ML19345F189
| ML19345F189 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 01/30/1981 |
| From: | WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML19345F185 | List: |
| References | |
| NUDOCS 8102090375 | |
| Download: ML19345F189 (57) | |
Text
ATTACHMENT 2 WESTINGHOUSE REPORT QIARACTERIZATION OF U-BENDS REMOVED FROM TROJAN STEAM GENERATOR "D" ABSTRACT Twenty-nine Inconal Alloy 600 U-bends sere removed from steam genera-tor "D" of the Trojan Plant, consisting of 26 Row I and three Row 2, during a May 1980 Plant shutdown. Nondestructive examinations gave no indications in U-bends from Row 2.
Three of 26 Row 1 bends had indica-tions in the cold legs at the transitions from curved to straight sections.
One of these three tubes had been a leaking tube. Each of these indica-tions was located on the leg with a readily defined intrados and extrados transition, oriented axially at the extrados, and ' contained to a large extent between these transitions (0.6 in. long).
On the leaking Row 1 tube examined destructively, through-wall cracking appeared to result from multiple initiation on the inside tube surface with propagation in an intergranular mode. Comparison of tubes with and without cracking indications did not yield any consistent or significant correlation with ovality, grain size, carbide distribution, minor element chemistry, or hardness. Preliminary residual stress measurements were completed on a virgin tube U-bend at a location where cracking was observed to initiate. The residual stress was measured to be compressive.
I l
One of the Row 2 tubes was awa=f ned destructively.
Both of the transi-i i
tions were smooth and no cracking was observed.
t P
h l
More information was obtained on cracking at the transitions by studying transitions on tight radius U-bends that had been removed previously from other plants to study cracking at the apex. Cracking at the apex was a consequence of tube support corrosion (tube denting) and inward movement of the tube legs. Five tubes from Turkey Point' 4 and three tubes from Surry 1 were armained.
Both transitions for each tube were
/
B102090 %. - -.
~.
~%
smooth, and 14 of 16 transitions exsuined were found to be free of cracking. Subsequently, four tubes from Surry 2 were examined. Each tube had a smooth transition and an opposite transition. One of the four opposite transitions had multiply-initiated cracks at the same location as the Trojan tubes. This tube had been reported previously to have cracking at the apex because of the leg inward displacement.
l l
l i
i
INTRODUCTION Twenty-nine Inconel Alloy 600 tube U-bends were removed from steam generator "D" of the Trojan plant in May 1980 for non-destructive and destructive examination. This examination is part of an overall program to determine the cause of leakage in the smallest-radius (Row 1)
(
bends.
The U-bends removed were from Row 1 - Columns 1 through 26 l
and the next-smallest-radius Row 2 - Columns 1 through 3.
These included one leaking tube (R1-06) and one tube (Rl-C26) with an in-plant l
l eddy current indication. The data base on Row 1 U-bends was augmented l
by a reexamination of U-bends removed from Turkey Point 4, Surry 1 and Surry 2, where cracking at the apex had been encountered as a conse-quence of tube support corrosion (tube denting) and inward displacement of the legs.
The Trojan U-bends were first received at Westinghouse for nondestructive examination which included visual, leg space measurements, double-wall I-ray radiography, and eddy current examination. The leg spacing measurements showed no significant leg displacement compared to nominal as-built dimensions. Three techniques identified indications at the cold-leg transitions from the curved to straight sections on tubes Rl-C6, Rl-C7 and El-C26. The indications appeared at the extrados.
Additional information was provided by the radiographs; they suggested that the indications were multiple, axially oriented cracks extending over a length of 1/4 to 7/16 in.
The results are presented under the following headings:
i 1.
Trojan tubes - Additional nondestructive examination and properties of straight length sections.
2.
Destructive aramination of Trojan U-bends.
3.
Comparison of Trojan tubes with tubes from Surry.1, Surry 2, and Turkey Point 4.
_3_
l l
4.
Residual stress measurements.
5.
Sunusary.
TROJAN PJ3ES - ADDITIONAL NON-DESTRUCTIVE EIAMINATION AND PRCPERTIES OF STRAIGHT LENGTH SECTIONS The.'.ocations of the intrados, flank, and extrados, as well a angular positions, are given in Figure 1.
A characteristic of the Trojan U-bends was that one transition from straight to curved was smooth; sometimes there was an identifiable intrados transition but always a smooth and diffuse extrados transition. The opposite transition had i
easily identifiable intrados and extrados transitions.
Diametral measurements were made for each transition of eight tubes including the three tubes with indications. These were made at.
l 1/4-in. intervals along the axis cf the tube through the transition and at 45-degree intervals around the circumference of the tube. Ovality versus distance from the intrados transition curves were similar for all the tubes. The data for the leaking tube (R1-C6) and a tube free from indications (Rl-C22) are shown in Figure 2.
The smooth transition i
has a higher degree of ovality than the opposite transition.
A 1/4-in. long ring was taken from the end of the longest leg of each tube. A comparison by light microscopy of the microstructures of the three tubes v1th indications with three other tubes revealed a finer grain size for El-C6 (Figure 3).
Carbide banding near the ID was observed on all tubes. The fine grain size appeared to be reflected in a slightly higher hardness for El-C6 (Table 1).
At higher magnifica-tions the grain boundary carbide distributions are similar (Figure 4).
Because there are few grain boundary carbides and those present are not interconnected, the structures are not sensitized.
-A portion of each ring was used for chemical analysis. A vet chemical analysis for tube Rl-C6 gave 76 percent Ni,15.5 percent Cr, and 7.12 percent Fe, all -
within ASTM SB-163 s peifications. Comparisons are made for some of -
the minor elements of particular interest for the three tubes with l
indications and several tubes free of indications by Spark-Source Mass Spectrometry (T1 is listed as a major) (Table 2).
No significant dif-ference in the concentratiocs of each element between tubes with and without indications is apparent.
DESTRUCTIVE EIAMINATION OF TROJAN U-BENDS Initial destructive anmination focused on the leaking tube (Rl-C6) and I
three tubes free of indications (RI-CIO, al-C22, and R2-C3). A print of the double-wall I-ray radiograph of Rl-C6 (Figure 5) has the intrados and extrados transirions identified for the opposite transition (right).
Included are locations of cuts, polished surfaces studied, and sections aumined for cracks af ter flattening to strain the ID in tension.
Section 6G was held for future studies, if necessary. Frints of single-vall I-ray radiographs of the extrados Section E6C are shown in Figure 6.
Identified are the additional sections prepared for metallographic anmination and the transition from curved to straight section (extrados l
t transition). <iply-initiated cracks extend at the extrados from the intrados transition level to just above the extrados transition. The l
lighter area above the extrados transition indicates a vall thickness reduction in the bend. These cracks were evident on the CD and on the ID I
where they were surrounded by a blue stain (analyzed later) (Figure 7).
The wall thickness reduction starting at the extrados transition is -8 mils (Figures 8 and 9).
As polished or etched surfaces through these cracks (Figures 10 through 14) show deposits on the CD surface 1.5 mils deep, a through-wall crack, and intergranular cracks emanating from the ID surface. Chemical analyses of the deposits within these cracks were performed with the energy-dispersive X-ray analysis and with electron beam microanalysis usina both crea scans and line traces.
Both techniques identified the main alloying elements present in Inconel Alloy 600 plus silicon.- The silicon would be expected because silicon carbide papers were used in polishing. In addition, energy-dispersive X-ray analyses weakly identified sodium.
An Auger spectrum was obtained on the previously mentioned blue deposit on the ID surface 10 mils away from the fracture surface. SiO, S, C1, x t I
Fe, Ni, A1, and the normally observed C and 0 were found (Figures 15 and 16). ;his bra deposit could not be seen on the ID at the location of the indications on Rl-C7 and R1-C26.
Additional hardness data were obtained on the polished sections at various angular, radial, and axial positions. Loads of 500g were used at mid-wall and of 200g at 5 mils in from the ID (Table 3).
At loads
(
<500g, the hardness readings are load-sensitive and increase with a reduction in load. Both sets of hardness data show that the hardness decreases as the distance away from the extrados transition and bend increases. These data will be compared later to those for tubes i
without indications.
l No cracking was observed anywhere else on Rl-C6 or for R1-C10 (Figure 17);
Rl-C22 (Figure 18); and R2-C3 (Figure 19). The start of the bend is evident on the extrados section of the cold leg of R2-C3 as evidenced by e reduction in wall thickness of about 1 in. at the extrados (Table 4).
l COMPARISON OF TROJAN TUBES WITH SQUEEZED TUBES
_FROM SURRY 1, SURRY 2 AND WRKEY POINT 4 An inventory of the tubes from Turkey Point 4, steam generator "B",
from Surry 2, steam generator "A", and from Surry 1, steam generator "A" l
was made and the experimental work plan established for these tubes and two additional Trojan tubes (Tables 5 through 8).
These plans were com-plated except for two changes:
l.
Trojan tube Rl-C26 was not examined and saved as an addy current standard.
2.
Both Surry 1 and Turkey Point 4 tunes had no opposite transitions (both were smooth), and therefore, a smooth transition was substituted for the opposite transition in metallography and both transitions were used when straining was performed.
l l
,.m.
i 1
(
Eddy current testing gave no indications. The diametral data were used to plot ovality versus distence along the tube and through the most well-defined transition for each trbe (Figure 20). The data for Trojan Rl-C6 were included for comparison purposes. The easiest-to-identify transitions for Surry 2 tubes were opposite transitions, whereas both transitions for Turkey Point 4 and for Surry 1 were smooth transitions. This is evident in the prints of double wall X-ray radiographs j
(Figures 21 and 22).
The use of straining to search for ID cracks was employed on both smooth transitions of four tubes from Turkey Point 4 and of two tubes from Surry 1.
Only the opposite transition was anaf ned on each of two tubes from Surry 2.
The only transition showing ID cracks was for Surry 2 (Rl-C7). Its opposite transition was cut transversely 'l/2 in. above the extrados transition, just below the extrados transition, and 1/2 in.
below the intrados transition (Figure 23). Each section was split longitudinally, flattened to strain the ID in tension, and examined at 10 to 501. Short intergranular cracks initiated on the ID at the extrados and between the transitions (Figure 24). They were completely intergranular and extended halfway through the wall (aspect ratio of ~4 l
(Figures 25 and 26). Typical light micrographs are compared for the materials tested to those from Trojan tubes (Figure 27). With regard to the two U-bends with ID cracks, Surry 2 (Rl-C7) had a microstructure similar to those of tubes free of cracks, whereas Trojan (R1-C6) exhib-l ited the finest grain size of all the tubes examined.
Additional metallography was performed on transverse cross-sections at the apex and at 0.2-in. intervals through the transitions.' No additional cracking was noticed in any of :he tubes studied. Meta 11ography alone was done only on Surry 2 (Rl-CL). Meta 11ography and Knoop hardness readings were done on transitions for Surry 2 (Rl-Cl), Surry 1 (Rl-C9),
and Turkey Point 4 (Rl-C83). The aper, smooth, and opposite transitions were studied for Troj a (R1-C8). Locations of the metallography speci-mens for the opposite transition are shown in Figure 28, and typical microstructures and hardness impressions in Figure 29. 'For a given I
angular position, a general decrease in hardness occurs as the distance l _
... ~.
i l
away from the band increases (Figure 30). These data at the extrados and _M5 degrees are compared to the envelope for the data on Trojan (El-C6) (Figure 31). They are quite similar. The data at the extrados or the extrados _45 degrees on Surry 1 (R1-C9), Surry 2 (El-Cl), and Turkey Pcint 4 (Rl-C83) worst centered (not plotted) on the envelope for Trojan (Rl-C6); 14 data points were within the envelope, four to the left, and seven to the right. There was no discernable difference in the hardness gradient or level at the extrados for the cracked Trojan tube i
l (Rl-C6) and for four crack-free transitions fica tubes from different l
plants.
RES DUAL STRESS MEASUREMENTS
. 211minary resiJial stress asasurements were made on a virgin 7/8-in.-
diameter small-radius Inconal Alloy 600 U-bend (Heat No.1202). A print of the double-wall I-ray radiographs of the chosen tube (Heat No.1202) is shown in Figure 32. Ovality seasurements (Figure 33) are sf=flar to those of the previously measured five opposite transitions, although the m imus appears more diffuse. Strain gauges were mounted on the extrados and the half section renoved with a jeweler's saw (Figure 34).
Af ter cutting. the strain gauges indicated a change in stress of 1.0 to 1.8 ksi. Additional strain gauges were placed opposite the other gauges on the D, and the positions of all strain gauges with reference to the transitions documented (Figure 35). The wall thickness measure-l ments show that Gauges 3 and 4 are at the extrados transition. Each i
strain-gauged section 5 s removed and the sectioning stresses obtained (Table 9); at the principal locations of interest -(3, 4, 5, and 6) no pronounced tensile stresses exist. These sectioning stresses and stress changes accompanying removal of material from the G were used to calcu-late the stress distributions in Figures 36 and 37. Again, no pro-nounced tensile stresses and aminly compressive stresses exist at or near the G surface at the location where cracking was encountered.
More residual stress data are required to define the stress distribution at this location...-
SUMMARY
1.
Trojan Row 1 U-bends were characterized by (1) a smooth transition without a well-defined intrados trao.ition, and (2) an opposite transition with well-defined intrac'ss ud extrados transitions.
2.
At the opposite transition, the extrados transition along the tube length was 0.6 in. above the intrados transition.
1 i
3.
Three of 26 tubes had cracks which occurred at the opposite transi-tion between the intrados and extrados transitions and at the extrados.
4.
These cracks were characterized as resulting from multiple inicia-tions on the ID with intergranular penetrations.
5.
No consistent and significant relationship could be established between cracking and ovality, grain size, carbide distribution, minor element chemistries, and hardness.
f i
6.
On a virgin tube, strain gauges and a layer-removal technique were used to measure residual stresses where cracking had been observed.
The stresses sere measured to be compressive.
7.
No cracking was found on the Row 2 tube studied.
8.
Row 1 bends from Surry 1 and Turkey Point 4 had only smooth transi-(
tions (no opposite transitions) and no cracking at the transitions l
was observed.
9.
Surry 2 Row 1 tubes had opposite transitions like the Trojan tubes and one of four had ID multiply-initiated cracks (aspect ratio of 4) at the same location as the Trojan tubes.
l l
l DRS/41m8A19.
~ : y,. -
s i
I i
/
s
\\
-f 45
/
g
/
/
1 l
/
/
i
.I t
//
/
+c
-s
/
//g
/
/
c4
/
I :
m 0
/
g
~_' '
w
~.
, 0*
l Colo 45 Le9 45e i
~.
~
s Hot Leg nsure 1
- Definition 0: enguur cositions
/
i b
i I
I i
1 i
l i
1 e
u C
l o
c l
a_
=
x c
l
.8 c-
- a.e 8.
l 2
1 0
."=
c m
.,c
=
s~
e.
.e
'9 a
=.-
c=
=
3
.no. @
.o== O
{
e o
- = #99 q'
_m u
=x-x._
6
.,5
.= -
os t
.C. o
/
- 9. =. -
~
t
/
~
\\
x c
e =
c
's o
~
= c t
~v e
.e a, x
H _, A
= 0
/
y y N *=
i
/
~~
/
J a
t P.=
~
e l
i
=
. = =
x l
~
g c-l U. T l
l h/-o m v i
ae
I
=.e e
.m. *m 1
~
o i
-e o =.
m
-=
w e
)
.e 2~
1 c.-
1
.4 e
2 a =
,=
=.a u
, ~
c.-
O =-
=
m l
c,,%,
g, oE o
>m o
ah i a w
g
~e e
a E 6
o u o i
t 1
I t
i i
s
[
w t
i
- r t
e e
a e
o a
o o
c m
v N
.-e l
u! E 01 ' I( o0 ) 'e!O - ( o06) 'e!Gl i
h'm.
b,*',
r"l,f f '$ {t{ J,f} f W),Y Q R
- t r*.M * 'w Q %,*'as: *.
b,.Y}S'}.
s*
% R ' j y ' A f ','f-
. *'* L.} :
J'*,}Rlht
- N5?.'*'?
{ 's' g
l$ { Ef,.-g f,'j 2QL f%
M'4(OPi@SA4'yitE$a%fyy, 6
-ii_;UGA'+-
K. MNU,j'.F
-GMs';?pwj.4+$Ed;#.am.c,w
. E s$ 3 2 4 4 L h 2,3 @4 %
@p n D
%iss kw $d w
ns
=~~
p.%Cc 3
+tg7 s.
y fM 3
n.. m.u.g.zn,c., fg;.,rew mp,m w
a er,.-.w
.,. < sn..
, 7.;.
w 3i'
"' k I
~*
- r c.
b Af5i$$$,i$$$2d !
I
$ S % n$ 8 5,E 5 5 $ $ m.'.. $ @m+ %
M,i f f., y G,E [ $
e.
mya;:.r.
- c. h.'2,%
. p,
e
..:p:
u W
qf;;:ia g49Yf.v y$$.hg$;%.-
e eoo s
m
$f.
5' I^ri:~st ?,kshr #
\\;} r 5.?G '[1,0't:'$
h 5C$
0 1.
- -4y.
- v4 n
gt
- 2n.s u
k ki M
dk w
i 7,M -.-rQ. e 7
L%c.C@r 2
.g &
W g @.g
- S. h, g %
--. >.o a
p
- '-e
't
-%.i g
g 3 5,.b.l:,
u<a QM e?]
4 a e in
.c h
. A< _
-$[gg$jpg- [ ~w,k-a 0 a 8'
z 1 H.,
j
-. LS..%' cA'.
M,
. _. s u
.+.!
.2 C
..p yi g
n.
.u.
n..
7 G
.C
\\
Id.F '
V,< b M
P' 4
~
.. g.-'1. 3.'.,
s.
c
-m
.u
=~-
I bc
%[;u-
'e~e'..-
% g
' ~3 4..
x o ;=
+**$' $','
8k
~'
A, r
h
~
., =
- p g..g a
w
. < " - 9 Yeb 5"
't.
?c. -
.ges
-2 d :-
^
c..
.. ; [
g- ;
r
,u
.00 =>
=
W S
f 6
j jb-
Table 1 Comparison of wall thickness (mils) and Knoop Hardness (500g) for three tubes with x-ray indications and three without at various approximate angular positions on a straight leg.
Tube / Leg Wall Thickness Ilardness (mid-wall) 135*
180 225 Avg.
90 135 180 225 270 Avg.
(Extrados)
(Extrados)
) h ' 57 RI-C6 o
55 54 55.3 219.8 233.9 245.0 234.7 216.7 230.0 Colda j
(Rb = 93)
Ia t
i R1-C7 k v4 57 57 55 56.3 186.7 187.3 178.2 178 183.2 182.7 Hot **
I g
Y R1-C26 x
56 57 56 56.3 199.6 190.4 185.2 201.8 206.3 196.6 unt**
)
R1-C10 59 60 57 58.66 179.1 186.4 183.1 180.4 186.1 183.0 Cold
- RI-C13
$2 52 54 52.7 187.8 183.1 187.8 186.5 190.8 187.2 H2t**
g Y
R1-C22 H
59 57 55 57.0 181.3 185.2 183.5 176.8 181.3 181.6 Cold ** gg (Rb = 83) 0 opposite Transition
~
03 Smooth Transition s%
4
.f k
^
f
.. w.4, *. N ' %,
_,_ O f% l
. MRiv-
' '*f
.?*'
f,*
',N
.m, * '..m. 4 gm 4. e.
-.. m *-
er.
w.
e.
- g b,
-E Jun.
x--
- -jf..
6.
O.. -
en A -
m y
y-
.+
.m.
~4.
.s
, e M e ne..w.-
x ps i
-My g
- -,.'*. ).
_, -.. ~, -. :. ; y[g...egf
' Egl@g W,
s s.
.y
+m. %e.
p
,y,?, p Q.
e
. %...g>
e
- s...s v
v
.' g L * -
... - y y '
. 3gQ,,,,
- 3 3..;a** V
- b A h *t yA,.,y %g. - [ '_ _,
3}e
%j. :;..v.,-' ", y gf #_f2,,g(.,
n pt,,+
-h, p *,.
-f e
'.. e g....y
. y p, g,,,,
y
.s
... Y g W.
lp... 3.T.
j..
g e
.g q
f.
. y.5.g% y R.+, _ fe., -.f r
. 9 S.-.p f.e a
,1* g.
yx&y.Q w,; n 3-..
- -5*- F l+E
.,g fst.,.
4*. j, -
... *y ay
. Q, ;g..;
2
.,.i Sk. }% :,,
Qe;
,,, -. h,*7.k..
., } g - [ " f ;.; i I=
I,N -.
y 3t
'.m p p & c ).f.[a..
- 4.
- '4, g 9.;.y 3. Q,-l *W =-.
{
s9
... M 4:p:$v J;,..
+(MMP s=. h?f :[:
?;a c
f
! G : M $ $ e Q } Q & "*
%qWCW%k
'd-C10 Rl-C6
.'... ~
- i
,s*6 k '- #
- f.....,.g 4
.,%.,'~
' ' A mg "4, W x n.: y +.
- [ v y *g, g '
d * '.
w.
h J h e-
- ?
- f
~
gr
, *4 A"
a
.s s
- g &
6 7
,...,y
. s
+
n Y.
~
f _A* }ip e E
?
w e-w f
'*k.*~
. s 4.y.
'R
' + M " j*
- E C &da
? s h*
. _t*,'
[4 y:. $.[4. 3. g v%..~g
' M
,k' s 3.. -MM_ d i$.A.; Q -[9
?.
g.
.ps
-,.
- megy. ;
Lf
?/ w. *-.J
+% *.
"M
~,
[' k d'
g,
%[.wl Q
Nh
[
/s. g g g ;tj/ g?'OjN Y~+{ g#& [gg g
%ig R e
? l f ll$'*
~ k -f f *,* f 4
- Q.Q;.%
6 ;' l
.ne 5
.y% gjhp3yQRq< y5$.f.:&C
[
[
{
11-C13 R1-C7 I
[
- h
[
I 4
.k'T,4 k 8.4 \\
[-k*
'. O N ;
- y.-.
-s y 4. s..,. 'uz
. *A. : ',,
.~*.,%
9p ' ',
1
-p
~wQ o..
.;w e
% $ ix p,.
3en.
\\
N a.-
%e-
.e4'
..: : c.;
at
- h. fIk'
]
'D 8@
N
[N h
s ;?
~.
[i& '
- D ~
- l1.<.Q ;,ll
?
w!-Ma s. %.?*-f& ^ % W i
-- h
~ ~
- e,,.4t*
- l}.;:;l-'t r
'e f?gy
,; l :. ;. 3
- . 6, y Y.,.- 'l Jl
\\
f f
Y._fN. ' y.
M*.
_, [ l{.
'l.', -
~
R.y
.. +..
.%A w.
... o-
,.1,;g ;
..,.o
- P 9
g..,-.
.af y.,
c
.en
- - r
- ?
- 4 x s
2
~.
s
..,u,.4 *f.
, p.;
+
,w
.~
~
3 m,.-
a. -.A s.
{&l?,[ '_
[l _ :s] l
. K.# f f l Y h '$*
y.
- l%h L.
C. - '. l Rl-C 6
,,o
.. m"a ",ns Indications Ccmparison of grain boundary carbide distributions for three tuces wl:h indications and three without indications fro Trojan.
Sections were extrados
'A" sections from straign: leg.
^q m) q h,
j-6 a z
,a
, a.j J ; M 3
L__._
Table 2 4
Spark-Source Mass Spectrometric Analyses (ppm wt.)
(First three tubeu have indications, balance do not)
Tube (Section)
S
_As
_ Pb B
Na K
V
_ Th No 81 Nb
, Zr Ca RI-C6 (C1) 33.0 18.0 4.1 50.0 0.47 0.10 330.0
<0.22 70.0
<0.10 28.0 0.33 3.6 (C2) 25.0 18.0-4.1 27.0 0.47 0.03 190.0
<0.22 47.0
<0.10 120.0 0.39 1.9 RI-C7 (17A) 13.0 5.2 3.1 50.0
<0.10 0.08 510.0
<0.22 1.3
<0.10 1.5 5.0 3.6
-RI-C26 (I26A) 17.0 18.0 2.2 27.0
<0.10 0.05 280.0
<0.22 180.0
<0.10 120.0 0.50 1.3 RI-C1 (IA) 25.0 5.2 4.1 100.0
<0.10 0.02 470.0
<0.22 2.0
<0.10 5.5 5.0 1.3 R1-02 (2A) 9.1 5.2 2.2 27.0
<0.10 0.05 190.0
<0.22 47.0
<0.10 28.0 4.0 1.1 RI-C3 (3A) 33.0 18.0 4.1 27.0 0.73 0.05 190.0
<0.22 47.0
<0.10 15.0 2.8 1.9 Rl-C4.
(4A) 33.0 28.0 8.2 27.0 0.27 0.05' 85.0
<0.22 130.0
<0.10 110.0 0.79 1.5 R1-C5 (SA) 25.0 10.0 3.1 27.0 0.27 0.05 140.0
<0.22 130.0
<0.10 200.0 0.79 3.6 RI-C8 (8A) 33.0 10.0 8.2 100.0
<0.10 0.05 510.0
<0.22 2.8
<0.10 2.8 6.9 2.7 RI-C9 (9A) 13.0 10.0 2.2 10.0
<0.10 0.10 190.0
<0.22 230.0 0.89 140.0 5.0 1.9 RI-CIO (Il0A) 5.9 2.8 0.82 27.0 0.27 0.03 190.0
<0.22 82.0
<0.10 55.0 5.0 7.1 (110B1) 59.0 10.0 4.1 27.0 0.13 0.05 94.0
<0.22 130.0
<0.10 55.0 9.2 1.9
.(1001) 33.0 18.0 4.1 27.0
<0.10 0.03 190.0
<0.22 100.0
<0.10 55.0 9.2 1.9 Rl-C11 (11A) 9.1 5.2 1.4 21.0
<0.10 0.13 190.0
<0.22 230.0 0.50 55.0 0.50 1.3 RI-C12 (12A) 17.0 5.2 4.1 100.0
<0.10 0.08 330.0
<0.22 1.8
<0.10 1.5 9.2 3.6 RI-C13 (113A) 17.0 10.0 8.2 100.0
<0.10 0.03 330.0
<b.22 4.7
<0.10 '
2.8 9.2 1.3 Rl-C14 (14A) 17.0 10.0 4.1 27.0
<0.10 0.13 190.0
<0.22 82.0
<0.10 28.0 0.33 1.9 RI-CIS (15A) 9.1 18.0 4.1 50.0
<0.10 0.13 330.0
<0.22 350.0
<0.10 120.0 0.50 1.3 kl-C16 (16A) 9.1 10.0 4.1 100.0
<0.10 0.08 330.0
<0.22 4.7
<0.10 9.8 18.0 1.5 RI-C17 - (17A) 33.0 10.0 1.4 50.0
<0.10 0.05 510.0
<0.22 180.0
<0.10 140.0 0.50 1.9 R1-C18 (18A).
17.0 18.0 4.1 27.0 - <0.10 0.13 330.0
<0.22 470.0 2.5 170.0 0.79 1.9 R1-C19 (19A) 13.0 5.2 3.1 50.0
<0.10 0.05 190.0
<0.24 230.0
<0.10 55.0 0.79 0.71 RI-C20 - (20A) 17.0 10.0 2.2 27.0
<0.10 0.08 330.0
<0.22 230.0
<0.10 120.0 0.50 1.3 RI-C21 (I21A) 17.0 4.4 3.1 38.0
<0.10 0.08 330.0
<0.22 2.0
<0.10 2.1 9.2 1.5 R1-C22 (122A) 9.1 10.0 2.2 10.0
<0.10 0.08 190.0
<0.22 230.0 0.89 120.0 0.50 5.3 (122B1) 17.0 18.0 4.1 21.0
<0.10 0.05 330.0
<0.22 230.0 1.4 120.0 0.69 3.6 (122C1)-
33.0 10.0 2.2 27.0
<0.10 0.05 190.0
<0.22 2 30.0 1.1 110.0 0.92 2.7 1
l l
I i
1 i
i i
}
6F E6E4 4
f l
6G 6c2 i
6c1
]
i M
16El i
E6C E611 g)
I 16C I 611
- Extrados Transition I
1611 168 E6B 1
Intrados Transition -
E6J1 i
i l
16A E6A 161
~
l
~
l E6.1 llot Leg Cold Leg i
Figure 5 Print of double wall x-ray radiograph of RI-C6. Solid lines show approximate locations of major cuts; sections are identified; surfaces polished for metallography are designated by A. Underlined sections were flattened and their ID surfaces examined at SX.
Further cuts of E6C are given in next figure.
l
1 l
6 l
.C
__a c
i i
1 1
/
l l
l l
< r ai mWW l
9 LOL; Y
1 A
26C2
.WL L
100 n
k
.2.-
\\
\\
\\
h CY. I N
Figur2 -
Or--.-
,e
,e
'-l,,,ain,e mai x-r3y "3C1CgP3Cns af **,e
- g]c 7,c 3
,--e e s i -c 0 SnCWing ;P3CXS at tae extP3CCS ' 'ON' 2nd 9
(eChiCn
~.n e *.Pa n S 1 *.i C h frCm Strai 00 00 Curvec.
as 2 "C Sec u e n t ; ;. ;g; +c f3 m ggg, ggg gg
.g
l l
.hi$'
i
'a.
1 l
IIA.
'.;;;q
(
..L d;
- k l
k,..,
k.<
1 F3 y
$5 e
. j k.
fI e
1 y
in' M
m 10 00 Figure 7 Photographs of ID and OD surfaces in the neighborhood of the crack on the cold leg of tube Rl-C6. A bluish color is evident on ID at location of crack.
Distance to end of crack from cut tube-sheet-and (bottom) was 0.71" on ID and 0.85" on OD.
oo m
14 M
l l
k c--.--,-..._
l 1
I C
n e
o=
o I
C d
mu
.e e,
=
.o o
=
a m m
=
o
.c o
m u
e m
o+
_=
l
~
.e o
e-o 2
%e3 m
e o.
m c.
=
e e~
m o_
ou o
~
88 o
~
-nu m
r sy v
(
_e -
~
a 1
5o J <:
C x
e9 e
c.
=_ m-c m Lu e
." o
- v l
~
sm
=_ o e
m o
> o
- 9 <: a o
MN s
s-m
- w l
=
m =.
e o
c.o
.-m m.-
m m e _
e~
),
o
-e ce c-~vu n
x e a.
.2 e
=
c=
= =m
~xxx o
e o
= 9
=
m o
E:: o o
l i
C n
7 n
T co s
X a v w
=.,
,=2.
t i
1 1
l
-s o
e S
S
(
.s l
i sI a 'sssup!U1 tie?.1
i s
?
r i
- .y. 7
- :5:::
I6 I6 T-48 51 I-~
o 51 48 51 I
l 51 47 50 1-E6C4 6
=
~
~
i d
i 53 46 49 i
6 l_ _ 49 46 1 1
r-ECC 3
=
1_ _ g. - I
- t Extrados i
54 55 r-E6C2 w;
,e Trans.t.i ion I-54 -L E6C1 l____..__a i
1 56 54l 54 = i
-E6 8 e
L I
55 54 54 i
,d c.,
____i_______i____
_b 57 i
57 55 54 i 54 57--I6A d
i I
.i. _ _
L____
\\
360 180 0
I6A E6A Figure 9 -Wall thickness measurements ( thread micrometer readings in mils) on OD surface layout in vicinity of crack on cold leg of Trojan Rl-C6. Cuts are designated by dashd lines; sections by letter-number-letter combinations.
Location of cracks on extrados is shown i
M~
4.
1
.g..;~..
y
' * *.4
% 3.;
2 i b7 -
[
??$2hh.. $,$,h)kk&
I I
s
' MP,* / '
/ Crack x 3,.- -
... ~
.l
~, -
.. ;l
~ '
.1-I L-
[,;[ g *
~..
e
~
~.y
~ ;.
E.;.
e, q-7 5
rs.
s.
~..
.c.
.n i
L
- 'f_ _._,
'su n '
J
+
g ~... 9 _ ' _ -
.;. qg..
< e. s...... +
.,. ~..
IG penetration on ID 1
d i
Figure 10 First plane of polish on section E6B showing through wall crack j
at extrados,1.5 mils of deposits on OD and intergranular (IG) penetration on ID.
i, l
)
00 4
I
/,
L i
~
6 t..
- o. 0/0
'J 3
on.
.;'g t,
g y,-
o
(
\\
'n
..s.
' ';y/f.
).<g l
4 4
8 4
s.
.v i
' ?.,
?s.\\.
.x 4
.l.
I
(
.,.,d' 4
l l
!U i;p n
M l
' W PE'P@
i 6
9
/
~
'cd50
.k.
I
@hP9 M
ID b
I Figure 11 Through wall crack observed on second plane of polish of Sect. E68.
o
[
h a
-w.---- - - - --
w
Y g W=.s.,.; ' *r: g'..'_i;.[.Q' Cv.
- . Q ; &P
&. A%O ?~ ?g.?
n
. y y*
.w
. Jeh. f. -
. )# * %u
.i d
g m.
h'
' ' '.'8*M%$M:-=AYgh'_,[ ' -
+p h. o ' y 7 -rt y -
y~.
l
_.1
. m- :- y
+
., y n%,.s y
- - t j
g 4 p}s;- ~-j l
b." M ';;;
...w
..... +
- '. m.g
. _ y +g, a;.,. > ?
~ A.,.,
- -. w 4..-
g.x..
g...g. 3.
... j
.<.;. j pcp.
y i
o, WA
- f.' $ %' Q % M,0 Q.?. g g, )
o
< +: c
+~.u N
9
,~
r
\\
- - }' -p&g.g j..g;;l
[
_ Q ;j.= 3 l..y.-
_4
_ :RG.;;yst -l;%;<g.
l j
l
.' ' ?. % *j.% % h ;
- l5. "
- b
g3,g jp.%w-,
3
- : *Q
. y q + ; y3.' Aec.,.3. g.
<r.*
- s. m a
~..y o
c.
-y
.g
... y y,. 7, y.
,,.u..
g*
- g v
g
't sp,
+g.,.
, ;, s a. g 5
g
.g
.,.e..
e
.... g;..
- ?
e
- p,V;...
e,, N.
.;. v e. ; ~ 1.; _..
l T N g.. 4.,
.... A. w- :
fh. ~
fY u
.. s
-v e.
n
. ~...
g %j'- h G 7 Y, e,,
.,,. %= ;
.e','-[*
, +;.y" ' ;.'
.ygs,'odef
[.',.
- * < ; '. v i; '
c,'
. ; Y :q.*
_ 'jQge
~
, l t.
- - ;' ~
w m
_ li_*.
- 4. > _- *l.;.
f w -- _ ~ { + +
- ,y.
&g '
7 3..
s. g. 4, 3 y.
f.
-d. 4 M y.
-s c._
. j, 3 t 's.* (g. 93ftp6M$',# ^/ b -[h..
4 s
.~
m h7,Y ?.,, ' :.,
I'
. '4 g
. A y : *
- q.
.b
~
-g_yi;
.. = *
=
w As Figure C SEM micrograpns of cracks in 2nd polisned plane of Sec E6B ('ll-C5 tube) showing areas examined with energy dispersive X-ray analyses (identified by numbers).
Cnly Si and Na were icentified as cresent clus alloying elements fcune in :nconel 500.
Wy,,!.mmm e c _a i
=..m.: v..~n_...w
. ~.
. &.:,... v _... w'*g.. x. 3.. ~&,'c
'f, ' J '-
- w 4 %..
~ j.c i,...
\\
~
va. q.~
i
(.w'h*k..m.,
-M.
- y;;,u
,,.,,,,..,.~~;.
n n% i-e
_y
%,~
bys ' *
~
&$..... $. 's. ~N
..m.,
~
.--^='-W_mn....,'jM..*.,,
l (LM) u.
T.,
1 1
I 9
- __i l
. ~
.'.\\*
(++
g
(.4 %..
~,,' %
j' r.
_g-(
N.. '
g.
v.
. '^-
q'q.".
N*4 %
r 'L f.
- g
- .%.6 L ;
w
'e-
+
y' N.
y p
7<
g,
. ; ; -. n..;,
4
.y
.n.
4 9
z E.,,.k.
3 11-1.% " % (* -
[,
']
w-O '. ~.
~
BSE Sik" l
Not Detected:
P, S, Pb, K, Na and As "M*
U Two areas on the 2nd polished plane of E6B (Rl-C6) were examined with electron beam microanalysis for Si, P, S, Pb, K, Na and As, and are shown in the photograph obtained with the light microscope (LM).
Similar results were obtainea for both areas.
Back scattered electron micrograph (SSE) is shown for lower area.
Si was detected -
in localized areas.
None of the other elements was detected by 3rea scans or by line traces through tight cracks.
b q.fnfa
,!e.jJ'
,' m l
, '\\,
)w s -,
g L b a O u t) J 'u J d C3 Qj y'
t s
t OD
,., r l
i.
\\
iP]Lf ' <. o_.
[
)
e, y
mm p-1 x- -
. a
..\\
>7.%
\\..
1.
.- }
{ ',
.,.e g,
.c n-yf,..-.
v,'c.,p. s.
\\
... ?,
l
..... ~:.
e.
e.n ;i..)s.q ;:..: ; w ;.t.,.,a..
~.
.a
>.z..
. yQ:.
c f)hef.p@.a... -
@ y$. $,....'.l..Y,..
.c
. y..
t!
ee 1
c.:..-.
\\,
i i
s V
{.
f y
- c.,;v:.,;)i '
..p ; ' -
~ :.q..
,f.
l'.
')
l l
Q
if.+t.
') #". '.,
.... j b M
'Y 1
- ,.. :..g,,.
+
,3
'M..:-g :- v-
. A -l.
e M, w q..
N m
ID c ____a Figure 14 cg Third plane of polishing (33 mils below second plane) of Section E6B at the extrados.
W P
l l
i e
Carve 7:'5403-A I
I I
l l
l l
l l
)
4 k
+m---
. r ;_
.x 3 0 S
N Fe Na l
Cl Fe fe Ni Si0 N.I x
C 0
I l
l I
I I
I I
l 0
100 200 300 400 500 600 700 800 900 1000 Electron Energy, eV Figure 15 - Port on of an Auger spectrum from the blue deposit (10 mils from fracture surface) on Section E6C1. A similar spectrum was obtained in a study from a neigh-
. boring areaof the blue deposit
\\
l e
.iYtthttUtw{'43l$y'yn" r 2y T.m*'Myj' vt:wl: *a' n
1 :.; -
., g.
- ,., s. a.
q
' i 1
' urse :;5107 - A l
1 I
I I
I I
I I
l gl hN 4
M p
p Q,,ag g h
- z w u
u Al Si0*
x8 i
i i
i i
i i
i 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 Electron Energy, eV
- Continuation of the Auger spectruni from the blue deposit on Section E6CI visure 16
~
=
0 m
Table 3' Comparison of Knoop hardnesses ut various locations on RI-C6 (coll leg)
Angular Position 135*
180-225 Depth from 25 25 5
2 ID mils Load,g 200 500 200 500 200 500 Sect / Distance above(+)
or below Extrados Tran sition, in.
6F % +4(Apex) 326.1 353.0 342.1 318.3 o
' E6C2 +0.1 314.1 366.6 337.9 "
375.2 322.0 gy362.6 au 320.0 310.8 x
E6B - 0. 5 272.2 g 312.0 288.6 2P4.2 U s310.0 E6A
-0.9 266.2 233.9 275.6 245.0 280.6 234.7 (Top)
R 2 96 b
(Bottom-1.3 248.2 225.5 249.0 211.5 240.0 215.7 R 390 b
ma
_Ama d--a A
am.-m---wA.-m.
m J
- aA 4A
.----.4
,at m.
4 ab s-._a
---#_~a
.o._d
.A u_m m.
a_i.
a,4 Aa a_ _.
,m-.
am._
i I
f i
e C
i 4
m CO w
emE 4
U i
i O
w l
- 3 o
W
.I g
C y'
+d W
3 C
y W
Y C
I 3
w.
C J
=
W r
t t
i l
l w
O N
e I
w E
b w,
E O
C C
U M
r c,
C
~
6 o
"C V1 "m
1 6
L T
C C
W U
C a
i m.
W M
j m
i tJ C
b l
~
m,o l
l l
~
~
C y
.,=
b @
I C
I C
.c 6 C-4 J
9 LT l
@c
.o= 7
~
5 C
=
M3 2
m cw c
kB
=
t W
e==
i ee he C3
% C C
Lo p
e
e
,3 h
A 3hO a
O l
O
- C0 Ca t t'J Ceste j
C%
W Oa g
9 m
CC O
Q y
bt W
y C.
- I an=
I W
.nn 009
22G Eg 4
E22E2 E22tli K
E220 122E3
~
E22!t 122E4 E22C 1220
[
\\
122I E22B1 122B1 E22dl I22J\\
122B g
122J
_m 122A r
i*5
' E22A Ilot le9 Cold Le9 Figure 18 Print of double Wall x-ray radiograph of Al-022. Solid lines identify cuts; locations and identifications of sections studied are given.
,m
l j
3G E3E6 3F l
E3E5_
1 j
3C1 i
I E3114 EP.
'I3E3.[{2 13El
~
13D 1 3112 j
E3113 i
E3C 1 3111 13C l
1 13C1
(
i I
I E3I 13I 13B E3B I
E3J 13J I3A l
E3A Ilot Leg Cold leg Figure 19 Print of double wall x-ray radiograpl! of Trojan tube R2-C3.
Explanations are given in Fig.
In addition, single wall x-ray radiographs were made of split i
sections about longitudinal 2" long immediately above the polished surfaces on the legs.
)
No indications were found in shots of the extrados, f,'zados and 145" from those pos t-(
tions for each leg.
Section E3B eras examined after metallography and straining.
I t
i I
Table 4 Wall thickness (mils) data for Trojan tube R2-C3 at and/or near the transitions from straight to curved sections.
Intrados Extrados Distance from Distance from Tube Tube l
Sheet Sheet End of.
End, in.
Designated Leg, in.
-90*
45 0
315
-270
-270' 273 180 135
-90 Hot Hot 3.2 55.9 59.5 57.9 58.3 58.7 3.0 55.1 55.1 58.3 58.7 58.3 2.8 55.1 e56.3 56.7 59.1 58.3 2.5 60.2 57.9 57.9 56.7 54.7 2.5 53.5 56.7 57.1 58.7 58.3 2.2 59.5 57.9 57.1 56.7 55.1 2.2 53.5 56.7 58.3 58.3 57.9 2.0 59.1 57.9 57.9 57.1 57.1 2.0 53.5 56.7 58.3 57.1 58.3 1.8 58.a 58.7 56.7 57.1 55.1 1.8 53.9 57.9 56.7 58.3 57.9 1.5 57.9 58.7 56.7 57.1 56.3 1.5 53.9 56.7 57.1 55.5 57.9 1.2 57.1 58.7 57.9 57.1 55.5 1.2 54.3 54.3 55.5 58.3 57.9 1.0 56.7 57.1 56.7 58.3 56.3 1.0 55.1 55.5 54.7 58.3 38.3
.8 57.1 56.7 55.1 54.7 56.3
.8 54.7 56.3 55.1 57.9 51.9
.5 56.7 55.5 53.5 56.3 55.5
.5 54.7 56.7 56.3 58.3 57.9
.2 55.1 55.1 52.4 53.2 53.5
.2 57.9 58.3' 57.9 56.7 56.3 53.5 52.4 52.0 52.4 53.2 58.3 57.1 57.9 58.3 56.7 Cold Cold 3.2 54.3 3.0 55e5 53.5 53.5 56.7 '57.9 2.8 56.3 54.7 54.3 57.5 57.9 2.5 54.7 54.3 53.9 59.1 58.3 2.2 61.8 61.4 58.7 58.7_ 54.7 2.2 53.5 53.2 53.5 57.9 58.3 2.0 60.6 60.6 59.1 58.3 55.1 2.0 53.5 54.3 53.5 56.3 -58.3 1.8 61.4 61.4 59.5 57.8 54.7 1.8 53.9 53.5 54.7 55.1 58.3 1.5 60.2 60.2 58.3 57.8 55.5 1.5 54.3 53.9 53.5 56.3 58.3 l
1.2 59.8 59.8 58.3 59.5 54.7 1.2 53.9 55.1 53.9 55.1 58.3-1.0 59.5 59.1 58.7 57.8 55.1
- 1. 0 53.5 52.8 54.3 56.7.57.9
.8 59.8 63.4 58.3 57.8 55.5
.8 54.7 53.2 53.9 56.7 57.1
.5 59.8 59.8 59.5 56.3' 55.1
.5 55.1 55.1 _56.7 58.7 59.8 l
.2 60.2 59.5 58.3 56.7 55.1
.2 55.1 58.3 57.9 58.7 59.8 59.8 59.1 58.3 57.8 54.7 56.3 58.3 58.3~ 58.7 58.3 l
1
,m
yr o
n
// //
// /
/
/
/
t evn I
g e _
n t
i i
n n
f s a
/
/
/
/
i oo r ar pT p
t O
S u
a e
e t
4 n
i n
d s
/
a
.o r o r a
pT N
i p
l o
tn dn i
o a
x P
y e
ys i
p lp A
ee a
kb r
r u e
g ut l
o h T
b l
t n
a n "B l
oswT_
a a
l a
r i
o" t
5 k r e S v
c r o H
a ot i
b wa s
a r
n T
l e l
o an a
i t e y
r a t
ng ad t t
//
/ /
i e
r ne s
a mm
- amD n
i a X
a a
re i
r et D
t pa x
e e
te d
l t
p e
ynt m. e.
n e s
//
n d
o r
a d
e cel E r f.
l b l
u P
C gb a nal t
il i o
nia l
iav
)
h s
ava t a qt w
n d
ne A
o oyn ol ib U3 SSSSSSSSSSRSS cl e l
t a N
f i H cl sd -
ei f
naU S a o
oe i re v
A e
t l
g ct o d
eoi t
E SNW E
=
=.
E 8901 2345678901 2 E
SAU 778888888888999 R
o N
C N
u
)
)
l b
E a
b u
(
(
T
yr u
/
// /
tnevn I
g e
n t
i i
n n
/
/
f s a i
o o. T r
a g
r p
t o
S ss e
e t
n in
/
/
d s a r
o r a
pT p
i l
2 o
dn y
a r
x r
y e
us h
p S e p
A b
a nu r
ot g
e o b k "A l
e l
t n
l r
l u a b e al b
o" a o r l b a
w mT t
r e S ia l
l o i
6 H
al at vi a
t a aa v
e nr s
v a
l b
ee t
na a
mn n
o s
l i e a
e i n n
T r g y
r e
//
/
//
t o o
m e
ad t i
ii pm rne r
st t
x a
- am i
u ni ee X
a s
as s
t a
rn n
i d s D
e t a ar e
H r.
n et t t
n t
c a
iha e
st t l
l t
. o oon P
nt
//
poieh y e s
/
d l w r
pm d
e ossb E rT nag u
t C
oggol n nniii t
l o
si it an nnivi l
wiisaa
)
s t
- b oaan l t t ad n w
nnrno ne o
f oot ec ol ib ppmNuS pu u l
O f
f cc B
t a oOS o
e-s cl nnl U n f
ei o
rooo o
S a ei ih ei e
t t t wl t v
A g
ncc oc eeeoh e dE cSSNWS
=
= - = =
e
.NI S l
E E
C 23456789 pm o
l OS N
C e
i
)
)
b H
a b
(
(
uT
yr o
tn
! // / /
/ / /
evn I
g e
n t
i i
n n
f s a
/
/
ia oo r r
pT p
t S
O ss e
e t
n i
n d
s a
/
r o
r a pT p
i l
I O
d y
n r
a r
x u
y e
S s p
e l
e p
A nb a
ou r
t g
e k
o h l
o "A r
l t
n b
l o a a
w" a o r
l mT 7
t i
l r e S a
e ao M
v l
t t a
b na a
er s
s T
me t
l n
i n a n o
re y
r e i
m eg adt
/
/
/
/
t e
p r ne i
xm r
- am s
u ea X
a n
e s
a i
d t Da r
es e
t n
H n
.e a
t t l
ce P
t al ynt t p e s r
/
/
/
nm dd e
ioe.
E rT cl e u
s bl C
ngab onl a t
iiil o
t nai
)
l iiva b
s saav nt a
ne AA w
and ol l b NNSSuUSSUuUUURR o
rony t a NN l
t cel f
Bi cl es - d ei f
l nl l a
S a o
oo e
v hi er A
e wtl g
cot d
oeh o E
NSWN
),
=
= =
g E
E 2345678901 2345 E
NSUA o
1 R
N C
1 11 1 1 1 N
e l
)
)
b R
a b
u
(
(
T
Table 8 Initial experimental work plannef on Trojan tubes X-ray Eddy and Metallography and Hardness Sectiong Current Diametral Smooth Opposite Tube No.
Availablega)
Test Measurements Tran.
Apex Tran.
Rl-C26 U
/
/
/
/
/
RL-C8 U
V
/
/
/
/
)
(a) U = Whole U-bend available i
I
2.:9:: :
i i
i i
i 100 90 80
' Nx -
s l
N? '
s's Plant Tube c
Surr/I 3
70
\\.
.5 5
y x
7 60 N 'x' ',N 9
o
\\
'N ( <
11
- s,jq Ns Turkey Point 78 30 x
79 o
s s
G ao
- ~ j;,
\\
~. :~ ~
83
.2 4
i 84
=
i
(
.t s
i 30
/
t
^
o
./
g 20
- I
.5-
/
~ %v
' Plant Tube c
10
/
/
l N
/.'
Surn; II 1
E'
'X".
(
2 0
L l.\\7 &'L 5
l 7
8 d
i 9
i 1
I
- 1. 0
- 0. 0
- 1. 0
- 2. 0
- 3. 0 Distance Above ( +1 and Below (-) Intrados Transition ngue 2
- Ovality measurements were made for the leg with the most well-defined transitions for each of 13 tubes. Five appeared like previously defined opposite transitions R4ttom) and eight like smooth transitions (top). Data on Trojan Rl-C6 smooth im and oposite(a) transitions are included for comparison purpos:s.
I i
r I
I I
j i
i l
Hot lag I
I i
l 1
+
)
l i
l l
i e
l l
Cold l
Leg
{
Surry 2 Surry 1 Rl-Cl Rl-C3 Figure 21 Comparison of Surry 1 transitions (Rl-C3) with an opposite transition from Surry 2.
.=w,eeww~--+--=--w.ww-pr=,=,w-wme,i-wm=.-=-e=www._-me,-e-w-_
~
l l
l 4
i l
Ho*
\\
i i
i 6
}
i j
l l
4 I
t r
P i
?
i Cold i
r i
t Say
~ Mark urkey pog3t Nc*
Turkey poing yo_
Surry :
Rl-CS3 RI-c34 RI-C1 i
Itsure 22 conparisc",o'..
.' f,#
rintsofdeudieval{t 1
s radiographe) fe ab urkey poir,g 3, 4
I
',.es surry x3, 2, i
and an orp asi:e transiti -
~.
- wg mu::
7EB 7IB Extrados Transition Intrados Transition 7EA 7IA 1"
i n r2n 23
- Locations of transitions and of cuts on the opposite transition of Surry 2. RI-C7 are shown.
Identified sections were strained on the ID and cracks were found on only Section 7EA at the extrados 1
1
~
m9 ar l
yy
,f
.1
- .s.-. m 3
v w, [uf
,_. ~i r
$,h C
1
~:-
~..,*s
- rr a
1 i
I d2
?3i 14 5
55 l
.5 b,
033 Yi3
'I
.=
i
<%5 e.4 in
.=
Z u
Laaaauudts h
A 2 ma
i I
./
i 4,
l l,
a i
\\
' ; y. lt..
.i..
.c:<r6-
.u
,..q: gc.., t.
.'a~Q
- . L-s,
.~
. '. i. fx. (. ' ;t((..b.1.','/llil..e l
,s,.
p,.
i i
-) /:.t'l:..y;M,'-[a[ x:n.g) l
., f,','. '.9 ;' s C.j P ' 1..~.
b..
i s.
7
,. N,.. :.
v.
c 1...
.. a,*
., y..
. _a r.
....'3
- . -~.
t
\\
.M '. :.\\,
- ,. t..k..',.n...
c.p. l,k..
/ q.:
'Q
?
l
.4
..%.%h
- 4..
~.: l
. i.. p, -i. e.g.,. a. q..
/va-
...k,
~.
.)
l
.... s,
._ G) v.:.
- /'::...
.v.
f
(.)
a.
..,r 7 -
o (E _)
by >_R.yM).~
- c. y g,..}
,h.>N c' F.
[
- e. : <
m
- 2, e) b r.h Figure 25 Near axially oriented cracks on the ID at the extrados of Sect. 7EA of Tube I
Surry 2 (RI-C7) after straining the ID surface in tension. Those cracks are l
g within an area 0.2" below the extrados transition and 0.1" above the intrados f
b transition. Energy dispersive X-ray analyses identified At and P plus the elements found in Inconel 600 on the fracture face. A designates plane l
studied by metallography.
j i
{
t l
f
-.~..
---l
1 1
enl1.-x,. M.,-
h/.
e& l, M 4
.*~%
p a,.c.c1~h k a <
.. fq ' <
/
.j. 'tv.
cm...wA.%~W:,.{\\ ?.,h.,:y%. -
/'s' r.
s' 3
.,' 4,
w %,.%n
.s A
.~
>g n >. q ;c fl$p$.u 4...-R$5%?8% ?%N&"
__s
. n.,, -
4%'%w.MP&WF a..n s
%.r y #r, a,
. g, f.
o c ~,
w.o,. n. s.y..
.;.s..
y.,
ru...,T.7*.3..,.
s.
s.
t.%
i
.,J,s
- m. m ~ s Y jg M ' %=(
u,
~ q (.M $.. j, f.fy M
k
.. c >. T %. p s
n b*
.Es '
m,' r"A "yl g,' W.; W.;.j r.% y..
.e m
1 v-A 4,.
S ir.-
h %c
- ~
e.
g f -
, * =p KT ~.e-j
<'r.- 4;4f'. } < ) g.-,v r-
- p.
m ni-m3
,.y Q, f.C"._..
-f.,
76-
',7 ~
%,C;.g%g. ' $23%. '
a.
ko
~
g g
cvw p z
' % Ya
%', g g.
o 3
0
~:
- O j
--. 1:2)
Y. 4
-y Y. h * ~'. %
[
~ ^ '
W,.
% p r m ;. m.F.3 L
'd s
a' t g
4 ss.1.
1 w r-p( y.f w,,. &.#..;h.
u p.'m,m m h-
.r....... ~[ g,, t)*
' ;*4 W,,.
.\\. % ~ 'g ), 'Q\\,
c-
~.
n.
=
-.* C','e t,,,r:--j.7
'y '(' '
- y..,; - m.-
- ~7.'
'r g
&.. qm.-[. i
- u
'I *+.
d:ff C._;r',, t'*o r
' D,7\\ g';; 7 v.
.f
- -) M:
w o
2.
,z.
& *q eQ,3 N. *. '
~
i
~'
p f.L'
'A
,s4.
2
- p.
- 2., '.>, : ~,/.
7
*v r.
c-W.;
.,N,' g'a;.
4 d
') /
.. y ' 1, L
- 5,-
.'._ Q '
,a
.h v..?w -@,%.e.,Q~.5 as, w m m
+w...n.,2 v
' w'
, n, s-ac~ n=mn S
.o c... y,-m 2
m.7
+
m&_2-,( #u..;,'yw g.py:y_g'f.,.g v g g.
I.*u.
, xq.
r
-g,
.w,.s W%.a
~3 4
b.. y
-p,-
.,. s
~
D ]
'D fij$
- U 7[W"B o
f d
&f L & K
.=?
y :.;*,,,m.. r,',v.g.s.',. m c, a
l g.
A:;. W;..?.2.x-W.4. :.
..E.v.#,g,m&x's..p.
c:.'...+..... *,... '
2,;'-
.t.
m
.,,. ~.,..
& W,g d.e : g ;1 9g n..F
.... y.., 4 :. W.... Mj.g
.... =. -., a c. :. ::. m. m4: '..M.F. %. 'g: v..,. :., m&...,. e.
- -e,.
=
.,. 3..
m'Rw.u9 o
-G. W.,'
r.p.~; n. ;. 2. :. m.a.:.':
=
.m
,r n
.s,
.., :C*~
r4 s-
,o'.,.,.
..t~.,s s_
....a
'*M
-e f
^*
a C
&b&&&hlf$.W%
U
~
I3
%w, b'h,3%
L..~Q,s.f$qmg'*tf45WN o~bpA, }. % x(k s.c.ww" a
- =c a
~f
[
0E" mWm.em
.r m w
T i.::.:l.t. p:.
q :-Y-y @ ' ' Wl.;c G" '*Q r.%Wtiy
,E
- =
,,'.:-W
- U h N'Q,'L g.
av YN b **, ' ~
5
.c e
P E. M. 9 8 d"""@. N.N*#@f =, C@'f, M,,'@w 7
I I
$3h
. H ' 3.*I.
E
.tj m
?3
!!~
~. s-...
' %* _a.w.n. N
.s.
% m-.~
-Bu
.,h s g.
(.. +.
. p. w. 4.
(.
\\..,,.;*
ev l'
M w Q M.D...y...i
.m
- s. -
=u2 n
s>... ~....r,.
~
IN.D J ff..gd$ %'.sd D g,.. y,. g.:r..
s;;<'.
E "."
. MS-%
y-
- ,. e..m - l,.,y.
- .t oy
?v
..A
,.g 3. e..
- c. m W ~~- Qg.? '.;c]&.?:y.@h..mQ:%:hB..Sdy :,'.$i*, d,
- 4 o
- >E.
.-<. r 2
~! ~ G~
Mh;.
g.p epsC a
"D2 F :.yN.P)pt:'b *~ W;A'M7eQ*s.p@q.:.f i..r.'v. p l
blir 5,s,, +
a 1 yk:2
- e y
y. :,:
s
~
uQ
.=== C 1
,.q'*3 '<*. C..iD. V Jg4st 'v
-- Q.Y
- ;RG
- T a'"
d~
th 30 E.4 y.m.n u'
.,.m:. m
.s.. w wy +..g.c.: :2.s
. r..
.~.
= a y y.A M. W.>. M O,Nm <>h,pr.
f, G z '".E t 3.h?<**.W, s.Y.'W..u i on-
- 5. '.
s.
a
%.f. y.
n' b.-
C ~~' 0
,c -
C
.n
- x.. ;s..%.... %., p-a..r f..._
-, u?,.e*d.. q
... k.[.
s..,.
e..'
- o u I'~ m, e,
.p.,!'..
.,g"....A,,.g gg,.O.- *.
?~~,. * ' ':[' E*fs W;.W',1flW4,g7.f ?-4Q f*;:.'j?
f.
&6
- }.,-
s..
s et,4.,TW.,,A,$
ypx
.s y ;. s.
- S v
c,g; w
y
~. / '.c
. frt n
3
.a
,,.N' *...b... f. y.'.,'
, ~ Y.
.i y b.*.;;'t.
-. XA
')
.[ *,,, h i._ '. w;;&u s
'T.+'-
u s aev c- > :..:
.. f.v
&...;. f..~ _'% >. Q* *<.m Q ke v,.p.Cv.t-*, g},.~ y.
--. f.
tt;,
,.. l' yyv y y.,
l
-l-t
's
..M y2
..... 4 ;,.~ s.'. 7.f m
c,
,A.
- s. ~,,,w.
e i r.
.y o a,.
h
,,Y h *h d
- k*
3 ;.,, :h:'.*,,#.- a.:,.*RJ *= ' Q; g g.;.',J.' %.,/ g-[#'g1~-
y
]
'.,:j a.7 j
r I
.s....
- -...f
- ; e
.c: ~
I' y
a g
5oxo
\\,:". ".. '.64 auuu
. 3. a ~..,..
-d k w y
.n~'.
r.:. 4
- . - ~
1
.. ~. -
N 2 M 3 t's
'%. - y ".Ju... g t. j.'
c... T y,, a ;,g {,'.
f v
g N.a:
a > r./3 m i
y.
ym s
,,,3 f
..[
J. ' *.
g m
% g..
., e ~;
.~
4
.:. '.-f..,. s
- e. -s.
wi.e tal O Q lQ =
.... '*r g.
.s.
t m.
g u= u
..,..;*'} f,.. '. g.,,,.*,..e
.;?
cw y
- g. s,..
,,e
.o a,. *
,.e s.
n y
,,.e ~ ' / g g
g ya f
9-
^. : <,:..- $
,'5 4
== 0 0 *3 uay n h '.
Ac C
C 3 ;04
.A
- g. %.3'j',* *d.I'le'.A.,4 h, E *. ' *. :'.l 6
E
- D "3
- E u
Lw
.1 r:+1
- y w,
...... t 3
.s.
1
> c. s..~. w m :. n# m. [ h l* *
- 0. ! -
m..p.v.nm;;> yy.,,3 g,. > e..c- <
4 39 3
w.
v <:
s.
.(
.s..
h dN
),.'
', ~. h.
- 7 s.
. I' tA.
N
- ".. +.*,gl,,.,,p.'f. a.'j. h e.", ]4,,fg** [* p -jf.,.'.l[e *P'.h4
" g@j r' a..
y M
- 6..
,.I.* *,
y e
r fr W g
.<8 j u-
)
-I, l.
..j,'.
- s.
- %. 'r.
A t
2
...o.e a
C DU 44 u
-3 V
.O L H U2
1 i
i 1
l l
1 l
1 I
r i
Extrados Tran.
Intrados i
Tran.
l l
h a
l i'
Figure 28' Locations of transverse cross-sections through the i
opposite transition at NO.2" intervals on the unpainted (cold leg) of Trojan Rl-C8 tube
.J "T--
a34 'yg.ph" % f;, ^
' 40 s
.,v $i b 3
- v. : *s
~
"f!Y* 'fU&?h
?
'l
['b _, ' VnQ[tM.Q:.?c. <
l
,2x M
,t
,Q W:, b M
'iM%S Q.
1 Qc.
y;. }$
~it'Q MQ !.' C,-
T 9
r yn..ue+f-p:s~-
s.-
. -is m fy r
w l' $s(.-f /f s
,e h* lYh'G48,YY':y 'Q g!i*
'b'Ql'$
j - N$Y M2Q+it '.
l rn "W- '
$= y A g.,g'.d.' :x.!St++.:'
p?hV
-c s
f-3 n.-
,. +
.u,Nw.j.T
.w s*
/,~e"-
- $.-h', J'y ;
f
^
l}
'W.
9
- t...-;.. -
,,.h, '.
~' ' '
,2 Ad
- [
'ihT$g
~
r
.. P,.RM
%[A%
~
h@t&fr$2
- %d $$$$'515@!p$5)$5$$$
l O.4" above intrados tran.
~
y-w j_' y-Y g
A
'M nN
~
h4 4,
N VE"
' $&,,p,
. ; 4-
. [g..d
~. '
... %, & * <,,g.-.
D 4
._._s,
.,, ; t.
dkM '
,k
.q 4,
te
~.,
~
~$
" rpf.'."
+
l ljLb.
-'n
[ $. w %.-n. 7-.
3
,,.i
+
~-r'.
-A.
,, e 0.2" below intrados tran.
Figure 29 Microstructures and hardness readings at the extrados for the opposite transition of Trojan Al-CS.
D**3D *D ~TY o o J\\\\ oId k N a
---~---.e 9-g,--,-
-,.-wg.
-y*-
v vg--e--
,.---.--mw-y
--,y+-
,w e-+---,o-
-w-e-g
-s*--
ee a----m. --a-
w, e 725h.-A g
i i
i i
j iiiiiiiiil i
i i
ii i
i i
i 330 -
320 310 0.2" Above 300 htrados Tran.
3 290 8
m
~
}280 0.4" Above f,270 a~~%
,/
Intrados Tran.
z
~
~
ks i f At Intrados1 an.
N'
'd
/
0.2" Above Intrados s
260
'o
/
'/ 'O/1 ran.
8= 250 s
s vi s
,4
/
At Strados Tran.
x'N g
s
/
\\
240 e
s\\'
O. 2" B elow Intrados l
230 e
T ra n.
220 210 200
,,,,,,,,,,,,i,,,,
0 100 200 300 400 Angular Position. *
"8" 3
-Knoop hardnesses near and within opposita transition of Trojan Rl-CS tube.
.e :;5c s.a I
i i
350 1rojan Angular Position Tube 135*
180 225 340 RI-C6 O
o O
330 Rl-C8 v
0 a
O 320 0
.310 o
v 300 290 8
m i
280 o
E j
270 yo
$~
260 0
=
o 5
Y.
250 0
240 7
ha 230 a
Envelopa {'
20 For Rl-C6 '\\
i Extrados 210 Transi! ion 200 I
I I
t
-1.0 0.0' 1.0 D! stance Below(-) and Abovet N Intrados Transition ngun 31
-Comparison of Knoop hardnesses for the leaking tube. Trojan Rl-C6 and a tube frae of indications. Troian Rl-C8
1 i
l' Extrados transition Intrados transition Smooth opposite Transition Transition l
Figure 32 Prints of double wall radiographs of transitions on a tube from Heat No. 1202. The opposite transition was used for residual stress measurements.
1 b
i i
i l
l t
i I
i l
I r
t i
Curve 725432-A I
i j
20 e
10 e
m e
e
=
E e
e 0
e a
-10
\\
i Envelope from Prior 5 Opposite Transitions l
- 20 I
i
-1.0
- 0. 0 1.O Distance from Intrados Transition, in.
-Comparison of ovality measurements on the opposite transition of n ure 33 a
the tube from heat No.1202 ( individual data points) and the erivelope from 13 previously measured transitions
~ - ~ - - -
l l
l l
a-wa M U U
G @
Nw l
m-t O d
}
C. E
\\
if m
{
O l
vo G
i j
U
- x oc ma C4 "o u 6 C
- 1 g
a-N
~3 d
5 7
e-o=
3::
=
di "1
8".
.I
- =
i Q^
Q N
W mC MM U Q-Q
- C 7J G 2 O :=
b E
M Q C to u w-%
@ N W Q m 4.4 4
m g' 3 a oo
. *.3
- D W =
m CH I
lrJ >v M
E=
00 i
L L
.h
l -?,,
O bk
, AG
' (
c..*'
L - _,_ A.,
1:~,. &
_. L T,2 g; O 1
v.
i g
. a %..,
.. w W e,?.*
l!y>m-.... _... ' '.. ; J hQ.?l '. *;& '
~
.m
_ g' L
~-
vs.. <.- n -
S
@gj %.. "t !.,,
~.,.:.
S 'Y s g
- -
- J ',}.<3....
3j ~y.;. ~ --
r s
^
l n;y': -
l
[h 3..d N-[ $,
I.' * [
f I
%q.gA. $.... A QQ# ' 3'C,v ? pen.;
e g
+
l k
i ff n, k'.
A Extrados f* Tg f'?
)'
g.f Transition g,g.
..-@"^
4*
s ',y
_ y,~:;-
~
4,
{
. e,
..4 t.:
fifl h,k
.! ? '
N k
Level of l i,I*[ *. +.
- ..N..
[*C. :h:.
q[,' ' -
,)
l l
Intrados i
Transition Y '. S, 7 Ef y>%* : ~ **i?
}
l 1.
J u~
3M li-
')
i '
y, pq 1
- " yg.
- Q i y.g %,
p(r r.Q -),4;yp
.n-Q:c -
}
p g
~.
- g.
1 l
!.X
- ,e
/, 1., '.
7 ft ~ '..., : '
(j x
. F.
- . x c[
,bifi'$
l.
E c
- ~
~
llJQQi; w.,;,. gy,y y
_.y; 9-
~
d 5 -
E
p vif.:.{,;Q.QX54:
~.
- . y.e 0,
- ^d
~
m-@ [y'K.,.,9,3 _ %... w. ~
w rp n
t
.a l.
p
- 9.'Q' b.- g o.
t
?
t
- ' e, ',.' fl. ; --.
,,y, h
- [,k
~
j 9*
~
OD C
"
- 3 ';
Locations of CD strain gauges on the extrados half section from a Tube (Heat 1202).
Numbers to the right strain gauge numbers are wall thicknesses in sils.
or After removal of the half section, strain gauges were placed at corresponding positions on the D
Table 9 Sectioning stresses (ksi) af ter removing strain gaged sections on the extrados section of a tube from heat no.1202 Circumferential Axial Location Remarks O.D.
I.D.
0.D.
I.D.
1&2 Above
+11.9
-18.6
-17.8
-12.2 Extrados Transition 3 &4 At Extrados
+ 8.4
-25.1
-68.5 8.8 l
Transition 5&6 Between
- 4.0
+ 1.1
-20.5
-- 8.2 Extrados and Intrados Transition 7&S Below
+ 0.8
- 0.9
-15.3
- 4.5 Intrados Transition i
l l
L
i i
20 l
0
~
=
=
~
- 20
- N s.1 + 2, above extrados transition 10 h.
s g
Nos. 3 + 4, at extracos tran
- 10
- 20 N
._j2-30 l
d U
g f - 10 c
c_
~
m
" - 20 oms. 3
- 6. elevation between intrados and l
- 30
- x..:cos transitions
- d0
- 50 10
^
0 l
Nos. 7 - 3
+
- 10
- below extrados transition
%m o
0 2
4 C
S 10 12 la 16 18 20 22 24 26 28 29
-3 Distance Below ID Surface, lej in.
Fi p e 36
_.gyj3l str"sses 3i i fiot 5 l0 Cations and deDihs fron the ID surface fcr the extrados section from a tube ( heat No.1202) d.lf*h[\\\\ "'DT99M D
1
..M v1 A110o
l 20 0
s
=
~
~
~
~
l
- 20
os.1 + 2. a,ove extrados trans,t,on e
ii 10 10 0
- 10 3-
/
Nos. 3 + 4. at extrados transition 4
- 30 n
10
-m
~
0
~-
7
_ [ -
- 20 [ hos. 5 A i vation vt'c;een i.1trados and extrados J
tran sitior s
- 30 o-10 c=
=
=
-u 0
+
=0
..os. 7 A 2. belov.'intrados transition
- 10 0
2 4
6 S
10 12 14 16 IS 20 22 24 20 28 29 Distance Below ID Surface.10_'2.
in.
"8"'* 37
- Circunfer: ntial str?sses :t various locations and depths f rom the ID surface for the extr3 dos section from a tube ( Heat 1202)
'J J
It is, however, possible that some percentage of the material in the suspect tubes in the Trojan stese geserators may be microstructurally susceptible to the type of intergranular stress corrosion cracking obse rved. This percentage is not known, nor is the oquality of aicro-structural susceptibility among tubes produced from the same heat known. Tube hollows of the same heat processed at different times will not necessarily produce tubing of equal susceptibility within a given exposure time frame (ignoring possible stress level variations).
Operating plus residual stress level is the other variable of impor-tance. Based on laboratory examinations and eddy current data inter-pretation, indications and leakage appear related to the presence of the " opposite" transition at the extrados tangent. Indications and cracking have been observed on both hot and cold legs at Trojan, in roughly equal proportions. There is also no reason to suppose other than a random orientation of the " opposite" transition relative to the hot and cold legs (12 to 14 distribution on Row 1 tubes from Trojan).
Therefore, the existence of differential thermal stresses from leg to leg during operation would not appear to be a determining factor.
(
Excessive ovality is known to increase operating stresses on tubing.
Such an increase could well explain the failure at Doel II (ovality >15 percent). The macroscopic ovality of the Trojan tubing is modest by comparison (<6 percent in all tubes removed from Row 1), and is gener-ally higher on the smooth transition, where leakage and indications do not appear to occur.
For the U-bends aramined here by Westinghouse, " opposite" transitions were noted on the Trojan and Surry II tubes, and not on the Surry I l
and Turkey Point 4 tubes. A frequent feature on the Trojan tubes was a change from positive (flanks widest) to negative (flanks narrowest) ovality in or near the transition region. R1-C6 was especially marked in this respect.
It thus appears that there may be a correlation l
between this configuration and an increased stress level (operating plus residual or residual or operating alone). This increased stress combined with the potential microstructural susceptibility may account for the observed location of leakage and cracking. 1
N ATTACHMENT 3 PORTLAND GENERAL ELECTRIC COMPANY'S DISCUSSION OF WESTINGHOUSE REPORT I
The mode of failure observed in the Trojan U-bends most closely resembles the " pure water" intergranular stress corrosion cracking observed in deoxygenated pure water in laboratory environments. This phenomenon has been reproduced a substantial number of times in heate of mill annealed Inconel 600 tubing typical of at least one manu-facturer's former nuclear grade production. Of 10 heats exposed as stressed U-bends at temperatures of 345'c and 365'C, five heats were observed to be susceptible to such cracking (l). Bulischeck and j
van Rooyen also noted that " Failure times in primary water environments appear to be unchanged from those in pure water"(1).
The variables related to such intergranular stress corrosion cracking in field service would appear to be the following:
time, temperature, residual plus operating stresses, and microstructural susceptibility.
Since time is an independent variable, and temperatures over the ID surfaces of all Row 1 U-bends would seem to be equal within any l
l meaningful range, variations in the behavior of individual U-bends would appear to be determined by the other factors mentioned.
l Assessment of variations in microstructural susceptibility is not l
feasible, since the causative factor (s) rendering a microstructure l
susceptible are not known. Microstructures can be rendered non-susceptible by " sensitization" or by " thermal stabilization" heat treatments.. The causative factor (s) that are changed by such heat treatments have not been determined.
Bu11scheck ard van Rooyen comment on the difficulty of judging susceptibility on the basis i
of microstructural analysis, and state that "small variations in processing history which occur within a mill or different mills must play an important role"(1). This is compatible with the findings in the Trojan case that no consistent and significant relationship could be established between cracking and grain size, carbide distribution, minor element chemistries, and hardness. l l
. -