west Mifflin, Pennsylvania 15122. Abstract. A criterion for the initiation of subcritical crack growth at blunt notches and sharp defects was developed and applied ...
111~ 1
Parkins Symposium on
Fundamental Aspects of
, I
Stress Corrosion Cracking Proceedings of a symposium sponsored by
TMSfASM·MSD Corrosion & Environmental Effects Committee
held at the 1991 Fall Meeting in Cincinnati, Ohio, October 21-24, 1991.
(
1I ' I
Edited by
S. M. Bruemmer, E. I. Meletis, R. H. Jones, W. W. Gerberich, F. P. Ford, and R. W. Staehle 11
I
A Publication of
TMS
Minerals' Metals • Materials
c (.
111 ~(
•• <
3. V.J.Gsdgil, S.Mambiej & S.H.Kolster: Hydrogen Elleets on Material Behavior (eds N.R.Moody &: A W. 1bpmpson), TMS-AlMB, Karrendsle PA (1990) p.375. 4. V.1.Gadgii. MSmithers &: G.Hochortler: Mikrostrokturelle und Milaoanalytische Charak terisierong (eds V.1bien et al), DVMBerlin (1990), p.21 I. STRAIN ENERGY DENSITY - DISTANCE CRITERION FOR THE 5. T.I.Marrow, C.AHippsley &: I.RKing: Acts Metall &: Mat. VoJ.J9 (1991) p. 1367. 6. S.Mandriej: Scripta Metall &: Mat, Vol. 25 (1991) p.213. 7. A W.SleesM:)'k &: S.Mandziej: Scripta Metall &: Mat, VoJ.24 (/990) p. 7.
INITIATION OF HYDROGEN - INDUCED CRACKING OF ALLOY X-750 8.MLHolzworth &: MR.l.outhan Ir: Corrosion; VoI.24 (/968) p.1 10.
9. J>.Maulik &: I. Burke: Scripta MetaJl; VoI.9 (/975) p. 17.
M. M.Hall, Jr., D. M. Symons, J. J. Kearns 10. N.Narita, C.J.Altstetter &: H. K. Birnbaum: Metall Trans; \6/.13A (1982) p.3155.
II. D.Eliezer, D.G.Chakrapani, C.1.Altstetter &: E.N.Pugh:. MetaIJ Trans; Vol. lOA (1979)
Bettis Atomic Power Laboratory p.935.
westinghouse Electric Corporation west Mifflin, Pennsylvania 15122 12. P.C.Maxwell, A GoldOOrg &: 1.C.Shyne: Metal/ Trans; \61.5 (1974) p. 1305.
13. T-P.Pemg &: C.J.Altstetter: Acta Metal/; Vol.34 (1986) p. 1771.
14. T-P. Perng &: C.J. Altstetter: Metal/ Trans; VoI.18A (1987) p. 123.
15. S.singh &: C.Altstetter: Metall Trans; Vol. 13A (1982) p.I799.
16. T-P.Perng &: C.J.Altstetter: Materials Sci Eng; VoI.Al29 (/990) p.99.
Abstract 17. N.AMelsen: lournal ofMaterials, AS7M,. \6/.5, No.4 (1970) p. 794.
18. R.Uu, N.Narita, C.Altstetter, H.Birnbaum &: E.N.Pugh: Metall Trans; Vol.IlA (/980)
A criterion for the initiation of subcritical crack growth at p.1563.
blunt notches and sharp defects was developed and applied to /.9. E.I.Meletis &: R.F.Hochman: Corrosion Sci; Vol.24 (/984) p.843.
hydrogen-induced cracking of the Ni-base superalloy X-750. The 20. S.lani, MMarek, R. F.Hochman &: E./.Meletis: Metall Ttans; VoI.22A (1991) p.1453.
onset of crack growth is shown to occur when a critical strain energy density is attained at a distance from the notch and crack 21. MR.Louthan Ir, G.R.Caskey Ir. I.ADonovan &: D.E.RaW! Ir: Materials Sci Eng, Vol.
tips that is characteristic of the microstructure along the 10 (1972) p.357. .
prospective crack path. Rising load crack growth initiation data 22. AI. ~t Ir &: MR.l.outhan Ir: Metall Trans; \61. lOA (1979) p.1675.
were obtained using homogeneous hydrogen precharged notched and 23. AI. ~t Ir&: MR.l.outhan Ir: MetalJ Trans; \61. 13A (1982) p.2049.
fatigue precracked bend specimens. The notch root radius, grain 24. S.Mandziej: FOMSlWproject GNS44.Q704 (1985) unpublished resellrch..
size and hydrogen concentration were varied. The crack growth initiation loads were found to be dependent on both the notch 25. P.MKelly, A/ostsons, G.R.Blake &: I.G.Napier: Pbys Status Solidi A; VoJ.3/ (1975)
root radius and the bulk prechargedhydrogen concentration. p.77I.
These data were shown to be correlated using a critical strain 26. RK.Ham: Phil Mag; Vo/.6 (1961) p.JJ8].
energy at-a-distance (SEOA-D) criterion. Furthermore, an elastic 27. I.Kolts: Duplex Stainless Steels (ed R.ALula), ASM; Metals Park OH (/983) p.233.
plastic analysis of the strain energy distributions showed that 28. H.C.Rogers: Acta Met; VolA (1956) p.ll4. the critical strain energy density value is attained at one grain diameter from the notch and fatigue precrack tips. The 29. RGibala: TransTMS-AlME; VoI.239(1967)p.1574. mechanical and microstructural aspects of the crack growth 30. I.P.Hirth; Metall Thins, Vol. JlA (1980) p.86I. process and the relevance to hydrogen induced cracking are 31. H.D.Solomon de T.MDevine Ir.: Duplex StailJless Steels (ed R.ALuLa). ASAI; Metals
discussed. Park OH (/983) p.693.
32. A w.SJeesM:)'k &: S.Mandziej: Strength of Metals and Alloys (ICSMA-8), (eds P.O.
Kettunen et al) Vol.lIl, Pergamon Press, OxfOrd (1988) p.155.
JJ. S.MandrJej &: A w.SJes":Yk: Proo 7th OMAE Conf(eds MMSalama et al), ASME, New
Yor.k (1988) p. 103.
34. MSmialowski: Hydrogen in Steel; Pergamon Press, OxfOrd (1962) p. 154.
35. P. Rozenak de D. Eliezer: Metall Trans; Vol. 19A (1988) p. 723.
36. P.Rozenak&: D.E/ieter: Acta Metall; Vol. 35 (1987) p.2329.
37. S.Mandziej: unpublished MOrk.
38. AH. Heiss &: H.-I. Eckstein: Korrosions-bestandige Stahle (ed H.-I. Eckstein): Deutscher
Verlag Grondstoflindustrie. Leipzig (/990) p.89.
P.rkins Symposium on FIIndanienW Aapec:to> orSt..... CorTOOion Cr..kitlfl' 39. A W.SleesM:)'k &: S. T.Mandriej: Strength of Metals and Alloys (lCSMA-9). (eds D.G.
UlOId by S. M. Bruemmer. E. I. Meleli•• R. H. Jo...... Brandon et al) Pergamon Press, London (1991) p.l7l.
W. W. Getbetich. F. P. Ford. and R. W. Sl.uhle The Minera". Metal. & M.lenal. Society. 1992
230
231
450 0 F (232°C» and low temperatures (130°F - 200°F (54°C 93°C». The low temperature cracking mechanism is apparently hydrogen embrittlement due to absorption of the hydrogen produced by corrosion of Alloy X-750 in the water environment (5). Low temperature cracking is of particular concern since, although the low temperature crack growth is subcritical, the onset of crack growth in a precracked fracture mechanics test specimen is essentially immediate and very rapid once the stress intensity factor for initiation of environment-induced crack growth is exceeded. The reactor designer must consider the potential for environment induced failures of Alloy X-750 components due to preexisting defects. Precracked fracture mechanics specimens are therefor used to determine the minimum stress intensity factors necessary to initiate .environment-induced crack growth. However, since reactor components are carefully fabricated to avoid defects, another case of design importance is the potential for crack growth initiated from a notched geometry such as the thread root of a threaded fastener. There is a need, therefor, for testing of notched test specimens and development of an appropriate notch fracture mechanics. The purpose of the work reported here is to develop a fracture mechanics parameter and crack growth criterion capable of encompassing the threshold conditions for initiation of environment-induced crack growth in both precracked and notched Alloy X-750 specimens. Materials and Procedure The Alloy X-750 compositions and heat treatments used in this study are given in Table 1. 1 Table 1.
Materials Compositions and Heat Treatments
MATERIAL CONDITION
HEAT TREATMENT
HTH (Solutionized and aged]
Hot worked + 2000°F (1094°C) for 1 hr + 1300 0F (704°C) for 20 hrs
AM (Equalized and aged)
Hot worked + 1625°F (885°C) for 24 hrs + 1300 0F (704°C) for 20 hrs 232
HEAT
Ni
Cr
Nb + Ta
Ti
Ai
Fe
C
8
HTH
70.7
15.04
1.01
2.58
0.75
7.90
0.04
0.002
AM
73.6
.15.44
0.96
2.49
0.77
6.45
0.03
---
Minor Elements HEAT
Mn
Co
S
Si
P
cu
As
HTH
.06
.07
.001
.01
.001 27 ppm
AM
.08
.11
.007
.04
---
.02
---
The room temperature air-environment tensile properties of these heat treatments are given in Table 2. Table 2. Heat
Test direction
UTS (ksi)
Tensile Properties
(\)
Gage Dia. (in)
Gage length (in)
34
.507
2.000
34
.505
2.000
Yield strength (ksf)
Elongation
RA
(\)
HTH
Lonqitude
171
113
26.7
AM
Lonaitude
172
109
24
Both the HTH and the AM heat treatments result in equiaxed grain microstructures with the mean intercept grain sizes being 0.005 inches (0.127 mm) for HTH and 0.0008 inches (0.019 mm) for AU. The HTH precipitate microstructure is characterized by a fine dispersion of matrix strengthening gamma prime throughout the matrix and Cr-rich carbides along the grain boundaries. The predominate intergranular precipitates are semi-continuous MZ1 C6 carbides. The AH precipitate microstructure is characterized by a duplex gamma prime structure throughout the grain interior. The coarse gamma prime precipitates are formed during the 1625°F (8850C) equalization heat treatment while the fine precipitates form during the aging heat treatment at 1300"' (704"C). There is a zone on both sides of a grain boundary that is denuded of coarse gamma prime, although the fine gamma prime has precipitated in this zone. The grain boundaries have widely spaced MC carbides and MUC6 carbi,des. For convenience of testing, hydrogen precharged specimens were tested in air at 200°F (93°C) in lieu of testing in a pressurized hot water environment. This testing procedure was considered acceptable since the purpose of this study is limited to development of a mechanical criterion for fracture initiation. Previous testing had established the precharged hydrogen concentrations needed to produce threshold stress intensity factors equivalent to those found by testing in a hot water environment. Specimens were hydrogen .precharged in a hydrogen gas phase environment at 1300 0F (704°C) for 1.50 hours. Diffusion calculations indicate that this exposure time is sufficient to produce a uniform concentration of hydrogen within 233
'J the specimens. Vacuum extraction analysis was used to determine . ,:he hydrogen concentrations.
Table 4.
'.'racture initiation testing was conducted using the three point land specimen shown in Figure 1. flCURE 1: THREE POINT BEND SPECIMEN
~1.75";;-I
I nO·:Jli:~120"· IT [
Dlm.A-r
fl I~OO"D
precrack 0.0025 0.010 0.030
X.
at Maj: l~ad) (ksi in 41.4; 43.2 42.2; 44.1 49.4; 50.5 59.9 234
H content (ppm) 52 52 52 52
precrack 0.005 precrack 0.005
(ksi [in) 38.2 52.4 28.4 40.6
33 33 62 62
i
Extensive use has been made of the linear elastic fracture mechanics stress intensity factor, X, for the description of the minimum stress conditions necessary for the onset of environment induced subaritical crack growth. The critical value of X, Which is often treated as a material parameter, provides a one parameter criterion for the' onset of subcritical crack growth from a preexisting sharp defect such as a fatigue precr~ck. Gangloff (6) has provided a critical review of the corrosion fatigue crack growth literature and provides justification for use of the stress intensity factor for correlation of environment-induced·crack growth rate data~
of the fracture initiation tests are given in Table 3 for 'ondition HTH and Table 4 for Condition AH.
(XI
H content (ppm)
R.:
peyelopment of a Strain Energy Density - Distance Criterion
~esults
Root Radius (inches)
(X, at Max. Load)
Discussion
Results
Results of Rising Load Crack Growth Initiation Tests of Hydrogen Precharged Alloy X-750 Condition HTH
Root Radius (inches)
The stress intensity factors at the onset of crack growth, Kc ' were calculated for both precracked and notched specimens using the maximum loads and the analysis methods found in ASTM standard test method E399-83. Stress intensity .factors for the notched specimens were calculated in the same manner as those for precracked specimens except that the notch depth instead of crack length was used in the analysis. The results found in Tables 3 and 4 show that X. monotonically increases with increasing notch root radius with the precracked specimen ·providing a lower bounding value. The mode of fracture was intergranular for all test specimens.
;pecimens of the HTH heat treatment condition were tested as atigue precracked and as-notched using notch radii of 0.0025 ,nches, 0.010 inches and 0.030 inches (0.064 mm, 0.254.mm and '.762 mm, respectively). Specimens of the AH heat treatment 'ondition were tested as-fatigue precracked and as-notched using t notch radius of 0.0050 inches (0.127 mm). Hydrogen :oncentrations for each specimen are indicated in Table 3 and j'able 4. Rising load tests were conducted at t- load point lisplacement rate of 0.002 inches/minute (8.5xl0· mm/second), Ihich previous testing had shown to be sufficiently slow to >roduce results independent of loading rate. The load at which 'rack growth initiated was taken in each case as the maximum load. All load versus load point displacement test records ;howed linear response to the maximum load, except for the 'pecimens with the 0.030 inch (0.762 mm) notch radius. These ;pecimens displayed measurable nonlinearity before maximum load.
('able 3.
Results of Rising Load Crack Growth Initiation Tests of Hydrogen Precharged Alloy X-750 Condition AH
.,
The uniqueness of X and, consequently, its usefulness for correlation of crack initiation and growth data, rests on the self similarity of crack tip stress-strain fields. When similitude does not exist, as is the case among notches and between a crack and a notch, stress intensity factors can no longer correlate crack initiation events. In order to discover a fracture criterion that can overcome this limitation of the stress intensity factor, we consider the more fundamental energy parameters that may control fracture initiation events. In this paper, we develop a criterion for the onset of subaritical crack growth, for application to both notches arid sharp cracks, based on ~he hypothesis that fracture initiation occurs whe,n the strain energy in an elemental volume of material along the prospective crack path exceeds a critical value. The distance aspect:, of this strain energy at-a-distance (SEDAD) 'hypothesis is motivated by our frequent observation that the 235
')
tension of a macrocrack is preceded by nucleation of a Dsurface microcrack in the vicinity of the macrocrack tip. The !of strain energy as the mechanical parameter is motivated by Ilelief that it is more fundamental to the fracture initiation ,nt than is e~ther stress or strain. rain energy density, W, is the work, per unit volume of cerial, that is absorbed during application of a mechanical ,'ain: W(r,8)
~
·lr.'1
1
0(r,8)de.
(1)
-ain energy density is a spatially distributed parameter that be determined for both sharp cracks and blunt notches using field equations for the stress, a, and strain, €, !tributions. In the engineering applications of interest, :alized yielding occurs at notch roots due to stress and strain lcentration. Use of Equation (1) would therefor in principle luire an elastic-plastic analysis for the local stress and -ain distributions. However, Glinka et. al. (7) have shown It the elastic-plastic strain energy density, W, can be Iculated by first calculating the elastic stra~n energy lsity, W., then applying a plastic correction factor, Cp : I
(2)
WI!' = CpW",
, elastic strain energy density may conveniently b8 written as iUrn of two components:
wI/J = 12
1-2\1
E
(12
v
+
12
1 +\1
E
.,..2
(3)
oct,
't
q
•
~( . 8 e 38 .f2rCi Sln2'cos2'cos 2
train 38) • 2
Creager and Paris point out that their notch stress field equations reduce to those of the mathematically sharp plane crack in the limit that the notch radius approaches zero. Therefor, the stress intensity factor appropriate for use with these notch stress field equations is simply that of a sharp crack having a length equal to the notch depth. Substitution of the Creager and Paris equations into Equation (3) gives the following expressions for Wv ' Wd and W.:
w. • v
W
d
l+v Ki 2'1'1: Er
[~(1-V-2v2)COS2..!!), 3
K2
•
.
l+V_z [(L)2 + cosd!{.! + sin 2 ! 2'1'1:
Er
2r
2
w • .!.!! Ki {(-2...)2 ..
(7)
2
2'1'1:
Er
2r
3
2
,luation of Equation (3) requires a knowledge of the stress ;tributions about a notch. Creager and Paris (8) have provided ess field equations that are applicable to either an elliptic Ie or notch. Their equation~ appropriate to Hode I opening, ,ch are reproduced below, utilize a stress intensity factor, 1(., t are very similar to those for a mathematically sharp crack: 0,,"
(1
y
=
8 (l-sin!sin 38) - Lcos 38 ~ {COS1
(4)
~ [cos! (l+sin!sin 3e) + Leos 38 1 , .j2'1'1:r 2 2 2 2r 2
(5)
v ... 'ltr
2
2
2
236
2r
2'
- .!\I(l-V»),
2
(8)
3
+ cos 2 ! (1-2\1 + ain2!») ,
(9)
2
In these equations, (r,B) is radial position measured from the origin of coordinates, which is the focal point of an ellipse that is located at a distance ,/2 behind the notch tip. The radial distance from the origin is: 1
r" ~ [(cos 2 8 + 3)"2 - cose1 + 2
're the first term on the right hand side is the specific -ain energy due to volume expansion, Wy , and the second term is -specific strain energy due to distortion, Wd of an elemental urne of material. II) this equation, C1y is the volumetric -ess, (a x+o y+o.)/3, and T~t is the octahedral shear stress.
(6)
X,
(10)
In this equation X is the distance from the the notch tip and , is the angular position about the notch tip. Equation (2) can now be used to calculate the elastic-plastic strain energy density. The correction factor for notch tip plasticity, Cp , as derived by Glinka is given by: C •
2 (-L) +(-L)2 2rp
2rp
(11)
1+{-L) 2Ip
I'
Glinka provided an expression for the size of the notch plastic zone, rp' A simplified approximate form was used in the present work: Ep =
£. ( 2
K3,,) 1/3
ps/ 231
'
(12)
(".
where Sy is the tensile yield stress. Equation (12) is to be used with notches only and has reasonable accuracy for notch sizes of engineering importance. The difference between the notch tip plasticity correction factors (Equation (11) calculated using Equation (12) and the more accurate expression for the notch plastic zone size given by Glinka is less than about 8\ for the smallest notch used in our study. According to the SEDAD criterion, the onset of crack growth occurs when the strain energy density exceeds a critical value, wc' at a characteristic distance, Xc' Therefor, the stress intensity f.actor· for initiation of crack growth at notches, Kc' can be obtained by combining Equations (2), (9) and (11) and evaluating at Xc (or r c)' Wc and e. where e. is the, direction of , the maximum strain energy density at the characteristic distance. The expression for Kc so obtained can be rearranged to obtain a normalized expression that is convenient for anlysis of the data: K." en
(,~ ,"-,---- i ~ 2nBW~(/F(P'X,,), I /J 1 c (l+V)C,", l'
fj
where
.ttt F(p,Xc )
,
Xc'
11
(J
I} 1
,r . ' I
(13)
.
".,'"
•
n'
.(
lill
I
.('~:-'~[(..L)2 XC , 2Ie
+ COS 2
(
0.) (l-2v + sin2 2
(
~"»
J,
(14) ,
and rc is obtained by evaluating Equation (10) at Xe and 8•• The angular position, e., about the notch where the maximum strain energy density occurs, is a function of the notch .root radius, p, and the characteristic distance, xc' Application to Test Results We begin' with Figure 2, which, for the Condition 8TH specimens, shows the critical stress intensity factor data from Table 3 plotted as a function of the square root of the notch root radii. Note that, although there is a very good fit of the notched specimen data, the data 'for the precracked specimens, when plotted at zero notch root radius, are not well correlated with those of the notched specimens. As shown in this figure, the precracked specimens behave as if blunted to an effective notch root radius of about 0.0018 inches (0.046 mm). However, this is an order of magnitude larger than the crack tip opening predicted by the
FIGURE 2, THE EFFECT OF NOTCH ROOT RAi'HUS ON THE CRITICAL STRESS INTENSITY FACTOR FOR H CHARGED CONI)'TlON HTH AllOY X- 750
·60
,z ~
"0
...'"
55
~
'iii 50 c:
£'"
., 45
o
~
~
results of elastic-plastic crack mecilanics (9). Therefor, crack tip blunting cannot directly account for the rather large effective notch root radius of the fatigue precracked specimens. Cottrell (IO) and others (11,12) have observed that the toughness of notched materials no longer decreases with decreasing notch root radius when the notch size becomes less than a size that i. characteristic of the material microstructure. Tetelman (ll) discusses these observations in terms of the microstructural features that control the fracture mechanisms. For example, Tetelman asserts that, for a cleavage fracture mechanism, the plastic deformation processes that initiate cleavage require a minimum volume ot the order of the grain size in which to operate. Based on this reasoning, Tetelman concludes that regardless of the notch size, the effective notch plastic zone size must be at least equal to the grain size for tracture initiation to occur. For a parallel sided notch the results ot elastic-plastic notch mechanics show that the effective notch radius must be about one-quarter of the grain size to obtained a plastic zone size equal to a grain diameter. The hydrogen precharged AlloyX-750 specimens in our study tail intergranularly. Assuming that the plastic deformation processes that initiate this mode of fracture require a minimum volume ot the order of a grain size, Tetelman's requirement on the plastic zone size and consequently the effectiVe notch size should apply to our Alloy X-750 data. The data shown in Figure 2 indicate an effective notch radius of 0.0018 inches (0.046 DUll) for the fatigue precracked Condition HTH specimen which is in reasonable agreement with the Tetelman requirement as theHTH material has a grain size of about 0.004 to 0.006 inches (0.102 to 0.152 _). OUr-SEDAD requirement for the onset of crack growth is similar to Tetelman's requirement for clevage fracture initiation. As we will see below, the strain energy density must exceed a critical value, which exceeds the strain energy tor plastic yielding, and that this must occur at a characteristic distance, which is one grain diameter. 1 FIGURE J: CORRELATION or CRITiCAl STRESS I INTENSITY fACTOR
DATA USING THE STRAIN
ENERGY DENSITY CRITERION Figure 3 shows the goodness of
fit provided by the SEDAD ~'
expression for a normalized'~
critical stress intensity factor, -E
Ken' as a function of the square ~
~Model_ root of a notch root function, ~
Predic.tion F(p,X...). Shown are the Condition iii HTH aata, Table 3, which were '3 first plotted in Figure 2, plus :'i' u~ the Condition AU data, Table 4. ]
~/
Opan -.ymbo1a ...
35
"E ~
o
II\ot~NO
tymbol. - Pf'Mto
