E F F E C T OF S T R E S S C O N C E N T R A T I O N IN V A R I O U S W O R K I N G M E D I A
ON L O W - E N D U R A N C E
FATIGUE
V. I. Tkachev, R. I. Kripyakevich, A. B. Kuslitskii, and G. I. Kreimerman Fiziko-Khimicheskaya Mekhanika Materialov, Vol. 2, No. 4, 464-467, 1966 When test pieces are cyclically deformed, plastic strains are usually produced near stress raisers. This happens par ticularly in low-endurance fatigue tests, in which the applied stresses are sufficiently high. Plastic flow has a substantial effect on stress raisers; the stress raiser geometry, which provides a basis for calculating the theoretical stress concentration coefficient, is changed, and work-hardening and other structural changes take place near the stress raiser root. Consequently, the theoretical stress concentration coefficient cannot be used as a criterion of the reduction in specimen life in low-endurance fatigue. For this purpose, the concept of the effective coefficient of stress concentration was introduced [ 1 - 4 ] in the form
Keff-
Nn
Ns
,
where Nn is the number of cycles to fracture of a test piece without a stress raiser and Ns the same characteristic of a speciman with stress raiser. The values of the effective and theoretical stress concentration coefficients approach when the nominal m a x i m u m stress is increased [1-5]. It was established in [6, 7] that the relationship between endurance of test pieces with stress raisers and the cyclic strain amplitude is linear (when plotted in logarithmic coordinates) and typical for low-endurance fatigue. There are published reports [5, 8 - 1 1 ] on the endurance of specimens with stress raisers tested in various working media in the low-endurance range, but no comparison has been made of the effect of these media in the presence and in the absence of stress raisers. The present investigation was concerned with the effect of total strain amplitude e and stress cycle frequency v on the low-endurance fatigue of specimens with stress raisers in air and in corrosive and hydrogen-charging media. The tests were carried out on fiat steel specimens, with stress raisers in the form of a 1 m m diam. hole (kteor ~ 1.8). The experimental technique and results of similar tests on specimens without stress raisers are described in the preceding article [12]. a
b
t ,-.~"L
g
8
2
t
61
...o
6
2 2
I 5
' 2
9
OL.
4
r
v,5
?
~ e,g
0.5
.
,
2
4
~,~
Fig. 1. Dependence of the effective stress concentration coefficient Keff on the strain amplitude e in tests in (a) air; (b) 3~ NaCl solution; and (c) 3% NaC1 solution with cathodic polarization at 10 A/dm z. 1) 100 cpm; 2) 10 cpm; 3) 1 cpm. It was found that the character of the fundamental relationship N(e) for specimens with stress raisers is qualitatively the same as for unnotched test pieces. In the former case, however, there is a sharp reduction in critical strain amplitude r [12] which can be attained even in a hydrogen-charging medium.
333
As shown in Fig. la, which relates to tests in air, Keff increases with increasing ~, i. e,, the detrimentai effect of stress raisers on endurance becomes more pronounced at large e. This is attributable to an increase in temperature of the specimen taking place at large e, particularly at high v.
F Io
< 2
-...d. 9 2 \"
2
',o,,
"9. ,
.+a
\
"-4
I
Io
,.-~::-
/oo
//J
'
~'.~1
' '
i
4
~
I
v, cpm
+ +
I0 V,
"j
Io0
cpm
Fig. 2. Dependence of the coefficient of the effect of working medium ~ on the stress cycle frequency v plotted for specimens with stress raisers (Roman numerals) and for tmnotched specimens (Arabic numerals), a) 3% NaC1 solution; b) 3% NaC1 solution with cathodic polarization at 10 A/dmZ; 1, I),~ = 0.5"]r 2, II) ~ = 4%.
+
a
C
8 I
3 ~,T
3 2 ..,
i
i
0
i
tO v, cpm
!
tOO
~
;0
u, cpm
100
tO
/00
P, cpm
Fig. 3, Effective stress concentration coefficient Keff plotted against stress cycle frequency v. a) Air; b) 3% NaC1 solution; c) 3% NaC1 solution with cathodic polarization at 10 A/dmg; 1) e = 4~ 2) e = 25~ 3) ~ = 0.59o.
384
Heat transfer conditions in corrosive and hydrogen-charging media are better than in air. Even so, one stress cycle at 100 cpm is not sufficiently long to allow for heat dissipation; as a result, the reiationship observed in tests in air holds also for specimens tested in liquid media (curves 1 in Fig. lb, c), although at different levels of Kef f. The variation of Keff in working media is, to a large extent, dependent on the coefficient of the effect of a given medium, which is the ratio of the endurance of a specimen in air to its endurance in the medium: Nnm a ~n Ken}f-- Arm = Keff" ~ s
'
where indices m and a relate, respectively, to tests in a liquid medium and in air. Consequently, the effect of the working medium on Keff depends on the ratio of the coefficients of the effect of the medium on specimens with and without stress raisers. This ratio may increase or decrease, depending on the cyclic load ing parameters (~, v) As foltows from Fig. 2, Bn/Bs increases with increasing a, which leads to a reduction in Kef f (see curve 3 in Fig. lb, and curves 2 and 3 in Fig. lc). The results of tests on the effect of v on Kef f are reproduced in Fig. 3. It will be seen that the effect of v in tests in air is similar to the effect of ~ (Fig. 3a) (and for the same reason). As a role, B for unnotched specimens in corrosive and hydrogen-charging media is larger than for specimens with stress raisers (see Fig. 2). As a result, Keff is slightly smaller in these media than in air. The increase in Keff at high v is attributable to two factors: a reduction in the difference between Bs and Bn, and an increase in specimen temperature (curves i and 2 in Fig. 3, b, c). The variation of Kef f in the hydrogen-charging medium at s = 0.5 is an exception: Here Kef f decreases with increasing v (curve 3, Fig. 3c). This is because at v = 1 cpm ~n is smaller than Bs, which leads to an increase in Keff; gn increases with increase in v, which leads to a reduction in Keff. SUMMARY 1. Stress raisers produce a sharp reduction in the number of cycles to fracture in low-endurance fatigue, this effect being most marked in tests in air. 2. The character of the relation N (s) for unnotched test pieces and specimens with stress raisers is similar. 3. The critical strain amplitude is smaller for specimens with stress raisers. 4. The dependence of the effective stress concentration coefficient on the strain amplitude and stress cycle frequency is determined by the ratio of the coefficients of the effect of working medium on unnotched test pieces and specimens with stress raisers. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8, 9. 10. 11. 12.
H. F. Hardarath and W. Illg, NACA TN, no. 3132, 1954. H. F. Hardrath, C. B. Landers, and E. C. Utley, NACA TN, no. 3017, 1955. E. b/. Evans, The Engineer, London, part 1, 203, 5274, 1952. W, Illg, NACA TN, no. 3866, 1956. I. H. Gross and R. D. Stout, J, Weld, 34, 161, 1955. W. G. Finch, Proc. ASTM, 52, 1952. T. H. H. Plan and R. D' Amato, WADC TN, no. 5 8 - 2 7 , 1958. T. b/, Crooker, R, W, Judy, R,E. Morey Jr, , and E, A. Lange, US Naval Res. Lab. Rept. NRL Progr,, March, 1965 Y. Minami and Y. Fukuda, Proc. 6th Japan Congr. Test. Mater., Kyoto, Japan Soc. Test Mater, 1963. Y. Minami and H. Itagaki, Proc. 7th Japan Congr. Test. Mater., Kyoto, Japan Soc. Test, Mater, 1964. G, Sheven, G. Sach, and K. Tong, Proc. ASTM, 57, 1957. V. I. Tkachev and R. I. Kripyakevich, FKhMM [Soviet Materials Science] no. 4, 1966.
3 March 1966
Institute of Physics and Mechanics AS UkrSSR, L 'vov.
335