ML20084Q686
| ML20084Q686 | |
| Person / Time | |
|---|---|
| Site: | Nine Mile Point  |
| Issue date: | 04/30/1970 |
| From: | Cheng C ARGONNE NATIONAL LABORATORY |
| To: | |
| Shared Package | |
| ML20084P985 | List: |
| References | |
| NUDOCS 8306130538 | |
| Download: ML20084Q686 (15) | |
Text
l 1
O O
i Attachment B INFORMATION NEEDED ON INTERGRANULAR CORROSION CRACKING C. F. Cheng i
l i,
Materials Science Division Argonne National Laboratory j
Argonne, Illinois 60439 i
i i
i I
,1 i
4
\\
4 4
i I
.i 1
i!
- i i
t k
~
[
April 1970 l
0306130538 700520 l '*
PDR ADOCK 0500022 P
- )
i;
I' O
O INFORMATION NEEDED ON INTERGRANULAR CORROSION CRACKING 1.
Introduction Attachment A, Intergranular Corrosion Cracking of Type 304 Stainless Steci in Water Cooled Reactors, is a review of the published literature.
This memorandum suppicments attachment A by discussing:
(a) unpublished reports on recent intergranular failures of components in water-cooled reactors, (b) the status of sensitized stainicss steci (Types 304,-304L,
^
316, and 308 wcld overlay) safe-ends, and i
(c) mechanical properties for evaluation of'intergranular i
corrosion cracking aided by low-cyclic and/or static stress.
I i
2.
Recent Intergranular Failures in Water-Cooled Reactors In a meeting on March 19, 1970, the Division of Compliance reported that the primary system piping of three boiling water reactors (BWR)'
recently exhibited intergranular corrosion fractures. These cracks propagated from the water-side outward and the failures were attributed.
[
to stress corrosion and/or low-cycle fatigue. 'In all three cases, the I
piping was fabricated from Type 304 stainless steel that was sensitized J.
l
- at a temperature of 621*C (1150*F) for 4-10 hr.
These reactors are l
{
located in (a) Elk River, Minn., (b) Lacrosse, Genoa,1Eis., and -
4 j
(c) Nine Mile Point, Oswego, N. Y.
1 2
2.1~ Elk River BWR I
?
Intergranular corrosion cracks at-three locations were observed in; 4.
the primary system piping in the Elk River BWR. The cracks were located at (a) the nozzic safe-end in the upper liquid icvel line, (b) the core.
.0 f
k t:
+
p
(',,
V V
spray safe-end, and (c) a steam nozzle, af ter 41 thermal cycles of operation. These areas were exposed to stagnant steam and/or droplets of steam condensate.
In each case, the cracks were in the vicinity of weld deposits, but they did initiate at the heat-affected. zone. The oxygen in the steam phase is estimated at 10-20 ppm.
The magnitude of l
the stress was not known.
2.2 Lacrosse BWR The transition piece in the schedule 80 feed water pipeline of the Lacrosse BWR also exhibited intergranular corrosion cracking after some service. The transition piece was exposed to the stagnant water
(<5 ppm oxygen). The cracks were located between the thermal sleeve l
and the pipe. The magnitude of the stress was reported to be well l
1 above yield strength.
2.3 Nine Mile Point BWR Intergranular fracture was observed at the core spray safe-end of I.
the Nine Mile Point BWR, af ter 11 thermal cycles of operation. The t
crack was adjacent to the thermal sleeve and exposed to the stagnant water (<5 ppm oxygen). The calculated stress after installation was estimated to be approximately 60,000 to 100,000 psi compared with the l
originally designed stress of about 14,000 psi.
The details of the t
j failure analysis are still in progress at the GE Laboratory and BNWL.
v I
3.
The Safe-end Status in Water-cooled Reactors i
j Tables I and II summarize the status of stainless steel safe-ends in water-cooled reactors (licensed and license-pending, respectively).
This information, furnished by the Division of Compliance, is tentative and will be verified shortly.
4
l l
g n
i
(
')
\\'
3.
In the case of the licensed BWR (Table I), intergranular corrosion cracking occurred in some of the welded safe-ends of Types 304 or 304L stainless steel. The material was stress relieved at a sensitizing temperature of 621*C after welding. There are no failures of safe-ends g.Jb in licensed EWB to date.
The welded safe-ends of Types 108L, 304, and ?'
s p
316 stainicss steel in the latter reactors were covered with~308/309 stainless steel weld _ overlay. The forged Type 316 stainless steel safe-ends in the licensed PWR were fabricated by svagging or by welding the end joints, llowever, it has not been resolved that the reported failures in the BWR are attributed to l
(a) chemical composition of the metal, (b) carbide morphology, 1
(c) radiolytic decomposition of water, l
(d) stress Icvel, or a combination of the variables.
If carbide morphology were an important factor, the welded Type 316 stainless steel safe-ends in the sensitized condition may also exhibit failure similar to that experienced with Type 304 stainicss steel.
In this event, an evaluation of welded Types 304 and 316 stainicss steel safe-ends in license-pending water-cooled reactors will be necessary.
l If oxygen concentration were the contributing factor, high nickel-base i
alloys (e.g., Inconel 600) must be avoided as alternate safe-end I
j material. The opinion is based on stress corrosion da'ta summarized in
~
Table III and illustrated in Fig. 1.
4.
Evaluation of Intergranular Corrosion Cracking by Mechanical Properties Intergranular corrosion cracking of austenitic stainless steel can be avoided only if the components were operating below the critical static and/or cyclic stresses for that specific environmental condition.
4
4
-l-
- g O
O r
4.
The following paragraphs will claborate on the mechanical properties relating to stress corrosion and low-cycic corrosion fatigue.
4.1 Stress Corrosion The kinetics of stress corrosion processes have been studied almost exclusively in terms of the total time to fracture, T. The term T f
f shows an Arrhenius relationship with temperaturc6,,f,77,wg (1) log T = log T, +
e f
where T, is some constant, T is the absolute temperature, R is the gas constant, and At is the activation energy for the rate-controlling process. T, however, is a summation of two further terms.
f T
=T
+T (2) f n
p, where T represents the induction period required to nucicate cracks, and T is the time required to propagate the crack to failure.
In P
t many systems T r pr sents a large fraction of T, whereas T often n
f p
does not vary from a small fraction of T. Kohl developed an elegant f
)
technique to determine T for austenitic stainless steci under constant loading in boiling magnesium chloride at atmospheric pressure. Time-
~
~
j clongation or " creep-curves" were utilized, when the extension of a i
cracking specimen c, or.the extension rate dc/dt was observed,8 o
f.
These principles ucre used by Wilde to study sensitized Type 304 9
x I
stainicss steel in 289*C water that contained 100 ppm oxygen. He l
conducted a series of tests' to determine the influence of the' applied L
. stress (aapp) on the induction period for crack nucleation.- A typical 4
trace is shown in Fig. 2 for o
= 29.1 kpsi. After the initial p
I w
d-
^
7 q
6.
5.
cxtension associated with applying the load, a quiescent period of 5.3 hr was observed (Fig. 2, I), af ter which periodic extension cecurred (Fig. 2,11 and III). This " stair-step" extension behavior was observed in all tests and appears to be characteristic of this intergrantilar corrosion cracking process.
In addition, he found the influence of a on T decreases markedly as o increases from app n
app 24 kai to 30 kpsi. The apparent independence of T below 24 kpsi is n
not understood at present, although it'has been observed that no cracking occurs at stresses below the engineering yield strength (21.2 kpsi)- over periods of many hundreds of hours.
For the present, this reviewer will l
assign the critical static stress for stress corresion cracking at the change of slope of a The above mentioned method of constant loading to determitic critical static stress for stress corrosion cracking is' practical only for exaggerated test conditions.
For example, if'the' solution-annealed Type 304 stainicss steel were tested in 289'c water containing 10 ppm i
oxygen, the T, for critical static stress may be several-thousands of hours. A more rapid test is definitely needed.
A short-term tensile k-test could be used, since the results could be correlated with the i
long-term result by the constant loading method. Essentially, the-I procedure for the rapid test consists of conducting a tensile test at 4-i i
constant strain rate in a controlled environment. A rating of stress i;
corrosion cracking susceptibility (S) can be made by comparing time, 4
['
ultimate tensile and clongation at fracture between the values.(Mgy) in
~
(
fnert atmosphere, and the corresponding. values (M
) in-the corrodent.
S is defined as S=
x 100.
-Mh w
L
i j
O O
t e.
10 This approach has been used successfuny by Orman to study stress corrosion cracking of Type 321 stainless steci in chloride and nitrate solutions.
In addition, the short-term tensile test can be used to 9
correlate with low-cycle fatigue data.
Berling and Conway recently developed an equation of " characteristic slopes," which satisfies the relationship of total strain range versus fatigue life. The tests were conducted in 650*C (1200*F) air at a strain rate of 4 x 10-3,gc-1 for Types 304 and 316 stainless steel.
This approach is particularly useful to predict corrosion fatigue properties of in-reactor surveillance specimens in postirradiation tests.
4.2 Low-Cycle Corrosion Fatigue Low-cycle fatigue can occur after relatively few cycles as a result of stress levels sufficiently high to produce measurable cyclic plastic s train. (12) For Type 304 stainless steel under 1% alternating strain, the number of cycles to failure at room temperature (computed from Manson's Universal Slope equation (13)) is 1.2 x 10. (14) 3 Normally, the effect of corrosion may not be detected by a standard low-cyclic fatigue test at 3-6 cpm.
For example, the Naval Research I
Laboratory (15) conducted fatigue-crack propagation tests on 9 wt% Ni-4 wt%-
Co-0.25 wt% C, 12 wt% Ni maraging steel, and 18 wt% Ni maraging steel I
in dry air and 3.5% Nacl solution.
Single-edge-notched specimens were i,
cycled zero-to-tension in cantilever bending.
Significant differences were observed in the fatigue-crack, growth-rate characteristics of the three types of steel at the same yield-strength ' level (180 ksi). The fatigue-crack growth in all three types of' steel was acceldrated by the Nacl environment, but only to a limited extent. The greatest increase in crack growth rate was less than one order of magnitude and occurred i
i j
(,
(y V
C) under low stress intensity (AK = 40 ksi/in. ).
The crack-growth rates were similar in air and Nacl at high stress intensity (AK = 100 ksi/in.2),
There was no correlation between the fatigue-crack growth behavior in l
Nacl and the stress-corrosion cracking parameter (KI
) for these steels. This was attributed to the fact that cyclic loading at 5 cpm did not allow sufficient time for stress-corrosion cracking, which is time dependent, to cause any significant effect on the fatigue-crack growth in these steels.
3 This is the reason, in the case of primary coolant system rupturc study, GE chooses to conduct the low-cycle fatigue test in reverse bending at only 2 cycles /hr.( ) The test is to be conducted in 286*C primary water at the Dresden-I m m.
The test will take 2-1/2 years to
{
complete and, therefore, will allow time for corrosion, if any, to j
affect crack initiation and propagation growth.
When low-cycle corrosion fatigue tests are made in the laboratory, it is only practical to accentuate the conditions conducive to the reduction of the fatigue life. The approach GE (NSP) used to investigate the numerous variables of f atigue lifn of Type 304 stain 1 css i
stcol is particularly suited to the study of low-cycle corrosion fatigue.
i These variables include strain amplitude, strain rate, temperature, I
j loading systems, hold time, and strain wave form.
For example, tests j
conducted at a temperature of 650*C in air have shown that a tension-hold-only wave form results in a greater reduction in fatigue life than otlier wave forms (e.g., triangle, compression-hold-only, or tension-hold and compression-hold); the longer the hold period the greater the reduction in fatigue life.
Furthermore, only specimens from Lension-hold-only wave
\\u
'l 1
e
r-4 l
/3
,O
(/
U 8.
form exhibited intergranular fracture, while those from other wave forms resulted in transgranular fracture.
In another analysis, linearity was noted in a logarithmic plot that related the time-to-fracture to (a) stress amplitude, and (b) length of hold at a fixed strain rate. These relation-ships enahic the time-to-fracture to be calculated for any stress amplitude or holding time.
In addition, there is a linear relationship (logarithmic coordinates) between the time-to-fracture and the time-to-one cycle at
~
a fixed strain rate.
In which case, time-to-fracture at other strain rates, with or without hold periods, can also be obtained.
It is proposed that stress and strain analysis be performed on the reactor component of interest, which will a.11ow the experiments on low-cycle corrosion fatigue to be related to in-reactor operating conditions. The objective is to set up a standard procedure to determine the low-cyclic corrosion fatigue strength, so that p, roper limits can be assigned for specific applications.
i-f l
l r
I i
- l r
s 4
j' O
O References 1.
- 11. R. Copson and S. W. Dean, Corrosion, Vol. 21, No.1, p.1 (1965).
2.
II. 'Coriou, L. Crall et al., Corrosion 21, p. 280 (1966).
3.
II. R. Copson and G.. Economy, Corrosion 24, p. 55 (1968).
4.
J. O. Edstrom and'L. A. Forsman, Proc. of 3rd Intl. Congress on Metallic Corrosion, Vol. 2, p. 292, MIR Publisher, Moscow, USSR (1969).
5.
J. S. Armijo, Corrosion, p. 319 (1968); Jersey Central Power and Light Co.,
USAEC Docket.No. 50-219, Amendment No. 37 (April 1968).
]
6.
- 11. Tornes, Werkstoffe und Korrosion, p. 729 (1963).
7.
- 11. Kohl, Workstof fe und Korrosion, p. 364 (1964).
8.
- 11. Kohl, Corrosion, Vol. 23, p. 39 (1967).
9.
B. E. Wilde, Corrosion, Vol. 25, p. 359 (1969).
10.
S. Orman, Corrosion Science, Vol. 9, p. 849 (1969).
11.
J. T. Berling and J. B. Conway, TMS-AIME, Vol. 245, p.1137 (1969).
12.
L. F. Coffin, Jr., Experimental Mechanics, Vol. 8, p. 218, May (1968).
13.
S. S. Manson and G. R. IIalford, Proceedings Internal Conference on Thermal and liigh-Strain Fatigue, Metals and Metallurgy Trust, London, p.154 (1967).
~
~
]
14.
C. G. Collins, J. Moteff and B. A. Chandler, GEMP-573 Revised, GE Co.
J June 1969.
15.
T. W. Crooker and E. A. Lange, NRL Report 6761 (Sept. 1968).
16.
Reactor Primary Coolant System Rupture Study, Progress Report No. 9 CEAP-5512 (August 1967).
[-
'17.
J. B. Conway, J. T. Bealing and R. II. Stentz, GEMP-702, GE Co. (NSP).
~
(June 1969).
- 18. - AEC Fuel and !!aterials Development Program - Progress Report, GEMP-1012, (March 1969).
9 4
m.
+
y C
O So U
STRESS CORROSION CRACKING IN WATER + OXYGEN TEMPERATURE-316 C CREVICED U-BEND SENSITIZED CONDITION AS V/ELDED A
A A
l' 7
(/)
LL IL z
x zX zy 3
ljj N
/
NF-NO FAILURE 9
z"
/
N
/
-(6 N
/
h io N
/
?
/
7 Q
N
/
X
/
)
/
2 6
N
/
x o
N
/
X
/
x
/
x ua x
y y
/
X m.
N f
'd N
y X
/
X m
x x
/
o N
/
X
'X x
y v-p co X
N N
/
g Es N
/
'(
x
/
$, l a Z
/
g 1
N j
W X
N X
/
w
\\
/
X N
X
/
8 N
/
X
/
X o
S X
\\
/
X
/
i w
X W
D
\\
/
X N
X o
o uh l o
o f m o N o /
m u) m N
o o /
u>
o g a m
o o
m m
e x m / e s
e
/
m e
i e
e / e e x d x $,y O id
$ / 8 d N
/ 8 U
d
. d /
,\\
g
/ E X B N "
/ g X g 5
1 g f
>. X.
o N 8 / g B N ss 6 X
? -\\
/
o o
> X o
o o
o t-E E /
E /
i-I z N M y i-X E K l
s
' i
~3 ppm O
~l4 ppm O
~54 ppm O Fig.i
p.
O:
4_ y:
e i
i II I
IS CON flNut 010 fg
- P6 2 hee
,~*
gg e
2fI E
- , S u
E tw F-.
i 1
1 I
4 '.
O IO 20 2S v iuc...
1 Fig. 2.
. Extension / time plot for-a sensitized Type 304 stainicas steci sampic, stressed to 29.2 ksi in water
+ 100 ppm 02 at 289'C.(9)-
l
- i....,....
w.N.
o e
i 4
g es t
c 4
io.
4 e a etial e
e i eeaei 3
400 to ucanioN iiuc.
Fig. 3.
The variation in~nuc1 cation j -
time with applied. stress. for the cracking of sensitized Type 304~
stainicss steci in' water + 100 ppm 0 at 289'C.(9) 7 1
3=
i.
4 5,
t:
r
-- i
-4,
-e
=
,w,,,
,,, -. < we,,-
.,-r-rreyer,e,.
ie=-r s
,*g eg
=
y
j O
Q s.g L
TABLE I.
Safe-End Status in Water-Cooled Reactors - Licensed Stainless Sensitized Weld Metal No. of
,N'ime, Location, Operato.r.
Steel 0 NJ150*F (Safe-end to Vessel)
Nozzles (a) Boiling Water Reactor Dresden I 304L' Yes 308 SS 34 Grundy County, Ill.
Com. Edison Co.
Big Rock Point 304 Yes 308 SS 15 (Vessel)
Charlevoix County, Mich.
11 (Steam Drum)
Consumers Power Co.
ilumbolt Bay 304 Yes 308 SS 20 Eureka, Calif.
Pacific Gas & Elec.
Lacrosse 304 Yes Inconel 182 11 i
~
Genoa, Wis.
Diary Power Corp.
Oyster Creek 1 304 Yes 308 SS
(?)
Oyster Crock, N. J.
clad with before Jersey Cen. Power & Light 308 cladding Nine Mile Point 304 Yes 308 SS 31 Oswego, N. Y.
j
. Niagara Mohawk Power Co.
j i
Dresden 2 316 Yes
(?)
27 Grundy County, Ill.
Com. Edison Co.
g Elk River 304 Yes 308 SS 1
Elk River, Minn.
e 4
Elk River Rural Coop. Power f
Assoc.
(b) Pressurized Water Reactor 1
i Indian Point 1 308L Yes 8
Buchmann, N. Y.
clad with (precoolant)'
Com. Edison Co. of N. Y.
308/309 4
Yankee 316.
Yes 308L SS-None i
Rowe, Mass.
i Yankee Atomic Elec. Co.
Conn. Yankee 316 Yes Inconel 182 8
Haddan Neck, Conn.
forging i
Conn. Yankee Atomic swagged i
Power Co.
j.
1 i
l.
v
i
- o o.'
TABLE.I.
Safe-End Status in Water-Cooled Reactors - Licensed (Contd.)
Stainleso Sensitized Weld Metal No. of
'Name,' Location,' Operator Steci
@ N1150*F (Safe-end to Vcosc1)_
Nozzles (b) Pressurized Water Reactor (Contd.)
Cinna 316 forging No
-316 SS 6
Ontario, N. Y.
(welded ' ends) -
Rochester Cas & Elec. Co.
Saxton 316 & 304 Yes 308-309 Berkshire County, Pa.-
clad with Saxton Nucicar Expt. Corp.
308/309 San Unofro 316 No Inconel 182 Camp Pendicton, Calif.
forging So. Calif. Edison Co.
(swagged) h, l
1 i
W i
'{'
j i
l~
L
,4,
!L i.
Ih o-l1:
'Intergranular failure of safe-ends' reported.
n
'i-I li:
!il -
< s.~
i
+
t-j r
l
',. s
%)
(!
TABLE II.
Safe-End' Status in Water-Cooled Reactors - License Pending Stalnicss Sensitized Weld Metal No. of Namn Location, Operator Steel
_0 N1150*F (Safe-end to Veasel)
Nozzles (c) J5 oiling Water Reactor _
D csden 3 316 Yes 308L 27
.,cundy County, Ill.
Com. Edison Co.
Milestone Point 1 304 Yes (3087) 27 Witerford, Conn.
Conn. Light & Power Co.
Quad Citics 1 & 2 304 Yes 308L
?
Rock Island, Ill.
Iowa-Illinois Cas & Elec. Co.
Monticello 304 Yes 308L 8
Monticello, Minn.
No. State Power Co.
3 Pilgrim 304 No Inconel 182 Nonc
{
P.lymouth, Mass.
?
Boston Edison Co.
4 Vermont Yankee 304 Yes 308 9
Brattleboro, Vermont Vermont Yankee Nuclear Corp.
(b) Pressurized Water Reactor I
Robinson 2 316 Yes Inconel 182
?
Ilartville, S. C.
Carlonia Power & Light Co.
Point-Beach 1 & 2 316 Yes 316 87 Manitowoc County, Pis, j
Wis. Elec. Power Co.
304 Turkey Point 3 & 4 316 Yes 316
?
1 Duke County, Florida
-l Florida Power & Light Co.
304 9
i Oconce 1 316 No Inconel 182
?
Oconee County, S. C.
Duke Power Co.
1 Fort Calhoun 316 Yes
,Inconel 182 6 (with 308L.
Wash. County, Nebraska wcld overlay {
Omaha Public Power District Indian Poinc 2 316 Yes Inconel 8 (with 308L j
- Buchanan, N. Y.
wcldoverlay) j Com. Edison Co. of N. Y.
(
p_
- )
TABLE III.
Strrr3 Corrorirn Crrcking in Wct r cnd Oxyg n Syr:te=
T:st Condition A_
B_
C D_
E_
F_
C_
H I_
y _._0xygen in cas Phase Degassed Degased Degass:d 1%
5%
20%(Air) 20%(Air)
Yes _
Yes Oxygen in Water, pps.
<1
<1
<1
%3 s14
%54 s54
(%40+Cl )c 100 Initial pH 7
7 7
10(NH 0H) 10(NH 0H) 10(NH 0H) 10(LiOH) 7 7
4 4
4 Tc=perature, *C 350 350 350 316 316 316 316 295 2S9 Stress-Loading Constant Constant Constant Constant Constant Constant Constant Constant Constant Method Strain Strain Strain Strain Strain Strain Strain Strain Loading
-Level Y.S.
1.5 Y.S.
- 1. 2 Y. S.
- 12. Y.S.
1.2 Y.S.
1.2 Y.S.
1.2 Y.S.
Y.S.
1.2 Y.S.
Specinen-Shape C-Bend C-Bend Tensile U-Bend U-Bend U-Bend U-Bend Bar, Tubing Tensile
-Creviced No No No Yes Yes Yes Yes Yes No
. Tctt-No. of Periods 1
1 3
9 9
9 9
1 Continuous
-Weeks / Period 35.5 12.8 12.8 2
2 2
2 11.8 Monitoring
-Total Weeks 35.5 21.5 38.4 18 18 18 18 11.S 1.8 R?.ference Copson &. Coriou et al.(2)
Copson & Econo =y(3)
Edstro: &
Ar=iio &
Dean (l)
Forsman(4)
G.E.15)
. rial First Observed Cracking Weeks solution Annealed d
Type 304 SS NF NFb 10 4
6 11.S 37 Type 304L SS NF 6
NFd yy Type 316 SS NF NFd d
Type 316L SS NF d
Type 347 SS NF NF 16 12 4
NF Type 321 SS 11.8 Incoloy 800 NF NF NF 12 18 11.8 Inconel 600 NF 21.5 25.6 NF 4
8 12 NF S nsitized b
Type 304 SS NF 10 2
2 0.1 Type 304L SS 2
Type 347 SS NF 16 10 8
Incoloy 800 NF NF 4
4 0.4 Inconel 600 (35.5)a NF 4
2 2
Id Rolled Inconel 600 NF NF NF 14 NF" W ld Overlay NF Type 308 SS Inconel 82 4
Inconel 182 2
182 NOTES: NF - No -failure for the test ation.
d - Corresponding stainless steel doped with Ta isotope or Ti during alloying cracked readily.
a - High carbon (0.1 C) In W J. 600.
e - S rings cracked during subsequent exposure of additional P
b - Tested up to 14 weeks.
11.8 weeks.
c - Water initially saturated with oxygen at room temperature, then added 5 ppm e,
- ~
~
s _
=,.
-- ~ )