Simulation Approaches for Sludge Rheology

1
Simulation Approaches for Sludge Rheology

• Martin Whittle Karl P. Travis
• University of Sheffield
• Scott L. Owens Mark Bankhead
• Nexia Solutions Ltd.

2
Legacy Sludge

• Legacy wastes at Sellafield Site
• Sludges in storage tanks and ponds
• Recovery/ storage operations
• Pumping/ Squeezing
• Requires an understanding of
• rheology

3
Model scales
Mesoscale methods 'bridge the gap' between
fundamental theory and continuum scale models
Use the mesoscale to develop constitutive equation
s for use In continuum models
4
Brownian Dynamics

• Particles interact with continuum fluid
• Free draining Brownian Dynamics
• Algorithm inclusion of shear flow

5
Periodic Boundaries
Characteristic time
6
Homogenous shear
Imposed linear velocity profile
Dimensionless shear rate
7
BD Problems with Imposed Shear
Fluid interaction with pores
Imposed linear velocity profile
Bridges weaker than the rest
8
Dissipative particle dynamics (DPD)

• Forces

• Conservative

• Dissipative

• Random

Each DPD particle represents a number of fluid
particles
9
Parameterisation

• Conservative terms originally through
  compressibility
• Recent methods (Travis et. al.) involve use of
  Hildebrand solubility parameters

10
Model 1 standard DPD

• Mark Bankhead Si02/H20
• a11 44.8
• a12 402.0
• a22 640.6
• X1 0.25
• X2 0.75
• Num. Density 3.1819

11
Lees-Edwards Boundaries
12
Couette flow Si02/H20 Velocity profiles at low
shear rate
13
Stress v Strain-rate Si02/H20
14
Viscosity Si02/H20
Can reach low shear rates because Viscosity is
high
15
Couette flow Si02/H20 velocity profiles at
high shear rate
Image shear rate 0.5 Sigmoid at steady state
16
Sxy vs. time at different shear rates Si02/H20
17
Viscoelasticity of Yoghurt
From Laurent Schmitt 3rd International
Symposium on Food Rheology and Structure 2003
18
Poiseuille Si02/H20
19
Poiseuille Si02/H20
20
Discussion

• Model demonstrates viscoelasticity qualitatively
  comparable with real systems
• Non-linear flow profile makes the accurate
  determination of viscosity unreliable by standard
  method.
• But this is probably realistic how is this
  taken into account in real measurements?

21
Model 2 Core-modified DPD
Distribution 50 type1 9950 type2 10000 total
Type 2 R2 0
R1rc
Type 1 R1 2rc
2R1 rc
First contact distance 1-1 5.0 1-2 3.0 2-2
1.0
h
Mass scales with particle size
22
Colloidal interaction
23
Model2 Core-modified DPD

• Colloidal system
• a11 25.0
• a12 25.0
• a22 25.0
• F0 800.0 k 10.0
• R1 2.0rc

24
Poiseuille CM-DPD
25
Poiseuille flowrate CM-DPD
26
Discussion

• Core-modified model adds realism, allows the
  possibility of interconnecting gel-forming bonds
• Awkward density profiles arise (may be realistic)
• Rheology and flow characteristics are
  viscoelastic but different to basic DPD
• Models taken together add richness to simulation
  possibilities

27
Acknowledgments
Thanks to
for funding.