Two Methods for Measuring Residual Strain in ISIS - PowerPoint Presentation

Slide1

Two Methods for Measuring Residual Strain in ISIS Targets

Dan Wilcox

High Power Targets Group, RAL, UK

ICANS XXIII, Chattanooga, October 16th 2019

Outline

Background

Asymmetric-clad strip method

Neutron diffraction method

Results and analysis

Implications for target design

Acknowledgements

Sample

manufacture –

Leslie Jones, Jeremy

Moor, Max Rowland, Peter

Webb

Asymmetric-clad

concept

– Peter

Loveridge

Instrument

scientists – Saurabh Kabra, Tung Lik Lee

ISIS management

– Ste Gallimore, David Jenkins

Background

ISIS operates two spallation targets; TS1 and TS2

Both have tungsten cores clad in tantalum for corrosion resistance

Any significant cladding breach means we have to replace the target

Causes activation of the cooling plant, which needs hands-on maintenance

Cladding is attached by Hot Isostatic Press (HIP)

Produces a strong bond with good thermal contact

High residual stress predicted from HIP – maybe more than beam heating

Exact amount of residual stress is a major unknown in current simulations*

*D. Wilcox, P. Loveridge, T. Davenne, L. Jones and D. Jenkins, "Stress levels and failure modes of tantalum-clad tungsten targets at ISIS," Journal of Nuclear Materials, vol. 506, pp. 76-82, 2018.

TS1

TS2

ISIS Target Manufacture

Tantalum can welded shut with tungsten core inside

Hot

Isostatic Press (HIP) process applies high temperature and pressure,

bonding

cladding

and core

Final machining to achieve dimensional

accuracy

Components of TS2 target HIP assembly

ISIS Target Manufacture

HIP cycle in detail:

Cladding deforms plastically under high pressure (140MPa)

Peak temperature is 1200°C, held for 2 hours; this relieves stress, but does not cause grain growth

Tantalum and tungsten are bonded

together

As the clad target cools below the annealing temperature, residual stress builds up due to different coefficients of thermal expansion:

Tungsten 4.5E-06/K

Tantalum 6.3E-06/K

A typical ISIS HIP cycle

Simplifying Assumptions

Simplifying assumptions for simulating residual stress:

All stress from welding and HIP pressure is relieved during HIPing

Stress due to final machining and plate welding is small compared to stress from HIPing, and can be ignored

Stress from HIP cooldown is relieved at first, but below a certain ‘lock-in’ temperature stress starts to builds up

Industry uses a stress relieving temperature of around 850°C

Lock-in

temperature estimated to be 500°C, but this needs validation

Asymmetric-Clad Strip Method

Provides a

simple, mechanical

method of measuring residual stress after HIPing

HIP a long strip

of

tungsten with asymmetric

tantalum

claddingThe strip should deflect in proportion to the residual stressCalculations predict a spherical surface, concave on the thick tantalum sideDimensions optimised to produce a measurable deflection without breaking the sample

Nominal dimensions of asymmetric-clad strip

Photos: Jeremy Moor

Manufacture

Manufactured

by Jeremy Moor and his team using the same materials and methods as ISIS

targets

Measurement

Sample was measured before and after HIP using a Faro Edge arm (laser scanner coordinate measuring machine)

Concave on thick tantalum side and almost spherical, as expected

Spherical surfaces fit the upper and lower faces with r

2

= 0.990

Cloud of points from Faro arm (above) and spherical fits to upper and lower surfaces (right)

Upper surface

Lower surface

Comparison with Theory

Compare fit and simulated radii of curvature

Estimates lock-in temperature at 385-450°C

To produce the measured deflection the tantalum must have yielded

Neutron Diffraction Method

ENGIN-X is an instrument on ISIS TS1 which uses neutron diffraction to make non-destructive strain measurements

“

The beamline is optimized for the measurement of strain, and thus stress, deep within a crystalline material, using the atomic lattice planes as an atomic 'strain gauge

'.”

https://www.isis.stfc.ac.uk/Pages/Engin-X.aspx

Good penetration depth even in Ta/W

1x1x18mm measurement volume

Cannot measure plastic strainStrains are calculated relative to an unstressed ‘d0’ measurement

The ENGIN-X instrument: measuring residual stress within friction stir welds on an Airbus prototype wing rib

Neutron diffraction experiment on ENGIN-X by Yanling Ma et al. on a TS1 plate*

Demonstrated that ENGIN-X can measure depths of up to 13mm in tantalum/tungsten in a reasonable length of time

*

Yanling

Ma et al., “An Experiment Using Neutron Diffraction to Investigate Residual Strain Distribution in a Hot Isostatic Pressed (HIPPED) Target Plate,” in Joint 3rd UK-China Steel Research Forum & 15th CMA-UK Conference on Materials Science and Engineering, 2014.

Prior Experiment by

Yanling

Ma et al.

Measured elastic strains are broadly consistent with the cladding being at the yield stress as predicted – however the sample was too thick to measure all three directions so this could not be confirmed

Engin

-X Strain Result (Y direction)

ANSYS Simulation of

Engin

-X Sample

≈ 600

Microstrain

662

Microstrain

-749

Microstrain

≈ -300

Microstrain

Prior Experiment by

Yanling

Ma et al.

ENGIN-X Measurements

Two experiments: October 2018 (3 days) and March 2019 (2 days)

Aims – measure a detailed through-thickness strain profile

–

check uniformity of strain profile with position

– measure all 3 strain components at every point

Clamp

this end

50mm

50mm

20mm

Z

X

Y

10mm

10mm

9 Points

3, 9, 3

Points

3

Points

ENGIN-X Measurements

Sample rotated part way through to measure all 3 strain components

Normal direction measured twice

Photos courtesy of Tung Lik Lee

ENGIN-X Results – Part 1

Similar profiles in x and y, as expected

Good repeatability between repeat measurements

Quite good agreement with simulation apart from through-thickness (z) strain in tantalum

Changes for Second ENGIN-X Run

Previous ‘d0’ samples may not have been truly unstressed

Tantalum d0 sample was rolled and annealed

Tungsten d0 sample was forged, but probably not annealed

Repeated measurement with new d0 samples

Tungsten

sample

cut into combs to relieve stress from forging

Previous W sample had higher normal strain – possibly due to forgingPrevious Ta sample the same to within the ≈100 microstrain uncertainty

Also re-measured old locations and added more measurement points

Strain in

previous

d0 samples

In Plane

Normal

Tungsten

91

-221

Tantalum

-70

20

All values in

microstrain

HIP Effects on Nominally Unstressed Samples

Second run also looked at the stress change after

HIPing

in pure tantalum and pure

tungsten

HIP may change the atomic spacing and therefore d0 measurement

An

exact copy of each d0 sample was put through a standard ISIS HIP cycle

HIPing induces a strain which is compressive in W, tensile in

TaChanges are anisotropic, which was unexpected and makes it difficult to apply d0 corrections to other geometry e.g. the strip

Strain change due to

HIPing

In Plane

Normal

Tungsten

-294

-58

Tantalum

309

83

All values in

microstrain

ENGIN-X Results – Part 2

Good

repeatability between different locations and

times

No choice of d0 measurement improves the fit for all materials and directions

Conclusions from Experiments

Presence of large tensile residual stress in cladding confirmed by physical test and neutron measurement

Simulations with lock-in temperature ≈ 400°C predict the shape of the deformed strip, but do not explain all the details seen by ENGIN-X

HIPing

produces unexpected changes in neutron diffraction strain measurements of pure

tantalum

and

tungsten

samplesDifferences in ENGIN-X data vs simulation may be due to:Large uncertainties relative to measurement (high Young’s mod of Ta/W)

Creep of tantalum during cooldown after HIPingDeformed grains or other microstructural effects => anisotropic propertiesNo stress relieving was observed over time. Residual stress may relieve in-beam, but this would be difficult to measure.

Implications for Target Design

Large residual stress confirmed, so this should be taken into account in future target analysis work

High tensile stress will reduce fatigue life of the cladding, although current ISIS targets have a large safety factor even with this included

Limited data on combined radiation damage and fatigue effects

If tantalum ductility is lost this may present a problem

Not currently an issue, but could be for potential ISIS-2 targets

Continuing to develop our understanding of failure modes will enable more optimised targets and higher beam powers in future