You are on page 1of 7

The Effect of Microstructure on the Fatigue Crack

Propagation Behavior of an AI-Zn-Mg-Cu Alloy


J. L I N D I G K E I T , A. GYSLER, A N D G. L U T J E R I N G

The effect of slip distribution on the fatigue crack propagation behavior in vacuum of a high
purity A1-5.9Zn-2.6Mg-1.7Cu alloy in various age-hardened conditions has been investi-
gated. The crack propagation resistance was observed to be significantly higher for
underaged microstructures containing shearable precipitates in comparison to overaged
conditions with nonshearable precipitates. The improved crack propagation resistance is
attributed in part to an increased amount of reversed slip in the plastic zone at the crack tip
due to a higher degree of planar slip for conditions with shearable precipitates. The observed
increase in fatigue crack propagation resistance with decreasing precipitate size for
microstructures containing a constant volume fraction of shearable precipitates cannot be
explained on the basis of such slip reversibility alone. The variation in ductility for the
different microstructures has also to be taken into account. It was found that the enhanced
crack propagation resistance can be correlated to the increased ductility with decreasing
precipitate size. This explanation was supported by the experimental observation that
microstructures containing different volume fractions and sizes of shearable precipitates but
exhibiting the same ductility showed approximately the same resistance against fatigue crack
propagation.

1. I N T R O D U C T I O N fatigue crack propagation behavior of aluminum and


titanium alloys also showed experimental evidence
B E S I D E S the modulus of elasticity ~also microstruc- which demonstrated improvements in crack propaga-
tural effects are now generally considered to affect the tion resistance with increased tendency for planar
fatigue crack propagation behavior of age-hardened slip. H-12 However, it was found in these studies, that
alloys, especially in the low and high cyclic stress such improvement was observed only when the fatigue
intensity factor regime. The microstructural parameters tests were performed in an inert environment.
which were found to have an influence on the crack Laboratory air for example was leading already to an
propagation resistance are the grain size2-5 and the slip opposite ranking with regard to the fatigue crack pro-
distribution, which might be changed by varying pagation resistance between age hardening conditions
stacking fault energy6-9 or degree of age-harden- with a planar as compared to a more homogeneous
ing. 1~ Fatigue crack propagation measurements for deformation mode. 11
copper alloys with a wide variation in stacking fault The purpose o f the present study was to investigate
energy showed that significantly lower propagation the effect of the slip distribution on the fatigue crack
rates were observed with decreasing stacking fault en- propagation behavior of an AI-Zn-Mg-Cu alloy in an in-
ergy. 6-9 However, no explanation for the decreasing ert environment. Various conditions with different heat
crack propagation rates with decreasing tendency for treatments were selected in order to obtain a wide
dislocation cross slip processes was offered by these spectrum of different slip distributions: microstruc-
authors. On the other side in a recent study Horn- tures with shearable or nonshearable precipitates of
bogen and Zum Gahr ~~explained the observed im- different sizes and volume fractions, but leading to
provements in fatigue crack propagation resistance of the same yield stress and microstructures with the
an underaged as compared to an overaged microstruc- same volume fraction of shearable precipitates but
ture for an Fe-Ni-AI alloy by a model of reversed slip different particle sizes. With the investigation o f these
of dislocations at the crack tip. They concluded that conditions it was hoped to further the understanding
the propagation rate in a material which promotes a of the influence of slip distribution on the fatigue
planar slip mode will be strongly retarded by such crack propagation behavior o f age-hardened alloys.
reversed dislocation movement during the unloading
part of a loading .cycle. ~~This also would explain the
lower propagation rates found in low stacking fault
energy materials. Other recent investigations o f the 2. EXPERIMENTAL PROCEDURE
The high purity laboratory aluminum alloy with the
J. LINDIGKEIT, formerly with German Aerospace Research basic composition A1-5.9Zn-2.6Mg-l.7Cu (wt pct) was
Establishment (DFVLR), Cologne, Germany, is now with Krupp prepared by Alcan Laboratories, Banbury, U. K., where
Research Institute, D-4300 Essen 1, Germany. A. GYSLER is with the cast ingots were rolled at 430 ~ to a thickness of 25
German Aerospace Research Establishment (DFVLR), D-5000Co- ram. The chemical composition of this alloy is shown in
logne 90, Germany. G. LOTJERING, formerlywith Ruhr-University, Table I. Blanks of these plates were solution heat
Bochum, Germany, is now with Technical UniversityHamburg-
Harburg, D-2100 Hamburg 90, Germany. treated at 450 ~ for 1 h, quenched in ice-water and
Manuscript submitted March 20, 1980. immediately rolled at room temperature to a final
ISSN 0360-2133/ 81/0911-1613500.75l0
METALLURGICALTRANSACTIONSA 9 1981 AMERICAN SOCIETY FOR METALS AND VOLUME 12A, SEPTEMBER 1981--1613
THE METALLURGICAL SOCIETY OF AIME
ments were applied which resulted in a constant yield
Table I. Chemical Composition of the Alumunlum Alloy (WI Pct) stress of 475 M N m - 2 : 1 h at 160 ~ and 100 h at 100 ~
(both shearable precipitates) and 70 h at 160 ~ (non-
Zn Mg Cu Cr Si Fe Mn Ti AI
shearable precipitates). To compare the effect of dif-
5.93 2.58 1.7 0.002 (0.01 (0.01 0.002 0.017 Balance ferent volume fractions and particle sizes of shearable
precipitates, the following age-hardening treatments
were chosen resulting in a constant yield stress of 280
thickness of about 12 ram. Tensile specimens, with gage M N m - 2 : 1 0 0 h at 20 ~ and 0.5 h at 100 ~ To study
dimensions of 5 mm diam and 25 mm length, and single the effect of different particle sizes at a constant volume
edge notched (SEN) specimens, with gage dimensions fraction the following age-hardening conditions were
of 2.5 mm thickness, 20 m m width and 50 mm length, selected: 0.5, 10 and 100 h at 100 ~ aging temperature.
were machined so that the stress axis upon testing The tensile properties of all these aging conditions are
would be parallel to the rolling direction. The specimens summarized in Table II.
were recrystallized at 450 ~ for 2 h, quenched in TEM studies of thin foils, prepared from samples
ice-water and aged at room temperature or stored at with the different age-hardening treatments (Table II),
room temperature for two days before aging them at showed that no semicoherent ~/'-particles were precip-
100 or 160 ~ for various times. The recrystallized itated within the matrix of specimens aged either at 20
material showed an equiaxed grain structure with a or 100 ~ Also no precipitate free zones were found
grain size of about 100 #m. The starter notch of the along grain boundaries for samples aged at both of
SEN specimens was machined by spark erosion cutting these temperatures as well as for samples aged 1 h at
after the final aging treatment. 160 C. The micrograph in Fig. 2(a) may serve as an
Tensile tests were performed at room temperature in example to characterize these microstructures. Aging
air employing an initial strain rate of 6 • 10 -4 s-1. All the alloy for 70 h at 160 C resulted in precipitate free
fatigue crack propagation tests were carried out at room zones along grain boundaries due to the precipitation of
temperature in vacuum ( ~ 10 -4 Pa) on a closed-loop equilibrium ,/-particles at the boundaries, as can be seen
servohydraulic machine under stress controlled sinusoi-
dal tension-tension loading with an R-ratio of 0.1 and a
frequency of 30 Hz. After initiating a fatigue crack at
the starter notch the load was reduced in several Table II. Tensile Properties
decrements until well established crack propagation E Oo 2 OF EF
rates of about 10 -6 m m / c y c l e were reached. No meas- Aging Condition ( G N m -2) ( M N m -2) ( M N m -z) ln(Ao/A~) n*
urements were utilized until the crack had propagated
100 h, 20 ~ 68.5 280 729 0.56 0.56
at last 1 mm beyond the region of plasticity associated 0.5 h, 100 ~ 68.8 280 704 0.56 0.55
with this initiation procedure. The crack length was 10 h, 100 ~ 68.5 413 752 0.44 0.37
measured optically with a travelling microscope. 100h, 100 ~ 68.7 475 756 0.40 0.38
The microstructures were investigated by thin foils, i h, 160 ~ 69.0 475 786 0.51 0.31
70 h, 160 ~ 68.9 475 665 0.31 0.33
prepared with a double-jet electropolishing apparatus,
using a JEOL electron microscope operating at 200 kV. *n = [d log(a - %)]/d log e.
The fracture surfaces were examined in a Cambridge
scanning electron microscope.
6OO
3. E X P E R I M E N T A L R E S U L T S
|
160~
3.1. Tensile Properties 4130

The dependence of yield stress 00.2 and true fracture


! RT
strain eF as a function of aging time is given in Fig. 1 for
aging temperatures of 20, 100 and 160 ~ At the lower //"
aging temperatures (20 and 100 ~ the yield stress 0.8
increased continuously within the time interval studied,
whereas the true fracture strain declined. At the higher
aging temperature of 160 ~ the yield stress passes
through a maximum (at about 10 h) indicating that the 0 6 ,.
alloy can be overaged within reasonable time at this "•"•'• RT
w
E
temperature. The ductility goes through a minimum,
reaching a relatively low value of about 0.27 after about
IOOeC
15 h (Fig. 1). The tensile fracture behavior was found to
be similar to the behavior of a comparable AI-Zn-
Mg-Cu alloy reported previously. 13 % O2
From the age-hardening curves in Fig. 1 different 0 10o 10~ 102 103
agelt~ t0me [h ]
heat treatment conditions were chosen for the fatigue Fig. 1--Tensile yield stress o 02 a n d true fracture strain cr as a function
crack propagation tests. To study the effect of shearable of aging time for three different aging temperatures of 20, 100 and
and nonshearable precipitates, different aging treat- 160 ~

1 6 1 4 - - V O L U M E 12A, S E P T E M B E R 1981 METALLURGICAL TRANSACTIONS A


Fig. 3--Fatigue crack propagation rate d a / d N as a function of cyclic
Ca)
stress intensity factor AK for microstructures with shearable (1 h
160 ~ 100 h 100 ~ and nonshearable (70 h 160 ~ precipitates
(002 = 475 MNm -2 for all conditions).

(b)
Fig. 2--Transmission electron micrographs for underaged and over-
aged microstructures. (a) 100 h at 100 ~ (b)70 h 160 ~

Fig. 4--Fatigue crack propagation rate d a / d N as a function of cyclic


in Fig. 2(b). Within the matrix of this overaged condi- stress intensity factor AK for microstructures with different sizes and
tion small semicoherent ~/'-particles were precipitated volume fractions of shearable precipitates (a02 = 280 MNm -2 for
both conditions).
with sizes of about 20 nm, as wasdetermined from dark
field electron micrographs.

3.2. Fatigue Crack Propagation On the other hand only minor differences in crack
The crack propagation results are shown as plots of propagation rates were found between the two under-
fatigue crack propagation rate d a / d N vs the cyclic aged microstructures which were aged to contain both
stress intensity factor AK. In Fig. 3 the results are shearable precipitates (100 h 20 ~ and 0.5 h 100 ~
shown for the comparison between shearable and and to have the same yield stress of 280 MNm -2, as can
nonshearable precipitates. This figure also shows the be seen in Fig. 4. The AK-values at low crack propa-
scatter band obtained from two tests of the same gation rates were almost the same for both conditions
age-hardening condition which can be considered rep- (about 10 MNm-3/2).
resentative for all the other d a / d N - A K curves. It can be Fatigue crack propagation results for underaged
seen that pronounced differences in crack propagation microstructures, aged at 100 ~ for various times, are
rates exist between the three different microstructures, given in Fig. 5. In the intermediate AK-region almost no
especially below AK-values of about 10 MNm -3/2. At difference in crack propagation rates was found be-
the lowest crack propagation rate (about 10 -6 mm/cy- tween the three different age-hardened conditions. At
cle) which could be measured within reasonable testing low AK-values some deviation among the three curves
times, the overaged condition (70 h 160 ~ showed the was observed, where the condition with the longest
lowest AK-value (6 MNm -3/2) whereas the underaged aging time (100 h) showed the highest propagation rate
conditions exhibited a higher resistance against crack whereas the highest resistance against slow crack prop-
propagation. The highest AK-value (8.5 MNm -3/2) was agation was observed for the microstructure with the
found for the condition aged for 100 h at 100 ~ shortest aging time (0.5 h).

METALLURGICAL TRANSACTIONS A VOLUME 12A, SEPTEMBER 1981--1615


Fig. 5---Fatigue crack propagation rate da/dN as a function of cyclic
stress intensity factor AK for microstructures with different sizes of
shearable precipitates.

rb)
3.3. Fracture Surface Observations
Scanning electron microscopy studies of the fatigue
fracture surfaces of all microstructures revealed that
crack propagation occurred mainly transgranularly
along slip bands in the low and intermediate
AK-regime. This is shown by the micrographs in
Fig. 6,7, and 8, taken from regions where the lowest
crack propagation rates were measured. However the
slip step heights, visible on these micrographs, showed
pronounced differences. It was observed that significant
differences in the fracture surface appearance occurred
between those microstructures for which the most
pronounced differences in propagation rates were meas-
ured (Fig. 3). For example the micrograph in Fig. 6(a)
shows only a few very large slip steps, the orientation of

(c)

these steps with regard to the crack front varying from


grain to grain. On the other hand the fracture surfaces
of specimens aged for 1 h and 70 h at 160 ~ which
resulted in a lower resistance against crack propagation
(Fig. 3), were exhibiting much finer slip steps within
each grain, as can be seen in Figs. 6(b) and (c).
Comparing the fracture surfaces at low AK-values for
the three different microstructures aged at 100 ~
where higher resistance against slow crack propagation
was found for specimens with shorter aging times
(Fig. 5), one can see that much finer slip steps were
observed for this aging condition (Fig. 7(a)) in com-
parison to the less resistant microstructure aged for
100 h at 100 ~ (Fig. 7(c)). In contrast to these observa-
tions, for microstructures which showed almost identical
(a) d a / d N - A K curves, i.e., for 100 h at 20 ~ and 0.5 h at
Fig. 6--Fatigue fracture surfaces at low crack propagation rates 100 ~ (Fig. 4), also the same fracture topography was
(da/dN = 10 -6 mm/cycle) for microstructures containing shearable found. For both of these microstructures fine slip steps
and nonshearable precipitates (SEM). (a) 100 h 100 ~ AK = 8.2
M N m -3/2, (b) 1 h 160 ~ AK = 7.0 M N m -3/2, (c) 70 h 160 ~ AK were observed in the low AK-region, an example is
= 6.0 MNm -3/2. shown by the micrographs in Fig. 8.

1616--VOLUME 12A, SEPTEMBER 1981 M E T A L L U R G I C A L TRANSACTIONS A


fa~
(a)

fb)
Fig. 8--Fatigue fracture surface at low crack propagation rate for an
underaged microstructure (SEM): 100 h 20 ~ AK = 10 MNm-3/2;
da/dN = 1.2 • 10 -6 mm/cycle (micrograph (b) higher magnification
of (a)).

4. D I S C U S S I O N
The results obtained in the present work will be
discussed on the basis of existing fatigue crack prop-
agation models which take microstructural parameters
into account. The discussion will be focused mainly on
the low AK-regime where plane strain conditions are
applicable. First a model will be applied, which was
proposed by Hornbogen and Zum Gahr ~~to explain
their fatigue crack propagation results for an Fe-Ni-Al
alloy in an underaged and an overaged condition. This
model is based on the reversibility of dislocation motion
within the plastic zone close to the crack tip. These
authors pointed out that in a microstructure containing
shearable precipitates, which promote a planar dislo-
Fig, 7--Fatigue fracture surfaces at low crack propagation rates
(da/dN = 10-6 mm/cycle) for microstructures with different sizes of
cation distribution, a certain number of dislocations will
shearable precipitates (SEM). (a) 0.5 h 100 ~ AK = 10 MNm -3/2, be able to move backwards during unloading in the
(b) 10 h 100 ~ AK = 9.2 MNm -a/2, (c) 100 h 100 ~ AK = 8.2 same slip plane as during the forward motion in the
MNm-3/2. rising part of a loading cycle. Some of these dislocations
METALLURGICAL TRANSACTIONS A VOLUME 12A, SEPTEMBER 1981--1617
will leave the material at the crack tip and therefore will behavior of age-hardened alloys. It should be men-
not contribute to crack propagation. The number of tioned that a similar discrepancy was also observed
dislocations moving in the opposite direction upon between this concept of slip reversibility and the results
unloading and which are able to leave the material of fatigue crack propagation measurements on age-
depends, of course, strongly on the degree of slip hardened Ti-A1 alloys?4
planarity. For example, if dislocation cross slip pro- In recent studies attempts have been made to predict
cesses take place, induced either by high stacking fault the fatigue crack propagation behavior from low and
energies or by nonshearable precipitates, some dislo- high cycle fatigue properties./5-~7 The equation proposed
cations will move out of the original slip planes during by Majumdar and Morrow ~6for example is:
the rising part of the loading cycle and will therefore
reduce the number of dislocations moving backwards da - 2 ( b + c) [ oy ]-l/(b+c)
on the initial slip plane. This would result in an dU = b + e + 1 _4(1 + n')o)c'y]
enhanced crack propagation rate.
This model, applied to the results obtained in the
• C--0--D]2O
]Cb+c+'
" )/(b+r
present study, provides an explanation in a qualitative
way for some of the microstructural effects on the
fatigue crack propagation behavior. The increased re- _ [4(l + n,)c,r]~b'c+wc~+~')c',AK2
sistance against crack propagation of microstructures
containing shearable precipitates in Fig. 3 (1 h 160 ~
and 100 h 100 ~ in comparison to the structure with The following parameters can be determined from low
nonshearable precipitates (70 h 160 ~ can therefore cycle fatigue tests: dr, c'v = cyclic uniaxial yield stress
be explained with a higher degree of reversed slip within and yield strain respectively, n' = cyclic strain hardening
the plastic zone for both of the underaged conditions. exponent, c = fatigue ductility exponent and
Further support for this explanation can also be drawn ~'j = fatigue ductility coefficient, determined from Cof-
from the fracture surface studies which showed that fin-Manson plots. From high cycle fatigue tests one
high slip steps are related to the microstructures con- obtains: b = fatigue strength exponent and a"I = fa-
taining shearable precipitates (100 h 100 ~ while tigue strength coefficient. The other parameters are: p*
much finer slip steps were found for the condition with -- microstructural size related to the mean free path
nonshearable precipitates (compare Figs. 6(a) and (c)). between major deformation barriers and COD = crack
A fracture topography with large slip steps is typical for opening displacement at maximum load.
microstructures with a highly planar slip distribution It was reported by these authors ~6that the parameters
and a large number of dislocations within a slip band dt and c, as well as o'j, E and p" all have a pronounced
while finer slip steps on the fracture surface are influence on the fatigue crack propagation rate. Other
indicative of a more homogeneous slip distribution. material properties, like the cyclic yield stress o'y and
The fatigue crack propagation curves shown in Fig. 4 the cyclic strain hardening exponent n' were found to
for the two microstructures, both containing shearable have relatively minor importance in determining the
precipitates (100 h 20 ~ and 0.5 h 100 ~ also can be fatigue crack propagation resistance, according to
explained qualitatively with the reversed slip model. At Eq. [1].
low AK-values almost the same resistance against crack Unfortunately for the microstructures studied in the
propagation was found for both aging conditions and present work no low cycle fatigue data are available and
also the fatigue fracture surfaces did not show signif- therefore only a tentative evaluation of the models
icant differences (Fig. 8). It is concluded therefore that mentioned above ~5-n7can be made with regard to our
the degree of slip planarity between both microstruc- experimental results. First of all no differences in crack
tures may be similar and the coincidence between both propagation rates between the different microstructures
fatigue crack propagation curves (Fig. 4) reflects the should be obtained due to changes in the elastic
similarity in slip reversibility within the plastic zone at modulus, since no significantly different values were
the crack tip between both microstructures. observed (Table II). It will now be assumed that the
However, if one applies the same model to explain true tensile fracture strain c~- (Table II) may be an
the results obtained for microstructures which were approximate measure of the cyclic ductility coefficient
aged at 100 ~ for various times (Fig. 5), not even a c'f in the Coffin-Manson relationship. Lin and Starke, 18
qualitative agreement can be deduced. At low extrapolating Coffin-Manson plots for strain ampli-
AK-values the highest resistance against crack propa- tudes greater than 1.2 pct, obtained c~f-values for
gation was found for the microstructure which exhibited A1-Zn-Mg-Cu alloys which were of the same order of
the most homogeneous slip distribution (0.5 h 100 ~ magnitude as measured tensile ductility values. The
Indications for such homogeneous slip can be derived tendency of the curves in Fig. 5 at low aK-values may
from the high tensile ductility of cr = 0.56 (Table II) then be explained in a qualitative way. The reported
and from fine slip steps found on the fatigue fracture models predict decreasing fatigue crack propagation
surface (Fig. 7(a)) for this age-hardened condition. The P
rates with increasing ~j-values. 16,17 From Table II one
lack in agreement between the proposed slip rever- can see that cF increases with decreasing aging time at
sibility model (10) and the experimental results in Fig. 5 an aging temperature of 100 ~ (see also Fig. 1).
demonstrates that also other factors besides reversed Therefore the microstructure aged for 0.5 h at 1130~
dislocation motion have to be considered in order to should exhibit the highest resistance against crack
explain all aspects of the fatigue crack propagation propagation due to the high oF-value (0.56) as compared

1618--VOLUME 12A, SEPTEMBER 1981 METALLURGICAL TRANSACTIONS A


to both of the other conditions with lower oF-values. 3) For microstructures containing a constant volume
The experimental determined d a / d N - h K curves fraction but different sizes of shearable precipitates, the
(Fig. 5) follow at low ilK-values indeed the prediction fatigue crack propagation resistance in the low
in Eq. [1]?6 AK-regime was found to increase with decreasing
A comparison of the fatigue crack propagation curves particle size or increasing tensile ductility. This result
for microstructures with shearable precipitates, as cannot be explained alone on the basis of reversed slip
shown in Fig. 4, with regard to the proposed model, ~6 of dislocations within the plastic zone. Ductility for the
also shows qualitatively agreement with the prediction. different microstructures have to be taken into account.
Assuming again a correlation between e'j and the true Existing fatigue crack propagation models as well as
fracture strain ~r, no significant differences in crack crack closure concepts seem to explain in a qualitative
propagation rates should be observed between both way the increased fatigue crack propagation resistance
microstructures (100 h 20 ~ and 0.5 h 100 ~ since with increasing ductility.
the measured eF-values are identical for both aging
conditions (Table II).
ACKNOWLEDGMENTS
However the large differences in resistance to fatigue
crack propagation at low ilK-values, shown in Fig. 3 for The authors wish to acknowledge the assistance of
the three different microstructures, cannot be explained Mr. J. Endrejat in the experimental work. Thanks are
on the basis of the different ductility values (Table II). also extended to Alcan Laboratories, Banbury, U.K.,
Crack propagation rates predicted from Eq. [1] on the for providing the alloy, and to the Deutsche For-
basis of the observed ~r-values alone would lead to a schungsgemeinschaft for financial support of this work.
different ranking of the three microstructures, as com-
pared to the experimentally measured curves (Fig. 3)
REFERENCES
A consideration of crack closure as the rate con-
trolling mechanism also would incorporate the ductility 1. M. O. Speidel: High-Temperature Materials in Gas Turbines, p.
of the different microstructures. The extent of crack 207, Elsevier, Amsterdam, 1974.
closure depends very strongly on the true tensile 2. J. L. Robinson and C. J. Beevers: Met. Sci.J., 1973, vol. 7, p. 153.
3. P. E. Irving and C. J. Beevers: Mater. Sci. Eng., 1974, vol. 14, p.
fracture strain and the retarding effect increases with 229.
increasing ductility of a material. ~9Therefore the exper- 4. K. H. Zum Gahr and L. J. Eberhartinger: Z. Metallkd., 1976, vol.
imental results in Figs. 4 and 5 also could be explained 67, p. 640.
on the basis of the crack closure concept, while it would 5. J. Lindigkeit, G. Terlinde, A. Gysler, and G. LOtjering: Acta
Metall., 1979, vol. 27, p. 1717.
not adequately describe the ranking of the d a / d N - i l K 6. J. T. McGrath and R. C. A. Thurston: Trans. TMS-AIME, 1963,
curves in Fig. 3. vol. 227, p. 645.
7. A.J. McEvily and R. C. Boettner: Acta Metall., 1963, vol. 11, p.
725.
5. CONCLUSIONS 8. G.A. Miller, D. H. Avery, and W. A. Backofen, Trans.
TMS-AIME, 1966, vol. 236, p. 1667.
1) A comparison between underaged and overaged 9. H. Ishii and J. Weertman: Metall. Trans., 1971, vol. 2, p. 3441.
microstructures of an A1-Zn-Mg-Cu alloy, aged to have 10. E. Hornbogen and K. H. Zum Gahr, Acta Metall., 1976, vol. 24, p.
the same tensile yield stress, showed that the fatigue 581.
crack propagation resistance in vacuum is higher for 11. J. Albrecht, J. W. R. Martin, G. Ltltjering, and J. W. Martin:
Proc. 4th Int. Conf. on Strength of Metals and Alloys, Nancy,
conditions containing shearable precipitates as com- France, vol. 2, p. 463, ENSMIN, 1976.
pared to overaged structures with nonshearable pre- 12. A. Gysler, J. Lindigkeit, and G. Ltltjering: Proc. 5th Int. Conf. on
cipitates. This improvement is attributed in part to an Strength of Metals andAlloys, Aachen, Germany, vol. 2, p. 1113,
increased tendency for reversed slip of dislocations due Pergamon Press, 1979.
to the more planar slip behavior for conditions with 13. G. Ltitjering, T. Hamajima, and A. Gysler: Proc. 4th Int. Conf. on
Fracture, Waterloo, Canada, vol. 2, p. 7, 1977.
shearable precipitates. 14. A. Gysler and G. Llltjering: unpublished results, DFVLR,
2) Microstructures containing different volume frac- Cologne, Germany, 1979.
tions and different sizes of shearable precipitates, but 15. H. W. Liu and N. Iino: Proc. 2ndlnt. Conf. on Fracture, Brighton,
the same tensile yield stress and true fracture strain, p. 812, Chapman and Hall, 1969.
16. S. Majumdar and J. D. Morrow: ASTM STP 559, 1974, p. 159.
exhibited the same fatigue crack propagation resistance 17. S. B. Chakrabortty: Fatigue of Engineering Materials and Struc-
at low cyclic stress intensity factors. This behavior can tures, 1979, vol. 2, p. 331.
be explained with similarities in slip reversibility within 18. F. S. Lin and E. A. Starke, Jr. Mater. Sci. Eng., 1979, vol. 39, p.
the plastic zone at the crack tip which also resulted in 27.
identical fatigue fracture topographies for the different 19. S. Purushothaman and J. K. Tien: Proc. 5th Int. Conf. on Strength
of Metals andAlloys, Aachen, Germany, vol. 2, p. 1267, Pergamon
microstructures. Press, 1979.

METALLURGICAL TRANSACTIONS A VOLUME 12A, SEPTEMBER 1981-- 1619

You might also like