visualising internal erosion mechanisms using transparent soil ·
TRANSCRIPT
![Page 1: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/1.jpg)
E. T. Bowman
University of Sheffield, UK
Visualising internal erosion mechanisms
using transparent soil
Elisabeth T. Bowman
Department of Structural & Civil Engineering
University of Sheffield, UK
Workshop on Seepage Induced Geotechnical InstabilityImperial College London, 31 August - 1 Sept 2017
![Page 2: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/2.jpg)
• Internal erosion via Skempton & Brogan (1984) experiments
• Experimental techniques
– Transparent soil via Refractive Index Matching
– Plane Laser Induced Fluorescence
– Box permeameter
• Comparison with tests on soil
• Visualisation of internal erosion
• Image processing
• Future work
Contents
![Page 3: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/3.jpg)
Internal Erosion – dams, canals, dikes & levees
Internal erosion (or suffusion) responsible for ~50 % of embankment dam failures globally
- Predominantly a problem in older or smaller earthworks
- Work focused on establishing erosion criteria
- Work focused on understanding mechanisms of internal erosion
Gap-graded soil in which fine particles can migrate
through voids = suffusion (Rosenbrand, 2011)
![Page 4: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/4.jpg)
Skempton & Brogan (1994) experiments
Performed tests on internally unstable sandy gravels to compare theoretical value of critical hydraulic gradient at which piping occurs ic under an upward flow (according to Terzaghi, 1925) with the actual hydraulic gradient icr.
Proposed that the erodible fine grains carry some (reduced) proportion of the overburden load
The critical gradient for piping in the fine grains is then:
𝑖𝑐𝑟 = 𝛼𝛾′
𝛾𝑤or 𝑖𝑐𝑟 = 𝛼𝑖𝑐 where icr is the critical hydraulic gradient
observed in the test
From this: a larger α will yield a greater resistance to the onset of seepage-induced instability.
![Page 5: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/5.jpg)
Visualising internal erosion
Laboratory based
Using a box-shaped permeameter
○ Initial experiments based on Skempton & Brogan (1994) design
Glass particles and optically matched immersion oil
○ Replicating soil and water
Variable particle grading and hydraulic head
Refractive index matching (RIM) & planar laser induced fluorescence (PLIF) techniques
(High) speed imaging
Image processing
![Page 6: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/6.jpg)
Refractive Index Matching & seepage scaling
Refractive
index at
589.3 nm
Density
at 25 °C
Kinematic viscosity
at 25 °C
(g/cm3) (cSt)
hydrocarbon
oil1.4715 0.846 16
Duran glass 1.4718 2.23 -
Matched!
Ratio Gs = 2.65!
Scale particles ~4
times up!
![Page 7: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/7.jpg)
RIM with transparent soil & PLIF
Transparent materials
– Solids & fluid with same refractive index (RIM)
Planar laser induced fluorescence (PLIF)
– Fluorescent dye in fluid
– ~1mm thick laser sheet (532nm) to illuminate plane
– Particles appear dark against bright fluid background
7
![Page 8: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/8.jpg)
Particles used
Duran borosilicate glass, irregular shape:
• Cut / crushed rods (4mm to 30mm)
• Crushed tubes (4mm to 150mm)
dparticle> 4mm dparticle< 4mm dparticle= 300mm
![Page 9: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/9.jpg)
Crushing / breaking & sieving
Saturating with immersion oil
Careful placement into permeameter(no air bubbles!)
Sample preparation
![Page 10: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/10.jpg)
Skempton & Brogan (1994) permeameter method
![Page 11: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/11.jpg)
Header tank
High speed camera
Permeameter
Laser
Fluid recirculation system
Winch for header tank
Manometer rule
11
![Page 12: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/12.jpg)
0
20
40
60
80
100
0.01 0.1 1 10 100
Per
cen
tag
e b
y w
eigh
t fi
ner
th
an
D (
%)
Grain Size, D (mm)
S&B-A (Soil)
Scaled up S&B-A
GS&B-A (Glass)
Gap graded susceptible soil – Grading “A”
![Page 13: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/13.jpg)
0.0
0.1
0.2
0.3
0.4
0.5
0 0.1 0.2 0.3 0.4 0.5
Flo
w V
elo
cit
y,
v (c
m/s
)
Average hydraulic gradient, iav
slight movement of
fines in voids and
along glass walls
slight general
movement
icr = 0.248
moderately strong piping
through centre and sides
of sample
k = 1.6cm/s
strong general
piping
k = 3.3cm/s
Test results – hydraulic gradient vs seepage velocity
Critical hydraulic gradient = 0.25; Alpha factor = 0.21Skempton & Brogan (1994), sample A
Critical hydraulic gradient = 0.2; Alpha factor = 0.18
![Page 14: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/14.jpg)
0.0
0.1
0.2
0.3
0.4
0.5
0 0.1 0.2 0.3 0.4 0.5
Flo
w V
elo
cit
y,
v (c
m/s
)
Average hydraulic gradient, iav
slight movement of
fines in voids and
along glass walls
k = 0.26cm/s
slight general
movement
icr = 0.248
moderately strong piping
through centre and sides
of sample
k = 1.6cm/s
strong general
piping
k = 3.3cm/s
Refractive Index Matching
a) iav = 0
b) iav = 0.153
c) iav = 0.248
d) iav =0.286.
e) iav = 0.381
![Page 15: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/15.jpg)
0
10
20
30
40
50
60
70
80
90
100
0.01 0.1 1 10 100
Pass
ing
by m
as
s (%
)
Grain Size D (mm)
S&B-A
S&B-B
S&B-D
F&M G4-C
GS&B-A
GS&B-B
GS&B-D
GF&M G4-C
GS&B A/B
Test suite - PSDs
![Page 16: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/16.jpg)
16Glass – oil tests GS&B-A GS&B-B GS&B-D G-G4-Cn 0.27 0.27 0.29 0.266kinitial (cm/s) 0.30 1.00 0.18 0.02
ic 1.19 1.16 1.16 1.20
icr 0.25 0.300 / 1.01 1.31 0.72
a 0.21 0.26 / 0.85 1.13 0.60
Failure mode Piping Piping / heave Heave Piping with suffusion / volume change
Soil - water tests S&B-A S&B-B S&B-D G4-C
n 0.34 0.37 0.365 0.24
kinitial (cm/s) 0.45 0.84 1.80 0.022
ic or igc 1.09 1.04 1.05 53
icr or igcr 0.20 0.34 1.0 9.1, 8.0
a 0.18 0.33 0.95 0.34, 0.30
Failure mode Piping Piping Heave Piping with suffusion / volume change
Test results – comparison with soil
![Page 17: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/17.jpg)
“Suffusion”:Internal erosion without structure collapseInternal erosion sequence
Pipe formation:Glass S&B – grading “A”
17
![Page 18: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/18.jpg)
“Suffosion”:Suffusion leading to structure collapse
Fannin & Moffat (2006)Material Glass “4C”
18
![Page 19: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/19.jpg)
Image issues: degradation with laser / image depth
10mm 20mm 30mm 40mm
Loss of overall clarity with depth due to mismatched RI and impurities (dirt, air bubbles etc)
Loss of brightness along laser due to mismatched RI and impurities
![Page 20: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/20.jpg)
Quantitative image analysis of “open” void space
1
2
3
a b c
(a) Original image (b) ½ image (c) Divisions (a) Original
image
(b) Large
particles
(c) Large
void space
![Page 21: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/21.jpg)
Void space analysis: GS&B – A/B results
0.0
0.2
0.4
0.6
0.8
1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Flo
w v
elo
city
, v (
cm/s
)
Average hydraulic gradient, iav
n = 28.2%(H/F)min = 0.27
Minor movement of
fines along glass
edge k = 0.033cm/s
Slight-moderate
movement of fines
throughout k = 0.042cm/s
Moderate movement of
fines, minor movement of
smaller, coarse particles
throughout k = 0.072cm/s
Initiation of piping along glass
wall. Washing out of fines
between clasts k = 0.142cm/s
12
3
4
5
6
0
2
4
6
8
10
12
14
16
18
1 2 3 4 5 6
Pe
rce
nta
ge a
rea
of
op
en
vo
ids
(%)
Phase
Section 1
Section 2
Section 3
Upward seepage velocity vs hydraulic gradient Open void vertical migration with hydraulic gradient
21
![Page 22: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/22.jpg)
E. T. Bowman
University of Sheffield, UK
Conclusions => “Seeing is believing”
Transparent soil permeameters
Allow internal erosion mechanisms to be visualised,
internal to the transparent soil
Similar results to those on real soil
Image analysis on particle fabric shows fines migration
References
Hunter, RP & Bowman, ET “Visualisation of seepage induced suffusion and suffosion within internally erodible granular media”
submitted to Géotechnique
Hunter, RP & Bowman, ET (2015) “Visualisation of internal erosion of a granular material via a new transparent soil
permeameter” 16th European Conference on Soil Mechanics and Geotechnical Engineering, 13-17 Sept. 2015, Edinburgh.
Hunter, RP (2012) “Development of Transparent Soil Testing using Planar Laser Induced Fluorescence in the Study of Internal
Erosion of Filters in Embankment Dams” University of Canterbury MSc Thesis, New Zealand
![Page 23: Visualising internal erosion mechanisms using transparent soil ·](https://reader035.vdocument.in/reader035/viewer/2022081600/6048e07988fbe118e1066861/html5/thumbnails/23.jpg)
E. T. Bowman
University of Sheffield, UK
Future work
Transparent soil rigid permeameter (with Jonathan Black and Nicoletta Sanvitale) Further work to refine equipment and methods
○ Precision slices using stage micrometer
Further work on material behaviour
○ Compare behaviour of spherical and angular particles (compare to DEM)
○ Fluid tracking using neutrally buoyant particles (PIV)
Transparent soil triaxial permeameter (with Fahed Gaber, Jonathan Black and Nicoletta Sanvitale)
More complex stress states
○ Influence of erosion on strength / deformation
Visualisation of erosion