Erect of Shot Peening on Surface

Fatigue Life of Curbutized and

HardenedAISI 9310 Spur Gears
NASB Technical Paper 2047,1982 shot-peened and non-shot-peened gears were then tested to fatigue by
Dennis P, Townsend and Erwin V. Zaretsky surface pitting under identical test conditions. These test conditions
Lewis Research Center Cleveland, Ohio included a gear temperature of 350 K (170" F), a maximum Hertz
stress of 1.71x109N/m2 (248 000 psi), and a speed of 10 000 rpm.
Summary
Gear surface fatigue endurance tests were conducted on two Apparatus, Specimens, and Procedure
groups of 10 gears each of carburized and hardened AISI 9310 spur Gear Test Apparatus
gears manufactured from the same heat of material. Both groups The gear fatigue tests were performed in the NASA Lewis
were manufactured with standard ground tooth surfaces. The sec- Research Center's gear test apparatus (fig. 1). This test rig uses the
ond group was subjected to an additional shot-peening process on four-square principle of applying the test gear load so that the input
the gear tooth surfaces and root radius to produce a residual surface drive only needs to overcome the frictional losses in the system.
compressive stress. The gear pitch diameter was 8.89 cm (3.5 in.). A schematic of the test rig is shown in figure I(b). Oil pressure
Test conditions were a gear temperature of 350 K (170" F), a maxi- and leakage flow are supplied to the load vanes through a shaft seal.
mum Hertz stress of 1.71x109N/mZ (248 000 psi), and a speed of As the oil pressure is increased on the load vanes inside the slave
10 000 rpm. gear, torque is applied to the shaft. This torque is transmitted
The shot-peened gears exhibited pitting fatigue lives 1.6 times through the test gears back to the slave gear, where an equal but
the life of the standard gears without shot peening. Residual stress opposite torque is maintained by the oil pressure. This torque on
measurements and analysis indicate that the longer fatigue life is the the test gears, which depends on the hydraulic pressure applied to
result of the higher compressive stress produced by the shot peen- the load vanes, loads the gear teeth to the desired stress level. The
ing. The life for the shot-peened gear was calculated to be 1.5 times two identical test gears can be started under no load, and the load
that for the plain gear by using the measured residual stress differ- can be applied gradually,without changing the running track on the
elice fun the &mdard and shotpeened gears. The measured residual gear teeth.
stress for the shotpeened gears was much higher than that for the
standard gears.
Introduction
Shot peening has long been used as a method for improving the
bending strength of gear teeth (refs. 1 to 3). However, shot peening
has not been considered as a means of extending the surface fatigue
life of gears. In essence, shot peening induces a residual compres-
sive stress below the surface of the gear tooth. Studies of residual
stresses in rolling-element bearings have shown that increased
residual compressive stress will increase rolling-element (surface)
fatigue life (refs. 4 and 5 ) . There is always a need to improve the
surface fatigue life of aircraft gears, especially in helicopter and Test-lubricant out

V/STOL aircraft , ial Cutaway view.

The objectives of the research reported herein were (1) to
-Slave qear
investigate the effects of shot peening of gear teeth on the surface
fatigue life of standard ground, case-carburized, and hardened MSI
9310 spur gears, (2) to compare the life of shot-peened gears to rest'
gea n 4
that of non-shotpeened gears manufactured with the same material
and specifications, and (3) to determine the residual stress pro-
duced by shot peening and its effect on the surface fatigue life.
pressureJ
To accomplish these objectives, 20 spur gears were manufac- (b) Schematic diagram.
View A-A

tured from a consumable-electrode-vacuum-meltedsingle heat of CD-11124-1,

AISI 9310 material. Ten of these gears were shot peened after finish Figure I. MSA Lewis Research Center's gear fatigue test apparatus
grinding. The gear pitch diameter was 8.89 cm (3.5 in.). Both the
 Separate lubrication systems are provided for the test gears and Tabh III
the main gearbox. The two lubrication systems are separated at the Heat Treatmentfor A N 5)310
gearbox shafts by pressurized labyrinth seals. Nitrogen is the seal Process I Temperature
gas. The test gear lubricant filtered through a 5-pm-nominal fiber-
glass filter. The test lubricant can be heated electrically with an
immersion heater. The tempergture of the heater skin is controlled
to prevent overheating the test lubricant. A vibration transducer Preheat i n a i r
mounted on the gearbox is used to automatically shut off the test rig Carburize
when a gear surface fatigue occurs. The gearbox is also automatically A i r c o o l t o room
shut off if there is a loss of oil flow to either the main gearbox or temperature
Copper p l a t e a l l over
the test gears, if the test gear oil overheats, or if there is a loss of
Reheat
seal gas pressurization. A i r c o o l t o room
The belt-driven test rig can be operated at several fixed speeds temperature
by changing pulleys. The operating speed for the tests reported Austenitize
herein was 10 000 rpm. O i 1 quench
Subzero c o o l
Test Materials
D o u b l e temper
The test gears were manufactured from consumable-electrode- Finish grind
vacuum-melted (CVM) AISI 9310 steel from the same heat of mater- Stress r e l i e v e
id. Both sets of gears were case hardened to a case hardness of
Rockwell C 58 and a case depth of 0.97 mrn (0.038 in.). The nomi- case carburized and heat treated in accordance with the heat treat-
nal core hardness was Rockwell C 40. One set of the gears was shot ment schedule of table In. Figure 2 is a photomicrograph of an
peened, after finish grinding, on the tooth root and the tooth profile etched and polished gear tooth surface showing the case micro-
according to the specifications given in table I. The chemical com- structure of the AISI 9310 material.
position of the material is given in table IZ. Both sets of gears were
Test Gears
Table I. Shot Peening Speczjkation
Dimensions of the test gears are given in table N AU gears have
a nominal surface finish on the tooth face of 0.406 pm (16 pin.)
r Shot-peened rms and a standard 20" involute profile with tip relief. Tip relief was
0.0013 cm (0.0005 in.), starting at the highest point of single-tooth
contact. Surface traces of the standard gear and the &ot-peened
gear are shown in figure 3.
Test Lubricant
- - --- --
hpecitication
- -.
. . . . . . . . , . . MIL-S-131658 AU the gears were lubricated with a single batch of synthetic
paraffinic oil. The physical properties of this lubricant are summa-
BPS F W 4409
S h o t s i z e . . . . . . . . . . . . . . . . . 070 rized in table V. Five percent of an extreme-pressure additive, desig-
Shot t y p e . . . . . . . . . . , , . . Cast s t e e l nated Lubrizol5002 (partial chemical analysis given in table V), was
I n t e n s i t y ( h e i g h t o f ,411~ien . . . . 0.18 t o 0.23 added to the lubricant.
s t r i p , t y p e A ) , mm ( i n . ) (0.007 t o 0.009)
Coverage ( s i d e s and r o o t . . . . . . . . . . 200
Test Procedure
only), percent
After the test gears were cleaned to remove the preservative, they
-- -- were assembled on the test rig. The 0.635-cm (0.25-in.) wide test
Table I1 gears were run in an o%et condition with a 0.30-cm (0.12-in.) tooth-
Nominal Chemical Composition of CVMAISI9310 Gear Material surface overlap to give a load surface on the gear face of 0.28 cm
(0.11 in.), thereby allowing for the edge radius of the gear teeth.
E 1 ement / Cornpos it ion, If both faces of the gears were tested, four fatigue tests could be run
wt% for each set of gears. AU tests were run in at a pitch-line load of
1225 Nlcm (700 Win) for 1 hour, which gave a maximum Hertz
C 0.10
stress of 0.756 x lo9N / m ~ l l l 0 0 0psi). The load was then
Mn .6 3
Si .2 7 increased to 5784 Nlcm (3305 Win), which gave a pitch-line
Ni 3.22 maximum Hertz stress of 1.71x109N/m2 (248 000 psi). At this
Cr 1.2 1 pitch-line load the tooth root bending stress would be 0.21 x lo9
M0 .12 N/m2 (30 000 psi) if plain bending were assumed. However,
Cu .13 because there was an offset load, an additional stress was imposed
P .005 on the tooth bending stress. Combining the bending and torsional
S .005 moments gave a maximum stress of 0.26 x lo9N/m2 (37 000 psi).
L J
 ! c ) 5 hor-peeneo C J W ~ case. !dl Shot-peenefl year core.

Figure 2. Photomicrographs of case and corefor standard and shot-peened spur gears
This bending stress does not include the effects of tip relief, which film thickness calculation, that the gear temperature at the pitch line
would also increase the bending stress. was equal to the outlet oil temperature and that the inlet oil temper-
Operating the test gears at 10 000 rpm gave a pitch-line velocity ature to the contact zone was equal to the gear temperature, even
of 46.55 mfsec (9163 fumin). Lubricant was supplied to the inlet though the inlet oil temperature was considerably lower. It is possi-
mesh at 800 cm3//minat 319 +_6K (llG"*lOo !?),The lubricant out- ble that the gear surface temperature was even higher than the outlet
let temperature was nearly constant at 350 2 3 K (170" +5" P). The oil temperature, especially at the end points of sliding contact. The
tests ran continuously (24 hr/day) until they were automatically shut EHD film thickness for these conditions was computed to be 0.33
down by the vibration detection transducer, located on the gearbox pm (13 pin.), which gave an initial ratio of film thickness to corn-
adjacent to the test gears. The lubricant circulated through a 5-prn posite surface roughness h/o of 0.55 at the 1,71x109-N/rn"248
fiberglass Uter to remove wear particles. After each test the lubri- 000-psi) pitch-line maximum Hertz stress.
cant and the filter element were discarded. Inlet and outlet oil tern-
Results and Discussion
peratures were continuously recorded on a strip-chart recorder.
Gears manufactured from CVM AISI 9310 material were tested
The pitch-line elastohydrodynamic (EID) film thickness was
in pairs until failure or for 500 hours. One-half of the gears were
calculated by the method of reference 6. It was assumed, for this
 shot peened on the tooth root and profile. Nineteen tests were run Table IC( Spur Gear Data. Gear To~eranceper ASMA c h s , 12.
with standard-finish ground test gears, and 24 tests were run with
Number of teeth . . . . . . . . . . . . , .A
,
standard finish ground gears that had been shot peened. Test results Diametral pitch .... .. .. ... ... . 8
were analyzed by considering the life of each pair of gears as a Circular pitch, cm (in,) . . . 0.9975 (0.3927)
system. Whole depth, cm (in.) . . . . . . 0.762 (0.300)
Surface (pitting) fatigue results for the standard-finishAISI Addendum, cm (in.) . . . . . . . 0.318 (0,125)
9310 gears are shown in figure 4(a). These data were analyzed by Chordal tooth thickness . . . . . 0.485 (0.191)
the method of reference 7. The 10- and 50-percent fatigue lives (reference), cm (in.)
were 18.8 x lo6and 46.1 x lo6stress cycles (31.3 and 76.8 hr), Pressure angle, deg
Pitch diameter, cm (in.)
. . . .. .. .. .. ,8.890
. . . . . 2(
(3.500,
respectively These results are summarized in table VI. The failure
index (i.e, the number of fatigue failures out of the number of sets Outside diameter, cm (in.) . . . 9.525 (3.750;
tested) was 18 out of 18. A typical fatigue spall is shown in figure Root fillet, cm (in.) . . . . . . 0.102 to 0.15;
(0.04 to 0.06)
5 (b). A cross section of a typical fatigue spall is shown in figure
5 (a). The surface pitting failure occurs slightly below the pitch line
Measurement over pins, . . . . . 9.603 to 9.63C
cm (in.) (3.7807 to 3.7915)
in the area of highest Hertz stress and is of subsurface origin. P i n diameter, cm (in.) . . . . . 0.539 (0.216)
Pitting fatigue life results for the gears that were shot peened Backlash reference, cm (in.) . . 0.0254 (0.010)
are shown in figure 4(b). The failure index was 24 out of 24. A typi- Tip relief, cm (in.) . . . . . 0.001 to 0.0015
cal fatigue spa11 for the shot-peened gears is shown in figure 6(a). A
cross section of a typical fatigue spall for the shot-peened gears is rooth width, cm (in.) . . . . .(0..OOO4
. 0.635to 0 (0.25)
.OO06)

shown in figure 6(b). The 10- and 50-percent surface pitting fatigue
lives were 30.1 x 106and 67.5 x 106stress cycles (50.3 and 112.6
hr), respectively. These results are summarized in table VI. The
shot-peened gears exhibited a 10-percent fatigue life 1.6 times that
of the standard ground AISI 9310 gears. The confidence number for
the difference in life was 83 percent. The mean life ratio for the
shotpeened over the standard AISI 9310 gears was 1.5, with a c o d -
dence number of 98 percent. The confidence number indicates the
percentage of time the relative lives of the material will occur in the
same order. The 90-percent confidence bands for the standard
la) Standard gear.
ground and shot-peened AISl9310 test gears are shown in figure 4.
A coride~icenumber of 45 percent is equvaient to a 2 0 conti-
dence level.
It is well known that shot peening produces residual sub-
surface stresses in steel in addition to the residual stresses pro-
duced by case carburizing, hardening, and grinding. It was theo-
rized that the additional residual stresses induced by shot peening Figure 3. Surfacefinish of standard
ground and shot-peened gears.
should account for the increased life of the shot-peened gears.
Therefore two shot-peened and untested gear teeth and two stan-
Ib) Shot-peened gear.
dard ground and untested gear teeth were subjected to X-ray dif-
fraction residual stress measurements to determine the magnitude peened or tested. The high compressive stress on the surface of the
of these residual stresses. Residual stress measurements were made gear tooth is the result of grinding and has a very shallow depth that
near the pitch point at the surface and at nominal subsurface depths has very little effect on the surface durability of the gear. The lower
of 5, 13, 25, 76, 127, and 254 pm (0.2,0.5, 1.0, 3, 5, and 10 mil). \compressive stress, which has much greater depth, is from the case
Material was removed for subsurface measurement by elec- carburizing and hardening of the gear tooth surface. This compres-
tropolishing in a sulphuric-phosphoric-chromicacid electrolyte in sive residual stress has a definite beneficial effect on surface fatigue
order to minimize possible alteration of the subsurface residual and bending fatigue life.
stress distribution as a result of material removal. All data obtained Figure 7(b) contains two plots of corrected X-ray diffraction
as a function of depth were corrected for the effects of the penetra- residual stress measurements as a function of depth below the
tion of the radiation employed for residual stress measurement into surface in the ground and shot-peened AISI 9310 gear teeth that had
the sub-surface stress gradient and for stress relaxation, which not been tested. The high grinding compressive stress on the surface
occurred as a result of material removal. The method used for the was reduced. A hook in the curve shows a high compressive stress
X-ray stress measurements and the calibration procedures used are 1.3 pm (0.5 mil) below the surface as a result of the shot peening.
described in references 8 and 9. The compressive stress at greater depths below the surface was also
Figure 7(a) shows two corrected X-ray diffraction residual increased as a result of the shot peening. It is the increased com-
stress measurements as a function of depth below the surface for pressive stress at the greater depths that has the major effect on the
the standard ground MSI 9310 gear teeth that had not been shot surface fatigue life. The depth to the maximum shear stress for the
 load conditions reported herein was 178 pm (7 mil). Table l?
Bgure 7(c) contain plots of the average of the two X-ray resid- Properties of Synthetic Paraflnic Oil
--
ual stress measurements as a function of depth below the surface for
both the standard ground and shot-peened gears. This figure shows
Additive .......... . . a~ubrizol5002
Kinematic viscosity, cm2/sec (cS) at-
the average increase in the residual compressive stress due to shot 244 K (-20" F) ..... .. 2500x10-~'(2500)
peening, At the maximum shear stress depth of 178 pm (7 mil) the 311 K (100' F) ....... 31.6x10-~ (31.6)
average residual compressive stress was increased from 0.186 x lo9 372 K (210' F) ........ 5.7x10-~ (5.7)
N/mZ(27 000 psi) in the standard ground AISI 9310 gear to 0.26 x 477 K (400" F) ...... .
, 2.0x10-~ (2.0)
lo9N/m"37 700 psi) in the ground and shot-peened AISI 9310 Flashpoint, K (OF) . .. ... .... 508 (455)
gear. From equation (A1 1) , taken from the analysis given in the Fire point, K ( O F ) .......... 533 (500)
appendix for maximum shear stress and residual stress, Pour point, K (OF) ... .... ... 219 (-65)
Specific gravity ....... ... .. 0.8285
Vapor pressure at 311 K (100' F), ..... 0.1
mm Hg (or torr)
Specific heat at 311 K (100' F), 676 (0.523) ..
where J/kg K (Btu/lb OF)

R = 7.62 mm (0.3 in.)

S,, = 1.71 x 109 N/m2 (248 000 psi)

Therefore for peened gears, in SI units,

in U.S. customary units,

= - 121 080 Number of cycles

(a) 5tandard gears. (b) Shot-peened gears
And for standard gears, in SI units, Pigure 4. Comparison of surface @itti& f a t e e lives of standard
ground and shot-peened carburized and hardened CTWAISI9310 steel
spur gears. Speed, 10 000 rpm; lubricant, syntheticparaflnic oil;.gear
temperature,350 K (170"P);maximum Hertz stress, 1.7 x 109 N/mZ
(248 0000 psi).

Table Irl.
in U.S. customary units, Fatigue results with AlSI 9310 Standard and Shot-Peener Test Gears
AND SHOT-PEENED TEST GEARS

Gears 10-Percent 50-Percent Slope F a i l u r e Confidence

life, life. indexa number ,b
cycles cycles percent

Standard --
Shot peened 83

a I n d i c a t e s numbers o f f a i l u r e s o u t of t o t a l number o f t e s t s .
b r o b a b i l i t y , expressed as a percentage, t h a t t h e 10-percent
l i f e w i t h t h e b a s e l i n e A1S1 9310 gears i s e i t h e r l e s s than,
o r g r e a t e r than, t h a t o f t h e p a r t i c u l a r l o t o f gears being
considered.
 l a ) iyelcal falljjiiY ~ p d l l
Figure 6 Patgue spa11for shotpeenedgeal:
 The surface fatigue life from reference 10 for gears is inversely
proportional to the maximum shear stress to the ninth power. The Residual stress, psi
calculated life ratio from measured residual stress is therefore

Residual stress. Nlm

This calculated ratio of the fatigue life of the shot-peened gears

to that of the standard gear compares favorably with the experimen-
tal fatigue life ratio of 1.6.
Summary of Results
Gear surface fatigue endurance tests were conducted on two
groups of carburized and hardened AISI 9310 steel spur gears o Gear A
A Gear B
manufactured from the same heat of material. Both groups were
manufactured with a standard ground tooth surface. One group was
subjected to an additional shot-peening process on the gear tooth
surface and root radius to produce a residual compressive surface
stress. The gear pitch diameter was 8.89 cm (3.5 in.). Test conditions 28 L
were a gear temperature of 350 K (170" F), a maximum Hertz stress l a ) Standard qears.

of 1.71 x lo9N/mZ (248 000 psi), and a speed of 10 000 rpm.

The lubricant was a synthetic paraffinic oil with an additive
package.
The following results were obtained:
1.The 10-percent surface (pitting) fatigue life of the shot-peened
gears was 1.6 times that for the standard test gears that were not 0 Gear C
A Gear D
shot peened. This was shown to be statistically significant.
2. The calculated 10-percent surface (pitting) fatigue life for the
shot-peened gears determined from measured residual subsur- \
face stress was 1.5 times that for the standard gears that were not
shot peened.
3. Measured residiial stresses for the stm&rird m! shot-neened
Y (bj ~ n o i - p e e n e dgears.
0
gears show an increase of 40 percent for the maximum shear
stress in addition to a 350 percent increase at a depth of 13.0 pm 4-
(0.5 mil).
8-
Lewis Research Center National Aeronautics and Space
12 - 0 Standard gears
Administration Cleveland, Ohio, March 5, 1982 A Shot-peened qears
Appendix - Derivation of Residual Stress Effect on Maximum 16 - \
Shear Stress \
20 - I
It is well known that classical rolling-element fatigue begins in
the subsurface zone of maximum shear stress (refs. 11 and 12).
Therefore, to determine the effect of residual stress on rolling- 28 L
element fatigue in gears, it is necessary to analyze the effect of resid- ( c ) Average principal residual stress for standard and
shot-peened gears.
ual stress on the maximum shear stress below the surface. The max-
imum shear stress at any point in a stressed volume below a Figure % Principal residual stress as afunction of depth below the
rolling line-contact load is surface of carburizecl, hardened, ground and untestedAISI9310 steel
(.mu,, -2 1
'SZ - SY' (A])
spur gear teeth.

where Sz is the principal compressive stress in a direction

normal to the contact area, Sy is the principal compressive stress
parallel to the direction of rolhg (ref, 13), and Slx, is the principal
stress normal to the direction of rolling.
 For rollers or gear teeth loaded staticdly, the maximum theo-
retical shear stress occurs in the y-z plane since the stress in they,
o r rolling, direction is less than the stress in the x direction. where Sqcan be either compressive or tensile. When gears are shot
Therefore the maximum shear stress is peened, the residual stress is compressive and therefore reduces the
1 maximum shear stress. Since the rolling-element fatigue life of gears
~ m , = 3 'SZ - SY' (A3 is inversely proportional to the maximum shear stress to the ninth
If the residual stresses are equal in the x and y directions, for power (ref. l4),
the line contact in the y-z plane the maximum shear stress including
the residual stress is

where Sry is the residual stress in the y direction and is positive for
tensile stress and negative for compressive stress. From reference 13 From equation (A4), where (T,,)~ = Tm, - 1/2 (Sry),
for line contact of rollers
b
T ~ , = - 0.30025 (A51
where is the half width of the Hertzian contact,
using Ilife
. ratio of L1 and L2

and A for a contact of two rollers of the same material is

When the residual stress developed by the shot peening the

gear teeth is known, the change in life produced shot peening can
If the rollers are of the same radius, be determined from equation (A14).
References
1. Moore, H. E: Shot Peening and the Fatigue of Metals. American
Foundry Equipment Co., 1944.
2. Straub,J. C.: Shot Peeningin Gear Design. AGMA Paper 109.13June
where 1964.
3. Valentine, K. B.: Recrystallization as a Measurement of Relative
PN normal load, N (lb) Shot Peening Intensities. Am. Soc. Met. Trans. Q., vol. 40, 1948 pp.
S
,, maximum Hertz stress, N/m2 (psi) 420-434.
R1,R2 radius of curvature of the two rollers, m (in.) 4. Zaretsky, Erwin V,; et al.: Effects of Component Differential: Hardness on
6 Poisson's ratio Residual Stress and Rolling-Contact Fatigue. NASA TN D-2664, 1965.
E Young's modulus, N/m2 (psi) 5. Zaretsky, E. y; Parker, R. J.; and Anderson, W. J: Component Hardness
Differences and Their Effect on Bearing Fatigue. J. Lub. Technol.,vol.
Substituting equations (A6) a n d (A8) into equation (A5) 89, no. 1,Jan. 1967, pp. 47-62. 6. Dowson, D.; and Higginson, G. R.:
f o r T,, results in Elasto-HydrodynamicLubrication. Pergamon Press, 1966, p. 96.
7. Johnson, Leonard G.: The Statistical Treatment of Fatigue
Experiments. Elsevier Pub. Co., 1964.
8. Christenson, A. L., ed.: Measurement of Stress by X-Ray SAE HS-182,
Aug. 1971.
9. Prevey, Paul S.: Method of Determining the Elastic Properties of Alloys in
If equation (A9) is substituted into equation (A4), Selected Crystallographic Directions for X-Ray Defraction Residual Stress
Measurements. Adv, X-Ray Anal., vol. 20, 1977,pp. 345-354.
10. Townsend, D. P.; Coy, J. J,; and Zaretsky, E. V.: Experimental and
Analytical Load Life Relation for AISI 9310 Steel Spur Gears. Journal of
Mechanical Design, Trans. ASME, vol. 100, no. 1,Jan. 1978, pp. 54-60.
11.Jones, A. B.: Metallographic Observations of Ball Bearing Fatigue
F o r steel gears E = 207 x 109 N/m2 (30 x 106 psi) and Phenomena. Symposium on Testing of Bearings, ASTM, 1947, pp. 35-
48; discussion, pp. 49-52.
6 = 0.30; therefore equation (A10) becomes for SI units 12. Carter, T. L.; et al.: Investigation of Factors Governing Fatigue Life with
the Rolling-Contact Fatigue Spin Rig. Am. Soc. Lubr. Eng. Trans.,vol. 1,
no. 1, Apr. 1958, pp. 23-32.
( T ~ , ) ~= - 21.74 x log--------PN - 1 (A1 la) 13.Jones, A. B.: New Departure - Analysis of Stress and Deflections. Vol. I,
LSm,R z~" New Departure, Div. Gen. Motors Corp., 1946, p. 22.
14. Lundberg, G.; and Palmgren, A.: Dynamic Capacity of Rolling
a n d for U.S. customary units Bearings. Acta. Polytech. Scand., Mech. Eng. Ser., vol. 1, no, 3, 1947.