The Electron Microscopy Unit

1
The Electron Microscopy Unit

• Started by Dr Sally Stowe 15 years ago
• Includes 10 columns across campus
• 8 staff mostly in centralised facility
• Interfaces with research infrastructure _at_ ANU
• now looking to consolidate and to bring in
  emerging technologies
• Case study X-ray micro-Tomography

2
The group effort is directed towards the
fundamental understanding of disordered and
porous materials with the aim of predicting
properties via imaging, visualisation,
theoretical and computational modelling.

• Foamed materials

• Sedimentary Rock

• Fibre-cement composite

• Bone

• Implants

• Wood

• Christoph Arns
• Ji-Youn Arns
• Tomaso Aste
• Holger Averdunk
• Andrew Kingston
• Fabrice Bauget
• Stephen Hyde
• Anthony Jones
• Alexandre Kabla
• Mark Knackstedt
• Shio Inagaki
• Shane Latham
• Yildiray Cinar (UNSW)
• Sacha Igorsky (UNSW)
• Munish Kumar

• Ajay Limaye
• Val Pinczewski (UNSW)
• Stuart Ramsden
• Vanessa Robins
• Peter Wood
• Mohammad Sadaatfar
• Arthur Sakellariou
• Tim Sawkins
• Tim Senden
• Adrian Sheppard
• Rob Sok
• Michael Turner
• Abid Ghous
• Trond Varslot
• Jill Turner

group leaders
15 experiment - 85 theory
3

• Computed Tomography
• 2-3 micron resolution
• high brilliance
• 30 - 225 kV polychromatic
• 1.1x to 120x magnification
• supports 120kg to milli-degree
• Pb lined room
• 16 bit, 2048 x 2048 cooled CCD
• max. sample 55 mm diameter

• Second facility finished end-07
• 4-port 2K CCD
• 400 fps camera
• lt1 micron source
• Vertical radiography
• triaxial cell (400 atm)
• Third facility just installed in India
• Forth to be installed at UNSW next year

4
Common garden skink Multiple radiographs (2D)
at different viewing angles. Typically, (2048)2
pixels proj. 3000 projections 8 Mbytes /
projection 15 - 48 hours 24 Gbytes (1024)2
pixels proj. 1600 projections 2 Mbytes /
projection 3 - 28 hours 3.1 Gbytes
DATA COLLECTION
5

• 2D radiographs are transformed into true 3D
  dataset.
• Beer-Lambert law
• Linearise data
• Modified Feldkamp
• 10243 dataset
• 32 CPUs 20Gb RAM
• 30 mins
• 20483 dataset
• 128 CPUs 170Gb RAM
• 1.5 hours
• APAC
• 1600 2GHz CPUs
• 1000 Gbytes RAM

RECONSTRUCTION
6
Neural Plasticity Bee brains grow when they
learn. 600 microns wide - 10 micron voxels
Collaboration with Universität für
Humanwissenschaften, Liechtenstein
7
Air breathing
Sensory organs in snout of Gogonasus 384 Mya 1 x
2 x 2 mm chunk
Collaboration with Museum Vic Monash Uni
8
Drishti - runs natively on PC/Mac OSX/Fedora
http//sf.anu.edu.au/Vizlab/drishti/
9
Motivation use tomograms as frameworks for
simulation of material properties in porous and
granular materials.

• Segmentation (binarisation of datasets)
• NMR diffusion
• Network modelling
• Medial surface/axis studies
• Topology (Minkowski measures)
• 2 and 3-phase fluid penetration simulation
• 2 and 3-phase Permeability
• Conductivity (heat flow simulators)
• Sonic velocities
• Linear elastic isotopic FEA
• Anisotropic FEA
• Upscaling statistics
• Visualisation (ray tracing / volume rendering)
• Tracking in deformabale/granular materials
• Fibre and irregular grain detection

Suite of parallel code
Quantitative imaging
10
Core/cut
5 mm3
1 mm3
Can we get multiple information from single sample
11
Permeability calculations from 6 reservoir
cuttings.
In collaboration with
Basis of Consortium with 14 multinational oil
companies
12
From solution to equations we can measure local
transport
Fluid velocity map in pores of coral (0.5 mm)
13
Grain IdentificationRock Fabric and
TexturePetrophysical modelling
Can partition both pore and grain space
14
An invitation

• Visit us
• Download Drishti
• help us to expand tools for 3D analysis