Page 1
1
Flow, disintegration and lubrication of clay – sandy debris flows: from
the laboratory to the fieldAnders ElverhøiFabio De Blasio
Dieter IsslerJohann Petter Nystuen
Peter GauerCarl B. Harbitz
Jeff Marr
International Centre for GeohazardsNorwegian Geotechnical Institute, Norway
Dep. of Geosciences, University of Oslo, Norway.St. Anthony Falls Laboratory, University of Minnesota
.
Page 2
2
SHELF EDGE (STAGING AREA)• Storage Capacity• Volume of Failure• Canyon Positioning relative to delta (sequence stratigraphy)
DEPOSITIONAL CONTROLS-Basin Relief-Local Gradient-Evolving Seabed Topography
SHELF
• WIDTH
• GRADIENT
• ACCOMODATION
• SEDIMENTS
TRAPS
Controls on submarine fan facies distribution
Adopted from Statoil, Frode Hadler-Jacobsen
Page 3
3
What is the origin of deep-water sand bodies?
• Turbidity currents? (low density, dominant turbulence)
• Debris flows? (high density, laminar, 1) cohesive? 2) non-cohesive?)
Page 4
4
Experimental settingsSt. Anthony Falls Laboratory
Experimental Flume: “Fish Tank”
Video (regular and high speed) and
pore- and total pressure measurements
Page 5
5
Debris flows- high clay content
A: 32.5 wt% clay, hydroplaning front Dilute turbidity current
B: 25 wt% clay hydroplaning frontD: Behind the head, increasing concentration in overlaying turbidity current
Page 6
6
Debris flows- low clay content (5%)
Turbulente front Deposition of sand
Page 7
7
Anatomy of a sand-rich experimental debris flow
Page 8
8
Low clay content – video record
Turbidity current
Sand waves
Dense flow
Deposition of sand.
Page 9
9
Close up pictures/ sand waves
High speed video showing:
1) The turbidity current (TC)
2) Debris flow/fluidized layer (DF)
3) Settling layer
New layer of sand, stacked sequences?
Sand
wave
Page 10
10
High clay content-- Plug flow- “Bingham”
High sand content-Macro-viscous flow?-Divergent flow in the shear layer
Page 11
11
Some examples of velocity field – one image
Page 12
12
Velocity for the whole flow (proportional to
colour)
Deposited materials
”Debris flow”
Turbidity current
Convection cells ?
Hei
ght
[pix
el]
1m =
640
0 pi
xel
Time [frame] 1 frame = 0.04s
Velocity m/s
Page 13
13
Thickness of sandy deposits – versus clay content
Page 14
14
Flow behavior:Mass flow at high sand fraction
Page 15
15
Numerical experiment 1
• Transport of a single sand particle through a free surface laminar flow
• Lift and Drag force• Results: a few km
runout at most: not an efficient process
Page 16
16
Fluidized layer: what is its potential for sand transport?
Page 17
17
Numerical experiment 2: A simple model for fluidized sand
• Calculation of settling velocity and runout of an inclined granular bed.
• Settling is hindered by water expelled from beneath (Richardson-Zaki).
• The material not yet settled is modeled as a laminar free-surface flow
• What are the boundary conditions at the top?
Page 18
18
1
ut y
Numerical experiment 2Basic equations
• Richardson-Zaki
• Viscosity
• Conservation of particle number
• Laminar velocity field
max
max2
0
1 if
1 if
v u
y D
y y
ydydygyu
0 ' )''(
)''(1'''sin)(
Page 19
19
Example I:Viscosity=0.02 Pa s ( 10 x water)Thickness=0.5 mSlope=0.1 degreesLimit velocity=1 cm/s
Result: Runout between 400 m and 2 km
Results from the calculations
Example IIViscosity=0.6 Pa sThickness=2 mSlope=0.8 degreesLimit velocity=0.2 cm/s
Result: Runout between 23 km and 100 km
Page 20
20
Processes that could potentially improve sand transport in the ocean
• Turbulence (generally accepted view).
• Fluidization.
• Importance of clay, even little clay increases viscosity/improves cohesion lubricated (hydroplaning) flow
• Sand waves.
Page 21
21
Conclusions and future plans
• CONCLUSIONS• Water increases mobility • Potential for sand
transport in the dense phase 1) sand packed in the clay matrix?) 2) Fluidization?
3) Sand waves?
• Scaling problems
• FUTURE PLANS• Experiments with
different volumes and different grain sizes
• Direct numerical simulation of high-density fluidization
Page 23
23
1
ut y
Model equations
• Conservation of particle number
• Laminar velocity field
• Viscosity
• Richardson-Zaki
20 max
max
( 1 )
( 1 )
if
if
v u
y D
0 y'
1 (y '')u(y) gsin dy ' dy '' dy ''
2 (y '')
Page 24
24
What model for clay-poor debris flows?
• Some existing models: Savage-Hutter (SH), Norem Irgens Schielthorp (NIS), Iverson Denlinger (ID)
• Granular: YES Turbulence: NO• Macroviscous regime not addressed only in NIS
• Direct Models ?
Page 25
25
Prompt disitengration of the debris flow
• 1) disintegration of the mass: the yield stress drops dramatically
• 2) settling and stratification
y k exp C
solid fraction in the slurry
dependent on the clay content
Reference solid fraction
Page 26
26
Existing models: e.g.: NIS model
• Mud with plug and shear layers– plasticity, viscosity, and visco-elasticity
• dry friction (no cohesion in code)• dynamic shear (thinning)• dispersive pressure
r
xexy
r
xuey
r
xuex
dy
ydvmpc
dy
ydvpp
dy
ydvpp
)(tan
)(
)()(
2
21
Page 27
27
I-D model
• Depth integrated, three-dimensional model • Accounts for the exchange of fluid between
different parts of the slurry due to diffusion and advection.
• Limitations for our purpose: water content of the slurry must not change, no cohesion, no turbulence
2
2
p ' p ' p ' p 'u v
t x z y