Laser Doppler Vibrometer tests

Goran Skoro

UKNF Meeting

7-8 January 2010

Imperial College London

UKNF Target Studies Web Page: http://hepunx.rl.ac.uk/uknf/wp3/

Current pulse – wire tests at RAL

Tantalum wire – weak at high temperatures

Tungsten – much better!!!

The Finite Element Simulations have been used to calculate equivalent beam power in a real

target and to extract the corresponding lifetime.

Energy deposition – current pulseAr

bitr

ary

unit

s

Time (ms)

FittedMeasured

Fit: exponential and linear functions,then analytic solution for current

density across the wire as a function of time

1 21

)(

02

0

20

0

)(

/2/

/1),(

2

n nn

n

ta

k

nt

aJ

a

earJa

eaJ

rJjtrj

n

teII 10

For example, exponential rise of

current:Current density is:

Energy deposition ~

integral of j2

Current pulse shape + Lorentz force induced pressure wave

Laser beam

Laser beam

Wir

e

We are in other room

Hole in the wall – to monitor…

…to measure temperature

Video camera to monitor laser beam position

Remote control to change it

Tungsten wire at 2000 K

Shock test Lab

Test wire - to illustrate a scale

Laser Doppler Vibrometer (LDV)

3 different decoders: VD-02 for longitudinal, DD-300 and VD-05 for radial oscillations

Wire diameter [mm]

Current < 10 kA

- excluded

Shock ~ NF target

Freq. and Amp. range

0.5 mm diameter tungsten wire

Wire Laser beam

Longitudinal oscillations

4.5 cm length

E

ln

lcnfn 4

)12(4

)12(

c – speed of soundl - wire length

Fundamental frequencies

After error propagation:

kHz )3.03.24(5.40 cmf

E – Young’s modulus - density

GPa )5368(5.4 cmlE

Corresponding frequency spectrum

• Oscillations are complicated Details not fully reproduced, but studies continuing

0.3 mm diameter tungsten wire

Wire Laser beam

3.9 cm length

Shorter wire – higher frequency:

kHz )3.03.28(9.30 cmf

Comparison between 2 different FFT algorithms.

Longitudinal oscillations

GPa )9376(9.3 cmlE

Our results for E vs. literature data?

Next slide

NuFact target will ‘oscillate’ in ~ this

frequency range

Tungsten Young’s modulus at room temperature

Comments about literature data:- mostly tungsten sheets

used- static (tension) techniques- dynamic techniques- no errors givenDifferent results even for the ‘same’ samples

For example, black points: 6 samples from the same manufacturer –> 10% difference between Young’s modulus values

(sheets rolled from ingots which are pressed from powder and consolidated by sintering)

We have tungsten wires: manufactured in different way

Radial displacement as a function of energy deposition (0.3 mm diameter wire)

Wire

Laser beam

Peak displacement value – nice agreement between experiment and simulationDifferent shape (as a function of time) – strongly depends on measurement’s position along the wire

Frequency of radial oscillations In experiment, we see it only here

Wire length = 3.9 cm

f = 11 MHz (crude estimate) f = 11.3 MHz (LS-DYNA) Hard to measure it for such a tiny wire!

Better for 0.5 mm diameter wire (next slide)

Radial oscillations

DD-300 decoder

Frequency of radial oscillations as a function of energy deposition (0.5 mm diameter wire)

Wire

Laser beam

In almost perfect agreement with expected value (from LS-DYNA)

2% d

iffer

ence

Radial oscillations

DD-300 decoder

But, DD-300 is in saturation at higher temperatures

(displacements outside the range)

Frequency of radial oscillations as a function of temperature (0.5 mm diameter wire)

VD-05DD-300

Radial oscillations

Different decoder…

0.5 mm diameter wire

Radial oscillations

Wire

Laser beam

• Correct velocity is reproduced (level of stress is correct)

• Lifetime results are valid

VD-05 decoder

• Details not fully reproduced

• Frequency is OK!

0.38 mm diameter wireRadial oscillations

Wire

Laser beam

• wire is being stressed at above NF levels

VD-05 decoder

per pulse1000 pulses – no problem

Then 1000 pulses at 3x higher stress than at NuFact, even higher…

Young’s modulus of tungsten as a function of temperature (0.5 mm diameter wire)

Doesn’t depend on shock!

If we know the frequency f, Poisson’s ratio m, density , root of

corresponding Bessel function and wire radius r then:

CT 5100893.1279.0m

2843 102448.2103786.23027.19/ TTcmg

mm

2112 22

22

rfE

Radial oscillations

Doesn’t depend on shock!

Radial oscillationsYoung’s modulus of tungsten as a function of temperature (0.75 mm diameter wire)

Bonus: Measurements of Young’s modulus of tungsten

E difference believed simply to be because different wire

samples have different E

J.W. Davis, ITER Material Properties Handbook, 1997, Volume AM01-2111, Number 2, Page 1-7, Figure 2

Concern: low

strength from static

measuremts at

high temp

Young’s modulus

remains high at high

temperature & high stress!

Conclusions

* Note the different time scale

Shock measurements:

Measurements of tungsten properties: comparable with existing;

LS-DYNA predictions: confirmed (details still being understood);oscillations are complicated, wire partially fixed to frame,

frame also moves, etc

Bottom line: Lifetime results are valid

Wire is being stressed at above NuFact levels.

Plans:

Continue detailed studies;

Repeat lifetime tests, but measure with LDV over time;

Use beams and measure with LDV.

Papers: 1st – LDV results (‘material’ journal – Journal of Nuclear

Materials?)2nd – lifetime/fatigue tests, shock at the NuFact, etc… (NIM

B ?)