STRENGTH CHARACTERISTICS

R E S I S T A N C E OF C A S E - H A R D E N E D S T E E L T O C O N T A C T F A T I G U E

(UDC 620.178.154.3 : 621.78.5)

M. A. B a i t e r a n d M. L. T u r o v s k i i

Khar 'kov Transportation Machine Construction Plant Translated from Meta l lovedenie i Termicheskaya Obrabotka Metal lov, No. 3, pp. 2-6, March, 1966

We studied the contact fatigue resistance of case-hardened steel , using four-rol ler machines to imi ta te (at the surface of the rolIers) fr ict ion conditions close to the ac tual working conditions of gear teeth [1]. The distr ibution of the sl iding veloci t ies over the c i rcumstance of rollers is shown in Fig. 1. The contact fatigue strength was m e a -

sured by the working l ife under stresses of 18,000 or 24,000 k g / m m 2 according to the Hertz method.

The tests were divided into stages in each of which the sample was subjected to a cer ta in stress under iden- t i ca l conditions. After each stage we measured the radia l wear of the rolling surface every 30 ~ of the c i rcumference (with a ve r t i ca l opt imeter) and the microhardness 20 microns from the surface of friction. We aiso examined the working surface under magnif icat ions of 10 and 100. The magni tude of the wear measured with the opt imeter is the

totaI crnmbling out by wear and pitt ing. Also, we studied the var ia t ion of residual f irst-order stresses on the working surface during friction. Tangential residual stresses were determined by the general ized Davidenkov method [2, 3].

We invest igated the influence of the structure and hardness on the contact fat igue strength of 18Kh2N4VA steel which, after case hardening, was heat treated as follows: 1) quenching in oil from 850~ and tempering at 140 ~ C (HRC 54); 2) high temperature tempering at 650~ quenching in oil from 800~ and tempering at 140~ (HRC

59); 3) quenching from 800~ cold t rea tment at -120~ and tempering at 140~ (HRC 63). The thickness of the case-hardened layer was 1.7-1.9 ram.

The working surface of the roller was heat treated and then 0.3 mm of the radius was ground off so that the friction surface corresponded to the case-hardened layer containing the max imum amount of residuai anstenite.

Standard connected rollers of 20Kh2N4A steel were case hardened and heat treated to a hardness of HRC 60-61.

It is widely be l ieved that the contact fatigue strength varies proport ionally with the hardness of the mater ia l . The results we obtained show that the contact fat igue strength of case-hardened al loyed s teel depends not so much on hardness as on the microhardness of the case-hardened layer , which varies with the amount of residual anstenite and carbides resulting from the hea t t rea tment conditions. Not only the amount of austenite but also its distribution

plays an important role. As the result, the var ia t ion of the contact fat igue strength as a function of hardness is not uniform. Thus, t rea tment of 18Kh2N4VA steel at subzero temperature (heat t rea tment No. 3), which ensures a con- s iderable increase in hardness as compared to heat treatments 1 and 2, does not increase the resistance to contact fat igue (Fig. 2).

X-ray and structural analyses of the case-hardened layer showed different amounts of residual austenite and a difference in the composit ion and the distribution of phases after the different heat treatments.

Heat t rea tment No. 2 induces uniform distribution of the structural componen t s -h idden needle structure of martensi te and smal l areas of austenite. These lat ter are so smal l that when the amount of austenite is 35% they cannot be seen under a magnif ica t ion of 1000. The structure resulting from t rea tment at subzero temperatures in - duces a combinat ion of needle martensi te formed during quenching and hidden needle martensi te formed during cooling to subzero temperature , i .e . , the inherited structural heterogenei ty pecul iar to the case-hardened layer of highly a l loyed steel (and due, apparently, to the eoncentrat ional heterogenei ty) is preserved.

The microhardness of austenite which is not decomposed during t rea tment at subzero temperatures (H 640-660) is considerably higher than the hardness of the residual austenite before the t reatment at subzero temperature (H 520).

X-ray analysis showed that the concentrat ion of carbon in residual austenite after quenching is 1.2-1.3%.

177

Vslid, m/sec

lX/__l i _ l l l , " ! l\,r-- _ L ~ ) t /

+i \ oL_l ), l i l t !

I i ~\1 ,Do F-5-a7~

: A7 Fig. 1, Distribution of sliding velocities

over the surface of the roller.

Diffusionless transformation of a large amount (65-70%) of high- carbon austenite and of anstenite alloyed with chromium and nickel is apparently responsible for the high microstresses produced during cool- ing to subzero temperature under conditions in which relaxation pro- cesses are difficult (this is also indicamd by the considerable increase in the microhardness of residual austenite). The increase in the micro- hardness of the residual ausmnite decreases the resistance to rupture in steel treated at low temperature, which explains the ineffectiveness of treatments at very low temperatures for increasing the contact fatigue resistance of Case hardened steel (similarly, the strength of untempered martensite in medium carbon steel is low). At the same time, the case hardened layer containing a considerable amount of austenite and, con- sequently, having a low hardness (HRC 54-56), has a high contact fa~ tigne strength (Fig. 2a).

The investigation showed that the prepitting cracks appear earlier and develop faster on the case -hardened surface treated at very low temperature than on the surface not treated at subzero temperature. This phenomenon can be explained by the dynamics of the formation of pits, which is the result of two simultaneous processes: 1) cold working of the surface (smoothing and removal of the peaks of the protrusions left by mechanical treatment) with the formation of cracks under the influence of alternating stresses ;

MK 2 75 " I~ 3,5 "lO ~ ~

~ 20 " 15 " 10 6 ~ . .~20.#0 ~

;e~a

50 120 #80 NO ~00 3~0 ~ ~p a

qO

~00

88

�9

50

#0

20

o

Fig. 2.

i

i 9 �9 10 6

x

,

50 #20 #80 2zsO 700 0 60 #20 #80 21~0 700 760"~o c Angle of rotation of the roller b

Wear of the working surface of the roller as a function of the number of test cycles, a) Heat

treatment No. i; b) heat treatment No. 2; c) heat treatment No. 3.

178

H

gs~ I T iT;7 850 / H~ m '

5sa ! - - : - ~ + - i - - - r - - - 4 - -

I , i ! : ! i J a

,oZ J J L 2 i Tql b

goo I / 5 ' - " ~ ! ~ i ~ I - -

~" ' " , i , ! ~ i L I

1/ I J ! xk J I I/i i\l ! ~ .

0 30 90 150 210 270 COO tpo C

Fig. 3. Variation of the microhardness of the friction surface along the working surface of the roller , a) Heat t rea tment No. 3; b) hea t t r ea t -

ment No. 2; c) heat t rea tment No. 1.

o kg/mrn '~

uJ I , i ' , 2 --~v-'~ : , " I I ~0 1 80 1 2 0 / ~ 6 0 I 200 1 24'O I ~ i n '

io . i " ~ ' ' - ~ 2 I ~ I

20 ~ ~ ' - -

Fig. 4. Variation of first-order residual stresses during friction. 1) Pure roiling; 2) roll ing aM

sliding (Vslid = 2.187 m/sec) .

TABLE 1

Condi- tion No.

Type of t rea tment

Heat t rea tment , grinding

Grinding, hea t t rea tment Heat t rea tment , grinding, shot

blast ing Grinding, hea t t reatment , shot

b lasting

Heat t reatment , grinding, e iec- t rolyt ic polishing

Heat t rea tment , grinding, shot blast ing, e lec t ro ly t ic polish- ing

Heat t rea tment , grinding, ~ea t - ment with rollers

Number of cycles before crumbling

out

1,359,000 2,300,000

945,000

1,800,000

2,234,000

1,249,400

3,026,000

2) wearing off of the cold-worked layer during which the surface cracks are e l imina ted and crumbling out is slowed down. In the presence of a large amount of anstenite in the case-hardened l ay -

er the wearing off of the surface layer , unlike that of martensi te , determines the process of development of prepit t ing cracks.

The study of the microhardness of the fr ict ion surface after the test showed addi t ional reasons for the difference in the behav- ior of the two typ ica l microstructures. The type of var ia t ion of the microhardness (Fig. 3) during fr ict ion turned out to be d i f fer - ent in principle. In the case of the in i t ia l structure rich in aus-

teni te (heat t rea tment No, 1) the hardness increases over the whole working surface of the roller during the friction process. In the case of the martensi t ic structure (heat t reatments 2 and 3) the hardness increases slightly on areas of max imum specific sliding, while in the case of pure roiling the increase is par t icular ly large after hea t t rea tment No. 3. It is character is t ic that for bothtypes

of structure the max imum change in hardness is in the zone of pure roil ing, i .e . , in areas of most intense pit t ing. Thus, during the friction process the austenit ic structure is strengthened in the areas most susceptible to pit t ing (which

is apparently due to the increased capac i ty of austenite for cold working), while the martensi t ic structure softens b e - cause of the rapid dec l ine in its susceptibi l i ty to plastic deformation. The structures resulting from treatment at subzero temperatures are much more rarely subject to softening.

However, the use of the structure of the case-hardened layer with increasing amounts of residual austenite is l imi ted by the low fatigue resistance on bending and torsion [4]. Also, the strength of toothed gears in use, which is the result of the in teract ion of a number of factors in addit ion to the microstructure (the type of stress, the precision of fitt ing, the s tabi l i ty of the profile of the teeth during use, the finish of the surface, etc .) , does not always co in - c ide with the results of laboratory investigations. Therefore, the presence of a considerable amount of anstenite in the case-hardened layer sometimes leads to crumbling and to the deformat ion of the teeth as the result of plastic deformat ion, and this is ~he determining factor in the working life of gears.

The study of the var ia t ion of the residual f irst-order stresses during friction (0.06 m m of the surface) showed that within the area of pure roi l ing the in i t ia l compression stress decreases and is transformed into tensile stress, while in the case of roil ing a M sliding the compression stresses remain almost constant (Fig. 4).

When one takes into account the fact that the magni tude and the sign of residual stresses affect the de formabi l - i ty by slip, it is c lear that they play a role in the deve lopment of contact fatigue. Indirect ly, this is confirmed by the known fact of preferent ial formation of pit t ing in the area of pure roil ing where the residual tensile stresses

179

formed during fr ict ion (see Fig. 4) fac i l i t a te deformation by slip, while the compression stresses in the head of the

tooth handicap deformat ion by slip.

On the basis of the existing concepts concerning the role of residual stresses in the resistance to destruction by contact fatigue, we made tests to de termine the contact fat igue resistance of case hardened 20Kh2N4A steel after different t reatments of the surface which created different residual stresses. We studied the influence of grinding, shot blasting, roll ing with rollers, and e lec t ro ly t ic polishing.

The tests were made under a stress of 24,000 k g / m m 2 by the Hertz method. The working l i fe of the samples after three to four tests is given in the table. The scattering of the results does not exceed 10%.

The results of the tests showed the following:

1. Grinding before hea t t rea tment when the unfavorable effect of grinding is e l imina ted (condition No. 2) ensures a higher contact fat igue strength than grinding after heat t reatment (condit ion No. 1).

2. Shot blast ing of unpolished surfaces (condit ion No. 4), in spite of a considerable increase in roughness (by a factor of 2-2.5) as compared to polished surfaces (condition No. 1), does not decrease but rather sl ightly increases the contact fatigue strength. Taking into account the very strong influence of the finish of the surface on the con- tact fatigue strength [5], one must assume that residual compression stresses created by shot blasting play a consider-

able role.

3. Electrolyt ic polishing has a favorable influence on the contact fatigue strength. The decisive factor in this case is apparently the removal of the thin layer modif ied by grinding,

4. The decrease in the degree of finish of the surface due to shot blasting of polished surfaces (condit ion No. 3) in the presence of the layer modif ied by grinding leads to a decrease of the contact fatigue strength as compared to that resulting from polishing (condition No. 1). Electrolytic polishing after shot blast ing (condition No. 6) decreases the roughness of the surface and increases the contact strength to the leve l of the ground unhardened samples (condi- t ion No. 1).

5. Roiling with rollers (P = 1000 kg; radius of the roller, R= 5 mm; S = 0.2 mm/revolut i0n) increases the contact strength the most. The effectiveness of this type of strengthening is due to the favorable residual compres-

sion stresses and the high degree of finish of the surface.

L I T E R A T U R E C I T E D

1. L .M. F e l ' d m a n and M. A. Baiter, Zavodskaya laboratoria , ~ No. 12 (1955). 2. M . A . Babichev, Methods of Determining Internal Stresses in Machine parts [in Russian], Izd. AN SSSR,

Moscow (1955). 3. M . A . Baiter, N. M. Grinberg, L S. Dukarevich, and M. L. Turovskii, Modern Scient if ic , Technical , and

Industrial Experience, No. 9, Subject 7, No. M-62-240 /9 [in Russian], GosINTI, Moscow (1962). 4. M . A . Baiter, MiTOM, No. 5 (1956). 5. A . I . petrusevich, Quali ty of the Surface and Strength of Materials under Contact Stress [in Russian], Izd.

, AN SSSR, Moscow (1946).

All abbreviations of periodicals in the above bibliography are letter-by-letter translitera- tions of the abbreviations as given in the original Russian journal. S o m e o r a l l o f t h i s p e r i -

o d i c a l l i t e r a t u r e m a y w e l l b e a v a i l a b l e in E n g l i s h t r a n s l a t i o n . A complete l is t of the cover-to- cover English translations appears at the back of the first issue of this year.

180