Effe,ct 0£ Shot Peening on Surface

Fatigue Life of Carburized and

Hardened AI'SI '93,10 'Spur Gears
by
Dennis P. Townsend and ,Erwin V. Zaretsky
NASA Lewis Research Center
Cleveland, Ohio

ting fatigue lives 1.6 times the life of the means o.f extending the surface fatigu
Su:mmary standard gears without shot peening. Re- life of gears. In essence .•shot. peening in-
Gear surface fatigue endurance tests sidual stress measurements and analysis duces a residual compressive 'Stress below
were conducted on ,two groups of 10 indicate that. the longer fatigue life is the the surface of the gear tooth, Studies of
gears each of carburizedand hardened result. of the higher compressive stress residual stresses in rolling-element bear-
Al:SI 9310 spur gears manufactured from produced by the shot peening. The life ings haveshown that increased residual
the same heat of material, Both groups to,r the shot-peened gear was calculated compressive stress will increase rolling-
were manulactued with standard ground to be 1.5 times that for the plain gear by element (surface) fatigue life.!4-S1There is
tooth surfaces. The second group was using the measured residual stress dif~ always a need. to impmve th urfao fa-
subjected Ito an additional shot-peening lerenoe for the standard and shot-peened tigue life of aircraft gears, especially in
process on the gear tooth surfaces and gears, The measured residual stress for helicopter and V ISTOLaircraft.
root radius, to produce a residual surface the shot-peened gears was much higher The objectives of the research reported
compressive stress. The gear pitch di- than that for the standard gears, herein were (1) to investigate the eHects
ameter was 8.89 em (3.5 in.). Test con- of shot peening ofgearteeth on. the sur-
ditions were a gear temperature of 350 Introduction face fatigue life of standard ground, ease-
K (170°F), a maximum Hertz stress of Shot peening has long been used as a earburized, and hardened A[SI 9310 spur
1.71 X 109 N/m2 (248000 psi)' and a method. for improving the bending gears. (2) to compare the lif . of shot-
speed of 10000 rpm, strength of gear teeth.n·ll However, shot peened gears to that of non-shot-peened
The shot-peened gears exhibited pit- peening has not been considered as a gears manufactured with the am ma-
terial and specifications. and (3) to deter-
mine the residual stress produced by sh t
peening and its effect on the surface
f.atigue life.
A.UTHORS:
honored by the ASME when he was made TO' accomplish these objectives, 20
DR. ERWIN V. ZARETSKY has been a Fellow of the Society, Dr. Zaretsky has a spur gears were manufactured from a
member of tile research staff of the NASA BSME from the lIlinois lnstiruie of r:echflology consumable-electrode~va,cuum~melt d
Lewis Research Center since 1957 where he is lind a Doctor of Law Degree from Clevel'cmd
Chief £ngitleer for Structures. He was head of State University, sing} heat of AlSI 9310' material. T enot
tile NASA Bearing, Gearing and Transmission MR. DENNIS P: TOWNSEND is the these gears were shot peened after finish
Section and waS manager of the NASA Manager of the Gear Research Program for grinding. The gear pitch diameter was
Helicopter Transmission System Technology Advanced Aircraft Applications at the NASA 8.89 em (3,5 in.). Both th shot-peened
Program. Zaretsky is author or co-author 01 Lewis Research (imler, He directs and CDIlducts and non-shot-peened gears were then
over 125 technical papers. books. and articles research on gear lubrication. gear mater:ials,
tested to fatigue by surface pitting under
in tile field of bearings. gears and lubrication and gear design" He has authored or co·
technology. He is Q Licensed Professional IWthored over 54 NASA technical sodety identical test conditions. These test con-
Engineer in Ohio and has been admitted to the publicatiollS. Townsend holds a B5 in ditions included a gear temperature of
Bars of Ohio and lIlinois as well as the Federal Mechanical Engineering from the University of 350 K (170°F), a maximum Hertz stress
Bar. Zaretsky has chaired several national West VA He served with the ASME Design Q
of 1.71 XIO N/m2 (248 000 psi). and a
commiUe,es for the A5ME and was a Director Division as ChJlirman of the Power Tnmsmi.s-
speed of 10 000 rpm.
of the Cl'roeland Society of Professional sion and Gearing Committee from 1978 to
Enginurs and is a lormer Chairman of the 1983. ,and he was Papers Review Chairman
Cleveland Section of A5ME. The .LubricaJion and Associate Editor of the journal for two .Apparatus, Specimens. and Proudure
Division of the ASME honored him in 1972. years. He is now a member of the ASME Geu Test Appuatus
.by bestowing UpOtl him their D.istinguished Design Engineering Division. Executive
Service Award. In 1979. he was further Committee. The gear fat:igue tests were performed

January/February 1986 .29

 TABLE I!.
SHOT-PEENING,
SPECI!FICATIION

r Shot-peened
/ area
/

- Loading vane

Specification ... , .. , .. , ... MIL·S·131658

,/ IBPS FW 4409
,
/ Shot size .........•... , , .. , ..... , .070
L Test-lubricant outlet
tempe ratu re measure- Shot type ' Ca.sl steel
ment locatian
Intensity (height of Almen .. ,0.18 to 0.23
tal Cutaway view.
strip, type Al, mm(in.) (0.007 to 0.009)
Coverage (sides and root ... , .. , .. , .200
-slave gear
only), percent

;Iave"<jear

~!~s-<,l~l3=~JE~.~
__
§.~~.;e§~g·
~ TABLE III.
NOMINAL CH'EMICAL

-t--loqd ~~~4Lt
pressure"
view 1\-1\
torque
COMPOSITION OF CVM AISI
9310 G'EAR MATERIAL

{bl Schematic di~ram.

CO-IIIZ4-b
iElement Composition,
wtOfo

Fig, 1- NASA Lewis Research Center's gear fatigue test apparatus. C 0.10
Mn .63
Si .27
in the NASA Lewis Research Center's vided for the test gears and the main Ni 3.22
gea:rtest apparatus (Fig. 1). This test rig gearbox. The two lubrication systems are Cr 1.21'
uses the four-square principle of apply- separated at the gearbox shafts by pres- .112
Mo
ing the test gear load, so that the input surized labyrinth seals. Nitrogen is the
Cu .13
drive only needs to overcome the fric- seal gas, The test gear lubricant is filtered
p .005
tional losses in the system. through a 5-,um-nominal fiberglass filter.
A schematic of the test rig is shown in The test lubricant Can be heated electric- S .005
Fig. l(b), Oil pressure and leakage flow ally with an immersion heater. The tem-
we supplied to the load vanes through perature of the heater skin is controlled
a. shaft seal. As the oil pressure is in- to prevent overheating the test lubricant
creased on the load vanes inside the slave A vibration transducer mounted on from consumable-electrode-vacuum-
gear, torque is applied to the shaft. This the gearbox is used to automatically shut melted (CVM) AISI 9310 steel from the
torque is transmitted through the test off the test rig when a gear surface fatigue same heat of material. Both sets of gears
gears back to the slave gear, where an occurs, The gearbox is also automatically were case hardened to a case hardness of
equal, but opposite torque is maintained shut off if there is a loss of oil flow to Rockwell CSB and a case depth of 0.97
by the oil pressure. This torque on the either the main gearbox or the test gears, mm (0.038 in.). The nomina] core hard-
test gears, which depends on the hydrau- if the test gear oil overheats, or if there ness was Rockwell C40. One set of the
lic pressure applied to the load vanes, isa loss of seal gas pressurization. gears was shot peened, after finish grind-
loads the gear teeth to the desired stress The belt-driven test rig can be operated ing, on the tooth root and the tooth pro-
level. The two identical test gears can be at several fixed speeds by changing file according to the specifications given
started under no load, and the load can pulleys, The operating speed for the tests in Table 1. The chemical composition of
be applied gradually. without changing reported herein was 10000 rpm. the material is given in Table II. Both sets
the running track on the gear teeth. Test Materials of gears were case carburized and heat
Separate lubrication systems are pro- The test gears were manufactured treated in accordance with the heat treat-

30 Gear fechnology
Fig. l-Photomicrographs of case and core for standard and shot-peened spur gears.
TABLE ill. .HEAT TREATIiIU :FOiI i\JSJ 9310
:Re
Step P"o-cess T"""lture Hilt,
61 Ihr
.. .. .. :61 K, Of

'flO
'.r. .•
I'
.~' ' 58
2
'Preheat In
CnbYl'he
&11"

I
~... -'"'
1112 16SO 8
57 '-I. to 'I"OCIii
(001
I
U.'pcrlty.·e
S6 4 tQPPfI" pllte In 0'11',1"
S Reheat 922 lZOO' 2.~
6 Air cool te rr~
t""rltur,
1 Austcnlt.he 1117 15SO 2.5 I

a I 011 quench
9 Subzero cool 19:) .120 3.5
10 ,
Double ttllOtr 4SO 3SO 2 IKh
.... _-
11 II F1nhh g"lnd
I. St...n rll'l¥e
I

,4SO 350
··----1I
2

TAkE I¥. - SPUlI Iil:AR DAtA

[GeAr tolerance per A'sw, clln 12.]

N"'er of teeth • • .• • • • • • • • • • • • • 28
DI_trll pl·tch ••••••••••••• , •• B
ClrcuW' pitch, 01( (In.' ••• , 0.9975 (0.3927)
(bl St!ndilrd gear core. Whole depL~.cl!! (In.) •• • • 0.7'62 [0.300)
Addendum, cm (In.) ••••• 0.318 [0 •.125)
Chordal tooth thickness. 0.485 (0.'190
(refen~e). ao (In.)
PressUl'e Ingle., deg ••••••••••.••• zo
Pitch' dl_t". ClIo ,[In.) •••• 8,.m (l.!.OO,)
Outside dl_ttr •. CII (In.) ••• 9.525 (3.1501
Root 11,lht. ell (In., •••• , • 0.102 to 0.152
(O.OCto 0.06)
Heuurem.mt over pIns, • ,.••.• 9.60] to 9.630
ell [I.d (3.19:)7 to 3.7915)
,PI.di_ter. til (In., ••••• D.549 (0.216)
hc·klash refe,.enee. til (In.) •• 0.02S4 10.01DI
TIp rillef, til [In.) ••••• , 0.001 to 0.0015
[0.0004 to Q.OIlO6)
Tooth width, til (In.) •• , ..... 0.635 (0.25)

Ti\III.E V•. - PROPERTIES Of S'~THETI C

PAlWFI Nle 01 L

Addlthe • • • • • • • • • • •• '\ubrhol 5002

K'lneotatlc viscosity, uJ.1r..K (es) tt-
244 K (·ZO· F) • • • • • • • 2500110-2 (2500)
311 K (100· f) •••.•••• 3I,61IO,2,(ll.6)
372 , (210' F) • _ • • • • •• S.•7dO-2 (5,,7J;
417 ,K (·400· F) • •• • • • •• 2,0010-2 (2.0)
FlunpQlnt. K (OF J • • • 5{)8 (.55)
re I Shot-peened gear case. 1!l1 SM!-peened gea.r core.
Fire pQl(1t, K (OF) ...... '.... " ,. 533 :(500):
POUT pOint. ~ (On ••.• " •• '. " •• 219 (-65):
SpKltic gravity' ••• ,. • • • • • • •• 0'.8285
ment schedule of Table III. Fig. 2 is a ¥ apor ,preuure It HI K (tOO· f I" ••.•• •. O. I
photomicrograph of an etched and pol- '... ~9 [or torr)
Specific ".at I,t 311 ~ [100· fI, • " 676 (0.521)
ished gear tooth surface showing the ease JI kg K (Btullb·F)
microstructure of the AJSI 9310 material.
'AddH he. lubrho 1 5002 (5 '011.).;, content,
sr od<l''II •• : pl>osl>-l>orus. 0.6 wU; 5ul,'u.,
~est Gears 18.5 wtl.

Dimenions of the test gears are given

in Table IV. All gears have a nominal batch of synthetic par.aff.inic oil. The
surface finish on the tooth face of 0 ..406 physical properties of this lubricant are
jUm (16 jUin.) rms and a standard 20° in- summarized in Table V. Five percent of
volute profile with tip relief. Tipl'elief an extreme-pressure additive, designated
was 0.0013 em. (0.0005 in.), starting at Lubriznl 5002 (partial chemical analysis
the highest point of single-tooth contact. given in Table V)" was added 'to the
Surface traces of the standard gear and lubricant.
the shot-peened gear ar-eshown in Fig. 3.

Test Lubricant Test Procedure

Fig" J - Surface finish of standard ground and
shot-peened. gears. AI] gears were lubricated with a single After the test gears were cleaned to

Jonuary/Fe'bruary 1986 3:1'

 T~au .1 •• Fm;1/E IlESI.US "!TH ~ISI 9310 STAIIlIOlID remove the preservative, they were
IHD SHllT.FW.EC TEST ;!/,RS
assembled on the test rig .. The 0.635 em
"Ifr, ID-Pell'(.ffIt 5O·P't'i"Ctftt :S'lope hil .....' CDftUCltI'lCI (0.25 in.) wide test gears were run in an
111,_ Hit,,, .nat;]l1 'i!i~,,1b
[,J'C. '11 etc1it1i plf'cnt
offset condition with a 0.30 em (O.U in.)
n •..r...o 19.~ Ol.~ 1.1 11'/11
tooth-surface overlap to give a load sur-
~""l _nod JO 61 l.l· Itnt l).
face on the gear Iace of 0.28 cm (0.11
Ill'IdtCltft n..., .. g'
".t h!""i OI!!~, ~of 't01;.,
it..,.r' of w;;tli. in.), thereby, allowing for the edge radius
trprDhll,-tllt7. u",",u'" ,U I poet'Citlll.llI. tl'ilt tlIt 1iQ...Mf'C!lftt
11,. .1\- eM .15.111_ ,li.1U 1)10 IIIit'S. 11 .1t:Jwto lnl tlwlt of the gear teeth. If both faces oE the
OJ' vut.1"r !tn'" tlHt • ., tM, ,..tit1ill.ar' tot 01 11_$ Mtll
c.,~dIIrtd.. 1

- gears were tested. four fatigue tests could

fig. 4 - Comparison of surface (pi tring) fal igue

lives of standard ground and shot-peened car-
burized and hardened C\IM: AlSI 9310 steel spur
gears. Speed, 10 000 rpm; lubricant. sYl1thetic
paraffinic oil: gear temperature, 350 K (170° F);
maximum Hertz stress, 1.7 X 10Q N/ml (248
000 psi).

be run for each set of gears ..All tests were

run in ata pitch-line load of 1225 N/crn
('700 lb/in.) for 1 hour, which gave a
maximum Hertz stress of 0.756 X 109
N/m2 (111 000 psi). The load was then
increased to 5784 N/cm (3305 lb/tn.),
which gave a pitch-line maximum Hertz
stress oJ 1.71 X 109 N/m2 (248 000 psi).
At this pitch-line load the tooth root
bending stress would be 0.21 X 109
N/m2 (30 000 psi), if plain bending were
assumed. However, because there was an
offset load, an additional stress was im-
posed on the tooth bending stress. Com-
bining the bending and torsional mo-
ments gave a maximum stress of 0.26 X
109 N/m2 (37 000 psi). This bending
stress does not include the effects of tip
relief. which woulda1so increase the
bending stress.
Operating the test gears at 10000 rpm
gave a pitch-line velocity of 46.55 mlsec
(9163 It/min). Lubricant was supplied to
the inlet mesh at 800 cm3/min at 319 ±
6 K (116" ± 10"F). The lubricant outlet
temperature was nearly constant at 350
±3 K (170° ±5°P). The tests ran con-
tinuously (24 hr I day) until they were
automatically shut down by the vibra-
tion detection transducer, located. on the
gearbox adjacent to the test gears. The
lubricant circulated through a 5-l"m fiber-
glass filter to remove wear particles.
Aftereach test, the lubricant and the fil-
terelement were discarded. Wet and out-
let oil temperatures were continuously
recorded on a strip-chart recorder.
The pi.tch-lineelastohydrodynamic

32 Gear Technology
(EHD) film thickness was calculated by hr), respectively. These resultsare sum- for the standard ground and hot-peened
the m.ethod ,of Reference 6. It was as- marized in Table VI. The shot-peened AlSI 931!Otest gears are shown in Fig. 41.
sumed, for this film thickness calculation. gears exhibited a 10-percent fatigue life A confidence number of 95 percent is
that the gear temperature at the pitch line of 1..6 times that of the standard ground equival:enlt to a 2u ,conJidence level.
was equal to the ,outlet oil temperanire AJS19310 gears ..The confidence number It is well known that shot peening pro-
and that the inlet oil, temperature to the for the difference in life was 83 percent. duoes residual subsurface stresses in,' teel
contact zone was equal to the gear Item- The mean life ratio tor the shot-peened in. addition to Ithe residual stresses pre-
perature, even though the inlet oil over the standard A1SI 9310 gears was duced by case carburiziag, hard rung.
tempera'tu:re was considerably lower. It 1.5, with a confidence number of 98 per- and grinding. It was theorized Ithat. the
is possible that the gear surface 'temper- cent. The conlidenee number ilndica'~es additional residual stresses induced by
ature was even higher than the outlet oil 'the percentage of time the re]ative lives shot peening should account for the in-
temperature, especially a't the end points of the material win occur in the same creased Me of the shot-peened gears.
,.of sliding contact, The EHD film order. T:he '90-percent confidence bands Therefore. two shot-peened and untested
thickness for 'these conditions was com-
puted to be 0.33 lID' (131lin.)', which gave
an initial ratio 01 film thickness to com-
posite surface roughness. hi oof 0.55 at
~he 1.71 X 109 N/m2 (248 000 psi)
.... '1:031. ,NILES BEAIR' BRINDER
pitch-line maximum Hertz stress.
___,~DEL· ,ZSTZ 63IJC3 INI sroCK!
Results and Discussion
Gears manufactured from ,~ AlSI
9310 material were tested in pairs until
failure or for 500 hours ..,One-half of the
gears were shot peened on the tooth root
and profile. Nineteen tests were run with
standard-f:inish ground test gears, and 24
Itests were run with standard-finish
ground gears that had been shot peened.
Test resuUs were analyzed by consider-
ing the life of each pair of gears as a
system.
Surface (pitting) fatigue results for 'the
standard-finish AlSI 9310 gears are
shown in Fig. 4(a). These data were ana-
lyzed by the method of Relerence Z The
10- and SO-pe::rcentfatigue lives were 18.8
6
X 10 and 46.1 X 100 stress cycles (31,3
and 76.8 hr), respecti.veJy. These results
are summarized in Table VI. The :failure
index (i.e., the number of fatigue failures
out of the number of sets tested) was 18
out of 18. A typical fatigue spall is shown
in Fig. 5(b). A cross section of a typica] SPECIFICATIONS:
fatigue span is shown in :Fig. 5(a)'. The om. de 'dl mete. ,Imax. • .." n. 29.5 IMaxmum, helix _Aale .. deg. •
surface pitting failure occurs slightly IRool'clrcle dlam t r, rnA. .n. 2 Stroke, length" .., , , , ".. In. 8.81
I umber of laetht, lmax. • , ..•. /I 1!.a IDoub' ,rarilltroke::
below the pitch line in the area of highest
INumber of leeth', mJn. .." •. # 1,2 (InflnHely war.) .,,""" 1'/mln!. 7W15
Hertz stress and is of subsurface origin. IDlmutnI IPHcn. mIn~ " •• ,," D'.R 11.2.7 IlaXJlrnu:m IIble Iludl "" lb... 1801
Pi.tting fatigue life results for the gears ID:I.metnll pHch, Imx~". "." ID.P.: .2.12 -ble bore .,,"""""""" In. 3.5
that were shot peened are shown in :Fig.
4(b). The failure index was 24 out of 24.
A typical fatigue spall for the shot-
peened gears is shown in Fig. 6(a), A
cross section of a. typical fatigu.e spall for
'the shot-peened gears is shown in. fig.
6(b). The 10- and 5O-percentsurface pit-
ting fatig-ue lives were 30.1 X Hi'and
67.S X lei' stress cycles (50.3 and 112.6

JanualiV/February 11986 II
 Residual stress, psi

o -40 -80 -120

I I l
Residual stress, Nlm J
-40 -61)

o Gear A
.e. Gear B

Fig. 5 - Fatigue spall for standard ground gear.

lal Stand;!rd ge4rs.

o Cea:r C
tJ. Cear 0

(bl Shot-peened gears.

o Standard gurs
Fig. 6 - Fatigue spall for shot-peened gear. .e. Snot-peened gea rs

gear teeth and two standard ground and untested gear teeth
were subjected to X-ray diffraction residual stress measure-
ments 'to determine the magnitude of these residual stresses.
Residual stress measurements were made near the pitch point
lei Average principal residual stress for staodard ;tnd
at the surface and at nominal subsurface depths of 5, 13,25, shot-peened gears.
76, 127, and 254 ftm (0.2, 0.5, 1.0, 3, 5, and 10 mil).
Material was removed for subsurface measurement by elec-
tropolishing in a sulphuric-phosphoric-chromic acid elec- Fig. 7 - Principal residual stress as a function of depth below the surface of
trolyte in order to minimize possible alteration of the sub- carburtzed. hardened, ground and untested AIS19310 steel spur gear teeth.

surface residual stress distribution as a result of material

removal. All data obtained as a function of depth were COf-
rected for the effects of the penetration of the radiation
employed for residual stress measurement into the subsur-
face stress gradient and for stress relaxation, which occurred which has much greater depth, is from the case carburizing
as a result of materia] removal. The method used for the and hardening of the gear tooth surface. This compressive
X-ray stress measurements and the calibration procedures residual stress has a definite beneficial effect on surface fatigue
used are described in References 8 and 9. and bending fatigue life.
Fig. 7(a) shows two corrected X-ray diffraction residual Fig. 7(b) contains two plots of corrected X-ray diffraction
stress measurements as a function of depth below the sur- residua! stress measurements as a function of depth below
face for the standard ground AISI 9310 gear teeth that had the surface in the ground and shot-peened AISI 9310 gear
not been shot peened or tested. The high compressive stress teeth that had not been tested. The high grinding compressive
on the surface of the gear tooth is the result of grinding and stress on the surface was reduced. A hook in the curve shows
has a very shallow depth that has very little effect on the sur- a high compressive stress 1.3 ftm (0.5 mil) below the surface
face durability of the gear. The lower compressive stress, asa result of the shot peening. The compressive stress at

34 Gear Technology
greater depths below the surface was also increased as a result calculated life ratio from measured residual stress is therefore
of the shot peening. It is the increased compressive stress at
the greater depths that has the major effect on the surface ~;; [(Tmax)r.s]9 = (0.872)( l(9). 9= 1 S
fa.tigue life. The depth to the maximum shear stress for the
LlOs . (Tmax)rp .. 0.835)( 109 . •
load conditions reported herein. was 1'78 ~m (7 mil).
Fig. 7(c) contain plots of the average of the two X-ray This calculated ratio of the fatigue life of the shot-peened gears
residual stress measurements as ill function of depth below to that of the standard gear compares favorably with the ex-
the surface for both the standard ground and shotpeened perimental fatigue life ratio of 1.6.
gears. This fig. shows the average increase .in the residual
compressive stress due to shot peening. At the maximum
shear stress depth of 1781-1m(7 mil) the average residual com- Summa.ry of Results
pressive stress was increased from 0.186 X 109 N/m2 Gear surface fatigue endurance tests were conducted on
,(27 (100 psi) in the standard ground AlSI 9310 gear to 10.26 X two groups of carburized and hardened A]S] 9310 steel. spur
109 N/m2 (37 700 psi) In the ground and shot-peened A[5I gears manufactured from the same heat ,of material. Both
931.0 gear. From equation (All), taken Irom the analysis groups were manufactured with a standard ground tooth sur-
given In the appendix for maximum shear stress and residual Iace. One group was subjected to an additional shot-peening
stress, process on the gear tooth suriace and roo'! radius to produce
a residual compressive surface stress. The gear pitch diameter
(TmaJ,;;;;; - 3. 15)( 106 (LS:uR) - ~ s.; was 8.89 em (3.5 in.]. Test conditions were a gear temperature
of 350 K (170" F), a maximum Hertzstres of 1..71 x 109
where
N/m2 (248000 psi), and a speed of 10000 rpm. The lubri-
cant was a synthetic paraffinic oil with an additive package.
7.62 mm (0.3 in.)
R ;;;;;; The following results were obtained:
1. The IO-percent surface (pitting) fatigue life of the shot-
Smax = L 71 x [09 N/m2 (248 000 psi) peened gears was 1.6 times that for the standard test
gears that were not shot peened. This was shown to

i..'" = $78 375 N/m (3,305 lb/in)

be statistically significant.
2. The calculated 100percent surface (pining) fatigue life
for the shot-peened gears determined :from measured
Therefore for peened gears, in SI units, residual subsurface stress was 1.5 times that for the stan-
dard gears that were not shot peen d.
I _ '. '9 578 3·75 3. Measu» d residual stresses for the standard and shot-
(Tmax),- -.21.74)( to 0.00762)( 1.71 x 109 peened gears show an increase of 40 percent for the shot-
peened gears over the standard gears at the depth of
- ; (0.26 x 109) = -.0.835 X ]09 maximum shear stress in addition to a 350 percen] in-
crease at a depth of 13.0 ILm (0.5 mil).
in U.S. customary units.

3105 1- Appendix - Deri.vation of Residual Stfess

(1max),= -3.15x 106 0.3)(248000 -2(-37 000) Effect on Maximum Shear Stress
It is weU known that classical rolling-element faHgue begins
;;;;;;- 121 08.0 in the subsurface zone of maximum shear stress.II1-IZ1
Therefore, to determine the ·effect of residua] stress on rolling-
And for standard gears, in 51 units, element fa.tigue in gears, it is necessary toanalyze the effect
of residua] stress on the maximum. shear stress below the sur-
9 578 375 face. The maximum shear stress at any point in a stressed
'('max), = - 21. 74)( W 0.00762)( L 71 X 1.09 volume below a rolling line-contact load i
1
- ~ (0.186 x 109) = - 0.8.72)( 109' ,( Tmaxl', = r(S~- Sy) (AI)

in U.S. customary units, I

(Tmax)x= i(S<:-Sx) (A2)
_ -·06 nos where 5z is the principal compressive stress in a direction
(1mu), =- 3d5)( L 0.3,x 248000
normal 'to the contact area, Sy is the prineipal comprehen-
sive parallel stress to the direction of Jlolling}131and S.r is 'the
- ~ (- 27 (00) = - 126 430
pnncipal stress normal.
For rollers or gear teeth loaded statically, the maximum
The surface fatigue [ifeUO) for gears is inversely proportional theoretical shear stress occurs in the y-z plane since the stress
to the maximum. shear stress to the ninth power. The in the)/. or rolling, direction is less than the stress in the x

January/February 1986, 3.5

direction. Therefore, the maximum shear stress is and for U.S. customary units
]
Tmax= r(SZ-Sy) (A3)
(Allb)

If the residual stresses are equal in the x and y directions,

for the Hne contact in the y-z plane the maximum shear stress where Sry can be either compressive or tensile .. When gears
including the residual stress is are shot peened, the residual. stress is compressive and,
therefore, reduces the maximum shear stress.
(A4) Since the rolling-element fatigue life of gears is inversely
proportional to the maximum shear stress to the ninth
power,(14)

. (1)9
L- -.-
Tmax
1

where 5ry is the residual stress in the y direction and is

positive for tensile stress and negative for compressive stress. (Al2)
or
From Reference 13 for line contact of rollers
b
Tmax = -0.30025 A (AS) L- [ ---1 ] 9
(Tmax), _
From equation (A4), where (Tm;uo)r =:Tmax -112 (5....)'
where b is the haU width of the Hertzian contact,

L- 1 ]9 (A13)
2P [_ T max - [12 I(Sry)
b=_...,,-N_._
1rLSmax
(A6)
using a life ratio of Ll and Lz
and A for a contact of two rollers of the same material is

(A14)
(A7)

If the rollers are of the same radius, When the residual. stress developed by the shot peening of
the gear teeth is known, the change in life produced by shot
A=2RC~(2) (AS) peening can be determined from equation (AU),

Where References
1. MOORE, H. F.: "Shot Peening and the Fatigue of Metals."
PN normal load, N (lb)
American Foundry Equipment Co.. 1944.
5max maximum Hertz stress, N/m2 (psi)
2. STRAUB, 1. C: "Shot Peening in Gear Design.. AGMA Paper
N

RIRz radius of curvature of the two rollers, m (in.) 109.13, June 1964.
o Poisson's ratio
3. VALENTINE, K. B.: "Recrystallization as a Measurement of
E Young's Modulus, N/m2 (psi) Relative Shot Peening Intensities." Am ..Soc. Met. Trans. Q.,
Vol. 40, 1948, pp, 420-434.
Substituting equations (A6) and (AS) into equation (AS) for
4. ZARETSKY, ERWIN V.: et al.: ~ects of Component Differen-
'"max results in tial Hardness OIl Residual Stress and Rolling-Contact Fatigue."
NASA TN D-2664, 1965.
P [ E ]-1 5. ZARETSKY, E. V.; PARKER, R.J.; and ANDERSON, W. J.:
Tmax = -0.30025 vri~ax RO -02) (A9) "Component Hardness Differences and Their Effect on Bear-
ing Fatigue." J. Lub. Technol., Vol. 89, No.1, Jan. 1967, pp,
47-62.
If equation (A9) is substituted into equation (A4), 6. DOWSON, D.; and HIGGINSON, G. R.: ".EJasto-
Hydrodynamic lubrication." Pergamon Press, 1966, p, 96.
PN
(Tmax),,,,-0.30025,rLSmax
[~ E
R(I. -02)1
]1 - 2ISry 7. JOHNSON, LEONARD G.: 'The Statistical Treatment of
(AIO) Fatigue Experiments," Elsevier Pub. Co., 1964.
B. CHRISTENSON,. A. L.,ed.: "Measurement of Stress by X-Ray.~
SAE HS-182, Aug, 1971.
For steel gears .E=207x 109 N/m2 (30 X lot' psi) and 0=0.30; 9. PREVEY, PAUL S.: "Method of Determining the Hastic Pro-
therefore, equation (Ala) becomes for Slunits perties of Alloys in Selected CrystalJographic Directions for
X-Ray Defraction Residual Stress Measurements." Adv. X-Ray
Anal., Vol. 20, 1977, pp, 345-354.
(AlIa)
(continued O'n page 48)

36 'Gear Technology
THE PROCESS Of GEAR SHAVING ... duction casts .. Moreover, further developments may be ex-
(continued from p.age 46) pected in the relatively young CBN technology, on the basis
of ongoing progress in fundamental knowledge.

Summary
One of the main problems in the application of CBN wheels
is the correct economical and technological design of the
dressing process, i.e. trueing and sharpening. This paper
presents methods for optimizing the dressing process, and in
particular, the sharpening process. A process mode] far
sharpening with a corundum sharpening stick is presented.
The chip space of the grinding wheel is described as a Iunc-
tian of wheel specification, setting parameters and duration
of the sharpening process. The model far description of
sharpening results can be used directly in practiealapplica-
tion, since it includes only variables that can be regarded as
known when the process design is made.
The technological advantages offered by the use of CBN
must be offset against the main disadvantage of high grind-
ing wheel. cost. As the tool casts per workpiece are mainly
influenced by wheel wear, the result of the present investiga-
tions show possibilities of improving wear behavior by adap-
tation of the grinding wheel specification. Possible measures
might be the selection of suitable grit size, the use of a harder
bond and an increase in grit concentration. An increase in
grit concentration makes the grinding wheel more expensive,
but in return it gives a clear improvement in the length .of
service life.
The machine concepts were also discussed as the prere-
quisite for economic application of this process. The fallow-
Supplementary 7
ing must be particularly stressed: high rigidity of the machine,
E-5 ON READER REPLY CARD high cutting speeds and drive powers and automated true-
ing and sharpening systems.
LOWER GRINDING COST . If the process is properly designed, it is at present possible
(continued from page 21) to reduce the production costs per drill flute by approximately
30 % as compared with corundum grinding.
the trueing costs (not included here) these constitute the pro-
IE-l ON READER REP,L V CARD,
duction costs per flute. This shows thatcorundum grinding
is about 30% more expensive than CBN grinding. The
superioity of CBN is likely to increase still further, assuming
a rise in the labor and machine casts which determine pro- EffECT OF SHOT PEENING ....
(continued from page 36)
Fig.. 17-CBN grinds more economically than corundum and gives better
quality

10. TOWNSEND, D. P.; COY, J. ].; and ZARETSKY, E. V.: "Ex-

perimental 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
Fa.tigue Phenomena." Symposium en Testing of Bearings,
ASTM, 1947, pp. 35-48; discussion, pp. 49-52.
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.
13. JONES, A B.: "New Departure-Analysis of Stress and Deflec-
tions," Vel.t New Departure, Div, Gen. Motors Corp.. 1946.
pp.22.
14. LUNDBERG, G.; and PALMGREN, A.: "Dynamic Capacity
of Rolling Bearings." Acta. Poly tech. Scand.. Mech. Eng. Ser..
Vel. 1, No.. 3, 1947.

E-4 ON READER REPLY CARD

48 Gear Technology