p fiber volume fraction (%)
TRANSCRIPT
Micro- and macro- elastic properties of tungsten fiber-reinforced tungsten
composites probed by nano-indentation and laser ultrasonics S. Ando, H.T. Lee*, J.W. Coenena, Y. Maoa, R. Kasadab, J. Rieschc, and Y. Ueda
Graduate School of Engineering, Osaka University, Suita, Osaka 565-0871, Japan a Forschungszentrum Jülich GmbH, Institut für Energie- und Klimaforschung – Plasmaphysik, Partner of the Trilateral Euregio Cluster (TEC), 52425 Jülich, Germany
b Institute of Advanced Energy, Kyoto University, Uji, Kyoto 611-0011, Japan c Max-Planck-Institut für Plasmaphysik, 85748 Garching, Germany
Presenter email : [email protected] * Corresponding author email : [email protected]
#308 17-22 June 2018, International Conference on Plasma Surface Interactions in Controlled Fusion Devices
Tungsten fiber-reinforced tungsten composites (Wf/W) are presently being
developed in the EU as next generation W materials for plasma-facing
components. They possess pseudo-ductility and can overcome some of the
limitations caused by the inherent brittleness of pure W. Material properties must
be well characterized to allow detailed component design and analysis.
The macroscopic mechanical properties of composites depend critically on the
microscopic interplay of the matrix, interface, and fiber.
If Wf/W composites are to be used as plasma-facing materials, the effect of
hydrogen plasma exposure on the mechanical properties needs to be clarified.
Introduction
Purpose
Conclusion
2) Nano-indentation method
Load Coil
sample
Capacitance gauge
Leaf spring
Berkovich tip
(SEM observation)
10 μm
To characterize elastic properties of Wf/W by elucidating the interaction of W-
matrix and W-fiber and to investigate the effects of hydrogen inclusion on
mechanical properties
• Agilent Tech, Nano Indentor G200
• CSM mode
• Depth-dependent hardness/modulus
measurements
Matrix of 3 by 4 points or line scan of 15 points were
taken with 100 μm separation
The hardness or modulus data between 1000-1500 nm
was averaged to avoid “surface” effects
1) Laser ultrasonic method
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0
5
10
15
20
25
30
17 mJ
5 mJ
Dis
pla
cem
ent (n
m)
Time (s)
8 mJ
Pulse response
delay time of
electronics ~150 ns
Front surface
Nd:YAG Pulsed Laser Laser
Vibrometer
sample
Vacuum chamber Photo Detector
Increasing excitation laser power results in shift from thermoelastic to ablation regime
Longitudinal and shear wave velocities are determined from time of flight
Propagating distance are determined from sample thickness measured by micro-meter
Propagating time are measured by delay time of first and reflected wave
Excitation laser (Nd: YAG 1064 nm)
• Pulse FWHM ~15 ns
• Beam focused to ≤ 1 mm diameter
Polytec vibrometer
• He-Ne laser (632 nm)
• Surface displacement ±75 nm
• Surface velocity up to 3 m/s
• Bandwidth up to 24 Mhz
0.0 0.5 1.0 1.5 2.0-18
-12
-6
0
6
12
18
3L wave
Dis
pla
cem
ent (n
m)
Time (s)
L wave
S wave
Ph
oto
dio
de
Sig
na
l (a
.u)
Back surface
Results & Discussion
0 500 1000 1500 2000100
200
300
400
500
600
60% fiber Wf/W
matrixModu
lus
(GP
a)
indentation depth (nm)
fiber
4) Deuterium irradiation effect on elastic properties of Wf/W composite samples
4) D irradiation experiments Incident flux :
Vacc:
Time :
Fluence :
Irradiation temp :
1.0 × 1020 𝑚−2𝑠−1
1 𝑘𝑒𝑉
1800 𝑠 1.8 × 1023 𝑚−2
20℃ 293 K , 500℃(773 K)
sample
Irradiated at 293 K
Irradiated at 773 K
Φ 5 mm
W powder grain: 5 μm
W-fiber length: 2.4 mm
W-fiber diameter: 0.24 mm
Multi fibers / random distribution
W-fiber volume fraction: 20% - 60%
Compared to reference W from
ALMT Corp, Japan
Method: Spark Plasma Sintering (SPS)
Pulsed DC current directly
passes through and heats
the graphite die
High heating or cooling
rate enables to sinter metal
composites
3) Wf/W composite samples
100 μm
(50 % fiber) Unirradiated
area
Irradiated area
(773 K)
Summary: The purpose of this research was to determine the effect of fiber volume fraction on the elastic
properties of Wf/W samples manufactured using SPS technique. Laser ultrasonic measurements indicate
change in bulk modulus between 20-40% fiber volume fraction, while no significant difference was observed
between 40-60% fiber volume fraction. To elucidate the dominant contribution on the micro-scale, nano-
indentation measurements were performed to measure the strength of fiber and matrix separately.
Conclusions
• Using a simple mixture model considering fiber shortness, we find good agreement between laser ultrasonics
and nano-indentation measurements – suggesting bulk properties can be described/estimated by linear
superposition of fiber and matrix strengths scaled by the fiber volume fraction.
• Using the present synthesis method, 40~50% fiber volume fraction is highest Young’s modulus due to poorer
matrix properties at higher fiber volume fraction. Improved synthesis method to increase the matrix strength or
using longer fibers may provide a realistic path towards creating stronger Wf/W composite materials
approaching pure-W properties in the elastic regime with the additional benefit of pseudo-ductility.
Further work required: The effect of D-irradiation on the elastic properties was within the scatter of the data.
0 10 20 30 40 50 60 70 80 90 100300
350
400
450
Considering finite fiber
length of 2.4 mm (l=0.88)
Ignoring finite fiber length (l=1)
Average value of matrix modulus
Laser Ultrasonic
Young'
s m
odu
lus
(GP
a)
fiber volume fraction (%)
Nano Indentation
Average value of fiber modulus
Difference due toweak matrix modulus of 60 % sample
1) Longitudinal and shear wave velocities
from laser ultrasonic
10 20 30 40 50 60 70 80300
350
400
Young'
s m
odu
lus
(GP
a)
fiber volume fraction (%)
detection
excitation
Data for all 9 measurement spots
were averaged
Wave velocities do not scale with
increasing fiber volume fraction
Wave velocity differences
between pure W samples arise
due to density (porosity)
Poisson’s ratio is expressed as a
function of CL/Cs
Poisson’s ratio of Wf/W are
similar to pure W samples
fiber
matrix
100 μm
0 500 1000 1500 2000200
300
400
500
600
40 % fiber
Modulu
s (
GP
a)
Displacement Into Surface (nm)
60 % fiber
2) Matrix and fiber modulus from nano-indentation
0 20 40 60 80 100250
300
350
400
450
500
Considering finite fiber lengthE = lEfVf+Em(1 - Vf)
matrix modulus (Em)Young'
s m
odu
lus
(GP
a)
fiber volume fraction (%)
E = EfVf+Em(1 - Vf)
average fiber modulus (Ef)
Ignoring finite fiber length
0 20 40 60 80250
300
350
400
450
500
Youn
g's
modu
lus
(GP
a)
fiber volume fraction (%)
fiber part
matrix part
440 GPa except for
40% sample
40 GPa lower for 40 %
fiber sample
Reasons for this
discrepancy is unclear
– Raw data qualitatively
different
Suggests fiber surface
properties are different
Fiber Matrix 310 GPa at all samples
Standard deviation by
variation of measurement
spot are 15 GPa
3) Comparison of modulus by two methods
as a function of fiber volume fraction in sample
→ indicates Wf/W samples can be
treated as isotropic at macro level
𝐸 = 𝜌𝑐𝑠23𝑐𝐿
2 − 4𝑐𝑆2
𝑐𝐿2 − 𝑐𝑠
2
Relationship between wave velocities and elastic constants for isotropic material
𝐸 ∶ 𝑌𝑜𝑢𝑛𝑔′𝑠 𝑚𝑜𝑑𝑢𝑙𝑢𝑠 𝜌 ∶ 𝐷𝑒𝑛𝑠𝑖𝑡𝑦
𝑐𝐿 ∶ 𝐿 𝑤𝑎𝑣𝑒 𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦
𝑐𝑆 ∶ 𝑆 𝑤𝑎𝑣𝑒 𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦
[1] Yukio SANOMURA, “Prediction of Young’s Modulus and
Strength of Extrudate in Short-Fiber-Reinforced Polypropylene”
𝐸𝑐𝑜𝑚𝑝 = 𝜂𝑙𝜂𝑜𝐸𝑓𝑉𝑓 + 𝐸𝑚 1 − 𝑉𝑓
Prediction of Young’s Modulus in
short-fiber composite [1]
𝑉𝑓 ∶ 𝑣𝑜𝑙𝑢𝑚𝑒 𝑟𝑎𝑡𝑖𝑜 𝑜𝑓 𝑓𝑖𝑏𝑒𝑟
𝐸𝑓(= 430 𝐺𝑃𝑎) ∶ 𝑓𝑖𝑏𝑒𝑟 𝑚𝑜𝑑𝑢𝑙𝑢𝑠
𝐸𝑚 (= 310 𝐺𝑃𝑎):𝑚𝑎𝑡𝑟𝑖𝑥 𝑚𝑜𝑑𝑢𝑙𝑢𝑠
𝜂𝑙 = 0.88 ∶ 𝑆𝑡𝑖𝑓𝑓𝑛𝑒𝑠𝑠 𝑟𝑒𝑑𝑢𝑐𝑡𝑖𝑜𝑛 ratio due to shortness of fiber 𝜂𝑜 = 1 ∶ 𝑆𝑡𝑖𝑓𝑓𝑛𝑒𝑠𝑠 𝑟𝑒𝑑𝑢𝑐𝑡𝑖𝑜𝑛 ratio due to fiber orientation
Laser ultrasonics
measurements probes
the integrated properties
along the transport path.
Modulus improvement
by fiber are observed
between 20 – 40 % fiber
volume fraction
No significant difference
between 40 – 60 % fiber
volume fraction
Averaged data for all 10 measurement spots at
fiber or matrix in each samples is shown
Each measurement spot was averaged at a depth
from 1 μm to 1.5 μm
5 mm
20 mm
The values from the two methods agree within experimental error
except for 40 % fiber volme fraction sample – suggests simple mixture
model is sufficient to characterize bulk properties
Monotonic increase in strength was not observed for 60 % fiber
volume fraction sample – suggests that optimum levels for present
synthesis method is between 40-50% fiber volume fraction
Improved bulk elastic properties of WfW composite samples may be
possible by increasing matrix strength or by using longer fibers
Experimental setup
0 200 400 600 800 10000
200
400
600
60% fiber Wf/W
Modu
lus
(GP
a)
Displacement into surface (d) (nm)
Ignored region(d < 50 nm)
Smoothed line by percentile filter
100 200 300 400 500200
300
400
500
20 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)
100 200 300 400 500200
300
400
500
50 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)100 200 300 400 500
100
200
300
400
500
600
700
20 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)
100 200 300 400 500200
300
400
500
50 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)
Raw data / smoothing Comparison between fiber / matrix
D irradiation conditions
D more limited
to surface
Enhanced D
bulk diffusion
Simple mixture model suggests
strength monotonically increases
with fiber volume fraction
matrix matrix
100 200 300 400 500300
400
500
600
Irradiated
Modu
lus
(GP
a)
Depth into Surface (nm)
20 % fiber sample
Unirradiated
Unirradiated
Irradiated
D-irradiated at 293 K
100 200 300 400 500300
400
500
600
50 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)
D-irradiated at 773 K
100 200 300 400 500300
400
500
600
20 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)
Unirradiated
Irradiated
100 200 300 400 500200
300
400
500
20 % fiber sample
Modu
lus
(GP
a)
Depth into Surface (nm)
fiber fiber fiber fiber
matrix matrix 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 --
200
300
400
500
600
D Irradiated at 293 KD Irradiated at 773 K Unirradiated
Modulu
s (G
Pa)
Displacement from surface (m)
There is great variability/scatter in the
material properties of a single sample
depending on location
To quantify the effects of D irradiation, an
attempt was made to examine relative
changes in the modulus in a local region
separating the irradiated and unirradiated
areas (both fiber and matrix)
Representative data of modulus
determined from nano-indentation is
shown
Irradiated and unirradiated areas
correspond to red and black lines,
respectively
D-irradiation could result in an increase
or decrease or no changes in the
hardness/ modulus measurements for
both fiber and matrix
2600 2650 2700 2750 2800 2850 29004700
4800
4900
5000
5100
5200
5300
(→poisson's ratio )
■ SPS-Wf/W
■ SPS-W
60 % fiber
2000 C annealed
Ref-W(ALMT)
1800 C annealed
Longi
tudin
al w
ave v
elo
city
(m/s)
Shear wave velocity (m/s)
30 % fiber
50 % fiber
40 % fiber
CS/CL = 0.55
fiber
Averaged values of modulus as
a function of depth are shown
for unirradiated and D-irradiated
samples (293 K, 773 K)
The error bar shows the range
of scatter between 10
measurement spots
Effect of D is within the scatter
of the data fiber
𝜈 =1 − 𝐶𝐿 𝐶𝑆 2
2{ 𝐶𝐿 𝐶𝑆 2−1}
Modulus as function of depth