Rail Research UK teamwork in action

1
Effects of Construction Method on Embankment
Stability
Owen Davies
2
Aims
To assess the influence of embankment
construction method on pore pressure distribution
and stability and to predict future climatic
influence
3
Hypotheses to test

• Historically constructed embankments undergo
  large deformations.
• Pore water pressures within these embankments
  will increase quickly due to increased
  infiltration of water (higher permeability)
  combined with consolidation.
• Historically constructed embankments are less
  stable.
• Compacted embankments have higher initial pore
  pressures than tipped embankments.
• Settlements of compacted embankments are lower
  than tipped embankments
• Compacted embankments are more stable.
• Future predicted climates will cause higher pore
  water pressures within embankments and therefore
  reduce infrastructure stability.

4
Pore pressures

• Pore water pressures within an embankment are
  recognised to be an important factor in assessing
  the condition of an embankment in particular
  embankments of clay
• Within a saturated soil pore pressures are
  positive and will increase with depth
• Within a partially saturated soil negative pore
  pressures will be present these pressures are
  termed soil suction.

5
Pore pressures
The profile of shallow, surface negative pore
pressures and deeper positive pore pressures
exists within most embankments.
A. Ridley, B. McGinnity and P. Vaughan (2004)
Role of pore water pressures in embankment
stability Geotechnical engineering. 157 GE4 pp
193-198
High soil suction developed during long dry
spells and subsequent high positive pressures due
to prolonged wet periods often trigger movements
that can lead to failures
6
Modeling
To study the effects of construction technique on
embankment stability we are developing a
staggered coupled hydrological and geotechnical
numerical model.
The hydrological model will use climatic data
input to predict the pore pressure changes within
the embankment.
The geotechnical model will be able to calculate
the mechanical deformation caused by the pore
pressure changes.
The two models used for this study are SHETRAN
and Flac tp flow
With such a model we will be able to model annual
shrink swell cycles and progressive failure
mechanisms for both compacted and tipped
embankments.
7
Why couple?

• The SHETRAN model is a well established
  hydrological model capable of modelling
• subsurface flow
• overland flow
• canopy interception and drip
• evaporation

Flac two phase flow is capable of modelling both
saturated and partially saturated subsurface
flow. However Flac tp flow cannot model the
surface boundary condition.
8
The modelling process
A simulation is first run within the SHETRAN
programme and a file created holding the daily
pore water pressures for the centre of each
surface cell (shown in red).
9
The modelling process
These pore pressures are then imported into a
flac tp flow model which has an identical initial
pore pressure condition and hydraulic properties.
10
Model test results
A provisional test model has been run for 4
years. The model had initial dry conditions and
the charts below show a history of pore pressure
within the embankment for both the SHETRAN
simulation and the Flac tp flow simulation.
11
Model test results
12
Model test results
A fully coupled simulation has also been run
within Flac tp flow using a simple Mohr coulomb
constitutive model. The following shows the
development of shear strains at the toe of the
slope after a 9 year simulation
13
Next stage
Use a more advanced constitutive model, either
the in-built strain softening model or Cam Clay
model Input BIONICS soil properties for the both
hydrological model and mechanical model Refine
grid for optimum modelling capabilities