

271
Calculation No. TBL-2

1. Calculation no.; TBL-2 5. Bearing geometry

2. Bearing type: Tilting pad, 4 pads,
LBP
3. L:D ratio; 1.0

4. Special items
mP: 0
P: 80°
rxfP: 0.5 y
Calculation method: FDM
Boundary conditions: Reynolds condition
Number of grid points: Axial direction: 18
(per pad) uniform spacing)
Circumferential di-
rection: 18 (uniform
spacing) X

Reynolds number: 0, 1000, 3000, 5000,

8000, 12000
Tubulent flow coefficient: The same as for cal-
culation no. TBL-1

Re= 0
Static characteristics Stiffness coefficients Damping coefficients
s
€ rp Q. Qe JJ.p-l Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

3.30 .050 0 • 0417 1.97 5 8.0 .545 0 0 .489 11.5 0 0 11.5

2.15 .100 0 .0579 1.96 38.0 .701 0 0 .669 8.05 0 0 8.03

1..28 .zoo 0 • 0784 L93 23.1 .975 0 0 .963 5.64 0 0 5.62

.868 .300 0 • 0918 1.91 16.1 1.22 0 0 LZZ 4.61 0 0 4.60

.615 • 400 0 • 101 L90 1L9 L46 0 0 L47 4.05 0 0 4.0 3

• 442 .500 0 . 107 L88 9.00 L72 0 0 1.7 3 3.69 0 0 3.6 8

• 317 .600 0 ·111 L86 6.91 z.oo 0 0 2.01 3.49 0 0 3.45

.225 .700 0 .112 L85 5.33 2.32 0 0 2.35 3.31 0 0 3.30

.155 .800 0 • 112 1.83 4.11 2.73 0 0 2.76 3.21 0 0 3.19

.103 .900 0 .109 L8Z 3.11 3.25 0 0 3.28 3.12 0 0 3.11

.0826 • 950 0 .108 1.81 2.71 3.5 7 0 0 3.63 3.12 0 0 3.10

272
 Calculation No. TBL-2

Re=lOOO
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Qs Q, JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

2.28 . 050 0 .0485 1.97 101. . 508 0 0 .464 10.7 0 0 1 Q.7

1.50 . 100 0 .0672 1.95 67.0 .623 0 0 .598 7.52 0 0 7.51

• 917 .200 0 • 09 04 1.93 41.3 .847 0 0 .838 5.32 0 0 5.29

.639 .300 0 .105 1.91 29.2 1.05 0 0 1.04 4.37 0 0 4.36

.465 .400 0 . 11 4 1.89 21.6 1.27 0 0 1.27 3.84 0 0 3.83

• 344 ,500 0 . 120 1.87 16.4 1.4 3 0 0 1.32 3.44 0 0 3.47

.254 .600 0 .123 1.86 12.5 1.74 0 0 1.7 5 3.36 0 0 3.35

• 18 5 .700 0 .123 1.84 9.56 2.06 0 0 2.08 3.17 0 0 3.16

•1 31 .800 0 .122 1.83 7.18 2.49 0 0 2.52 3.09 0 0 3.07

,0893 . 900 0 .118 1.82 5.31 3.04 0 0 3.08 3.04 0 0 3.04

.0725 • 9 50 0 • 116 1.81 4.48 3.38 0 0 3.43 3.03 0 0 3.0 2

Re =3000
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Q. Qe JJ.p-1 Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

1.37 .050 0 .0545 1.97 to a .480 0 0 .457 10.4 0 0 10.3

. 907 .100 0 .0752 1.95 71.7 . 567 0 0 .581 7.31 0 0 7.27

.559 .zoo 0 • 102 1.92 44.4 .765 0 0 .792 5.16 0 0 5.13

.392 .300 0 .119 1.90 31.5 .988 0 0 .984 4.22 0 0 4.21

.288 .400 0 .130 1.88 23.4 1.18 0 0 L18 3.71 0 0 3.70

.215 .500 0 .138 L87 17.8 1.38 0 0 1.38 3.39 0 0 3.38

. 1 62 .600 0 .142 1.85 1 3. 7 1.62 0 0 1.63 3.18 0 0 3.17

.120 .700 0 .142 1.84 10.5 1.89 0 0 1.90 3.0 5 0 0 3.04

.0874 .800 0 .140 1.83 7.96 2.27 0 0 2.29 2.97 0 0 2.96

.0615 . 900 0 . 135 1.81 5.92 2.74 0 0 2.78 2.92 0 0 2.91

.0507 .950 0 . 132 1.81 5.00 2.74 0 0 3.09 2.91 0 0 2.8 5

273
Calculation No. TBL-2

Re =5000
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Qs Qe JJ,p-l Kxx Kxy Kyx Kyy Cxx Cu Cy x Cyy

LOO . 050 0 .0571 1.96 108 .480 0 0 . 451 10.1 0 0 10.1

.667 . 100 0 .0789 1.95 71.8 .551 0 0 .509 7.15 0 0 7.16

.413 .zoo 0 . 107 1.92 4 4.7 . 746 0 0 .758 5.06 0 0 5.03

.292 .300 0 . 125 1.90 31.8 .943 0 0 .940 4.15 0 0 4.14

• 216 .400 0 .137 1.88 2 3.8 1.1 2 0 0 1.12 3.64 0 0 3.63

.163 .500 0 • 14 5 1.86 18.2 1.32 0 0 1.32 3.34 0 0 3.3 3

.123 .600 0 .149 1.85. 14.1 1.5 3 0 0 1.5 7 3.14 0 0 3.13

.0920 .700 0 • 150 1.84 1 0.8 1.84 0 0 1.85 3.01 0 0 3.00

.0675 .800 0 .148 1.82 8.19 2.18 0 0 2.21 2.92 0 0 2.91

. 0480 .900 0 .143 1.81 6.09 2.61 0 0 2.64 2.86 0 0 2.85

.0398 .950 0 .140 1.8 0 5.19 2.90 0 0 2.93 2.84 0 0 2.83

R e=BOOO
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Qs Qe JJ,p-• Kxx Kxy Kyx Kyy Cxx Cxy Cyx Cyy

.737 .050 0 .0590 1.96 10 7. .475 0 0 .438 10.0 0 0 9.99

. 491 .1 00 0 .0818 1.95 71.5 .564 0 0 • 541 7.04 0 0 7. 03

.305 .200 0 . 111 1.92 44.6 . 735 0 0 .726 4.9 7 0 0 4.96

. 217 .300 0 .130 L90 32.0 . 909 0 0 . 906 4.10 0 0 4.09

.161 . 4 00 0 .142 1.88 24.0 1.09 0 0 1.10 3.62 0 0 3.61

.122 .500 0 . 151 1.86 18.4 1.29 0 0 1.30 3.31 0 0 3.30

.0924 .600 0 .155 1.85 1 4.2 1.48 0 0 1.46 3.10 0 0 3.10

.0696 . 700 0 . 15 6 1.83 10.9 1. 70 0 0 1.79 2.98 0 0 2.96

• 0514 .800 0 . 155 1.82 8.30 2.11 0 0 Z.H 2.90 0 0 2.90

.0368 .900 0 .150 1.81 6.19 2.44 0 0 2.58 2.85 0 0 2.82

.0306 .950 0 . 147 1.80 5.2 8 2.82 0 0 2.86 2.81 0 0 2.80

274
 Calculation No. TBL-2

Re= 12000
Static characteristics Stiffness coefficients Damping coefficients
s
€ ¢ Q., Q,. !J<P- ' K.rx Kx.v K.vx Kyy Cxx Cxy Cyx Cyy

• 54 7 .050 0 .0609 1.96 10 5. • 4 64 0 0 . 429 9.80 0 0 9.79

•3 65 . 10 0 0 .0843 1.9 4 6 9.9 .545 0 0 .525 6.9 0 0 0 6.9 0

.229 .200 0 • 1 14 1.92 44.0 .703 0 0 . 695 4.92 0 0 4.91

• 16 4 .300 0 . 133 1.90 31.7 . 876 0 0 .873 4.06 0 0 4.0 5

• 123 .400 0 . 14 6 1.8 8 2 3.9 1.0 6 0 0 1.0 6 3.57 0 0 3.56

.0932 .500 0 • 15 4 1.86 18.4 1.2 6 0 0 1.26 3.27 0 0 3.26

• 0 712 .600 0 .159 1.85 14.3 1.37 0 0 1.4 8 3.11 0 0 3.07

.0537 . 700 0 . 161 1.83 11.0 1. 7 3 0 0 1.7 4 2.9 5 0 0 2.9 4

. 0399 .800 0 • 16 0 1.82 &36 2.05 0 0 2.08 2.86 0 0 2.85

.0287 .900 0 . 156 1.81 6.23 2.48 0 0 2.5 1 2.80 0 0 2.79

.0240 . 9 50 0 .153 1.80 5.31 2.7 4 0 0 2.7 8 2.7 8 0 0 2.76

275
Calculation No. TBL-2

2.0 1.0 too·

1.9 0.8 so·

1.8 0.6 60 .

~ (<) -e..

0.2 0.4 40 .

0.1 0.2 20.

0 0 0
10

200~----------------------------------------------------------~

10
s

276
 Calculation No. TBL-2

50r---------------------------------------------------------------.

10

50~----------------------------------------------------------~

....,
G

10
s

277
4 Explanations

4.7 Analysis of Bearing Characteristics with Due Consideration

for Viscosity Variation

4. 7.1 Thermohydrodynamic Lubrication u, v, w: film velocity components in the x, y and

Theory z directions, respectively

In fluid film bearings, heat is generated in the film When the dimensionless film temperature
due to journal rotation and is dissipated by heat T=lyKifi 2 (T- T.n)f(pincv) is introduced, the di-
transfer. The hydrodynamic lubrication theory mensionless energy equation can be expressed as
dealing with the effect of heat generation and dissi- follows:
pation is called the thermohydrodynamic theory _, ar _,ar _, ar _ o2r
ao + v a;; + w a[ = P., a,2 + <P
1
(THD theory). This theory can give the precise per- u (3)
formances of fluid film bearings. The isoviscous the-
ory (IV theory) in contrast, cannot give the bearing where
temperatures, by definition, because it is based on
the assumption of constant temperature and there-
fore also of constant viscosity throughout the film.
The effective film temperature has been usually as-
sumed in the IV theory to obtain the bearing char-
v' = lf/- 1 H(v- ii'llfl ~;) (4)

!G:
acteristics. Although this method is very simple, the
prediction of bearing performance is inaccurate, be-
cause no generally accepted method of estimating v =- If/ H + ~~ - ; ~: ~~) d'l
the effective temperature has yet been proposed.
Recent investigation [1, 2] has shown that an effec-
tive viscosity applicable to the calculation for all
bearing characteristics does not exist and that ap-
plication of the THD theory is necessary for precise and j1 = fll flin (dimensionless viscosity coefficient),
prediction of the bearing characteristics. An outline P., = yKC 2 cvj2 (Peclet number), and !lin is the vis-
of the THD theory is described in this section. cosity coefficient of the supplied oil.
The steady state Reynolds equation for THD For the THD case the Reynolds equation, the
theory can be obtained by settings' = 0 and¢' = 0 energy equation of the lubricating film and the heat
in Eq. (10) of Sect. 4.1.1: conduction equations in the bearing and the jour-
nal should be solved simultaneously. In the case of
aea (H 3
E
aA)
ao + a(a (H 3
E
aA)
a( = aea (H G) (1) unsteady state condition, Reynolds [Eq. (9) in
Sect. 4.1.1] is simplified by using the two-dimen-
sional viscosity distribution flm(O, () which can be
where E and G are parameters expressing the effect calculated by averaging jl(O, 17, ()in the film thick-
of the film viscosity distribution and can be calcu- ness direction and given as follows [3]
lated from Eqs. (6) and (7) in Sect. 4.1.2. When E
1
and G are given, the film temperature is calculated
by solving the energy equation of the film. The
flm(O, () = Jfl(O, 17, () d17. (5)
0
energy equation for the fluid film bearings is given
by the following equation: E and G are expressed as follows:

yKJ (UaT aT aT)

ax- + Voy- + Woz- E = ---=---
1
(6)
12flm(O, ()'

ay + (ow)
ooy2T + fl {(ou)
2 2
= , 2
JA ay } (2) Substituting these values forE and G into Eq. (9) in
Sect. 4.1.1, the following Reynolds equation for the
where unsteady state case can be obtained:

(H (H
y: specific gravity of the lubricant
K: specific heat of the lubricant a 3
aA) a 3
aA)
J: heat equivalent of work ae flm . ae + o( flm . o(
2: heat conductivity of the lubricant
T: film temperature
ft: viscosity coefficient of the lubricant
= 6 ~; + 12 (s</J' sine- 8 1
cos 0). (7)

278
 4.7 Analysis of Bearing Characteristics with Due Consideration for Viscosity Variation

4. 7.2 Examples of Temperature ing obtained by THD theory. In all figures, the
Distribution in the Bearing Bush Sommerfeld number S* = fl;nNLD /(W'f/ 2 ) is plot-
ted along the abscissa. In Fig. 4.7.2, fl;n is the viscos-
ity coefficient of the supplied oil. The eccentricity
Figure 4.7.1 shows the typical temperature distribu- ratio and the attitude angle are shown against S*
tion at the central plane of the bearing bush. The with Peclet number Pe as a parameter. The friction
experimental isotherm (i) is found to be in good coefficient fJ '1'- 1 and the dimensionless film stiff-
agreement with the theoretical one (ii) by the THD ness and damping coefficients K, C are shown in
theory [8). On the metal surface, the point of maxi- Figs. 4.7.3, 4.7.4 and 4.7.5, respectively. The Peclet
mum temperature is located near the position of the number is one of the most important parameters
minimum film thickness and the minimum temper- which affect the temperature gradient in the film
ature, near the oil groove at the crown. The heat thickness direction.
which flows to the normal direction of the Figure 4.7.6 shows the experimental and theoret-
isotherms is conducted in the bearing bush and also ical results [4] for dimensionless load-carrying ca-
dissipated to the surroundings. The temperature on pacity - the eccentricity ratio curve for the circular
the outer surface of the bearing bush varies consid- journal bearing with one oil groove at the crown of
erably in the circumferential direction. the upper metal part. The experimental condition is
as follows:

4.7.3 Bearing Characteristics Calculated Bearing diameter: 100mm

by THD Theory Bearing width: 70mm
Radial clearance: 78Jlm
Figures 4.7.2 to 4.7.5 show the theoretical results of Lubricant: # 90 turbine oil
bearing characteristics for the circular journal bear- Supplied oil temperature: 40 °C

a Experimental b Theoretical

Fig. 4.7.1 a, b. Isotherms in the bearing bush. Eccentricity ratio = 0.7, Attidue angle = 43o . _ .
N = 2502 rpm; W = 5.61 N; P;n = 98 kPa; T;n = 40.0 °C; I;, = 29.2 oc; Clearance ratio = 0.00157; Lubncant = 90 turbme otl

0.8 75'

0.6 - - THD 60'

45' -e-.
0.4
30'
0.2
15'
0 0.
0.~~rT~~~~~~
005 0. 1 0. 02 0. 05 0.1 0. 2 0.~~~~~~~~
5 1 2
S* = p.,nNLD/(Wrp 2 )
Fig. 4.7.2. Eccentricity ratio and attitude angle according to the THD theory and the isoviscous theory

279
 4 Explanations

100r-------------------,

50

20

10

Pe =7 .05

0.5

0. 2~.;rr-_..,...,...~"';;;!-~~-";<"'-:;-'~~.......~~~._._"7'"~ Fig. 4.7.3. Friction coefficient according to the THD theory

0.01 .02 0.05 0.1 0.2 0.5 1 2 5 10
S*=p;nNLD/(W¢' 2 )

100

50
Kxx
,, , .
, ;

,, , ,
20 ;

, ,, ,
10 Kxy ,,, ,,
,
:::;-
,,
~
,
, ;",
,
5 ,' ;

- -·
2
Kyy

0.5
----- Isoviscous theory
Fig. 4.7.4. Dimensionless stiffness coefficients of oil film
0.2 according to the THD theory and the isoviscous theory
0.005 0.01 0.02 0.05 0.5 1 2 5 10
S*=p.;nNLD/(W¢ 2 )

100r-------------------~

50

20

10
·~

1:...) 5

- - THO theory
Pe=3.52
0.5
----- Isoviscous theory

0. 2~~-:-:-....,...::-::-'---';:-~~'-:--::'-..,..-'......L::i+.I...LL~---,!,--"0......0..~-LUJ
0.005 0.01 0.02 0.05 0.1 0.2 0.5 2 5 10
Fig. 4.7.5. Dimensionless damping coefficients of oil
S*=p;nNLD/(W¢' 2 ) film according to the THD theory and the isoviscous
theory

280
 4.7 Analysis of Bearing Characteristics with Due Consideration for Viscosity Variation

I.Or---------------------------------------,
• 500rpm
+ 1750rpm
THD theory ~.,

(2\~
0.8
0 • 2750rpm
·~
.A. 3500rpm
.... 0 6
a ·
:g
1l 0.4 THD theory
4~~~
./_,.""'
8 (17 SOrpm) ,.:-.
,. ""- Isoviscous theory
w ~z

0.2

0~~~~~~~~--~~~--~----~~--~
Fig. 4.7.6. 0.01 0.5 2 5 100.1 0.2
Experimental and theoretical eccentricity ratios
P~ = W¢' /(p.;n (J)LD)
2

X I0-3
The experimental results are obtained by varying
25 the bearing load under the constant revolutional
0 1000rpm speed. The THD theory agrees well with the exper-
<> 1500rpm iment. With the increase of the speed, the eccentric-

20
•... 1750rpm
2500rpm 0
ity ratio against the same Po* is found to increase
because of the decrease in the film effective viscosi-
0 2750rpm ty.
6 3000rpm Figure 4.7.7 shows the experimental and theoret-
ical frictional coefficients. The frictional coefficient
..:::: decreases with the increase in speed because of the
..... 15
c:
OJ
·u same effect as described above.
e
OJ Little data on the stiffness and damping coeffi-
0
u
cients of the film for the THD case has been provid-

:f<
c:
0
ed so far [5, 6]. Recent research work showed that
:~ 10
~ the theoretical film coefficients deriving from the
THD theory were in good agreement with the ex-
perimental ones [7].
, TI!Dthomy
(2500rpm)
5
,!3
~~ References for Sect. 4. 7

0~~----~----~------~----~ [1] Seireg, A., Ezzat, H., Trans. ASME, Ser. F, Vol. 95, No. 2
0.05 0.1 0.2 0.5 1.0 (1973), 187.
[2] Yu, T.S., Szeri, A.Z., Trans. ASME, Ser. F, Vol. 95, No. 3
s• =p.;nNLD/(W<p2) (1973), 298.
(3] Mitsui, J., Yamada, T., Bull. of the JSME, Vol. 22, No. 172
Fig. 4.7.7. Experimental and theoretical friction coefficients
(1979), 1491.
[4] Mitsui, J., Trans. of JSME, Ser. C, Vol. 49, No. 438 (1983),
199.
(5] Craighead, I.A., Dowson, D., et al., I. Mech. Engrs., Paper
II (i) (1980), 47.
[6] Boncompain, R., Frene, J., I. Mech. Engrs., Paper I (iii), 33.
[7] Mitsui, J., Trans. of JSME, Ser. C, Vol. 48, No. 428 (1982),
565.

281
4 Explanations

4.8 Unbalance Response of Rotors in Journal Bearings

4.8.1 Introduction becomes the centre of gravity, which means that the
phase lag becomes about 180°. This phenomenon
The vibration of a rotor supported by journal bear- becomes clearer at (6). At the previous rotating
spee~ it is mainly unbalance vibration that appears,
ings is different from that of one supported by ball
bearings: the rotor becomes unstable in the high- but m modes (7) and (8) the unstable vibration
speed range and its vibration is nonsynchronous mode whose frequency is the same as the natural
and different from the unbalance vibration. This frequency is beginning to be superposed on the un-
section is devoted mainly to synchronous vibration , balance vibration mode. In modes (9) and (10) the
. unstable vibration mode becomes more prominent
I.e. unbalance vibration.
Figure 4.8.1 shows the difference between unbal- and the unbalance vibration mode is eliminated.
ance vibration and unstable vibration. Fig- This unstable vibration is called "oil whip".
ure 4.8.1 a shows the phase angle, which is defined In this section the method of obtaining the unbal-
as the angle between the unbalance location and the ance response is shown for a simple model, and
maximum amplitude. Figure 4.8.1 b shows the un- resonance amplitude, vibration mode and whirl or-

t~:~:~:
balance response curve, unstable vibration curve (1)
and measured vibration curve. Figure 4.8.1 c shows
the frequency for the two vibration modes. Fig-

r"7t"7t"71l
ure 4.8.2 shows vibrational modes for the point
number on the response curve shown in Fig. 4.8.1 b. (2)
A pulse mark on the abscissa means an unbalance
point fixed in the rotor, i.e. the pulse is marked once
per revolution.

n~~~
The vibrational mode (1) shows the case of low (4)
rotating speed and in this case the amplitude is
small and the phase angle¢ is almost zero. In mode
(2) the amplitude becomes large and the deflection

b:::~
of the rotor can not instantly follow the unbalance

r-\C: ~~=:
(6)
force. In mode (3) the rotating speed coincides with
the natural frequency of the rotor system (resonan-
ce frequency), so that the phase angle is 90°. In
mode (4) the phase angle becomes more than 90° (7) (8) A (\ 0 (\ 0 (\
and in mode (5) the rotor deflection cannot follow
the unbalance force and a centre of spin of the rotor

(9)0!~
i'"'"l v:
a
j 9: :_======:?=====~:==============- 10
y-
Fig. 4.8.2. Vibration wave

Unbalance vibration
4
/Measured vibration
8
5 67 •
b
m : Mass of disc
k : Shaft stiffness
l : Span length
o· : Static equilibrium point
0 J : 1ournal centre
WI
rpm 0 • : Center of disc
G : Mass centre
Fig. 4.8.1 a-c. Relations among phase angle, amplitude and fre-
quency Fig. 4.8.3. Model rotor and coordinate system

282
 4.8 Unbalance Response of Rotors in Journal Bearings

bit are theoretically represented. From the view- 4.8.3 Unbalance Response
point of rotor-bearing system design the method of
selecting the bearing type and bearing configura-
In this analysis, the force of gravity on the rotor is
tion so as to reduce the amplitude of the critical
neglected. Then unbalance response is obtained as
speed is discussed.
a particular solution of Eq. (3). The unbalance force
in Eq. (3) is represented as
4.8.2 Equation of Motion
f(t) = {0, 0, eRw 2 coswt, 0, eRw 2 sinwt, Of
For the sake of simplicity, the Jeffcot rotor model
= {0, 0, eRw 2, 0, 0, O}r coswt
shown in Fig. 4.8.3 is considered, and it is assumed
that the rotor is located at the centre of the shaft + {0,0,0,0,eRw 2,0}T sinwt
and that bearing characteristics on both sides are =fc coswt + fs sinwt (4)
equal. The equilibrium point of the journal centre 0'
is selected as the origin of the rotor coordinate thus the particular solution of Eqs. (3) and (4) is
system. The coordinate of mass centre of the rotor
is selected as (x, y) and the journal centre is selected
x = ~ coswt + ~ sinwt
as (x 0 , y 0 ). The equation of motion is represented in = {~ 1 , ~ 2 , ... , ~ 6 f coswt
ref. [2] as + {F. 1 , F. 2 , ... , F. 6 } T sin w t . (5)
mx + k(x- x 0 ) = meRw 2 coswt + mg,
Substituting Eq. (5) into Eq. (3) we get
my+ k(y- y 0 ) = meRw 2 sinwt, (1)
wAFS + BPc =fc,
k(x- x 0 ) = J>,;, k(y- Yo)=~. (6)
- wAPc +BPs= fs·
where m is rotor mass, k shaft stiffness, w rotating
speed of rotor, eR eccentricity of the rotor and P,; This equation is rewritten in matrix form as
and ~ the journal bearing forces which are repre-
sented as
[ B wAJ
- wA B
[Pc] {/c}.
Ps
=
fs
+ kxy Yo+ cxxXo + CxyYo + mg,
P,; = kxxXo
P k (2)
y = yxXo + k yyYo + CyxXo + CyyYo,
0 0

Thus, the coefficients of the particular solution (5)

where kii and C;i are stiffness coefficients and damp- are obtained as

w:]-1{~:}.
ing coefficients of the journal bearing, respectively.
These values are listed in the tables and figures in
this book in the dimensionless form as
[~] = [-~A (7)

K.. = CPkii The coordinates of rotor disc centre due to unbal-

I] w ' ance are represented as

{ x} =
y
{PC2
cos w t
Pcs coswt
+ IS 2 s~n w
+ Pss smwt
t} . (8)
where CP is a radius clearance of the journal bear-
ing, W is the static load of the bearing and w, the The coordinates of the shaft at the bearing are rep-
rotating speed of the rotor. Equation (1) and (2) can resented as
be rewritten in matrix form as
Ai+Bx=f(t) (3) {x0 } = {Pc
3 coswt + f's 3 s~nwt}.
(9)
Yo Pc 6 coswt + f's 6 smwt
where
x1 x 0 0 Cxx 0 0 Cxy In the numerical calculation the bearing coeffi-
x2 X 0 0 Cyx 0 0 Cyy cients must be calculated for each rotating speed,
x3 Xo 1 0 0 0 0 0 because the bearing coefficients are functions of the
{x} = (A)=
x4 y 0 1 0 0 0 0 rotating speed.
Xs y 0 0 0 1 0 0
x6 Yo 0 0 0 0 1 0

0 - k kxx + k 0 0 kxy 4.8.4 Unbalance Vibration Mode

0 0 kyx 0 - k kyy + k
0 kjm - k jm 0 0 0 The whirling orbit of the rotor is obtained by com-
(B)=
-1 0 0 0 0 0 position of the x and y coordinates, which are ob-
0 0 0 0 k jm - k jm tained from Eqs. (8) and (9). The method of compo-
0 0 0 -1 0 0 sition is illustrated in Fig. 4.8.4.

283
4 Explanations

The direction of whirl is:

a) forward precession for r + > r _
b) backward precession for r + < r _
c) linear orbit for r + = r _
y, Im Usually the maximum amplitude A is used as the
amplitude of the unbalance response from the view-
point of shaft design.
Orbit r. ei"'

Rotating direction
4.8.5 Numerical Calculation
x,Re x, Re of Unbalance Response
Fig. 4.8.4. Calculation of elliptic whirling amplitude
In this section unbalance response is calculated for
a simple rotor system to illustrate the effect of bear-
The motion of a disc centre in the horizontal and ing types and some bearing parameters. The data of
vertical directions is represented as the model rotor is shown in Table 4.8.1. Figure 4.8.5
x(t) = Pe 2 cosOJt + f's 2 sinOJt, shows the unbalance response and whirling orbit
(10) for a cylindrical bearing. The axis of the abscissa
y(t) =Pes cosOJt + Pss sinOJt. being a rotating speed nondimensionalized by the
If the motion of the centre of the journal is to be natural frequency of a simply supported rotor OJ 8 ,

represented, ~ 2 and P. 2 are replaced with ~ 3 and and the axis of ordinate, an amplitude nondimen-
Jl. 3 , and ~sand P.s with ~ 6 and }J. 6 • Equation (10) sionalized by a mass eccentricity eR.
is then written as In this case, eigenvalues of the rotor/bearing sys-
tem (A.= o: + ift) are o: 1 =- 0.0965, P1 = 0.707 OJ8 ,
x(t) = JPr!2 + N 2 sin(OJt + qJx), (11) o: 2 = - 0.0514 and P2 = 0.966 OJ so that peak val-
8,

y(t) = JPr!s + Ns sin(OJt + qJy), ues for the response are near the first and second
natural frequencies. As I o: 1 1 is larger than Io: 21, the
where
qJ =tan- 1 -Pc2 Table 4.8.1. Basic numerical model
x Ps2'
Rotor mass m (kgf) 70
The whirling orbit of the rotor supported by jour- Shaft stiffness k (kgfjmm) 1.428
nal bearings is usually elliptical, and it is represent-
Stiffness ratio of shaft-bearing y 0.904
ed by the semimajor axis A, the semiminor axis B,
the direction of the principal axis and the whirling L:D 0.5
direction. A and Bare obtained from Eq. (10) as
r(t) = x(t) + jy(t) U=F)
= Pe 2 cos OJ t + Ib sin OJ t 100
+ j (Pes cosOJt + Pss sinOJt). (12)
Applying the relation e;o = cos 0 + i sinO to Eq. (12),
we get 10

r(t) = eiwt {t (Pc2 + Pss) + j! (Pes - Psz)} ......

.,
.....
+ {t(Psz - Pss) + j ! (Pes+ Ps2)}
e - iwt «:
.g I
=! J(~2 + Jl.s) 2 + (~s- JJ.2) 2 · ei<»t (13) .€
+ i j(~2- Jl.s)2 + (~s + Jl.z)2. e-iwt ~
-<
= r + ei<»t + r _ ei<» t . (14) .1

Equation (13) represents the complex amplitude of

the shaft by using the components of the forward
and backward precession. The A and B of the el- 0 .5 1.5
lipse are obtained as Dimensionless rotating speed w/ ws
A=r++r_,
Fig. 4.8.5. U nbalance response and whirling locus for cylindri-
B=r+ -r_ . cal bea ring

284
 4.8 Unbalance Response of Rotors in Journal Bearings

amplitude of the first peak is smaller than that of

100 the 2nd peak.
Figure 4.8.6 shows the unbalance response for
various kinds of journal bearing. The response
curve of the rotor supported by the cylindrical bear-
ing is shown as a reference response curve by the
unbroken line in each figure to allow comparison of
the response curves. The numbers and data illus-
trated refer to the data numbers of the bearing in
- - (1) Circular bearing
this databook and the bearing parameters. It is seen
............ (6)120' Partial from this figure that as the preload factor becomes
.1 circular bearing large, the amplitude at the critical speed becomes
------ (IU2 lobe mp =2/3 large. The amplitude of the bearing of TP4, LBP,
---(1?)3 lobe mp =2/3 mP = 2/3 is predominantly large, because the com-
----{2U4 lobe mp =2/3 bination of the stiffness of the rotor and of the
0 .5 1.5 bearings is not good. That is, the ratio of the bear-
a Dimensionless rotating speed (J)/ (J)s ing stiffness and shaft stiffness is important for the
unbalance response, and this ratio is different from
that in the other cases.
When the bearing stiffness in the horizontal and
vertical directions is nearly equal, as in LBP-type
100 tilting pad bearings, the unbalance response has
only a single peak.
Figure 4.8.7 shows the influence of the preload
10
factor on the unbalance response for two-lobe bear-
.." ing and tilting-pad bearings of TP 5-LBP and TP 5-
5--- LOP. It is seen from this figure that as the preload
<!)
't:l factor increases the response curve near the critical
.€ 1 speed increases, because the damping coefficient de-
0.
~ creases when the preload factor increases. In prac-
- - (1} Circular bearing tice, if the preload factor is small, instability occurs
.1 ............ (2~P4,LBP,mp=O at low speeds. Thus, when the rotor is operated at
----- (26)TP 4, LBP, mp=2/3 a high rotating speed the high preload factor taken
- - - CJ?)TP 4, LOP, mp= O from Fig. 4.8.1 0 should be used.
---- Clm'P 4, LOP, mp=2/3 Figure 4.8.8 shows the unbalance response for
0 .5 1.5 the stiffness ratio parameter y (y = Ff C k, where
b Dimensionless rotating speed (J)/ (J)s
F = bearing reaction force, C = bearing clearance,
k = shaft stiffness under the condition of simple
support). S0 in the figure represents S0 = S wsfw,
where ws is the critical speed under the condition of
a simple support and w the rotating speed of the
shaft. If y is large, the shaft is more flexible than that
100
of the journal bearing, so the deflection at the bear-
ing is small and energy dissipation at the bearing is
small. Therefore the damping of the whole system
..
-.,1o becomes small. On the other hand, if y is small, the
5--- deflection at the bearing· becomes large and the
<!)
't:l
energy dissipation at the bearing becomes large.
.a Therefore the damping of the whole system be-
'a 1
E comes large.
<
- - (1} Circular bearing
.1 ............ (40)TP5,LBP,mp= O
------ (47)T P 5, LBP, mp= 2/3 4.8.6 Damping Effect of Bearing
- - - (48)TP5,LOP,mp=O
- · - - (SO)TP 5, LOP,m,=2/3
The damping effect of the rotor/bearing system de-
0 .5 1 1.5
pends on (1) the stiffness ratio of the bearing and the
c Dimensionless rotating speed (J)/ (J)s
shaft and (2) the bearing itself. For case (1), if the
Fig. 4.8.6. Unbalance response for different types of bearing bearing is designed as the damping coefficients of

285
4 Explanations

100 100

~10
.. 10
--
-.::

---- r=5.0
.1 - - (1) Circular bearing =2.0
............ (IO)mp=l/2 =r.o
------ (10mP=2/3 =o.5
- - - (J2)mp=3/4 s; =1.62 =0.2
0 .5 1.5 0 .5 1.5
a Dimensionless rotating speed w/ ws Dimensionless rotating speed w/ w,
Two-lobe bearing Fig. 4.8.8. Unbalance response as a function of y

100 the bearing are designed too large, deflection at the

bearing becomes small compared with the shaft de-
flection. Therefore, the energy dissipated in the
bearing becomes small, because the dissipative en-
10 ergy is represented by the multiplication of the
.... damping force and the time derivative of amplitude
--
.:5
(velocity) and when the velocity is low the dissipa-
tive energy becomes small. Therefore, the damping
~ 1
.€ coefficient of the bearing should be chosen with due
c. consideration for the shaft stiffness. Figure 4.8.9
E - - (1) Circular bearing
-< ············ (48)mp=O
shows the logarithmic decrement versus stiffness
.1 ratio parameter y. It is seen from the figure that if
----- (49)mp 112 =
- - - (50)mp = 2/3 y becomes large (that is, the bearing restrictive force
- - - - (50mp =3/4 becomes large) the damping coefficient of the whole
system becomes small.
0 .5 1.5
c Dimensionless rotating speed wlws For case (2), the damping coefficient of the whole
Tilting pad bearing (TP 5, LOP) system is changed by the size and the type of the
bearing. That is, the stiffness coefficients and damp-
ing coefficients of the bearing are changed by the
viscosity of oil, the bearing clearance, bearing
100
length and bearing radius, because the equilibrium
point of the journal changes with these factors. Al-
so, these change with the bearing type, that is, ac-
10

E =0.6
<I)
-o s; = o.1535
.c..e 1
E
-<
- - (1) Circular beanng
2
.1 ············ (40)mp =0
---- - (40mp = 1/2
- - - (42)mp =2/3
---- (43)mp=3/ 4

0 .5 1.5
b Dimensionless rotating speed w/ ws
Tilting pad bearing (TP 5, LBPJ r : Relative flexibility of shaft
Fig. 4.8.7. Influence of preload factor on unbalance response Fig. 4.8.9. Logarithmic decrement J

286
 4.8 Unbalance Response of Rotors in Journal Bearings

cording as whether circular bearings, two-lobe 4.8.7 Application to a Complicated System

bearings or tilting pad bearings are involved, for
example. Logarithmic decrements of a rotor system
The unbalance response of the simple rotor/bearing
supported by two-lobe bearings and three-lobe
system has already been illustrated, but in practice
bearings are shown in Fig. 4.8.10 as an example: for
rotating machinery is not so simple as this. To ob-
the two-lobe bearing with mP = 1/2, the mode-1
tain the unbalance response of the shaft with an
vibration becomes unstable at wfws = 1.15. It is
arbitrary cross section or supported by multi-bear-
also seen from this figure that if the preload factor
ings, the transfer matrix method or the finite ele-
is properly selected instability dose not occur until
ment method must be applied. These methods are
a high rotating speed is reached. These results are
not described, in this section, because a great deal
qualitatively useful for the three-lobe bearing. The
has already been published on them.
logarithmic decrement for three-lobe bearings is
comparatively higher than that for two-lobe bear-
mgs.
4.8.8 Summary

The unbalance response of a rotor supported by

1.5 Mode journal bearings is characterized as follows:
Mode 2
1) A simple rotor supported by journal bearings
'<> has two critical speeds, because journal bearing
<:: stiffness in the horizontal and vertical directions
a
(!.)

is different. Since the bearing stiffness in the hor-

b 0.5
(!.)
"0 izontal direction is less than that in the vertical
u
'§ direction, the critical speed in the horizontal di-
] 0 0.1 0. 5 20 rection is lower than that in the vertical direc-
tion. The whirling orbits of the unbalance re-
3-o. 5 r : 0. 5 sponse near the critical speeds are ellipses.
s~ : o.3 2) The unbalance response differs widely with bear-
~ : Dimensionless rotating speed ing type, which should therefore be selected care-
- 1 w,
fully.
a Logarithmic decrement 8 of two-lobe bearing (L: D= l) 3) The unbalance response depends heavily on the
preload factor. If the preload factor is small the
unbalance response at the critical speed is usual-
ly small.
4) The unbalance response depends heavily on the
Y. M Mode 1
stiffness ratio parameter of the bearing to the
Mode 2
1.5 Ys shaft, y. If y is large, the resonance amplitude
becomes large and the logarithmic decrement
mp=Yz becomes small.
'<>
, .... ---~ : , ...... 5) As the rotating speed increases the damping ef-
E . . . . . .u ',,
a -__ . . -4---:::::-~, fect becomes small and the rotor system be-
(!.)

...
- __Ys.,.---.....~~ ,_
(!.)

--
u
(!.)
0.5
'~-
comes unstable. The threshold speed for insta-
"0
u bility is large as the preload factor become large.
'§
0
:E 0.1 0.5 w
m, (l)
s
.3
-0.5 r : 0.5 References for Sect. 4.8
s~ : o.3
- 1 : : Dimensionless rotating speed [1] Gasch, R . and PI t; ncr, H ., Rotordynamik, Springer-Ver-
s
lag, 1975, Cha p. 7.
b Logarithmic decrement oof three-lobe bearing ( L: D= 1) [2] lwatsubo, T., Tra ns. of JSME (C), 46-405, May 1980, p . 493.
[3] Kuro hashi, M . e t a!, Tra ns. of JSME (C), 47-422, Oct. 1981,
Fig. 4.8.10. Rotating speed vs. logarithmic decrement p. 1277.

287
4 Explanations

4.9 Stability of Rotors Supported by Plain Bearings

4.9.1 Introduction The oil whip phenomenon is basically character-
ized as follows:
Plain bearings are superior to rolling element bear- 1) It occurs when the shaft speed w reaches about
ings in high-speed operation. They have high twice the first critical speed (wk) and does not
damping ability, high load-carrying capacity and disappear even if the speed increases further,
low friction. In addition, silent operation and long quite unlike unbalance vibration. The shaft vi-
life can be attained. Therefore, plain bearings of brates violently, sometimes resulting in bearing
various types are widely used in high-speed rotating failure and damage of the rotor itself(Fig. 4.9.1).
machinery, such as turbines, pumps, and compres- 2) Its frequency and vibrational mode correspond
sors. The advantages are attributed to the mechan- to those at the first critical speed and do not vary
ical characteristics of the thin oil film formed in the much with changing speed of the rotating shaft
small clearance between journal and bearing sur- (Fig. 4.9.1 ).
face. Some of the complicated characteristics have 3) The centre of the rotor whirls in the direction of
already been revealed experimentally and theoreti- shaft rotation.
cally. 4) The smaller the load applied, the more readily it
In many cases, the spring and damping charac- occurs.
teristics of the lubricating oil film have a most sig-
In addition, each of the following phenomena
nificant effect on the vibration of a shaft rotating in
can be observed, depending on the combination of
journal bearings. For example, if a rotor and bear-
rotor and bearing.
ings are properly designed, the amplitude of unbal-
ance vibration of the rotor can be well suppressed 5) It does not occur before the increasing shaft
even at critical speeds. Inappropriate design, in con- speed w considerably exceeds 2wk , but once it
trast, may not only increase the vibration ampli- occurs it does not disappear until the decreasing
tude but bring about oil whirl or oil whip, that is, speed reaches nearly 2wk. This phenomenon
a self-excited lateral vibration of rotating shaft is often called the inertia effect of oil whip
caused by the oil film in journal bearings. Careful (Fig. 4.9.2).
attention should be paid to the prevention of oil 6) A vibration of small amplitude with the frequen-
whip in the design of high-speed turbomachines cy of wj2 occurs long before the increasing speed
equipped with fluid film bearings. The recent pro- reaches 2wk. This vibration is called oil whirl
gress in lubrication theories and rotor dynamics and is differentiated from large-amplitude oil
enables design engineers to predict the occurrence whip. The frequency of oil whirl increases with
of dangerous vibration and take effective steps to shaft speed and the vibration ultimately be-
avoid it.
Besides oil whip, high-speed machines sometimes
exhibit other self-excited vibrations, such as steam Oil whip
whirl, hysteresis whirl and subsynchronous whirl.
Journal bearings should be properly designed to
avoid these vibrations also when their occurrence is
predicted. This problem will be briefly discussed
later in Sect. 4.9.7 and in more detail in Sects.
4.10-12.
(J)K 2 (J)K

Rotational speed (J)

Fig. 4.9.1. Occurrence of oil whip

4.9.2 Oil Whip Phenomenon

The name "oil whip" was first coined by Newkirk

and Taylor [1], who experienced violent shaft vibra-
tion in a test rig about 60 years ago. They conclud- "'
-o
ed that the vibration was caused by the action of the .B
0.
bearing oil film because it subsided when the oil E
-<
supply to the bearings was cut off by accident. Since
then, oil whip has been studied extensively and was
soon confirmed to be a self-excited vibration with (J) K 2 (J) K

quite different features from that of any kind of Rotational speed (J)

forced vibration, such as unbalance vibration. Fig. 4.9.2. Inertia effect of oil whip

288
 4.9 Stability of Rotors Supported by Plain Bearings

Frequency OJK

~ Oil whip

2 (J)K
(J)K
Fig. 4.9.3.
Rotational speed OJ Oil whirl and oil whip

comes large-amplitude oil whip at w = 2wk

(Fig. 4.9.3). Fig. 4.9.5. Coordinates
Familiarity with the above descriptions allows
easy identification of oil whip or oil whirl in actual
As shown in Fig. 4.9.5, xR and YR denote each of the
machines. A similar self-excited vibration is some-
vertical and horizontal displacements of the rotor
times observed in high-speed, high-performance
centre from its equilibrium, respectively, and x and
turbomachinery, and it is believed to result from the
y, those of the journal centre. fx and fY denote the
action of the fluid film in the sealing clearance. This
x and y components of the oil film force in one
vibration is now under thorough investigation and
journal bearing, respectively, and are obtained by
it is expected that it will soon be identified.
integrating the pressure developed in the bearing
clearance, which is determined from the Reynolds
4.9.3 Theory of Oil Whip equation. Note that 2fx excludes the component
which is equal to gravity force, mg.
Hori solved the Reynolds equation, assuming in-
Many research workers have attempted to formu- finitely long bearing and n film, and obtained ana-
late theories to explain the various phenomena and lytical expressions for fx and fY. The oil film forces
mechanisms of oil whip, and Yukio Hori [2] is the can be linearized when x and y are assumed to be
first to have presented a theory which agrees with very small. Then fx and fY can be expressed as
experimental observations in nearly every point. follows:
His theory is outlined as follows:
A symmetrical rotor is supported in identical f x = kxxX + kxyY + Cx xX + CxyY ,
(2)
journal bearings at both its ends (Jeffcot rotor). The fy = kyxx + kyyY + cyxx + cYYy,
rotor is assumed to consist of a massless, elastic
shaft (bending stiffness k) and a single, centred mass where kxx• kxy• ... and cx x• cx y• . .. are the spring
rn (Fig. 4.9.4). When the load and the shaft speed do and damping coefficients of the oil film, which are
not fluctuate with time, the journal centre main- defined uniquely by the Sommerfeld number, that
tains a static equilibrium position in the bearing is, by the eccentricity ratio of the journal centre at
clearance, defined uniquely by the Sommerfeld its equilibrium in the bearing clearance. Conse-
number, which represents the operating condition. quently, these coefficients are determined for a giv-
If no mass unbalance is assumed, the rotor centre en operating condition, such as bearing load, lubri-
also keeps to its own static equilibrium point. In a cant viscosity, and shaft speed.
real rotating system there always exist some distur- Now Eq. (1) could be solved for xR and YR. How-
bances which perturb the journal and the rotor at ever, the solution itself is less important than
their equilibrium. Then the equation of motion is whether or not xR and YR increase with time. If the
given as follows: perturbation is found to grow with time it means
that the rotor-bearing system is unstable for the
rnxR + k(xR- x) = 0, operating condition and therefore the vibration
rnyR + k(yR- y) = 0, amplitude increases owing to self-excitation.
(1) Let xR = XRo exp(st), YR = YRo exp(st), where
k(xR- x) = 2fx,
xRo and YRo are constant and t represents time.
k(yR- y) = 2jy- Equation (1) can be expressed without x and y by
m
using the first two expressions, and the above equa-
tions are put into Eq. (1). Then a set of simulta-
neous, linear equations of xRo and YRo are ob-
tained. The value of the determinant of the
coefficients must be zero for nontrivial solutions of
xRo and YRo· Then the characteristic equation for
Fig. 4.9.4. Rotor model the system is obtained in the form of an algebraic

289
4 Explanations

polynomial of s : A3 = CxxCyy- CxyCyx

C 0 s6 + C 1 ss + C 2s4 + C 3 S 3 + C 4 s 2
A4 = Kxx + K YY
As= Cxx + CYY
+Css+C 6 =0 (3) Kii = kiiCP jW: dimensionless spring coefficient
The coefficients, C 0 , C 1 , C 2 ... are functions of Cii = ciiCPw jW: dimensionless damping coeffi-
kxx> kxy> ... and cxx> CXY' .. . ,etc. The complex eigen- cient.
values of the system are obtained by solving Eq. (3) Equation (6) represents the situation in which a
as follows: small perturbation increases self-excitedly over
s; = a; ± j b; (i = 1 "' 3) (4) time for the rotational shaft speed over v0 which is
determined by K;j, Cii and ex. The threshold speed
where a; and b; are real numbers and} is the unit of falls with increasing a (more flexible shaft).
an imaginary number. Then Routh-Hurwitz criteria can only give the stability
limit of vibration that is of such small amplitude
xR = XRo exp(a;t) exp(±jb;t),
that it does not prevent the assumption of a linear
YR = YRo exp(a;t) exp(±}b;t) . system. Therefore, the linear stability theory cannot
The term exp ( ± j b; t) represents a harmonic oscil- explain why, above the threshold speed, the shaft
lation with the circular frequency b; . On the other actually exhibits two unstable vibrations of small
hand, exp (a;t) determines whether the vibration amplitude (oil whirl) and large amplitude (oil whip),
amplitude increases (a; > 0) or decreases (a; < 0) depending on geometrical and operating condi-
with time. If a; = 0 the rotating system lies on the tions. In the case of oil whip, the journal centre
boundary between stable and unstable regions of whirls with a large amplitude in the bearing clear-
operation. ance and we can no longer use f x and h expressed
Summing up, stable running of a rotor free from by Eq. (2), which was derived under the assumption
self-excited vibration requires every a; to be nega- of small perturbation. Hori obtained the expression
tive, and vice versa. Application of the Routh-Hur- of the oil film force for this case, assuming a steady
witz criteria allows us to know whether or not state circular orbit of the journal centre in the clear-
Eq. (3) has a complex root with positive real part, ance, and then derived the stability limit for oil
without solving it. Let the coefficients of Eq. (3) be whip as follows:
arranged in a 6 x 6 matrix form [Eq. (5)]. Note that (7)
C 1 , C 2 , ... , C 6 are arranged on the diagonal in that
order. Routh-Hurwitz criteria show that every prin- which indicates that large-amplitude oil whip can-
cipal minor determinant of the matrix is positive if not exist at less than twice the first critical speed.
each a; is negative, and vice versa. Hori calculated
the minor determinants and obtained Eq. (6) for the
dimensionless threshold speed of self-excited vibra-
tion.
4.9.4 Occurrence and Growth of Oil Whip

~c3 Cs 0 0 0 Figure 4.9.6 shows the stability limit boundaries

represented by Eqs. (6) and (7). The abscissa de-
Co C2 c4 c6 0 0 notes the square of the reciprocal of dimensionless
shaft speed, v = w j(g /Cp) 1 12 , and the ordinate, the
0 cl c3 Cs 0 0 eccentricity ratio of the journal, e. The solid line
(5)
0 Co C2 c4 c6 0 represents Eq. (6), and the rotating system is stable
below this line, that is, small perturbation definitely
0 0 cl c3 Cs 0 decreases with time and disappears eventually. On

0 0 Co C2 c4 c6 llv
0 10 15
vc = (A 1 A 3 AV{(Ai + A 2 A;- A 1 A 4 As) WK
· (As+ cxA 1 ))) 112 (6) \ 12WK

where
\
1oil
\whip
''A
j,' '
Oil whirl

.5 \ , '- Stable
Vc
wcf(g /Cp) 112
=
(g: acceleration o~ gravity, CP: mean radial
\ I ....... A, L'

\ ' .......... • .._ __ a

I
B2
clearance of beanng)
a= W j(k Cp) (W = mg)
Al = KxxCyy- KxyCyx- KyxCxy + KyyCxx
B, ..... _
---------b
1.0 ...__ _ _ _ _ _ _ _ _ _;:___ _

A2 = KxxK yy - KxyKyx Fig. 4.9.6. Stability chart

290
 4.9 Stability of Rotors Supported by Plain Bearings

the other hand, small perturbations increase self-ex-

(a)
citedly in the unstable running region above the
line. The dot-dash lines represent the critical speed
of the system wk (arbitrarily given here) and also
2wk. Large-amplitude oil whip is unstable in the
region represented by the left-hand side of the line
2wk.
The eccentricity ratio of journal e, which was 1.0 a WK

at zero speed, decreases with increasing shaft speed

v. Therefore, the state point (v, e; the point 1fv2 , e) (b)
moves along the path shown by the broken line,
from lower right to upper left. The path, that is, the
relationship between v and e, is determined uni-
quely by the following dimensionless bearing mod-
ulus A
(8) b

Fig. 4.9.7a, b. Oil whip and oil whirl

where f1 is the viscosity of lubricant, R is the radius
of the journal and Pm is the mean bearing pressure.
The value of A for path (a) is larger than that for
path (b), and the eccentricity ratio e of the journal
centre is therefore smaller than that for the path (b) 4.9.5 Stability Criterion for Generalized
at the same shaft speed.
Consider the case along path (a) first. With in-
Rotor-Bearing System
creasing shaft speed the rotor becomes unstable at
the point A 1 where path (a) crosses the stability The stability limit of Eq. (6) is sometimes inapplica-
limit curve, and self-excited vibration occurs. The ble to actual machines, because actual rotor-bear-
vibration amplitude does not increase so much and ing systems cannot always be represented by a sin-
remains small (oil whirl) until the shaft speed reach- gle, centred mass, a uniform massless shaft and two
es 2wk (point A 2 ). The vibration frequency of oil identical bearings symmetrically arranged. The sta-
whirl is w/2 and it increases with w. At A 2 , the bility of more general systems can be determined
frequency reaches wk and large-amplitude oil whip numerically by means of transfer matrix method [3].
starts. Beyond that the vibration remains violent An actual rotor is assumed to consist of concen-
with a constant frequency of wk. (See Fig. 4.9.7(a).) trated masses, discs with the moment of inertia, and
On the other hand, in the case along path (b) the massless elastic beams. Each element is numbered
rotor enters the unstable running region of oil whip continuously from, say, number one for the left end
at B 1 . However, the rotor is still stable in the sense of the rotor. The state vector Z; for the element i is
of linear stability, and therefore self-excited vibra- a column vector, and its eight components are the
tion does not occur at B 1 . When the speed increases shearing forces Vx,y the bending moments Mx,y• the
further, to point B 2 , the condition for linear insta- slopes ex,y and the deflections X, y. The state vector
bility becomes satisfied. Oil whip starts instantly components at the element n + 1 can be expressed
because the stability limit of Eq. (7) has already in terms of those at the element n and in matrix
been cleared at B 1 . Once it occurs, it does not dis- notation:
appear until the speed becomes lower than 2wk at
(9)
B 1 . [See Fig. 4.9.7 (b).] This explains the inertia ef-
fect of oil whip. Some violentshock sometimes trig- where T, is the transfer matrix for the element n. Let
gers oil whip when it is given to the shaft between x = x 0 exp(st), etc. and then Eq. (9) is rewritten as
B 1 and B 2 during the increasing speed stage. follows:

1, 0, 0, 0, 0, 0, - (ms2 + Zxx), - Zxy

0, 1, 0, 0, 0, 0, - Zyx> - (ms 2 + zyy)
l, 0, 1, 0, Irs 2 fpWS, - l(ms 2 + zxx), - lzxy
0, l, 0, 1, - fpWS, Irsz, - fzYX' - l(ms 2 + zyy)
(10)
b, 0, a, 0, 1 + alrs 2, afpWS, - b(ms 2 + Zxx), - bzxy
0, b, 0, a, - a/pWS, 1 + alrs 2, - bzyx> - b(ms 2 + zyy)
d, 0, b, 0, l + blrs 2 , bfpWS, 1 - d(ms 2 + Zxx), - dzx y X
0, d, 0, b, - bfpWS, l + blrs, - dzYX' 1 - d(ms 2 + zyy)

291
4 Explanations

where To avoid obtaining the roots already found

again, Eq. (13) should be modified as follows:
a= 1/(EI) , b = F/(2EI), d = P/(6EI),
s(N +1) = s(N>- f(s(N>) /(f'(s(N>)- f(s(N>) /
E: Young's modulus of shaft material
J
I: geometrical moment of inertia of the beam · L: (s(N) - sJ) (16)
l: length of the beam i= 1
m: mass where J is the number of the roots already found.
I P: axial moment of inertia of the disc
Id: equatorial moment of inertia of the disc
zii = ciis + kii (i, j = x, y).
T, in Eq. (1 0) corresponds to the element which con-
sists of as many components as possible, and there- 4.9.6 Stabilization Measures
fore some of the components should be zero for a
particular element. Equation (6) and Fig. 4.9.6 show that the threshold
Let the total number of elements be r. Equa- speed of oil whirl increases with the decrease in r:x or
tion (9) is applied to each of the adjacent elements with the increase in e. On the other hand, Eq. (7)
of the rotor in turn, and eventually zr + 1' the state shows that the threshold speed of oil whip increases
vector for the right end can be expressed as follows: with the increase of wk . Higher wk corresponds to
(11) lower r:x which, as mentioned above, leads to higher
threshold speed of oil whirl.
where T= T, · T.- 1 · ... · T2 · T1 . The following countermeasures are often taken
The shearing forces Vx ,y and the bending mo- to eliminate oil whip or oil whirl in turbomachines
ments Mx ,y are zero at both ends when the rotor is supported in circular journal bearings with fixed
freely supported as usual. Substituting this load and speed.
boundary condition into Eq. (11), we get a set of
1) Increase of first critical speed wk by increasing
simultaneOUS, linear equations of Ox, 0y, X andy at
shaft diameter or shortening shaft length.
both ends. The determinant of the following sub-
2) Increasing journal eccentricity ratio e at equilib-
matrix T, components of which, taken from the
rium by decreasing bearing modulus A. In partic-
overall transfer matrix T, must be zero for nontrivial
ular,
solutions, and the characteristic equation f(s) is
(i) Shortening bearing width L,
obtained as follows:
(ii) Increasing bearing clearance CP,
t1s t16
(iii) Decreasing lubricant viscosity f.l-·
t1 7 t18
tzs t26 t27 tzs =f(s) = 0. The bearing length can be easily shortened by mak-
ITI= t35 t36 t37 t38
(12)
ing a circumferential oil groove at the bearing mid-
t45 t46 t47 t48 width, and this tactic is commonly adopted as "first
aid" for urgent suppression of oil whip in actual
Equation (12) is numerically solved for s for exam- machines, with a high degree of success. However,
ple by means of the Newton-Raphson method in precautions must be taken to see that an excessive
Eq. (13), and the sign of the real part of each root increase in e does not occur, which sometimes in-
obtained is checked directly creases temperature rise dangerously in the oil film,
s(N+1) = s(N>- f[s(N>]/f'[s(N>] (13) leading to bearing failure.
If a bearing replacement is available, the follow-
where N is the number of repetitions. ing measures will be often effective.
The derivative f'(s) can be calculated as follows.
I;' is a matrix whose components are those of I; 3) Installation of multilobe bearings (noncircular-
differentiated with respect to s. T', the derivative of bore bearings).
the global transfer matrix T, is obtained in Eq. (14) 4) Installation of tilting pad bearings.
Figures 4.9.8(1) - (7) are the stability charts for
T'= L (T, · T.- 1 · ... ·I;'· ... · T2 • T1 ). (14)
various bearing types. The stability limits are calcu-
i= 1
lated from Eq. (6) with oil film coefficients for each
Then bearing taken from this databook. Generally, mul-
4
tilobe bearings are more stable than circular bear-
f'(s) = L: Itl (15)
ings. A tilting-pad bearing is always stable, because
k=1
it has no cross-coupling terms of oil film coeffi-
where the elements of :1;, are the same as those ofT, cients, unless the shaft speed is so high as to cause
except for the k-th column, which is the same as the pad flutter. Tilting pad bearings are commonly
corresponding one for T'. used for high speed, lightweight rotors.

292
 4.9 Stability of Rotors Supported by Plain Bearings

(1) Circular bearing (Z) Two-lobe bearing

Oil groove 10' X 2 Oil groove 10- X 2
L/ D=O.S LID= O. S
mp= O. 5

\J:;::/.
6

5
"'
;.

1/ 4

~1
0 0
.02 . 05 .1 .2 .5 .02 . 05 .1 .2 .5 2

A A

10

9
a= 0
12
(4) T hree-lobe bearing

v
(3) Two-lobe bearing
Oil groove 10' x 2 Oil groove 10' x 3
8 L!D=O . S
LID=0.5 ~ 1/16 10
1/lp =0.5
mp=0.75 ~
6 8

1/ 16

~1/4
;. 6

4 1/ 4

2 ~1 2

o~~~~~~~~~~~~~~
. 01 . 02 . OS .1 . 2 .5 1 2 5 10
A

14 (6) Four-lobe bearing

Ja=O
(5) Three-lobe bearing
Oil groove 10' x 3 Oil groove 10' x 4
12 L!D=O.S L!D=O . S
mp=O. 75 mp= O.S

10" 10

8 ~1/16 8

6 6

4 ~1/4 1/4

OL-~~~~~~--~~~~~--~
.01 .02 .05 .1 .2 .5
,\
..\
Fig. 4.9.8. Sta bility charts (1-6)

293
4 Explanations

rotating system. Therefore, this vibration is

a= 0 (7) Four-lobe bearing sometimes called subsynchronous whirl.
12 Oil groove 10' x 4
LID=0.5 This vibration is presumably caused by fluid
10
mp=0.75 forces acting on impellers and blades, for example.
Such forces, sometimes called destabilizing forces,
are represented by cross-coupling spring forces as
~""
8
follows:
"'
6 Destabilizing force f=- (k~x k~Y) (~) (17)

4 ~ 1/4
where kxy= - kyx
= Q is often assumed. A consid-
..__._ erable part off is still unknown and further inves-
2 • • • •I tigation is needed .
The stability of a rotating system which has such
0 a destabilizing force can be checked by means of the
.005 .01 .2 .05 .2
above-mentioned eigenvalue analysis of the equa-
A
tion of motion derived with due consideration of
Fig. 4.9.8. Stability chart (7) the force. The sign of the real part a; of each eigen-
value s; shows whether or not the system is stable.
Furthermore, the stability margin can be evaluated
by calculating the damping ratio (; or logarithmic
4.9.7 Flow-Induced Vibration decrement <5; as follows:
(; = - a;/(a? + b?) 112 , (18)
Oil whip or oil whirl is caused by the action of the <5; = - 2na;/b;. (19)
oil film in a journal bearing. In addition, modern
high-performance turbomachines often experience As it is usually difficult to estimate the magnitude
another type of self-excited vibration of the rotor, of the destabilizing force fin actual machines, engi-
possibly due to the action of fluid flow in the ma- neers are advised to design machines so as to keep
chine. (; or <5; positive and well above zero at the rated
While the operating pressure of fluid in the ma- load and speed.
chine has increased rapidly in recent times, rotors The following measures are recommended for
have become more lightweight and the operating suppression of the subsynchronous whirl [4, 5].
speed higher. Therefore, the ratio of power to rotor 1. Raising first critical speed by increasing shaft
mass is increasing, and such rotors readily become diameter or decreasing bearing span.
subject to various vibration problems. As will be 2. Enhancing the asymmetry of rigidity of both oil
seen later in Sects. 4.10- 12, steam turbines, pumps film and bearing support in the horizontal and
and compressors have different causes of such self- vertical directions or increasing oil film damp-
excited vibrations. The most effective countermea- mg.
sure against the vibrations in all types of machinery 3. Increasing the system damping by adding
is to improve the damping capability of the rotating squeeze film dampers.
system by designing better journal bearings.
According to recent reports [4, 5], the self-excited
vibration experienced in high-speed turbocompres-
sors are characterized as follows:
1) It depends on load. When load is low it does not
occur at all or the vibration amplitude is negligi-
References for Sect. 4.9
ble. As load increases to a critical point the vi-
bration suddenly starts or the vibration ampli- (1] B.L. Newkirk and H.D. Taylor, General Electric Review,
Vol. 28, No. 8 (Oct. 1925), pp. 559-568.
tude suddenly increases to a violent level. A (2] Y. Hori, Trans ASME, Ser. E, Vol. 26, No.2 (June 1959),
further increase of the load would endanger the pp. 189-198.
machine. [3] E.C. Peste! and F.A. Leckie, Matrix Method in Elastome-
2) The vibration often occurs when the shaft speed chanics, McGraw-Hill, 1963.
[4] J.C. Wachel, NASA CP-2250 (1982), pp. 1- 19.
is higher than the first critical speed and the [5] M. Kurohashi eta!., Proceedings of 3rd Int'l Conf. on Vi-
frequency of vibration is lower than the rotating brations in Rotating Machinery, I Mech E, (Oct. 1984),
speed and near the first natural frequency of the pp. 51-58.

294
 4.10 Application to Steam Turbine Generator

4.10 Application to Steam Turbine Generator

4.10.1 Introduction
3000
_i:r-
J.4 poles
-t &7
2 poles,'/ / I
2000 ~I
A significant requirement of a large steam turbine
generator is that it should remain reliable for a long 1000
2 poles r-iV
period to ensure the public electric power supply.
For safe operation of the generator unit over a long
~VJIV lj;V
period the rotor vibration is important. Recently, ~ ~v
the problem of vibration has become more impor- r~v~v 0 3000rpm EL

tant, since turbine generators are larger and their ~ /' C. 3000rpm PAD
)( 3600rpm EL
output higher, in addition to which frequent start
X
-
0
100
0 3600rpm PAD
up and shutdown are necessary in accordance with • 1500rpm EL
the fluctuation in power demand. In addition, im- ~
mediate regulation and stabilization of rotor vibra- X
JV x --EL (2 poles)
;>;: ---·PAD(2 poles)
tion levels is required to shorten the test period, ---EL (4 poles)
which implies high fuel costs. Although there are 10 I I I I I I I I
various kinds of reliable analysis and evaluation 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
methods for rotor vibration, it is necessary to un- Journal diameter in
derstand the rotor supporting conditions and the
dynamic characteristics of the bearing before effec- Fig. 4.10.2. Relationship between unit output and bearing size
tive utilization of these methods is possible. In this
section the application of the dynamic characteris- Table 4.10.1 Main geometry of bearings
tics of bearings in large steam turbine generators Type
will be introduced, with examples of unbalance vi- Item
Pad bearing Elliptical bearing
bration and stability in the rotor-bearing system.
LjD 0.5- 1.0 0.5- 1.0
CjD 1.3/1000- 2/1000 0.9/1000- 1.5/1000
m o- o.s 0.5
4.10.2 Bearing Types and Operating m, preload; C, diametral clearance; D, diameter; L, length
Conditions
generator rotors. Furthermore, the bearing size has
been inevitably increased in proportion to the in-
a) Bearing Types crement in output, as shown in Fig. 4.10.2. The
A large steam turbine generator consists of a num- main dimensions and an outline structure of typical
ber of rotors, as shown in Fig. 4.10.1. The bearing bearings are shown in Table 4.1 0.1 and Fig. 4.1 0.3.
types for this multi-span system are selected with
due account taken of the stability of each rotor and b) Operating Conditions
the load capacity of each bearing. A pad bearing is
generally applied to a high/intermediate pressure The position of a journal centre expressed by an
rotor and an elliptical bearing, to low-pressure and eccentricity ratio and an attitude angle has a pro-

P, PAD: Pad bearing

EL: Elliptical bearing
GEL: Grooved elliptical
bearing

700MW G

Fig. 4.10.1. Type of turbine generator and bearings

295
 4 Explanations

nounced influence on the dynamic characteristics of (1) Between pad bearing (2) Load on pad bearing
a bearing, but the position of this centre changes
naturally with the operating conditions, which vary
during the process from start up to attainment of
the rated output of the generator. For instance, the
alignment of bearings varies significantly with the
operating conditions, causing variations in bearing
load, as shown in Fig. 4.10.4. Therefore, each bear-
ing level must be set with prior consideration of the
variations in alignment. Also the rotor-bearing sys-
tem must be designed in such a way as to prevent Oil outlet Oil inlet
Oil inlet
too much variation in a bearing load. However,
some variations will generally take place, amount- <D Carrier ring CD Outer ring

ing to about 20%-40% of the planned load. Fur-

Q) Pad CD Pad
Q) Pivot Q) Locking pin
thermore, the Reynolds number of the bearing ex-
ceeds the critical value of turbulent flow as the Fig. 4.10.3. Sectional view of tilting pad bearing
bearing size increases, and the lubricant flow in the
bearing transfers from the laminar to the turbulent X !03
state, causing some change in the bearing charac-
teristics. 6
The Reynolds number for each bearing size is
shown in Fig. 4.10.5 for a constant value of lubri-
cant viscosity. The state of turbulent lubrication
occurs at a rated speed, as in the case of a bearing 4

ranging in size from 15 in. to 16 in. (381 ,...., 406.4 mm)

for a two-pole machine.
In addition, the position of the journal centre
changes in response to the mean bearing pressure 2
and the rotational speed as shown in Fig. 4.10.6a, b.
Since the position of the journal centre, which
affects the bearing characteristics, can vary with
operating conditions, it is very important to under- 10 20 30
Journal diameter in
Fig. 4.10.5. Bearing size vs. Reynolds number

0.8
Relative level changes ( # 4- # 5 )
0. 7

0. 6

"i 0.5

G) 0.4
>
~

"'>
·~
0.3

G) 0.2
~
0.1

"'
E al:;::
o::;:
::>""2- •-L
g=jg_~ 500
>-cn- 440MW
760 4000 400 730mmHg

600 3000 300

400 2000 200

200 1000 100

6 8 10 12
(hr) Fig. 4.10.4. Alignment changes during operation

296
 4.10 Application to Steam Turbine Generator

Eccentricity < recently and the residual unbalance of a rotor has

o,. . .___,o.,z___o,._4___,~--~--~Lo been minimized.
90
However, many balancing shots may sometimes
be required on the site if the rotor shows poor
vibration characteristics, in which case it is likely to
respond to even a small unbalance. Therefore, the
most important requirement for elimination of any
problems caused by vibration is to calculate the
n=3600rpm unbalance response correctly and understand the
--l!r--EL vibration characteristics of the rotor-bearing sys-
-o-GEL
tem properly; namely, the critical speed, the unbal-
ance response and the vibration mode. If the rotor
~PAD
system has been correctly designed on the basis of
a
the calculated unbalance response in the design
stage, the rotor vibration problem can be treated
correctly and the balancing shot number can be
reduced.

a) Calculation Model
The rotor-bearing system of a large steam turbine
generator has a complicated configuration, consist-
ing of many rotors, discs and blades, operating as a
rotating system. The transfer matrix method is use-
ful for analysis of the vibration of such a rotating
system. The configurations of a low pressure rotor,
a support system, and a rotor-bearing system, are
shown in Fig. 4.10.7 a-c, respectively. The rotating
system is roughly classified into a rotor body and a
bearing system. The rotor body can be replaced
with an equivalent lumped mass and a weightless
0 : Measured values } elastic shaft. The bearing system is divided into a
e : Calculated values 20-in-dia tilting pad bearing bearing section and a bearing support section, each
0 : Measured values } of which consists of a spring and a damping ele-
+: Calculated values 32-in-dia elliptical bearing
ment.
Fig. 4.10.6. a Mean bearing pressure vs. journal centre.
b Rotor speed vs. journal centre b) Support System
The support system is one of the important factors
for determination of the vibration characteristics of
stand the characteristics based on practical condi- a rotating machine. A bearing system can be divid-
tions. ed into a bearing section and a support section, as
illustrated in Fig. 4.1 0.7 b. The support section is
either installed directly on the concrete foundation
or attached to the casing. One example of the verti-
4.10.3 Unbalance Response cal stiffness ratio for each of these cases is shown in
Table 4.1 0.2 with regard to the bearing support,
The admissible maximum amplitude of rotor vibra- rotor, and bearing oil film, and implies that stiffness
tion is in a range from 4/100 to 6/100 mmp-p, in is at almost the same level at the LP section, in each
operating conditions. case. This means that the stiffness of the support
Thorough investigation and quality control are section is also an important factor for determina-
required for rotor vibration at each stage of design, tion of the characteristics of vibration.
manufacture, and assembly, so that the vibration
data of the rotor system extended by 20- 60 m in Table 4.10.2 Stiffness ratio of support elements
total length, weighing 1000-8000 kN, and consist- LP section HP/IP section
ing of 3-6 rotor units may be below the admissible
amplitude regardless of the above-mentioned varia- Rotor 1.00 2.00
Bearing 1.29 2.10
tions in the supporting conditions. Fortunately, ro- Pedestal 1.40 17.9
tor-balancing technique has been greatly improved

297
4 Explanations

c, : Oil damping >, , Oil •<ilfi"~ } Booriog

ma: Equivalent mass

ca : Damping
,, ' '"""'~ }'''""''
a b

Fig. 4.10.7. a Low-pressure rotor.

b Model of support system.
c Model of rotor-bearing system

} Bearing
Equivalent mass
} Pedestal
c

Each constant of the bearing support and equiv-

alent mass of the total bearing system can be ob-
tained by actually exciting the bearing support sec-
tion by means of an excitation system, or by
calculation in the form of the finite element method.
Figure 4.10.8 shows the mechanical impedance ex-
pressed in the following formula, by actual excite-
ment of the bearing support of a low-pressure rotor
section:

IZijl = ~~~= (;;- m ;jw X+ c~. (2)

The spring coefficient, the damping coefficient and

the equivalent mass md = (m;) of the bearing sup-
port can be obtained by the above process. The
cross-coupling terms can be virtually disregarded at
Frequency (Hz) the bearing support.
Fig. 4.10.8. Mechanical impedance of bearing foundation
c) Calculated Examples

(i) Simulation of Unbalance Response

The constants for determining the characteristics
of a bearing support can be expressed in the same The calculated result and the actually measured
way as in the case of the bearing section. The spring value of the unbalance response of the secondary
and the damping coefficients are given in the fol- section of a cross-compound turbine generator as a
lowing formula: result of this simulation are shown in Fig. 4.10.9. A
three-dimensional display of the vibration mode is
(1) also shown in Fig. 4.10.10.
In this rotor-bearing system, the intermediate-
pressure rotor section and low-pressure/generator
k;j : spring coefficient, rotor section are supported by a tilting pad bearing
C;/ damping coefficient, and an elliptical bearing, respectively. It is possible
x: horizontal direction, to confirm and investigate various characteristics of
y: vertical direction. vibration, such as vibration mode, unbalance, sensi-

298
 4.10 Application to Steam Turbine Generator
lOOr---r--------.-----T--------. Table 4.10.3 Influence coefficient (10- 3 11m/g- mm, degrees)
---- Measured before improvement
- - Calculated
- - - Calculated after improvement
Bearing no. *1 #2 3* #4
Measured 0.04, 3° 0.31, 130° 1.11, 31 ° 1.08, 213°
Calculated 0.17, 3W 0.29, 91 o 0.99, 60° 1.07, 242°

measured values and calculated values of the influ-

0~~--~~--~~--~~--_J
1500 2000 2500 3000 3500
ence coefficient at each journal section of the LP
rotor, shown in Fig. 4.10.10.
Speed (rpm)
On the other hand, Fig. 4.10.11 shows the change
Fig. 4.10.9. Example of unbalance response between the vibration vectors observed at each
journal section before and after a balance shot. This
balance weight and the position has been calculated
by means of an influence coefficient obtained
J;L~~ through the process of calculating the unbalance
~~;15 #6 response.
Both the influence coefficient and the vibration
vector present practical values, although there are
some differences between calculated and measured
5
values. As stated above, the number of balancing
shots can be substantially reduced, provided that
the unbalance response is calculated accurately.
3000rpm

(iii) Q Factor Design

Fig. 4.10.10. Three dimensional vibration mode of multi-span It is well known that not only the critical speed, but
rotor also the response sensitivity at the critical speed is
very important. If the calculation of an unbalance
response can be carried out accurately, Q factor
design is possible. This is the design of a rotor-bear-
tivity, etc., by simulating the unbalance response of ing system in which the Q-factor representing the
an actual machine in this way. resonance magnification of a vibration mode, given
in the following formula, or in Fig. 4.10.12, is used
(ii) Influence Coefficient as a value for evaluating the response sensitivity

The influence coefficient method can be applied to (4)

the balancing of an actual machine. The objective of
this method is to achieve the most effective balanc-
ing after checking the influence of each balance i: i-th mode
weight affecting vibration by the following formula: (cfcc);: damping ratio of i-th mode

Biik A;k (Xijk uj (3)

t
Vibration after
t
Initial
t
Influence Trial
t
addition of a vibration coefficient weight
trial weight

i: position of measurement
j: corrected position
k: balancing speed
(a) i · k = j; [W] = - [a] - l [A]
(b) i · k > j; [aV [a] [W] + [aV [A] = o
W; corrected weight
Balancing on an actual machine can be facilitat-
ed if the influence coefficient has been obtained in
advance by calculating the unbalance response.
Table 4.10.3 shows the comparison between the Fig. 4.10.11. Change of vibration vector

299
4 Explanations

X max. o·

(C/Cc <1)

WI / wn w lwn

Fig. 4.10.12. Relationship between response curve and Q-factor

,,
Jso·
5 Fig. 4.10.14. Change of vibration vector due to load

4
v
v
• • On the other hand, Fig. 4.10.14 shows the change
in vibration vector due to load for the same rotor-
• bearing system as in Table 4.1 0.4. Table 4.1 0.4
shows that the Q factor is decreased to a large ex-

• • •
tent and the measured value is close to the calculat-
ed value. Furthermore, Fig. 4.10.14 shows that the
change in the vibration vector due to load is also
,, ', '
• small. These results show the effect of a lowered Q
factor. Meanwhile, this unit has been put into oper-
0 ----
- - 5 10 15 20
ation without field balancing.

Q-factor
4.10.4 Self-Excited Vibration
Fig. 4.10.13. Q-factor vs, number of balancing shots
Self-excited vibration of a steam turbine generator
The approximate relation between the number of is of two kinds: oil whip caused by the characteris-
balance shots applied for an actual machine and the tics of bearing oil film and steam whirl caused by
Q factor is shown in Fig. 4.10.13, in which the num- steam force. In addition, unstable vibration due to
ber of shots tends to increase in proportion to the shortage of oil supply to a bearing with a large bore
increase in the Q-factor. This shows the significance can be regarded as an unusual example. The dy-
of the Q factor. namic characteristics of bearing oil film relate to
Various methods of decreasing the Q factor have both cases.
been tested by calculation for the rotor-bearing sys- Some typical cases are introduced.
tem having the characteristics of vibration shown
by the broken line in Fig. 4.1 0.9. a) Oil Whip
As an example of applying the results of the test
to an actual machine, the effect of increasing the Oil whip is a well-known phenomenon; its causes
pedestal stiffness is introduced. The chain line in have b een clarified and preventive measures estab-
Fig. 4.1 0.9 shows the characteristics of vibration af- lished. Accordingly, it now rarely occurs in an actu-
ter the increase in pedestal stiffness, providing an al machine. With sufficient prior investigation into
example of comparison between a Q factor based stability attempts are made to apply a pad bearing
on the measured data and the calculated value, a s with excellent stability to a high-intermediate-pres-
shown in Table 4.1 0.4. sure rotor subjected to a comparatively low bearing
load. Figure 4.10.15 shows the results of subjecting
those bearings used in a large steam turbine-gener-
Table 4.10.4 Improvement in Q factor ator to their stability limit values. A certain design
Before improvement After improvement margin is maintained on the stability limit value for
each bearing, while both operational conditions
Measured 14.5 4,9
and change of the bearing load due to ageing effect
Calculated 13.6 5.6
are taken into account.

300
 4.10 Application to Steam Turbine Generator

4.0 PAD GEL D. EL

I
,Y ~(~EL
,
0 GEL
D PAD
a
_, / I. .I
...__ GEL
a 2.0
I
, joo~ ' D
0

'-.._,I PAD

0+---~~--~----~----~----~--~~
0 0.4 0.8 1.2

So=;%(%)'
Fig. 4.10.16. Model of flexible rotor
Fig. 4.10.15. Stability limit of bearings in operating condition

If only the dynamic characteristics of bearing oil I . I I •

Supply 011 temperature ..;.:49'-"'C--!
film are well known, sufficiently practical investiga-
tions concerning stability against oil whip can be 46'c l
carried out in the simple model of a rotor-bearing
system shown in Fig. 4.1 0.16. Figure 4.10.17 shows
a case in which the oil whip generated in an ellipti-
cal bearing as a result of reduction of a bearing load
caused by a perceptible change in the alignment
was eliminated by increasing the temperature of the
supply oil to a certain extent. The state of stability
of each bearing in this case is shown in Figs. 4.10.18
and 4.10.19.
Vibration wave CASE-1 CASE-2
Figures 4.10.18 and 4.10.19 present the stability
diagram obtained from Hurwitz's stability analysis Fig. 4.10.17. Example of oil whip
and the solution of a characteristic equation ob-
tained with the substitution ?c = a + i b. The
threshold speed is 1.7 - 1.8 times the natural fre- 3.0
quency in case 1, but it is stabilized in case 2, where
the temperature of the supply oil is increased by
3 oc or 4 oc. Thus, the above example seems to
simulate practical conditions accurately. The other 2.1
Unstable Stable
example relates to self-excited vibration due to in-
adequate oil supply of a bearing. Figure 4.10.20 q t::. OCASE- 1 r--
shows a measured example of unstable vibration
due to this phenomenon and the result of frequency 1.2 I N = 3000RP M
PM = 0. 87MPa f---

analysis. I Tout = 60 . O'C

-
I
ACASE-2
These data show a kind of self-excited vibration. \ N =3000RPM
PM=0 . 87MPa -
.........
As the quantity of supply oil decreases the circum- Tout = 64. O'C

ferential width of the oil film ep also decreases, as

0.3
0.0
shown in Fig. 4.10.21. The stability of this case ex- 0.1 0. 2 0.4 0.6 0.8 1.0
hibits its tendency as shown in Fig. 4.10.22. Thus, So =~(*) '
the existence of the limit value of eP relating to the
stability is proven. According to the information Fig. 4.10.18. Stability chart-1
given above this phenomenon may also be caused
w/ wn
by the dynamic characteristic values of the oil film,
1.0
and the extreme shortage of oil supply may reduce
the circumferential width of oil film and induce an
unstable vibration. - 0.0025
>:
a
b) Steam Whirl ci -0.005

A self-excited vibration dependent on a load has

- 0.0075
recently been found to occur in the high-pressure
rotor of a steam turbine. This unstable phe- Fig. 4.10.19. Stability chart-2

301
4 Explanations

mmp-p Record of vibration meter nomenon is called steam whirl, and is generated by
steam force exerted on the rotor. It is similar to the
= 0.09 oil whip caused by the properties of a bearing and
= 0.06 is characterized by its whirl with the natural fre-

= 0.03
= :::::o Time
(Normal) I==
quency of the rotor-bearing system.
The force which generates self-excited vibration
is caused by the action of a rotor which rotates
I

+Vibration wave eccentrically in steam, and two different kinds of

mechanisms in their physical properties are taken
into account.
One is an unbalanced torque force caused by
uneven tip leakage at a rotor blade in the circumfer-
ential direction, and the other is a force due to
XIO-' mm P-P
uneven pressure distribution in the circumferential
30
direction. This force is generated in a narrow space,
namely in the labyrinth seal section for sealing
20 Frequency analysis
c. steam.
s
«: An outline sketch of the area where the destabi-
10 lizing force take place is shown in Fig. 4.1 0.23. The
same phenomenon has also been experienced in
0
other turbomachines such as turbocompressors
0 25 50 75 100 and aeronautical gas turbines.
I (Hz)
Primarily, this phenomenon occurs under condi-
Fig. 4.10.20. Measured example of vibration due to starved lu- tions in which the fluid force is relatively stronger in
brication comparison with the elastic force of the rotor.
Taking this into account, in the high-pressure
rotor of a steam turbine with a high output density,
it is essential to check in advance whether this sort
of self-excited vibration is possible or not. In the
180 following, the stability of rotor-bearing system, in-
cluding the exciting force, will be introduced.
.., 140
(Seal force) (Seal force) + (Torque unbalance force)
~"
<>.
100
""
60
L 0.5 1.0
q,/q, ,p••

Fig. 4.10.21. Relationship between circumferential width of oil

film and supplied oil flow
Fig. 4.10.23. Destabilizing forces

(i) Destabilizing Force due to Torque Unbalance

This force depends largely upon blade tip leakage
and stage output that can be expressed by the tip
" 10 clearance and the number of labyrinth fins.
11 As seen in Fig. 4.10.24, the destabilizing force can
Op-(deg)

60 be found in the following way, with its eccentricity

"' 0 I
~~~' (e) in x direction. Provided that a local tip clearance
~! - 10
is given, a local torque can be expressed as
= rm (1 + r:xe cos 0)
io (5)
rm: average values at no eccentricity
-20 r:x: coefficient, depending upon tip clearance,
Fig. 4.10.22. Stability depending on circumferential width of oil number of fins, and height of blade, and rep-
film resenting the amount of leakage

302
 4.10 Application to Steam Turbine Generator

The above formula is expressed by means of the

dynamic characteristics of the bearing, ratio of
destabilizing coefficient to rotor stiffness kefks, ratio
of flexibility of shaft to bearing clearance IJ, and
Sommerfeld number S0 , which are thought to influ-
ence the stability of a rotor system on which the
destabilizing force is generated.

(ii) Stability of Rotor System Subjected

to Destabilizing Force
The r:x value is needed to find out the destabilizing
Fig. 4.10.24. Mechanism of destabilizing force coefficient (ke) to allow investigation of the stability
under the condition defined by the aforementioned
homogeneous equation. An example of a trial cal-
A torque expressed in the above formula shall be culation for a high-pressure turbine rotor is shown
decomposed into the separate components in the x in Table 4.10.5. Meanwhile, the value obtained by
and y directions, and each shall be subjected to summating the destabilizing coefficients of each
integral operation: stage presents the whole (ke), in a multi-stage rotor.
r 2
Results of investigations made on the influence
Fy = ___!!~_ J(1 + r:xe cos 0) ·cos OdO = nrmr:xej r, given by the destabilizing coefficient on the stability
r o
r 2 (6)
F x = ___!!~_ J(1 + r:x e cos 8) · sin 8 dO = 0.
r o
The solution of the above equation shows that a
force proportional to the eccentricity is exerted only
in y direction when the rotor is shifted in x direc-
tion. At this stage, equations of motion, including
the destabilizing force, can be expressed as follows,
provided that Fy is expressed as ke = Fy/e and a
simple system, including mass (m) and shaft stiffness
(k8 ), as shown in Fig. 4.10.16:
mxs + ksxs- ksxb +keYs= 0,
mys + ksYs- ksYb- kexs = 0, 1. 72
1.41
kxxxb + kxyYb + cxxxb + CxyYb (7) 1.49
- ksf2(X8 - xb) = 0, Stable D rotor \l 0.012 1.49

kyyyb + kyxxb + CyyYb + cxyxb OL---r---r---~--~--~--~------

0.2 0.4 0.6
- ks/2(ys - Yb) = 0. So
In addition, if substitutions xs = X8 e;. 1, ys = ys eM, Fig. 4.10.25. Stability chart of rotor subject to destabilizing
xb = xbe;. 1, Yb = ybe;. 1 are assumed to be acceptable, force
the following homogeneous equation is obtained.

w;
___!_ A 2 +1 k e /k s - 1 0 Xs

0 -1 Ys
=0 (8)
A So A
0 - Pxx
w
+ Yxx +-
'7
W Pxy + Yx y .xb
A A S
0 WPyx+ Yyx - pyy
w
+ Yyy + -I]0 Yb

where substitutions w;
= ksjm, 1J = qjC, and defi- of several high-pressure rotors, supported by six-
nitions q = static deflection of rotor, C = bearing pad bearings, is shown Fig. 4.1 0.25. It can be seen in
clearance, S0 = (pmff.lW) (C/R) 2 , Yii = 2kii · S0 · Cf this figure that stability tends to decrease as kefks
(mg), pii = 2 cii · S0 · C · wj(mg) are applied. increases. That is, the destabilizing force increases

303
4 Explanations

Table 4.10.5 Destabilizing coefficient of high-pressure turbine ~ =0.6

Stage 1st 2st 3st 4st Sst

Blade height (mm) 25 55 65 80 95
C( 0.05 0.02 0.019 O.Q18 0.016
k. kg/mm 445 230 200 180 130

.,
.
"0
.~
4/4 "i5.
E a
"0 ~
Load '"
"'0 !roo
...l 2/4 Amplitude
(( P-P)
7
'
X.o __) $
0 mm

Fig. 4.10.26. Vibration record of steam whirl of a rotor

Fig. 4.10.28. System damping of rotor system subject to desta-
bilizing force

mentioned characteristic equation. Although the

Wn =34Hz data shown in Fig. 4.10.25 were obtained by Hur-
witz's stability analysis, it is also possible to express
its stability by assigning the system damping of
a rotor-bearing system obtained by direct solu-
0 60 100 Hz 150
tion of the characteristic equation, as shown in
Fig. 4.10.27. Frequency analysis of steam whirl of a rotor Fig. 4.1 0.28.

against the rotor elastic force. Rotor A in this figure

stays near the stability limit line, while involving the
4.10.5 Concluding Remarks
possibility of generation of unstable vibration.
Figures 4.10.26 and 4.10.27 show the record of The dynamic characteristics of a bearing oil film
actual vibration of rotor A and the frequency anal- and their relation to rotor vibration have been in-
ysis of this respectively. The vibration amplitude troduced, with reference to the example of a large
increases to some extent on overload, and thus, steam turbine generator. As stated in this section,
depend on load. the dynamic characteristics of an oil film seems to
The frequency of whirl stays almost at the natu- be increasingly important, in keeping with the im-
ral frequency of the rotor system. provement of recent vibration analysis technology.
As can be seen so far, even in a rotor-bearing In particular, an improvement in its accuracy may
system subjected to steam whirl it is possible to clarify new problems of rotor vibration in a quanti-
investigate the stability by analysing the above- tative and practical manner.

4.11 Application to Turbo-Compressors

4.11.1 Introduction pots, and fertilizer plants. They are, however, the
focus of debate from the viewpoint of vibration
Turbo-compressors, which include centrifugal and technology, since they are inherently susceptible to
axial flow types, are used in such diverse fields as the vibration of vanes, shafts, and the frame sup-
petrochemical and refining plants, natural gas de- porting the main body. Generally, a rotor-bearing

304
 4.11 Application to Turbo-Compressors

system readily causes forced vibration due to the a) For Critical Speed
unbalance force or a self-excited vibration due to
the oil film force in bearing parts. Designers are, The actual critical speed shall not encroach upon
therefore, warned to give adequate consideration to the specified operating speed ranges, the margin of
these properties of the turbo-machine. which is at least 15% below or 20% above the
The lateral vibration system is made up of shaft, continuous speed. The amplification factor shall
impellers, couplings, bearings, etc. The weight of the not exceed 8 while going through criticals. (Values
impeller and coupling, and the rigidity of the shaft less than 5 are preferable.) Previously, undamped
and bearings are the determinants for the critical critical speeds were used to judge whether critical
speed of the system, whereas the internal friction of speed did not encroach upon the specified operat-
the system, the fluid dynamic exciting force induced ing speed range. However, damped natural frequen-
at sealing points and impeller and the reaction force cies and amplification factors obtained by consider-
of the oil film provided for journal bearings, etc., ing the oil film effect of bearings were introduced
have a marked influence on the resonance level and after revision.
threshold of self-excited vibration as a damping or
destabilizing effect. Therefore, control of damping
influence is indispensable for reduction of the am-
plitude in the resonance frequency range, protec- b) For Allowable Vibration Level
tion against unstable vibrations, and avoidance of
resonance. However, quantitative determination of The double amplitude of vibration in any plane
the internal friction and the fluid dynamic exciting measured on the shaft near the bearings shall not
force is almost impossible at present. For this rea- exceed 25.4 x j12000fN flm, where N is the maxi-
son, it must be pointed out that the dynamic char- mum continuous speed in revolutions per minute.
acteristics of the bearing oil film play an important
role.
Particularly in the case of a centrifugal compres-
sor, which is intended for higher speeds and pres- c) For Balancing
sure, the rotor is operated at more than the critical
speed in many cases. In such machines the vibration The rotating element shall be dynamically balanced
amplitude of rotor should be suppressed each time in multiple planes during assembly. The maximum
it passes through the critical speed at startup or admissible unbalance force on any bearing at the
shutdown. In addition, the impeller and the maximum continuous speed shall not exceed 10%
labyrinth seal induce a destabilizing force in these of the static loading of that bearing.
machines. Accordingly, it is necessary to positively
These requirements have the following implica-
suppress the vibration resulting from such a desta-
tions for the design and manufacturing process.
bilizing force by means of a damping action provid-
ed by an oil film. (i) In the design stage the rotor system, including
In this section we discuss the design procedure shaft, impeller and bearings, is modelled to suit
for rotor-bearing systems and the related applica- the vibration system, and damped natural fre-
tion of bearing data, which is mainly relevant to quencies and amplification factors are calculat-
centrifugal compressors. Briefly, "Rotor Dynam- ed by the complex eigenvalue analysis tech-
ics" as described under API Standard [1], which nique. These parameters are then evaluated.
was widely made available as the design and fabri- If these values are not desirable, design
cation standard for turbo-machines, are outlined parameters such as shaft diameter or bearing
first, followed by the analytical method which satis- dimensions are modified.
fies the API Standard. (ii) At the manufacturing and assembly stage, ro-
tating parts are balanced by using balancing
machines. Before release from the workshop, a
test is carried out with the machine running to
check that the vibration amplitude is lower
than the admissible values.
4.11.2 Dynamic Design of Rotor Systems (iii) In test operations at the users' factories, unex-
According to API Standard pected vibration may occur due to the different
load or stiffness of base conditions, or other
causes. Therefore, vibration values are moni-
The American Petroleum Institute Standard, tored and checked again. If unexpected vibra-
API 617, for turbo-compressors, which was revised tion occurs an investigation is promptly insti-
in 1979, specifies the following in the section "Dy- tuted; vibration analysis can be effectively used
namics" : to decide on corrective action.

305
4 Explanations

4.11.3 Method of Vibration Analysis

A simplified method and a finite element method

(FEM) for analysis of such vibration characteristics
as critical speed and amplification factor are de-
scribed in this section. A simplified method is
preferable for easy determination of approximate
characteristics, selection of appropriate bearings, or
for modification of the shaft stiffness, while FEM ~
analysis is effectively used for confirmation when
design parameters have been tentatively deter-
mined.

a) A Simplified Analysis

(i) Jeffcott Rotor System Supported by Tilting Pad

Bearings Without Fluid-Dynamic Cross-Coupling 0. 4 0.6
Forces 12/ f2s
Fig. 4.11.3. Chart for critical speed Q and amplification factor
Let us consider the Jeffcott rotor system shown in AF
Fig. 4.11.1 as the basic model of a turbo-compres-
sor. Since the coupling effect is disregarded vibra- where Qs =~'X= kb/k5 , 11 = cbwfks> and ks,
tion in the x-direction and that in the y-direction kb, cb denote the spring constant of the shaft, spring
are considered separately and can be modelled to coefficient and damping coefficient of oil film of
suit the vibration system shown in Fig. 4.11 .2. Thus, bearing, respectively. In the equation 11 = cbwfks
the angular natural frequency Q, logarithmic decre- the value of w is initially set by the undamped nat-
ment J and amplification factor AF are approxi-
ural frequency w = Jkskb/{(ks + kb) m}. Then, if
mately derived as the difference between w and Q is wide Eq. (1) is
Q x(1 + x) + 17 2 recalculated replacing w with Q.
J = n -·--~11-------;;: Figure 4.11.3 called Balda's diagram [2], is ob-
Qs - (1 + x)2 + 112 ' x(1 + x) + 17 2
(1) tained by using AF instead of J in Eq. {1). It is easy
AF = njJ to understand the influence of the oil film effect.
Here, a specimen calculation for a turbocompressor
with the following dimensions is worked:
m = 148 kg, ks = 43.4 MN/m,
k, kb = 53.7 MN/m, cb = 48.7 KNs/m .

(
When Eq. (1) is used, the following parameters are
calculated
2m
Qs 542 radjs, w = 403 radjs,
=
X = 1.237, 17 = 0.452.
Fig. 4.11.1. Jelfcott rotor system
Then, substituting these values into Eq. (1),
Q jQs = 0.755, Q = 409 radjs, J = 0.478
and AF = 6.6
are obtained. Alternatively, in the case of Fig.
4.11.3,
k,
kbj(2ks) = x/2 = 0.619, cbwjkb = 11/X = 0.365
give the values QjQs = 0.75, AF = 7. Either may be
k, k, = k, + i we, used. In this example the effect of the oil films low-
ered the natural frequency by 25% compared with
k,k.
-K=---= the simply supported condition, and the amplifica-
ks + k, tion factor is 7. In addition, the optimum damping
b
a value that minimizes AF is given as 11 = (1 + x)/2
Fig. 4.11.2. Equivalent vibration system from Eq. (1).

306
 4.11 Application to Turbo-Compressors

(ii) Jeffcott Rotor System Supported by Circular force

or Two-Lobe Bearings without Fluid Cross- W = J kskxx/{(ks+ kxx) m} = 357 radjs.
Coupling Force
Then kb + iwcb = (34.3 + 42.9i ± 53.3i) MN/m by
Since circular bearings or multilobe bearings have using Eq. (3).
cross-coupling terms in the oil film coefficients, the For the unstable mode corresponding to the neg-
vibration modes are coupled with the x- and y-di- ative sign in the above equation, kb = 34.3 and
rections and show a forward and backward whirl. w cb = - 10.4. These values give
In this case the model shown in Fig. 4.11.2 can be
used by introducing the equivalent complex stiff- Qs = 542 radjs, X= 0.791 , lJ = - 0.237
ness [3] and then
kb ± iwcb = t [(kx x + kyy) + iw(cxx + Cyy) Q jQS = 0.671, Q = 364 rad/s, J = - 0.506
± {(kxx- kyy + iwcxx- iwcyy)21 are obtained from Eq. (1). In this case self-excited
+ 4(kxy + iwcxy) (kyx + iwcyx)} 2] (2) vibration is assumed to occur because J is negative.
where kxx• kxy• kyx• kYY are oil film spring coeffi-
(iii) Jeffcott Rotor System Supported by Tilting-Pad
cients, cxx• cxy• cyx• cYY are oil film damping coeffi-
Bearings with Fluid Dynamic Cross-Coupling
cients, w is angular frequency, and kb, cb are the
modal stiffness and damping coefficient defined in Forces
Eq. (2). Substitution of kb and cb into Eq. (1) yields Although destabilizing forces generated at the im-
QjQ8 , J and AF. Double signs of Eq. (2) mean both peller or labyrinth seal section are not yet sufficient-
the forward and backward whirls. For simplifica- ly understood, their effect can be equivalently repre-
tion, the imaginary term in the bracket in Eq. (2) sented by the cross-coupling stiffness Q, which is
has been disregarded, as below: equal to kx y or - ky x• in a study of the rotor stabil-
kb + iwcb = t [(kxx + kyy) + iw(cx x + Cyy) ity. Then the natural frequency Q and logarithmic
± {(kxx - kyy)2 - W2(cxx - Cyy)2 decrement J are approximately expressed [4] as fol-
1 lows:
+ 4kx ykyx- 4w 2 cxycyxJ2]. (3) Q jQs = ) Real(- s2),
1
When the value of the term { }2 in Eq. (3) becomes J = n x I mag. (- s 2 )/Real (- s2 )
imaginary, the modal damping value cb decreases
where
and then becomes negative, which means that the
system is unstable or self-excited vibration occurs. ( -s 2) = {(Kx + Ky) ± j (Kx - Ky) 2 - 4Q 2 }/(2m)
Therefore, it is found that the stability of the
rotor system increases with the increase of the dif- and
ference in the principal stiffness (kxx - kyy) and with
K = k 8 (kxx + iwcxJ
the decrease in the difference in the principal damp-
ing coefficients (cx x - cy)· The cross-coupling stiff- x ks+ kxx + iwcxx'
ness terms having a different sign and cross-cou- K = k.(kyy + iw cyy)
pling damping terms having the same sign tend to Y k8 + kyy + iwcYY
destabilize the system. In particular, a bearing in
which the signs of kxy and kyx are not the same and Kx and KY are the resultant complex spring con-
their absolute values are large has a tendency to stants composed of the spring and damping of shaft
destabilize the system. and bearings of the model in Fig. 4.11.2. From this
Next, a specimen calculation is worked. The cir- equation, the maximum cross-coupling Qmax which
cular bearing with two grooves is used in the model makes J equal to zero is obtained in the following
in Fig. 4.11.1. The constants for calculation are as dimensionless form [4].
follows:
qmax = Qmax/ks
m = 148 kg, k8 = 43.4 MN/m, qmax = ± j'"(B-;:::-..-.+-B-~:::-,-)-(1---B-=~-)-B-=;,,...--;)
kxx = 33.5 MN/m, kxy = 60.3 MN/m,
where
kyx = - 38.4 MN/m, kyy = 35.2 MN/m,
cxx = 171 KN sjm, cxy = 13.8 KN sjm, BDr = {(Qx/Qsf - (Qy/Qsf} /2 (4)
cyx = 13.8 KNs/m, cYY = 69.2 KNsjm. Bmi = {(Qx/Qs) 2 Jx + (Qy jQs) Jy}/(2n)
Bv; = {(Qx/Qsf Jx - (Qy/Qs) Oy}/(2n)
First, the angular frequency w is calculated, which
is the undamped natural frequency in x-directional and (Qx/Qs) and Jx are dimensionless frequency and
vibration without the cross-coupling destabilizing logarithmic damping ratio in x-directional vibra-

307
4 Explanations

tion, while (Q)Qs) and JY are those in y-directional The eigenvalues s and eigen vectors V can be nu-
vibration, which can be obtained from Eq. (1). merically obtained by the QR method from Eq. (6).
As Q increases, J corresponding to forward whirl, For this purpose, the dimensions of shaft, weight
which has a negative sign, decreases, then becomes and moment of inertia of the impeller, dynamic
zero at Qmax' and eventually becomes negative. characteristics of bearings, etc. must be fed in, so
Since a negative J means the occurrence of whirl, that M, C and K can be automatically arranged in
the greater Qmax is desirable. the computer program and that the pair of s and V
An example of the model in Fig. 4.11.1, having can be calculated through the QR method. When s
is a complex value expressed as rx + fJ i the following
m = 148 kg, = 43.4 MNjm
ks
equation becomes applicable:
kxx = 69.8 MNjm, kyy = 5.41 MNjm
x =Vest= Ve" 1 (cosfJt + i sin[Jt). (7)
c xx = 84.2 KNs jm, cy Y = 24.4 KNs jm
The above equation implies a free vibration, as seen
is shown as follows. For x- andy-directional vibra- in Fig. 4.11.4, whose amplitude varies with e"1 and
tion, whose angular frequency is equal to fJ. When rx > 0,
Qx/Qs = 0.806, Jx = 0.531 unstable vibration occurs.
fly /Qs = 0.343, JY = 2.12 In order to express the vibration characteristics,
the logarithmic decrement J = - 2 n rxj [J, amplifica-
are obtained by applying Eq. (1).
Substituting these values into Eq. (4) provides
tion factor A F = J
rx 2 + fJ 2 / ( - 2 a) and natural fre-
quency f = fJ/(2 n) are used instead of rx and fJ.
Bvr = 0.266, Bmi = 0.0946, Whirl directions of vibration and distribution of
amplitude can be judged from the vector V repre-
BDi = 0.0152, qmax = 0.278,
senting vibration mode. An example of the above
and then Qmax = qmax X ks = 12.1. process is shown in Fig. 4.11.5 b.
It is also possible to determine the amplitude
caused by unbalance forces by solving the following
equations
b) Finite Element Method Analysis
Mx+Cx+Kx= f . (8)
When gyroscopic effect, difference in characteristics
between right and left bearings, and asymmetrical Equation (8) can be reduced to the following form:
distribution of mas or stiffness along with the axial [K- M w 2 + iwC] X= U w 2 (9)
direction must be taken into account, or modes of
a higher order must be determined, the simplified where
method is not considered adequate; FEM can then X: amplitude
effectively be used. w: angular velocity
Where a rotor system is modelled as a multi-mass U: vector representing unbalance
point system, the free vibration equation can be
expressed as follows: At the design stage, it will become possible to check
whether the limiting value of amplitude specified in
M X+ c X+ KX = 0 (5) the API Standard is complied with by obtaining X.
where
x: displacement vector
M: mass matrix
C: damping matrix
K: stiffness matrix
all of which are determined through the finite ele- time
ment procedure. M comprises the mass of each
point and the moment of inertia of the impeller. C
involves the gyroscopic effect and the damping co-
efficients at the bearings. K involves the spring coef-
ficients of bearings and the fluid dynamic cross-cou-
pling stiffness, in addition to the stiffness of the
shaft. When X = vest is introduced in Eq. (5), it can
be converted to

(6)
Fig. 4.11.4. Wave form in free vibration

308
 4.11 Application to Turbo-Compressors

Impeller vice speed of 7620 rpm and driving power of

Shaft
3000 kW. The rotor system at the initial design
stage is shown in Fig. 4.11.7. The mass of the rotor
system involving the impellers is 296 kg and the
bearing load is W = 1450 N. In addition, a bearing
y with L : D = 0.5 and mp = 0 corresponding to TP 5,
LOP (data no. 48) was adopted, since the tilting pad

~·
Rotating speed
bearing characterized by a higher stability is pre-
ferred in the high-speed centrifugal compressor.
Also, bearing diameter D = 76.2 mm, viscosity
10400rpm f.1 = 10.8 MPa · s and clearance ratio of 2 CPI D
Log. decrement
= 2.5/1000 were used.
-0.01 For the purpose of vibration evaluation, an ei-
Nat ural frequency genvalue analysis with reference to the FEM model
59.95Hz
shown in Fig. 4.11.7 was conducted to assess the
(b) Results
rigidity of the shaft. The critical speed under simply
supported condition Q s was found to be 542 radjs
Fig. 4.11.5. Analysis model and calculated results (5126 rpm).
The mark 0 represents the node where mass and
moment of inertia of impeller is taken into consider-
In this chapter only the finite element method is ation. The resulting Qs shows that the rotor is flex-
applied to the analysis, but the transfer matrix ible, since it is found to be lower than the service
method is also available. speed.
Next, the damped critical speed and damping
ratio, which take into account the oil film charac-
4.11.4 Application to a Centrifugal teristics of bearings, are discussed using the simpli-
fied method. The bearing data given in dimension-
Compressor less form has been converted to dimensional form,
as shown in Table 4.11.1.
A centrifugal compressor shown in Fig. 4.11.6 is Assuming that the equivalent mass m is
used as an example of a situation in which the 148 kg = 1/2 the total mass, the equivalent spring
bearing data and the analytical procedure are ap- constant ks can be calculated as ks = m, Q; =
plied. Usually, from the given specifications, such as 148 x 542 2 = 43.4 MNjm. The natural frequency
those involving pressure and flow rate, the machine
type is determined first and then the number of
stages, dimensions of the impellers and speed are
fixed in the performance design. Further, the di- 1960
1510
mension of impellers and consideration of the seal 0

....,_
0 N
;::;
....
N
and arrangement of bearings make the design con- Q I--... t~ .n.
cept of the rotor system more concrete. As a result, t--riM" - ~ ~

it can be said that the major vibration design ~ ' - ...J FJ:I 1.I
parameters are type of bearing, configuration of
sealing parts, and shaft dimension.
The example discussed here assumes the rotor
i wf~ ' --;:?p14
' -------------------- W=1450N
system design of a centrifugal compressor with ser- Q, =542 rad/s

Fig. 4.11.7. Shaft dimension and simple supported natural fre-

quency

Table 4.11.1 Oil film coefficients of tilting-pad bearings

Rotor speed N Spring coefficient Damping coefficient

(rpm) (MN/m) (kN · sjm)

k xx kyy cx x Cyy
speed
13200 53.2 7.23 48.7 22.2
7620 69.9 5.41 84.2 24.4
Suction Discharge Power 3880 95.0 3.54 169 26.9
3000kW 2586 114 2.65 257 28.1
1602 137 1.84 414 29.6
Fig. 4.11.6. Centrifugal turbo-compressor

309
4 Explanations
2nd 1.0
Vertical 2.4
200 1.7 l.S 2.0

- - FEM analysis FEM analysis

-----Simplified method
( )
"'c
~

0.5
Ci
II
1 -so
_2"N
"'
E
'-. ...u"'
>.
p"'
~ 100
.;,
"c:r 1 st 0

.::"' Vertical AF=5 . 9 ...l

AF=(6.6)

Cross-coupling stiffness
Q MN/m
-0.5
5000 10000
Rotor speed N" rpm

Fig. 4.11.8. Natural frequency, amplification factor vs. rotor Fig. 4.11.9. Decrease in log decrement due to cross-coupling
speed stiffness

and the amplification factor relevant to each speed curve for Q - () in Fig. 4.11.9 falls at the top right of
are calculated on the basis of m, ks and the bearing the graph.
data using the procedure specified in paragraph A verification through FEM should be made in
4.11.3 a.(i), and they are plotted with broken lines, the final design stage, in addition to the simplified
as shown in Fig. 4.11.8. Numerical values in this method discussed above. Results obtained from
figure represent the amplification factor. The solid FEM are plotted in Figs. 4.11.8 and 9. The results
diagonal line indicates the frequency equivalent to of FEM analysis shown in Fig. 4.11.8 confirm that
the rotor speed. Point <D, at which this line crosses in the first mode, the natural frequency and ampli-
the broken line, indicates the critical speed. The fication factor predicted by the simplified method
requirements of the API Standard will be adequate- agree well with those yielded by FEM, while in the
ly satisfied, since the critical speed of 4500 rpm is second mode the natural frequency is equal to
equal to 59% of the service speed and the amplifica- 153% of the service speed. Thus, the provisions in
tion factor is estimated to be 5.8. the API Standard are adequately complied with.
Meanwhile, in the case of a high-pressure cen- The logarithmic decrement under the influence of a
trifugal compressor, the possibility of unstable vi- destabilizing force is shown in Fig. 4.11.9. In this
bration during operation under load must be taken case, although Qmax is a little smaller than that ob-
into account. Whenever a disturbance occurs due tained through simplified analytical procedure, it is
to the fluid dynamic cross-coupling stiffness Q at well proved that Qmax has a sufficient margin, in
the impeller or labyrinth,() is reduced and vibration contrast with Q obtained through the Lund equa-
readily occurs. However, where Qmax specified in tion. Altogether, it can be confirmed that the rotor
paragraph 4.11.3 a.( iii) exceeds Q, the condition of and the bearing system of the present turbo-com-
() > 0 can be maintained, and stability is always pressor have been designed to complete satisfac-
retained. In the present example Qmax is equal to tion.
12.1 MN/m, as shown in specimen calculation (iii). If the compressor in Fig. 4.11. 7 had been
As for Q, however, Lund [6] introduced Q = p T/ equipped with circular bearings or two-lobe bear-
(2 r h) using the so-called Alford force, where T, r, h ings instead of the tilting pad bearings, the system
and pare torque, effective radius of impeller, outlet stability would have been as shown in Fig. 4.11.10.
width of impeller and factor, respectively. Assuming Here the former is designed in accordance with data
that T= 3.68 kN · m, r = 150 mm, h = 10 mm, and no. 2(i.e. 10° x 2, L: D = 1.0 and 2 CP/D = 2/1000)
P= 3, then Q is 3.68 MN/m. As Q < Qmax> the and the latter with data no. 13 (i.e. 10° x 2,
design therefore proves to have an adequate mar- L: D = 1.0, mp = 0.5, p = 170°, 2Cp jD = 4/1000
gm. (2Cb/D = 2/1000)). D, Wand 11 of the tilting pad
However, the destabilizing force is quantitatively bearing are used in these cases without change. As
not sufficiently understood. For instance, there is the oil whip attributable to the oil film characteris-
some possibility of the existence of a negative tics must be considered in these cases, () is plotted
damping force, in addition to the cross-coupling against speed in Fig. 4.11.1 0, referring to the first
stiffness force. Consequently, not only the condition forward mode. [Point ® in the diagram corre-
of Qmax > Q but also an additional margin obtained sponds to that illustrated in the calculation exam-
by elevating() near to Q = 0 is desirable, so that the ple of paragraph 4.11.3 a.(ii).] It appears that an oil

310
 4.11 Application to Turbo-Compressors

2.0
----- Simplified method
- - FEM analysis

Self-fexcited
1.0 -::1.. whirl
~ 61Hz
"0
::l
-~ {7

[
'<> P..
;: E
"'8 0
..:X: A Pd =4.9
c
"'... I
Unbalance
""' ¢

'J
Q response
.0 173Hz
0
-1 Pd=5.5 ¢
-1.0
& A
d 0 100 200
Frequency Hz
bearing
Fig. 4.11.11. Relation between discharge pressure and vibration
-2.o ~o.. _ _ _ __.__ _ _ _ ___
0
~...- .....~

5000 10000 15000

Rotor Speed rpm
Unbalance
Fig. 4.11.10. Dependence of log decrement on rotor speed

whip can occur in either of these cases, since b in

one of these modes is negative in the higher speed O.ls
range. Fig. 4.11.12. Wave form at whirl onset

4.11.5 Trouble Shooting in the Event of

Centrifugal Compressor Vibration

Finally, unstable vibrations will be discussed. A

high-speed centrifugal compressor is introduced
here as a typical case of vibration successfully over-
come through appropriate selection and replace-
ment of journal bearings. As related in the reports Fig. 4.11.13. Locus of vibration
published by Ferrara [7], Fowlie [8], Wachel [9] and
Smith [10], and also in the Keybob Southplant
Ekofisk case [11 ], self-excited vibration, which is the graphs, an unstable vibration with approxi-
likely to be attributed to fluid dynamic force of mately 60 Hz began to appear at the discharge pres-
centrifugal compressor with high speed and pres- sure of around Jl.J = 3.9 MPa.
sure, has attracted interest in many industries. Also, It gained in amplitude with increasing pressure,
an international conference concentrating on such resulting in an outbreak of large-amplitude vibra-
themes has been held every other year since 1980 in tion at P.J = 5.5 MPa. In Fig. 4.11.12 the waveform
the United States of America [12]. We now cite an is depicted, which shows the development of large-
example of unstable vibration that was experienced amplitude vibration. Figure 4.11.13 illustrates the
with a centrifugal compressor similar to that depict- whirl orbits which occurred just before the out-
ed in Fig. 4.11.6. break of large-amplitude vibrations.
The machine speed was increased to the service Since the machine was equipped with tilting-pad
level of 10400 rpm. However, in the course of in- bearings on both suction and discharge sides and
creasing the subsequent load, the machine suddenly the oil film coefficients involved no cross-coupling
became subject to substantial vibration. The inci- terms, it was unlikely that oil whip would have
dence of vibration can be identified in Figs. 4.11.11 occurred. The cause of vibration had to be sought
and 4.11.12. Figure 4.11.11 shows the result of the in the fluid dynamic destabilizing force caused by a
frequency analysis of the shaft vibration in the hor- swirl flow occurring around the impeller and
izontal direction on the discharge side. As seen from labyrinth seal. Accordingly, the damping ability of

311
4 Explanations

Table 4.11.2 Dimensions of the bearings a b

® Before remedials ® After remedials i: @Before remedials

"'e
Diameter D 76.2 76.2 ..."'
0
L:D Ratio 0.278 0.417
0"' @Aft er remedials
Clearance ratio 2 CP/D 2.5/1000 2.5/1000
.;,
Preload factor mP 0.5 0.1 0

Oil # 90 T ....J 200

# 140T
2Q MN/m ZQ
Cross-coupling stiffness
the bearing part had to be increased to ensure that
the energy dissipated at the bearings remained Fig. 4.11.14. Calculated log decrements and measured vibra-
tions of the rotor before and after remedial work
larger than the vibration energy brought in by the
destabilizing force. This measure increases the loga-
rithmic decrement of the system as a whole, which
can be studied by using the simplified method and specimen applications in turbo-compressors de-
FEM analysis described in Sect. 4.11.3. Conse- scribed. As turbomachines are used at increasingly
quently, a model was applied to the rotor system as high speeds and loads, interest in the problem of
in Fig. 4.11.7. Eight oil film coefficients (for practi- vibration will continue to grow. In particular, the
cal purposes four, as no coupling terms were in- journal bearing supporting the high-speed rotor
volved) were first determined by changing, in incre- plays an important part in influencing vibration.
ments, the width, clearance, preload ratio of the Conventional bearing technology has concen-
bearing, and oil viscosity, and FEM analysis then trated on the prevention of oil whip. Now, however,
followed. Through FEM analysis, it was found that technology is forced to adopt a more positive strat-
the mode in question, corresponding to 61 Hz, was egy, involving prevention of the causes of vibration,
the forward whirl mode. such as fluid dynamic destabilizing force, through
Ultimately, the specification ® in Table 4.11 .2 the damping effect provided by the bearings.
was selected as being pertinent to bearings giving a
larger logarithmic decrement to the mode. The re-
sults of calculation are illustrated in Fig. 4.11.14 a.
The fact that not only J at Q = 0 but also, the value References for Sect. 4.11
of Qmax was larger in bearing ® than in bearing ®
convinced us of the resistance to the vibration. Fig- (1) API Standard 617, Fourth Edition, 1979.
ure 4.11.14 b reveals the results of the operation test [2) Balda, M., IUTAM, Dynamics of Rotors, (1974-8), 30,
using bearings ® and ®. The diagram suggests Springer-Verlag.
[3) Kurohashi, M. et al., Trans. Japan Soc. Mech. Engrs. (in
that bearing ® will be capable of reducing the Japanese), Vol. 47, No. 422 (1981-10), 1269.
unstable vibration component at 60 Hz down to a [4] Kurohashi, M. et al., Trans. Japan Soc. Mech. Engrs. (in
substantially lower level under the same condition Japanese), Vol. 48, No. 433 (1982-9), 1389.
of discharge pressure. [5] Togawa, H., Vibration Analysis by Finite Element Method,
(in Japanese), (1978), 82, Science.
In fact, when bearings ® were used, the vibration [6) Lund, J.W., Trans. ASME, J. of ng. E for Industry ( 1974-5),
was successfully suppressed. 509.
[7) Ferrara, P.L., ASM E Paper No. 77-Det-15.
(8] Fowlie, D.W. et al., ASME Paper No. 75-Pet-28.
[9) Wachel, J.C., ASME Paper No. 75-Pet-22.
4.11.6 Concluding Remarks (10) Smith, K.J ., Proceedings of the Fourth Turbo machinery
Symposium, 115.
[11) Yokoyama, E., Turbo Machnery, (in Japanese), Vol. 7,
Applications of bearing data have been discussed No. 12 (1979-12), 21.
with reference to dynamic characteristics, and some [1 2] NASA Conference Publication 2133.

4.12 Application to Pumps

4.12.1 Introduction thermal and nuclear power plants require reliability

as high as that needed for main units such as tur-
bine generators. In these pumps with relatively high
Pumps are important rotating machines used in capacity, journal bearings are used. Thus it is no
various industries and requiring a high level of reli- exaggeration to say that properly designed bearings
ability. For example, pumps utilized in chemical, decide the reliability of the pumps [1).

312
 4.12 Application to Pumps

Design and vibration analysis of the bearings Table 4.12.1

used in pumps necessitate different considerations Bearing types Variation Applications
from those applied in the case of steam turbines and Small pumps
electric machines. This section discusses how these
considerations are different.
Antifriction bearing ~(jg
oooO
Ball and
roller bearing
Medium pumps
Radial bearing
Pump constructions are classified into two main Thrust bearing

Q
types: horizontal and vertical installations. Typical
Resin
pump structures are shown in Fig. 4.12.1. Cylindrical bore All pumps
Plastic

4.12.2 Bearings Installed in Pumps

s
~
Partial arc
bearing
Q Pressure dam
Oil ring feed
Horizontal
pumps

0
Q)
.D Lemon bore
As the applications and types of pump vary over an Horizontal and
·a
u
(multilobe)
vertical pumps
extensive range, various types of bearings are ap- "'

G
0 bearing
>,
plied. Further, bearings are made by many manu- Ci Axial grooved
facturers and almost all types of bearings are used. Grooved bearing Circumferential Vertical pumps
Typical bearing types installed in pumps are shown grooved
in Table 4.12.1. The types most frequently used in Spiral grooved

0
medium-sized and small pumps are ball and roller Tilting pad Number of Horizontal
bearings, while journal bearings are mostly used in bearing pads varies pumps

Hydrostatic bearing
0 Pocket type Sodium pumps

Radial bearing

Discharge pipe

Suction pipe Water-lubricated bearing

Fig. 4.12.1. Examples of pump structure

313
4 Explanations

large pumps. However, journal bearings are also tional conditions. Therefore, estimation of the force
applied in pumps used for liquids precluding the is difficult at the design stage. Another feature of
use of ball and roller bearings. pumps is that water-lubricated bearings are used.
Various configurations of bearings are selected There are three major differences between water-
according to the desired applications from the view- lubricated and oil-lubricated bearings.
points of performance, cost and maintenance. Hy- 1. Turbulent flow can easily be generated in water-
drostatic bearings are used in special cases. In the lubricated bearings since water has low viscosity.
selection of any bearings the emphasis is on operat- 2. Environmental pressure can be higher than at-
ing experiences.
mospheric pressure when used internally in a
In addition, liner ring seals used in pumps can
pump.
also be considered as bearings from the aspect of
3. The virtual mass of water can have a negligible
vibration characteristics [2 -4]. influence in oil-lubricated bearings, in relation to
point 1 above [7].
All these points can affect both the static and the
4.12.3 Differences Between Pumps dynamic characteristics of the bearings. They will
and Other Rotating Machines be discussed in later sections.

Since pumps handle liquids, the characteristics are

completely different from those of air/gas and elec-
tric machinery. Specifically, hydraulic forces are 4.12.4 Method of Design Study
dominant factors among static and dynamic loads for Pump Bearings
related to vibration and noise, and this is one of the
major features of pumps. Table 4.12.2 shows typical a) Flow Diagram of Design Study
examples of the loads acting on various shaft sys- for Pump Bearings
tems. It can be seen from the table that the weight
of the rotor can be considered a static bearing load The shaft-bearing system of a pump is studied ac-
in the case of conventional horizontal machines cording to the flow chart shown in Fig. 4.12.2. The
such as steam turbines and compressors. In the case problems noted in the double boxes are peculiar to
of pumps, however, consideration must also be giv- pumps. Hydraulic radial load, correction of turbu-
en to hydraulically induced radial loads. This trend lent flow in water-lubricated bearings, bearing
is similar to that of the torque-dependent load act- characteristics of annular seals and hydraulic exci-
ing on the pinion shaft of a drive gear shown in tation force are the new items to be considered.
Table 4.12.2. These inherent problems are under study [1 0, 11]
Further, for vertical pumps, the hydraulic loads throughout the world. Complete date and guideli-
are the major static loads to be considered, since no nes have not yet been published.
weight is imposed on the rotor system by the shaft
[5-9). In the case of dynamic loads, the hydraulic b) Estimation Method for Hydraulic Radial Force
excitation force, which can be disregarded in other
types of machinery, is dominant in pumps [5 - 7]. Generation of the hydraulic radial force acting on
Hydraulic excitation force is by nature not so a pump shaft depends on the pump type, i.e.
simple as unbalance force, which can be determined whether it is horizontal or vertical, and the magni-
by the rotating speed of the rotor, but varies signif- tude and direction of the hydraulic force vary de-
icantly in dependence on the flow rate and opera- pending upon the type and design of the casing. The

Table 4.12.2. Typical bearing loads of various rotating machines

Liquid machine Gas machine Electric machine Gear bearing
(pump) (steam turbine, blower) (generator, motor) (pinion bearing)
Static load Rotor weight • Rotor weight a Rotor weight • Rotor weight a
Misalignment force Misalignment force Misalignment force Misalignment force
Hydraulic load Nozzle force (some high- Magnetic pull Torque force
pressure turbine)
Dynamic load Unbalance force Unbalance force Unbalance force Unbalance force
Fluid exciting force Exciting force Gear mash force

Factors printed hnldface are dominant.

• Horizontal machine

314
 4 .12 Application to Pumps

Determination of static load

{
F min F max Horizontal shaft - reaction force due to rotor mass
Rigid coupling - misalignment force

Type and size of bearing Hydraulic radial load

Calculation of Sommerfeld number

S = pN(R/C) 2 L-0
F
F :Bearing load
J.1 : Viscosity of lubricant
N :Rotating speed
In case ofwaterlubricated bearing:
R :Journal radius
Correction of turbulent flow
C :Bearing radial clearance
L :Bearing width
D :Bearing diameter
Eccentricity calculation
Calculation of oil film thickness
Check of loss, tempature rise etc.

NO

Bearing spring ciefficient Annular seal: Calculation of spring coefficient,

bearing damping coefficient damping coefficient and added mass

Study of critical speed,

stability and unbalance response

Calculation of response against hydraulic

The items contained in double boxes are

peculiar to pumps

Fig. 4.12.2. Design study flow for pump bearing

hydraulic radial force is usually given by the equa- Imaich's research [5] and can be regarded as typical
tion below. The dimensionless radial force used in data of single-volute casing pump. In the figure, the
the equation is obtained experimentally value shown by Stepanoff is frequently used as a
conservative design evaluation. As shown in
F,. = Kr · y · H · D · B'. (1)
Fig. 4.12.3, Kr becomes 0.36 near the shutoff point
where in a pump of this type. Therefore, to reduce the
radial load, balanced casings are used. Typical bal-
F,.: Hydraulic radial force anced casings are double-volute and multivane dif-
Kr: Dimensionless radial force coefficient fuser casings. As shown in Fig. 4.12.4 [9], the radial
y: Specific gravity of pumping fluid forces in these casings are significantly smaller than
H: Head of impeller under scrutiny in the single-volute casings.
D: Impeller outlet diameter Moreover, the direction of the radial force is sim-
B': Impeller outlet width (including shrouds) ple in single-volute casings, as illustrated in
Fig. 4.12.5, while in double-volute and multivaned
Some research workers have investigated hy- diffuser casings the direction exhibits a complicated
draulic radial force. Figure 4.12.3 derives from variance, as shown in Figs. 4.12.6 and 4.12.7.

315
 4 Explanations

0. 40 ..----r---.----.,..- -..,--""T""---r---- ,
-+-Tongue I
(Static component)
--...-Tongue III
0.351--=...d--+- (Static component)
--o-- Tongue I
(static and dynamic) max.
0. 30 1-..I..L:.ol-~-fr- - - - Stepanoff

Volute
Tongue I III

AA/A 0.071 0.177

0 0.2 0 .4 0.6 0.8 1.0 1.2 1.4
Oi 0 17.8 36.1
Q / Qr

Fig. 4.12.3. Radial force coefficients in single volute casing pump [5]. "Tongue" refers to a tongue-shaped insert in the casing of the
pump in this experiment

0. 40

0. 35
o Double-volute casing
• Multi vane diffuser casing
0. 30 90"

0. 25

:..: 0. 20

0 . 15 0.792
0. 898
0.982
0. 10 1.036
1.095
1.171

0.05
0

~
p= ~ !'....
0 0. 2 0. 4 0. 6 0.8 1.0 1.2 1.4
Q / Qn
270"
Fig. 4.12.4. Radial force coefficients in double-volute casing Fig. 4.12.5. Radial force vectors in single-volute casing pump
pump and multivanes diffuser casing pump

The greater part of the hydraulic excitation force, volute and multivane diffuser casings [9]. Typical
which plays an important role in the excitation results are shown in Fig. 4.12.8. For single-volute-
forces acting on pumps, has not yet been clarified, type casings, the data obtained by Imaichi et al. is
and few research studies have been conducted. presented in Fig. 4.12.11. It should be noted that
Kanki et al. [9] have obtained the overall values of different models and ordinates are used in
the hydraulic excitation force, which varies as the Figs. 4.12.9, 4.12.10 and 4.12.11. In these figures, Q,
flow rate changes, from model pumps with double- represents the rated flow rate. They can only be

316
 4.12 Application to Pumps

270"
270"
Values noted in the figure show flow rate Values noted in the figure show flow rate
Fig. 4.12.6. Radial force vector in double-volute casing pump. Fig. 4.12.7. Radial force vector in multivane diffusor casing
Values noted in the figure show flow rate pump. Values noted in the figure show flow ra te

"'
0

"
DiffUSER CASING
i
NUMBER Of IMPELLER VANES : l : 7
u
,;:
: CENTER
,.
; ECCENTRICITY 0.20
~

100 ROTAT ING SPEED : 1500RPM ~

u
~ :::
~ 0.1 5 ~
g <.)
~
u Vl
:::>
:ii
"'~ "'
0
so
0.10
~
..
~

i5

~~~
<
"'
1 MINUIT l
"'
"'
0.05 ~
z:
0
;;;
z:

0
0
1171
20
UJI
40
ISOJ
60
1671 1831
80
11001
100
[1171

120 "i5

flOW RATE Q/Qr 1%1

"'0
:I:

i
DOUBLE VOLUTE CASING

NUMBER OF IMP(LLER VANES : l = 6 II

ECCENTRICITY : CENTER 0.20 ~

100 ROTATING SP[(O : I SOORPM ~

~
~ u
0 0.1$ ~
g< 0
u

..,..
u 1 MINUIT{ ~
V>
:ii :::>
z: "'
~
0 0.10
~

50 <
i5
<
"'
Vl
0.05 "'
w
~
z:
Q
"'
~

00 100 120
0 "i5
20 40 60

flOW RATE Q/Qr (%)

Fig. 4.1 2.8. Examples of unfiltered hydraulic exciting force

317
 4 Explanations

IIl
0
LOW CYCLE COMPONENT =£
40r---. .----------------------~ "
-<>DOUBLE VOLUTE+ 7 c.';
IMPELLER Z= 6 O.O II

+DOUBLE VOLUTE+
"- IMPELLER Z= 7 0.06 ~
6 w
~ 301-------\-----,-6-DIFFUSER CASING+ u
~ IMPELLER Z= 6 r;:
w 0.05 I'.
0 _..DIFFUSER CASING+ w
::::> 0
E-< IMPELLER Z= 7 u
:::1
p..
::a 20 1-----+-----\1----.------.-----r-----i
< 0 0.2 0. 4 0.6 0.8 1.2 1.4
0 Q! Q.
<
0
....l Fig. 4.12.11. Examples of dynamic hydraulic radial force mea-
sured in single-volute casing pump

Table 4.12.3

Cavitating bearing with Noncavitating bearing w ith

0 ol--~2LO___j40~~~~~~~~==~~~ low ambient (normal) pressure high ambient pressure
Wload
FLOW RATE Q/Qr ( %)

Fig. 4.12.9. Example of low cycle component

NZ COMPONENT
40 -o-DOUBLE VOLUTE+ IMPELLER ~
z= 6 0.07 k;
II

2n n
Negative pressure part
Water-lubricated bearing
Ex. Normal bearing
Oil film seal (compressor, etc.)

however, the ambient pressure of water-lubricated

bearings is higher than the atmospheric pressure,
depending on the suction and discharge pressures.
In these cases special bearing characteristics devel-
op, since the cavitating condition varies in the bear-
FLOW RATE Q/ Qr (%) ings. These are summarized in Table 4.1 2.3. That is,
the load-carrying capacity of full journal bearings
Fig. 4.12.10. Example of NZ component becomes about twice as large and the trace of the
shaft centre is perpendicular to the force if the am-
interpreted as examples, since the hydraulic excita- bient pressure is sufficiently high, as shown in
t~on force varies depending on the pump detail de- Figs. 4.12.12 and 4.12.13. As a result, spring and
sign. damping coefficients vary significantly, as shown in
Figs. 4.12.14 and 4.12.15.

c) Characteristics of Water-Lubricated Bearings d) Dynamic Characteristics of Annular Seals

Normally, the ambient pressure of bearings is ap- It is well known from experience that pump shaft
proximately atmospheric pressure. In some cases, systems have low sensitivity in vibration response,

318
 4.12 Application to Pumps
0 so
~~--
0.2 .,
Non
v
Cavi;/ ... 360" Brg.

\
"' 1/ , ~·

v ,'
0 0. 4 LID=l.O
·c:;
·.:
c
"
(.)
(.)
0. 6 ~' Cavi. ~
t:Ll ~.,

~~~., '' '"""'- ...... A

'
·~ r.;;: Cxx -:.,.;:;7 7

"'
0.8 /_ Kxy
~, ...
' v
c'
f...- ~
~
."-. ~Cyy /
........... ~"--
1.0
10 - 3 ro-' ro-• '' ~

s Ky~ ~ a, ~
Sommerfeld Number p.N(R /C)'
P,.
..)
~
.""'-,
Vb~ ~
lj.
Fig. 4.12.12. Eccentricity versus Sommerfeld number for 360°, Kxxlli.._L
L : D = 1 bearings \
Cxy ~Kyx

1
\ I
I
0.5 II
0 0.2 0. 4 0.6 0.8 1.0
Eccentricity <

Fig. 4.12.14. Dynamic characteristics of cavitating 360°

L:D = 1 bearing

100

a bearing centre
b rotation
.\ 360" Brg.
LID=l.O
c direction of eccentricity

Fig. 4.12.13. Journal centre loci in 360° L : D = 1 bearings ',\\ J

and in most cases no critical speed is observed on
the system. It became clear that the seal known as
a liner ring, which is located in the pump internally,
&\ [~ \.
\
"'
Cxx /
l/
Kxy-=

acts as a bearing and provides moderate support \. 1'. " / ""

' v
for the shaft [2 - 4, 17]. Therefore, it is necessary to
estimate the bearing effect of the liner rings accu- '""'
""-'....._ ~ yy
............ vf'-.....
/
rately when the vibration characteristics of a pump
shaft system are analysed. Research of this kind is ' ..... ... ~
currently being encouraged throughout the world.
In this section the well-known analytic studies con-
~
- ~-
Kyx
ducted by Black are described. Black derived an
approximation which can be used for determining
spring and damping coefficients and virtual mass 0 0. 2 0. 4 0. 6 0.8 1.0
Eccentricity <
by assuming that a liner ring can be considered as
a bearing with spiral flow, and confirmed the accu- Fig. 4.12.15. Dynamic characteristics of non-cavitating 360°
racy of the equation by experiments. L:D = 1 bearing

319
 4 Explanations

Black obtained the following equation for dy- flow (0 for zero rotation and 1 for high-
namic coefficients: speed rotation)
C =Radial clearance of annular seal
A
nRP FY
[Fx] L = Axial length of annular seal
Ra = Axial flow Reynolds number in annular
seal
= -[flo-tfl 2W2 T 2
- 2flt w T
t~ 1 WT
flo - 4fl2 w T
2 2] [X]
Y
R, = Circumferential flow Reynolds number in
annular seal
T= Average fluid passing time
_ [ flt T fl 2w T 2] [xy:] w = Angular velocity of shaft rotation
- fl2W T 2 - flt T

J [x]
R = Seal radius
_ [fl 20T
2 P = Seal differential pressure
fl2 0T2 ji . (2)
flo, flp fl 2 are given in Fig. 4.12.16 for inlet loss of
In Eq. (2), A is the friction coefficient for axial c; = 0.5, and P= 0, 0.5, 1.0.
flow, and therefore A varies with liner ring configu- The analysis mentioned above is performed on
ration and Reynolds number. As for annular seals, the basis of short bearing theory. The following
a semi-empirical equation proposed by Yamada approximations are therefore given to take account
et al. [16], in which the Reynolds number for cir- of seal length, which has not been sufficiently con-
cumferential flow is taken into account, is used to sidered hitherto [2]:
determine the coefficient:
A = 0.079 R;; 114 { 1 + (7 /8 R,/ RY} 318 (3) flo
(RL) = 1
flo
+ 0.28 (L /R) 2 '
a= A(C/L) = Parameter of resistance for axial
flow flt
(RL) = flt
1 + 0.23 (L/R) 2 ' (4)
P= (7 /8 R,/Ra) 2 / {1 + (7 /8 R,/RY} = Parameter
of the ratio of axial and circumferential
fl 2
(L)R=
fl2
1 + 0.06 (L/R) 2

-
0

0.8
~ I--
1.0
0. 7
0.6
v
v ..........
/3 t-- 4.12.5 Guidelines for Evaluation
0. 5 v 1-- 1--
0. 5

1-'o
0.4 / ./
v ..,_.. 1.0
Recently, various analysis programmes have been
I ~v -- - -- developed for shaft system vibration, and vibration
0.3
0. 2 ~, ,-
,
-- -- constant A
analysis can be performed easily. However, in
0.1 Ill pumps, few shaft systems exhibit high sensitivity to
vibration except some high-energy pumps, and
0
II' highly accurate vibration analysis need not neces-
2 3 4 5 6 7 8 10
9 11
sarily be applied. Moreover, in most cases vibration
I
on the top part of the motor is attributable to the
2.0 constant A vibration characteristics of the structure system. On
--p~
l 0 the other hand, in the development of a new high-
~ I-- - ~/3
1.5
,_,, f-- energy machine, vibration analysis of the shaft sys-
- ~ 1---
0.5
1.0 1.0 tem is indispensable and the method of study de-
~ ~ :::::: ~
F::::: 1--
0.5 ~
~
scribed in the previous section must be followed in
~ F" detail [10, 12].
0 2 3 4 5 6 7 8 9 10 11 Examples of calculation models and evaluation
of the results are described. In practical terms
1. 0 r: t--1.0
0.5~/3
judgement entails consideration not only of the cal-
culated result but of the actual operational data,
Ill +-- 0

- -- - ·r - -- -- since uncertain factors may be contained in the

0. 8

!-',
0 .6 ~ '::"""- applied data.
~...

0.4 ~ "' constant A

0. 2 II' a) Critical Speed

'I
In the case of horizontal pumps, analysis can be
0 2 3 4 5 6 7 8 10
9 11
performed by using a model (shown in Fig. 4.12.17)
Fig. 4.12.16. Parameters for dynamic coefficient in which the shaft, bearing and virtual mass are

320
 4.12 Application to Pumps

considered and sufficient stiffness is assumed for the

casing structure. As far as conventional types of
pumps are concerned, the critir-al speed hardly cre-
ates any problem, since sufficient damping is pro-
vided. Vibration characteristics of the shaft system
can be evaluated from the critical speed separation
margin shown in Fig. 4.12.18. In vertical installa-
tions, natural frequency analysis is required for the
Journal bearing Journal bearing casing structure, since few problems occur on the
shaft system. In this case, the damping effect is com-
Fig. 4.12.17. Vibration model of rotor system paratively small. Thus, the operating and critical
speed ranges must be separated from the natural
frequencies of the structure system by an adequate
margm.
API
Flexible shaft
0.2
l r API
Rigid
shaft
2.5
b) Vibration Response Characteristics

As important vibration responses of a pump shaft

system, the responses to unbalance and hydraulic
excitation force are the subjects of study. The un-
balance can be assumed on the basis of the allow-
able residual unbalance criteria for the balancing
operation. The hydraulic excitation force can be
estimated with reference to 4.12.4 b). However, it is
0.
CRITI CAL SPEED (% )
rather easier to consider the allowable excitation
OPERATING SPFFD force by obtaining a unit excitation force response.
A cannot be found; It is sufficiently conservative if a margin is
B low sensitivity; checked in the minimum clearance of the bearing
C moderately sensitive; and annular seals for ordinary exciting force.
D highly sensitive (needs field balancing);
E v e ry highly sensitive (must b e avoided)
c) Stability

Fig. 4.12.18. Critical speed separation margin with due consid- Stability can be a problem in a high-speed, high-
eration of damping effect pressure pump. The annular seals, the impeller, and

• "' .........bc=I.O%
2C
0 6 ····· ····D,=4 .6%

v ----- ,
?- 100
c. ......,, Diffuser
s 80 I
3 I
150
60
<::
s" 40
I
"u
"'
c.
V)

cs
20
0
l u
0~ 100
0 20 40 60 80 100 120 3
a Flow rate Q% ~
"sc.
E
·g.
"0
50
;.:::l

Fig. 4.12.19. a Vibration change as a function of flow rate.

b Influences of diffuser gap and working liquid temperature on
self-excited vibration
0
b Flowrate Q%
321
4 Explanations

0.5
FULL 0 4
JOURNAL 0.3

co
BEARING 0.2
0.1
..,. 0.05
.g
8
~ 0.025
·o.
6
Q 0
\
Unstable Stator mode \
-0.025 in y direction \
\
- 0. 05 ';;-----;-f;;;:;;;-----;:;f;;-;;----;=:-----'''-;;t;;;;--~~
0 1000 2000 3000 4000 5000
SHAFT SPEED(RPM)
a
0.1 10
Sommerfeld number ( S )
n, : First natural frequency of the system
Fig. 4.12.20. Stability chart of various bearings
(MOTOR)
1 1000
2" 500
the oil film bearing can all be sources of destabiliz-
~ 750
ing forces acting on the pump. Figure 4.12.19 shows 6
an example of a self-excited vibration that occurred 3
f-o
t'
in a vertical pump. This is a complex phenomenon z
~ 500
contributed to by the temperature of the pumping ::E
~
fluid and impeller radial clearance. The stability of u
<
the pump is generally evaluated from a stability ....l
ll.. 250
chart such as that shown in Fig. 4.12.20 for the Ul
5
bearings and by conducting complex eigenvalue
analysis. Furthermore, Fig. 4.12.21 shows examples o~lt-~-----~~---~~-_j
0 500 1000 1500
of measurement and analysis with consideration for (n,) ( 2 nJ
the oil whip developed at natural frequencies of the b SHAFT SPEED (RPM)
casing [15).
Fig. 4.12.21. a Complex eigenvalues calculated for vertical
pump. b Vibration change as a function of speed motor solo
operation

4.12.6 Concluding Remarks

Important points relating to the design and analysis References for Sect. 4.12
of a pump shaft system with journal bearings have
been described. These can be summarized as fol- [1) Makay, E. et al., Survey of Feed Pump Outages, EPPI
lows: Report, FP-754, (1978).
[2) Black, H.F. et al., Effects of High Pressure Ring Seals on
1) Bearings for hydraulic machines such as pumps Pump Rotor Vibrations, ASME Paper, No. 71-WA/FE-38.
must be designed with larger margins than for [3] Manabe, N. et al., A study on the Dynamic Characteristics
machines treating gas and electric machines. of Seals in Pump, Misubishi Juko Giho, Vol. 16, No.6, (in
Therefore, it is necessary to give sufficient con- Japanese) (1979).
[4] Takagi, M. et al., Analysis and Design of Centrifugal Pump
sideration to the actual data of similar machines, considering Rotor Dynamics, Proceedings of I. Mech. E,
since hydraulic excitation force is difficult to pre- (1980).
dict accurately. [5] Imaichi, K. et al., Experimental Study on the Radial Thrust
2) The following three items are important for fu- in Centrifugal Pump, Transaction of JSME Vol. 27, No.
ture works. 294 (1971) (in Japanese).
[6] Stepanoff, A.J., Centrifugal and Axial Flow Pumps. (1957)
• Hydraulic radial force John Wiley & Sons.
• Hydraulic excitation force [7) Agostinelli, A. et al., Trans. ASME, Ser. A, 82-1, (1960-4),
• Characteristics and effects of annular seals 120.

322
 4.12 Application to Pumps

[8] Biheller, J.J., Trans. ASME, Ser. A, 87-3, (1965-7), 319. [13] Shiraki, K . eta!., A New Vibration Criteria for High Speed
[9] Kanki, H. eta!. , Experimental Research on the Hydraulic Large Capacity Turbomachinery, Proceedings of the
Excitation Force on the Pump Shaft. ASME Paper, 81- Eighth Turbomachinery Symposium, (1979).
DET-71 , (1981). [14] Manabe, N., Vibration of High Speed Pump in Low Flow
[10] Matthew, C.E .K, Solution of the Subsunchronous Whirl Region Operation, Turbomachinery, Vol. 8, No. 10 (1980)
Problem in High Pressure Hydrogen Turbomachinery of (in Japanese).
Space Shuttle Main Engine. AIAA/SEA 14th Joint Propul- [15] Tomisawa, M. eta!., Vibration Analysis of Vertical Electric
sion Conference, (1978-7). Motor, Mitsubishi Electric Giho, Vol. 55, No.3 (1981) (in
[11] Kanki, H. eta!., Experimental study on the Dynamic Char- Japanese).
acteristics of Pump Annular Seals, Proceedings of Rotor [16] Yamada, Y. , Flow Resistance of Rotating Cylinder, Trans-
dynamic conference I. MechE. C297/84, (1984). action JSME, Vol. 27, No. 180 (1961) (in Japanese).
[12] Manabe, N. eta!., Development of Super High Pressure [17] Kaneko, S. eta!., Static and Dynamic Characteristics of
Centrifugal Pump for Water Jet Bedrock-Crushing, Mitsu- Annular Plane Seals, Proceedings of Rotor dynamics con-
bishi Heavy Industries, Ltd. Technical Review 1981. ference I. Mech E. C278/84, 1984.

323
 R. E. Roberson, R. Schwertassek

Dynamics of Multibody Systems

1988. 100 figures. XV, 460 pages. ISBN 3-540-17447-8
Multibody systems are the appropriate models for predicting and
evaluating performance of a variety of dynamical systems such as
spacecraft, vehicles, mechanisms, robots or biomechanical systems.
This book actresses the general problem of analysing the behaviour of
such multibody systems by digital simulation. This implies that
pre-computer analytical methods for deriving the system equations
must be replaced by systematic computer oriented formalisms, which
can be translated conveniently into efficient computer codes for
- generating the system equations based on simple user data describing
the system model
- solving those complex equations yielding results ready for design
evaluation.

Emphasis is on computer based derivation of the system equations

thus freeing the user from the time consuming and error-prone task of
developing equations of motion for various problems again and again.

J.P. Gupta

Fundamentals of Heat Exchanger

and Pressure Vessel Technology
1986. XV, 607 pages. ISBN 3-540-15619-4*
The unique question-and-answer approach of this volume facilitates
the location of information on specific heat exchanger design
problems. Because of the need to conserve energy and resources and
to operate plants safely and efficiently, heat exchangers must be
designed to perform optimally. This primer provides thorough and
simple answers to over 1200 questions about thermal and mechanical
designs, vibration analysis, and the operation and maintenance of heat
exchangers. Physical concepts explainyd in the text are illuminated by
more than 200 lavish figures. The boo~ is of interest to students in
advanced courses, process design engineers, project engineers in
evaluating the designs, operators who have trouble with heat exchang-
ers in the field, continuing education courses, and to those sitting
professional examinations. With over 540 literature references it also
serves as a useful source of research.
Springer-Verlag Berlin
Heidelberg New York London * Distribution rights for USA, Canada and Mexico: Hemisphere
Paris Tokyo Hong Kong Publishing Corporation
 F. F. Ling, C. H. T.Pan (Eds.)

Approaches to
Modeling of
Friction and
Wear
Proceedings of the Workshop on the Use of Surface
Deformation Models to Predict Tribology Behavior,
Columbia University in the City of New York,
December17-19, 1986
1988. 59 figures. XIII, 173 pages. ISBN 3-540-96656-0

From the contents: Invited Area Lectures. - Workshop

Summary.- Contributed Brief Position Papers.- Contri-
buted Brief Scientific Papers.

As the title suggests, this book is about approaches to

modeling of friction and wear. It is an outcome of a work-
shop which drew experts from industry and universities,
both in the United States and abroad. These experts from
multiple disciplines spent two days listening, lecturing and
debating in a workshop environment. As such this book
should be of interest to all who are concerned with tribol-
Springer-Verlag Berlin
Heidelberg New York London ogy at the research, advanced technology, and graduate
Paris Tokyo Hong Kong teaching level.