0
STEAM
CHEMICAL
RE24C"Y
OF
PLASMAS-SPRAYED BERYLLIUM
Robert
A
hderl, Robert
J.
Pawelko and Galen R. Smolik
Idaho National Engineering and Environmental Laboratory
P.O.
Box
1625, Idaho Falls, Idaho, 83415-7113
(208) 526-4153
ABSTRACT
Plasma-spraying with the potential for in-situ repair
makes beryllium a primary candidate for plasma facing and
structural
components
in
experimental magnetic hion
machines. Deposits with good thermal conductivity and
resistance to thermal cycling have been produced with low
pressure plasma-spraying (LPPS). A concern during a
potential accident with steam ingress
is
the amount
of
hydrogen produced by the reactions of steam with hot
components.
In
this
study
we measure the reaction rates
of
various deposits produced
by
LPPS with steam from 350°C
to above 1000°C. We correlate these reaction rates with
measurements
of
density,
open
porosity and I3ET surface
areas. We find the reactivity to
be
largely depndent
upon
effective surface area. Promising results were obtained
below 600°C fiom a 94% theoretical dense
('I'D)
deposit
with a BET specific surface area of
0.085
m*/g Although
reaction rates were
higher
than those for dense consolidated
beryllium they were substantially lower, i.e.. about
two
orders of magnitude,
than
those obtained
from
previously
tested lower density plasma-sprayed deposits.
I.
INTRODUCTION
Most investigations into the plasma :spraying of
beryllium
are
those
by
Union
Carbide and the Atomic
Weapons Research Laboratory (England) during the 1960-
7O's[l]
and later by Battelle. Historically
the
cpality
of
plasma-sprayed deposits has
been
limited by having to spray
in an inert gas near atmospheric pressure.
Dunmur[l]
reported densities (-90%)
,
porosities (10-12%0), and
gas
adsorption surface areas (-0.8 m2/g) for as-sprayed deposits.
Deposits produced by Battelle
in
1990's
for the Fusion
Program
and used in earlier steam reactivity experiments
[2]
had densities (86-92%) and open porositits (6-12%).
Advances in plasma spray technology allowing spraying
under reduced pressures, Le., less
than
400
torr, has
contributed to capabilities to produce better
@ity
deposits.
The Beryllium Atomization and Thermal SFray Facility
(BATSF) at
Los
Alamos
National Laboratory has such
capabilities
along
with
a -erred-arc cleaning system.
This
integrated system using improved powder feed stock
and
small
amounts of hydrogen in the plasma
gas
has
hard
G.
Castro
J
u
L
2
0
1998
Los
Alamos
National Laboratory
Los
Alamos,
New Mexico 87545
Q
s
I
1
(505)667-5191
produced some high quality beryllium deposits. Deposits
with
good substrate adherence, densities,
and desirable
crystallographic orientation have been produced with good
thermal conductivity and resistance to thermal cycling [3,4].
This
system
was
used to proctUce
the
deposits studied in
this
investigation.
II.
SAMPLES
AND
PROCEDURES
Two
types
of deposits were prepared
in
BATSF using a
SGlOO
Miller
Thd
torch mounted
on
an
automated
X-Y
manipulator. The coatings were produced from
Brush
Wellman spherical atomized powder (-38 pm to
+10
pm,
0.36
wtY0
BeO).
Beryllium was used as the substrate for
Plate
A
The other deposit, Plates B and C, was sprayed
onto a CuNiBe alloy. The Wen-ed-arc system shown in
fig.
1.
was
used
to clean and preheat the substrates before
spraying. Other cleaning and spraying parameters are
shown in Table
I.
The only notable differences are 1) a
higher substrate temperature for Plate
A
(600°C) compared
to (450-550°C) for Plates
B
and C and 2) higher power input
into the torch for Plate A (15.8-16.6
kw)
compared to
12.8
kW
for Plates B and C.
Fig.
1
Negative transferred-arc cleaning system used for
plasma spraying beryllium deposits.
This
work
is
supported by the
U.S.
Department
of
Energy, Office of Energy Research, under the DOE Idaho Operations
Contract DE-ACO7-94lD 13 223.
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, completeness, or use-
fulness of
any
information, apparatus, product, or process disclosed, or represents
that
its
use would not infringe privately owned rights. Reference herein to any
spc-
cific
commercial prcduct, prcwss, or service by trade name, trademark, manufac-
turer, or otherwise
does
not necessarily constitute or imply its endorsement,
mm-
menhtion, or favoring by the United States Government or any agency thereof.
The views and opinions
of
iiuthon exptessed herein
do
not necessarily state or
reflect those of the United States Government or any agency thereof.
DISCLAIMER
Portions of this document may be illegible
electronic im(age products. Images are
produced froim the best available original
document.
Table
2.
Density, porosity, and surface area parameters
of
plasma-sprayed coupons
Specimen Bulk
%
Theoretical
Open
Closed Total
BET
Type density Density
Porosity
Porosity
Porosity SSA
&/cm3)
(“A)
(Ye)
(“Yo)
(%)
(m2/g)
PS-Be-A
1.744
94.1.
1.4 4.2
5.6 0.085
+I-
0.009
+I-
(1.5
+I-
0.6
+I-
0.8
+I-
0.5
+I-
0.009
PS-Be-B
1.702 92.1
7.7
0.2 7.9
1.214
+I-
0.003
+I-
0.2
+I-
0.2
+I-
0.3
+I-
0.2
+I-
0.097
PS-Be-C
1.706 92.3
7.6
0.1
7.7
1.070
+I-
0.008
+I-
0.4
+I-
0.4
+I-
0.6
+I-
0.4
+I-
0.054
(b)
Fig. 3. Microstructures of beryllium plasm sprayed
deposits. (a) Plate
A.
(b)
Plate B. Magnrfication:
300X.
and separations between individual splats. The large
amount of void space revealed by the cross section in
fig.
3(b)
and the SEM
image
of the EDM surface, in
fig
2(b)
indicates high permeability for Deposits B and
C.
The high
total porosity and low closed porosity measurements in
Table 2 confirm
this.
The micrograph for Deposit
A
shows
far fewer unmelted particles, more isolated pores, and very
obscure boundaries
at
splat interfaces. We would expect
this
fiom a higher energy input that would provide higher
temperature, more fluid parbcles. The evidence of the
isolated porosity agrees with data in Table
2
showing that
most of the porosity for
this
deposit
is
closed porosity.
These observations show why access of
Kr
gas
and resultant
BET
SSA
was lower for Deposit
A
compared to
B
and C.
The samples were tested in the Steam-Reactivity
Measurements System
(SRMS)
shown in
fig.
4
and
described
in
detail elsewhere [5,6]. Typical operating
parameters were as follows; line pressure (675-685 torr),
Ar
carrier-gas flow rate (100 std cclmin), steam flow rate (2500
stdcclmin for a water throughput
of
2
cclmin), steam-
generator temperature (350°C). The reaction-chamber tube
furnace can operate between 25°C to 1200°C. The system
response time for these conditions is about
6
minutes and
H2
detection sensitivity with the quadrupole
mass
spectrometer
(QMS)
is about 3 ppm
of
H2
in
Ar.
This
allowed us to
detect hydrogen release
from
the steam reaction with the
plasma sprayed Samples down to 350°C. Oxidation rates
were
also
confirmed by measurements of mass change.
Specimens were suspended near the center
of
the
hot zone
in
the sample furnace via a platinum hanger attached to a
quartz
support rod. Specimen temperatures were based
on
prior calibration between a Type-K thermocouple (TC) at
the sample position and the
furnace
controlling
TC.
Itt.
EXPERIMENTAL RESULTS
Only samples from Plates
A
and B were tested due to the
very similar nature
of
the plasma-sprayed deposits
on
Plates
B
and C. Hydrogen and weight
gain
measurements for
these tests are summarized
in
Table
3.
In
general there was
very good agreement
between
the amounts of hydrogen
measured with the
QMS
and those calculated fkom
mass
change. There were some differences for samples heated to
6OOOC and higher. Some of these samples underwent
temperature excursions due to the exothennic energy
generated from the steam reaction. Oxide formed
on
these
specimens was very loose and friable resulting in
non-
recoverable amounts during weighing. Data in Table3
show that a sigmfkant
portion
of these samples reacted.
Table
I.
Spaying parameters for beryllium deposits
The as-sprayed deposits of
nominal
50-mm
widths were
8-mm, 10-mm and 12-mm thick for Plates A, B, and
C,
respectively. Milling was used to remove the substrates and
provide flat parallel surfaces. Electro-dischargc: machining
(EDM) was then used to remove oxidation coupons that
were nominally 15-mm wide
x
15-mm long
x
3-mm
thick
The milled surfaces
on
the back of the coupons had a finish
of
63
micro-inch while front surfaces of the coupons
underwent erosion from EDM. Illustrations of
kloth
types
of
surfaces
as
revealed by SEM at
SOOX
are shown in fig.
2.
Immersion density methods were used to measilre the bulk
densities and
open
porosities for the various deposits.
Measurements
on
bulk
sections of the deposits and on
individual
coupons
provided essentially the same
results.
Average values for the density and porosity parameters
obtained
from
the eight to ten coupons from each condition
are summarized in Table
2.
Theoretical density
is
considered
as
1.85 g/cm3. Surface
areas
wex measured
using a Micromeritic
ASAP
2010 gas-adsorption system
with
Kr
gas.
One coupon from each plate was split into
two
sections. Average BET specific surface areas obtained from
several measurements from each section are included
in
Table
2.
Metallographic sections were prepared from each of the
plasma-sprayed conditions. Light microscopy using both
bright-field and polarized light at 1002: provided
information about grain size, porosity and crystallographic
orientation.
Porosity
levels determined by image analyses
of
4.0%
for
A,
5.5%
for B, and
6.2%
for
C
art?
in
general
agreement
with
the total porosities reported
in
Table
2.
Bright field images at higher magmfication
(400X)
are
shown for Deposits A and B in fig.
3.
The micrograph for
B
shows a si&icant number of melted particles, porosity,
Fig.
2.
Surfaces
of
oxidation coupons. (a) Milled surface
on
the back of
a
Type
A
coupon.
(b)
EDM surface on the front
of
a Type
C
coupon. Magnification:
SOOX.