ORNL-TM-4272.txt

 

ORNL-TM-4272

 

COMPATIBILITY OF BRAZING ALLOYS AND
THE MOLTEN SALT NaBF,—8 MOLE PERCENT
NaF AT 610°C

J. W. Koger

 

 

 

oRcHne

N RN R

 
 

 

 

 

This report was prepared as an account of work sponsored by the United
States Government. Neither the United States nor the United States Atomic
Energy Commission, nor any of their employees, nor any of their contractors,
subcontractors, or their employees, makes any warranty, express or implied, or
assumes any legal liability or responsibility for the accuracy, completeness or
usefulness of any information, apparatus, product or process disclosed, or
represents that its use would not infringe privately owned rights.

 

 

 

-

-

")
ORNL-TM-4272
€
%
| Contract No. W-7405-eng-26
f METALS AND CERAMICS DIVISION
!
COMPATIBILITY OF BRAZING ALLOYS AND THE MOLTEN SALT
NaBF ,—8 MOLE PERCENT NaF AT 610°C
J. W. Koger
-
9
December 1972

 

| NOTICE :
This report was prepared as an account of work

sponsored by the United States Government, Neither
. the United States nor the United States Atomic Energy
‘Commission, nor any of their employees, nor any of
their contractors, subcontractors, or their employees,
makes any warranty, express or implied, or assumes any
legal liability or responsibility for the accuracy, com-
pleteness or usefulness of any information, apparatus,
product or process disclosed, or represents that its use
would not infringe privately owned rights.

 

 

 

 

Ll

OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37830
operated by
UNION CARBIDE CORPORATION

y for the

U.S. ATOMIC ENERGY COMMISSION

«)

 

DISTRIBUTION OF THIS DOCUMENT IS UNLIMIT&?

 
 

i

 

(a
 

i

QO

a)

9

 

CONTENTS
Abstract . ... ... S 1.
L IntrOdUC I ON .ottt ittt ittt e et 1
2. PreviOUS WOTK ... i i i e e e et e 2
3. Experimental Procedure . ........ .. ... .t e i i 5
4. Results ................ e e e ettt et e 7
4.1 Braze Alloy BNi-7 (Ni—13% Cr—10% P) . ...ttt e ene 10
4.2 Braze Alloy BNi-3 (Ni—3.5% Fe—4.5%S8i-29%B) ......ccouire i, 10
4.3 Braze Aloy BAu4 (Au—18 N1} . ... ..ot i e e 10
4.4 Braze Alloy BAg-8 (Ag—28 CU) . ..o vttt ittt et it et e i i 10
4.5 Braze Alloy BNi-4 (Ni—4.5% Si—3%B) ... .o i it it e i 10
4.6 Braze Aloy BCu(100% Cu) . .....coii ittt ittt e 26
4.7 Braze Alloy BNi-2 (Ni-7.0% Cr—3.5% Fe—4.5%Si-2.9%B) . . . ... ..., 26
4.8 Braze Alloy BNi-2 (Ni—6.5% Cr--2.5% Fe—4.5%Si—-3%B) ....... .. ..o iiiiiiinn.. 26
T B 1703 1T (e ¥+ O 26
6. CONCIUSIONS & o ittt ittt ittt ene et aeeee e toensssoarecasssnsnoanssnensnenasenananas 39

iii

 
 

 

 
 

i

<

“

=)

n

a)

COMPATIBILITY OF BRAZING ALLOYS AND THE MOLTEN SALT
NaBF, —8 MOLE PERCENT NaF AT 610°C

J. W. Koger

ABSTRACT

A comprehensive review of the compatibility of braze alloys with molten fluoride salts was
performed. Eight different braze alloys, used to braze Hastelloy N, were exposed to NaBF4—8 mole %
NaF at 610°C for 4987 hr. On the basis of weight changes, microstructural changes, and electron
microprobe analysis all alloys tested were compatible. The Ag—-28% Cu and 100% Cu braze alloys were
considered the most resistant. The prediction of corrosion resistance on the basis of free energy of
formation data was in reasonable agreement with our results. Nickel transferred to non-
nickel-containing braze alloys through an activity gradient mass transfer mechanism. Some deposits
were noted on the Hastelloy N base material. Significant interdiffusion of the Hastelloy N and the
braze alloy occurred only with those braze alloys whose composition was near that of Hastelloy N.

1. INTRODUCT ION

The most recent designs for the molten-salt breeder reactor (MSBR) call for two salts: one containing
both fertile and fissile material (ThF, and UF,;) and one that will transfer heat from this fuel salt to a
steam generator. The tentative choice for this latter salt is the eutectic sodium fluoroborate—sodium
fluoride mixture NaBF4 —8 mole % NaF, chosen because of its low melting point, low cost, and acceptable
heat-transfer properties. Until recently, little was known about the compatibility of this salt with alloys
considered for use in a molten salt reactor system. The nicke!l-based Hastelloy N alloy has been tentatively
chosen for use in molten fluoride salts, and many tests have been conducted on its compatibility with the
sodium fluoroborate mixture ! 1! _ | |

Welded, and back-brazed tube-to-tube-sheet joints are normally used in the fabrication of heat
exchangers for molten-salt service. The back-brazing operation serves to remove the notch inherent in
conventional tube-to-tube-sheet joints, and the braze material minimizes the possibility of leakage through a
weld failure that might be created by thermal stresses in service. Thus the braze material must be resistant
to corrosion caused by the molten salt. Also, as larger engineering loops are built, other uses for brazes that

will contact the sodium fluoroborate mixture will surely be found. However, no tests have been conducted

to establish the compatibility of braze alloys (specifically those that could be used with Hastelloy N) with
NaBF,—8 mole % NaF. This report details an experiment conducted to determine the compatibility of
various Hastelloy N braze alloys with the sodium fluoroborate mlxture at 610 C, the maximum

 

w Koger and A. P. Litman, MSR Program Semiannu. Progr. Rep. Feb. 29, 1968, ORNL-4254, pp 218-25.
W. Koger and A, P. Litman, MSR Program Semiannu. Progr. Rep. Aug. 31, 1968, ORNL-4344, pp. 257—66.
W. Koger and A, P, Litman, MSR Program Semiannu. Progr. Rep. Feb. 28, 1969, ORNL-4396, pp. 243-53.
W. Koger and A. P. Litman, MSR Program Semiannu. Progr. Rep. Aug. 31, 1969, ORNL-4449, pp. 195 208.
W. Koger, MSR Program Semiannu. Progr. Rep. Feb. 28, 1970, ORNL-4548, pp. 240-52.
W. Koger, MSR Program Semiannu. Progr. Rep. Aug. 31, 1970, ORNL4622, pp. 165-78.
W. Koger, MSR Program Semiannu. Progr. Rep. Feb. 28, 1971, ORNL4676, pp. 192-215.
. J. W. Koger and A. P. Litman, Compatibility of Hastelloy N and Croloy 9M with NaBF 4-NaF. KBF4 (90-4-6 mole
%) FluoroborateSaIt ORNL-TM-2490 (April 1969). .
9. H. E. McCoy, R. L. Beatty, W. H. Cook, R. E. Gehlbach, C. R. Kennedy, J. W. Koger,A P. Litman, C. E. Sesswns,

]
}
J
1.
]
5.
..

BN R W=

-and J. R. Weir, “New Developments in Materials for Molten Salt Reactors,” Nucl. Appl. Technol., 8, 156 (1970).

10. J. W. Koger and A. P. Litman, Compatibility of Fused Sodium Fluoroborates and BF3 Gas thh Hastelloy N
Alloys, ORNL-TM-2978 (June 1970).

11. J. W. Koger and A. P. Litman, Mass Transfer Between Hastelloy N and Haynes Alloy No. 25 in a Molten Sodium
Fluoroborate Mix ture, ORNL-TM-3488 (October 1971).

 
 

Table 1. Braze alloys, brazing temperature, and composition

 

 

 

N AWS-ASTM Brazing Composition (wt %)
0. . . temperature ,
classification co - Cr Fe  Ni P C Si B Au Cu Ag
1 BNi-7 1010 13 7685 10  0.15
2 BNi-3 1040 35  89.1 4.5 . 29
3 BAu<4 1010 18 : 82
4 BAg8 816 ' - 28 72
5 BNi4 1040 : 92.35 015 45 3 .
6 BCu 1125 ‘ 100
7 BNi-2° 1030 7 3.5 821 45 29
8

BNi-2* 1040 65 25 8335 015 45 3

 

 

%Two different manufacturers.

temperature proposed for the salt use. Eight fairly typical brazes were selected for this experiment. The
braze alloys, their composition, and the temperature of brazing are listed in Table 1. '

2. PREVIOUS WORK

‘Many experiments have been previously conducted to determine which braze alloys (suitable for use
with nickel alloys) would be acceptable in a high-temperature molten-salt environment containing UF,.
Some of these results will be reviewed in order to compare like-brazing alloys in different fluoride salt
experiments, keeping in mind temperature, time, and test differences.

Brazing alloys of composition 91.25% Ni—4.5% Si—2.9% B (quite similar to the BNi-4 we tested) and
93.25% Ni—3.5% Si—1.9% B in button form were subjected to 100-hr corrosion tests in the fuel mixture
NaF-40.0 mole % ZiF,—6.5 mole % UF, in a seesaw apparatus (a rocking furnace combined with a
temperature gradient, described by Vreeland et al.!?) at a hot-zone temperature of 816°C.'3 Minor
constituents of the alloys were leached to a depth of 1 to 2 mils, as shown by metallography, and weight
losses of the alloy specimens varied from 0.03 to 0.06%. |

The brazing alloys Au—18% Ni (which we tested, BAu-4) and Au—20% Cu were corrosion tested in
LiF—41.0 mole % KF—11.2 mole % NaF—2.5 mole % UF, and LiF—37.0 mole % BeF, —-1.0 mole % UF,
for 2000 hr at 650°C under static conditions.'* The alloys were also tested in the latter fuel for 500 hrin a
seesaw apparatus with a hot-zone temperature of 650°C. No attack was observed on either alloy in any of
the tests. A layer high in nickel was found on the Au—18% Ni specimen after tests in both salts.

Four silver-base brazing materials, ranging in composition from pure silver to alloys with 42 wt % silver,
and a gold-base alloy (Au—20% Cu-—5% Ag) were tested in static LIF-37.0 mole % BeF,; —1.0 mole % UF,
at 700°C for 500 hr. 15 As shown in Table 2, subsurface voids, as deep as 50 mils in the case of pure silver,

. were observed in all the sllver-contammg alloys. The gold-based alloy showed no corrosion.

In order to obtain long-term dynamic corrosion data on brazing alloys in LiF—37.0 mole % BeF,—1 |

mole % UF,, a series of five alloys were tested in duplicate by inserting brazed lap joints (nickel-base alloys,
similar configuration to that which we used) in the hot legs of thermal-convection loops.'® (We tested all

 

. C. Vreeland, E. E. Hoffman, and W. D. Manly, Nucleonics 11(11), 3639 (1953).

12. D.C
13. D. H. Jansen, ANP Quart. Progr. Rep. Dec. 31, 1957, ORNL-2440, p. 169.
14. D. H. Jansen, MSR Quart. Progr. Rep. Jan. 31, 1958, ORNL-2474, p. 59.
--15. D. H. Jansen, MSR Quart. Progr. Rep. Oct. 31, 1958, ORNL-2626, p. 64.
16. E. E. Hoffman and D. H. Jansen, MSR Quart. Progr. Rep. June 30, 1958, ORNL-2551, p. 62.
 

Table 2. Results of corrosion tests of silver- and gold-base brazing alloys exposed

to static LiF--37.0 molg % BeF; ~1.0 mole % UF, for 500 hr at 700°C

 

Braze materiat

Metallogriphic results

 

 

 

 

 

 

 

5 Pure silver Spotty, heavy attack; stringers to a depth of 50 mils in some places
Ag—10% Cu ‘Rather uniform, heavy attack to a depth of 16 mils
Ag—33.3% Au—16.7% Cu Spotty attack to a maximum depth of 5 mils
Ag—40% Au—18% Cu—0.6% Zn  Uniform attack to a depth of 15 mils
Au—20% Cu-5% Ag No attack -’
Table 3. Results of metallographic examinations of brazing materials tested in thermal-convection loop
- circulating LiF—37 mole % BeF,; —1 mole % UF,
Test conditions: hotleg temperature, 700°C; cold-leg temperature, 593°C
Alloy \ s . _
AWS-ASTM . . Time Metallographic results: alloy brazed to
- . composition 3
classification (%) (hr) Inconel Hastelloy N
BNi-3 89 Ni-5 Si—-4 B-2 Fe 5,000 No attack No attack
10,000 Diffusion voids to 3 mils No attack
below surface
BNi-2 81 Ni—-8 Cr—4 B—4 Si-3 Fe 5,000 No attack No attack
10,000 Severe porosity 15 mils No attack
deep
¥ BNi-§ 70 Ni-20 Cr—-10 Si 5,000 Heavy attack of 16 mils 3 mil attack
10,000 Complete attack; severe No attack
porosity through fillet :
T BAu4 82 Au-18 Ni 5000  Noattack No attack
10,000 No attack No attack
BCu 100 Cu 5,000 No attack; diffusion No attack; diffusion voids
' . o . voids 2 mils deep on 2 mils deep
A _ fillet ,
5 10,000 No attack; small dlffusmn No attack
: voids present in fillet
but the BNi-5 alloy) A series of three identical loops were operated for 1,000, 5 ,000, and 10,000 hr with
the temperature of the circulating salt in the region of the test specimens at 700°C. '
Metallographic ‘examination after the 1000-hr test showed that all the brazmg alloys had good
flowability on both Inconel and Hastelloy N.17 There was a tendency for the formation of diffusion voids
in the fillets of the joints brazed with the gold-nickel alloy. The BNi-5 alloy was heavily attacked on
Inconel. The BNi-3 and BNi-2 were depleted at the fillet surface to a depth of 1 to 2 mils. Some slight
- crackmg of the alloys occurred at the brazmg-alloy-—base-metal mterface Pure copper showed good
corrosion resistance and no cracking. ' '
The results of the 5000- and 10,000-hr tests are outlined in Table 3 (refs. 18 and 19). Copper,
% gold-nickel, and the BNi-3 and BNi-2 alloys showed good corrosion resistance after 5000 hr. A depleted
. 17. MSR Quart. Progr. Rep. April 30, 1959, ORNL-2723, pp. 60—61.
o’ 18. MSR Quart. Progr. Rep. Oct. 31, 1959, ORNL-2890, pp. 46-47.

19. MSR Quart. Progr. Rep. Jan. 31 and April 30, 1960, ORNL-2973, pp. 62—63.

 
 

 

Table 4. Results of static corrosion tests on
refractory-metal-base brazing alloys in
LiF-37 mole % BeF,; —1 mole % UF,

in nickel containers

Test conditions: time, 100 hr; temperature, 700°C

 

 

Alloy Acomposition (%) Weight change (%)
48 Ti—48 Zr—4 Be _42
95 Ti—5 Be -6.5
95 Ti-$ Be | -9.8

 

Table 5. Results of static corrosion tests on refractory metals in
- LiF-37 mole % BeF; —1 mole % UF, in containers of
several materials

Test conditions: time, 100 hr; temperature, 700°C

 

 

 

Material ' Weight change (%) when tested in —
Nickel Hastelloy N Titanium Zirconium
Ti -12.1. ' -7.6 -0.71
Zr ~11.7 —4.2 - —0.08
Be Excessive, portion
of sample dissolved

 

region to a depth of 3 mils was observed along the BNi-3 and BNi-2 alioy fiilets after the test. Results of
earlier corrosion tests on these alloys in NaF-ZrF,-base fuels and in liquid metals indicated that this
depletion was due to the leaching of the minor constituents boron and silicon from the alloy by the bath,
and it appears that boron and silicon are also being leached by the salt. This depletion has no detrimental
effect on the alloy, since a nickel-rich, corrosion-resistant matrix is left.

After 10,000 hr all the brazed joints showed good corrosion resistance when used to join Hastelloy N.
However, Inconel joints brazed with BNi-3, BNi-2, and BNi-5 were attacked. Thus the results from the
1,000, 5,000-, and 10,000-hr corrosion tests indicate that several brazing materials suitable for joining
Hastelloy N or Inconel have adequate corrosion resistance to the type of fluoride salt represented by
LiF—37.0 mole % BeF; —1.0 mole % UF,. '

Some refractory-metal-base alloys being developed by the Welding and Brazing Group at ORNL for -

possible application in joining graphite to graphite were given a static corrosion test in LiF-37.0 mole %
BeF;—1.0 mole % UF, for 100 hr at 700°C (r_ef. 20). Because of its comparative inertness to fluorides,
nickel was used as the test-container material. Test results listed in Table 4 showed large weight losses for

each alloy. In order to determine the alloying element or elements responsible for the heavy attack, the.

following static corrosion tests in LiF—37.0 mole % BeF,—1.0 mole % UF, were conducted at 700°C: (1)
titanium was tested in containers of titanium, nickel, and Hastelloy N; (2) zirconium was tested in

~ containers of zirconium,- nickel, and Hastelloy N; and (3) beryllium was tested in a nickel container. Large

weight losses (Table 5) were observed for specimens tested in nickel or nickel-base containers, while small

‘weight losses were found when specimen and container were of the same metal. Substituting a Hastelloy N

(containing 70% Ni) container for one of pure nickel resulted in a weight-loss decrease on the titanium and

 

20. MSR Quart. Progr. Rep. Oct. 31, 1959, ORNL-2890, pp. 46—47.
0

&)

9

)

)

R

 

Table 6. Results of fuel corrosion tests on newly developed brazing alloys
Test conditions:  time, 100 hr; temperature, 700°C

 

 

Fuel  goioht
Alloy composition Test container Type ?f test used change Metallographic results
(wt %) specimen as _
" bath® (mg)
60 Au-30 Ta—10 Ni Nickel capsule Alloy button 130 ;3' Slight, scattered subsurface
. _ voids to 1 mil in depth
62 Au-26 Ta—12 Ni Nickel capsule Graphite-Mo .. 130 +36 Slight, scattered subsurface
: o _T-joint . voids to 2 mils in depth
58 Au—27 Ni-8 Ta— . Inconel capsule Alloy button 30 0.9 No attack observed
3 Mo-2Ct-2Fe ' L :
63 Au—29 Ni—-3.5 Mo- Inconel caps_ule/ Alloy button 30 NA? No attack observed

2.5Cr-2Fe

 

9130: LiF—37.0 mole % BeF,—1.0 mole % UF4;
30: NaF-46.0 mole % ZrF4—4.0 mole % UF,.

bNot available.

zirconium specimens. Metallographic examination and spectrographic results on the nickel capsules revealed
surface layers containing titanium, zirconium, or beryllium, corresponding to the major component of the
brazing alloy or metal tested. Results of the tests indicate that the usefulness of titanium-, zirconium-, or
berylhum-contalmng brazing alloys in LiF—37.0 mole % BeF,—1.0 mole % UF, would be limited in
systems constructed of nickel or nickel-base alloys. Work was also done by the Welding and Brazing Group
at ORNL on gold-base alloys containing various amounts of tantalum, nickel, and other minor constituents
to provide proper flowability and melting point.?! The results of corrosion tests on these alloys in different
fuel salts are listed in Table 6. A moderate concentration of tantalum was detected spectrographically on
the inside walls of the nickel container from the Au—30% Ta—10% Ni test. This dissimilar-metal mass
transfer was not intense enough to be detected on the nickel capsule by metallographic examination. The
relatively large weight gain on the graphite-to-molybdenum T-joint was due to pickup of fuel by the
graphite. The two alloys listed last in Table 6 showed good resxstance to NaF—46 0 mole % ZrF4—4.0 mole
% UF,.

Tables 7 and 8 summarize the data obtained from numerous other braze-alloy—salt compatibility
tests.2? Examination of these data indicates that BCu, BAu-4, 60Pd-40Ni, BNi-3, BNi-1, and BNi-7 are
satisfactory for molten-salt service.

3. EXPERIMENTAL PROCEDURE

Sixteen brazed specimen.s (two with each braze alloy) were prepared in the lap joint configuration
shown in Fig. 1. The two joined pieces of base alloy were Hfiételloy N, nominal composition 16% Mo, 7%
Cr, 5% Fe, bal Ni. The test was conducted on eight specimens (one specimen of each braze) in a nickel pot
at 610°C for 4987 hr. At various times the specimens were removed for weight-change measurements, and
salt samples were taken for analysis of impufities. The compatibility was evaluated through weight-change
measurements, salt analyses, metallographic observations, and micgdprobe analysis.

 

21. MSR Quart. Progr. Rep. Jan. 31 and April 30, 1960, ORNL-2973, pp. 62—64.
22. H. G. MacPherson, “Molten Salt Reactors,” pp. 815-16 in Reactor Handbook, Second Edition, Volume 1V
Engineering, ed. by S. McLain and J. H. Mortens, Interscience, New York, 1964.

 
 

 

Table 7. Brazing alloys on Inconel T-joints seesaw tested in fluoride mixture
NaF—40.0 mole % Z1F4—6.5 mole % UF, for 100 hr at a hot-zone temperature of 816°C

 

AWS-ASTM classification and Weight changeb
braze alloy —_— Metallographic notes .

 

composition? Grams  Percent
BCu, 100 Cu —0.0002 -0.026 0.5 mil surface attack along fiflet
BNi-3 o -0.0008 —0.052 0.5 mil nonuniform surface attack along fillet
BNi-2 ‘ ‘ -0.0008 —0.063 0.5 mil nonuniform surface attack along fillet
BNi4 -0.0014 --0.085 0.5 mil uniform surface attack along fillet

70 Ni-13 Ge—11Cr-68Si  -0.0011 -0.067 Nonuniform attack of 1.5 mils along surface of braze fillet
, - -0.0009 -0.092 1.5 mil uniform surface attack along braze fillet
BNi-1 ~0.0005 -0.030 1.5 mil erratic surface attack along braze fillet

BNi-2 ‘ -0.0011 -0.092 1.5 mil nonuniform attack along surface of braze fillet
BNi-5 -0.0008 —0.067 3.5 mil attack along surface of braze fillet
65 Ni~-25 Ge-10Cr ' -0.0019 —-0.056 Stringer-type attack to a maximum depth of 4 mils, few localized areas

 

“Brazing alloys listed in order of decreasing corrosion resistance to the fluoride salt.
bWeight-change data for brazing alloy and base material.

Table 8. Results of static tests of brazing alloys on nickel T-Joints in fluoride mixture

 

NaF-40 mole % Z1F 4—6.5 mole % UF, at 816°C for 100 hr

 

 

 

AWS-ASTM classification Weight change” Metallographic notes
and braze alloy™ composition Grams Percent _

BAu4, 82 Au—18 Ni -0.0010 -0.036 Braze fillet unattacked

60 Pd—40 Ni -0.0016 -0.06 No surface attack along braze fillet

60 Pd—-37 Ni-3 Si - +0.0008 +0.027 . No attack along surface of braze fillet

BNi-7, 10 Cr—10 P—80Ni \ 0.0 00 No attack on braze fillet

50 Ni-25 Mo—25 Ge - 0.0 0.0 No attack along surface of braze fillet

BNi-1, 73 Ni—3.5 B-14 Cr—4.5 Fe-4 Si —0.0004 -0.016 No attack along fillet

BAu-2, 80 Au—-20 Cu =0.0007 —0.026 No attack on braze fillet _

75 Ni-25 Ge ~0.0001 -0.01 Max attack of 0.5 mil along surface at fillet

BCu, 100 Cu —-0.0006 °  -0.019 - 0.5 mil surface attack along braze fillet

65 Ni—-25 Ge-10Cr 0.0 0.0 Small subsurface voids to a depth of 0.5
mil along braze filtet

BNi-3, 90.5 Ni—-3.25 B-1.5 Fe-4.581 -0.0004 - ~0.05 Nonuniform attack of 6 mils along surface

' ' ' of fillet

BNi-5, 81 Ni—10 8i-13 Cr -0.0003 -0.012 Nonuniform attack of 12 mils along fillet

35 Ni=55 Mn—10 Cr -0.0111 - —0.48 Complete attack of braze fillet

40 Ni—60 Mn - —0.0159 -0.59 Complete attack of braze fillet

68 Ni—328n - , —0.0998 -3.49 Joint partially dissolved at fillet surface

 

 

“Brazmg alloys listed in order of decreasmg corrosion resistance to fluoride mixture.
bWelght-change data for brazing alloys and base material of joint.
 

n

A)

o

Ll

0

*}

 

could be determined.

ORNL-DWG 72-1124

       

HASTELLOY N

s

v

BRAZE
FILLET

Fig. 1. Lap joint configuration of braze alloy specimens.

4. RESULTS

The weight changes of the specimens as 2 function of time are given in Fig. 2 and Table 9. Over the
entire test, all specimens gained weight: a maximum of 0.024 g and a minimum of 0.014 gor 0.5 to 1.0%.
Weight losses for some specimens were seen in the time period between 1141 and 1471 hr and between
2978 hr and the end of the test. A defective pressure gage between a helium cylinder and the test vessel was
found at 1471 hr and replaced. This failure probably allowed moisture into the system, which would have
caused increased attack (weight loss). No cause for increased oxidation was found for the latter time period.

The salt analysis as a function of time is given in Table 10. Very small changes in impurity content are
noted. Little if any significant changes were noted diiring the periods of increased weight loss.

Figure 34 shows a macrograph of the specimens after test '(on bottom) as compared with identical
specimens untested:(on top). Figure 3b shows a close-up of a tested (right) and untested (left) specimen.
The deposits leading to the overall weight gains can be seen at this magnification (4'4 X).

" As mentioned earlier, two brazed specimens (lap joint between two Hastelloy N pieces) were made for
each braze alloy. Thus we were able to compare the braze microstructure after test with that of the
specimen prepared at the same time but not tested. Since the fillet size for each specimen before test may
not have been identical, we must necessarily restrict our observations to 'grbss changes. In making judgments
on the appearance of the microstructure, two effects must be considered: (1) the corrosive action of the
salt and (2) the aging of the braze due to its prolonged exposure at temperature. Percentages of elements in
phases were determined by electron probe microanalyzer. No element below the atomic number of sodium

 

 
 

 

'WEIGHT CHANGE (mg/cm?)

ORNL-DWG 72-1123

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5
4 /
-
1 /
6
3T
/ 3
5
2
0 1000 2000 3000 4000 5000 €000
TIME OF OPERATION (hr)
Fig. 2. Weight change of braze alloy specimens as a function of time.
Table 9. Specimen weight changes as a function of time
Specimen Weight changes (g)
No. First Next Next Next Next Total
594 hr 547 hr 330 hr 1507 hr 1798 hr 4776 hr
1 +0.0056 +00031 -0.0001 +0.0160 -0.0025 +0.0221
2 +0.0019 +0.0014 -0.0006 +0.0188 -0.0036 +0.0179
3 +0.0041 +0.0023 -0.0003 +0.0098 -0.0023 +0.0136
4 +.0072 +0.0038 +0.0002 +0.0127 +0.0005 +0.0244
5 +0.0028 +0.0027 +0.0001 +0.0128 —0.00i3 +0.0171
6 +0.0047 +0.0029 +0.0003 +0.0157 +0.0005 +0.0241
7 +0.0041 +0.0033 -0.0001 +0.0181 —0.0018 +0.0236
9 +0.0032 +0.0026 -0.0001 +0.0150 -0.0013 +0.0194
Table 10. Analysis of salt
Time Percent - Parts per million
(hr) Na B F ¢t Fe Ni Mo O, H+
Asreceived 21.6 978 689 12 294 32 <2 480 22
570 221 973 69.1 7 192 33 <2 490 19
1199 229 954 693 15 193 41 <2 640 29
1632 211 993 687 14 193 26 2 532 19
2978 209 924 890 16 252 48 65 609 24_
4991 225 946 6876 15 264 82 5 699 16

 

All other elements <2 ppm.
 

 

 

 

 

 

 

 

 

 

 

"

a)

A

H

a)

 

Y-109690

 

 

 

Fig. 3. Braze alloy specimerrl.‘ (@) As-brazed - top, exposed to NaBF ;—8 mole % NaF at 610°C for 4987 hr — bottom,
1% X ; (b) as-brazed — left, tested — right, 4'4X.

 
 

10

4.1 Braze Alloy BNi-7 (Ni—-13% Cr—10% P)

Figure 4 shows a braze fillet (untested and tested). Both braze alloys have dendrites extending from the
Hastelloy N. The major difference is the color of dendrites, which could be an artifact. Large deposits, seen
in Fig. 5, were found on the Hastelloy N after test. Figure 5 shows a microstructural analysis of an untested
braze and a braze exposed to the salt, and Fig. 6 shows the electron beam scanning images. The largest
deviation from the basic alloy composition occurred in dark dendrites, which were poor in phosphorus and
rich in chromium, extending from the Hdstelioy N. They were similar to the light-colored phases found in
the center of the tested braze alloy.'The overall appearance of the braze microstructure after the test was a
little different than the microstructure of the untested braze alloy, but the overall composition had not
changed.

42 Braze Alloy BNi-3 (Ni-3.5% Fe—4.5% Si—2.9% B)

Figure 7 shows an untested and a tested braze fillet. Certain phases are more defined in the tested
specimen. Small deposits were found on the Hastelloy N after test, and there was some interdiffusion at the
Hastelloy N—braze-alloy interface. Figure 8 shows the microstructural analysis of an untested and a tested
braze, and Fig. 9 shows the electron beam scanning images. The phases rich in iron were generally poor in
silicon. Boron could not be analyzed. Again, even though there was a difference in the appearance of the
tested and untested braze alloys, the coinpositions and amounts of each phase remained about the same.

4.3 Braze Alloy BAu-4 (Au—18% Ni)

Figure 10 shows an untested and a tested braze fillet. The exposure surface of the tested fillet scemed
to be somewhat roughened, with a slight difference in appearance of the as-brazed and the tested braze
alloys. There appears to be little interaction between the braze alloy and the Hastelloy N. Again, small
deposits were found on the Hastelloy N after test. Figure 11 shows the microstructural analysis of an
untested and a tested braze, and Fig. 12 shows the electron beam scanning image. The tested braze had a
very thin nickel deposit on its exposed surface. Some dark nickel-rich phases were found in both brazes.

4.4 Braze Alloy BAg-8 (Ag—28% Cu)

Figure 13 shows an untested and a tested braze fillet. A.deposit was evident along the exposed surface
of the tested braze alloy. Again, small deposits were found on the Hastelloy N after test, and there was little
interaction between the braze alloy and the Hastelloy N. Figure 14, shows the microstructural analysis of
an untested and a tested braze, and Fig. 15 shows the electron beam scanning image. Both phases making
up the eutectic braze alloy are easily identified along with nickel-containing phases formed during brazing.
The composition of the surface deposit of the tested fillet analyzed as 9% Ag, 41% Cu, and 50% Ni, which
means that the deposited material was essentially nickel which diffused into the braze alloy. :

4.5 Braze Alloy BNi4 (Ni—4.5% Si—3% B)

Figure 16 shows the as-brazed and the tested fillet. A large amount of fillet was missing in the tested
specimen, but the actual fillet size before test was not known. Diffusion between the braze alloy and the
Hastelloy N is evident. Rather large deposits were found on the Hastelloy N; Fig. 16¢ shows a typical
deposit. The microstructural analysis is seen in Fig. 17, and one phase with 11% Si is noted. Boron could
not be analyzed. The areas of lérge concentrations of silicon are seen in Fig. 18.
 

 

 

»n

F

oy

3

 

 

0,035 INCHES
N 100X

Jot

=

 

I

0.007 INCHES"
® 500X

e

 

 

Fig. 4. Braze alloy BNi-7 (Ni—-13% Cr—10% P)-Hastelloy N. As-polished. 100X. (4) As-brazed; () exposed to
NaBF;—8 mole % NaF for 4987 hr at 610°C; (c) deposit on Hastelloy N after test.

 
 

 

|
|
i
i
f
{

12

AS BRAZED

- Y=112352

 

TESTED

12 Cr
11P
75 Ni

 

12 Cr
10P
71 Ni

13Cr
1P
70 Ni

17 Cr
1P
74 Ni

17 Cr
1P

78 Ni

11 Cr

14 P

73 Ni

13 Cr
10 P
73 Ni

Fig. 5. Microstructural analysis of BNi-7 (Ni—13% Cr—10% P) braze fillet. As-polished. 500X. Top — as-brazed;

bottom — exposed to NaBF 4—8 mole % NaF for 4987 hr at 610°C. Reduced 19%.
      

 

     

a » | - " , ) . i

 

AS BRAZED Y-112360

 

BACKSCATTERED ELECTRONS - . PKa | T ke

Fig. 6. Electron-beam scanning images of BNi;T (Ni—13% Cr-10% P) braze fillet. Top — as-brazed;bottom - exposed to
NaBF4 -8 mole % NaF for 4987 hr at 610°C.

 
 

14

 

X001 oyt

 

v-110022 [

 

SIHINI SE00

atl|

il

 

 

-—

X000 o)
S3HIONI GE00
t

    

]

(b) exposed

.
»

100X%. (@) As-brazed

Fe—2.9% B) Hastelloy N. As-polished.

3 (Ni-4.5% Si-3.5%

Fig. 7. Braze alloy BN

C

NaF for 4987 hr at 610°

to NaBF4—8 mole %
 

 

 

 

 

 

 

 

of

»

(]

3

 

 

   

15

   

Y-112351

 

3 Fe
4 Si
90 Ni

N~
Za

TESTED

3-&&
zaym

 

Fig. 8. Microstructural anafysis of BNi-3_(Ni—4.5% Si—-3.5% Fe—2.9% B) braze fillet. Top — as-brazed; bottom —
exposed to NaBF;—8 mole % NaF for 4987 hr at §10°C. - '

 

 
AS BRAZED

 

   

BACKSCATTERED ELECTRONS

Fig. 9. Electron-beam scanning images of BNi-3 (Ni—4.5% Si—3.5% Fe-—2.9% B) braze fillet. Top — as-brazed; bottom
— exposed to NaBF4—8 mole % NaF for 4987 hr at 610°C.

Y-112356

 

91

 
 

 

 

 

 

 

 

 

 

 

 

17

"}

a)

|-

 

0.035 INCHES
IN 100X

o

b

 

0.035 INCHES
I 100X

Jou

 

 

 

"

Fig. 10. Braze alloy BAu4 (Au—18% Ni) Hastelloy N. As-polished. 100X. (a) Als-brazed‘l; (b) exposed to NaBF4—8
mole % NaF for 4987 hr at 610°C. ' ' S S

 

 

 
 

 

83 Au
11 Ni

 

18

AS BRAZED

TESTED

Y-112350

 

 

22 Av
78 Ni

15 Au
78 Ni

Fig. 11. Microstructural analysis of BAu-4 (Au—18% Ni) braze fillet. Top — as-braied; bottom — exposed to NaBF4—8

mole % NaF for 4987 hr at 610°C.

 

 
 

         

 

-y "

 

| Y-112359
AS BRAZED - _

61

 

BACKSCATTERED ELECTRONS. Nk e

Fig. !2 Electron-beam scanning images of BAu-4 (Au—18% Ni) braze fillet. Top — as-brazed; bottom — sxposed to
NaBF4~8 mole % NaF for 4987 hr at 610°C.

 

 

 

 
 

20

(b)

 

Fig. 13. Braze afloy BAg-8 (Ag-28% Cu) Hastelloy N. As-polished. 100
% NaF for 4987 hr at 610°C. :

 

0.035 INCHES
I 100X

e

-

 

0.035 INCHES
In 100X

fw

 

 

X. (a) As-brazed; (b) exposed to NaBF4—8 mole
21

   

 

 

 

 

 

 

 

   

. SR © y<i12349
'AS BRAZED Y-l
4 Ag
91 Cu
6 Ni
94 Ag
5Cv
15 Ag
48 Cu
% 9 Ag
5 Ag 4] (r:\lu
, 88 Cu '
6 Ni
91 Ag
; 6 Cu
| 58 Ag
38 Cu,
an 14. ch:rostructural analysxs of BAg-S (Ag—28% Cu) braze fillet Top —as brazed bottom - eXposed to NaBF4-8
z ' mole % NaF for 4987 hr at 610°C )

 

   

 

   

 

 
 

AS BRAZED

 

BACKSCATTERED ELECTRONS

Fifi. 15. Electron-beam scanning images of BAg-8 (Ag—28% Cu) braze fillet. Top — as-brazed; bottom — exposed to
NaBF4—8 mole % NaF for 4987 hr at 610°C.

NiKa

Y-112355

 

2

 
 

 

i
i
t
i

 

 

 

 

 

 

 

23

“v-110028

 

0.035 INCHES
I~ 100X

[

=

 

N 100X

 

e (), 3% INCHES
[ T

Y-110964

 

)

" 1] = 1

T27500X

jon

I

 

Fig. 16. Braze alloy BNi4 (Ni—4.5% Si—3.0% B) Hastelloy N. As-polished. 100X. (a) As-brazed; (b) exposed to
NaBF4—8 mole % NaF for 4987 hr at 610°C; (c) deposit on Hastelloy N after test.

 
 

24

 

Y-112348

AS BRAZED

 

.
&

 

 

TESTED

 

S —
vy
—
[

 

94 Ni

Fig. 17. Microstructural analysis of BNi4 (Ni—4.5% Si—3.0% B) braze fillet. Top — as-brazed

NaBF, -8 mole % NaF for 4987 hr at 610°C.

bottom — exposed to

 

 
 

 

25

     

Y-112353

  

- BACKSCATTERED ELECTRONS

 

o | | o AFTER TEST ) | S

 

 

 

.

S

* AR
LT PR W

)

-;"‘ir .
P
»

 

BACKSCATTERED ELECTRONS

 

+

- Fig. 18. Electron-beam scanning images of BNi4 (Ni—4.5% Si—3.0% B) braze fillet. Top — as-brazed; bottom —
exposed to NaBF4—8 mole % NaF for 4987 hrat 610°C.

 

  

 

 
 

26

4.6 Braze Alloy BCu (100% Cu)

Figure 19 shows the as-brazed and the tested fillet. A deposit was hofed along the exposed surface of )

the tested braze alloy. Little interaction between braze and base metal is noted. Larger deposits were found
on the Hastelloy N after test; Fig. 19¢c shows a typical deposit. The microstructural analysis shown in Fig.
20 shows constituents of the Hastelloy N base material mixed with the copper for both the tested and the
as-brazed material. The composition of the surface deposit of the tested fillet consists of equal amounts of
copper and nickel, with a smaller amount of nickel as you go from the surface. This probably means that
nickel is deposited on the surface during exposure and diffused in. The extent of the mixing of the nickel
and molybdenum in the matrix of the fillet before exposure is seen in the electron beam scanning images of
Fig. 21.

4.7 Braze Alloy BNi-2 (Ni—7.0% Cr—3.5% Fe—4.5% Si—2.9% B)

Figure 22 shows a little difference in appearance of the as-brazed and the tested fillet. A large amount
of diffusion between the Hastelloy N and the braze alloy is noted. A large amount of deposited material
was found on the Hastelloy N (Fig. 22¢). No compositional changes of the braze were found after test as
seen in Fig. 23. Figure 24 shows the electron beam scanning images. -

4.8 Braze Alloy BNi-2 (Ni—6.5% Cr~2.5% Fe—4.5% Si—3% B)

Figure 25 shows the as-brazed and the tested fillet. Considerable interaction between the Hastelloy N
and the braze alloy is seen. Small deposits were seen on the Hastelloy N. Few compositional changes of the
microstructure (Fig. 26) were seen after testing. Some phases very high in chromium are noted as seen in
Fig. 27. |

5. DISCUSSION

The corrosion resistance of metals to fluoride fuels has been found to vary directly with the “nobility”
of the metal, that is, inversely with the magnitude of free energy of formation of fluorides involving the
metal. Accordingly, corrosion of multicomponent alloys tends to be manifested by the selective oxidation
and removal of the least-noble component. In the case of Hastelloy N, corrosion is selective with respect to
chromium. Examples of chemical reactions which can cause this selective removal of chromium are: '

1. Impurities in the melt,

FeF, (NiF, or HF) + Cr=CrF, + Fe )

2. Dissolution of oxide films from the metal surface,
2Fe3* (or Ni?*) + 3Cr « 2Fe + 3Cr2*
3. Constituents in the fuel,

Cr+2UF, =2UF, +CiF, .

However, conditions could exist whereby soluble impurities in the salt could be reduced, resulting in
deposition on the metal surface.
27

 

 

 

N 00X

0,035 INCHES ————————————eeee e
[= {

[

 

I_

 

0.035 INCHES
iN 100X

|09

 

 

I

007 INCHES
anny

i

 

T ——

   

Fig. 19. Braze-alloy BCu (100% Cu) Hastelloy N. As-polished. 100X. (a) As-brazed; (b) exposed to NaBF 4—8 mole %
NaF for 4987 hr at 610°C; (¢) deposit on Hastelloy N after test.

 
 

 

_ Y-112347
? - AS BRAZED

91 Cu
: 7 Ni
1Cr

 

22 Cu
11 Ni
4Cr
48 Mo

 

 

 

 

1
i
3
{
i

1
i
i

 

Fig. 20. Microstructural analysis of BCu (100% Cu) braze fillet. Top — as-brazed; bottom — exposed to NaBF4-8
mole % NaF for 4987 hr at 610°C. ' ‘

 

 

 

 
 

 

 

 

 

 

 

 

 

Mola

bottom — exposed to

NiKa

 

 

 

BACKSCATTERED ELECTRONS

(2]
)y
™
- o~
—
o—
!
>
2
Pl )
Reedioh
=
N
g o
1
Oy
Pt ....w... ..r.u.,.. -,
: .“.ww..w\._‘.
[ —
” L 9
N g =
o = _
@ N .-."— ) .J»Jhwww‘gu fim%&.
AR T ».V.WJ*
(%] W e w“ .
c < <
-~y

 

BACKSCATTERED ELECTRONS

Fig. 21. Electron-beam scanning images of BCu (100% Cu) braze fillet. Top — as-brazed

NaBF4 -8 mole % NaF for 4987 hr at 610°C.

 
 

|-

~T 160X

j [

 

Y-110933

 

0.035 INCHES ~
N 100%

o

 

¥-110976

o
3
T
U
z
™~
O
b
o
5
6

 

|

500X

 

 

Fig. 22. Braze alloy BNi-2 (Ni—7.0% Cr—4.5% Si—3.5% Fe—2.9% B) Hastelloy N. As-polished. 100X. () As-brazed;
(b) exposed to NaBF 4—8 mole % NaF for 4987 hr at 610°C; (¢) deposit on Hastelloy N after test.

u
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

whk

-y

31

     

AS BRAZED T

88 Ni
/2 Fe
4 Cr

84 Ni
3 Si
4 Fe
6Cr

 

Fig. 23. Mlcrostructural analysis of EN 2 (Nl—‘i 0% Cr—-4.5% Si-3.5% Fe-2. 9% B) braze fillet Top — as-brazed;
bottom — exposed to NaBF,4 -8 mole % NaF for 4987 hrat 610°C. - -

 
 

 

- o | AS BRAZED

 

Crka

 

 

CrKa

BACKSCATTERED ELECTRO NS , FeKa

Fig. 24. Electron-beam scanning images of BNi-2 (Ni-—?.O% Cr-4.5% Si—3.5% Fe-2.9% B) braze fillet. Top —
as-brazed; bottom — exposed to NaBF 4 -8 mole % NaF for 4987 hr at 610°C. ‘

.

Y-112357

 

 

[A

 
 

 

 

 

 

 

 

 

1
}
i
!
|
1
J

T e S T

 

4

 

33

-

 

0.035 INCHES
™ 100X

|7

 

0.035 INCHES
I 100%

Jn

 

 

" Fig. 25. Braze alloy BNi-2 (Ni—6.5% Cr—4.5% Si—3.0% B—2.5% Fe) Hastelloy N As-polished. 100X (a) As-brazed
(b) exposed to NaBF;—8 mole % NaF for 4987 hr at 610°C.

 
 

 

 

 

 

34
Y-112345 ,
AS BRAZED
82 Ni 72 g'
2 Fe 4 Fe
SCr 10Cr
2 Ni
77 Ni ,
3 Fe 73 Cr
35Si
7 Cr .
85 Ni
2 Fe
6 Cr
83 Ni
3 Fe
3si
7 Cr

 

Fig. 26. Microstructural analysis of BNi-2 (Ni—6.5% Cr—4.5% Si-3 0% B-2.5% Fe) braze fillet. Top — as-brazed; .
bottom — exposed to NaBF4—8 mole % NaF for 4987 hr at 610°C. -
 

 

 

" ) - ¥

~ AS BRAZED

; ; .
BACKSCATTERED ELECTRONS

Fig. 27. Electron-beam scanning images of BNi-2 (Ni-6.5% Cr—4.5% Si-3.0% B-2.5%
as-brazed; bottom — exposed to NaBF4—8 mole % NaF for 4987 hr at 610°C. ‘

Fe) braze

 

 

fillet, Top

 

 

-t

Y-112354

 

 

gE

 
 

 

 

36

Another consideration in this experiment-is the effect of joined dissimilar metals. A potential difference
usually exists between two dissimilar metals when they are immersed in a corrosive or conductive solution.
If these metals are placed in contact (or otherwise electrically connect'ed),~ this potential difference
produces electron flow between them. Corrosidn of the less resistant metal (or alloy) is usually increased,
and attack of the more resistant material is decreased as compared with the behavior of these metals when
they are not in contact.

The potential differences between metals under reversible, or ‘nonco'rroding,' conditions form the basis
for predicting corrosion tendencies. These potentials can be indicated by taking the absolute differences
between the free energies of formation of the corrosion products. Table 11 gives the free energy of
formation of various fluorides at 600°C and includes metals present in the Hastelloy N or in the braze alloy.
No series of potentials or free energies exist‘fot alloys in molten salts like that found for alloys in seawater.

The potential generated by a galvanic cell consisting of dissimilar metals can change with time. As
corrosion progresses, reaction products or corrosion products may accumulate at either the anode or
cathode or both and reduce the speed at which corrosion proceeds. The polarizability of one of the metals
of the galvanic couple also may affect the galvanic action; thus it is quite difficult to predict the galvanic
behavior, and experiments necessarily must be conducted.

We will discuss the nobility of the braze alloy as compared with the Hastelloy N, considering the
constituents of each and the free energy of formation of the various fluorides strictly on a theoretical

thermodynamic basis with no regard for kinetics. Braze alloys 1, ‘7,' and 8 contain chromium, silicon, and-

boron, which form rather stable fluorides. Thus we assume that the braze alloy is probably less noble than
Hastelloy N. Braze alloys 2 and 5 also contain sufficient amounts of silicon and boron, so again these alloys
are probably less noble than the Hastelloy N. Because of the presehce of the relatively noble gold, silver,
and c0ppér, braze alloys 3, 4, and 6 should be more noble than the Hastelloy N. If there is corrosion, braze
alloys 1, 2, 5, 7, 9 should be selectively attacked with respect to Hastelloy N, and Hastelloy N should be
attacked relative to braze alloys 3, 4, and 6. We point out that all the brazed specimens have Hastelloy N in
common. Thus, those brazes less noble than Hastelloy N should corrode more than the brazes more noble
than Hastelloy N if they are in the same system. Therefore, again emphasizing that there is no consideration

Table 11. Relative thermodynamic stabilities of fluoride compounds formed
by elements employed as alloying additions at 600°C?

 

 

Moststable . Standard free energy of formation
fluoride compound per gram atom of fluorine (kcal per gram-atom of F)
BF; | - 87
SiF4 -85
Cl‘Fz } - =76
FCFQ —-69
NiF; - . ‘ -62
MOFs - ) o -'58
CuF, - . . -50
PF; . —50
AgF - -38
CF, o : -32
© AuFy L - -19

 

9A. Glassner, The Thermochemical Properties of the Oxtdes, Fluorides, and
Chlorides to 2500° K, ANL-5750 (1957).
 

 

 

v

»

wy

37
for kinetics, we would predict the following order of corrosion resistance based on alloy composition and
free energy of formation of the fluorides:

Braze alloy  No.

[ BNi-7 [ 1 Least corrosion resistant

BNi-2 7
BNi-2 9
‘BNi-3 2
BNi4 5
'BAu4 3
BAg-8 4
BCu 6 Most corrosion resistant

Brackets indicate élloys in the same order

Since the final weight-change results of this test showed deposition and since deposition is often a function
of nucleation sites and times, it is quite difficult to quantify Weiglltfgéin data and make comparisons
between specimens on weight-gain data. However, since there were two time periods when the oxidizing
conditions of the salt were such that some material was lost from the alloy, we decided that these points

- were sufficient for an experimental comparison of corrosion resistance.

During both time periods when some material was lost from the specimens (Table 9), the BNi-3 braze

“alloy (No. 2) lost the most weight while the BAg-8 (No. 4) and BCu (No. 6) braze alloys showed the most

corrosion resistance. Thus, on the basis of these two time periods when oxidation was greatest the following
list is given in the order of increasing corrosion resistance as determined from our experiment:

Braze allojr No.

BNi-3 2 - Least corrosion resistant
BNi7  [1
BAu4 3
BNi-2 7
BNi-2 9
BNi4 5
[BAg8  [4
BCu 6 Most corrosion resistant

Brackets indicate alloys in the same order

. The experimental results agree with our theoretical listing in the case of specimens 4 and 6; however,
specimens 3 and 2 were not as resistant as expected. No explanation can be given for the behavior of
specimen 3 (the Au-Ni braze) since, .in the past, it has proven to be quite resistant in molten salts.
Specimens 1, 2, 7, and 9 all contain from 10 to 14% Si, B, and Cr,>all of which form stable fluorides, so
their position in the table is not surprising. It is also gratifying that the BNi-2 brazes from different
manufacturers acted about the same. Specimen 5 has smaller amounts of these elements and improved
corrosion resistance. Thus consideration of the free energy of formation of corrosion products by the
alloying elements in the braze can be used torpredict the corrosion resistance of most of the brazes.
Another area to consider is the nobility of the braze as compared with the Hastelloy N. In'most cases it
would be preferable to have a braze alloy more noble than the Hasté!loy N since there is a much smaller
quantity of braze alloy that could be attacked. All in all, none of the corrosion rates are excessive.

 
 

 

38

ORNL-DWG 65-13233R2

  
       

'6.."0'

0‘0.0

 
   
 
     
  

A2

  

NN

         
      

  
  
   
    

© ALLOY A —_— - b ALLOY B
CONTAINING /7 - TE=—— KX CONTAINING
COMPONENTS MOVEMENT BY : R COMPONENTS

>

55

ey

  

)
9

e

    
   
 

NN

M+ Myt +M; - My _ DIFFUSION OR CONVECTION
= — - >

Wyt N+ 4N 4

(52
oo

         
  
   
 
    

W

-, v
000.

. — = = .

REACTION OF M, ==
WITH LIQUID =
. =

\\\\\\\\\\\\

SN

DIFFUSION OF M; TO
SURFACE OF ALLOY A
————————

SURFACE REACTION

DIFFUSION INTO
b% ALLOY B

 

DRIVING FORCE FOR THE TRANSFER OF ANY COMPONENT
IS THE DIFFERENCE_IN CHEMICAL POTENTIAL OF A% IN ALLOY A
COMPARED WITH' ALLOY B.

Fig. 28. Dissimilar-alloy mass transfer.

It is also interesting to note that specimens 4 and 6 had deposits on the braze fillets. These deposits
appear to be due to activity-gradient mass transfer. Mass transfer of this type only requires the presence of
different alloys in the same fluid. Figure 28 shows a schematic of this process. The sequence of events in
dissimilar alloy mass transfer involves removal of material from one metal and deposition on a second
metal, that is, movement from a region of high activity to.one of low activity. Examples of dissimilar metal
interactions have been seen in niobium and type 316 stainless steel systems exposed simultaneously to
sodium-potassium alloy,?® other liquid metal-alloy systems,* and in a Hastelloy N—Haynes alloy No.
25—molten sodium fluoroborate mixture system.2® In our case, assuming the products to be deposited
consisted mainly of the constituents of Hastelloy N (which is mostly nickel), the only materials in the
system that do not contain nickel are the braze alloys of specimens 4 and 6. Thus the deposits of nickel
(which diffused into the braze alloy) were a result of transfer to an area of low activity.

The fact that very little salt chemistry change was seen also was evidence that we had activity-gradient
transfer between the braze alloy and the Hastelloy N with the salt only acting as a carrier. However, the
. weight gains are evidence that some material is transferred from the salt or the pot to the alloys.
Calculations show that only 50 ppm of material needed to be removed from the salt or pot to equal the
measured weight gains. :

The micrographs substantlate the small measured weight gams and show no real evndence of attack by
the fluoroborate mixture. Of most interest is the interaction between the braze material and the Hastelloy
N in certain cases. The nickel-base braze alloys, with the exception of specimen No. 1 which differed from
the others since it contained 10% phdsphorus, showed areas of diffusion between the braze alloy and the

 

23. L R DiStefano, Mass flansfer E‘ffects on Some Refractory Metal-Alkali Metal Stainless Steel Systems,
ORNL-4028 (November 1966). ‘ :

'24. J. H. DeVan, Compatibility of Structural Materials with Bozlmg Potassium, ORNL-TM-l 361 (Apnl 1966) _

'25. 1. W. Koger and A. P. Litman, Mass Transfer Between Hastelloy N and Haynes Alloy No. 25 in aMolten Sodium
Fluoroborate Mixture, ORNL-TM-3488 (October 1971). .
 

 

ok

. into the braze alloy. The International Nickel Company

39

Hastelloy N. No interaction was seen between the gold-, silver-, or copper-based braze alloy and the
Hastelloy N. In specimen No. 1, dendrites of composition Ni—15% Cr—1% P extended from the Hastelloy N
26 recommends that brazing alloys containing
phosphorus never be used with any nickel alloys because of the possibility of formation of a phosphide of
nickel at the bond. If impact or bending then occurs, failure would be imminent. However, in our case,
according to the electron beam scanning images the phosphorus seems to be fairly well dispersed.

The Stellite Division of Cabot Corporation?” warns against using brazing alloys that contain copper
with Hastelloy alloys. They say that the infusion of copper into grain boundaries can affect both the
corrosion resistance and the mechanical properties. We saw no evidence, in either of our copper-containing
brazes, of copper interaction with the Hastelloy N. We did see mixing of the Hastelloy N constituents with
the copper. Copper does altoy quite readily with nickel but will generally not flow too far before it has
picked up enough nickel to raise its liquidus and reduce its fluidity.

Because of strength and oxidation characteristics BAg-8 alloy No. 4 (Ag—28% Cu) is generally not used

‘at temperatures over 400°C, and the copper braze alloy should probably not be used at temperatures above

500°C.2% We observed no problem with these alloys at 610°C in a molten fluoride salt.

6. CONCLUSIONS

1. On the basis of corrosion resistance, all the braze alloys tested under our conditions are compatible
with NaBF, —8 mole % NaF at 610°C. The Ag—28% Cu and 100% Cu braze alloys were the most resistant.

2. Prediction of corrosion resistance on the basis of free energy of formation data is reasonable.

3. Nickel transferred to non-nickel-containing braze alloys through an activity-gradient mass transfer
mechanism.

4. Some deposits were noted on the Hastelloy N base material.

5. Diffusion between the Hastelloy N and the braze alloy occurred only with those braze alloys whose
composition was near that of Hastelloy N.

 

26. Huntington Alloy Products Division of the International Nickel Co., Inc., Joining the Huntington Alloys,
Technical Bulletin T-2, 1967. ‘
27. Stellite Division of Cabot Corporation, Fabrication of Hastelloy Alloys, F-30, 126F, 1970.

 
 

 

  
 

  
 

 

i

 

ORNL-TM-4272
INTERNAL DISTRIBUTION
(79 copies)

(3) Central Research Library | (5) J. W. Koger
ORNL — Y-12 Technical lerary E. J. Lawrence
Document Reference Section A.L. Lotts .

(10) Laboratory Records Department T. S. Lundy
Laboratory Records, ORNL RC R.N. Lyon
ORNL Patent Office H. G. MacPherson
G. M. Adamson, Jr. R. E. MacPherson
C.F. Baes W. R. Martin
C. E. Bamberger R. W. McClung
S. E. Beall H. E. McCoy
E. G. Bohlmann C. J. McHargue
R. B. Briggs H. A. McLain
S. Cantor B. McNabb
E. L. Compere L. E. McNeese
W. H. Cook A. S. Meyer
F. L. Culler | . R. B. Parker ?
J. E. Cunningham P.Patriarca r
J. M. Dale A. M. Perry
J. H. DeVan M. W. Rosenthal
J. R. DiStefano H. C. Savage
J. R. Engel J. L. Scott
D. E. Ferguson J. H. Shaffer
J. H. Frye, Jr. G. M. Siaughter
L. O. Gilpatrick G. P. Smith
W. R. Grimes R. A. Strehlow
A. G. Grindell R. E. Thoma
W. O. Harms D. B. Trauger
P. N. Haubenreich A. M. Weinberg

(3) M. R. Hill J. R. Weir
W. R. Huntley J. C. White
H. Inouye L. V. Wilson
P. R. Kasten

EXTERNAL DISTRIBUTION
(24 copies)
BABCOCK & WILCOX COMPANY, P. O. Box 1260, Lynchburg, VA 24505
B. Mong
BLACK AND VEATCH P O. Box 8405, Kansas City, MO 641 14
.C. B. Deering
BRYON JACKSON PUMP, P. O. Box 2017, Los Angeles, CA 90054
_G. C. Clasby
CABOT CORPORATION, STELLITE DIVISION, 1020 Park Ave., Kokomo, IN 46901
T. K. Roche

41
 

 

42

CONTINENTAL OIL COMPANY, Ponca City, OK 74601 - | -
J. A. Acciarri |

EBASCO SERVICES, INC., 2 Rector Street, New York, NY 10006
D. R. deBoisblanc
T. A. Flynn .

THE INTERNATIONAL NICKEL COMPANY, Huntmgton WV 25720
J. M. Martin

UNION CARBIDE CORPORATION, CARBON PRODUCTS DlVlSlON 12900 Snow Road, Parma, OH 44130
R. M. Bushong

USAEC, D{VISION OF REACTOR DEVELOPMENT AND TECHNOLOGY Washington, DC 20545

David Elias
J.E. Fox
Norton Haberman
C. E. Johnson

-T.C. Reuther
S. Rosen

" Milton Shaw
J. M. Simmons

USAEC, DIVISION OF REGULATIONS, Washmgton DC 20545
A. Glambusso
USAEC, RDT SITE REPRESENTAT[VES Oak Ridge National Laboratory, P. O. Box X, Oak Ridge, TN 37830
D.F.Cope
Kermit Laughon
C. L. Matthews

USAEC, OAK RIDGE OPERATIONS, P. O. Box E, Oak Ridge, TN 37830
Research and Technical Support Division :

USAEC, TECHNICAL INFORMATION CENTER, P. O. Box 62, Ozk Ridge, TN 37830 ~
(2) _