Professional Documents
Culture Documents
Thin-Section
Bearings Are
Created Equal:
KAYDON NEW CAPACITY CALCULATIONS
(Courtesy Kaydon Corp. Bearings Division)
Robert Roos and Scott Hansen
Management Summary
American Bearing Manufacturers Association (ABMA) Standard 9 and ISO 281 give equations for calculating
the basic dynamic radial load rating for ball bearings. These equations are based on a number of assumptions, many
of which are not valid for thin-section bearings. (Thin-section bearings are described in ABMA standard 26.2.)
Nevertheless, many thin-section bearing catalogs report load ratings based on these equations.
Kaydon has developed a new method for calculating the dynamic radial load rating for thin-section ball bearings.
The new method uses the contact stress and the number of stress-cycles-per-revolution to calculate the capacity. The
new numbers are based on five years of actual test results. These equations can also be used to calculate the dynamic
radial load rating for four-point contact ball bearings, which are not covered in ABMA standard 9 or ISO 281.
Introduction
The year was 1944 and the problem that needed to be
solved required the development of a new type of bearing.
The U.S. Army Air Corp contacted Kaydon Engineering
Corporation to engineer and build a light-weight, thinsection ball bearing for a ball gun turret to be used on an
aircraft. This bearing became the inspiration for a catalog
line of thin-section ball bearings known today as REALISLIM (REALI-SLIM is a registered trademark of Kaydon
Corporation).
In many applications, shafts supported by bearings are
lightly loaded. Shaft position with respect to the housing or
other components is critical. These designs do not need big,
heavy bearings and can be supported adequately by thinner
bearing races manufactured to close tolerances. Kaydon had
www.powertransmission.com
continued
february 2011
powertransmissionengineering
15
300
Medium series
bearings
Extra light
series bearings
200
Extreme light
series bearings
100
Thin-section
bearings
with 25.4-mm
radial cross section
200
400
600
800
1000
Bearing Bore Diameter (mm)
Bore
diameter
Radial
cross section
Ball
diameter
TEST
REPORT
BASE
PART
NUMBER
MANUFACTURER
RP545
RP487
RP582
RP487
RP534
RP529
RP584
KC040CP0
KC040CP0
KC040CP0
KC040CP0
NC040CP0
NC040CP0
NC040CP0
Kaydon
Kaydon
Competitor ( A )
Competitor ( B )
Kaydon
Kaydon
Competitor ( C )
16
powertransmissionengineering
B10
EQUIVALENT CATALOG
LIFE
CAPACITY CAPACITY
(hours)
(lbs)
(lbs)
76.0
1,059
880
71.2
1,036
880
23.2
713
1,290
12.5
580
884
59.9
978
880
120.4
1,234
880
19.1
668
1,290
february 2011
www.powertransmission.com
CATALOG
CAPACITY
COMPARISON
ACTUAL
B10 FATIGUE
PERFORMANCE
+20%
+18%
-45%
-34%
+11%
+40%
-48%
+73%
+64%
-83%
-72%
+36%
+174%
-87%
L 10 =
Eg. 1
Equation
Eg. 1 These
Cr
L 10 =
Cr
3
106 million revolutions ...................... ISO 281 par 5.3.1
P
3r
106 million revolutions ...................... ISO 281 par 5.3.1
two standardsP give the following equations for calculating the Basic Dynamic
Load Rating (C r): r
0.41 10 / 3
1.72
ISO1.39
281 p a r 5.1 0.41
0.3
Equation 2 E g . 2
ri 2 ro D b
1
(1
)
2 ri
Eg. 3
f c 4 . 1 1 1 . 04 0.7 2 / 3 01..78 2 / 3 1.8
1
/
3
C
=
3
.
647
b
f
(
i
cos
)
Z
D
ballsize
(Db) (>1 1 inch
C r = brm fc ( i cos m)1 c Z
Drb 2for
(Db)
1 inch.
Dbbsizefor
)
2 ri D b
o ri ball
Eg. 2
ISO 281 p a r 5.1
DbCos
where:
Palmgren Table 3.2
b f ( i cos ) 0.7 Z 2 / 3 D 1.8 for ball size
0.3 > 1 inch
Cr = 3 . 647
(Db)
0.41
m Dc
b
0.41 10 / 3
1.72
39
0
.
3
1
.
p 1
ri 2 ro D b
(1
)
2 ri
Equation
Eg. 33
f c 4 . 1 1 1 . 04
1/ 3
ro 2 ri D b
1
(1 )
2 ri D b
where:
Equation
Eg.44
DbCos
Dp
f c 38 . 20 1 1 . 04
1
1
1.72
fi 2 fo 1
fo 2 fi 1
0.41 10 / 3
0.3
0.3
(1
(1 )
0.41 10 / 3
0.3
1.39
1/ 3
2 fi
2 fi 1
0.41
0.41
1.72
0.3
1.39
fi 2 fo 1
1
(1
)
2 fi
Eg. 4
f c 38 . 20 1 1 . 04
1/ 3
f o 2 fbi 1 = Material factor for
f i 1(b = 1.3)
1
(1 contemporary
)
2 steels
million revolutions.
m
m
These standards also define the L10 bearing fatigue life as: f
= Geometry factor from tables
c
The basic rating life in millions of revolutions for 90% reliability. i
Ref. Harris eq. 18.106
= Number of rows
The L10 life is then calculated from the load rating (Cr) a
= Contact angle
using Equation 1. Please note that the basic dynamic radial Z
= Number of balls per row
load rating, as defined in these two standards, is not the maxi- D = Ball diameter
b
mum operating load for the bearing; it is simply a constant P
= Applied radial load
r
used in the life equation. It is often greater than the static
radial load rating (Cor). Loading the bearing beyond the static
Both standards provide tables for the geometry factor (fc).
rating will cause permanent deformation, or brinelling. The material factor (b ) was added to ISO 281 in 1990. It
m
Both standards state:
equals 1.3 for radial and angular contact ball bearings made
The life formula gives satisfactory results for a broad range of contemporary steels. The 1990 version of ABMA standard
of bearing loads. (However), the user should consult the bearing 9 does not include this factor. However, the tables for factor f
c
manufacturer to establish the applicability of the life formula in are 1.3 times higher than the ISO 281 tables. Therefore, the
cases where (the applied load) Pr exceeds (the static capacity) Co or two standards give exactly the same capacity.
(one-half the dynamic rating) 0.5 Cr whichever is smaller.
The factor fc can also be calculated using the equation
below:
(See above for Equation 1)
(See above for Equation 3)
These two standards give the following equations for calculating the basic dynamic load rating (Cr):
(See above for Equation 2)
where:
L10 = The basic life rating in millions revolutions for
90% reliability
Cr = Basic dynamic radial load rating
where:
The units in this equation are kg and mm. For radial and
angular contact ball bearings l= 0.95. For 4-point contact
bearings l= 0.90. If we let fi and fo equal the inner and outer
curvature ratio (rx/Db) and convert the units to N and mm,
this equation can be rewritten as:
(See above for Equation 4)
www.powertransmission.com
continued
february 2011
powertransmissionengineering
17
Part No.
KAA10CL0 KA020CP0 KC040CP0 KF080CP0 KA120CP0 KG120CP0 KG400CP0
Kaydon Catalog
150
320
880
4,100
980
8,510
18,310
Catalog 300
ISO/ABMA
1990 fc Tables:
558
1,012
2,321
8,081
1,904
14,133
21,630
1978 fc Tables:
430
778
1,785
6,216
1,465
10,872
16,638
Calculated fc:
11,766
1,316
18,006
1990:
379
701
1,656
6,363
1978:
291
539
1,274
4,895
1,015
9,050
13,851
(1)
INA
558
1,012
2,316
8,094
14,164
21,582
NSK(1)
558
1,012
2,316
8,094
14,164
RBC(2)
300
560
1,290
5,140
1,060
10,690
16,230
SKF(1)
555
1,009
2,338
8,049
1,915
14,029
21,493
INA, NSK and SKF all use the ABMA/ISO ratings taking fc directly from the tables without adjusting for
the curvature ratio or diametral clearance.
(2) RBC numbers appear to adjust for curvature, but not clearance.
(1)
18
powertransmissionengineering
february 2011
www.powertransmission.com
Eg. 5Equation 5
C
Cr
1
6
r
10
Lh
rev
min
Pr
Pr
60 hr N min
16, 667
( hrs )
Life (hrs.)
Eg. 6
life (Eq. 1).
150
tomer to use a quick (Cr/Pr)3 equation to estimate
C
N
L
h
r
100
The ABMA/ISO equations are not adequate for this purpose
50
because they dont consider the curvature or clearance in the
0
basic dynamic load rating. Therefore, alternate methods must
-0.0010
-0.0005
0.0000
0.0005
0.0010
0.0015
0.0020
0.0025
Diametral Clearance (Inch)
be used to calculate the published ratings for these bearings.
Test Method
diametral
clearance.
2 Life vs. Diametral Clearance
Preparation: The test parameters (i.e., load and speed) Figure 3Life vs.Figure
must first be calculated. These are selected to cause the
bearings to fail by fatigue in a reasonable time period (typically less than 100 hours), while not subjecting the bearing
to excessive loads. Therefore the testing speed is set to 1,780
rpm.
The desired radial test load is calculated from the life
equation (Eq. 1). To get the life in hours, simply divide the
life in revolutions by the rotating speed as shown in Equation
5.
(See above for Equation 5)
0.0030
Where:
Lh is the L10 life in hours under the test conditions
Cr is the dynamic load rating for the bearing (from a
catalog or calculation)
Figure 4Test arrangement.
P r is the applied
radial load
3
3
FIGURE 1
C r 16, 667
Cr
1
ing witness
marks to indicate orientation in housings, and
6
in RPMs.
bearing under test,
L h Nis the speed
10 ofthe
(
hrs
)
then mounted on a driven spindle with the axis horizontal as
rev
min
N shown
in Figure 4.
Pr
Pr
60 hr N min
We can then rearrange Equation 5 and solve for the
A steel load strap is then attached to the outer race and
applied radial load Pr. For an L10 life of approximately 50 connected to an air cylinder arranged to apply radial load to
hours, the radial load should be a little less than 60% of the bearing. The centerline of the air cylinder, the steel strap
the rated capacity (Eq. 6). Keep in mind that the L10 life is and the lower tooling are aligned in the same plane as the
based on 90% reliability. This means that 90% of the test bearing centerline.
samples are expected to exceed the 50 hours. For the Kaydon
The force exerted by the load cylinder is measured by a
KC040CP0 bearing, a radial load of 525 lbs. was selected.
portable load cell and digital indicator. The air pressure to the
load cylinder is adjusted to produce the desired radial load
1/ 3
16, 667
on the bearing. Note: In calculating the total radial load on the
Pr
(6)
bearing, the weight of the tooling is also considered.
february 2011
powertransmissionengineering
19
ing is connected and turned on. The cooling water to the oil
heat exchangers is turned on. Air to the load cylinder is set
to off.
The drive motor is jogged to ensure that no components
are loose and that there is no unwanted interference between
the rotating and stationary parts. The motor is then turned
on and the reading (in hours and tenths)Eg.
of7the elapsed time
meter is noted. All test notes are recorded on the test data
collection sheet.
Oil flow to the bearing is adjusted to be as high as possible
without causing excessive foaming or leakage from the bearing housing. Oil flowing to each bearing should not exceed a
temperature of 130 F. Air to the load cylinder is then turned
on.
The baseline radial vibration is measured. The auto shutoff system is then set for this baseline vibration level. The
automatic shut-down occurs when vibration increases to
150% of the baseline level.
Stopping test. When bearing failure is suspected, either
powertransmissionengineering
february 2011
1073 (7)
16667
16667
These values are much greater than the old Kaydon ratings of 880 lbs. for both. However, they still fall short of the
ABMA/ISO ratings for radial bearings calculated using the
fc values from the tables. (Neither ABMA or ISO publish
ratings for four-point contact ball bearings.) The main reason
for the apparent discrepancy is the poor assumptions these
standards make regarding race curvature and diametrical
clearance. Based on the results of these tests, it is apparent
that a new method is required for calculated the dynamic
1 bearings.
8
1
radial capacity for thin-section
.7
L10 1
3 P 10 / f 3 P 100 3 Q max Zi Cos
r
r
6
6
Equations
S
10
10 Kaydon
f
New
www.powertransmission.com
Z
Db
(164 . 84 )( 35 )( 1 )
Q max Z cos
9
1
Eg.
f
cos .......................................................
5 . 3769
=
=
ref. Harris= eq.
25.14
ref.SHarris
Z 2 Db D
.......................................................
eq.
25.14
(9)
P
1073Q
Pr
1
.9 f
cos .......................................................
cos
(164 . 84 )( 35 )( 1 )
Z
max
ref.
DP
= 5 . 3769 (12)
S = Harris eq. 25.14=
2
1073
Pr
Equation for ball distribution factor (S): The relationship
between max ball load (Qmax) and applied radial load (P r) is
calculated using the Stribeck equation (Eq. 10). The constant
The number of stress-cycles-per-revolution is calculated
S is the load distribution factor. For bearings with zero diam- using Equation 9 and equals 16.75 for the KC040CP0, as
etral clearance, S equals 4.37. For bearings with nominal shown in Equation 13. This means that there are 16.75 milclearance, an approximate value of 5 is often used for S.
lion stress cycles in one million revolutions.
Q max Z cos
10
cosEg.
Pr Q
35
Z
Db
. 1875
Z
cos
S
...ref.
Harris
eqs.
6.23/6.24
max
. 10 ...ref. PHarris
f
1
cos
Eq.
13
eqs.
6.23/6.24
1
= 16 . 75
(13)
r
S
DP
2
2 Z 4 . 375
...ref. Harris eqs. 6.23/6.24
35
Db
. 18
(10)
1
f =
cos =
Eq. 13
1
is= 262,500
C r = a KC040CP0 outer
1073 =psi.
1
or Q max
Eq. 14
S
f
5S..3769
16
(11)
B)
The
ball
load
distribution
For
the KC040CP0
100
3. 75
Eq. 11
S = .027D p + 5.260
Q maxfactor
ZCos
Qm
assumed
C r = load of 1,073 lbs., and an
= 1073 =
bearing
Eq. 14under an applied
S inch, the
f load
5 . 37
ball
distriCalculating capacity. It takes four steps to calculate the clearance after installation of .0016
new dynamic radial capacity:
bution factor S was calculated to be 5.3769 (Eq.12.).
Step 1: The first step in calculating the capacity is to
For the KG400CP0, an outer race stress of 262,500 psi
calculate the normal ball load (Qmax) that gives an outer race corresponds to a normal ball load of 801.23 lbs. The stressmean contact stress of 262,500 psi. This is calculated using cycles-per-revolution (f) for this bearing equal 60.75. As
the standard stress equations. This gives the ball load capacity shown in Equation 8, the capacity is a function of the load
distribution factor, which is in turn a function of the diamfor 108 stress cycles.
Step 2: The next step is to calculate the ball distribution etral clearance and the applied load. Equation 8 was then
solved using an iterative solution for the capacity Cr and the
factor S from Equation 11.
Step 3: The number of stress-cycles-per-revolution is then load distribution factor S. For an assumed clearance after
(164 . 84 )( 35 )( 1 )
Q max Z cos
calculated
Equation 9.=
installation of .0072 inch and a radial load of 18,307 lbs., S
= 5 . 3769
S =using
1073
Prmaximum normal
Step 4: Once the
ball load, distribution was then calculated using Equation 10 to be 6.3562.
factor and number of stress-cycles-per-revolution
areN known,
factor S varies with the installed
80 stress
1780 distribution
L 10
1/ 3
The
1/ 3
Eg. 7
P
525
1073with the pitch diameter. The equa
the radial capacity can be calculatedCusing
8.
clearance,
which increases
r Equation
16667
for
continued
164.84 lbs. Solving Equation 10 for S gives a load distribution
Equation
Eg. 8 8
Equation
Eq. 14 14
8
1
10 / f
L10 3
C r Pr
6 Pr
6
10
10
Cr =
S
f
1
1
1
3
100 3
Q max Zi .7 Cos 100 3
Pr
f
S
f
1
1
3
Q max 35 100 3
= 1073 =
5
.
3769
16
.
75
www.powertransmission.com
february 2011
powertransmissionengineering
21
D p1 D p
41 4 . 375
y = mx + b where:
m =
S1 S 2
D p1 D p
=
2
6 . 3562 5 . 3769
41 4 . 375
= . 02674
L 10
Cr
=
16 , 667
1417
=
525
N
16 , 667
= 184 (hrs.)
1780
bearing and is lower than the life for the individual rows. This already factored into the maximum allowable stress level,
3
factor is consistent with both the ISO/ABMA
calculations as which
was3established
by testing.)
Cr
16 , 667
1417
16 , 667
well as the old Kaydon method.
E)
Four-point
contact
bearing.
Using
the new capacity cal(hrs.)
L 10 =
= 184
D) The stress factor (l). The
8),
radial capacity
of a KC040XP0 (4-point)
culation
the
Ptables
in bothN ABMA
525 (Eq.
1780
Standard 9 and ISO 281 show multiple columns for factor bearing is 1,417 lbs. If we plug this number into the life equafc. The first column is titled Single Row Radial Contact tion (Eq. 5), the L10 life under a test load of 525 lbs. at 1,780
and Single and Double Row Angular Contact. The second rpm is 184 hours (Eq. 16). In Kaydon testing, the L10 life of
column is titled Double Row Radial Contact and has a a standard KC040XP0 was actually 189.3 hours. Therefore,
lower value for fc. The reason for this is described on page there is excellent correlation between the theoretical and
81 and table 3.3 of Ball and Roller Bearing Engineering measured capacity.
by A. Palmgren (1959). Palmgren shows a l of .95 for
single-row radial bearings, and for single- and double-row
(See above for Equation 16)
angular contact bearings. He also shows a l=.90 for doublerow radial bearings. It is described as a stress factor. This
Conclusions
factor allows for uneven load sharing between the two rows
The new Kaydon capacity equations are based on the
of contact. Therefore, Kaydon has chosen a de-rating factor maximum contact stress and the number of stress cycles per
of (l=.90/.95) for four-point contact (X-type) ball bearings. revolution. They consider the actual curvature of the races.
(The 0.95 factor for radial and angular contact bearings is They also consider the diametral clearance. The new capacities are supported by actual test data. The new radial capacity
for radial (C-type) bearings ranges from 36% higher in the
Robert Roos has been a senior prodKAA10CL0 to no change in the KG400CP0. The radial
uct engineer with Kaydon Corporation
capacity of four-point contact (X-type) bearings increases
Bearings Division for 10 years. He
from 31% to 77% depending on the bearing size. The new
earned a bachelors degree in mechanicapacities are still lower than the ABMA/ISO numbers,
cal engineering from the University
which are based on very poor assumptions. The new Kaydon
of Michigan and a masters degree in
equations provide a more accurate estimate of actual bearing
mechanichal engineering from Western
life when used in the basic L10 = (Cr/Pr)3 equation.
Michigan University. Roos is a liThese equations apply to Kaydon Catalog 300 bearings
censed professional engineer in the state
with standard clearance only. Preload and clearance have a
of Michigan. He holds two patents and
significant influence on bearing life. Clearance after installais a member of the ASTM F34 comtion will vary depending on fits and shaft/housing materials.
mittee for rolling element bearings.
The Catalog 300 capacity should be used for an initial bearing selection only. Life can then be confirmed using more
advanced Kaydon programs. These programs use the ISO/
Scott A. Hansen is vice president of
ABMA capacity, but also take the curvature and clearance/
manufacturing planning for Kaydon
preload into consideration.
Corporation Bearings Division. His
30 years with Kaydon include posiReferences:
tions in product design, manufactur1. Load Ratings and Fatigue Life for Ball Bearings, ANSI/AFBMA
ing and operations. He holds a bachStd. 9, (1978).
2. Load Ratings and Fatigue Life for Ball Bearings, ANSI/AFBMA
elors degree from Western Michigan
Std. 9, (1990).
University and two patentsfor a
3.Rolling BearingsDynamic Load Ratings and Rating Life, ISO 281,
wind turbine pitch bearing and a robot
(1990).
pivot arm assembly for semiconductor
4. T. Harris, Rolling Bearing Analysis, 3rd Ed., J Wiley & Sons USA
wafer handling. Hansen belongs to the
(1991).
ASTM F34 rolling element bearings
5. Palmgren, A. Ball and Roller Bearing Engineering, 3rd Ed., SKF
group and the ABMA engineering specification group.
Industries, Inc., Philadelphia, PA. (1959).
22
powertransmissionengineering
february 2011
www.powertransmission.com