Joe Kelleher Presentation (May 27th 2014)

Using  neutrons  for  in-­‐situ  observation  of  engineering  

material  behaviour    

Joe  Kelleher  Instrument  Scientist,  ENGIN-­‐X  beamline,  ISIS  

Neutrons  at  ISIS  

H-­‐  ion  accelerator  

Proton  (H+  ion)  synchrotron  

Second  target  station  

First  target  station  

Engin-­‐X  beamline  

IMAT  beamline  

The  ENGIN-­‐X  beamline  

Incident beam

South detector bank

North detector bank

Engin-­‐X  layout  Radial

collimator (defines

outgoing beam size)

Incident slits

Strain direction

(bank 1)

Diffraction detector (bank 1)

Diffraction detector (bank 2)

Strain direction

(bank 2) 45º 45º

Sample on translation / rotation table

Strain direction (transmission

detector)

Time of flight

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

16

18

20

22

24

26

28

β-Sn powder

Total cross section (barns)

Neutron wavelength (Å)

Experiment Calculation

Transmission detector

Types  of  experiment:  trends  from  1999  to  2011  

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100%

1999  (PEARL)

2002 2005 2008 2011

Strain scanning: welds

Strain scanning: other

In-situ loading, room temperature

In-situ loading, high temperature

Other in-situ processes

Physics of neutron measurement

In-situ loading, cryo-temperature

Shar

e of

use

r ex

per

imen

t ti

me

From  materials  to  processes  Magnetic fields

Electrochemical reactions

Mechanical deformation

Heat treatment

Phase transformations

Welding

Fuel cells

Corrosion

Shape memory alloys

Material forming

Stressrig  (up  to  100kN)  

Cryostat  Down  to  -­‐200C  

Optical  furnace  

Up  to  1100C  

Resistance  furnace  (small  samples,  up  to  ~1600C)  

Sample  environments  for  in-­‐situ  tests  

Ceramic  heating  pads  (larger  samples)  

In-­‐situ  heat  treatment  

Linear  weld   Ni  superalloy   Circumferential  pipe  weld  

Single  crystal  Ni  superalloy  

Anna  Paradowska  demonstrates  proof  of  principle  

James  Rolph  et  al.  Comptes  Rendus  

Physique  13(3):307–15.  (2012)  

Bo  Chen  et  al.  Acta  Materialia  submitted  (2013)  

See  γ’  misfit  as  function  of  temperature  

In-­‐situ  heat  treatment    of  pipe  weld  

2.865

2.87

2.875

2.88

2.885

2.89

2.895

2.9

0 200 400 600 800

Atom

ic  la*ce  sp

acing  /  Å  

Temperature  /   ° C

Weld... Stress-­‐free reference...

Hea8ng Hea8ng Cooling Cooling

Bo  Chen,  Alexandros  Skouras,  Yiqiang  Wang,  Joe  Kelleher,  Shu  Yan  

Zhang,  David  Smith,  Peter  Flewitt,  Martyn  Pavier  

Cyclic  voltage  on  PZT  ferroelectric  

100  MPa  applied  

Zero  load    

The  MANTID  platform  §  Data  reduction  and  visualisation  

for  all  ISIS  instruments  §  Supports  event  mode  and  

stroboscopic  data    

David  Hall,  2013  

                 Loading  direction                  Transverse  direction    

500  V  /  mm  AC  electric  field    

Cyclic  electric  field  causes  straining  of  poled  PZT,  but  applied  load  

depoles  the  PZT    

Plots  show  difference  between  +  and  –  half-­‐cycles  

–  

+  

In-­‐situ  welding  

Fatigue  crack  growth  

Practical  considerations  

• Sample  environment  –  Sufficiently  non-­‐interacting  with  neutron  beam  –  ‘Contains’  the  process  for  steady-­‐state,  safety  

• Timing  for  dynamic  effects  –  Synchronise  clocks  or  use  trigger  pulses  – Get  event  mode  acquisition  to  collect  other  data  

• Can  we  record  more  than  just  the  neutron  data?  

Supplementary  analytical  methods  

Things  that  might  change  in  a  process  

•  Mechanical  deformation,  stress  and  strain  

•  Material  ‘damage’  •  Phase  changes  •  Diffusion  •  Temperature  

Methods  that  might  reveal  these  changes  

•  Image  correlation  •  Acoustic  emission  •  Thermoelastography  •  Dilatometry  •  Calorimetry  •  Ultrasonic  and  magnetic  

methods  

•  Sometimes  possible  to  measure  these  ‘for  free’  with  existing  sensors  

Full-­‐spectrum  imaging  

• Neutron  detectors  not  intrinsically  sensitive  to  wavelength,  but  to  time  of  detection  – …hence  velocity,  hence  wavelength  

• Each  pixel  of  a  2D  detector  can  record  a  full  wavelength  spectrum  

• We  can  thus  see  both  spatial  and  temporal  variation  in  several  physical  parameters  

Time-­‐of-­‐Klight  neutron  imaging  for  in-­‐situ  studies  

Bragg  edges  show  crystal  structure  –  how  those  atoms  are  arranged  

Resonance  peaks  show  which  atoms/isotopes  present  

Wavelength  

Detected  intensity  

Height  →  Texture  

Height  →  Concentration  

Width  →  Temperature   Position  →  Strain  

Steven  Peetermans  &  Joe  Kelleher  (2013)   17

Transmission  spectrum  from  single  crystal  

Σabs+Σinc+Σinel,coh

Σel,coh Position = Orientation and strain

Width = Mosaicity

Conclusion:  Some  future  directions?  

• Broader  array  of  sensing  and  actuation  on  beamlines  –   Users  won’t  need  to  bring  their  own  

• Flexible  data  chopping  with  event  mode  –  Especially  cyclic  or  highly  dynamic  processes  

• Sensor  /  actuator  bus  for  real  time  measurement  and  control  (e.g.  CANBUS  for  vehicles)