ORNL-TM-4271.txt

   
 
 
 
  
 

 

ORNL-TM-4271

 

13*

MASS TRANSFER BETWEEN
HASTELLOY N AND A MOLTEN SODIUM
FLUOROBORATE MIXTURE IN A
THERMAL CONVECTION LOOP

 

J. W. Koger

 

 

RIDGE

TN R AT R LR AL LN

 
 

"4

 

 

This report was prepared as an account of work sponscred 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,

 

 

 

 

 

 
N

2

 

 

ORNL-TM-4271

Contract No. W-7405-eng-26

METALS AND CERAMICS DIVISION

MASS TRANSFER BETWEEN HASTELLOY N AND A MOLTEN SODIUM FLUOROBORATE

MIXTURE IN A THERMAL CONVECTION LOOP

J. W. Koger

DECEMBER 1972

 

NOTICE—mMm——

This report 'was prepated as an account of work
.| sponsored by the United States Government, Neither
_the United States nor the United States Atomic Energy

Commission, nor sny 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.

 

 

 

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

U.S. ATOMIC ENERGY COMMISSION MASE‘ER

-

BISTRIBUTION OF THIS DOCUMENT IS UN-LIMIng'

 

 
 

 

 

 
 

»

¥

 

CONTENTS
ADSTACE & e e ettt e e e e e e e e e e e e e e 1
| (18 (0T LT T2 4 o ¢ e 1
2. Experimental Details .......... ettt it e e e e 1
2.0 Materials . ..ot it e e et 3
2.2 SaltPreparation .. ... ... . it i e 3
2.3 Loop Operation . .........ciiutiiitiniintinn ittt e 3
K T 2= 31 1 2 e 5
3.1 Operation, Weight Changes, and Salt Chemistry ............. .. .. . i iiiiiaiiianl, 5
3.2 Specimen Analysis . .......c.iiiiiiii i e i it it e 6
4. Discussion .......... ..., [ it ittt ettt 17
T 03 To3 11 .3 (0 ¢ T 19

iii

 
 

 

"

 

 

 
 

 

 

 

¥

 

MASS TRANSFER BETWEEN HASTELLOY N AND A MOLTEN SODIUM
FLUOROBORATE MIXTURE IN A THERMAL CON VECTION LOOP

J. W. Koger

" ABSTRACT

We showed that temperature-gradient mass transfer occurred in a Hastelloy N thermal convection
loop circulating a sodlum fluoroborate-—-sodmm fluoride mixture (NaBF4—8 mole % NaF) at
temperatures of 687 °C maximum and 438°C minimum for 19,930 hr. The maximum calculated
corrosion rate (based on uniform removal) was 0.29 milfyear, Overall, the compatibility in this system
was quite good. Impurities such as air inadvertently admitted to the loop system increased the mass
transfer rate. Material (chromium, iron, and nickel) removal resulted in voids in the hot leg. Iron and
nickel deposited in the cold section, but chromium ions remamed in the salt, Some of the deposits in
the cold leg appeared to have cubic symmetry.

1. INTRODUCTION

A thermal convection loop, designated NCL-20 and similar to that pictured in the foreground of Fig. 1,
began operation in December 1969 and was drained in March 1972, after 19,930 hr at design conditions.
The purpose of the test was to determine the compatibility of standard Hastelloy N with a sodium
fluoroborate—sodium fluoride eutectic mixture, NaBF,;—8 mole % NaF, at the most extreme temperature
conditions considered (687°C maximum and 438°C minimum) for the molten-salt breeder reactor (MSBR)
secondary circuit. The temperatine extremes are the estimated wall temperatures in the secondary circuit of
an MSBR. | |

The corrosion of Hastelloy N by the sodium fluoroborate mixture has been studied for several years in
both thermal convection and pumped loops but usually at temperatures of 610°C and lower.! ~!3 The
experiment discussed in this report was conducted to determine the mass transfer characteristics at higher
temperatures with a larger temperature gradient.

2. EXPERIMENTAL DETAILS

- The test system consisted of a thermal convection loop in a harp configuration, with surge tanks atop
each leg for sample and specimen access. Flow was generated by the difference in density of the salt in the
hot and cold legs of the loop, and the salt flow velocity was approximately 7 ft/min. The loop was operated

 

1. J. W. Koger and A. P, Litman, Compatibility of Hastelloy N and Croloy 9M with NaBF 4-NaF-KBF4 (90-4-6 Mole
%) Fluoroborate Salt, ORNL-TM~2490 (April 1969).
2.-J. W. Koger and A. P. Litman, Catastrophic Corrosion of Type 304 Stamless Steel in a System Circulating Fused
Sodium Fluoroborate, ORNL-TM-2741 (January 1970).
3. J. W, Koger and A. P, Litman, C’ompatzbzl:ty of Fused Sodium Fluoroborates and BF3 Gas with Hastelloy N Alloys,
ORNL-TM-2978 (June 1970). -
‘4. J.: W, Koger, Corrosion and Mass Transfer Charactenstzcs of NaBF4-NaF (92-8 Mole % } in Hastelloy N,
ORNL-TM-3866 (October 1972).
5. J.W.Koger and A.P. Litman, MSR Program Semiannu. Progr. Rep. Feb. 29, 1968, ORNL-4354, pp. 221--25.
6. J. W. Koger and A. P. thman MSR Program Semiannu. Progr. Rep. Aug. 31, 1968, ORNL-4434, pp. 264—66 and
285-—89.
W, Koger and A.P. thman MSR Program Semiannu. Progr. Rep Feb. 28, 1969, 0RNL-4369 pp. 246-53.
. J.W.Koger and A.P. Litman, MSR Program Semiannu. Progr. Rep. Aug. 31, 1969, ORNL-4449, pp. 200-208.
. J. W. Koger, MSR Program Semiannu. Progr. Rep. Feb. 28, 1970, ORNL-4548, pp. 242-52 and 265-72.
. J.W.Xoger, MSR Program Semiannu. Progr. Rep. Aug. 31, 1970, ORNL-4622, pp. 168~78.
. J. W. Koger, MSR Program Semiannu. Progr. Rep. Feb. 28, 1971, ORNL-4676, pp. 192-215.
. Y. W. Koger, MSR Program Semiannu. Progr. Rep. Aug. 31, 1971, ORNL-4728, pp. 138-53.
. J. W. Koger, MSR Program Semiannu. Progr. Rep. Feb. 29, 1972, ORNL-4782, pp. 174--82.

 
 

 

T PHOTO 75125 A

   
   

Fig. 1. Hastelloy N thermal convection loop NCL-ZO-containing NaBF4 -8 mole % NaF at a maximum temperature of 687°C
with a temperature difference of 249°C.

 
 

e

 

at a maximum temperature of 687 C and a temperature dlfference of 249° C Forced air cooling of the cold
leg was required to obtain this AT. A schematic of the loop is given in Fig. 2.

2.1 Materials

The harp portion of the loop system was fabricated from 0.750-in. OD, 0.072-in. wall Hastelloy N
tubing. This material was part of heat 5095, which was purchased from Wall Tube and Metal Products
Company, Newport, Tennessee. The Oak Ridge National Laboratory Inspection Engineering identification
number is IR-8402-2. The finished loop system was stress relieved at 900°C for 6 hr in hydrogen. '

The loop contained 14 standard Hastelloy N specimens and two titanium-modified Hastelloy N

| specimens 0.75 X 0.38 X 0.030 in., each with a surface area of 0.55 in.? (3.5 cm?). Modifications to the

standard Hastelloy N composition are being made to test for improved mechanical properties after
irradiation and were included in this test to check their compatibility with the sodium fluoroborate
mixture. The compositions of the standard and modified Hastelloy N specimens are given in Table 1. Eight
specimens were attached at different vertical positions on each of two %-in. rods. One rod was inserted in
the hot leg and another in the cold leg. The rods were placed into or removed from the loop from
standpipes atop each leg. The rods were moved through a Teflon sliding seal compression fitting at the top
of the standpipe and a ball valve at the bottom. Another ball valve on the loop above each leg assured
removal or insertion without dlsturbmg loop operation or introducing air contamination.

“Table 1. Nominal compositions of corrosion specimens ‘

In percent by weight

 

 

Cr Fe Mo Ni Ti
Standard Hastelloy N 7.4 4.5 17.2 70.0 0.02
Ti-modified Hastelloy N 7.3 <0.1 13.6 77.0 - 0.5

 

2.2 Salt Preparation

The fluoroborate salt mixture was furnished by the Fluoride Processing Group of the Reactor
Chemistry Division, and its composition before test is given in Table 2. To mix and purify the salt, the raw
materials were first heated in a nickel-lined vessel to 150°C under vacuum and held for 15 hr. Then the salt
was heated to S00°C, agitated with helium for a few hours, and transferred to the fill vessel.

Table 2. Salt analysis before test

 

 

' Content : - Content
Elementr % Element (ppm)
Na 214 Cr 58
B 9.47 " Ni 5
F 68.4 Fe 227
' 0. 600

Mo 2

 

 

2.3 Loop Operation

The loop was heated by pairs of clamshell heaters placed end to end, with the input power controlléd
by silicon controlled rectifier units and the temperature controlled by a current-proportioning controller.

 
 

b e et e v e AL

ORNL-DWG 68-3987

     
 
 

STANDPIPE

   
 
 

TWiN
BALL VALVES

S
prm——

=
_—__”‘—_-—___-—_’II‘/
FI'_II

r'.“ j

; 1

CLAMSHELL -
HEATERS i

 

 

i
-INSULATION II

CORROSION

SPECIMENS
\.fln :Ii

| 2

$ SAMPLER

 

 

|
I
|
|
|
I
|
30 in. I
I
I
|
I
|
|

 

 

 

 

/

TN

 

 

INCHES

“FREEZE
VALVES

 

 

 

FLUSH
TANK DuMpP
TANK

 

 

 

 

 

 

Fig. 2. MSRP natural circulation loop and salt sampler.

 
 

 

 

n

0

h

 

Temperatures were measured by Chromel-P vs Alumel thermocouples that were spot welded to the outside
of the tubing, covered by a layer of quartz tape, and then covered with stainless steel shim stock. Tubular
electric heaters controlled by variable autotransformers furnished the heat to the cold leg portions of the
loops. ' ' _ -

Before filling with salt, the loop was degreased with ethyl alcohol, dried, and then heated to 150°C
under vacuum to remove any traces of moisture. A helium mass spectrometer leak detector was used to
check for leaks in the system, |

The loop was filled by heating the harp section, the salt pot, and all connecting lines to approximately
550°C and applying helium pressure to the salt supply vessel to force the salt into the system. Air was
continuously blown on freeze valves leading to the dump and flush tanks to provide a positive salt seal. All
fill lines exposed to the fluoroborate salt were Hastelloy N, and all temporary connections from fill line to
loop were made with stainless steel compression fittings.

A flush salt charge, intended to remove surface oxides and other impurities, was held for 24 hr in the
loop at the maximum operation temperature and then dumped. The loop was then refilled with fresh salt.
Once the loop was filled, the heaters on the cold leg of the loop were turned off. As much insulation as
necessary was removed to obtain the proper temperature difference by exposing the cold leg to ambient air.
Helium cover gas of 99.998% purity and at a slight positive pressure (approx 5 psig) was maintained over
the salt during operation.

Corrosion specimens were withdrawn periodically along with salt samples to follow corrosion processes
as a function of time. During the removal periods, all specimens were weighed and measured. Portions of
the hottest and coldest specimens were removed and examined metallographically during test, and all
specimens were examined metallographically after test. After the end of the experiment, the hottest and
coldest specimens were analyzed by scanning electron microscopy (SEM) and x-ray fluorescence. The x-ray
fluorescence was induced by bombardment by the electron beam of the SEM.

3. RESULTS
3.1 Operation, Weight Changes, and Salt Chemistry

- For the first 11,900 hr, weight-change measurements showed losses in specimens from the hot section
and gains from those in the cold section. The maximum weight loss was 6.2 mg/cm? (corrosion rate 0.2
mil/year, assuming uniform removat), and the maximum gain was 2.8 mg/cm?. Table 3 gives the weight loss
of the hottest specimen and the concentration of the major impurities in the salt as a function of time.

_Table 3. Weight change of specimen at 687°C and
~impurity content of salt of NCL-20

 

 

 

Time Maximum weight .Incre{nental Chemical analysis (ppm)
(hr) loss corrosion rate - -
: (mg/ sz) - (mils/year)® .CI Fe Oxide
o o e 58 227 900
624 03 0.19 90 198 550
1,460 0.7 0.18 99 . 110 590
2,586 1.0 0.15 109 104 470
3,784 1.7 0.17 131 91 670
8,500 4.3 0.2 175 90 550
11,900 - 6.2 0.2 198 90 720

 

2 Assuming uniform loss.

 
 

 

Operation of the loop was uneventful during these 11,900 hr, and the weight changes of the specimens

and the chemistry changes of the salt indicated that the mass transfers were the lowest attained yet with the

fluoroborate mixture in a temperature gradient system. At 12,429 hr the normal 20-psia helium
overpressure' dropped to atmospheric pressure. Investigation disclosed that the safety valve on the helium
regulator had failed and had released helium from the loop. To determine if the salt had become
contaminated with air or moisture, the specimens were immediately removed from the loop and weighed.
The maximum metal loss durifig this additional time period (525 hr) was fairly low (0.34 mil/year) with a
correspondingly low weight gain on cold leg specimens. Thus, the mass transfer rate increased but did not
greatly accelerate. A salt analysis did not disclose any gross changes. The distance from the regulator to the
loop was approximately 20 ft, and the gas line was Y -in. tubing; therefore, movement of air into the loop
system was somewhat difficult but not impossible. The valve was replaced, and loop operation continued.

During the next 1850 hr, the maximum corrosion rate (assuming uniform loss) was 0.7 mil/year. Part of
this increased mass transfer was associated with the regulator problem and some impurity inleakage, but
further inspection disclosed a leaking mechanical pressure fitting. The fitting was repaired, and the loop
continued operation.

During the last few hundred hours of operation, variations in flow were noted. The specimens were
removed and the salt was frozen in the loop. On removal of the heaters we found that the loop tubing had
been severely damaged and pitted because of direct contact with the heater wire. The insulator bushing
normally used to separate the heater and the tubing had slipped out of place. We decided to make repairs
by weld overlaying Hastelloy N on the damaged portions of the tubing. Because of the possibility of
penetration through the tubing while overlaying, the heaters were temporarily wired back in place and the
loop was drained. This terminated the operation of NCL-20 after 19,930 hr of operation.'*

Table 4 lists the specimens, their position in the loop, the temperature of the salt at the various
positions, thickness measurements, and weight changes after 19,300 hr exposure time. The specimens in the
hottest position showed the greatest loss, 14.4 mgfcm? for the standard Hastelloy N and 15.5 mg/cm? for
the modified Hastelloy N. The maximum corrosion rate (assuming uniform loss) was 0.29 mil/year. The
maximum weight gain in the cold leg was 4.0 mg/cm?. Thickness measurements were made before and after
test, and the changes in thickness corresponded quite well to the weight changes, that is, weight losses were
associated with decreases in thickness and weight gains occurred with increases in thickness. Figure 3 shows
the weight changes of specimens in the loop as a function of time and temperature, and Fig. 4 shows the
overall picture of the mass transfer in the system. '

3.2 Specimen Analysis

Figure 5 shows optical micrographs of all the corrosion specimens and some of the loop tubing. Note

the attacked areas on the pieces exposed at the higher temperatures and the deposits on the cold leg

specimens and the bottom specimen of the hot leg, which also gained weight. It is interesting to note that
the weight changes correlate well with the observed microstructural changes. Figure 6 shows micrographs of
specimens in the hottest and coldest positions of the loop after 11,900 hr exposure and at the end of the
test (19,300 hr). The attacked area is a little more defined after the longer exposure times, and perhaps a
few more deposits can be seen.

The two Hastelloy N specimens exposed to the hottest and coldest conditions (Figs. 6¢ and 6d) were
analyzed by scanning electron microscopy (SEM) and x-ray fluorescence (x-ray fluorescence was induced

 

14. The repair was successful and the loop was filled with another charge of salt. This “revised” loop was designated
NCL-20A and was operated over 2000 hr before termination. However, this report will deal specifically with the results
from loop NCL-20.. )

o

 
 

 

 

 

©

H

#

 

Table 4. Corrosion specimen data after 19,300 hr exposure

 

 

) ) Original Final Obi.e“'ed :"pfl‘
Spe mmeél Position® T em%erature Weight chgnge thickness®  thickness® o attac' or
number ‘ (C) {mg/cm*) (mils) (mils) deposit

(mils)

H-01 Hot leg surge tank, vapor BF 3-He 687 . =03 30.2 30¢ 0
H-02 Hot leg surge tank 687 ‘ -5.6 293 28 0.5
H-03 Hot leg 687 , -144 294 28¢ 1.0
HM-03" Hotleg | 687 ~155 28.7 26¢ 1.0
H-04 Hot leg : . 655 -89 30.2 29.5 0.5
H-05 . Hot leg 620 -2.7 30.9 30.5 0.5
H-06 Hot leg ' 582 : 0 - 30.1 30¢ 0
H-07 Hot leg 555 +2.3 303 305 04
C-01 Cold leg surge tank, vapor BF3-He 587 - -04 304 30 0
C-02 Cold leg surge tank 587 +0.3 30.0 30 -0
C-03 Cold leg 587 +2.8 30.5 30.5 0.5
C-04 Cold leg 544 +4.0 309 31.5 0.5
C-05 Cold leg - 504 : +34 30.5 30.5 0.5
C-06 Cold leg ' 482 +2.7 31.0 31.5 0.5
CM-077  Cold leg 460 +2.0 29.2 29 0.5
C-07 Cold leg _ : 460 +2.8 30.6 31.5 0.5

 

“Spei:lmens H-02 and C-02 in nonflowing salt,
DAl specimens exposed to molten salt unless noted otherwxse
' "‘Measured with a micrometer. :
4 Spec:men thickness measured with an optxml rmcroscope after specimen mounted for microstructural examination.
€Uneven surface.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Modified Hastelloy N.
ORNL-DWG 71-8059A
10 I
NCL-20

_ | - | 460°C COLDEST
85 — SPECIMENS, ———
& p— — |
g __o..-—l——"-—-o'—:‘-—-‘,..i ® ®
Lt
o
=
<
o
Q
-
S

-

685°C HOTTEST e«
| SPECIMENS .= \
-15 - i I _l ) *

o 2000 4000 6000 8000 10,000 12,000 14,000 16,000 18,000 20,000
" ' TIME OF OPERATION (hr)

Fig. 3. Weight changes of Hastelloy N specimens from NCL-20 exposed to NaBF ;—NaF (92-8 mole %) as a function of
time and temperature.

 
ORNL-DWG 7T1-8061R

AIR- ,
l- COOLED .IH HEATED AND INSULATED
SECTION

450

 

 

&é 500 _
X &
E 550 W
~ =]
o g
z T
g THERMAL CONVECTION LOOP NCL-20 600
5 | :
- | z
g_ o 3784 hr - : €50
g.l ¢ 6059 hr
a 8387 hr
4 11,883 hr N 700
8 14,300 hr
m 19,300 hr
=20
0 10 20 30 40 50 60 70 80 90 100
DISTANCE FROM HOTTEST PORTION OF THE UPPER CROSSOVER (in.)
1. | I e e o
i il e T 1
UPPER COLD LEG BEND LOWER BEND HOT LEG™
CROSSOVER {VERTICAL) CROSSOVER : (VERTICAL)

Fig. 4. Weight changes from Hastelloy N specimens in NCL-20 exposed to NaBF ;—NaF (92-8 mole %) as a function of
position and time.

by bombardment by the electron beam of the SEM).! 3 A view of the hot leg specimen at three different
magnifications is shown in Fig. 7. The surface has a semipolished appearance. There are many areas which
have been undermined with labyrinths of holes and channels. The stereo pairs (Fig. 8) provide a

three-dimensional view of the holes and grooves. All areas of Figs. 7 and 8 were composed primarily of '

nickel and molybdenum. The concentrations of iron and chromium were extremely low. The
semiquantitative estimates of concentrations are 71% Ni, 29% Mo, 0.3% Fe, and <0.1% Cr. There were a
few areas such as those in Fig. 9 that contained large amounts of debris. The debris is mostly iron,
presumably an oxide. Microprobe analysis'® of specimens after 11,900 hr exposure showed no detectable
concentration gradients.

Selected views of the cold leg specimen are shown in Fig. 10. The photos suggest that there has been
both dissolution and deposition on this surface. The photos were taken in the area of a plateau on the
surface. The grains in this plateau have a flat truncated appearance on top, which suggests that this plateau
is part of the original specimen surface that has been only lightly etched. The stereo pairs in Fig. 11 show

that the rest of the specimen surface is lower than this plateau. Thus it appears that there has been

considerable dissolution of the surface. The concéntrations of elements on this surface were 65% Ni, 25%
Fe, 0.2% Cr, and 9% Mo. Microprobe analysis of the specimen after 11,900 hr exposure showed that the
deposit on the surface was rich in Fe (>25 wt %) and Ni (>64 wt %), but contained very little Cr or Mo.

 

15. Analysis performed by L. D. Hulett of the Analytical Chemistry Division. Quantitative analysis nuiabers are subject
to errors as high as 20%. '
16. Performed by H. Mateer, T. J. Henson, and R. S. Crouse of the Metals and Ceramics Division.

 
"

 

 

 

There is also evidence of deposifion on the cold leg specimen surface. Figure 12 was taken from an area
remote from the plateau. Many of the grains appear to have cubic symmetry. Figures 12z and 12b show
possible growth steps and grainé with cubic symmetry. One would not expect the grains of the original alloy
substrate to have this much cubic symmetry. The concentrations of elements from the area of Fig 12 are
65% Ni, 23% Fe, 0.5% Cr, and 11% Mo, | | ’

Y8771

 

.~ STANDPIPE

& -STWIN
|
I

 

  

/ BALL VALVES

 

 

 

 

 

 

 

< ~CLAMSHELL
rEATERS

    
   
 

 

 
 
 
 

 

INSULATION

  

CORROSION
SPECIMENS -

    

 

 

 

 

 

hasi

 

 

 

 

 

 

 

DumP
TANK

 

 

 

 

 

 

 

 

 

Fig. 5. MSRP natural circulation loop and salt sampler.

 
 

 

 

10

§ PHOTO Y-107939 ; PHOTO Y-107341

 

15001 .

 

10,003 in.

0.007 INCHES
500X

16,005 n.

 

 

e
10.007 in.

 

  

.... " = *a L . £

c%“1g 6. Optical mlcrographs of Hastelloy N specimens from NCL-20 exposed to NaBF;—NaF (92-—8 mole %). (@) 11,900 hr,
685°C weight loss 6.2 mg/cm etched with glyceria regia; (b) 11,900 hr, 460° C, weight gain 2.8 mglcm as polished; (¢) 19,300 hr,
685°C, welght loss 144 mglcm etched with glyceria regia; (d), 19, 300 hr, 460°C, weight gain 4.0 mg/cm as polished.

 
 

 

o

 

 

11

PHOTO Y-115072

PHOTO Y-115071

 

Fig. 7. Scanning electron micrographs of hot leg specimen from NCL-20. (@) 5000X; (b) 1000X; (c) 500X.

 

 
 

12

>
O
-
o
I
a

115068

PHOTO Y-115069

"=

-115073

PHOTO Y

-115070

PHOTO Y

 

 

; (b) 2000X.

. (@) 5000X

from NCL-20

Fig. 8. Scanning electron micrograph stereo pairs of hot leg specimen
 

 

 

i
{
i

 

 

13

 

   

PHOTO Y-115074

-
¥

PHOTO Y-115075

»
-

 

Fig. 9. Scanning electron micrographs of hot leg specimens from NCI-20. (2) 5000X ; (b) 2000X.

 

 

 
 

 

14

PHOTO Y-115079

 

PHOTO Y-115080

 

Fig. 10. Scanning electron micrographs of éold leg specimens from NCL-20. (2) 10,000X ; (b) 5000X; (c) 2000X.

»

w

 
 

 

 

C

y

 

15

HOTO Y-115082

PHOTO Y-115083 BEAY PHOTO Y-115086
WETnR

 

Fig. 11. Scanning electron micrograph stereo pairs of cold leg specimen from NCL-20. (z) 1000X ; (7) 2000X

 

 
 

i
|
i
i

 

 

 

|
|
t
|
t
{
i

 

16

! PHOTO Y-115087

PHOTO Y-115088]

 

. .

 

 
 

 

.

17

4. DISCUSSION

Temperature-gradient mass transfer, schematically presented in Fig. 13, was seen in this test. As noted in
Fig. 4 and the various nficroéiéiifis, we detected material loss in the hot portions of the loop, material
deposition in the cold portion, and a point of little change (balance point) at about 580 or 590°C. Qur
microprobe analyses showed that iron and chromium were selectively removed in the hot zone. If iron and
chromium had been completely removed, the remaining alloy would have contained 80.5% Ni and 19.5%
Mo. Analysis showed 71% Ni and 29% Mo; thus some nickel may also have been removed. In the cold leg,
assuming that the molybdenum concentration in the alloy had not changed, it appeared that iron and nickel
deposited. We also showed that deposition was not entirely a function of temperature, since the weight
gains throughout the cold leg varied only a small amount. Thus our picture of mass transfer in this system is
that chromium, iron, and smaller amounts of nickel were removed from the hot leg; iron and nickel
deposited in the cold leg, and chromium remained in the salt. 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 the least-noble constituent is chromium. Iron is more noble than chromium but less
than nickel and molybdenum.

In this work, we removed the least-noble constituents, and under conditions where all constituents were
removed (formation of Ni, Mo, Fe, and Cr fluorides), the fluorides formed by the noble elements, such as
nickel, were so unstable that they were quickly reduced, resulting in iron and nickel deposition in the cold
leg.

ORNL-DWG 67-68B00R

 

 

HOT SECTION —— —— — —

DIFFUSION TO SURFACE

 

SOLUTE ESCAPE THROUGH
NEAR-SURFACE LIQUID LAYER

 

—_— - DIFFUSION INTO BULK LIQUID

‘ - =

: TRANSPORT
TO COLD PORTION OF SYSTEM

 

 

COLD SECTION — -
— — { X SUPERSATURATION — ————

 

     
 

* NUCLEATION - T T
¢ GROWTH TO STABLE CRYSTAL SIZE

o —OR . - _

SUPERSATURATION AND DIF FUSION
THROUGH LIQUID

 

- ONMETALLIC WALL — —— - -
OR DIFFUSION INTO WALL

Fig. 13. Temperature-gradient mass transfer,

 
18

A large part of the corrosion that occurs in fluoroborate salt is due to impurities, particularly moisture.
One of the proposed reactions involves water and fluoroborate as indicated below:

H,O + NaBF, = NaBF;0H + HF. | )

The reactions between HF and the metal components of Hastelloy N can be represented by

2HF + M = MF, + H,, | S 2

where M = Cr, Fe, Ni, or Mo. In our system the most predominant expression of Eq. (2) is, therefore,
6HF + 2Cr + 6NaF = 2Na;CrFg + 3H,, | (2a)

although nickel and iron also react with HF. _

The effect of HF in a salt system can be quite virulent. A previous Hastelloy N loop that circulated
equimolar NaF-ZrF, salt saturated with HF operated only 200 hr before it became plugged with nickel
corrosion product.’ 7 Attack of the Hastelloy N in this case was uniform throughout the loop.

We noted increased attack when we had impurity inleakages of air and moisture.

It is also thought that the hydroxide compound from Eq. (1) may dissociate in the following manner:

NaBF;0H = NaBF,0 + HEF. | 3
Combining reactions (1), (2a), and (3) we get

6NaF + 3NaBF, + 3H,0 + 2Cr = 3NaBF, 0 + 2Na,CrF, + 3H,. | (@)
Reactions (2a) and (3) can be combined tb obtain

6NaF + 6NaBF; OH + 2Cr = 6NaBF, O + 2Na,CrF¢ + 3H,. - 5)

Removal of hydrogen from the system by diffusion would drive reactions (4) and (5) to the right. ~

In fluoride salt systems, the temperature-gradient mass transfer process begins with material removal at
all temperatures, and later, as the corrosion product concentration increases, deposition begins in the colder
sections. We assume that the mass transfer in our system occurred in the same manner as described above,
with primarily chromium removal with some iron dissolution also. Thus we can explain that the dissolution
and deposition that occurred in the cold leg (as seen in the SEM) stemmed from material removal all over
the loop before equilibrium was attained at the lowest temperature point. After equilibrium, the chromium
then was returned to the walls in the cold part of the systéin. '

Since the weight losses of the hottest specimen are functions of time, they can be fitted to an equation
of the type

W=atb,

 

17. MSR Program Semiannu. Progr. Rep. Feb. 28, 1962, ORNL-3282, p. 72.

 
 

e i et e’

o

19

where
W = weight loss in milligrams per square centimeter of surface area,
a = a constant,
t = time in hours, and

b = a kinetic time constant.

For asolid-state diffusion-controlled process, b = ', and for a dissolution-controlled process b = 1 (ref. 18).
Calculations showed that for the first 11,900 hra = 2 X 107* and b = 1.1. The almost straight-line
dependence of weight loss with time (b = 1.1) also points to a fairly general type of attack, as opposed to
selective removal of only one element. The salt analysis showed evidence of reaction (1) in that an ionic
species of iron was removed from the salt while the concentration of chromium ions increased. Thus in
keeping with our other findings it would appear that more iron deposited than was removed.

Subsurface voids are often formed in alloys exposed to molten salts, and we see them in some of our
hot leg specimens. The formation of these voids is generally initiated by the selective oxidation of an alloy
constituent (usually chromium) along exposed surfaces through oxidation-reduction reactions with
impurities or constituents of the molten fluoride mixture. As the surface is depleted in chromium (or the
least-noble constituent), chromium from the interior diffuses down the concentration gradient to the
surface. Since diffusion occurs by a vacancy process and in this particular situation is essentially
unidirectional, it is possible to build up an excess number of vacancies in the metal. These precipitate in
areas of disregistry, and voids tend to agglomerate and grow in size with increasing time and/or
temperature. Examinations have demonstrated that the subsurface voids are not interconnected with each
other or with the surface. Voids of this same type have also been developed in Inconel by high-temperature
oxidation tests and high-temperature vacuum tests in which chromium is selectively removed.'? Voids
similar to these have also been developed in copper-brass diffusion couples and by the dezincification of
brass.2® All of these phenomena arise from the so-called Kirkendall effect; that is, solute atoms of a given
type diffuse out at a faster rate than other atoms comprising the crystal lattice can diffuse in to fill
vacancies that result from the outward diffusion. In our case the attack was selective toward iron and
chromium, and still voids were seen. Thus void formation does not mean that only one element has been
removed from the alloy matrix.

5. CONCLUSIONS

1. Temperature-gradient mass transfer of Hastelloy N, exposed to a sodium fluoroborate mixture of
NaBF;—8 mole % NaF in a thermal convection loop system that operated at temperatures of 687°C
maximum and 438°C minimum for 19,930 hr, does occur. The maximum calculated corrosion rate (based
on uniform removal) was 0.29 mil/year.

2. Material removal (chromium and iron) that resulted in voids occurred in the hot leg while iron and
nickel deposited in the cold section with chromium ions remaining in the salt.

3. Some of the deposits in the cold leg appeared to have cubic symmetry.

4. Impurities in the loop system (probably stemmmg from air inleakage) increased the mass transfer
rate.

5. Uve:all, the compatibility in this systein was quite good.

 

18. 1. W. Koger, C’orroszon and Mass Trensfer Characteristics of NaBF4-NaF (92-8 Mole ?’ ) in Hastelloy N,
ORNL-TM-3866, October 1972, pp. 66—-68. .

19. A. DeS. Brasunas, “Sub-Surface Porosity Developed in Sound Metals During ngh-Temperature Corrosion,” Metals
Progr. 62(6), 88 (1952).

20. R. W. Balluffi and B. H. Alexander, “Development of Porosity by Unequal Diffusion in Substitutional Solutions,”
SEP-83, Sylvania Electric Products (February 1952).

 
 

 

 

 
 

Hy

(3) Central Research Library
ORNL - Y-12 Technical Library
Document Reference Section -
(10) Laboratory Records Department
Laboratory Records, ORNL RC

ORNL Patent Office
G. M. Adamson, Jr.
C. F. Baes

~C. E. Bamberger

E. Beall

G. Bohlmann

B. Briggs

Cantor

Compere
Cook

Cunningham

Gilpatrick
Grimes

S.
E.
R.
S.
E.
W.
F.
J.
J.
J.
L.
J.
D.
J.
L.
W.
A. G. Grindell
W.

P

L.
H.
L.
E.
M.
H.
R.
R.
E.
H.
0.
R.
G.
0.
N.

Haubenreich

 (3) M. R_Hill

W. R. Huntley
H. Inouye
P. R. Kasten

BABCOCK & WILCOX COMPANY P. O. Box 1260, Lynchburg, VA 24505

B. Mong

BLACK AND VEATCH, P. O. Box 8405, Kansas City, MO 64114

C. B. Deering

BRYON JACKSON PUMP, P. O. Box 2017, Los Angeles, CA 90054 _

G. C. Clasby

 

INTERNAL DISTRIBUTION

(79 copies)

" EXTERNAL DISTRIBUTION

(24 copies}

ORNL-TM-4271

(5) J. W. Koger

J. Lawrence
L. Lotts

S. Lundy

N. Lyon

G. MacPherson
E. MacPherson
R. Martin
W. M

E.

J. McHargue
A. McLain
McNabb

E. McNeese
S. Meyer

B. Parker
Patriarca

M. Perry

W. Rosenthal
C. Savage
L. Scott

H. Shaffer
M. Slaughter
P. Smith
A. Strehlow
E. Thoma
B. Trauger
M. Weinberg
R. Weir

C. White:

V.

E.

A.
T.
R.
H.
R.
w.
R.
H.
C.

H.
B.
L.
A.
R.
P.

A.
M.
H.
J.

J.

G.
G.
R.
R.
D.
A.
J.

1.

L. V. Wilson

CABOT CORPORATION, STELLITE DlVISION 1020 Park Ave., Kokomo, IN 46901

T. K. Roche

21

 
 

 

 

 

22

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, Huntington, WV 25720
J. M. Martin

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

USAEC, DIVISION OF REACTOR DEVELOPMENT AND TECHNOLOGY, Washmgton 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 Washington, DC 20545
A. Giambusso :
USAEC, RDT SITE REPRESENTATIVES, 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, Oak Rldge TN 37830
(2)

W

(