STUDY ON CHEMICAL INTERACTION OF FUEL CYCLE AND

MANAGEMENT

U3Si2 DISPERSOID AND ALUMINUM

MATRIX IN PLATE FUEL ELEMENTS AND KEYWORDS: plate
element, dispersoid,
fuel
bond

ITS INFLUENCE ON THE MECHANICAL strength

PROPERTIES
V. P. SINHA,a* D. KOHLI,a R. RAKESH,a P. V. THAKAR,a and A. KUMARb
a
Bhabha Atomic Research Centre, Nuclear Fuels Group, Metallic Fuels Division, Mumbai, India
b
Bhabha Atomic Research Centre, Nuclear Fuels Group, Mumbai, India

Received July 1, 2014

Accepted for Publication February 5, 2015
http://dx.doi.org/10.13182/NT14-59

Dispersion-type plate fuel elements are being fabri- examinations. During the course of study, it was observed
cated with U3Si2 dispersoid (prepared by an innovative that U3Si2 dispersoids in actual plate fuel elements were
powder processing route) in aluminum matrix and clad in enveloped by a different phase while the dispersoid
Al alloy for the modified core of the APSARA reactor by a of Y2O3 remained inert in the surrogate plate under a
standard picture frame technique followed by hot roll similar fabrication history. The study concludes that
bonding operation at the Bhabha Atomic Research Centre limited exposure of the actual fuel plate at 5008C for 5 h
Metallic Fuels Division. The fabrication regime allows results in improvement of bond strength mainly due to
the fuel elements to be exposed at 5008C for almost 5 h chemical interaction between fuel dispersoid and alumi-
(total duration including hot roll bonding and blister test num. The study also concludes that the tensile strength
operation). Therefore, it is expected that during hot roll and ductility of the fuel plates did not show any adverse
bonding and blister test operation, U3Si2 will chemically effects during dispersoid-matrix chemical interaction;
interact with aluminum and form an intermediate phase. however, the modulus of elasticity was found lower than
Hence, the chemical interaction behavior of fuel dis- the theoretically estimated value calculated by composite
persoid (U3Si2, prepared by powder metallurgy route) theory. The observations derived in the study are critical
and matrix (aluminum) in plate fuel elements and its effect from the viewpoint that a decrease in the elastic modulus
on mechanical properties is studied in the present paper. of the plate would adversely affect its flow-induced
Therefore, a comparative study between an actual vibration properties during reactor operation. It may also
plate fuel element (i.e., U3Si2 dispersed in aluminum be concluded that exposing the plate fuel elements at
matrix and with Al alloy clad) and a sandwich plate with 5008C for longer duration (i.e., 30 h) will result in
chemically inert material (i.e., Y2O3) as dispersed in excessive swelling because of the accelerated interaction
aluminum matrix with Al alloy clad was carried out. The between dispersoid and matrix, which will eventually
roll bonded samples were investigated through pull and deteriorate the desired properties.
peel tests, microhardness, tensile test, optical microscopy,
scanning electron microscopy, electron probe microana- Note: Some figures in this paper may be in color only in the electronic
lysis, and X-ray diffraction for various metallurgical version.

*E-mail: [email protected]

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 35

Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

I. INTRODUCTION (i.e., melting and casting) followed widely elsewhere in the

world.7–10
The inception of the Reduced Enrichment for Hence, the impetus was to investigate/characterize
Research and Test Reactor (RERTR) program by the the chemical interaction behavior of U3Si2 compound
U.S. Department of Energy in the late 1970s was mostly (prepared by powder metallurgy) with aluminum metal
focused on reducing the number of research reactors and powder under an out-of-pile test condition from the points
isotope production facilities operating on highly enriched of view of fuel design and operation.11,12 It is also
uranium (HEU)–based fuel (HEU: 235U $ 20 wt%) around worthwhile mentioning that the fabrication regime allows
the world.1–3 The objective of RERTR was primarily to the plate fuel element to be soaked at ,5008C for 5 h
discourage commerce in HEU and therefore make fissile (includes soaking time for hot rolling operation and blister
material more proliferation resistant for peaceful usage and testing). It is well understood from the literature that
societal benefit. However, in 2004 the RERTR program soaking fuel plates at 5008C for 5 h does favor chemical
was brought under the umbrella of the Global Threat interaction between U3Si2 compound and aluminum metal
Reduction Initiative, with greater comprehensive/broader powder and results in an intermediate phase product with
objectives.4,5 Therefore, in synergy with the global efforts measurable dimensions.13,14 It is also worthwhile men-
and also as part of the Department of Atomic Energy’s tioning that during flow-sheet development of plate-type
policy shift in accordance with the international stipulation fuel for the modified APSARA reactor, several rolling
on HEU (HEU: 235U . 85 wt%) fuel utilization, the temperatures were tested and characterized. However, it
APSARA research and test reactor at Trombay, Mumbai, is was observed that roll bonding carried out at a lower
undergoing a core conversion program from HEU-based temperature other than 5008C is insufficient to form
fuel to low-enriched uranium (LEU)–based fuel (LEU: bonding and the plates were having entrapped gas
235
U , 20 wt%). During the reactor core conversion pockets; hence, the optimized fabrication temperature
process the basic reactor design was not altered; however, was found to be 5008C. Therefore, keeping this aspect in
other attributes of fuel like the chemical composition of mind, the samples were not prepared with a lower heat
dispersoid, heavy metal loading in the plate, number of treatment history other than 5008C.
fuel plates in the assembly, water gap, standard and control It is also reported in the literature that formation of
fuel assembly configuration, etc., were largely changed/ intermediate phase product due to chemical interaction
modified. In the process the reactor power was also will lead to a progressive compositional variation and also
increased from 1 to 2 MW(thermal) and in turn has resulted result in a lower-density product as compared to the
in higher thermal, epithermal, and fast neutron flux. dispersoid phase, thereby adversely affecting the thermal
The fissile atom loading in the APSARA LEU core conductivity of fuel elements.15 Hence, minimization of
was compensated by replacing U-Al alloy (density: interaction product volume is always desired for better
6.8 g/cm3) with U3Si2 compound (density: 12.2 g/cm3) fuel performance. However, very limited literature is
and also increasing the heavy metal density in the fuel available in open access that correlates the effect of fuel
meat by almost six times (0.7 to 4.4 g/cm3). The major dispersoid and matrix chemical interaction on the
reason fuel designers chose the U-Si system over the U-Al mechanical properties of plate fuel elements. Hence, the
system was mainly because the compounds in the U-Si objective of this paper is to characterize the effect of
system have higher heavy metal density. The LEU plate chemical interaction between fuel dispersoid (U3Si2
fuel elements for the modified APSARA core with U3Si2 synthesized by powder metallurgy route) and matrix
dispersed in aluminum matrix and with Al alloy clad are (aluminum) on the mechanical properties of plate fuel
being fabricated by a standard picture frame technique elements by evaluating tensile test and bond strength
followed by the hot roll bonding process at the Bhabha results. In order to examine the effect on bond strength of
Atomic Research Centre (BARC) Metallic Fuels Division irrevocable soaking of fuel plate elements at 5008C for
(MFD). It is also worthwhile mentioning that an innovative 5 h, a comparative study between the actual fuel plate
powder processing route has been developed to synthesize (U3Si2 dispersed in aluminum matrix) and the sandwich
uranium silicide compound (i.e., U3Si2) and is successfully plate with inert surrogate meat (Y2O3 dispersed in
adopted in the regular production activity of fuel plates aluminum matrix) was carried out.
for the BARC modified APSARA core.6 The powder This paper also characterizes the chemically interacted
metallurgy route developed for the synthesis of U3Si2 product by different metallurgical tools such as optical
compound has distinct advantages such as excellent product microscopy, backscattered electron image by scanning
homogeneity without undergoing an additional step of heat electron microscopy (SEM), electron probe microanalysis
treatment or homogenization, lower impurity pickup, lower (EPMA), and X-ray diffraction (XRD). In addition, this
processing temperature, negligible liquid phase formation paper summarizes the results of microstructural character-
as compared to a conventional technique, and added ization of the intermediate phase formed due to different
intragranular porosity in the compound when compared heat treatment schedules/histories (i.e., 3008C, 4008C, and
with the product prepared by another conventional technique 5008C for 25, 50, and 100 h) by means of stereo

36 NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015

 Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

microscopy and XRD analysis. A detailed analysis on

mechanical property evaluation by peel and pull tests in a Cover Plate (AA6061)
universal tensile testing machine and microhardness
analysis for the actual fuel plate specimen and the
surrogate sandwich plate specimen under as-fabricated
conditions were also carried out and are discussed. Fuel meat compact

II. EXPERIMENTAL WORK Picture Frame (AA6061)

For clarity, the experimental work is divided into two

Cover Plate (AA6061)
sections. Section II.A briefly describes the process details
for fabricating the fuel plate and the surrogate sandwich
plate, while Sec. II.B describes various methodologies (a)
adopted to prepare different types of samples and also
different metallurgical characterization techniques employed Cover Plate (AA6061)
for their evaluation.

II.A. Fabrication of Plate Fuel Elements

Two different types of sandwich plates were
prepared. In the first type, called case-A, the plate was
prepared with U3Si2 dispersoid, while in the second type,
called case-B, Y2O3 was used as dispersoid; however, the
matrix and clad material remained the same for both
cases. The preparation of sandwich plates for both case-A Fuel meat compact (Al &
and case-B was carried out under similar fabrication dispersoid) Picture Frame (AA6061)
parameters. The dispersoid (U3Si2 prepared by powder
metallurgy for case-A and Y2O3 for case-B) in 42 vol % (b)
was first mixed with aluminum metal powder (58 vol %)
for 6 h in a cubical blender. The mixture was then Fig. 1. Picture framing of fuel meat: (a) schematic and (b)
compacted at 926.4-MPa pressure in a single-action actual.
hydraulic press to prepare a rectangular compact (i.e.,
meat). The compact was then picture framed (see Fig. 1)
in aluminum alloy frame AA6061 and welded all along longitudinal and transverse directions as per the standard
the edges to make the sandwich. The typical sandwich ASTM E8-04 (Ref. 16). The plates for case-A and case-B
dimension was 150 mm (length) 6 90 mm (width) 6 were also used to prepare samples for the pull and peel
9.6 mm (thickness). A detailed chemical analysis of tests as per the standards ASTM C633-01 (Ref. 17) and
AA6061 is shown in Table I. The sandwich was prepared ASTM D1876-01 (Ref. 18), respectively. Photographs of
with an opening at the tail end side to allow gaseous the tensile test specimen, pull test specimen, pull test
product release during hot rolling (see Fig. 2). The entire specimen mounted with sample holder, peel test specimen
operation was carried out inside a glove box train under with fixture, and the specimen after peel testing are shown
once-through flow of high-purity argon cover gas. in Fig. 3.
The sandwich was then subjected to the hot roll The tensile test for case-A and case-B was performed
bonding process where a gradual reduction in thickness in an Instron tensile testing machine under similar strain
was achieved (40% in each step) until the final thickness rates. The experiment was carried out in a well-ventilated
of 2.5 mm. These hot roll bonded plates were then area with the required number of air changes at 258C and
subjected to cold rolling operation where rolling was 60% relative humidity to ensure radioactivity contain-
performed in steps to achieve a final thickness of 1.5 mm. ment. The fractured specimen was further examined under
The rolling operation was performed with extreme care to SEM for detailed fracture surface analysis. Similarly, pull
achieve clad and meat thicknesses of 400 and 700 mm, and peel tests were carried out in the same Instron
respectively, in both case-A and case-B. These plates were machine under identical environmental conditions; how-
then blister tested to characterize the nonbond region. ever, the sample holders for the respective tests were
customized and specially fabricated. As shown in Fig. 3c
II.B. Sample Preparation and Characterization the sample for pull testing was first glued with high-
strength epoxy to polished and cleaned surfaces of
The as-rolled sandwich plates for case-A and case-B stainless steel platen (see Fig. 3c). A constant crosshead
were used to prepare tensile test specimens in both rate was applied to pull the sample, and the induced load

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 37

Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

TABLE I
Chemical Composition of AA6061 Grade Aluminum Alloy

Element Mg Si Mn Fe Cu Cr Zn Ti Al

Concentration (wt%) 1 0.67 0.19 0.20 0.30 0.024 0.015 0.022 97.57

EPMA for quantitative and qualitative analyses of the

interaction product. The samples prepared from the
as-fabricated plate for case-A were cut in the required
proportion and were vacuum sealed in a quartz tube under
partial pressure of helium gas to perform the desired heat
treatment (see Fig. 4). The as-fabricated and heat-treated
samples were further evaluated by XRD for qualitative
and quantitative phase analyses using the Powder Cell
software program. The program uses an integrated
intensity method20 to calculate relative phase percentage.
These samples were also examined by Leica Metallogra-
phy Work Station software for quantitative phase analysis.

III. RESULTS AND DISCUSSION

Tail end opening
III.A. Microstructure and Phase Characterization
The stereo and optical microscopy results of as-rolled
Fig. 2. Photograph of fuel meat sandwich with opening at tail
end.
sandwich plates (both case-A and case-B) are shown in
Fig. 5. The results clearly indicate that in case-A, an
interaction layer product between U3Si2 dispersoids and
was measured. The bond strength was then calculated aluminum was observed; however, the sample in case-B
from Eq. (1) (Ref. 17): did not show any sign of chemical interaction between the
Y2O3 particle and aluminum. Therefore, the results clearly
Average Bond Strength ¼ suggest that the Y2O3 particle has remained chemically
Average Load inert with surrounding aluminum under the given set
: ð1Þ of processing parameters. The microstructure shown in
Cross-Sectional Area of Sample
Figs. 5a and 5b also indicate interparticle cracks in U3Si2
Similarly, for the peel test, the bent, unbonded ends of the granules that reiterate their extreme fragility. These cracks
test specimen were clamped in the test grips of the tension might have developed either during compaction or rolling.
testing machine (see Fig. 3d), and an autographic The samples were further characterized by EPMA, and the
recording of load versus head movement was plotted. results are shown in Fig. 6. The qualitative analysis of the
The peel resistance over 31-mm length of the bond line backscattered electron image shown in Fig. 6a reconfirms
after the initial peak was measured, and the bond strength dispersoid-matrix chemical interaction. However, a quan-
was calculated from Eq. (2) (Ref. 18): titative analysis was also performed, and the elemental
line profile of U, Al, and Si across the interaction layer is
Average Load shown in Fig. 6b. As shown, the line profile between
Average Bond Strength ¼ : ð2Þ points B and C is a logarithm plot; therefore, the above
Width of Sample
case represents the condition for non-steady-state diffu-
The metallography and microhardness characterization of sion (i.e., Fick’s second law) where the concentration of
the sandwich plates in both case-A and case-B was elements changes exponentially across the interface and
performed at long transverse (containing rolling direction therefore is represented by Eq. (3):
and normal direction) sections of the sandwich plates by a   2 
stereo and optical microscope and a Vickers microhard- B 2x
Cðx,tÞ ¼ pffiffiffiffiffiffiffiffiffi exp : ð3Þ
ness tester, respectively. The samples for microstructural pDt 4Dt
examination were prepared by standard metallography
practice.19 These samples were further examined under The boundary conditions are

38 NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015

 Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

Cladding
Meat

25 mm

(a) (b)

(d)

(c) (e)

Sample for peel test

(f) (g)

Fig. 3. (a) Schematic for peel test sample, (b) schematic for pull test sample, (c) photograph of samples for tensile test, (d)
photograph of samples for pull test, (e) photograph of pull test sample glued with holder, (f) photograph of peel test specimen
with fixture, and (g) photograph of sample after peel test.

where

B ¼ amount of solute added in the system

x ¼ distance at which concentration is
10 mm measured
D ¼ diffusivity
Fig. 4. Photograph of quartz encapsulated sandwich plate
sample. t ¼ time
C(x,t) ¼ concentration of particular element at a
distance x and time t.
Cðx ¼ 1,tÞ ¼ 0
The reports available in the literature also suggest that
and ð1 growth of the U(Al,Si)3 layer at the U3Si2-Al interface
Cðx,tÞdx ¼ B , obeys the classical diffusion kinetics model, and the
0 interaction layer width Y is represented by Eq. (4) (Ref. 15):

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 39

Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

(a) (b)
Fig. 5. Stereo and optical micrographs of as-fabricated samples: (a) U3Si2 dispersed in aluminum matrix (case-A) and (b) Y2O3
dispersed in aluminum matrix (case-B).

Y 2 ¼ K0t ð4Þ

and
U3Si2
  
0 Q th
Interaction layer K ¼ K exp 2
th
, ð5Þ
AB C D phase U (Al, Si)3 RT
Aluminum
where
K th ¼ interaction layer growth coefficient
t ¼ time
(a)
Q th ¼ activation energy required by the process.
A B C D

In this study the calculations are carried out for the

out-of-pile test condition; therefore, the growth of the
interaction product is primarily governed by a thermal
process. Hence, in order to calculate Y, the value for Qth
and K th was taken as 354 kJ/mol and 2.6 £ 1021 mm2·s-1,
respectively, from the literature.15 The value of Y was
calculated from Eqs. (4) and (5) and was found to be
7.48 mm for the temperature 5008C and time 18 000 s.
Similarly, the U(Al,Si)3 growth layer was also measured
experimentally using an optical micrograph, and the
(b)
average value was found to be 8.53 mm, which is very
close to the theoretically estimated value. In the exper-
Fig. 6. (a) Backscattered electron image of U3Si2 particle imental result nine numbers of data points were taken with
showing envelope of interaction layer and (b) elemental a standard deviation of 0.17.
line profile of uranium, silicon, and aluminum across The qualitative and quantitative elemental mapping
the line segment AB (0 respresents point A while 18 overreaction product zone, dispersoid particle, and matrix
mm represents point D). were also performed and are shown in Fig. 7. The results

40 NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015

 Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

F
B

D
E
C A

(a) (b)

(c)

Fig. 7. Elemental area mapping by EPMA: (a) uranium, (b) silicon, and (c) aluminum.

confirm the presence of reaction product at various are also marked in Fig. 7b. These points correspond
cracked regions in U3Si2 dispersoid suggesting uniform directly to the region over which mapping was performed
chemical interaction. These results were further examined and the data were evaluated. The data confirm the
for quantitative elemental analysis and are shown in Table II. chemical formula of the reaction product as U(Al,Si)3
The points are illustrated as A through F in Table II and intermetallic. However, it could also be observed that the
concentration of various elements at different locations
was found progressive from higher to lower concen-
TABLE II
tration, indicating diffusion control reaction kinetics. The
Quantitative Elemental Analysis Results by EPMA* samples for both case-A and case-B were also character-
ized by XRD, and the results are shown in Fig. 8. The
Point Al Si U Total Probable Phase result shown in Fig. 8a indicates the presence of a
U(Al,Si)3 intermetallic phase in addition to U3Si2 and
A 43.61 33.14 23.26 100 U(Al, Si)3 aluminum for the case-A sample, while for the case-B
B 45.5 31.44 23.06 100 U(Al, Si)3 sample (as shown in Fig. 8b), only Y2O3 and aluminum
C 44.85 32.23 22.92 100 U(Al, Si)3 phases were observed. The XRD results were further
D 0.92 42.04 57.03 100 U3Si2 analyzed by the Powder Cell program to evaluate the
E 1.13 42 56.88 100 U3Si2 quantitative phase percentage of U(Al,Si)3. The phase
F 0.35 42.44 57.22 100 U3Si2
percentage of U(Al,Si)3 in the as-fabricated sample was
*Quantitative elemental analysis results are in units of atomic found to be ,11 vol %. The result obtained from XRD
percent. The points A, B, C, D, E, and F are as shown in Fig. 7b. analysis is in close agreement with the microstructural

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 41

Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

TABLE III

U3Si2(201)
Interaction Product (U(Al, Si)3)

Al(111)
Average Microhardness Results of Different Phases

U3Si2(211)
Average
Hardness Standard
Description (VHN) Deviation

Al(200)
U3Si2(220)
Cladding 81 1.76
U3Si2(210)

Al(220)
U3Si2(411)
U3Si2(331)
U3Si2(111)

U3Si2(312)
Matrix (aluminum) 53 1.23

U3Si2(511)
U3Si2(212)
Dispersoid (U3Si2) 813 2.91
U3Si2(002)

U3Si2(410)
U3Si2(001)

Interaction product [U(Al,Si)3] 358 2.50

Interaction Product
U3Si2
500-100 Hr
(a) 500-50 Hr
500-25 Hr
Y2O3(222)

Relative Intensity
400-100 Hr

400-50 Hr

400-25 Hr

300-100 Hr
Y2O3(440)

300-50 Hr
Y O (622)
Y2O3(400)

300-25 Hr
Al(111)

Al(200)

Al
Y2O3(136) 2 3

Al(220)
Y2O3(211)

Al

Al
Y2O3(134)

500-30 min.
Y2O3(411)

Y2O3(332)

Y2O3(125)

Y2O3(444)
Y2O3(611)
Y2O3(145)
Y2O3(433)

Y2O3(721)
Y2O3(543)

2θ (°)

Fig. 9. X-ray diffraction pattern of U3Si2 dispersed in aluminum

matrix sample with different heat treatment profiles.
(b)
product; however, heat treatment at 5008C even for
Fig. 8. X-ray diffraction profile of sample: (a) U3Si2 dispersed minimum duration (i.e., 25 h) does not show any sign of
in aluminum matrix (case-A) and (b) Y2O3 dispersed in the U3Si2 phase in the diffraction pattern. It could further
aluminum matrix (case-B). be observed that heating the sample at 3008C and 4008C
for all the three durations does not reveal any significant
phase analysis result since the U(Al,Si)3 phase percentage change in the interaction product volume percentage (see
was calculated as 10.2 vol %. The sample was further Fig. 10). It is also worthwhile mentioning that the optical
characterized by microhardness testing, and the results are microstructures of heat-treated samples with excessive
shown in Table III. The hardness value of U(Al,Si)3 swelling could not be established. Therefore, in those
intermetallic (e.g., chemically interacted phase) was found particular samples the results derived from stereo
to be around 2.28 times lower than the dispersoid phase microscopy are given for the understanding of readers,
(i.e., U3Si2 compound), therefore confirming it to be a and a representative stereo micrograph is shown in Fig.
distinct phase resulting from the chemical interaction. The 11. The micrograph clearly indicates that the sample has
hardness data results obtained in the present study are in delaminated and bended due to excessive swelling.
close agreement with the literature value.13
The XRD analysis was also carried out for the case-A III.B. Mechanical Property Characterization
sample prepared under different heat treatment histories
(i.e., 3008C, 4008C, and 5008C for 25, 50, and 100 h It was clearly mentioned in the beginning that one of
each), and the results are shown in Fig. 9. The results the objectives of the paper is to characterize the effect of
clearly indicate that heating the samples at 3008C and chemical interaction between dispersoid and aluminum on
4008C even for the maximum duration does not bond strength. Therefore, the samples derived for the pull
completely eliminate the U3Si2 compound from the end test were characterized as per schematic shown in Fig. 12

42 NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015

 Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

(i.e., 4 MPa). However, the quantitative analysis of the

bond strength for the case-A samples was separately
carried out by the peel test method described in Sec. II.B,
and the average values (after taking six measurements in
each case) were calculated as 8.34 and 7.88 N/mm in the
longitudinal and transverse directions, respectively.
A typical plot between the load and crosshead movement
and the schematic are shown in Fig. 13. Similarly, the
average bond strength value for the case-B sample was
also calculated by this methodology and was found to be
2.25 N/mm. Therefore, the higher bond strength in case-A
could be attributed to the interaction layer growth between
U3Si2 dispersoid and aluminum matrix.
The tensile test analyses in both the longitudinal and
transverse directions were also carried out for the case-A
sample, and the results are shown in Table IV. The results
Fig. 10. Bar chart representation on quantitative evaluation of clearly indicate that tensile strength along the rolling
U(Al,Si)3 phase after different heat treatment cycles for direction (longitudinal direction) was higher as compared
case-A samples. to the transverse direction, which is also in line with the
predictions made by composite theory.21 The higher
tensile strength in the longitudinal direction as compared
with the transverse direction could well be supported by
Cladding
Meat
the fact that the samples were prepared by hot roll
Cladding bonding operation followed by cold rolling, which would
result in piling up of dislocations in the rolling direction.
It is also important to note that the blister test operation
performed during sample preparation has not resulted in
Fig. 11. Stereo micrograph of actual fuel meat after annealing at complete recrystallization.22 Therefore, it could safely be
5008C for 30 h. submitted that the dislocation pileup in the rolling
direction has resulted in higher strength in the longitudinal
for both case-A and case-B to calculate bond strength. direction as compared to the transverse direction.
It was found that the average bond strength (after six The fractography results shown in Fig. 14 indicate
measurements for each case) measured for the case-B that the mode of fracture in the cladding was ductile while
sample was 1.02 MPa while the actual bond strength for it was a mixture of both ductile and brittle in the meat
the case-A sample was not derived by the present region (i.e., dimples in aluminum and cleavage in U3Si2
methodology because of the failure of epoxy (before particles). Table IV collates the different tensile properties
actual bond breaking). Although the actual bond strength measured for both cases. The tabulated results clearly
was not measured for case-A by the pull test, it can safely suggest that the strength and ductility of the plate fuel
be submitted as more than the strength of epoxy remain unaffected during dispersoid-matrix chemical

Fig. 12. (a) Schematic of pull test specimen with fixture and (b) a typical load versus time plot obtained after the test.

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 43

Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

(a) (b)

Fig. 13. (a) Schematic of peel test specimen with fixture and (b) a typical load versus crosshead movement plot obtained after the test.

Fuel meat region

(a) (b)
Dimple in aluminum
Shearing in U3Si2 matrix
particle

Fig. 14. (a) Cross-section view of fractured plate fuel element and (b) fractured surface of actual fuel meat.

interaction. The results shown in Fig. 14 also indicate that below. The EPF in the longitudinal and transverse
during tensile loading, the fuel compact has failed by directions can be shown by Eqs. (6) and (7), respectively:
brittle fracture and therefore the flow properties were
primarily controlled by cladding characteristics. Hence,
this could be attributed for similar tensile properties in EPF ¼ ECladding £ V Cladding þ EMeat £ V Meat ð6Þ
both case-A and case-B.
Similarly, as one of the objectives of the present and
paper, the effect of interaction on intrinsic bulk 1 V Cladding V Meat
mechanical properties EPF of the sandwich plate was also ¼ þ , ð7Þ
EPF ECladding EMeat
characterized. However, it was also realized that modulus
of elasticity being a material property cannot be compared where
for these two cases. Therefore, in order to evaluate the
effect of chemical interaction on EPF, the modulus of Ecladding ¼ elastic modulus of cladding, and its
elasticity was theoretically calculated by layered compo- value was obtained experimentally
site material mechanics theory,23 which is described as 72 GPa

44 NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015

 Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

Vcladding, VMeat ¼ volume fractions of cladding and meat

Deviation
(these two values were calculated to

Standard

1.12
1.29
0.25
0.69
be 0.54 and 0.46, respectively)
EMeat ¼ elastic modulus of meat.
The elastic modulus of meat without porosity is
termed E0Meat and can be expressed by Eq. (8):

Transverse
Direction

114.18
77.49

73.58
9.89
E0Meat ¼ EAl £ V Al þ EU3 Si2 £ V U3 Si2 , ð8Þ

where
Case-B

EAl, E U3 Si2 ¼ elastic modulus of aluminum powder

Deviation
Standard

(i.e., 69 GPa) and U3Si2 compound,

1.06
1.32
0.33
0.61 respectively
VAl , V U3 Si2 ¼ volume fraction of aluminum powder
(0.58) and U3Si2 (0.42), respectively.
Average Tensile Test Results for Case-A and Case-B Samples

Longitudinal

It is also understood that since U3Si2 was prepared

Direction

129.75
98.28

12.12
73.62

by powder metallurgy in the present study, the porosity

of 5% was also accounted in the calculation. It is also
well understood from the literature that porosity does
adversely affect the elastic modulus.24 Equation (9)
represents the correlation between elastic modulus and
porosity:
Deviation
Standard

 
1.11
1.33
0.29
0.49

p f
TABLE IV

E ¼ E0 1 2 , ð9Þ
pc
where
E0 ¼ modulus without porosity
Transverse
Direction

78.98

59.67
116.13
9.68

p ¼ pore fraction
f, pc ¼ constants.
Case-A

The values of f and pc are taken as 2.23 and 0.652 in

the present calculation.24 If the value of E0 is taken as 131
Deviation
Standard

1.08
1.72
0.34
0.58

GPa (elastic modulus of U3Si2), then by using Eq. (9) the

value of E is calculated and found to be 109.64 GPa
(elastic modulus of U3Si2 with 5% porosity). The values
were then substituted in Eq. (8) to calculate E0Meat and
were found to be 86.07 GPa. The relationship between
Longitudinal

the porosity in the fuel meat compact and the volume

Direction

130.53
99.96

12.24

fraction of the fuel phase is also shown in Eq. (10)

60

(Ref. 13):

V p ¼ 0:072 £ V f 2 0:275 £ V 2f þ 1:32 £ V 3f , ð10Þ

Tensile strength (MPa)

Elastic modulus (GPa)

Yield strength (MPa)

where Vf is the volume fraction of the fuel phase and its

value is taken as 42% in the present calculation.
Property

Elongation (%)

Therefore, by using Eq. (10), the porosity percentage in

the fuel meat compact was calculated and found to be
7.94%. Equation (9) was once again used to calculate the
value of Emeat and was found to be 64.43 GPa. Therefore,
the elastic moduli of the plate fuel in the longitudinal and
transverse directions were calculated from Eqs. (6) and (7)

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 45

Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

and were found to be 68.51 and 68.30 GPa, respectively. from the viewpoint of the effect of flow-induced vibration
Similar calculation was also carried out for case-B by on fuel plates.26
taking the value of the elastic modulus of high-density
Y2O3 as 173 GPa (Ref. 25) and considering 5% porosity. ACKNOWLEDGMENTS
Based upon the calculations the theoretical values of the
elastic modulus of the plate fuel for the case-B sample in The authors are grateful to the authorities of BARC, in
the longitudinal and transverse directions were found to particular, S. Anantharaman, MFD, and R. Keswani, MFD
be 73.58 and 73.53 GPa, respectively. Alloy and Plate Fuel Fabrication Facility, for their support
A comparison between the theoretical values and the and encouragement toward the development of the high-
experimental data (see Table IV) for case-B clearly density uranium system as a potential fuel material for our
indicates that the moduli of elasticity are similar. On the research and test reactor program. The authors also thank
the technical staff members of MFD, in particular, S. S. Gotad,
other hand, the elastic moduli of the composite plate for
M. R. Shaikh, A. B. Chowdhury, S. S. Prasad, R. K. Singh,
case-A in the longitudinal and transverse directions were and B. N. Pisal, for their help during the course of work.
determined experimentally and found to be 60 and 59.67 The authors also acknowledge the help and support provided by
GPa, respectively (see Table IV). It shows that the J. S. Dubey, Post Irradiation Examination Division (PIED), on
experimental values for case-A are somewhat lower than performing mechanical test analysis on plate fuel samples, as
the theoretically estimated values. Hence, the lower value well as G. K. Dey, Materials Science Division (MSD), and
of the modulus could be related to the chemical interaction A. Laik, MSD, for providing all the necessary help and support
of dispersoid and matrix. This result is significant for fuel in carrying out the EPMA studies. The authors also thank
plate performance because a reduction in elastic modulus S. Kumar and B. N. Rath, PIED, for their help in obtaining the
will decrease the plate stiffness and in turn would SEM analysis results.
adversely affect flow-induced vibration in plate fuel during
actual operation.26 REFERENCES

1. “Announcement: International Meeting on Reduced Enrich-

ment for Research and Test Reactors (RERTR): Petten (North-
IV. CONCLUSION Holland), The Netherlands, 14–16 October, 1985,” J. Nucl.
Mater., 132, 203 (1985); http://dx.doi.org/10.1016/0022-3115
The present study was carried out with the objective (85)90415-5.
to evaluate the effect of chemical interaction between 2. J. P. DURAND, Y. LAVASTRE, and M. GRASSE,
dispersoid and matrix on different properties of the plate “Proceedings of the 20th International Meeting on Reduced
fuel element. In view of this extensive study and the Enrichment for Research and Test Reactors,” ANL/TD/TM99-
present investigations, it could be safely concluded that 06, p. 28, Argonne National Laboratory (1999).
fuel dispersoid (U3Si2) and matrix (aluminum) undergo 3. V. P. SINHA et al., “Development, Preparation and
chemical interaction at fuel fabrication conditions (i.e., Characterization of Uranium Molybdenum Alloys for Dis-
5008C for 5 h) to form a low-density compound [i.e., persion Fuel Application,” J. Alloys Comp., 473, 238 (2009);
U(Al,Si)3]. The interaction layer growth for U3Si2 http://dx.doi.org/10.1016/j.jallcom.2008.05.061.
dispersoids prepared by the powder metallurgy route
4. A. J. YOUKER et al., “Aqueous Processing of U-10Mo
was found to be consistent with the calculated value by Scrap for High Performance Research Reactor Fuel,” J. Nucl.
the correlation found in the literature (i.e., U3Si2 prepared Mater., 427, 185 (2012); http://dx.doi.org/10.1016/j.jnucmat.
by the melting and casting route). 2012.05.002.
The conclusion is substantial because U3Si2 prepared
by powder metallurgy has intragranular porosity with 5. D. M. WACHS, C. R. CLARK, and R. J. DUNAVANT,
“Conceptual Process Description for the Manufacture of
distinct advantages,6 and it (,6%) does not adversely
Low-Enriched Uranium-Molybdenum Fuel,” INL/EXT-08-
influence the interaction characteristics under the given set 13840, Idaho National Laboratory (Feb. 2008).
of experimental conditions. It could be further concluded
from the present investigations that controlled interaction 6. V. P. SINHA et al., “Development of Powder Metallurgy
between U3Si2 and aluminum will result in higher bond Technique for Synthesis of U3Si2 Dispersoid,” J. Nucl. Mater.,
strength. However, exposing the fuel plates at higher 383, 1–2, 196 (2008); http://dx.doi.org/10.1016/j.jnucmat.2008.
08.046.
temperature for longer duration results in excessive
interaction and therefore will lead to swelling and 7. Y. S. KIM, “3.14 – Uranium Intermetallic Fuels (U-Al,
delamination. The effect was found more influential with U-Si, U-Mo),” Compr. Nucl. Mater., 3, 391 (2012);
temperature rather than with time. It could also be http://dx.doi.org/10.1016/B978-0-08-056033-5.00112-9.
concluded that although U3Si2-aluminum chemical inter- 8. J. MARÍN et al., “Synthesis and Clad Interaction Study
action does not influence the tensile properties of plate of U3Si2 Powders Dispersed in an Aluminum Matrix,”
fuel elements, the modulus of elasticity in the composite J. Nucl. Mater., 228, 1, 61 (1996); http://dx.doi.org/10.1016/
fuel plate has deteriorated, and the result was significant 0022-3115(95)00167-0.

46 NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015

 Sinha et al. CHEMICAL INTERACTION OF U3Si2 AND Al MATRIX IN PLATE FUEL ELEMENTS

9. S. NAZARÉ, “Low Enrichment Dispersion Fuels for 17. ASTM C 633-01, “Standard Test Method for Adhesion or
Research and Test Reactors,” J. Nucl. Mater., 124, 14 (1984); Cohesion Strength of Thermal Spray Coatings,” ASTM
http://dx.doi.org/10.1016/0022-3115(84)90005-9. International, West Conshohocken, Pennsylvania.
10. J. GAN et al., “Microstructure of the Irradiated U3Si2/Al 18. ASTM D 1876-01, “Standard Test Method for Peel
Silicide Dispersion Fuel,” J. Nucl. Mater., 419, 1–3, 97 (2011); Resistance of Adhesives (T-Peel Test),” ASTM International,
http://dx.doi.org/10.1016/j.jnucmat.2011.07.030. West Conshohocken, Pennsylvania.
11. A. LEENAERS et al., “Post-Irradiation Examination of 19. ASM Handbook, Vol. 9, “Metallography and Microstruc-
AlFeNi Cladded U3Si2 Fuel Plates Irradiated Under Severe tures,” ASTM International, West Conshohocken, Pennsylvania
Conditions,” J. Nucl. Mater., 375, 243 (2008); http://dx.doi.org/ (2004).
10.1016/j.jnucmat.2008.01.013.
20. B. D. CULLITY and S. R. STOCK, Elements of X-Ray
12. G. L. HOFMAN, G. L. COPELAND, and J. L. SNEL- Diffraction, 3rd ed., Prentice Hall (2001).
GROVE, “Postirradiation Evaluations of Capsules HANS-1
and HANS-2 Irradiated in the HFIR Target Region in 21. G. E. DIETER, Mechanical Metallurgy, SI Metric Edition,
Support of Fuel Development for the Advanced Neutron McGraw Hill (1988).
Source,” ORNL/TM-13061, Oak Ridge National Laboratory 22. D. KOHLI et al., “Fabrication of Simulated Plate Fuel
(1995). Elements: Defining Role of Stress Relief Annealing,” J. Nucl.
13. J. L. SNELGROVE et al., “The Use of U3Si2 Dispersed in Mater., 447, 1–3, 150 (2014); http://dx.doi.org/10.1016/j.jnucmat.
Aluminum in Plate-Type Fuel Elements for Research and Test 2014.01.002.
Reactors,” ANL/RERTR/TM-11, Argonne National Laboratory 23. A. K. BHARGAVA, Engineering Materials: Polymers,
(Oct. 1987). Ceramics and Composites, p. 224, Prentice Hall of India (2004).
14. K.-H. KIM et al., “Thermal Compatibility Studies of U3Si2 24. A. P. ROBERTS and E. J. GARBOCZI, “Elastic Properties
Dispersion Fuels Prepared with Centrifugally Atomized Pow- of Model Porous Ceramics,” J. Am. Ceram. Soc., 83, 12, 3041
der,” J. Nucl. Mater., 270, 315 (1999); http://dx.doi.org/ (2000); http://dx.doi.org/10.1111/j.1151-2916.2000.tb01680.x.
10.1016/S0022-3115(99)00005-7.
25. O. YEHESKEL and O. TEVET, “Elastic Moduli of
15. Y. S. KIM and G. L. HOFMAN, “Interdiffusion in U3Si- Transparent Yttria,” J. Am. Ceram. Soc., 82, 136 (1999);
Al, U3Si2-Al, and USi-Al Dispersion Fuels During Irradiation,” http://dx.doi.org/10.1111/j.1151-2916.1999.tb01733.x.
J. Nucl. Mater., 410, 1–3, 1 (2011); http://dx.doi.org/10.1016/
j.jnucmat.2010.12.031. 26. M. HO, G. HONG, and A. N. F. MACK, “Experimental
Investigation of Flow-Induced Vibration in a Parallel Plate
16. ASTM E 8-04, “Standard Test Methods for Tension Reactor Fuel Assembly,” presented at 15th Australasian Fluid
Testing of Metallic Materials,” ASTM International, West Mechanics Conf., The University of Sydney, Sydney, Australia,
Conshohocken, Pennsylvania. December 13–17, 2004.

NUCLEAR TECHNOLOGY VOL. 192 OCT. 2015 47