A15094, Effect of an Aluminum Alloy on Sanicro 31 at Elevated Temps. Final Rept on Metallurgical Examination Re Potential of Aluminum Alloy Chip.Steam Generator Leak Cracking from Liquid Metal Embrittlement Unlikely

Effect of an Aluminum Alloy on Sanicro 31 at Elevated Temps. Final Rept on Metallurgical Examination Re Potential of Aluminum Alloy Chip.Steam Generator Leak Cracking from Liquid Metal Embrittlement Unlikely

ML19274D338
Person / Time
Site:	Fort Saint Vrain
Issue date:	11/30/1978
From:	Holko K GENERAL ATOMICS (FORMERLY GA TECHNOLOGIES, INC./GENER
To:
Shared Package
ML19274D332	List: A15094, Effect of an Aluminum Alloy on Sanicro 31 at Elevated Temps. Final Rept on Metallurgical Examination Re Potential of Aluminum Alloy Chip.Steam Generator Leak Cracking from Liquid Metal Embrittlement Unlikely ML19274D331
References
GA-A15094, NUDOCS 7901230204
Download: ML19274D338 (45)

Text

-,.

GA-A15094 THE EFFECT OF AN ALUMINUM ALLOY ON SANICR0 31 AT ELEVATED TEMPERATURES by K. H. HOLK0 F

DATE PUBLISHED: NOVEMBER 1978

~

GENERAL ATOMIC COMPANY 790123020L[

.j, GA-A15094 THE EFFECT OF AN ALUMINUM ALLOY ON SANICR0 31 AT ELEVATED TEMPERATURES by K.H.HOLK0 GENERAL ATOMIC PROJECT 1900 DATE PUBLISHED: NOVEMBER 1978 ag'

-Y' 5 ', -

.,hs,"

l..,".

s -.

'h g

0 GENERAL ATOMIC COMPANY

6 ABSTRACT As a result of the steam generator tube leak at the Fort St. Vrain generating station, an investigation was conducted to determine the effect, if any, of an aluminum alloy tool chip that was left in Module B.1.1 during construction and installation. Tests on stressed Ni-Fe-Cr alloy and 2-e/* Cr - 1 Mo (principal materials of steam generator construction) showed that, while some erosion of both materials could result, no rapid cracking or grain boundary penetration type liquid metal embrittlement occurred.

It was concluded that it was unlikely that the aluminum alloy chip causr3 the steam generator leak either by erosion or liquid metal embrittlement.

O 9

iii

CONTENTS ABSTRACT.

iii 1.

INTRODUCTION.

1 2.

LITERATURE REVIEW 4

3.

EXPERIMENTAL PROCEDURE.

5 3.1.

Aluminum Alloy Identification 5

3. 2.

Test Cycle Determination.

5 3.3.

U-Bend Tests.

7 3.4.

Stress Rupture Testing.

13 4.

ALUMINUM ALLOY CHARACTERIZATION 15 4.1.

Aluminum Alloy Characterization 15 4.2.

U-Bend Tests.

18 4.2.1.

Specimen Qualification.

18 4.2.2.

Cracking Examination.

18 4.2.3.

Erosion Measurements.

27 4.2.4.

Phase Analysis 30 4.2.5.

Stress-Rupture Tests.

34 5.

DISCUSSION.

42 6.

CONCLUSIONS 43 44 7.

REFERENCES.

ACKNOWLEDGMENTS 47 FIGURES 1.

Steam generator module arrangement 2

2.

U-Bend test specimen 10 14 3.

Typical stress-rupture specimens 4.

Microstructure of cast tooling plate. Etchant:

1/2% HF in H O 17 2

V

FIGURES (Continued) 5.

Heating curve for cast tooling plate (heat input rate held constant).

19 6.

U-bend qualification specimen after exposure to a silver base brazing allcy 20 7.

Typical U-bend specimen after helium control case test cycle exposure 21 8.

Cross section of a typical U-bend specimen (S-1-10) 23 9.

Cross section of restraint bar/U-bend weldment after helium control case test cycle exposure (Specimen S-1-9) 24 10.

Appearance of typical U-bends during and after metal control case test cycle exposure 25 11.

Cross section of a typical U-bend after metal control case test cycle exposure (Specimen A21) 26 12.

Erosion from helium control case test cycle exposure (Specimen A13) 28 13.

Erosion from future SH II test cycle exposure (Specimen S-1-29) 28 14.

Erosion of Sanicro 31 by aluminum alloy.

29 15.

Cross section of welded 2-1/4 Cr - 1 Mo U-bend specimen (CM-30) after exposure to future SH I test cycle 31 16.

Scanning electron microscope energy dispersive X-ray analysis location on typical U-bend cross section exposed to helium control case test cycle (Specimen S-1-8) 32 17.

Region examined by scanning electron microscope energy dispersive X-ray analysis on cross section of Specimen A-20-1 exposed to metal control case test cycle.

35 18.

Stress-rupture tests on solution-annealed Alloy 800 at 1050*F in helium af ter coating with cast tooling plate aluminum alloy 38 19.

Photographs of typical stress-rupture fracture surface (Specimen B-20-2) taken with a scanning electron 40 microscope 20.

Cross section through typical stress-rupture fracture 41 surface (Specimen B-20-2) vi

TABLES 1.

Module B.2.3 temperatures and times in Superheater II Region 6

2.

Test temperatures and time cycles used 8

3.

Chemical composition of Sanicro 31 9

4.

U-Bend test matrix 12 5.

Chemical analysis and melting range of cast tooling plate compared to similar aluminum alloys 6.

Chemical composition of coating phases, cast tooling plate, and Sanicro 31 by scanning electron microscope energy (

dispersive X-ray analysis for helium control case test cycie (Specimen S-1-8) 35 7.

Chemical composition of coating phases and solution-annealed Alloy 8, by scanning electron microscope energy dispersive X-ray analysis for metal control case test cycle (Specimen A20-1) 36 8.

Stress-rupture test results of solution-annealed Alloy 800 at 1050"F after coating with cast tooling plate at 1180*F for 10 minutes 37 vii

1.

INTRODUCTION In November 1977, a steam generator leak developed at Fort St. Vrain.

Testing isolated the leak in the Superheater II section of Module B.1.1.*

Investigation of the quality assurance records on this module revealed that, during construction, the tip of a pry bar made of an aluminum alloy broke off and became lodged in the Superheater II section of Module B. I.1 at the approximate location shown in Fig.

1.

Testing at the time of construction was done with the assumption that the pry bar had been made f rom 6061 aluminum. Test work conducted by simulating predicted temperatures and times for Module B.1.1 with the aluminum alloy 6061 in contact with Sanicro 31 concluded that there was only a general erosion of Sanicro 31, the attack would be minimal, and no liquid metal embrittlement would occur.. The decision was made to leave the aluminum chip in place because of the minimal and acceptable pre-dicted damage. The aluminum alloy chip was approximately 0.75 x 1.5 in.

with a thickness of 0.25 in. tapering to a thickness of 0.06 in.

This R

represents a total weight of about 7.9 grams of aluminum alloy.

When Module B.1.1 developed the leak, the investigation into this incident was reopened. This was done to determine if the aluminum alloy had caused the leak.

During re-examination of the incident, suspicion was cast on the assumption that the alloy type was 6061 when conversations with site personnel involved during manufacture and installation of Module B.1.1 revealed that the aluminum alloy pry bar had exhibited brittle behavior. This behavior is not typical of an alloy such as 6061 with approximately 17% elongation displayed in a room temperature

• The Superheater 11 section is made of solution-annealed Alloy 800 (Ni-re-Cr) purchased from Sandvik under the trade name Sanicro 31.

The Superheater I section is made of 2-1/4 Cr - 1 Mo low alloy steel in the annealed condition.

I

-u.. n......

WPett StaL

.l,

'~

rC t,,....

dI'IN_

I atat av ta -

l l a

Ni if i%

i y

fl8"I j % )

LOCATION OF LEAK h

j M

IN MODULE B.1.1 I

I 5#PER, eta,tB '

(v areta,0R Ecomonellt

1. eos ceDr toute sta Ik) 9

'" " ~'M EE

= t, L.

,,E,_<t,

k. lr, fi'f4' l'

/

v l

3 r.. d a is.

reinas, cLosuar.

7, a i -m

~

I i.

\\ l l

^

6

!!,'k...

b j.

I it.":;*it'"'

3"i>! L t.

[-l.

}

.I t

y Pce,\\ t d p1.i s

i

, s.,E -

Q*NJ""

aei gr.-

_s -

[if f,'

"J!';*

V 'l l

,,n.,,..

. s(+ f_.

ca. u,.., ist,

. =, -

i h : ef g g

,.E..,

nu

= s-Fig. 1.

Steam generator module arrangem2nt 2

tensile test.

Because of this, the man who made the pry bar was contacted.

He indicated that he still had some of the original ti.terial that he recollected as having been used to make the pry bar. He described this material as cast tooling plate.

Since this raised the possibility that Module B.1.1 contained an alloy other than 6061, another investigation was initiated to evaluate the new alloy.

In this investigation, as in the first, primary attention was directed to potential liquid metal embrittlement attack.

Remnants of the material that the tool maker described as the starting stock, or cast tooling plate, were obtained and used in this investigation to determine if liquid metal embrittlement of stressed Sanicro 31 was possible. Conditions of past Superheater II exposure, as well as future anticipated Superheater II times and temperatures, were simulated. A literature review was also conducted to determine if liquid metal embrittle-ment had been encountered in any similar base metal compositions.

4 0

3

2.

LITERATURE REVIEW Some 30 books, reports, and articles were reviewed in preparation for and during this work. Of these, 20 were considered to be significant; they are listed as Refs. 1 through 20.

In summary, the literature review indicated that it was possible for liquid metal embrittlement to occur in austenitic alloys in combination with elements commonly present in aluminum alloys under certain conditions.

The specific combination thought to be present in Module B.1.1 was not addressed in the literature. Based on metallurgical experience and sug-gestions from the literature, the ocroected material combinations and operating conditions were investigated for liquid metal embrittlement.

e a

0 4

3.

EXPERIMENTAL PROCEDURE 3.1.

ALUMINUM ALLOY IDENTIFICATION A portion of the cast tooling plate that was identified as the raw material for the pry bar was obtained for analysis, and its microstructure was determined to confirm that it was in the cast condition. A temperature /

time heating curve was run on this material with an embedded thermocouple to determine the solidus and liquidus temperatures.

The chemical composition was determined by two different test laboratories. This was done by the atomic absorption technique along with a wet chemistry tec'..tique for smaller percentages of elements by chemistry laboratory No.

1.

It was done entirely by atomic absorption analysis by chemistry laboratory No. 2.

The more accurate analysis then should be the one performed by chemistry laboratory No.

1.

3.2.

IEST CYCLE DETERMINATION Time-temperature test cycles were established for testing in this pro-gram.

Cycles were established to represent " worst case" service prior to the steam generator tube leak, and also to represent " worst case" future anticipated steam generator module service. The temperature records kept for instrumented Module B.2.3 were used to establish the operating history for Module B.1.1.

This was done because it has been shown that Module B.1.1 performs within 10*F of Module B.2.3.

Table 1 is a summary of the operating history for Module B.2.3 in terms of helium and metal temperatures prior to the leak and the anticipated temperature-time maximum profile for future service.

5

TABLE 1 MODULE B.2.3 TEMPERATURES AND TIMES IN SUPERHEATER II REGION

[

r Max Temp Period Location Helium Metal (hr)

Prior to tube leak Superheater II inlet 1157 1030 7.5 Superheater II outlet 1110 1065 2.8 After tube leak Superheater II inlet 1400 2.8 repaired (a)

Superheater II outlet 1215 2.8 Superheater I inlet 1210 2.8 Superheater I outlet 1050 2.8 (a) Predicted future temperatures under transient conditions.

O 6

From the temperatures given in Table 1, several test cycles were developed; they are given in Table 2.

Dealing first with the operating history prior to the tube leak, two cycles were identified: a helium control case and a metal centrol case.

In the helium control case, it was assumed that somehow the helium could entirely control the temperature of the cast tooling plate in Module B.1.1 and, in the metal control case, the metal temperature would control.

By comparing Tables 1 and 2 it can be seen that both the tempera tures and the times for the test cyc; 7s are conservative in that they are both higher and longer than the temperatures and times actually experienced.

This was intended to maximize any deleterious effects of the cast tooling plate on the steam generator material. Also shown in Table 2 are two future cycles, one for Superheater II and the other for Superheater I.

These cycles are based on future anticipated metal temperatures. Although longer times at lower temperatures will certainly be experienced in future operation, it was judged f rom the times described in the liquid metal embrittlement literature that the cycles selected here represent a critical evaluation for liquid metal embrittlement.

3.3.

U-BEND TESTS U-bends with a 2t radius (i.e., bend radius = 2 x wall thickness) were made from approximately 0.5-in.-wide slices of Sanicro 31 cut from Fort St. Vrain tubing. The chemistry of the U-bend material is given in Table 3.

The shape of the U-bends is shown in Fig. 2.

In order to introduce stress during the U-bend test, the open ends of the U were closed approximately 0.015 in., which was calculated to cause at least yield point stress in the U-bend material. To hold this restraint, a bar made of Sanicro 31 was welded across the open end of the U, as shown in Fig. 2.

7

TABLE 2 TEST TEMPERATURES AND TIME CYCLES USED Cycle Temperature Time (hr)

Helium control case 1180*F

<10 Metal control case Coated (*

+ 1065'F

<1000 Future - Superheater 11 1215'F 3

Future - Superheater I Coated + 1050*F 10

(* Coating consisted of immersion in liquid cast tooling plate at 1180*F for 10 min.

e D

8

TABLE 3 CHEMICAL COMPOSITION OF SANICR0 31 Specimen Type Heat No.

Fe Ni Cr C

Ti 41 Mn Si Cu P

S U-Bend 5-67864 Bal(a) 33.2 20.67 0.055 0.37 0.41 1.15 0.64 0.03 0.01 0.005 Creep Rupture A Series 6300 44.06 32.97 20.64 0.04 0.46 0.5 0.79 0.28 0.23 0.007 B Series 6295 44.34 31.38 21.57 0.06 0.5 0.49 0.98 0.42 0.23 0.007 e

C Series 6624 44.61 33.95 19.02 0.03 0.49 0.53 0.77 0.24 0.33 0.007

(*) Balance.

2t BEND RA'010S

~

j-[gu-1/2fN.-WID E STRIP FROM pg,

SANICR0 31 TUBING

'g 7-

\\';,

~

INCONEL 82 WELDMENT

~ SANICR0 31

- ' 2 RESTRAINT BAR

.J Fig. 2.

U-bend test specimen 10

U-bends were also made from SA-213 Type T-22 (2-1/4 Cr - 1 Mo) tubing that had been butt-welded with Type 521 (2-1/2 Cr - 1 Mo) filler as in Module B.1.1 Superheater I construction. The weld was located approximately in the center of the bend with the size, shape, ano location of the 2-1/4 Cr - 1 Mo restraint bar as described for the Sanicro 31 U-bend.

These U-bends were used to evaluate the potential liquid metal embrittle-ment of the Superheater I material.

To qualify the Sanicro 31 U-bend specimen, a typical specimen was coated with the silver base brazing alloy bag-1 (45% silver, 15% copper, 16% zinc, 24% cadium) which is known to cause severe liquid metal embrittle-ment cracking in Sanicro 31 at temperatures in the 1100* to 1200*F range.

A typical U-bend specimen was heated to the brazing temperature (1145*F) in an air furnace with the flux and braze alloy preplaced and held for 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. Af ter cooling, excess braze material was removed by acid leaching.

Dye penetrant testing and metallography were used to evaluate the specimen.

Actual U-bend testing was done with the test cycles identified in Table 2.

For the constant temperature test cycles, the U-bends were immersed in the liquid cast tooling plate and held for various times.

Immersion was done under a partial pressure of several different inert atmospheres as shown in Table 4.

Cast tooling plate quantities varied from 60 to 100 grams which, compared to the actual pry bar chip weight of 7.9 grams, added conservatism to the test.

After the predetermined time was reached, the specimen was withdrawn from the liquid melt and allowed to cool. Evaluation for liquid metal embrittlement was done by sectioning the U-bend and examining it metallographically. Cracking, grain boundary penetration, and erosion were searched for at magnifications to 500X.

For the test cycles, which involved first coating the U-bend specimen and then exposing it to a lower temperature, the U-bend specimens were immersed in the liquid metal for approximately 10 minutes and then both were cooled to room temperature. To complete this type of cycle, the specimens were removed and placed in a separate furnace where aging was 11

e 21.BLE 4 U-BEND TEST MATRIX Results Coated with Grain Aging CTP(8) Prior Exposure Erosion Soundary Material Treatment to Exposure to CTP Exposure Depth Cracking Penetration Specimen Type

('P/hr)

(*F/ min)

(*F#hr)

Environment (in.)

Observed Observed S-t-7 Saniero 31 None No 1800/1 CP argon-51 M

<0.001 No No at 0.6 torr S-I-8 1180/2

0. 002

$-!-9 1180/4 S-I-10 1180/9 S-1-13 1180/4 CP argon at

').015 0.6 tarr

$-1-14 1180/10 0.015 S-1-18 1180/2 0.007 5-1-16 1180/4 0.011 5-t-20 1180/10 0.011 t

5-1-17 1180/10 1065/100 CP helium at 0.0005 1 atm 5-1-19 1065/1000 O.002 S-1-21 1065/1

, 0.001 S-1-22 1065/10 0.001 S-1-29 1215/3 CP argon at l0.036 I

i 0.6 torr A13 1100/200 No 1180/10 0.010 A21 1100/200 1180/10 1065/10 CP helium at 0.002 I atu j

A22 1100/200 1180/10 CF argon at 0.015 0.6 torr A23 1100/200 1065/100 CP helium at 0.002 1 ata A24 1100/315 None 0.002 A25 1100/)f5 1065/1 CP helium at I sta A26 1100/315 1065/10 0.002 A27 1100/315 1065/100 0.002 A29 1100/314 No 1215/3 CP argon at 0.007 0.6 torr Ot)O SA21) Type T-22 1275/t *I 1180/10 1050/10 CP helium at 0.001 I

(2-1/4 Cr - 1 Mo) 1 sta CM11 None

<0.001 i

(a) Cast tooling plate.

(")Commarcially pure.

'Postweld heat t rea tmen t.

4 12

done for longer times in a chemically pure helium environment at 1 atmosphere pressure. This particular case represents the instance one might postulate where the liquid alloy is deposited on a cooler tube, undergoes partial solidification, and is exposed to a lower temperature for some longer time.

Evaluation was again done by metallography.

3.4.

STRESS RUPTURE TESTING

\\

Since the literature describes the low strain rate test as being most discriminatory for revealing liquid metal embrittlement, the stress-rupture test was selected to evaluate the effect of the cast tooling plate alloy on solution-annealed Alloy 800. Specimens for this test were cut from the wall of solution-annealed Alloy 800 tubing with the chemical composition given in Table 3 and the shape shown in Fig. 3(a). This was not actual Fort St. Vrain tubing, but was tubing produced in a similar manner and with a similar chemical composition.

Since the literature also states that the hardened condition of a given material is most susceptible to liquid metal embrittlement, the stress-rupture specimens were tested in the as-received condition, prestrained 20% in tension, and prestrained 20% in tension plus being aged'at 1100*F.

This processing was done to the specimens before coating with the cast tooling plate.

Coating was done by immersing a gage section of the creep-rupture specimen into liquid cast tooling plate at 1180*F for approximately 10 minutes.

Approximately 50 to 60 grams of cast tooling plate were used per group of three specimens, which were allowed to solidify and cool.

Speci-mens were cut from the solidifed structure as shown in Fig. 3(b).

Stress-rupture testing was conducted in a commercially pure helium environment.

Specimens were deadweight loaded and tested either to rupture or until the base metal stress-rupture life was exceeded.

Specimens were evaluated on the basis of tine to rupture, light metallography on cross sections, and scanning clectron microscopy on the fracture surfaces.

13

AS-RECEIVED CONDITION DOUBLER PRESTRAINED 20%

(a) BEFORE COATING s.. _.

CAST TOOLING PLATE ay COATING t'

n. a

,d t_~

m

..w-..

(b) AFTER COATING Fig. 3.

Typical stress-rupture specimens 14

4.

RESULTS 4.1.

ALUMINUM ALLOY CHARACTERIZATION The analyzed chemical composition of the cast tooling plate is given in Table 5.

As seen, the analyses from the two chemistry laboratories agree fairly well. As previously described, it is believed that the composition determined by chemistry laboratory 1 is more accurate.

Com-parison of the cast tooling plate to several other common aluminum alloy chemistries (Table 5) shows that the chemistry of the cast tooling plate is most similar to a 7079 aluminum alloy composition.

It may be that when the cast tooling plate was manufactured it was intended to be close to a 7079 or a 7075 aluminum alloy composition, but did not exactly meet the specified composition range.

It is common industrial practice to not require the producer to meet specific composition limits when manufacturing or selling such a product.

The fact that the cast tooling plate is indeed a cast product form is shown by the microstructure of the cast tooling plate in Fig. 4.

This is a typical cast alusinum structure having a segregated solidification structure with several types of intermetallic compounds in the as-cast morphology.

It is seen that no hot or cold working was introduced to this structure to break up these compounds.

It may also be noted that this type of structure, along with the chemistry given in Table 1, explains the low ductility of the pry bar used at Fort St. Vrain.

Cast aluminum alloys typically have room temperature tensile elonga-tions as low as 2% with alloy compositions leaner than those determined here (Ref. 21). Thus, it would be expected that the cast tooling plate machined into a pry bar shape could have room temperature tensile ductility as low as 2% and possibly even less.

15

TABLE 5 CHEMICAL ANALYSIS AND MELTING RANGE OF CAST TOOLING PLATE COMPARED TO SIMILAR ALUMINUM ALLOYS l

Melting Range Material Source Al Mg Zn Cu Fe Si Mn Cr Pb Ti Ni

(*F)

CTP

• Chem lab 1 Bal(

2.53 4.93 0.86 0.62 0.22 0.18 0.09 0.02 0.02 0.01 1005-1180(c)

CTP Chem lab 2 91.03 2.81 4.05 0.76 0.38 0.31 0.26 0.11 0.001 0.12 ND(d) 1005-1180(c) 7075 Nominal (*)

Bal 2.1-5.1-1.2-0.7( )

0.5 0.3 0.18-0.1 890-2.9 6.1 2.0 0.4 1175(*)

5 7079 Nominal

• Bal 2.9-3.8-0.4-0.4 0.3 0.1-0.1-0.1 900-3.7 4.8 0.8 0.3 0.25 1180(e) 7277 Nominal

• Bal 1.7-3.7-0.8-0.7 0.5 0.3 0.18-0.1 2.3 4.3 1.7 0.35

(* Cast tooling plate.

( } Balance.

(

Measured.

(

Not detected.

(* Ref. 21.

(

Single values are maximum.

N

.k

.3

,it r-

., Q r..

I 4.

7 o

'J*.

hh n

k t

t sJ'y:p.

.y f

,$.u-s*7 ; 7 j

$N

~

y' N

. i.],9

~

3 fy}s.

~

x j/.,

k*

. ('

,a/'

6. Wj j' }A?. s t u

1~

'/

e 1'

c-: 4

\\y

,,r,

$ n. \\

. m:

./g j' s., ;,3 c

e *, ^ '/P

s-

+,.

M

' d)d I 6

J 50 V5 Fig 4-Microstructure of cast tooling P ate-Etchant.

1/2% HF in H O l

2 6

17

The melting range determination made is shown in Fig. 5.

It may be seen that from slope changes in the heating curve, a solidus temperature of 1005'F and a liquidus temperature of 1180*F were determined. Compared to the literature for similar alloy compositions, as presented in Table 5, the

~

liquidus temperature agrees very closely. However, the solidus temperature determined here is about 100*F higher than the literature values. Since the techniques for determining solidus temperatures in the literature are more sophisticated than used here, it may be inferred that the solidus temperature is closer to 900*F, as shown in Table 5.

4.2.

U-BEND TESTS 4.2.1.

Specimen Qualification As described in Section 3, -he U-bend specimen was qualified by coating with a silver base braze alloy that is known to cause liquid metal embrittlement of Sanicro 31.

The results of this test are shown in Fig.

6.

Figure 6(a) shows an isometric view of the specimen after dye penetrant testing to expose the large crack. A cross section of this crack is shown in Figs. 6(b) and 6(c). This test, then, does show that the specimen is capable of revealing liquid metal embrittlement and, in the particular case of the silver base braze alloy, at least, the attack is rapid and inter-granular.

4.2.2.

Cracking Examination As previously described, the U-bend test was used to determine if the cast tooling plate material could cause cracking or intergranular peneera-tion. Also as previously described, several test cycles were used to represent past and future steam generator performance. Figure 7 shows a typical U-bend specimen that has been exposed to the helium control case cycle given in Table 2.

As can be seen, the specimen retains significant coating thickness, which in this case is approximately 0.050 in.

18

1300 0

1200 LlQUIDUS - 1180 F 1100 1

SOLIDUS - 1005 F 000 C

900 g

E Q

800 5

o.s 700

'o s

600 500 400 e

e 300 I

I I

I I

I I

I I

I I

I O

20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 TIME (SECONDS) i Fig. 5.

IIeating curve for cast tooling plate (heat input rate held constant)

~-~ -

. ' ', NMI;34.. ~ -

, -L%e '%i,Q.. %

,- j

. 'p

,b,

.'..,,..s'.

. ;. Llh,:

...- o ; 4.,m%

._y-m.

.\\

e

-e ge+

~ AjN.

Y.

~

.e b

- ry (a) ISOMETRIC VIEW OF DYE PENETRANT TEST

-. Q $.,.'v.

• a y/..

?.----

'e*

g.

^

'. 5,...

- - ~~ *f =*

,),, f l'.. _ - -'7-.

ti '

5

~

g-3

.y.--

ja..

s. -- 3 2 g.

..t-,

,s

?., :r g'

.< a -.s,4,. y &

d*.t% 3r /; a 9< ' T =

3,4 _.g-s s

• [.

N.'.r,.

em,,: 4.

} *4

% f > ;b

. E5 M g

E csh

.t..,

w,.

59 %,,. >?. '_;

.d4*y q',',;';,lt 4.y'

~.'

'k

4 t*

4 h

_Q-

^

me

_.__.m_

g op.

h hi 7

..i,

pr '

~

l n'

A 4

S.

g4

.g' ;

1000 pn m

{

\\

N (c) CROSS SECTION ETCHED WITH 10% OXALIC g-ACID, ELECTRO LYTsCALLY s

s,

I i

'\\

Wm.1 e

M 1000 un (b) CROSS SECTiON OF SPECIMEN IN (a). UNETCHED Fig. 6.

U-bend qualificacion specimen after exposure to a silver base brazing alloy 20

RESTRAINT BAR SANICR0 31 U-BEND

~ ' y,~< y ~ q

^

? *-

CAST TOOLING

(

- PLATE CO ATING

.g-k *Y

~

q

'i..

Fig. 7.

Typical U-bend specimen after helium control case test cycle exposure 4

21

The U-bend specimens were sectioned after exposure and examined at various magnifications to determine if cracking or intergranular penetra-tion was present. A typical example of the metallographic investigation la shown in Fig. 8.

As can be seen, there is a significant phase formation near the Sanicro 31 surface from the dissolution of Sanicro 31 by the cast tooling plate, but there is no cracking or intergranular penetration present.

For all specimens tested, as summarized in Table 4, cracking or inter-granular penetration were never observed.

It is interesting to note that the literature indicates that in systems where mutual solubility is pre-sent, as we have here, liquid metal embrittlement is less probable.

Weldments were also tested in Sanicro 31 in the helium control case test cycle. This occurred when U-bend specimens were coated to the level of the weldment that held the restraint bar to the legs of the U-bend.

The Inconel 82 (ER Ni Cr-3) filler metal used is the same as used to join Sanicro 31 tubes in Module B.1.1.

Such a test is shown in Fig. 9.

No evidence of liquid metal embrittlement was found in any of the weld metal or heat-affected zones examined.

Typical specimens exposed to the metal control case test cycle are shown in Fig. 10.

This test cycle is actually more representative of the temperatures that are experienced in the actual steam generator. Also, it was intended to make sure that liquid metal embrittlement was not prevalent at lower temperatures compared to higher temperatures where dissolution rates were higher.

Figure 10(b) shows evidence of the flow of a low taulting point phase from the cast tooling plate material at the lower exposure temperature. A typical cross section after full exposure is shown in Fig. 11.

It was also postulated that the metal control case test cycle might be worse since the lower temperature exposure might favor a higher concentra-tion of potentially embrittling alloying elements at the surface of the 22

SANICR0 31 PHASE FORMATION CAST TOOLING PLATE COATING f

0.5 cra (a) MAGNIFIED VIEW

-~

?. r,n

.1 vry;, x,

(/,. ;f.

E>

f

-i' t,',..-

3 y

'.i#~

.-N

' ~'

i

.,S"trff f%.;,c :.- &

M 5'

t

d'

a. k).pfJ:fvib:' 'pf

/'

.p 's h ' CAST TOOLING L,,

s

.s f.'f Vq PLATE COATING j

ff:

5

- - -t 9s.

O. _-

.a

~c PHASE

,, (5 -

f,.~ , / --'

g.+

- ' ~ - -

~

' g y'Yfp,~,dv-

_ [~'

. g y,o FO RMATl0N cy 7.,s yr- - [-9 f., % N '-

,- c- ;g INTERFACE r

n

._ e y,.;;;3. ~;-- m. b., w-4 o

.~.,

t

• h t*t-4. ?.,' '. Y: ' W

. i,' ;

i Q :.

'^

4.,d

,r.

o' w>

,%. r.., y. 4. ',

' mi SANICR0 31 J

n~<

.nr

[,,

...:,y 'y g..,... %y S d. -

',~

!;\\', ~

1.

.' - n.- W.

=a 4 L e---?q[dJ. %

.. S t. s

, c,.m..

-f16

4 f e#.'. df ~a.;..r, G A.. b',;>ri,.~.1 2 % *$

w 1000 pn (b) VIEW OF TENSION SIDE IN CURVED PORTION Fig. 8.

Cross section of a typical U-bend specimen (S-1-10) after exposure to helium control case test cycle 23

m s

N 7

)

RESTRAINT 1

r s

~'

=

~

.a

/

BAR t^

r' ~Qq 4

s

, +. _

,i<...,

,3

c.,

~..J, o

• )

^ -

.

: ~.-'lC k

, ghs

^

+ +

.. 'l }

L. -

~

1.

o.

s 0.

c.. -- ?

NCONEL 82 WELD METAL rlf.Ly,;'

fQ p ',

c i., '

s.

's Q('y,

CAST TOO LING y,,.(

..b PLATE C0 ATING j.'

U-BEND LEG

=, M G..,/i,,,%,.

' W.

lf:;. s,.

V?; -

' ~

)

,..,..+~

c

'.,,,,f-I p' : ;< *4.. Lp.,

~ - (3 p:

s g;fj :1

n'

_'./;

U.d; f.ls!). m.

.1; -

a e

A,. ;. A,c lis:k,c.' u

^,tu-

s s

1000 pm Fig. 9.

Cross sect.lon of restraint bar/U-bend weldment af ter helium control case test cycle exposure (Specimen S-1-9).

Etched electrolytically with 10% oxalic acid I

24

-2

. - - _ ~ -. - -.,.

-n.-,.,-

n -

Mi d, f;

~ =:. n. sy. a mj o ;,r s t ri.y.l *

(S-1-21)

(S-1-22)

(S-1-17)

(S-1-19)

(a) AS-COATE0 AT 1180 F FOR 10 MINUTES ALLOY FLOW OCCURRING OURING 1065*F EXPOSURE l

.f.

(b) SPECIMEN S-1-17 (ABOVE) AFTER 1065 F EXPOSURE FOR 100 HOURS Fig. 10.

.ippearance of typical U-bends during and af ter metal control case test cycle exposure 25

INTERFACE CAST TOOLING SANICR0 31 PLATE C0ATING =

r R ' ';

, l' ;

9 l

1D

• Q.

.f

,s i t.,.

rd g

1 k, '. S',k PHASE ACCOUNTING FOR ALLOY FLOW OBSERVED IN FIG.10 E

a I

As

• ..,*ya r e*

,n.'

s :,

3 t

. 2 6.

Xi

['q.f,

(c D'-

Fig. 11.

Cross section of a typical U-bend after metal control case test cycle exposure (Specimen A21) 26

U-bend from a simple binary constitutional diagram argument. Again, no evidence of cracking or intergranular penetration was ever observed for any of the specimens tested and summarized in Table 4.

For the future test cycles given in Table 2, both for Superheater II (tested with a Sanicro 31 U-bend) and for Superheater I (tested with a 2-1/4 Cr - 1 Mo welded U-bend), no evidence of cracking or intergranular penetration was observed. However, for the Superheater II future test cycle case, some erosion was measured. This is discussed in Section 4.2.3.

4.2.3.

Erosion Measurements Specimens were examined for depth of erosion (high-temperature cor-rosion) attack by the cast tooling plate alloy. This was done to determine if excessive erosion could have been the cause of the leak, or be a cuase of future leaks. Most of the specimens listed in Table 4 were examined for erosion.

For the helium control case test cycles and the future Super-heater II test cycle, typical results are shown in Figs. 12 and 13, respectively.

For the tests conducted with the helium control case cycle, the results of erosion measurements are plotted in Fig. 14.

It is seen that erosion reaches a maximum of about 0.015 in, penetration after 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> at 1180*F and does not appear to increase beyond that depth for times up to 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> at 1180*F.

Evidently the phases that form in about 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> at the interface of the Saniero 31 and cast tooling plate retard further attack at this temperature.

For the future superheater test cycle, the depth n' sttack was some-what greater and was measured at 0.036 in. at one location in the U-bend specimen, as shown in Fig. 13.

In general, the depths of attack were lower than this on the rest of this specimen and another, similar specimen.

For the metal control case, the depth of erosion was significantly less than the helium control case. Typically, for the times shown in Table 4, the depth of erosion was always less than 0.002 in.

27

a

+..-.

9.

+

C0ATING j

0.010 lN.

.....-v,3.~. n

-x r:p ;c,,a.m$j.,s p.. ag '.

• 4 > p., -

. 3.y g +. %

+.

a

-5 i--

v p4 t. r.

s e

t

,)*Yy,%.?5'.[.,,t-' ? ' : %.. -, l~l.'

.-y%. &sA e-s.

s

.c....

y

'. %u % n a

. ~.

'r. n.,.Q. e.Y,,.h,.&l..N:~ h.$. e '. f.eM,,,,' -

e W'}rl;rM : X..Q *:

<w

+f

$4he&,f5

.'er?.',

!?

5ANICR0 31 c

. g -y.

• I <.3. /lv*Q*bi j.. p::?.v,4.ca.. ;Q.- i c 33g*' X.

g F, o U,- vs.s -. va. A.. A -

. v.

• k %

p

=

vs M.* */" +b IM.* --p. e* f, q l3

' M -g ( 'A

, e A.

• b.

+x *n..y A H*Q + v&...; ~h*,. g;g 1,~;M. a *

• W.1# h *4.e.i,c

e. <s v -

n, N.

p y

y.<

. q.

4',,G ls.y :*Lr,.at w.,

t. '.s.g. ;f.q

.:.e s.i V~ 4. k..s;.

m,

- = ;

' x.. n,g v y' '

74'..v.

.m

__-.a

- x.

C0ATING Fig. 12.

Erosion from helium control case test cycle exposure (Specimen A13) l-SANICR0 31

-l l' j :.,' k ? h. f f u f, l,,

(

w i j %+,Qy '. %..[: '+k,Y Ng I..

C0ATING

~

((h s.k..%. -

~5 4

~

l i. t T e,US-M' 's < '.iQ}s.

, o,M c > 4. 0 s. : 4

.. y 3

i'

?r -,;

• p e.;#

y r.

n -

.v m w

a;

. I.

^ " =: ---ORIGINAL SURFACE

((f[$&,R y w %~d[(;p

.39

hll * '

w h.

.{ 6 $.',.g

?,

% ww,.,y h {'f-

.Q y W.;:3q::'S ).: u.E d,1 0.036iN

,t p 5l g ;.:-

s-+ -q%

9.;.m $g,4. /r

-'i' y.;- %p +Q. s..g?~ f, y%-

i 3.

..., =.

,J

@ a.

',..r i.C',u N'....,

3;.a, ggS;f.,

t 5.

--*s e,

m. : <.... 3,::

-m:-

n 7. :.

v.~

.? A,.., 4 s-s s..

c. #. ps

. n

.ft: :". s e;,.

j WQ.~:f: t 1.,V. &13,.

.. gs

.a :

.y-

x. w:.~,..,

y.,+

Nh,d

• 3

{g

,i 7.(*7p3.:,iY.

f 4 k,.

. -CO ATING

'i iJ.g %

. xyp. y g

b

[

hk

-'l y:

c -

-t.

w w. :

w. A =.: n t_....

t

... _ ' " ::- w _- '

.m 1000 pm Fig. 13.

Erosion from future SH II test cycle exposure (Specimen S-1-29) 28

0.016

=,

0.014 0.012 1

0.010 5

5s E

$ 0 008 o

5 O

5 0.006

~

O AS RECEIVED SANICR0 31 O AGED SANICR0 31 0

(200-315 HOURS AT 1100 F)

A 20% PRESTRAINED SANICR0 31 0.002 d/

O a

I I

I I

0 O

2 4

6 8

10 0

TIME AT 118G F (HOURS)

Fig. 14.

Erosion of Sanicro 31 by aluminum alloy 29

For the future Superheater I test cycle on the 2-1/4 Cr - 1 Mo welded sample, the depth of erosion was also very low and typically was measured

~

at less than 0.002 in.

An example of this type of specimen is shown in the cross section in Fig. 15.

~

4.2.4.

Phase Analysis Scanning electron microscope energy dispersive X-ray analysis was used to determine the c(mposition of the phases formed at the interface between the cast tooling plate alloy and the Sanicro 31.

As seen in Fig. 16(b),

obtained by back scattered X-ray mapping, significant chromium is present in the phases that are formed at and near the surface of the Sanicro 31 for the helium control case test cycle.

Table 6 shows that the composition of the phases reflects elements present in both the cast tooling plate and the Sanicro 31.

This is evidence that the phases are formed as a result of e'

-'ing between the Sanicro 31 and the cast tooling plate.

It is interesting to note c very little zine is present in the phases that have formed. During testing, zinc evidently vaporized from the coating material. After this loss was noted, several of the experiments were analyzed for the presence of zinc in the cast tooling plate material that had been used for coating. Based on a random sample, it was found that the amount of zinc remaining varied from 0 to approximately 4%.

The explanation for this appeared to be whether or not sufficient oxygen was available in the test to form an oxide film on the surface of the melt.

If an oxide film had formed, zine was trapped and did not leave the melt. Thus, some experiments had more zinc present at termination than others, but in no case was there any evidence of cracking or grain b.ndary penetration.

It is likely that high vapor pressure elements, such as zinc, also vaporized from the aluminum alloy in Module B.1.1.

Reference 14, for example, describes the vaporization of alloying elements in a high pressure, flowing helium environment.

30

.j, -

,e.

-s

/f.'-

'i.g isM j Y.

~ ~

\\ R

g? +

g_

gg.

=

g....

.cw;,.

'^

~

T, ;~ "

~

y..

r.g -

.,T.

%j m

9 :.

v,.

C

\\

.r *

'g.

4 g

,7 4

~.

y A'if *,.

A,;.

s.-

/

V.

~ :

• r

cn%vi ;;i.+

t..,,

s.\\ *

~:

YIkh;I 1,

,f

• 4$s&;w[,h.,.!

'l$

7 s.

i.u:.cy

%;-A. ', ; b..,

at..'

s.: \\

~, ~

\\

.y N.

+-

k",

I q

\\

s

'$7

_ CAST TOOLING TYPE 521 PLATE COATING WELD METAL (2-1/4 Cr - 1 Mo)

INTERFACE Fig. 15.

Cross section of welded 2-1/4 Cr - 1 Mo U-bend specimen (CM-30) after exposure to future SH I test cycle. Etchant: nital 31

- MATRIX PHASE o

)

(

h PLATELET PHASE CAST TOOLING PLATE C0ATING l'. f

/

'j.

MMM

\\.

~

^

BORDER PHASE

.c

=

SANICR0 31 f

50 pm (a) REGION EXAMINED me

.,Y-

, ':y:

.~.

.;]i ',.

<: n y.g,

-l %

Axide.

. r;na c:. A$:YT

,,..<u 50 pm (b) BACKSCATTERED X-RAY MAP 0F CHROMlUM IN (a) AB0VE Fig. 16.

Scanning electron microscope energy dispersive X-ray analysis location on typical U-bend cross section exposed to helium control case test cycle (Specimen S-1-8) 32

TABLE 6 CHEMICAL COMPOSITION OF COATING PHASES, CAST TOOLING PLATE, AND SLNICR0 31 BY SCANNING ELECTR0; MICROSCOPE ENERGY DISPERSIVE X-RAY ANALYSIS FOR HELIUM CC iROL CASE TEST CYCLE (SPECIMEN S-1-8)

Location

• Al Mg Zn Cu Fe Si Cr Ni Matrix phase 91.79 4.28 0.00 0.95 0.24 0.39 0.23 0.12 Platlet phase 76.36 2.86 0.03 0.64 5.16 0.22 12.10 0.62 Border phase 67.85 3.16 0.00 0.46 13.63 1.50 6.56 4.83 Cast tooling 87.69 5.51 3.36 0.38 0.31 0.71 0.05 0.00 plate base material i

Sanicro 31 1.23 0.00 0.00 0.38 41.2' 1.94 20.44 32.81 base material

• See Fig. 16.

33

The Sanicro 31 near the coating interface was examined for the presence of cast tooling plate elements by scal 41ng electron microscope energy dispersive X-ray analysis. None were found, indicating no significant diffusion of the coating elements into the solid substrate h.d occurred.

For the metal control case test cycle, a typical specimen, shown in Fig. 17, was analyzed; th' results are given in Table 7.

Again, tha coating compositions show the presence of Sanicro 31 elements, indicating that alloying occurred between the cast tooling plat. and the Sanicro 31.

Compared to the helium control case anlysis, the border phase compositions in the two coatings are similar. The matrix phase compositions are sig-nificantly different.

The platelet phase is absent from the coating produced by the metal control case cycle. A low liquidus temperature of the matrix phase composition given in Table 7 accounts for the alloy flow observed in Fig. 10(b) after the 1065'F exposure. Again, no zine was present in the coating after the test cycle. No significant diffusion of the cast tooling plate elements into the solid Sanicro 31 occurred, as Table 7 shows.

4.2.5.

Stress-Rupture Tests Stress-rupture test results are given in Table 8 and Fig. 18.

The uncoated, solution-annealed Alloy 800 base material stress-rupture scatter band is shown in Fig. 18 for comparison to the coated specimen test results.

As can be seen, the coated, solution-annealed Alloy 800 stress-rupture test failure times compare favorably with the scatte'. hand, and either exceed or fall close to it.

This was true for all three material condi-tions coated and tested.

It can be concluded from this that the presence of the cast tooling plate alloy did not affect the rupture life of the base material.

34

INTERFACE CAST TOO LING SANICR0 31 c

= PLATE COATING

.R.

'L C

w&~{,^.'",

~

• • y

', q;

?

y.*t

~~

gc

. ' +

a.:.

yy

• r y

r~

l

~K y...

e SYaNY,. *

c' -l.

• ?s,.

.t s

% ' ?,.

?

^ }IY%.,jW e.#

. ' dT %

.7 * <

~

4

. MYl+ O %.'

hk$C.. f. U.'.

h*

. (

. g af 7

.t

^

&f. >.f6 ^ _l-k.~

l 1

50 pm M10D LE OF EDGE OF BORDER PHASE MATRIX PHASE SANICR0 31 SANICR0 31 Fig. 17.

Region examined by scanning electron microscope energy dispersive X-ray analysis on cross section c Fpecimen A-20-1 exposed to metal control case test cycle 35

TABLE 7 CHEMICAL COMPOSITION OF COATING PHASES AND SOLUTION-ANNEALED ALLOY 800 BY SCANNING ELECTRON MICROSCOPE ENERGY DISPERSIVE X-RAY ANALYSIS FOR METAL CONTROL CASE TEST CYCLE (SPECIMEN A20-1)(a)

Location Al Mg Zn Cu Fe Si Cr Ni Matrix phase 63.05 4.08 0.00 0.10 12.37 5.33 5.93 7.15 Border phase 57.04 4.12 0.00 0.00 16.29 1.95 6.92 11.68 Edge of 2.64 0.04 0.00 0.27 41.99 1.43 19.46 32.17 Sanicro 31 Middle of 2.39 0.03 0.00 0.13 42.29 1.57 19.58 32.01 Sanicro 31

• Specimen was coated in cast tooling plate at 1180*F for 10 min and stress-rupture tested at 1050*F for 307 hr.

36

TABLE 8 STRESS-RUPTURE TEST RESULTS OF SOLUTION-ANNEALED ALLOY 800 AT 1050*F AFTER COATING WITH CAST TOOLING PLATE AT 1180*F FOR 10 MINUTES Tension Time to Heat Prestrain Aging Stress Rupture Specimen No.

(%)

Treatment (ksi)

(hr)

Comments A20-1 6300 0

None 45 307 Grinding undercut on gage section caused premature failure A20-4 6300 0

None 45 2301 Discontinued B20-4 6295 0

None 50 164 C20-1 6624 0

None 55 71 C20-2 6624 20 None 55 455 Discontinued C20-4 6624 20 None 60 747 Discontinued B20-2 6295 20 200 hr 60 201 at 1100*F B2 0-3 6295 20 200 hr 60 308 Discontinued at 1100*F 37

100 90 80 70 60 Z

A 0+

v D-50 O

5 40

=

uy 30 BASE MATERIAL CONDITION BEFORE COATING:

g SCATTER BAND FOR AS RECElVED t;

O AS RECEIVED SOLUTION ANNEALED ALLOY 800 N

~

O 20% PRESTRAIN IN TENSION BASE MATERIAL (N0 COATING)

A 20% PRESTRAIN IN TENSION PLUS AGED 200 HOURS AT 1100 F

--- TEST DISCONTINUED 10 I

I l

0 100 1000 5000 TIME (HOURS)

Fig. 18.

Stress-rupture tests on solution-annealed Alloy 800 at 1050'F in helium after coating with cast tooling plate aluminum alloy

Several of the fractured specimens were examined by scanning electron microscope and light microscopy. A typical fracture surface examined by scanning electron microscope is shcwn in Fig. 19.

The appearance of the f ractured, solution-annealed Alloy 800, which in this case had been pre-strained 20% and aged for 200 hours0.00231 days <br />0.0556 hours <br />3.306878e-4 weeks <br />7.61e-5 months <br /> at 1100*F. is quite normal as compared to previous examinations of similar uncoated material. The fracture is largely intergranul

, with some tearing evidenced by the fibrous character of the fracture fa ats.

In fact, the fibrous feature is an indication of the fairly good rupture ductility of this particular heat for this test temperature. No indication that the cast tooling plate coating influenced the fracture was found.

A cross section of the same specimen fracture surface is shown in Fig. 20.

Again, examination of the fracture surface near the interface between the cast tooling plate coating and the solution-annealed Alloy 800 by light microscopy shows no indication that the coating had an influence on the fracture. This is a very typical fracture surface cross section as compared to previous examinations of uncoated, solution-annealed Alloy 800 fracture surfaces tested in this material condition near this temperature.

The cracks that are apparent in the coating in Fig. 20 probably occurred on cooling. Near the test temperature cast tooling plate alloy flow was observed in U-bend tests (1065*F), which indicates the coating was probably liquid or semiliquid during the stress-rupture test at 1050*F.

4 39

SANICR0 31 CAST TOOLING PLATE COATING s *. -3

~73 ih.%j

$y,.,.D

• .N

._..,3.gs; +

...w M

200 pm (a) MAGNIFIED VIEW

_ INTERFACE i

as

2. -

't l'N'[

'k.,

CAST TOOLING

• 't

'i. A '

• ) PLATE COATING J4

~.i)W!

.)

SANICR0 31 b

(b) VIEW AT INTERFACE Fig. 19.

Photographs of typical strea -rupture fracture surface (Specimen B-20-2) taken with a scanning electron microscope 40

SANICR0 31 w.-

p-

.-." c," -

-g; t..,

-. :1

/

k.c

.g

~s

, h'

')

2

.s 44 J:h

~

4 H..

~

3 y

Mt i

m

-4 k37

~,

N-

,4.

p N'[

CAST T00 LING e

j.

^

,/.

~

v.

,,'r..

y PLATE COATING

g.. ' Ji

-?

^

I h,,

j, ' y~.? ;

[f '

e

?lf.,,.se,,* N;~y.

T e

^ A ',

e

., ' 3.;

e.

(

u

.r.,

k f

t ; s, e

I f'

E r.

3 C

t;.

j.

3; z,

i.-

t

, Vi,.

=.

l mq q

qp~'

r t,..s 1

; '.s

.4 N.;.. ~ % S A'

l t

1000 pm 20 pm (a) MAGNIFIED VIEW (b) HIGHER MAGNIFICATION OF (a)

Fig. 20.

Cross section through typica' stress-rupture fracture surface (Specimen B-20-2)

J 5.

DISCUSSION Even with the conservative factors introduced in this test program in both the test cycles used and the quantity of cast tooling plate alloy used, no cracking or grain boundary penetration type liquid metal embrittle-ment was observed. This was true for both the.canicro 31/ Alloy 800 and the 2-1/4 Cr - 1 Mo materials tested. The erosion depths measured with the test cycles representing the operating history of Module B.I.1 up to the time of the leak were small.

It is not likely that these erosion depths could account for the leak that occurred. Erosion depths were higher, at least in one instance, for the future predicted Module B.1.1 service.

Even with this depth, however, it is unlikely that a leak will be caused because of the conservative factors that were introduced in the test.

It should, of course, be pointed out that this test program and its results are predicated on the temperatures and times described. For example, it was observed from this program that erosion depth is very sensitive to metal temperature.

Small increases in temperature above the test temperatures cited here caused marked increases in erosion depth in relatively short times. But even in these higher temperature experiments, liquid metal embrittlement of the cracking or intergranular penetration type was never observed.

4 42

6.

CONCLUSIONS It may be concluded that for the test temperatures, times, cast tooling plate alloy, and base materials evaluated in this program:

1.

Cracking or intergranular penetration type liquid metal embrittle-ment is highly unlikely.

2.

The erosion depths measured indicate that it is unlikely that the tube leakage in Module B.1.1 resulted from the presence of the aluminum chip.

e e

43

7. REFERENCES 1.

Radeker, W., and Muhlhelm-Ruhr, "Intercrystalline Attack on Steel by Molten Metal," Werkstoffe und Korrosion 21, No. 10, October 1973, pp. 851-859 (in German).

2.

Horstmann, D., and F-K. Peters, "The Attack of Iron-Saturated Zinc Metals on Iron in the Temperature Range 540-740*C," Arch.

Eisenhuttenwesen 40, (8), August 1969, pp. 621-626 (in German).

3.

Radeker, W., and Mulheim a.d. Ruhr, " Die Erzeugung von Spannungsrissen in Stahl durch flussiges Zink," Stahl u.

Eisen 73, 1952, pp. 654-658 (in German).

4.

Smith, W. R., and P. E. J. Forsyth, " Grain Boundary Penetrations by Liquid Metals, I.-Some Service Failures," Metallurgia, August 1946, pp. 186-188.

5.

Stoloff, N. S., " Liquid Metal Embrittlement," Surf aces and Interfaces, Vol. II, Syracuse University Press, Syracuse, New York, 1968, pp. 157-182.

6.

Dinda, S., and W. R. Warke, "The Effect of Grain Boundary Segregation on Liquid Metal Induced Embrittlement of Steel," Materials Science and Engineering 24, 1976, pp. 199-208.

7.

Westwood, A. R.

C., and M. H. Kamdar, "Concerning Liquid Metal Embrittlement, Particularly by Zine Monocrystals by Mercury," Phil.

Mag.J3, 1963, pp. 787-804.

8.

Stoloff, N.

S., and T. L. Johnston, " Crack Propagation in a Liquid Metal Environment," ACTA Metallurgica 11,, April 1963, pp. 251-261.

9.

Eborall, R., and P. Gregory, "The Mechanism of Embrittlement by a Liquid Phase," Journal of the Institute of Metals 8JL, 1955-56, pp. 88-90.

10.

van Ewijk, L. J. G., "The Penetration of Steel by Sof t Solder and Other Molten Metals at Temperatures up to 400*C," J. Inst. Metals, Vol. LVI, 1935, pp. 241-248.

44

11.

Stoloff, N.

S., "Recent Developments in Liquid Metal Embrittlement,"

presented at Symposium on Environment Sensitive Fracture of Engineer-ing Materials, at the Metallurgical Society (TMS-AIME) Annual Fall Meeting, Chicago, Illinois, 24-27 October 1977.

12.

Austen, G. Wesley, "The Effect of Molten Solder on Some Stressed Materials," J. Inst. Metals Sp_, 1936, pp. 173-185.

13.

Lampe, V.,

H. Roos, and M. Svensson, "The Attack of Molten Zinc on Steels," Werkstoffe and Korrosion 28, 1977, pp. 226-232.

14.

Charlot, L.

A., R. A. Thide, and R. E. Westerman, " Corrosion of Superalloys and Refractory Metals in High Temperature Flowing Helium," Battelle-Northwest Laboratory report BNWL-SA-1137, March 1967.

15.

Elliott, D., "Zine Embrittlement of St ainless Steel - the Problem in Perspective," Process Engineering, July 1976, pp. 67-71.

16.

Genders, R., "The Penetration of Mild Steel by Brazing Solder and Other Metals," J. Ins t. Metals jl7, 1927, pp. 215-240.

17.

Shunk, F.

A., and W. R. Warke, " Specificity as an Aspect of Liquid Metal Embrittlement," Scripta Metallurgica 8_, 1974, pp. 519-526.

18.

Preece, C. M. " Liquid Metal Embrittlement of Iron Alloys," presented at NACE Conference for Stress Corrosion Cracking and Hydrogen Embrittlement of Iron Base Alloys, Unieux-Firminy, France, June 12-16, 1963, preprint number E-2.

19.

Rostoker, W., J. M. McCaughey, and H. Markus, Embrittlement by Liquid Metals, Reinhold Publishing Corporation, New York,1960.

20.

Berry, Warren E., Corrosion in Nuclear Applications, Chapter IV -

" Corrosion by Liquid Metals," John Wiley and Sons, Inc., 1971.

21.

Alcoa Aluminum Handbook, Aluminum Company of America, Pittsburgh, Pa.,

1967.

s 45

ACKNOWLEDGEMENTS The author wishes to express his appreciation to R. F. Stetson and H. D. LeRoy for their work in conducting experiments. The efforts of R. E. Villagrana and D. R. Wall for the scanning electron microscope work and interpretation, and to J. E. Knipping and R. H. Speas for the metallography work are acknowledged. The assistance of L. D. Thnapson and C. C. Li in conducting the literature review was a valuable contribution.

1 a

46