Snow Characterization Workshop 2008 3D (Lecture 2) Martin Schneebeli.

Snow Characterization Workshop 2008

3D (Lecture 2)

Martin Schneebeli

3D microstructural analysis

two methods for 3D microstructural analysis:

– polycut = serial sectioning– microCT = micro computer

tomography

polycut

microCT

2D-section: 3D image: serial

2D-section: 3D image: tomography

chloro-naphthalene (three-phase)

3D image: tomography-BET

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

Principle of the new method

working point

-20°C

5×10-2 mbar

Principle of the new method

snow

snow

phthalate

raw CTimages

segmented images

image processing

image inversion

"digital replica"

new method

Sample preparation: casting

• casting technique is identical as for polycut• replace the air in the pores by diethyl phthalate• diethyl phthalate: melting point -3°C, but

solidification temperature is much lower (c. -30°C)

• freeze with dry ice

Sample preparation: trimming

• at -20°C to prevent melting of the phthalate

• cut and trim samples to cylinders of 18 mm diameter

Experimental setup: vacuum

• store samples in vacuum desiccator

• high vacuum pump: final pressure ~5×10-2 mbar

• optional: put drying agent in desiccator

Experimental setup: microCT

• when sublimation terminated: microCT measurement

• same measurement protocol as for snow

Image processing

• raw CT-images: Gaussian filtered (3D)

• image segmentation: local minimum of the histogram as threshold

-2000 0 2000 4000 6000 8000 10000 12000 140000

20000

40000

60000

80000

100000

120000

140000

160000

180000

Min.4203

icecounts

grey level

air

threshold

-2000 0 2000 4000 6000 8000 10000 12000 140000

20000

40000

60000

80000

100000

120000

140000

160000

180000

Min.3691

phthalate

counts

grey level

air

threshold

air

ice air

phthahlate

Sublimation process

9

10

11

12

13

14

15

16

0 12 24 36 48 60 72 84 96 108 120

time (h)

sample weight (g)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 12 24 36 48 60 72 84 96 108 120

time (h)

normalized weight (g)

Weight loss

• vacuum pressure is a good indicator of the progress

• drying agent significantly accelerates the sublimation process

0

20

40

60

80

100

0 12 24 36 48 60 72 84 96 108 120

time (h)

pressure (a.u.)

vacu

um

pre

ssu

ren

orm

aliz

ed

weig

ht

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 12 24 36 48 60 72 84 96 108 120

time (h)

normalized weight (g)

with drying agent

Validation: samples

• special sample holder for validation measurements

• CT-measurement of snow before casting

• casting and vacuum sublimation in the CT sample holder

• measurement of the phthalate after sublimation digital replica

• comparison between digital replica and orginal snow structure

Validation: visual comparison

original snow digital replica difference image

Validation: structural parameters

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.500

5

10

15

20

25

30

35

40

45

50

55

60

65

70

frequency (%)

thickness (mm)

(a)

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.500

5

10

15

20

25

30

35

40

45

50

55

60

65

70

frequency (%)

spacing (mm)

(b)

structural parameters: density, SSA, trabecular thickness, trabecular spacing, etc.

relative errors are very small (a few percent)

trabecular thickness trabecular spacing

anaglyph images of rounded snow

anaglyph images of new snow