Download 0229.PDF

CP620, Shock Compression of Condensed Matter - 2001
edited by M. D. Furnish, N. N. Thadhani, and Y. Horie
© 2002 American Institute of Physics 0-7354-0068-7/02/$ 19.00
SHOCK INDUCED MELTING OF LEAD (EXPERIMENTAL STUDY)
Catherine Mabire, Pierre L. Hereil
Centre d Etudes de Gramat - 46500 GRAMAT - FRANCE
Abstract. To investigate melting on release of lead, two shock compression measurements have
been carried out at 51 GPa. In the first one, a pyrometric measurement has been performed at the
Pb/LiF interface. In the second one, the Pb/LiF interface velocity has been recorded using VISAR
measurement technique. VISAR and radiance profile are in good agreement and seem to show
melting on release of lead.
by optical pyrometry, the other with a velocity
measurement. The aim of these two shots was to
detect melting of lead and also to compare the
velocity signal with the color temperature signal.
EXPERIMENTS
The single-stage powder gun ARION was used
to perform two plate impact experiments on lead.
The experimental configuration is depicted in
figure 1 and the parameters of these two experiments
are presented in table 1. Impact velocity and tilt
were measured by mean of self-shorting pins located
at the periphery of the target. Typically accuracy on
impact velocity was ± 1 % and planarity was better
than 5 mrad. The impactor disks, buffer plate, target
and window all present a 0.01 mm flatness on their
two faces and the parallelism difference between the
two faces is smaller than 0.02 mm. The face of the
lead disk aimed by the pyrometer presents a
polished surface with a roughness of 0.01. The
window material is characterized by an optical
polish. The lead used in these experiments had a
purity of 99.9% and a density of 11.346 g/cm3.
INTRODUCTION
The determination of melting curve under shock
compression is very important in shock physics.
Several experimental techniques have been used to
detect
melting
under
shock
conditions:
measurement of the release wave on dense impacted
materials [l]-[2] and porous materials [3],
measurement of the shock temperature by mean of a
pyrometer [4]. Recently, a study on tin [5] showed
that particle velocity measurements with accurate
metrology could display melting on release.
However, theoretically, temperature measurement is
a more sensitive test of melting than a velocity
measurement which provides only pressure/volume
information. But, unlike a velocity measurement,
temperature measurements of shocked heated metals
is still an important unsolved problem of high
pressure physics. It would be ideal to obtain
temperature and velocity measurements for the same
experiment. Unfortunately our experimental device
does not allow that. This paper presents two
experimental results on lead at shock stresses of
about 51 GPa, one with a temperature measurement
Table 1. Parameters for the impact experiments on lead
Shot
impactor buffer plate
target
window
Number
A104
A105
Ta
(3.03mm)
Ta
(3.016 mm)
Cu
(3.024mm)
Cu
(3.054 mm)
Pb
LiF
(3.788mm) (15.039mm)
Pb
LiF
(3.755 mm) (15.022 mm)
229
impact velocity impact stress interface stress
(m/s)______(GPa)______(GPa)
1895 ±4
51.1 ±0.1
30.2 ±0.1
1990 ± 2
54.5 ±0.1
32.1 ±0.1
Interface velocity (m/s)
Cu (3 mm
Pb (4 mm)
_LiF(15rnm)
Ta (3 mm).
VISAR
or
PYROMETER
self shorting pins
(impact velocity and tilt)
Space
Experimental setup
setup
FIGURE 1
1 :: Experimental
FIGURE
g
2000
jL,\J\J\J
A
B
C
1800
1800
1600
1600
650 nm
D
1800
1800
850 nm
ig
g
1600
1
1600
QJ
a
I
1100
nm
HOOnm
1270
nm
1270 nrri
1510
nm
1510 nm
1400
1400
I
1200
"o 1200
U
1000
1000
0.0
0.5
0.5
1.0
1.0
1.5
1.5
Velocity (m/s)
Color temperature (K)
2200
2200
2000
2000
Luminance temperature (K)
Comparison wave
wave diagram
diagram // experimental
experimental results
results
FIGURE 22 :: Comparison
FIGURE
1400
1400
1200
1200 1000
1000 800 _
800
L,
iA
t TX
B
C
N.
l^s^^D
D
o
600 _
600
5
> 400
2.0
2.0
(µs)
Time (ps)
:
_
400 200 200
00 A | , , , , i , , , , i , , , , i , , , ,
0.0
0.5
1.0
1.5
2.0
0.
0
0.5
1.0
1.5
2.
Time (µs)
(ps)
Interface color
color temperatures
temperatures measured
measured for
for shot
shot A104
A104 FIGURE 4 : Interface
velocity for
profile
FIGURE 33 :: Interface
FIGURE
4 : Interface
FIGURE
velocity
profile measured
shot measured
A105
for shot A105
In the
the first
first experiment
experiment (shot
(shot A104),
A104), the
the
In
diagnostic used
used was
was an
an optical
optical pyrometer
pyrometer [6].
[6].
diagnostic
Thermal radiation
radiation emitted
emitted at
at the
the Pb/LiF
Pb/LiF interface
interface
Thermal
was
collected
with
an
optical
head
then
transmitted
was collected with an optical head then transmitted
to the
the pyrometer
pyrometer by
by optical
optical fiber.
fiber. The
The pyrometer
pyrometer is
is
to
designed to
to provide
provide electrical
electrical signals
signals proportionnal
proportionnal
designed
to the
the collected
collected flux
flux at
at six
six wavelengths
wavelengths (500
(500 nm,
nm, 650
650
to
nm,
850
nm,
1100
nm,
1270
nm
and
1510
nm).
nm, 850 nm, 1100 nm, 1270 nm and 1510 nm).
Measured voltages
voltages were
were converted
converted into
into flux
flux from
from
Measured
calibration realized
realized before
before experiment
experiment with
with aa
calibration
blackbody cavity.
cavity. The
The flux
flux was
was then
then converted
converted into
into
blackbody
color
temperature
using
Planck’s
law.
The
color temperature using Planck's law. The
measurement surface
surface was
was located
located at
at the
the focal
focal
measurement
distance of
of the
the optical
optical device.
device. To
To avoid
avoid possible
possible
distance
parasitic light
light radiation,
radiation, the
the edge
edge of
of the
the window,
window, as
as
parasitic
well
as
a
part
of
its
free
face,
were
coated
with
black
well as a part of its free face, were coated with black
paint and
and aa black
black cardboard
cardboard tube
tube was
was put
put on
on the
the
paint
target up
up to
to the
the optical
optical device.
device. The
The glue
glue between
between the
the
target
window and
and the
the lead
lead target
target was
was UV-hardened
UV-hardened
window
LOCTITE
358.
The
glue
thickness
was
less than
than
LOCTITE 358. The glue thickness was less
10
µm. This
This glue
glue remains
remains transparent
transparent under
under shock
shock
10 pm.
for the
the studied
studied range
range of
of stresses
stresses and
and
for
temperatures [7].
temperatures
[7].
The time-resolved
time-resolved color
color temperatures
temperatures for
for 55
The
wavelengths is
is presented
presented in
in figure
figure 3.
3. The
The
wavelengths
temperature in
in lead
lead is
is too
too low
low to
to obtain
obtain an
an
temperature
exploitable signal
for the
the 500
500 nm
nm wavelength.
wavelength. The
The
exploitable
signal for
five signals
signals have
have aa first
first ramp
ramp when
when the
the shock
shock
five
reaches
the
Pb/LiF
interface.
After
0.5
µs,
there
is aa
reaches the Pb/LiF interface. After 0.5 ps, there is
change in
in slope
slope (point
(point B).
B). This
This change
change corresponds
corresponds
change
to aa reflected
reflected wave
wave from
from copper
copper as
the (x,
to
as shown
shown in
in the
(x,
t)
diagram
of
figure
2.
The
decrease
of
signal
t) diagram of figure 2. The decrease of signal (point
(point
230
2200
2200
1.4
3000
3000
1.2
release due to LiF
2500
2500
0.8
1600
0.6
1400
0.4
calculated temperature (hyp. graybody)
calculated temperature (hyp. graybody)
measured luminance temperature (650 nm)
1200
I
............... calculated
calculatedemissivity
emissivity(hyp.
(hyn. graybody)
graybody)
0.0
0.5
0.5
1.0
1.0
1.5
1.5
^ 2000
calculated
calculatedmelting
meltingcurve
curve.
650 nm
2000
calculated
calculated shock
shock in
in lead
lead
850 nm
1100 nm
1270 nm
1510 nm
1500
1500
1000
0.2
measured luminance temperature (650 nm)
1000
1000
Temperature (K)
ε = 0.68 ± 0.1
emissivity
Temperature (K)
1.0
1800
calculated temperature
(hyp. graybody)
^ interface
interface Pb/LiF
Pb/LiF
ca
lcu
lta
ted
Hu
go
nio
t
T = 1860 ± 50 K
2000
2000 -
500
A &
& ^f
experimental melting curve [16]
experimental melting curve [16] —
0.0
2.0
00
10
10
20
20
30
30
40
40
50
50
60
60
70
70
80
80
Stress
Stress (GPa)
(GPa)
Time
Time (µs)
(ps)
FIGURE 5:
FIGURE
5: Calculated
Calculated true
true temperature
temperature and
and emissivity
emissivity
FIGURE
FIGURE 6:
6: phase
phase diagram
diagram of
of lead
lead
C)
C) corresponds
corresponds to
to the
the arrival
arrival of
of the
the unloading
unloading wave
wave
at
the
Pb/LiF
interface.
During
release,
at the Pb/LiF interface. During release, we
we observe
observe
an
increase of
an increase
of signal
signal (point
(point D).
D).
In the
In
the second
second experiment
experiment (shot
(shot A105),
A105), the
the
diagnosic
used
was
a
VISAR
[8].
The
acquired
diagnosic used was a VISAR [8]. The acquired timetimeresolved velocity
profile is
resolved
velocity profile
is displayed
displayed in
in figure
figure 4.
4.
The velocity
The
velocity increase,
increase, point
point A,
A, is
is caused
caused by
by the
the
arrival
compression shock
Pb/LiF
arrival of
of the
the compression
shock at
at the
the Pb/LiF
interface.
The
small
decrease
of
velocity,
point
interface. The small decrease of velocity, point B,
B, is
is
due
to
the
return
wave
from
copper
(figure
2).
Point
due to the return wave from copper (figure 2). Point
C
C corresponds
corresponds to
to the
the arrival
arrival of
of the
the unloading
unloading wave
wave
at
Pb/LiF interface.
at the
the Pb/LiF
interface. As
As we
we will
will see
see in
in the
the next
next
section,
the change
change in
in slope
section, the
slope (point
(point D)
D) is
is consistent
consistent
with the
the increase
increase of
of color
color temperature
temperature in
with
in shot
shot A104
A104
and
may
be
interpretated
as
a
phase
change
and may be interpretated as a phase change from
from
solid
to liquid.
solid to
liquid.
ANALYSIS
ANALYSIS
Shock
stresses
Shock stresses generated
generated in
in lead
lead targets,
targets,
indicated
in
table
1,
have
been
evaluated
indicated in table 1, have been evaluated by
by the
the
impedence
matching
method
knowing
the
equations
impedence matching method knowing the equations
of state
of
state of
of materials
materials [9]
[9] and
and the
the impact
impact velocity.
velocity.
Despite the
Despite
the difference
difference of
of impact
impact velocities,
velocities, these
these
shock
shock stresses
stresses are
are close,
close, so
so we
we can
can consider
consider that
that
velocity and
velocity
and color
color temperature
temperature signals
signals are
are indicative
indicative
of the
release path
path in
of
the same
same shock
shock -– release
in lead.
lead.
(point
A to
point C)
Compression
wave
Compression wave (point A
to point
C)
followed
followed by
by lead.
lead. The
The shape
shape of
of the
the color
color
temperature
profiles
is
not
so
easy
to
temperature profiles is not so easy to analyse
analyse
because
because we
we measured
measured thermal
thermal radiation
radiation emitted
emitted by
by
lead
lead through
through LiF
LiF window
window and
and not
not the
the absolute
absolute
temperature
temperature reached
reached in
in lead.
lead. We
We observe
observe aa ramp
ramp
instead
of
a
plateau
and
an
increase
instead of a plateau and an increase of
of color
color
temperatures
temperatures (point
(point B)
B) instead
instead of
of the
the expected
expected
decrease.
decrease. These
These phenomena
phenomena have
have already
already been
been
observed
observed by
by Bass
Bass [10].
[10]. ItIt isis still
still an
an unresolved
unresolved
problem,
problem, but
but several
several explanations
explanations can
can be
be proposed
proposed ::
1)
a
change
in
optical
properties
of
LiF
1) a change in optical properties of LiF window
window
[11]
[11] ;; 2)
2) aa heterogeneous
heterogeneous interface
interface temperature
temperature [12][12][13]
[13] ;; 3)
3) an
an emissivity
emissivity increase
increase caused
caused by
by damage
damage at
at
the
the Pb/LiF
Pb/LiF interface
interface (because
(because of
of shock
shock and
and
unloading)
unloading) or
or by
by melting
melting of
of lead.
lead. If
If we
we consider
consider the
the
Hugoniot
Hugoniot melting
melting zone
zone defined
defined by
by Lalle
Lalle [15]
[15] (50
(50 to
to
62
62 GPa),
GPa), lead
lead was
was partially
partially shock-melted
shock-melted in
in our
our
experiments.
experiments. So,
So, an
an emissivity
emissivity change
change due
due to
to
melting
on
the
release
wave
(point
B)
is
a
plausible
melting on the release wave (point B) is a plausible
explanation
explanation for
for the
the increase
increase of
of color
color temperature
temperature
instead
of
the
expected
decrease.
instead of the expected decrease.
True
True temperature
temperature and
and emissivity
emissivity have
have been
been
calculated
calculated using
using the
the graybody
graybody approximation
approximation (figure
(figure
5).
5). The
The calculated
calculated emissivity
emissivity value
value is
is about
about 0.68
0.68 on
on
the
plateau.
It
is
less
than
expected
liquid
emissivity
the plateau. It is less than expected liquid emissivity
(about
(about 1)
1) but
but higher
higher than
than initial
initial values
values of
of 0.1
0.1 [14].
[14].
This
This high
high emissivity
emissivity is
is consistent
consistent with
with partial
partial
melting.
melting. However,
However, in
in release,
release, there
there is
is an
an unrealisitic
unrealisitic
true
temperature
increase
and
a
strong
true temperature increase and a strong emissivity
emissivity
decrease.
decrease. These
These unrealistic
unrealistic values
values are
are due
due to
to the
the
graybody
graybody approximation.
approximation.
We notice
velocity profile
profile
We
notice in
in figure
figure 22 that
that the
the velocity
and
interface
color
temperature
profiles
present
and interface color temperature profiles present
singularities
The velocity
profile
singularities at
at the
the same
same time.
time. The
velocity profile
(A-C)
is
consistent
with
the
compression
(A-C) is consistent with the compression path
path
231
Release wave (from point C to the end)
During the unloading part of the signals, we can
notice a strong increase of color temperatures (point
D) and a small change in the shape of velocity
profile. This color temperature increase cannot be
explained by the arrival of a compression wave at
the Pb/LiF interface. In release, we know that the
absolute temperature in lead is decreasing, so point
D corresponds to a strong emissivity increase. We
think that this is due to the end of the melting
process of lead in release.
Melting curve of lead has been determined by
means of a laser-heated diamond cell up to 100 GPa
[16]. To compare our experimental results with this
experimental melting curve, we calculated the stress
profile corresponding to our color temperature
profiles. Simulation of the experiment have been
performed with a 1-D lagrangian code
UNIDIM [17]. The behaviour of materials is
described by a Mi'e-Griineisen equation of state
based on the Hugoniot and the constitutive relation
used is perfectly plastic. This calculation does not
take into account melting but stress is little changed
by melting.
We represented the release path follow by lead
in the phase diagram using calculated stress and
measured color temperatures (figure 6). Hugoniot
and melting curve indicated in figure 6 have been
calculated using a mixture law between solid and
liquid phase of lead [5]. We observe that the
temperature increase occurs when the release path
moves away from melting curve. It seems to confirm
our hypothesis that lead is partially shocked-melted.
The release path into LiF window follows the
melting curve of lead and is not sufficient to achieve
melting. Melting is achieved during the unloading
wave from the back of the impactor. The shape of
velocity profile is consistent with this explanation.
We have already observed the same kind of velocity
profile in tin when the Hugoniot point reached under
shock was on the melting curve [5].
CONCLUSIONS
Comparison between velocity and color
temperature profiles may make easier the analysis of
temperature profiles. According to our analysis of
pyrometric signals, it seems that upon initial
shocking, lead is partially melted and remains
partially melted upon partial release into LiF
window. Complete melting occurs upon release
(coming from the impactor), that induces an
emissivity increase and an increase of color
temperature signals. It would be interesting to do the
same experiments at lower shock levels with VISAR
and the appropriate pyrometer to display melting on
release of lead.
ACKNOWLEDGMENT
The authors gratefully acknowledge the technical
assistance of P. Bouinot. This work was supported
by the French Ministry of Defense.
REFERENCES
1. Al'tshuler L.V., Kormer S.B., Brazhnik M.I.,
Vladimirov L.A., Speranskaya M.P. and Funtikov A.L,
Sov. Phys. JETP 11, 4 (1960)
2. Brown J.M. and McQueen R.G., /. Geoph. Res. 91, B7
(1986)
3. Asay J.R. and Hayes D.B., /. Appl. Phys. 46, 11 (1975)
4. Urtiew and Grover, /. Appl Phys. 45, 140 (1974)
5. Mabire C., Ph D Thesis, Universite de Poitiers (1999)
6. Leal B., Ph D Thesis, Universite de Poitiers (1998)
7. Hereil P.L., Mabire C., J. Phys. IV France 10, Pr9-799
(2000)
8. Barker L.M. and Hollenbach R.A., /. Appl. Phys. 43
(1972)
9. Marsh S.P., LASL Shock Hugoniot Data
10. Bass J., Ahrens T., Abelson J., Hua T., /. Geophys.
Res., vol. 95, N°B13, 21,767-21,776 (1990)
11. Tan H., Ahrens T., High Pres. Res., vol. 2, pp 159182 (1990)
12. Svendsen B ; Ahrens T., Bass J., High Pres. Res. in
Mineral Physics, pp. 403-423 (1887)
13. Grady D.E., /. Geophys. Res., 85, 913-924 (1980)
14. Gaussorgues G., la thermorugosite infrarouge (1999)
15. Lalle P., Courchinoux R., Melting on the Hugoniot,
High Pressure Science & technology, 207-210 (1995)
16. Godwal B.K., Meade C., Jeanloz R., Garcia A., Liu
A., Cohen M., Science, vol. 248, 462-465 (1990)
17. Internal report CEG (1982)
232