Skip to main content

Full text of "Cumulative creep fatigue damage in 316 stainless steel"

See other formats


N91-24312 

CUMULATIVE CREEP FATIGUE DAMAGE IN 316 STAINLESS STEEL 



Michael A. McGaw 
NASA Lewis Research Center 



The cumulative creep-fatigue damage behavior of 316 stainless steel at 
1500 "F has been experimentally established for the two-level loading cases of 
fatigue followed by fatigue, creep fatigue followed by fatigue, and fatigue 
followed by creep fatigue. The two-level loadings were conducted such that 
the lower life (high strain) cycling was applied first for a controlled number 
of cycles and the higher life (low strain) cycling was conducted as the second 
level, to failure. The target life levels in this study were 100 cycles to 
failure for both the fatigue and creep-fatigue low life loading, 5000 cycles 
to failure for the higher life fatigue loading and 10 000 cycles to failure for 
the higher life creep-fatigue loading. Results of the fatigue followed by fa- 
tigue loading experiments follow the behavior predicted by the double linear 
damage rule and deunage curve approach of Manson and Halford (ref. 1), while 
the remainder of the two-level experiments produced results that were inade- 
quately predicted by this approach. These experiments all involved creep- 
fatigue loading and likely introduced a different damage mechanism from a 
cumulative damage stai^dpoint. The failed specimens are being examined both 
f ractographically and metallographically to ascertain the nature of the damag- 
ing mechanisms that produced failure. Models of creep-fatigue damage accumula- 
tion are being evaluated and knowledge of the various damaging mechanisms is 
necessary to epsure that predictive capability is instilled in the final fail- 
ure model. 



EXPERIMENTAL RESULTS AND DISCUSSION 

A strainrange partitioning (SRP) characterization of this material was 
performed for the PP and CP damage modes at 1500 "F. While a complete SRP 
characterization for this material at this temperature has been established by 
Kalluri (ref. 2), the emphasis in that work was on the effect of exposure time 
on the low-cycle fatigue life; the life data so developed are over a range 
that was considered too narrow for the purposes of the present cumulative 
creep-fatigue damage study. Since, in fatigue, the nonlinear cumulative ef- 
fects depend strongly on the relative life levels involved in the loading, the 
present study is focused on obtaining failure data at two widely separated 
life levels, both under pure fatigue and creep-fatigue loading. To this end, 
baseline SRP experiments were performed to establish the inelastic strain 
range levels for 100 and 5000 cycles to failure in PP and 100 and 10000 cycles 
to failure in CP. Results of these baseline characterization experiments are 
shown in figure 1. The resulting life relationships expressed in terms of in- 
elastic strain range versus cyclic life are substantially different in the low- 
er life regime and appear to approach each other in the longer life regime. 



PRECEDING PAGE BLANK NOT FILMED " 



These life relationships were used to determine the life levels corresponding 
to the loading levels applied in the two-level loading experiments. 

The results of the two-level loading experiments involving PP followed by 
PP loadings are shown in figure 2. Substantial cyclic hardening was observed 
at the low life level, which served to influence the cyclic behavior observed 
at the (subsequent) high life level. The material response at the second load- 
ing level was different from that observed for experiments involving that load- 
ing level alone, in the respect that it exhibited much higher stresses for the 
applied strains. As the second level cycling progressed, the material cycli- 
cally softened and tended to approach the stabilized cyclic response observed 
in the baseline fatigue characterization experiments. The failure data in fig- 
ure 2 are well represented by the damage curve approach of Manson and Halford 
(ref. 1), and are unconservatively represented by the linear damage rule of 
Miner (ref. 3). 

The results of the two-level loading experiments involving PP followed by 
CP loadings are shown in figure 3. In this case as well, the influence of the 
substantial amount hardening induced by the first level, lower life cycling 
manifested itself in terms of a somewhat different cyclic response at the sec- 
ond, higher life level, CP loading. In the second level the material exhibit- 
ed much higher cycle times, which, as the cycling went on, tended to approach 
the cycle times observed in the baseline CP behavior. The data in figure 3 
show a trend that is reasonably represented by the linear . damage rule. A pre- 
diction based on the damage curve approach is also shown, which, while provid- 
ing a very conservative estimate for the results, does not adequately repre- 
sent the observed behavior. The presence of creep damage in the second load- 
ing level appears to influence the cumulative damage behavior, implying that 
another damage mechanism is operative. Fractographic and metallographic 
analysis of the fractured specimens is currently underway to. ascertain the 
nature of the damaging mechanisms operative in this type of cumulative fatigue 
loading. 

The results of the two-level loading experiments involving CP followed by 
PP are shown in figure 4. In this series of tests the effect of the CP low 
life level cycling on the second, high life level PP cycling was to introduce 
a significant tensile mean stress as well as to produce a "harder", cyclic 
stress strain response than was observed in the baseline PP experiments at the 
same applied total strain range. This mean stress tended to relax somewhat as 
the second load level cycling progressed, but a significant mean stress was 
nevertheless maintained throughout the test. The life data shown in figure 4 
were calculated assuming R = -1, and the results of these experiments indi- 
cate that the influence of the earlier CP cycling on the subsequent PP cycling 
is, at worst, virtually insignificant, and, at best, an enhancer of fatigue re- 
sistance at the second level. These results also indicate the possibility of 
another damage mechanism. Fractographic and metallographic analyses are in_ 
progress. 

The results obtained thus far indicate that for this material in cumula- 
tive creep fatigue loading, the interaction between PP and CP loadings is not 
as deleterious as the interaction that occurs between PP and PP loading. This 
result suggests that another cumulative damage mechanism is operative when CP 
loading is followed by PP loading or vice versa. While this' is not surprising 
(the SRP approach is predicated on the possibility of different damaging modes 
of straining), the direction of the influence in the cumulative case, is an 



34 



unexpected experimental result. Future work will include characterization of 
the PP and PC and the PP and CC cumulative damage behaviors so that the devel- 
opment of a physically based model that describes the general cumulative da- 
mage behavior for this material under conditions of fatigue and creep-fatigue 
loading is achieved. 

REFERENCES 

1. Manson, S.S.; and Halford, G.R.! Practical Implementation of the Double 

Linear Damage Rule and Damage Curve Approach for Treating Cumulative 
Fatigue Damage, Int. J. Fract., vol. 17, 169-192, 1981. 

2. Kalluri, S. : Generalization of the Strainrange Partitioning Method for Pre- 

dicting High Temperature Low Cycle Fatigue Life at Different Exposure 
Times, Ph.D. Dissertation, Case Western Reserve University, Cleveland, 
Ohio, 1987. 

3. Miner, M.A. : Cumulative Damage in Fatigue, J. Appl. Mech., vol. 67, 

A159-A164, 1945. 



35 



SRP PP AND CP CHARACTERIZATIONS 

316 SS AT 1500 op 



.1000 



.0100 



STRAIN 



.0010 



.0001 



10 



O PPDATA 
A CP DATA 




100 1000 10 000 100 000 

LIFE, Npp OR Nqp 



Figure 1 

PP PLUS PP INTERACTION 

316 SS AT 1500 "F 



C0-e9-39293 



1.0 



\ 



O PP + PP DATA 

DAMAGE CURVE APPROACH 



.6 



REMAINING 

LIFE 
FRACTION, 

n2/N2 .4 



.2 




.2 .4 .6 .8 1.0 

APPLIED LIFE FRACTION, n^m^ 



CD-89-39294 



Figure 2 



36 



PP PLUS CP INTERACTION 

316 SS AT 1500 op 



1.4 
1.2 

1.0 k 

REMAINING q _ 

LIFE 
FRACTION, 

n2/N2 



O PP+CP DATA 

DAMAGE CURVE APPROACH 



LINEAR DAMAGE RULE 




.2 .4 .6 .8 1.0 

APPLIED LIFE FRACTION, n^fU^ 



C0-8»-3»2»5 



Figure 3 

CP PLUS PP INTERACTION 

316 SS AT 1500 »F 



3.2 
2.8 
2.4 

REMAINING ^^ 

LIFE 
FRACTION, 1.6 

n2/N2 

1,2 

.8 

.4 



O 



o 



o 



CP + PP DATA 

DAMAGE CURVE APPROACH 



LINEAR DAMAGE RULE 




.2 .4 .6 .8 

APPLIED LIFE FRACTION, n^/N^ 

Figure 4 



CD-e?-3?2M 



37