ORNL-TM-12925.txt

 

 

_ Rmm |
N N Cl | .
3 =25 mwaME
g Mmm DH«RW
M m m — by by
~ Smo
o mAc

O
5
NO
mmm
o

 
 

 

 

| This report has been reproduced directly from the best available copy.

Available to DOE and DOE contractors from the Office of Scientific and Techni-
cal Information, P.O. Box 62, Oak Ridge, TN 37831; prices available from (615)
876-8401, FTS 626-8401.

Available to the public from the National Technical information Service, U.S.
Department of Commerce, 5285 Port Royal Rd., Springfield, VA 22161,

 

 

 

This report was prepared as an account of work sponsored by an agency of
the United States Government. Neither the United States Government nor any
agency thereof, nor any of 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 dis-
closed, or represents that its use would not infringe privately owned rights.
Reference herein to any specific commercial product, process, or service by
trade name, trademark, manufacturer, or otherwise, does not necessarily consti-
tute or imply its endorsement, recommendation, or favoring by the United States
Government or any agency thereof. The views and opinions of authors
expressed herein do not necessarily state or refloct those of the United States
Government or any agency thereof.

 

 

 

 
 

DISCLAIMER

Portions of this document may be illegible
in electronic image products. Images are
produced from the best available original
document.

 

 

 

 
ORNL/TM-12925/R1

MATERIALS CONSIDERATIONS FOR MOLTEN
SALT ACCELERATOR-BASED PLUTONIUM
CONVERSION SYSTEMS

J. R. DiStefano, J. H. DeVan, J. R. Keiser,
R. L. Klueh, and W. P. Eatherly

Date Published: March 1995

Prepared by the
OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37831-6285
managed by
MARTIN MARIETTA ENERGY SYSTEMS, INC.
for the
U.S. Department of Energy
under contract DE-AC05-840R21400

DWTRIBUTION OF THIS DOCUMENT 1S UNLIMITED g«fi

DISTRIBUTION CF Ti#@ DOOUMENT 1S UNLIMITED MAS TE R

 

 
 
 

 

TABLE OF CONTENTS
‘ .
Page
. LISTOFFIGURES ............... e e e e e \
LISTOFTABLES ... ittt i e e e vii
INTRODUCTION ...t ettt ettt et ettt e e e e e 1
SELECTION OF CONTAINMENT MATERIALS .. ... ... ... ... ... 2
CORROSION BY FLUORIDE MIXTURES ......... .. i, 8
TRANSMUTATION EFFECTS . ... ettt e e e e 11
FISSION PRODUCT EFFECTS ........... e e e e 12
RADIATION DAMAGE PROCESSES .. ... i i e 15
Nickel-Base ALlOys . ..........c. ittt it 16
Alternative Materials for Molten Salt Operations .. ......................... 20
Refractory Metals and Alloys . ... ... i i i 20
X GRAPHITE . ... . e e e e e et 22
Radiation Damage .......... ... it i, 22
Salt and Gas Entrapment . ... ... ... ... . it e e 23
' SUMMARY AND RECOMMENDATIONS .. ... .outieiineteaienanaenns, 23
REFERENCES . ... e et e et e e et 25

i

 

 
LIST OF FIGURES
Figure Page
* 1 Oxidation states of actinides in LiF-BeF,-ThF,-UF, ..................... 9

2  Hastelloy N used in experimental molten salt reactor showed intergranular
cracks when tensile specimens were strained to failure . .. ... ... ... .. ... 13

3 Effect of niobium in modified Hastelloy N on grain boundary cracking when
exposed in salt-CryTe, + CrsTeg for 250 h at 700°C ... ... ... ... ... ... 14

 

4  Extent of tellurium embrittlement of Hastelloy N is strongly affected by
oxidation potential of salt ......... ... ... ... . .. L i, 15

S Ductility behavior of the austenitic superalloys D25, D68, Inconel 706,
and Nimonic PE-16 irradiated in Experimental Breeder Reactor II up to
T x L10P D/CII? « o oottt e e e e 17

6  Variation of fracture strain with strain rate for Hastelloy N specimens
irradiated in the Molten Salt Reactor Experiment and tested at 650°C ...... 18

7  The effect of niobium additions on the creep-rupture behavior of
irradiated Hastelloy N .. ... .. . o 19

 

 
 

 

Table

LIST OF TABLES

Corrosion of 2% Cr-1 Mo steel in thermal convection loop tests (ref. 1) ... ..

Corrosion of low-alloy steels by Pb-Bi eutectic in thermal convection
Jooptests (ref. 1) ... .. . i e e

Standard free energies of formation of fluorides in a molten salt system . . . . . .

Operating conditions of stainless steel thermal convection loops involving
LiF-BeF,-based molten salts ............ ... ... ... ... ... ... ......

Nominal composition of standard and modified Hastelloy N . . ....... ... ..

vii

 

 
 

 

-

 

 

MATERIALS CONSIDERATIONS FOR MOLTEN SALT ACCELERATOR-BASED
PLUTONIUM CONVERSION SYSTEMS'

J. R. DiStefano, J. H. DeVan, J. R. Keiser.
R. L. Klueh, and W. P. Eatherly

ABSTRACT

Accelerator-driven transmutation technology (ADTT) refers to a concept
for a system that uses a blanket assembly driven by a source of neutrons
produced when high-energy protons from an accelerator strike a heavy metal
target. One application for such a system is called Accelerator-Based
Plutonium Conversion, or ABC. Currently, the version of this concept being
proposed by the Los Alamos National Laboratory features a liquid lead target
material and a blanket fuel of molten fluorides that contain plutonium. Thus,
the materials to be used in such a system must have, in addition to adequate
mechanical strength, corrosion resistance to molten lead, corrosion resistance
to molten fluoride salts, and resistance to radiation damage.

In this report the corrosion properties of iquid lead and the LiF-BeF,
molten salt system are reviewed in the context of candidate maternals for the
above application. Background information has been drawn from extensive
past studies. The system operating temperature, type of protective
environment, and oxidation potential of the salt are shown to be critical
design considerations. Factors such as the generation of fission products and
transmutation of salt components also significantly affect corrosion behavior,
and procedures for inhibiting their effects are discussed. In view of the
potential for extreme conditions relative to neutron fluxes and energies that
can occur in an ADTT, a knowledge of radiation effects is a most important
factor. Present information for potential materials selections is summarized.

 

INTRODUCTION

A technology for the conversion of highly reactive spent fuel and nuclear waste to
non-radioactive materials or materials with much shorter half-lives has been proposed for
development. In a concept proposed by the Los Alamos National Laboratory (LANL), the
system uses a high flux of neutrons to bombard the material to be converted. The neutron
source is an external accelerator that delivers protons to a target surrounded by a "blanket”
of fissile material. This concept is a part of what is referred to as accelerator-driven

transmutation technology (ADTT). In the application of this technology to accelerator-based

 

"Research sponsored by the Office of Fissile Material Disposition, U.S. Department of
Energy, under contract DE-AC05-840R21400 with Martin Marietta Energy Systems, Inc.

 

 

 
 

 

 

2

conversion of plutonium, or ABC. the nuclear fuel is carried by a molten fluoride mixture
containing LiF-BeF..

The ABC design as presently envisioned uses a flowing lead target for generating
spallation neutrons upon being bombarded by the accelerator proton beam. The lead
temperature ranges from 400 to 500°C except at the proton beam interface where it can rise
to 1000 to 1100°C. Since materials appropriate for containing the molten lead (Fe- or
refractory-metal-based alloys) are not necessarily suitable for containing the molten fluoride
fuel salt (Ni-based alloys), different materials either in contact with each other or separated by
a helium or NaK space may be required. The containment material for the liquid lead target
will see the highest neutron flux in the system and, therefore, its resistance to radiation
damage will be critical. Because of its complexity, the target and containment material
subsystem probably cannot be annealed or repaired and, therefore, will have to be replaced
periodically.

Fused fluoride mixtures containing PuF; are being considered as the blanket material for
the accelerator-based system. Because of their high boiling points, these mixtures can be
contained at low pressures even at relatively high operating temperatures. Their chemical and
physical properties impart additional advantages such as stability under irradiation and
chemical compatibility with a variety of containment and neutron moderator materials. The
reference salt mixtures under consideration for ABC systems are modelled after mixtures that
were developed for the Molten Salt Reactor (MSR) Program, conducted at the Oak Ridge
National Laboratory (ORNL) during the 1950s to early 1970s. These mixtures were based on
the eutectic LiF-BeF, system containing small percentages of UF, and, for some concepts,
larger percentages of ThF,. This report reviews the materials developments underlying the
MSR Program and relates them to the materials requirements for the ABC PuF;-containing
blanket. The selection of structural and moderator materials is discussed in terms of their
chemical compatibility with the fluoride fuel mixtures and their neutron radiation damage
characteristics. The report then addresses the ABC materials issues and needs in the

implementation of a molten fluoride blanket.

SELECTION OF CONTAINMENT MATERIALS

Target. Selection of a structural material for containment of flowing liquid lead is limited

by the relatively high temperature at the proton beam interface. In a flowing system

 

 
 

3
containing a temperature differential (AT). dissolution and mass transfer of material from high
to low temperature is an important concérn. For example, thermal convection loops made of
low-alloy steel plugged during exposure to Pb-2.5% Mg or Pb-Bi eutectic in 1000 to 2000 h at
a hot leg temperature of ~550°C as shown in Tables 1 and 2 from ref. 1. However, in pure
lead, a 2¥4 Cr-1 Mo steel loop was operated for over 27,000 h at this temperature with a AT
of approximately 100°C. At higher temperatures (800, 300°C AT), the refractory metals
niobium and tantalum showed no mass transfer, while other conventional alloys showed smalil
to heavy amounts.” Thus, it would appear that selection of a material for containment of
liquid lead may be restricted to a refractory metal or alloy unless the maximum temperature
can be limited to 700°C or lower.

If a refractory metal is chosen, special care must be taken to exclude oxygen and nitrogen
from the system, especially in the case of niobium and tantalum or their alloys. Both of these
materials, in contrast with molybdenum and tungsten, readily pick up interstitial elements that
can result in severe embrittlement. Thus, deoxidation of the lead would be an important
requirement. Furthermore, purifying and maintaining low levels of impurities in an inert gas
(helium) or alkali metal (NaK) environment to prevent this type of embrittlement will be a
formidable task.

At the detection limits of analytical methods (~1 ppb), the activities of nitrogen and
oxygen impurities in helium are far above those needed to form oxides of tantalum or
niobium. Because of the obvious contamination problems, very few experiments have been
conducted to evaluate the reaction rates of refractory metals in helium. The earliest attempts
to operate niobium and Nb-1 Zr corrosion loops made use of helium as a protective
environment, but all fell short of protecting the loop components from contamination.> In the
mid-1960s, Battelle Northwest Laboratories* conducted a test of refractory metals using a
conventional alloy recirculating helium loop (< 1 ppm total impurities). Specimens of
molybdenum, tungsten, niobium, and tantalum were exposed for a total of 504 h at 1150°C.
Results were summarized by Battelle as follows:

Niobium and tantalum readily pick up impurities from (helium) loop atmospheres

pure enough to permit the evaporation of superalioys, while molybdenum and

tungsten remain unaffected. The mechanical properties of Nb and Ta can be

drastically altered by contaminant absorption during "clean run" conditions (i.e.,
without purposeful impurity additions).

 
 

 

‘paydde usaq o4

wu pey judtpesd dooy -a)-a,]
3q 01 punuj jeudjew fnyg adid
(Ww-£-Z1) "uFsp paddngg o9
18 UOIS0LI0D 219438 “UONN|OS

 

 

OIUL 03 JoUu pIRoM NG pappe 17 JUON SLL'T 01§ 05% Apana FW-qd 8t
‘061 = .08t
- al)L8 O) poseanur |y ~adid
(ww-£z) ‘u-gg paddnig auON o1l 0Er 0P ondana SN-qd 0t
"189YJ3A0
0) 391 104 pasneds ainje]
1mod Jaipe umop Jnys doo §9/°12 ;Jeio], SHi 0$s Bay
"89] ploa
w udd y3ys  uoisosoo oN FE9's Siy 0ss
$L6'6 oty 08§
001°6 09t 058
I/ |8 180 uofInjos
QU 0F 10U PINOM )G pappe 17 8 wdd ¢zz 950t 01¢ 0s¢ qd 8L
S{ICWY S2ANIPPY Yy Ly UN xR [e1ow pinbyy 'oN doo]
Do ‘L

 

(1 "3o1) 51591 doo| UOII2ALOD [ewLIDY) UT [99)s O [-1D *AT JO UOISOLIO) T 9[qe].

 

 
 

“Kjeowydefoferaw paunuexs doo,

 

"(3a1 pjoo ur papsasur diys J7 woyj ag ~ SWose  (papnnxaoo)
wdiw) ‘uidd apqissod uoisodoo oN 35 ploo ur papuadsns diyd 7 618L 00% 049 IZ e§  [991s uoqle) +£C1
_ SN 08¢ (parpy-1s)
“utejuiewt 03 ynagip -ouod ], “uidd Jo uoIs01I03 ON 0222 00¢ 00$ [L €T [291s uoqie) 121
JW 0st  (papuaixano)
'dosp 11, 1y3Ns -widd Jo uoisosi0d oN v18¥ 00S 059 IL SE€1 {9318 uoqie) +CC1
W 0sg  (pspnnxaod)
‘s80} aatippe oN 'uidd 10 uoIs0sI0d ON STPE osPy 009 Z Sy 9915 UoQaeD) 811
S 0ss  (papnixa00)
"$80] aanippe ON "uidd 10 UoISO1I00 ON 8/¢6 008 089 IZ 001  [991s uogde) <611
. 3N osE
‘wdd (g 03 dozp 1), wdns widd 1o uoisossod oN 9967 00  0$9 1L St %1 4ojo1)) 0Z1
"£TI-L1T ‘€11 sdoo| 3N ose
ur pasn g Kund-yadipy ssof sanppe oN uidd 1o uolso1109 oN 8¢S 50 0%9 1Z 001 Wi 40jo1) +£11
‘nidd 10 uoiso1I00 aN 1966 00z 00t JUON %] £OjoID) «L11

‘9] 304 Ul UOISOLIND
212438 "4)-2:] Jo pasodwond @nyg -adid (ww-£-Z1) "u-go padinig I8 00§  $79 QUON %l £0j01) «0F1

‘Fa1 104 Wl UOISOLIOD
UIA0S  O%d Larssod pue 1)-2 jo pasodwoo Inid adid (ww

 

 

-L°T1) "ul-6p padingg "uonnjos oy of jou pinom Jnq pappe 17 0061 00t 0ssS 3N ost %l Aojoi) +C6
YLy UN  xep wdd jelate ‘oN doop
SYJeWway SAIAMNPPY
Do ‘L

 

 

 

 

 
 

6

The Japanese>S investigated the oxidation rates of TZM (Mo-0.5% Ti-0.1% Zr).
molybdenum, and Nb-1 Zr in commercially purified (> 99.995%) helium and the same helium
containing 13 ppm oxygen. The Nb-1 Zr specimens had Nb,N surface layers after exposure to
the commercial-grade helium but formed NbO and NbO, films when exposed to the He plus
13 ppm oxygen. Concentrations of oxygen in the Nb-1Zr reached 1% in either environment.
However, molybdenum showed no measurable effects in commercial-grade helium but did
show evaporation of molybdenum oxide above 800°C in the He plus 13 ppm oxygen. The
oxygen content and microhardness of the TZM increased in both environments.

The extent of oxygen contamination of Nb-1 Zr in vacuum environments is determined by
the system pressure, temperature, and time. In a recent study,’ the rate of contamination was
reported to be proportional to the first power of pressure and an exponential function of
temperature at P02 < 103 Pa (10° Torr) and T > 773 K (500°C). During long-term
operation, significant oxygen increases (hundreds to thousands of ppm) can occur unless the
partial pressure of oxygen is maintained at 10® Pa (10 Torr) or lower. On the other hand,
the weight increase in molybdenum was low compared with Nb-1 Zr at a system pressure of
3.6 x 10° Pa (2.7 x 107 Torr) [ref. 8].

Although the compatibility of niobium and tantalum with alkali metals can be seriously
degraded by oxygen in the refractory metal,” oxygen in sodium, potassium, or NaK has also
been reported to cause corrosion of niobium, tantalum, and to a lesser extent, molybde,num.g
Cold-trapped sodium (~40 ppm oxygen) results in a much higher weight removal rate of
niobium or tantalum compéred with hot-trapped (~10 ppm oxygen) sodium {or NaK).

Based on corrosion considerations, molybdenum would probably be a satisfactory material
for containment of the liquid lead target; however, fabricability would be a significant issue.
The Nb-1 Zr alloy would provide good resistance to lead and has excellent fabricability; but
extreme care would have to be taken to avoid contamination and/or corrosion by system
environments. For some heat exchanger coolants, e.g., helium, significant contamination
probably could not be avoided.

Blanket. The development of systems that incorporate a circulating molten fluoride fuel
is predicated on the availability of a construction material that will contain the salt over long
time periods and also afford useful mechanical properties. The container material must also
be resistant to air oxidation, be easily formed and welded into relatively complicated shapes,

and be metallurgically stable over a wide temperature range. For initial MSR studies, several

 

 
 

7

commercially available. high-temperature alloy systems were evaluated with respect to the
aforementioned requirements.’®! As a result of these initial studies. Inconel 600, a nickel-
based alloy containing 15% Cr and 7% Fe, was found to afford the best combination of
desired properties and was utilized as a container for the construction of the Aircraft Reactor
Experiment.’> Corrosion rates encountered with this alloy at 700°C and above, however,
were excessive for long-term use with most fluoride fuel systems. Ultilizing experience gained
in corrosion testing of commercial alloys, an alloy development program was conducted to
provide an advanced container material for MSR systems. The reference alloy system was
composed of nickel with a primary strengthening addition of 15 to 20% Mo. This composition
afforded exceptional resistance to salt attack but lacked sufficient mechanical strength and
oxidation resistance at the desired service temperature. To augment these latter properties,
additions of various solid-solution alloying agents were evaluated, among them Cr, Al, Ti, Nb,
Fe, V, and W. An optimum alloy composition was selected on the basis of parallel
investigations of the mechanical and corrosion properties which were imparted by each of
these additions. The composition best suited to reactor use was determined to be within the
range 15 to 17% Mo, 6 to 8% Ci, 4 to 6% Fe, 0.04 to 0.08% C, balance Ni. This alloy was
commercialized under the trade name Hastelloy N.

A 7.4-MW test reactor, the Molten Salt Reactor Experiment (MSRE), was constructed of
Hastelloy N and became critical in 1965. The reactor operated successfully for 4 years and
verified the excellent corrosion resistance of the alloy to the UF,-containing LiF-BeF, salt
mixture. However, operation of the MSRE revealed two potential problem areas that were of
concern in the utilization of the Hastelloy N alloy for more advanced Molten Sait Breeder
Reactors (MSBRs), where greater concentrations of fission products and greater neutron
fluences would be encountered. Radiation damage, in the form of helium embrittlement from
(n,et) reactions, was found to have reduced the creep ductility of the Hastelloy N based on
post-operation examinations. Also, grain boundaries of the Hastelloy N directly exposed to
the fuel salt were shown in post-operation tensile tests to have been embrittled to depths of
0.15 to 0.25 mm. Subsequent studies showed the embrittlement to be associated with the
presence of fission product tellurium on grain boundaries that intersected with the salt-
exposed surfaces. Accordingly, the earlier Hastelloy N alloy development program was

extended to improve the alloy’s radiation damage characteristics and resistance to penetration

by sulfur-like fission products, such as tellurium. (These alloy modifications are discussed in

 

 

 
 

8

later sections of this report.) In addition, compatibility tests were conducted to re-examine
the possibility of using iron-based alloys as containment materials for the MSBR. These alloys
are more resistant to tellurium penetration and also generate less helium from (n,a) reactions
than nickel-based alloys. However, their use would severely limit the oxidation potential of
the salt (in effect, the fluoride ion activity), whereas the Hastelloy N composition could resist
attack by the reference MSBR salt components under virtually all sustainable reactor
operating conditions. As will be discussed, these same materials concerns are key issues in
selecting the ADTT salt blanket structural material but with the condition that PuF; has

replaced UF, in the salt mixture.

CORROSION BY FLUORIDE MIXTURES

The selection of a structural material for containment of molten fluoride salts begins with
a consideration of the reduction-oxidation (redox) potentials of the component elements of
the material with respect to components of the salt. Table 3 lists the standard free energies of
formation, per gram atom of fluorine, of fluoride compounds at 527 and 727°C associated
‘with salt constituents and metals commonly used in high-temperature alloys. It is evident that
the major salt components, LiF and BeF,, are considerably more stable than any of the
fluorides of the structural metals listed, so they are unreactive with the structural metals. The
corrosion of a variety of structural alloys by reference MSR salt mixtures has been found’® to
result from a combination of reactions involving impurities in the salt, 2HF + M = MF, + H,
(M = Ni, G, Fe) and XF, + Cr = GiF, + X (X = Nj, Fe), and reactions involving fuel
components, e.g., 2Cr + 2UF, = CtF, + 2UF;. In all cases, the corrosion product fluorides
are readily dissolved by the molten salt mixture. The impurity reactions can be minimized by
maintaining low-impurity concentrations in the salt and clean alloy surfaces. However, the
reaction with UF, is intrinsic and depends on the redox potential of the salt, which can be
described in terms of the ratio of activities of quadrivalent-to-trivalent uranium ions.
Although this redox potential can be adjusted to make the salt less oxidizing (for example, by
exposing the salt to beryllium or chromium metal), the activity of UF, must be maintained
high enough to avoid its disproportionation and subsequent alloying of metallic uranium with
the containment alloy. The relationship between the U** and U*> ratios and the effective
fluoride ion oxidation potential is given in Fig. 1. The cross-hatched area in the figure

indicates the ratios that were generally studied in corrosion tests and that were maintained in

 

 
 

9

Table 3. Standard free energies of formation of
¢ fluorides in a molten salt system

 

 

; Fluoride —AG; (800 K) ~AG; (1000 K)
kj/g-atom ~ {kcal/g-atom kj/g-atom (kcal/g-atom
of fluorine of fluorine) of fluorine of fluorine)
LiF 539.7 128.9 5204 124.3
ThF, 466.4 1114 451.8 107.3
RuF; 453.0 108.2 4375 104.5
BeF, 451.3 107.8 437.1 104.4
UF; 436.7 104.3 422.5 100.9
UF* 415.7 99.3 399.8 | 95.5
TiF, 3525 84.2 337.9 80.7
CrF, 334.1 79.8 320.7 76.6
MoF, 310.7 74.2 302.7 723
Nb; 308.6 73.7 297.3 71.0
FeF, 297.3 71.0 284.3 67.9
NiF, 266.7 63.7 251.6 - 60.1

 

ORNL-DWG 73-6975

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P 4+
" v 100
. 2 ]
§ Poo* ~ 108
'lb. o - ]
o 4 ] 6
i ue+ — 10
& _
" 44 -— 104 o
5 Th : Pt "é
-— 5
£ WW ///////////////////AV/////////////////)(////////////////J% 102 %
B -1 Pod* -
& 1o
3+ —
e — 10
u _
o O
Po® v 10~
-2 ThY Pyl
Th Po U Py

ACTINIDE ELEMENT

Fig. 1. Oxidation states of actinides in LiF-BeF,-ThF,-UF,.

 

 
 

 

10

the MSRE. Relative to the reactions listed above, assuming equilibrium is attained. those
structural metals with less stable fluorides, such as nickel and molybdenum. are less susceptible
to oxidation, while those with the more negative free energies, such as chromium, are more
prone to oxidation. The chemistry of Hastelloy N was tailored to afford compatibility with
UF,-containing salt mixtures by simply allowing the salt to "equilibrate” with the alloy. In any
heat-transfer system, true equilibrium between the salt and the alloy is precluded by
temperature differences in the system. Nevertheless, in a Hastelloy N closed-loop system
operating at 600 to 700°C, a steady-state U**/U*? ratio is reached within a few hundred
hours** and has been shown experimentally to be in the range 100 to 350. Corrosion proceeds
by the selective oxidation of chromium at the hotter loop surfaces and the reduction and
deposition of chromium at the cooler loop surfaces. In the case of the MSR fuel salts, the
resultant maximum corrosion rate of Hastelloy N measured in extensive loop testing was
below 4 pm/year, and a similar rate was measured on surveillance specimens in the MSRE."
If a 300-series stainless steel (18% Cr-10% Ni-balance iron) is exposed to UF,-containing
salts under the same closed-system conditions as described above, corrosion again is
manifested by the selective removal of chromium from hotter loop surfaces with concomitant

chromium deposition at cooler surfaces. However, because of the much higher chromium

 

activity in the stainless steel compared to Hastelloy N, the extent of chromium mass transport
is considerably greater than for Hastelloy N. Table 4 lists the operating conditions for two
stainless steel thermal convection loops that circulated LiF-BeF, salts containing 1 and

0.3 mol % UF,, respectively.'® The 304L loop operated successfully for 9 years at a maximum
temperature of 688°C and exhibited an average corrosion rate equivalent to 21.8 pm/year,
based on uniform wall reduction. However, corrosion was actually manifested by subsurface
void formation which, 1n the case of the loop wall, reached depths up to 1.2 mm. A second
loop was constructed of type 316 stainless steel and operated for 4490 h. The corrosion rate
of this loop at 650°C was slightly less than the 304L loop, but corrosion again was manifested
by subsurface voids, in this case to a depth of 76 pym. A second 316 stainless steel loop,
whose conditions are also shown in Table 4, was operated with an LiF-BeF, salt that
contained no uranium, so that the oxidation potential of the salt could be lowered by buftering
with metallic beryllium without concern for the disproportionation of UF;. Before adding
beryllium, the corrosion rate of the loop averaged 8 pm/year over 25,000 h at 650°C. After
contacting the salt with a small beryllium rod, the corrosion rate was lowered to 2 pm/year

over a 2000-h period.”

 
 

11

Table 4. Operating conditions of stainless steel thermal
convection loops involving LiF-BeF,-based molten salts

 

 

Salt Maximum
Loop composition temperature AT Time
material , (mol %) (°C) (°C) (h)
304L SS LiF-BeF,-ZrF,-ThF,-UF, 688 100 83,520
(70-23-5-1-1)
316 SS LiF-BeF,-ThF;-UF, 650 110 4,490
(68-20-11.7-0.3)
316 SS LiF-BeF, 650 125 25,100
(66-34)

 

Based on the above observations, the corrosion characteristics of an LiF-BeF, salt
containing PuF; in place of UF,, i.e., the reference ABC blanket salt in the LANL concept,
would depend on the redox potential set for the salt system. If the potential is maintained in
the same range as for the MSR Program, the corrosion rate of the containment material, in
terms of chromium and iron oxidation, should be unaffected by the replacement. However,
as shown in Fig. 1, the oxidation potential at which the plutonium activity associated with
PuF; becomes unity is lower than the potential at which the activity of uranium associated
with UF; becomes unity. Accordingly, it may be possible to maintain a lower oxidation
potential in the PuF,-containing salt than in the case of the uranium-containing salt without
encountering corrosion from alloying of metallic Pu with the containment. If this can be
demonstrated, then the corrosion rate of an austenitic stainless steel may be in an acceptable

range to be used as the structural material for the ABC blanket.

TRANSMUTATION EFFECTS

Depending on the neutron energies and fluxes in the ABC blanket, transmutation
reactions may affect the salt chemistry and, accordingly, the redox potential of the salt.” The

overall reactions can be represented as:

| °LiF + n -~ ‘He + TF

iF+n-%He + TF + n

 

 

 
 

 

 

12
BeF.+n-~2n +2*He +2F
1BeF., + n -~ ’He + *He + 2 F followed by
‘He - °Li. and °Li + F ~ °LiF
YF + n -~ N - “He followed by

5N - 1°0O" + B

The TF and F produced by the transmutation of lithium and beryllium, respectively, will act to
increase the oxidation potential of the salt. The effect is similar to that encountered in the
fission of UF, in the MSR salt, where effectively one fluorine atom per fission is free to
oxidize or corrode the container (i.e., is not tied up with fission products). In the MSRE, the
small concentration of UF; in the fuel provided a reductant to buffer any oxidants resulting
from transmutation, and a redox couple similar to UF,/UF; could be added to the ABC
blanket to control the transmutation effect. The selection and use of redox additions has
beén discussed in a recent paper.”” The transmutation of fluoride ions ultimately results in the
formation of O ions in the salt; however, whatever reductant is used to reduce fluorine to

fluoride should also reduce the oxygen.!”

FISSION PRODUCT EFFECTS

As previously described, the corrosion rate of Hastelloy N measured during operation of
the MSRE was relatively low and of the same order as that for forced-convection loops
operating under similar temperatures. However, Hastelloy N tensile specimens that were
tested to failure after they had been exposed to the MSRE fuel salt showed shallow surface
cracks along the gage length at grain boundaries that connected to the salt-exposed surfaces
(see Fig. 2) [ref. 20]. These cracks generally extended to depths of about 0.13 mm but were
as deep as 0.33 mm in parts removed from the salt plenum region of the pump bowl. Because
this grain boundary embrittlement was found in the heat exchanger tubes where the neutron
flux was insignificant, as well as in samples exposed in the reactor core, it was concluded this
cracking was not due to radiation effects. Controlled dissolution of samples detected a
number of fission products within the material; tellurium was the fission product found at the
highest concentration.” Subsequent tests showed that grain boundary embrittlement could be

caused in Hastelloy N samples when tellurtum was applied to the sample surface. Since the

 

 
 

 

 

13

ORNL-DWG 95-5767

Hastelloy N exposed -
21,000 h to salt 3

containing tellurium

oo 00 fll?l
'soos o -3 8020

§§

Hastelloy N exposed
21,000 b to vapor

above salt containing _
tellurium - .

 

Fig. 2. Hastelloy N used in experimental molten salt
reactor showed intergranular cracks when tensile specimens
were strained to failure.

depth of cracking observed in the MSRE would not have been acceptable when extrapolated
to the 30-yeér design life of an MSBR, a program was undertaken to identify ways to prevent
tellurium embrittlement of Hastelloy N.

Investigation of the problem was approached in several ways. In-reactor experiments
were used to evaluate the embrittlement resistance of various alloys. In another study, over
60 alloys were electroplated with tellurtum and annealed for extended times before being
tensile tested. No cracks formed on a number of alloys including stainless steels and
nickel-base alloys containing more than 15% chromium. Most heats of Hastelloy N developed
cracks, but modifications that contained 2% niobium showed better crack resistance than
standard Hastelloy N.”* Other tests were developed that involved exposure of alloy samples to
molten fluoride salts containing telluride compounds.” Telluride compounds used included
Li,Te, LiTe,, Cr,Te,, Cr;Te,, CrsTeg, and Ni,Te,, and cracking resulted in Hastelloy N samples
exposed to salt solutions of all but the first of these compounds. Modifications of Hastelloy N
were independently being developed to address the problem of radiation embrittlement, so
samples of alloys modified with different combinations of titanium, niobium, and chromium
- were also exposed in the salt solutions containing the tellurides.

Results indicated that niobium as an alloying agent reduced embrittlement of Hastelloy N,
but neither chromium nor niobium exhibited as strong an effect when titanium was included.
The extent of the effect of niobium on reducing grain boundary embrittlement is shown in

Fig. 3 (ref. 20).

 

 

 
 

 

 

14

ORNL-DWG 95-5768

 

(CRACK FREQUENCY) 1 (AVERAQGE DEPTH)

 

 

— 3

2 4
Q . ? 3 4 3

 

 

NIOBIUM CONCENTRATION (wt %)

Fig. 3. Effect of niobium in modified
Hastelloy N on grain boundary cracking
when exposed in salt-Cr;Te, + CrsTeg for
250 h at 700°C (refs. 21 and 22).

Although the effort to prevent tellurium embrittlement of grain boundaries was primarily
directed at alloying modifications of Hastelloy N, some work was done to determine if
chemical changes in the salt could also alter tellurium behavior. On the basis of
electrochemical studies conducted at ORNL? that indicated tellurium could exist as a telluride
rather than as elemental tellurium in a relatively reducing melt, standard Hastelloy N samples
were exposed to molten salt solutions containing Cr;Te, with variable oxidation potentials.
The oxidation potential of the salt solution was described in terms of the U**/U*? ratio (see
Fig. 1), and this ratio was varied between 10 and 300. Results of these studies are shown in
Fig. 4 where it is apparent that below an oxidation potential of approximately 70, tellurium
embrittlement of grain boundaries is greatly reduced.? This finding may be particularly
significant for the PuF;-containing ABC blanket salt which, as discussed above, appears more

amenable to operation at lower redox potentials than the UF,-containing MSR fuel salts.

 

 

 

 
 

 

 

 

15
ORNL-DWG 77-4680
* T L r
900 REDUCING OXIDIZING —
B /.—_
. —
CRACKING 600 -
PARAMETER N _
[FREQUENCY (cmi )
X AVG. DEPTH (um)] — —
300 [~ -
| J _
o l-t—t——T1" | | 1

 

 

 

10 20 40 70 100 200 400
SALT OXIDATION POTENTIAL [UllV)/ullIl)]

Fig. 4. Extent of tellurium embrittlement of Hastelloy N is strongly affected
by oxidation potential of salt (ref. 21).

RADIATION DAMAGE PROCESSES

When an alloy is exposed to a high-energy neutron field, neutrons collide with atoms in
the alloy and knock them from lattice positions. The displaced atoms displace other atoms to
form a "damage cascade” of vacancies (vacant lattice sites) and ari equal number of interstitials
(displaced atoms wedged in the interstices between atoms). Although most interstitials and
vacancies recombine, it is the disposition of the unrecombined defects that determines the
radiation effects. Interstitials and vacancies not eliminated by recombination diffuse to
"sinks” (surfaces, grain boundaries, dislocations, and existing voids) where they are absorbed.
Defect clusters also form: clusters of interstitials evolve into dislocation loops; vacancy
clusters develop into voids (small cavities).

For irradiation temperatures below about 0.35 T,,, where T, is the absolute melting
temperature of the irradiated material, interstitials are mobile relative to vacancies.
Interstitials combine to form dislocation loops that increase the strength (“irradiation
hardening”) and decrease ductility. Above 0.35 T, vacancies become mobile and agglomerate
into voids; the irradiated alloy swells. Excess vacancies cause enhanced diffusion of elements

i the alloy, which can promote precipitation of new phases that can harden and embrittle the

 

 

 
 

 

16

alloy. At irradiation temperatures above about 0.6T . defect clusters are unstable, and the
high equilibrium vacancy concentration and rapid diffusion enhance vacancy-interstitial
annihilation, such that displacement damage has little effect on properties.

Besides causing displacement damage, an atom of the alloy can absorb a neutron and
undergo a nuclear transmutation reaction. Transmutation produces a new atom (usually
another metal) and hydrogen and/or helium gas atoms inside the alloy. New metal atoms have
little effect on properties because relatively small quantities form. Hydrogen also has little
effect for it will diffuse from the alloy. Helium, however, is insoluble in metals and alloys and
will be incorporated into bubbles or voids that form within the metal.

Nickel-Base Alloys. Nickel-base alloys of the type proposed for an ABC structure to
operate with a molten salt working fluid for a blanket are relatively resistant to void
swelling. 2% The neutron spectrum in the blanket region will be highly thermalized, and a
high cross section for (n,a) reactions exists between thermal neutrons and ¥Ni (68% of
natural nickel is *Ni). Furthermore, high-energy neutrons (up to 200 MeV) will be generated
in the target. If the neutron spectrum of the blanket contains high-energy neutrons
(- > 6 MeV), they could contribute to further helium generation by (n,«) reactions. At

elevated temperatures, helium can diffuse to grain boundaries and embrittle an alloy. Helium

 

can also cause swelling by bubble formation at elevated temperatures. Finally, in certain
nickel alloys, irradiation enhances precipitate formation, which can cause properties
degradation.

Several high-nickel superalloys were studied in the Liquid Metal Reactor (LMR)
Program.>* When irradiated in the Experimental Breeder Reactor II (EBR-II) at 465 to
575°C and tensile tested at 234 to 785°C, ductility decreased to zero at test temperatures of
600 to 800°C, depending on the alloy (see Fig. 5) [ref. 23]. The temperature where zero
ductility occurred decreased with increasing neutron fluence. This ductility loss was attributed
to an irradiation-enhanced formation of a grain boundary precipitate. However, the alloys
were dropped from the LMR Program before the mechanism was completely understood.
Embrittlement was not due to elevated-temperature helium embrittlement because little
helium is produced by irradiation in EBR-II.

Since the ABC is proposed to have a molten salt working fluid, it is of interest to examine
the radiation-damage experience in the MSRE.? Nickel-based Hastelloy N alloy was used in

the MSRE and demonstrated excellent performance. The major difference between the

 

 
 

17

ORNL-DWG 95-5769

 

 

S0 [ T ! ! ! T T
40 — O D25ST —
® D25STA
D D68ST
B D68STA
- © INCONEL 706ST
® 30 — - A PE16ST —
e
=
-l
: —
2 20— T—0. —_
10 — —
0 | 1 L

 

 

 

o 100 200 300 400 500 600 700 800 900
TEST TEMPERATURE (°C)

Fig. 5. Ductility behavior of the austenitic superalloys D25, D68, Inconel 706, an
Nimonic PE-16 irradiated in Experimental Breeder Reactor IT up to 7.1 x 10% n/cm®.

MSRE and the ABC involves the neutron fluxes. The high-energy (> 0.8 MeV) neutron flux
in MSRE was only 1.2 x 10" n/cm?s, and the thermal flux was only 6.5 x 10" n/cm®-s

(ref. 15). Much larger fast and thermal fluxes are expected in the ABC, and neutrons with
much higher energy will be present in the ABC spectrum.

During irradiation at 650°C in the MSRE, no swelling or hardening was observed in
Hastelloy N. However, helium formed in the Hastelloy N by (n,«) reactions of thermal
“neutrons with nickel and residual boron, and the helium caused elevated-temperature helium
embrittlement, as exhibited by a loss of ductility measured in tensile and creep tests at 650°C.
Fracture strains depended on strain rate and thermal fluence (see Fig. 6) [ref. 22]. Creep-
rupture strains of less than 0.1% were observed for the highest fluence, compared to strains
that exceeded 16% at all strain rates for specimens aged in fluoride salt for 15,189 h at
650°C. Embrittlement was attributed to helium gas bubbles that were observed on grain

boundaries.?

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

18
ORNL-DWG 68-117772R
5, . : - 4
' ' .9 25.8 .
o o224
‘e e ® 196
3t : e 7.0 . .
O C © 3.7
1L : S .
= MSRE SURVEILLANCE '
HEAT VESSEL CORE
2t S06S ° . -
5085 o . o . B
TH 0 il
| ERM;L/:;LL:ENCE 1350 2610 1 32402 9411020 "o H'i
"t TIME AT ; e
i S50°C (hr) 14,000 20800 4800 15300 ; ; i H'?i
: Vh RRE
E ! i i;:' i ‘.;
| ‘ H l%%” . '11 -Hl:?l
S 9 — : e e
> | B T
o} boe i . .
$ . -l il ’ *
T g ' : -
» L ! I}U Hig!
W | il
2., ! il ] i
U o ! :
- . HE Y I
e l | :}1‘; “ | :i.
¢ | ( il i
5 ] il
t o [ i . .
Ll o H i L
5 — . : —_
il i ¥ e
";!l f]“ | | AVERAGE, ORR IRRADIATION, fl}“:
a [;J et —~ 2-'5!1020 nfem? ——r]
el ,[:'i |l ' . f'i Coy e =|
=”iE' e ] ':Eil : : vy
3 ‘Hv L1 4 ; L
i ] | T, , T
T O ;HI }sl i
2 -. ngzilb i : 4” lf‘ ‘ il”.;'
bi}lf' o T
1 * |1}l creee | TensiE Gl i,
o ‘?Fil -~ TESTS @ TESTS —
'F'|i f‘t | .‘ Pt : li‘] 1| ::1;
> fl‘t-’" ’fli "! l i :||H ' 'llH Py
1o—3 1g-e 10~4 100 10 102 10°
STRAIN RATE {%/hr)

Fig. 6. Variation of fracture strain with strain rate for Hastelloy
N specimens irradiated in the Molten Salt Reactor Experiment and
tested at 650°.C. .

 

An alloy development program demonstrated that the radiation resistance of the
Hastelloy N could be improved by alloying.” A niobium addition increased creep strength and
fracture strain (see Fig. 7). Titanium aiso proved beneficial. This work resulted in the
modified Hastelloy N composition in Table 5, where it is compared to the standard
Hastelloy N. The damage resistance of this new alloy to the conditions in an ABC would

need to be determined.

 

 

 

 

 
 

 

   

19

ORNL-DWG 95-5770

 

 

 

 

5
N CONTENT (%)
CREEP 20
STRAIN
100 b UNIRRAD.
creep ‘O e50°C
TEsT | ros*c
0l 760°C
o 1 ] 1 } i
0 1 2 3 a 5

Nb CONTENT (%)

Fig. 7. The effect of niobium additions. on the
creep-rupture behavior of irradiated Hastelloy N
(ref. 25).

Table 5. Nominal composition of standard and modified Hastelloy N

 

 

Standard Maodified
Ni bal bal
Mo 16 12
Cr 7 7
Fe 4 0.5
Mn 0.5 - 0.2
Si 0.5 0.1
C 0.05 : 0.05
» Ti+Nb — 2
Re — 0.01

 

 

 

 

 
 

 

 

20

Alternative Materials for Molten Salt Operations. As an alternative to high-nickel allovs,
certain improved stainless steels may offer enhanced irradiation resistance for an ABC. Less
helium will be generated by thermal neutrons in an austenitic stainless steel than in a nickel
alloy because there is less nickel present. Although conventional austenitic stainless steels
(e.g., type 316) are extremely sensitive to helium embrittlement, embrittlement-resistant
austenitic stainless steels were developed in the U.S. fusion reactor program.”* Helium
embrittlement was minimized by the precipitation of a fine dispersion of carbide particles in
the matrix and grain boundaries. Fine helium bubbles formed on the particles, and because of
their large number and small size, bubbles did not grow to the minimum size required for
embrittiement. After irradiation at 600°C to 17 dpa and 1750 appm He in the mixed-
spectrum High Flux Isotope Reactor (HFIR), little change in tensile ductility was observed in
one of the earlier alloys. Behavior at > 600°C has not been determined. More advanced
alloys have also been developed but have not been irradiated.

Refractory Metals and Alloys. Since the molten salt blanket region of the ABC will be
moderated by graphite and the maximum temperature will be below 700°C, the use of a
nickel-base alloy or an austenitic stainless steel should be possible. However, the structural
material for the target region will be exposed to much different temperatures and irradiation
conditions: temperatures as high as 1100°C are possible in a radiation field with neutron
energies up to 20 MeV. The use of conventional nickel-base alloys or stainless steels will not
be possible, and a refractory metal alloy based on Nb, Ta, Mo, or W will probably have to be
used. Limited irradiation-effects data are available for these materials; the most extensive data
are available for niobium and molybdenum and their alloys. The following discussion, which is
based on a review of the subject by Wiffen,>’ will emphasize those alloy systems.

Irradiation of refractory metals and their alloys to ~ 3 to 6 x 102 n/cm® (< 0.1 MeV) has
been conducted in fast reactors, such as EBR-I1, and in mixed-spectrum reactors, such as
HFIR. Under such irradiation conditions, the cross section for (n,e) reactions is small.
Therefore, little helium forms and it had no effect on the properties determined for these
irradiations. However, for the high-energy neutrons (up to 20 MeV) in the target region of
the ABC, much larger helium concentrations will be generated, and helium could greatly
affect the properties.

Of the refractory metals of interest, niobium and niobium alloys have been studied most

extensively. Swelling of over 2% with a peak near 600°C was observed for unalloyed niobium

 

 

 

 
 

 

21

-

over the range 400 to 1050°C (the range over which data are available). The Nb-1 Zr allov.
the only niobium alloy investigated extensively. is more radiation resistant than unalloyed
niobium. Swelling occurs over a much narrower temperature range with a peak near 800°C.
Ion bombardment simulation of neutron irradiation indicated that both dissolved oxygen and
helium can enhance swelling. Therefore. the high helium concentrations that could form in an
ABC from transmutation reactions with the high-energy neutrons of the spectrum could be
important to the swelling behavior.

Irradiation of niobium and Nb-1 Zr in the HFIR from 50 to 950°C caused an increase in
yield stress and ultimate tensile strength, with the magnitude of the strength increase
decreasing with increasing temperature. Total elongations were 5 to 10%, but uniform
elongation decreased to zero at the lower temperatures. Fractures of specimens tested at
room temperature and the irradiation temperatures remained ductile for all test temperatures,
indicating that the ductile-to-brittle transition temperature (DBTT) remained above room
temperature—at least in slow strain-rate tensile tests. (Like all body-centered cubic metals and
alloys, the refractory metals and their alloys show a DBTT that can be increased by neutron
irradiation.) Limited studies of specimens with accelerator-injected helium indicated that
elevated-temperature helium embrittlement is pbssible for helium contents of 200 appm or
greater. |

Molybdenum, Mo-0.5% Ti, and TZM alloy have been studied, and similar results were
observed for the three materials. Swelling occurred over the range 450 to 1300°C with peaks
of up to 4% observed at 600 to 800°C. Where direct comparisons were made, swelling in the
molybdenum alloys exceeded that for unalloyed molybdenum. The most important irradiation
effect involved hardening, which was evident in a slow three-point bend test. Hardening for
rradiation temperatures in the range 400 to 1000° C caused the DBTT (determined by the
bend test) to increase to as high as 650°C for irradiation at 585°C and to be above room
temperature for all irradiation conditions.

Fewer data are available for tantalum and tungsten and their alloys. Tantalum showed a
similarity to niobium with a peak swelling temperature around 600°C. Tungsten additions to
tantalum reduced swelling. Tensile data for irradiated tantalum and tantalum alloys are scarce
(no data above 650°C), and the general conclusion was that the behavior of irradiated

tantalum and T-111 alloy is similar to niobium and Nb-1 Zr (ref. 27).

 

 
 

 

272

e s

There are even fewer data on tungsten and tungsten alloys than there are for tantalum.
A swelling peak was observed at around 800°C for tungsten. On the other hand. little or no
swelling was observed for W-Re alloys (5 to 25% Re). The effects of irradiation on the
mechanical properties of tungsten were similar to the behavior of molybdenum: the limited

data indicated that hardening caused a large shift in DBTT.

GRAPHITE

Graphite in an MSR serves no structural purpose other than to define the flow patterns
for the salt and, of course, to support its own weight. The requirements on the graphite
derive almost entirely from nuclear considerations—namely, stability of the material against
radiation-induced distortion, resistance to penetration by the fuel-bearing salt, and non-
trapping of Xe within the free-volume of the graphite pores.”

Radiation Damage. Neutron bombardment produces atomic displacements which result in
vacancy-interstitial pairs. Vacancies migrate to dislocations and collapse; interstitials combine
mto ordered interlaminar planes which assume a graphite structure with stacking faults. The
result is that individual crystallites contract in the a-axis direction and expand in the c-axis .
direction. This is reflected in an initial macroscopic bulk contraction of the polycrystalline
graphite, followed by bulk expansion as damage proceeds. Eventually, the expansion induces -
macroscopic cracking of the graphite until mechanical disintegration occurs. During this I
progression, all of the thermo-mechanical properties are affected, and radiation-activated
creep occurs, resembling in some detail its high-temperature thermal analogue.

During reactor operation, internal stresses develop within the graphite due to temperature
and flux gradients. These internal stresses are relaxed due to the radiation creep. At
shutdown, the thermal gradients disappear and a reversed stress field immediately appears in
the graphite. The most detailed creep data exist on the U.S. and German graphites for the
high-temperature gas-cooled reactor, but these graphites, because of their coarse granularity
and large pore size, are unsatisfactory for molten salt applications. Conversely, the IG-100
graphite grade produced in Japan for their gas-cooled reactor program is generically much
closer to the MSR requirements.”® However, the dominant term in the constitutive equations
describing mechanical properties in the presence of neutron and thermal gradients is the creep
term,” and the creep behavior of the fine-grained, low-porosity graphites needs further

evaluation.

 

 
 

 

23

The operating temperature of the ABC blanket. which is on the order of 600 to 700°C,
will be advantageous in terms of the graphite lifetime. It is precisely this temperature range
over which graphite 1s the most long-lived, 1.e.. the time interval during irradiation when after
initially contracting, graphite has expanded back to its original unirradiated volume. For
graphites of the type needed for ABC service, this fluence will be of the order of 25 to 30 x
10% nvt (E > 50 keV) [ref. 30]. |

Salt and Gas Entrapment. The unique character of the graphite required for an MSR
relates to the necessity in preventing the molten salt from "soaking" into the graphite.
Artificial graphites are generally fabricated from petroleum cokes and pitches or polymers
which, on heat treatment to high temperatures (> 2500°C), leave a relic carbon structure
capable of crystallization. By necessity, the evolving gases, predominantly hydrogen, must
escape the structure, and this produces a pore texture within the graphite. Completely sealing
these pores is impractical—the material will simply "blow up" due to internal gas pressure
developed during heat treatment. However, since the molten salts are non-wetting to graphite
and possess a high surface tension, it is only necessary to reduce the entrance pore diameter
to < 1 um to prevent salt intrusion. Fortunately, the industry is quite capable of producing
materials with the very small pores required. Also, in the last decade, isostatic molding has
become a standard forming technique. As a result, quality has been significantly improved,
and cost has decreased for materials of the type used in the MSRE some 20 years ago.

Gas entrapment in the graphite, specifically Xe' and Xe!¥’, was an additional problem in
the MSBR because of its deleterious effect on breeding uranium from thorium and the
resultant increase in the fuel doubling time. Fortunately, this problem should be of lesser

significance in a Pu-burner reactor.

SUMMARY AND RECOMMENDATIONS

Based on corrosion considerations, molybdenum would probably be the best choice as
containment material for a liquid lead target system operating at 1000 to 1100°C. However, if
neutronic or other considerations, e.g., radiation damage, eliminate its use, Nb-1 Zr could be
considered, but reactivity with other environments containing interstitial impurities could be
significant. Testing will have to be conducted to define these limitations as well as the

practicality of constructing such a system.

 

 
 

 

24

Corrostion studies and reactor operation have shown that Hastelloy N (both standard and
modified) has excellent resistance to molten fluoride salts. As a result of studies conducted
for the MSR Program from 1956 through 1976, an extensive corrosion data base exists on
both nickel- and iron-base alloys for UF,-containing salts. This data base provides a roadmap
for establishing the corrosion properties for PuF;-containing salts, but additional corrosion
testing under ABC conditions will be required to quantify corrosion rates of candidate
structural alloys.

Use of Hastelloy N as a containment material allows a wide latitude in possible system
operating parameters such as the temperature and salt redox potential. However, operating
experience in the MSRE did show that Hastelloy N was embrittled at elevated temperatures
due to radiation-induced internal helium generation, and shallow intergranular cracking was
associated with interactions with fission products (especially tellurium). Alloy development
studies showed that Hastelloy N modified with 1 to 2% niobium had better resistance to
irradiation embrittlement and to intergranular cracking by tellurium. It was also found that
the severity of cracking by tellurium could be diminished by lowering the redox potential of
the salt. However, since the spectrum of products from transmutation of plutonium or
actinide wastes would be different from that investigated in the MSR Program, corrosion
reactions specific to the ABC salt chemistries must be evaluated. Included in this evaluation
should be an assessment of lower salt redox potentials from the standpoint of allowing the use
of austenitic stainless steels as structural materials and establishing the potentials that must be
maintained to avoid alloying metallic plutonium with the containment.

The radiation fluence and spectrum in an ABC pose a serious challenge for any structural
alloy. For the part of the system containing molten salt, the combination of both high-energy
neutrons and thermal neutrons at high fluences introduces conditions that are different from
previous applications or previous irradiation-damage studies. Calculations of neutron energy
spectra and fluxes are required, so damage parameters can be determined and a better
assessment of the type of damage to be expected can be made. For the nickel-base alloys,
stability of potential structural alloys under high-energy neutron irradiation needs to be
determined (i.e., whether any irradiation-enhanced precipitation causes embrittlement, such as
that observed in fast reactors). The combination of irradiation-enhanced precipitation and
high helium concentrations must be evaluated for nickel-base alloys. A program to determine

this could be carried out in a mixed-spectrum reactor, such as HFIR. Alloy development

 

 
25

techniques used to improve resistance to helium embrittlement in austenitic stainless steels
’ may be applicable to modified Hastelloy N to develop a nickel-base alloy with properties
suitable for the molten salt environment of an ABC.

From the limited data available on refractory metal alloys, 1t appears that niobium or
tantalum would provide the best choice of a structural material for the target region, given the
indications of irradiation embrittlement in molybdenum and tungsten alloys. Again,
appropriate neutron energy spectra and fluxes are needed to determine how appropriate the
existing data are to the ABC. These data are also needed to determine the type of irradiation
tests required to qualify a refractory alloy for an ABC. Irradiation experiments would be
required to evaluate a chosen alloy over the appropriate operating conditions. None of the
presently available data have been obtained under conditions where large transmutation
helium concentrations are generated, such as might be expected for the target region of the
ABC due to the presence of the high-energy neutrons in the spectrum. Such data are
important to evaluate the propensity of the materials toward elevated-temperature helium
embrittlement and to determine whether helium can adversely affect swelling and increase
embrittlement due to irradiation hardening (i.e., affect the shift in DBTT).

Graphite has long been successfully employed in nuclear reactor applications, and
commercial graphites are available with adequate physical, corrosion, and structural properties

‘ for ABC service. Based on available data, it should be possible to estimate the lifetimes of
these graphites at ABC neutron fluences once the ABC design parameters have been
finalized.

In summary, the selection of materials for an ABC system will require trade-offs between
the system operating parameters and material properties in order to achieve adequate

resistance to lead and salt corrosion, fission product cracking, and radiation damage.

REFERENCES

1. A. J. Romano, C. J. Klamut, and D. H. Gurinsky, The Investigation of Container Materials
for Bi and Pb Alloys, Part 1. Thermal Convection Loops, BNL 811 (T-313), Brookhaven
National Laboratory, Upton, N.Y., July 1963.
2. J. V. Cathcart and W. D. Manly, "Technique For Corrosion Testing in Liquid Lead,"
* Corrosion 12, 43-47 (February 1956).

 

 
 

10.

11.

 

12.

13.
14.

15.

16.

26

E. A. Limoncelli, Use of Helium and Argon as Protective Environments for SNAP-50/SPUR
Program, TIM-824, Pratt-Whitney Aircraft. CANEL, Middletown. Conn., November 1965.
L. A. Charlot, R. A. Thiede, and R. E. Westerman, Corrosion of Super-Allovs and
Refractory Metals in High Temperature Flowing Helium, BNWL-SA-1137. Battelle
Northwest Laboratories, Richland, Wash., September 1967.

T. Noda, M. Okada, and R. Watanabe, "The Compatibility of First Wall Metallic Materials
with Impure Helium," J. Nucl. Mater. 85-86, 329-33 (1979).

T. Noda, M. Okada, and R. Watanabe, "Corrosion Behaviors of Iron-Base Alloy, Nickel-
Base Alloy and Retractory Metals in High Temperature Impure Helium Gas," J. Nucl. Sci.
Technol. 17(3), 191-203, Atomic Energy Society of Japan, 1980.

J. R. DiStefano and J. W. Hendricks, Oxidation/Corrosion Studies on Nb-1 Zr for Space
Reactor Applications, ORNL-6720, Martin Marietta Epergy Systems, Inc., Oak Ridge Natl.
Lab., August 1992.

H. Inouye, Contamination of Refractory Metals by Residual Gases in Vacuum Below 10°°
Torr, ORN1f3674, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1964.

J. R. DiStefano, "Review of Alkali Metal and Refractory Alloy Compatibility for Rankine
Cycle Applications," J. Mater. Eng. 11 (3), 215-25 (1989).

G. M. Adamson, R. S. Crouse, and W. D. Manly, Interim Report on Corrosion by Alkali-
Metal Fluorides, ORNL-2337, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab.,
1953.

G. M. Adamson, R. S. Crouse, and W. D. Manly, Interim Report on Corrosion by
Zirconium-Base Fluorides, ORNL-2338, Union Carbide Corp. Nuclear Div., Oak Ridge
Natl. Lab., 1961.

W. D. Manly et al,, Aircraft Reactor Experiment - Metallurgical Aspects, ORNL-2349,
Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1957.

W. D. Manly et al.,, Progr. Nucl. Energy, Ser. IV 2, 164 (1960).

J. H. DeVan and R. B. Evans, Corrosion Behavior of Reactor Materials in Fluoride Salt
Mixtures, ORNL/TM-328, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1962.
M. W. Rosenthal, P. N. Haubenreich, and R. B. Briggs, The Development of Molten Salt
Breeder Reactors, Vol. 11, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1972.
J. W. Koger, Alloy Compatibility with LiF-BeF, Containing ThF, and UF,,
ORNL/TM-4286, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1972.

 

 
27

 

17. J. R. Keiser, J. H. DeVan, and D. L. Manning. The Corrosion Resistance of Type 316
Stainless Steel to Li,BeF,, ORNL/TM-5782, Union Carbide Corp. Nuclear Div., Oak Ridge
Natl. Lab., 1977.

18. S. Cantor and W. R. Grimes, Nucl. Technol. 22. 120 (1974).

19. L. M. Toth, "Molten Fluoride Fuel Salt Chemistmry,” Invited Paper-Session XII, 1994
International Conference on Accelerator-Driven Transmutation Technologies and
Applications, Las Vegas, Nevada, July 1994.

20. H. E. McCoy, Jr., and B. McNabb, Intergranular Cracking of INOR-8 in the MSRE,
ORNL-4829, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., November 1972.

21. 1. R. Keiser, Status of Tellurium-Hastelloy N Studies in Molten Fluonde Salts,
ORNL/TM-6002, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1977.

22. H. E. McCoy, Jr., Status of Materials Development for Molten Salt Reactors,
ORNL/TM-5920, Union Carbide Corp. Nuclear Div., Oak Ridge Natl. Lab., 1978.

23. P. S. Sklad, R. E. Clausing, and E. E. Bloom, "Effects of Neutron Irradiation on
Microstructure and Mechanical Properties of Nimonic PE-16," pp. 139-55 in Jrradiation
Effects on the Microstructure and Properties of Metals, ASTM STP 611, American Society
for Testing and Materials, Philadelphia, 1976.

24. S. Vaidyanathan, T. Lauritzen, and W. L. Bell, "Irradiation Embrittlement in Some
Austenitic Superalloys,” pp. 619-35 in Effects of Radiation on Materials: Eleventh
Conference, ASTM STP 782, American Society for Testing and Materials, Philadelphia,
1982.

25. H. E. McCoy, Jr., An Evaluation of the Molten-Salt Reactor Experiment Hastelloy N
Surveillance Specimens - Third Group, ORNL/TM-2647, Union Carbide Corp. Nuclear
Div., Oak Ridge Natl. Lab., 1970.

26. P. J. Maziasz and D. N. Braski, "Modification of the Grain Boundary Microstructure of
the Austenitic PCA Stainless Steel to Improve Helium Embrittlement Resistance," J. Nucl.
Mater. 141-143, 973 (1986).

27. F. W. Wiften, "Effects of Irradiation on Properties of Refractory Alloys with Emphasis on
Space Power Reactor Applications,” pp. 252-72 in Refractory Alloy Technology for Space

Nuclear Power Applications, ed. R. H. Cooper and E. E. Hoffman, Technical Information
- Center, Office of Scientific and Technical Information, U.S. DOE, 1984.
28. T. Oku, M. Eto, and S. Ishiyama, J. Nucl. Mater. 172(1), 77 (1990).
. 29. B. T. Kelly, Physics of Graphite, Applied Science Publishers, London, 1981, p. 114.
30. T. D. Burchell and W. P. Eatherly, J. Nucl. Mater. 179-181, 205 (1991).

 

 
1-2.

4-5.

8-10.
11.

12.

13.

14.

15.
16-20.
21-25.
26-30.

52.

53.

. 54.

55.

56.

 

29

ORNL/TM-12925/R1

INTERNAL DISTRIBUTION
Central Research Library 31. T. A. Gabriel
Document Reference Section 32. U. Gat
Laboratory Records Department 33. S.R. Greene
Laboratory Records, ORNL-RC 34. H. W. Hayden
ORNL Patent Section 35. D. T. Ingersoll
M&C Records Office 36-39. 1. R. Keiser
B. R. Appleton 40-44. R. L. Klueh
E. E. Bloom 45. L. E. McNeese
R. H. Cooper 46-47. G. E. Michaels
B. S. Cowell 48. P. R. Rittenhouse
G. Del Cul 49. A F. Rowcliffe
J. H. DeVan 50. J. E. Rushton
J. R. DiStefano 51. L. M. Toth
W. P. Eatherly

EXTERNAL DISTRIBUTION

ARGONNE NATIONAL LABORATORY, 9700 South Cass. Ave., Building 360,
IPNS Division, Argonne, IL 60439

J. Carpenter

BARTHOLD & ASSOCIATES, INC., 132 Seven Oaks Drive, Knoxville,
TN 37922-3407

W. Barthold

BROOKHAVEN NATIONAL LABORATORY, Building 130, MS 5000, Upton,
NY 11973-5000

H. Ludwig

LAWRENCE BERKELY LABORATORY, Nuclear Science, 500-106, 1 Cyclotron
Rd., Berkely, CA 94720

L. Schroeder

LAWRENCE LIVERMORE NATIONAL LABORATORY, P.O. Box 808, L-036,
Livermore, CA 94551

J. N. Kass

 
 

30

57-67. LOS ALAMOS NATIONAL LABORATORY, LER/ADTT, P.O. Box 1663,
Los Alamos, NM 87545

E. Arthur (MS-H854)

C. Beard (MS-F607)

J. J. Buksa (MS-K551)

L. L. Daemen (MS-H&805)
J. W. Davidson (MS-F607)
R. A. Krakowski (MS-F607)
J. Park (MS-K551)

W. C. Sailor (MS-F607)

W. Sommer (MS-E546)

J. W. Toevs (MS-F628)

M. A. Williamson (MS-G730)

 

68. Monroe Wechsler, 106 Hunter Hill Place, Chapel Hill, NC 27514-9128

69. PACIFIC NORTHWEST LABORATORY, P.O. Box 999, MS P8-15, Richland,
WA 99352

F. Garner
70. PAUL SCHERRER INSTITUTE, CH-5232 Villigen-PSI, Switzerland

G. Bauer

 

71. SANDIA NATIONAL LABORATORIES, P.O. Box 5800, Albuquerque, NM 87185

-

R. J. Lipinski (MS-1145)
D. L. Mangan (MS-0656)

72-73. SANDIA NATIONAL LABORATORIES, P.O. Box 969, Livermore, CA 94551

M. Abrams (MS-9015)
C. A. Pura (MS-0969)

74. UNIVERSITY OF ILLINOIS, Nuclear Engineering Department, 103 S. Goodwin
Ave., Urbana, IL 61801

J. Stubbins

 

75. WESTINGHOUSE SAVANNAH RIVER CO,, P.O. Box 616, Aiken, SC 29808

M. Louthan

   

 
31

76-77. DOE, OAK RIDGE OPERATIONS OFFICE. P.O. Box 2001, Oak Ridge,
TN 37831-6269

E. E. Hoffman, National Materials Programs |
- Deputy Assistant Manager for Energy Research and Development

78-80. DOE, Office of Fissile Materials Disposition, MD-2, FORS/3F043,
1000 Independence Avenue, SW, Washington, D.C. 20585

H. R. Canter, Technical Director
A. L Cygelman
P. S. Gibson

81. DOE, OAK RIDGE OPERATIONS OFFICE, P.O. Box 2008, Oak Ridge,
TN 37831-6269

H. E. Clark, Fusion and Nuclear Technology Branch

82-83. DOE OFFICE OF SCIENTIFIC AND TECHNICAL INFORMATION,
P.O. Box 62, Oak Ridge, TN 37831

For distribution as shown in DOE/OSTI-4500, Distribution Category UC-704
Office of Defense Programs (Materials)

<