You are on page 1of 16

SETTLEMENT CHARACTERISTICS OF DEEP ENGINEERED FILLS

Peter J. Waddell and Patrick K. Wong


Coffey Geosciences Pty Ltd

1 INTRODUCTION
The long-term settlement characteristics of engineered fills are generally not a significant concern where the fill
thickness is less than a few metres. With the increasing scarcity of land for urban and commercial development and the
availability of large volumes of excavation materials, sites are being considered for development that previously would
have been filled with refuse of other uncontrolled material and left undeveloped. For such sites, including backfilled
quarries, the long-term settlement characteristics of the fill becomes an important consideration.
In this paper the settlement of engineered fill has been characterised as having four potential components:
1. Short-term Settlement, which occurs due to self-weight as the fill is placed and for a relatively short time
after fill has reached full height.
2. Elastic Settlement, which occurs in the fill when subjected to loads from footings and floor slabs.
3. Long-term or Creep Settlement, which occurs over a period of years. In the case of deep fills with light
building loads, the creep due to the self-weight of the fill will be the major component of the long-term
settlement.
4. Hydroconsolidation (Collapse) Settlement, which can occur and is due to saturation of the fill.
This paper presents data on the settlement characteristics of deep fill with emphasis on the characteristics of engineered
fill (i.e. fill placed and compacted in relatively thin layers to an engineering specification).
Data derived from laboratory testing and field monitoring is provided for a variety of materials placed as deep
engineered fills for a number of projects in the Sydney Region.
Settlement estimates derived from parameters obtained from a desk study of international literature and some data from
the Sydney Region are presented and compared with the results of more recent laboratory testing on materials from the
Sydney Region.

2 DEFINITIONS
The following sources of information have been utilised in assessing potential settlement for projects involving
significant thicknesses of fill:

Published literature on compressibility of compacted fill;

Literature relating to settlement of dam embankments;

Laboratory testing and field settlement monitoring conducted by Coffey Geosciences Pty Ltd (Coffey) for
various projects in the Sydney Region;
Field plate bearing tests carried out by Coffey and others on existing engineered fill.
Published literature reports data on fill compressibility in various forms and expresses the compressibility parameter in
a variety of ways. Typical parameters reported are explained in Table 1 below, together with their relationship to each
other.
The compressibility of fill is dependent on the material type, level of compaction, moisture content and applied
pressure. Settlement due to self-weight of the fill will occur under 1D conditions. Following completion of filling, the
settlement due to applied stress due to building loadings will be under 3D conditions over the footprint of the building.
In addition to load-related settlement, long-term creep of fill and potential hydroconsolidation settlement due to
inundation of fill compacted dry of Optimum Moisture Content (OMC) is also a potential risk.

Australian Geomechanics Vol 40 No 4 December 2005 57


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

Table 1: Definition and Relationship of Compressibility Parameters.


Parameter Definition Relationship with Other Parameters
D Constrained Modulus for one- D = v / vs
dimensional (1D) settlement
calculation due to self-weight of fill. 1D Settlement = (v x Fill Thickness)/D.
For estimating Eu a constrained modulus, D, has
been calculated based on strain on initial loading.
For estimating E a constrained modulus, D, has
been calculated based on the strain rate at the
estimated end of primary consolidation.
E Drained Elastic Modulus for E = {(1+)(1-2)D} / (1-)
calculation of consolidation settlement
under three-dimensional (3D) loading where = Poissons Ratio ( = 0.3 may be
conditions and/or geometry. adopted for most soils)

Eu Undrained Elastic Modulus for Often measured using plate bearing test or
calculation of initial settlement under pressuremeter test in situ.
three-dimensional (3D) loading May be determined from initial strain of
conditions and/or geometry. laboratory triaxial test.
For idealised elastic material, Eu / E= 3/[2(1+)]
but this ratio is often higher for real soils.
Creep strain rate (per log cycle time) Creep settlement = x Fill Thickness x log
under no further increase in loading (t2/t1) where time t2 is later than time t1 (usually
for calculation of long-term log (t2/t1) may be taken as 2 over the design life
settlement. provided construction does not occur until at least
6 months (i.e. 0.5 year) after filling. For a design
life of 50 years (i.e. log(t2/t1) = log(50/0.5) = 2)
Creep strain over design life of project = 2 (assuming 2 log cycle time over design
life)
vs Vertical strain due to self-weight of vs = 1D Settlement / Fill Thickness
fill
H Hydroconsolidation strain due to H = 1D Hydroconsolidation Settlement / Fill
inundation Thickness
L Long-term post-construction vertical L = (vs + H) x Fill Thickness + B
settlement (after settlement from self- Where B = 3D settlement due to building loads
weight of fill completed) (usually small for residential structures)

Table 2 summarises short-term and long-term fill compressibility and hydroconsolidation settlement data from
published literature and some Coffey experience from previous projects. It is probable that some of the long-term field
monitoring data of fill settlement may include some components of creep and/or hydroconsolidation settlement due to
inundation by rising groundwater level following filling.

Table 2: Summary of Data Review on Fill Compressibility.

Source Material Type And Test Condition Deformation Parameter


Sydney Project Sandy Clay (laboratory testing data) D = 13.9 MPa at 100kPa applied pressure
small samples compacted to 100% D = 11.1 MPa at 200kPa applied pressure
Standard.
Sydney Project Sandy Clay (field monitoring data) D = 10.8 MPa at 36 kPa trial load
17m thick scraper compacted fill
with Density Ratio ranging from 80%
to 103% (median value of about
95%)

58 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

Source Material Type And Test Condition Deformation Parameter


Charles and Various granular materials D = 14.1 MPa for Sandstone Rock fill at
Skinner (2001a) compacted to 80% Density Index 200kPa
(approx. equivalent to 98 to 100% D = 48 MPa for Gravel fill at 200kPa
Standard Density Ratio) D = 52 to 125 MPa for Sand fill at 200kPa
Charles and Various materials values appropriate D = 50 MPa Sandy Gravel (ID = 0.8)
Watts (2002) to relatively small foundation loads. D = 25 MPa Sandy Gravel (ID = 0.5)
D = 200 MPa Sandy Gravel (preloaded)
D = 12 MPa Sandstone Rock fill (ID = 0.8)
D = 6 MPa Sandstone Rock fill (ID = 0.5)
D = 40 MPa Sandstone Rock fill
(preloaded)
D = 6 MPa Colliery Spoil (compacted)
D = 3 MPa Colliery Spoil (uncompacted)
Sydney Project Mixture of Clayey Sand, Sand, Sandy Eu = 30 to 75 MPa from 8 tests ignoring 2
Clay and Ripped Sandstone with unusually high results
brickbats, concrete and sandstone E assumed to be about 50% Eu (i.e. E = 15
fragments (field plate bearing tests) to 38 MPa assumed)
Sydney Project Mixture of clay, brickbats, clay pipe Eu = 24 MPa (91.3% SMDD)
fragments compacted to average Dry Eu = 30-58 MPa (94.5%SMDD)
Density Ratios ranging from 88.5% Eu = 20 MPa (88.5% SMDD)
to 94.5% (field plate bearing tests Eu = 22-31 MPa (93.6% SMDD)
with a 0.6m diameter plate)

Earth and Compacted Clayey Sand and Clayey During Construction


Rockfill Dams Gravel vs = 0.003 to 0.008 at 70 kPa
Sherard et al., vs = 0.019 to 0.033 at 700kPa
1963 Post Construction
vs < 0.002 in the first 3 years
vs < 0.004 in periods up to 14 years
Charles and Creep settlement monitoring of large = 0.01 for very soft puddle clay core dam
Skinner (2001a) embankment dams embankments.
= 0.0017 for the 73m high
Scammonden dam constructed with
sandstone and mudstone rock fill dam
(equivalent to about 0.0023 x average stress
in MPa).
Only one-third the above value for the 56m
high Megget dam constructed of sandy
gravel.
Penman et al Internal deformation measurement of = 0.0013 x vertical stress in MPa
(1971) a large dam
Charles and Collapse (Hydroconsolidation) Hydroconsolidation Strain, H values:
Watts (2002) settlement of non engineered fills  Mudstone/sandstone 2%
subjected to inundation  Clay/shale 5%
 Stiff Clay 3 to 6%
 Colliery Spoil 7%

3 GENERAL BACKGROUND ON FILL MATERIALS


Materials commonly available within the Sydney Region for bulk filling operations include:
Soils derived from alluvial and residual deposits. These include silty and sandy clays (CL and CH), silty and
clayey sands (SM and SC).

Australian Geomechanics Vol 40 No 4 December 2005 59


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

Rock fill derived from sedimentary rocks generally consisting of gravely, silty and clayey sands (SM and SC),
gravely, sandy and silty clays (CL) derived from sandstone and gravely silty and sandy clays (CL and CH)
derived from shale.
Recycled materials such as concrete, brickbats, clay pipe and tile, road materials including asphalt, roadbase
and sub-base materials derived from quarried igneous materials.
Engineered fill derived from the above materials typically has the following characteristics:
Maximum particle size is less than the layer thickness, which is generally less than 300 mm;
The sedimentary rocks commonly available are sandstones and shales, generally with unconfined compressive
strengths of less than 50 MPa. The resulting fills are generally gravely soils rather than rock fill;
Fills derived from shale excavations are likely to be more variable than those from sandstone excavations.
Clay contents, strength and compressibility varies greatly depending on the degree of weathering and
susceptibility to breakdown.
The bulk density of compacted fills generally ranges from 2 t/m3 to 2.4 t/m3 with a Standard Optimum
Moisture Content ranging from about 8% to 18%.
The more clayey fills can have relatively high potential for shrink/swell volume change with variations in
moisture content. Some silty clays may be dispersive.
Specified minimum levels of compaction range from 95% to 100% Standard Maximum Dry Density at
moisture contents within a range of 3% of Standard Optimum Moisture Content.

4 RECENT TESTING ON SYDNEY REGION MATERIALS


For this paper information from Coffey laboratory testing program on samples derived from sandstone, shale and
alluvial soils has been used to assess the following settlement parameters:
 Initial constrained Modulus 1-D loading modulus on initial loading;
 Constrained 1-D Modulus, D after primary consolidation;
 Undrained 3-D Elastic Modulus, Eu (derived from the constrained 1-D modulus, D);
 Drained 3-D Elastic Modulus, E(derived from the constrained 1-D modulus, D);
 Creep Strain Rate, , per log cycle time;
 Hydroconsolidation Strain Due to Inundation, H.
As it is uncommon for long duration testing to be carried out in the commercial environment, special test rigs were
constructed to test 75mm diameter, 150 mm thick samples at vertical pressures of up to 400 kPa. The samples were
compacted into PVC cells, which had holes drilled through the side walls to allow the samples to be saturated during
testing.
For testing samples at vertical pressures of up to 160 kPa a simple test rig was constructed using a plate fitted to a rod to
apply a vertical load by direct application of kentledge. The test rig is shown diagrammatically in Figure 1.

For testing samples at vertical pressures of up to 400kPa a purpose built rig was constructed to apply the vertical load
through a lever arm system. The test rig shown in Figure 2 is similar to a conventional odometer test cell. The rig was
purpose built so that it could be dedicated to long-term testing of compacted fill samples, rather than the alternative of
modifying existing odometer cells, which could not be dedicated to such a long-term project in a commercial
laboratory.
The samples were tested in a dry (at compaction moisture content) and saturated (inundated) condition.

60 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

Dial gauge to monitor


settlement

Weights to achieve
required surcharge load
Soil sample compacted
to required density

Plate to distribute load

75 mm diameter x
150 mm high mould with
holes to allow water
penetration

Container to hold
water for saturation of
soil sample

Figure 1: Schematic of Test Apparatus for Testing Samples up to 160 kPa.

Figure 2: Photograph of Test Apparatus used for Testing Samples up to 400 kPa.

5 MATERIAL PROPERTIES

5.1 SANDSTONE AND SHALE


Samples of Hawkesbury Sandstone, Weathered Ashfield Shale and Fresh Ashfield Shale were pre-treated by repeated
compaction and then tested for:
 Atterberg limit;

Australian Geomechanics Vol 40 No 4 December 2005 61


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

 Particle size distribution;


 Standard compaction.
Table 3: Summary of Laboratory Test Results on Sandstone and Shale Samples.

% Passing 2.36

Maximum Dry
Shrinkage (%)

Density (t/m3)
Liquid Limit

Content (%)
% Passing
mm Sieve
Index (%)

Optimum
0.075mm
Plasticity

Moisture
Standard

Standard
Linear

Sieve
(%)
Sample

Sandstone 22 NP NP 58 19 2.03 8.9


Weathered Shale 40 23 10 52 35 1.94 13
Fresh Shale 25 10 4.5 30 13 2.13 7.5
Note: NP Non Plastic
The test results presented in Table 3 indicate that sandstone is a non-plastic Silty Gravely Sand, the weathered shale is a
Gravely Sandy Clay and the fresh shale is a Silty Sandy Gravel.
The sandstone and shale samples were compacted to a target density of 98% Standard Density Ratio at moisture
contents 1% dry of Standard Optimum Moisture Content.

5.2 ALLUVIUM
Samples of alluvium were tested for
 Atterberg Limit;
 Particle size distribution;
 Standard Compaction.
Table 4: Results of Laboratory Testing on Alluvium Samples. 0.075mm Sieve
Plasticity Index

% Passing 2.36

Maximum Dry
Shrinkage (%)

Density (t/m3)
Liquid Limit

Content (%)
% Passing
mm Sieve

Optimum
Moisture
Standard

Standard
Linear
(%)

(%)

Sample

A. Sandy Silty
26 11 6.5 99 73.3 1.89 13
Clay
B. Silty Clayey
NP NP NP 96.5 47.5 1.88 13.8
Sand
C. Sandy Silty
22 7 3.5 100 53.7 1.89 13
Clay
D. Sandy Silty
22 7 3.5 100 53.7 1.89 13
Clay
Note: NP Non Plastic
The test results presented in Table 4 indicate that the alluvium tested was either a Sandy Silty Clay or a Clayey Silty
Sand and was tested after compaction to Standard Maximum Dry Density Ratios or 90%, 95% and 100% at moisture
contents 3% dry of Standard Optimum Moisture Content. The Sandy Silty Clays were tested at vertical pressures of 80
kPa (Sample A), 120 kPa (Sample C) and 160 kPa (Sample D). The Silty Clayey Sand was tested at 120 kPa (Sample
B).

6 TEST RESULTS

6.1 SANDSTONE
Samples of compacted sandstone fill were tested at vertical pressures of 200kPa, 300kPa and 400kPa over 156 days.
After 57 days the samples were saturated. At the end of testing, total strains of between 1.8% and 3.3% were recorded
for samples subjected to vertical pressures of between 200kPa and 400kPa.

62 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

Strain versus time and stain versus log time plots are shown on Figure 3 and Figure 4.
4 4

3.5 3.5

3 3

2.5 2.5

STRAIN %
STRAIN %

2
2

1.5
1.5

1
1
Samples saturated
0.5
0.5

0
0 1 10 100 1000
0 50 100 150 200
TIME (DAYS)
T IME (D A YS)
400kPa 300kPa 200kPa 400kPa 300kPa 200kPa

Figure 3: Strain versus Time Sandstone. Figure 4: Strain versus Log Time Sandstone.

Figure 5 shows the portion of the strain versus log time plot that is approximately linear, which has been used for the
estimation of the creep strain rate (). Figure 6 shows the portion of the strain versus log time plot at saturation (57
days).
4
3.5
y = 0.0832Ln(x) + 2.8916
3.5 3.3

3.1
3
y = 0.0621Ln(x) + 2.218
2.9
2.5
STRAIN %

2.7
STRAIN %

y = 0.059Ln(x) + 1.5146
2 2.5

1.5 2.3

2.1
1
1.9
0.5
1.7

0 1.5
10 100 1000 10000 40 45 50 55 60 65 70 75 80
TIME (DAYS) TIME (DAYS)
400kPa 300kPa 200kPa 400kPa 300kPa 200kPa

Figure 5: Creep Strain versus Log Time Figure 6: Hydroconsolidation upon Saturation at 57
Sandstone. Days Sandstone.

6.2 WEATHERED SHALE


The weathered shale was tested at initial vertical pressures of 200 kPa, 300 kPa and 400 kPa. After 164 days the
samples were saturated. After 130 days the sample loaded to 400 kPa was unloaded to 250 kPa and monitoring
continued.
Strain versus time and Strain versus log time plots are shown on Figure 7 and Figure 8.
Figure 9 shows the portion of the strain versus log time plot that is approximately linear, which has been used for the
estimation of the creep strain rate (). Figure 10 shows the portion of the strain versus log time plot at saturation (164
days).

Australian Geomechanics Vol 40 No 4 December 2005 63


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

7 Sample unloaded 7
6.5 to 250kPa 6.5
6 6
5.5 5.5
5 300kPa sample 5
4.5 possibly disturbed 4.5
STRAIN %

STRAIN %
4 4
3.5 3.5
3 3
2.5 2.5
2 2
Samples saturated
1.5 1.5
1 1
0.5 0.5
0 0
0 100 200 300 400 500 1 10 100 1000
TIME (DAYS) TIME (DAYS)
400kPa 300kPa 200kPa 400kPa 300kPa 200kPa

Figure 7: Strain versus Time Weathered Shale. Figure 8: Strain versus Log Time Weathered
Shale.

6.5
8
6.3
7.5
6.1
7
5.9
6.5
y = 0.209Ln(x) + 4.88 5.7
6
5.5 5.5
STRAIN %

5 5.3
STRAIN %

4.5 5.1
y = 0.246Ln(x) + 2.58 4.9
4
3.5 4.7
3 4.5
2.5 4.3
2 4.1
1.5 3.9
1 3.7
0.5 3.5
0 150 155 160 165 170 175 180
100 1000 10000 TIME (DAYS)
TIME (DAYS)
250kPa 300kPa 200kPa
400kPa 200kPa 300 KPa

Figure 9: Creep Strain versus Log Time Figure 10: Hydroconsolidation upon Saturation at
Weathered Shale. 164 Days Weathered Shale.

6.3 FRESH SHALE


The three samples were tested at a vertical pressure of 200 kPa. In the testing on the sandstone and weathered shale
samples there were concerns about the influence of moisture variation in the unsaturated samples due to the potential
for drying out of the samples via the holes drilled in the PVC test cells.
To assess the possible influence of the holes in the test cells the three samples were tested as follows:
 Sample saturated at time of initial loading and holes in cell left open;
 Sample unsaturated and holes in cell taped;
 Sample unsaturated and holes in cell left open.
Strain versus time and strain versus log time plots are shown in Figure 11 and Figure 12.
Figure 13 shows the portion of the strain versus log time plot that is approximately linear, which has been used for the
estimation of the creep strain rate ().

64 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

2.5 2.5

2 2

STRAIN (%)
STRAIN (%)

1.5 1.5

1 1

0.5 0.5

0 0
0 50 100 150 200 250 300 1 10 100 1000
TIME (DAYS) TIME (DAYS)

200kPa - Cell Holes Exposed 200kPa - Cell Holes Taped


200kPa - Saturated 200kPa - Cell Holes Taped
200kPa - Saturated
200kPa - Cell Holes Exposed

Figure 11: Strain versus Time Fresh Shale. Figure 12: Strain versus Log Time Fresh Shale.

2.5

y = 0.1174Ln(x) + 1.1399
2
STRAIN (%)

1.5

y = 0.068Ln(x) + 0.9096
1

0.5

0
10 100 1000 10000
TIME (DAYS)

200kPa - Cell Holes Exposed 200kPa - Cell Holes Taped


200kPa - Saturated Log. (200kPa - Cell Holes Taped)

Figure 13: Creep Strain versus Log Time Fresh


Shale.

6.4 ALLUVIAL SOIL


The samples were compacted to target densities of 90%, 95% and 100% Standard Maximum Dry Density 3% dry of
Standard Optimum Moisture Content.
Strain versus time and strain versus log time plots for the various vertical pressures are shown in Figure 14 to Figure 17.
These plots also show the hydroconsolidation strains that occurred on saturation.

Australian Geomechanics Vol 40 No 4 December 2005 65


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

7
1
6.5
0.9
6
0.8 5.5
5
0.7
4.5
0.6
STRAIN %

STRAIN %
4
0.5 3.5
3
0.4
2.5
0.3
2
0.2 1.5
1
0.1
0.5
0
0
1 10 100 1000 10000 100000 1000000
1 10 100 1,000 10,000 100,000 1,000,000
TIME(SEC)
TIME(SEC)
A - DDR 90% A - DDR 95% A - DDR 100% C - DDR 90% C - DDR 95%

Figure 14: Strain versus Log Time 80kPa Silty Figure 15: Strain versus Log Time 120kPa Silty
Clay Sample. Clay Sample.

1 7
6.5
0.9
6
0.8 5.5
5
0.7
4.5
STRAIN %

0.6
STRAIN %

4
0.5 3.5
3
0.4
2.5
0.3 2
1.5
0.2
1
0.1
0.5
0 0
1 10 100 1,000 10,000 100,000 1,000,000 1 10 100 1,000 10,000 100,000 1,000,000
TIME (SEC) TIME (SEC)
D - DDR 90% D - DDR 95% D - DDR 100%
C - DDR 90% C - DDR 95% C - DDR 100%

Figure 16: Strain versus Log Time 120kPa Silty Figure 17: Strain versus Log Time 160 kPa Silty
Sand Sample. Clay Sample

7 SUMMARY OF RESULTS

7.1 SHORT TERM SETTLEMENT


For the estimation of D and Eu the applied vertical stress has been plotted versus strain values on initial loading for the
sandstone and shale samples, as shown in Figure 18.

66 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

0.45
0.40

VERTICAL PRESSURE (MPa)


0.35
y = 18.4x + 0.03
0.30
0.25
y = 32.8x y = 26.4x - 0.12
0.20
0.15
y = 15.1x
0.10
0.05
0.00
0.0% 0.5% 1.0% 1.5% 2.0% 2.5%
STRAIN ON INITIAL LOADING (%)

Sandstone Weathered Shale Fresh Shale

Figure 18: Vertical Pressure versus Strain on Initial Loading,


Sandstone and Shale Samples.

For the estimation of constrained modulus, D and E the applied vertical stress versus strain at 20 days has been plotted
for the sandstone and shale samples, as shown in Figure 19.

0.45

0.40
VERTICAL PRESSURE (MPa)

0.35
y = 16.5x - 0.07
0.30

0.25
y = 18.1x y = 10.5x - 0.1
0.20

0.15

0.10
y = 13.3x
0.05

0.00
0.0% 1.0% 2.0% 3.0% 4.0% 5.0%
STRAIN AFTER 20 DAYS (%)

Sandstone Weathered Shale Fresh Shale

Figure 19: Vertical Pressure versus Strain after 20 days,


Sandstone and Shale Samples.

For the estimation of D and Eu the applied vertical stress has been plotted versus strain values on initial loading for the
alluvium samples, as shown in Figure 20.

Australian Geomechanics Vol 40 No 4 December 2005 67


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

0.2
y = 129x y = 11x
0.18

VERTICAL PRESSURE (MPa)


y = 18x
0.16
0.14
y = 13x
0.12
Non -Linear Respon se
0.1
0.08
0.06
Secan t Modulus
0.04
0.02
0
0.0% 0.5% 1.0% 1.5% 2.0%
STRAIN (%)

90% SMDD Sandy Clay 95% SMDD San dy Clay


100% SMDD Sandy Clay 90% & 95% SMDD Clayey Sand

Figure 20: Vertical Pressure versus Initial Strain, Alluvium.

Modulus values derived from Figure 18 to Figure 20 are summarised in Table 5.


Table 5: short term Settlement Parameters Derived from Recent Laboratory Testing.
Alluvium
Weathered Alluvium Alluvium
Parameter Sandstone Fresh Shale 100%
Shale 90% SMDD 95% SMDD
SMDD
D (MPa) 26.4 18.4 15.1 32.8 11 13 13 18 129
D (MPa) 16.5 10.5 13.3 18.1 - - -
Eu (MPa) 19.6 13.6 11.2 24.4 8 10 10 13 96
E (MPa) 12.3 7.8 9.9 13.4 - - -
The modulus values calculated are surprisingly low for sandstone and shale fill compacted to 98% Standard Density
Ratio. In Figure 18 it can be seen that the trend lines do not pass through the origin for the sandstone and weathered
shale samples indicating that there could have been seating errors in the tests. If the trend lines were forced to pass
through the origin the calculated modulus values would be even lower than the tabulated values. Comparing the
modulus values obtained from field plate bearing tests reported in the literature with values obtained from laboratory
testing indicates that higher modulus values are generally obtained from field testing.
The low modulus values may also be influenced by factors such as sample variability. Samples were monitored after
applying the full vertical load. Incremental loading of each sample may have enabled initial seating errors to be
assessed.
For the fresh shale the influence of the testing procedure can be seen from the variation in modulus. Where the sample
is sealed by taping the drainage holes, Figure 18 shows the initial modulus values was significantly lower than where
the holes are left open and the sample is unsaturated or where the sample is saturated before loading. Figure 19 shows
that after primary consolidation has occurred there is significantly less difference between the modulus values for the
fresh shale.
The testing indicates Eu/E ratios ranging from 1.1 to 1.8. This implies that a large percentage of the settlement due to
self-weight of the fill will occur immediately as the fill is placed. Subsequent consolidation settlement would depend
on the rate of fill placement, but given that deep fill sites usually take months to construct, the majority of the
consolidation settlement is also expected to take place as the fill is placed or soon afterwards.
The high modulus values for the alluvium are probably a result of heavy overconsolidation due to soil suctions
developed by compacting the samples 3% dry of Standard Optimum Moisture Content, and possibly a non-linear
response resulting in high modulus values at low vertical test pressures. Figure 20 shows a non-linear response trend

68 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

line for comparison with the tangent lines that were weighted towards the strains derived from vertical test load of 160
kPa to derive the modulus values in Table 5.

7.2 LONG-TERM CREEP SETTLEMENT


The creep strain per log cycle time is plotted versus vertical stress in Figure 21. This has then been used to calculate the
creep strain rate, .

0.9%

0.8%
CREEP STRAIN OVER ONE LOGCYCLE TIME (%)

y = 0.027x
0.7%

0.6%

0.5%

y = 0.015x
0.4%

0.3%
y = 0.014x
0.2%
y = 0.008x
y = 0.005x
0.1%

0.0%
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
VERTICAL PRESSURE (MPa)

Sandstone W eathered Shale W eathered Shale After Surchage Removal Fresh Shale

Figure 21: Summary of Creep Strain Rate () versus Vertical


Stress for Sandstone and Shale.

The measured creep strain rate, , appears logically to be stress level dependent and is assessed to be:
0.005 v when the vertical stress is measured in MPa for the sandstone;
0.027 v when the vertical stress is measured in MPa for the weathered shale;
0.008 v to 0.014 v when the vertical stress is measured in MPa, for the fresh shale.
These creep strain rates versus stress level relationships for the sandstone and shale fill are higher than the range of
0.0013v to 0.0023v reported by Skinner (2001a) and Penman et al (1971). While a creep strain rate of 0.005v for
sandstone is not expected to cause significant long-term creep problems, the values for shale are of more concern.
Figure 22 shows predicted creep settlements for various fill thicknesses over one log cycle time for sandstone,
weathered shale, fresh shale and weathered shale after removal of a surcharge. A bulk density of 22 kN/m3 has been
assumed.

Australian Geomechanics Vol 40 No 4 December 2005 69


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

140

CREEP SETTLEMENT OVER ONE LOG CYCLE TIME


120

100

(mm) 80

60

40

20

0
0 5 10 15 20 25
FILL THICKNESS (m)
Sandstone Weathered Shale Fresh Shale Weathered Shale (150kPa Surcharge Removed)

Figure 22: Predicted Creep Settlement Over One Log Cycle Time.

From Figure 22 it can be seen that while the creep settlement of a 20 m thickness of sandstone is only about 20 mm
over one log cycle time, for weathered shale creep settlement is over 120 mm. Creep settlements for fresh shale at 35
mm to 60 mm per log cycle time lie somewhere between those of sandstone and weathered shale.
The weathered shale sample loaded to 400 kPa was unloaded to 250 kPa after 130 days to assess the impact of removal
of a surcharge loading on long term creep settlement. Figure 21 and Figure 22 indicate that the removal of 38% of the
load on the sample resulted in a 44% reduction in creep strain over one log cycle time. Although Clients and
Contractors often view surcharges as impractical, this test result indicates the potential value of surcharging in reducing
long term settlement. The authors consider that rolling surcharges should be considered as a practical long term
settlement management tool in deep fills.

7.3 HYDROCONSOLIDATION
Table 6 summarised the hydroconsolidation rates for the various materials tested. As reported by other researchers (e.g.
Charles and Skinner (2001a), Jeffery and Thorne (1980), hydroconsolidation (sometimes referred to as collapse
settlement in uncontrolled fills) is highly dependent on compaction density and moisture content.
Table 6: Hydroconsolidation Strains Derived From Recent Laboratory Testing.
Material v (kPa) Strain (%) Strain (%) Strain (%) Strain (%)
90% SMDD 95% SMDD 98% SMDD 100% SMDD
Sandstone 200 - - 0.05 -
300 - - Nil -
400 - - 0.16 -
Weathered 200 - - 0.2 -
Shale 250 - - 0.16 -
400 - - 0.3 -
Sandy Silty
80 0.05 0.2 - Nil
Clay
Silty Clayey
120 0.47 0.017 - Nil
Sand
Sandy Silty
120 3.4 0.15 - Nil
Clay
Sandy Silty
160 3.1 0.85 - Nil
Clay

70 Australian Geomechanics Vol 40 No 4 December 2005


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

The relationships between hydroconsolidation and compaction conditions (density ratio and moisture content) for the
testing described in this paper are presented in Figure 23 and Figure 24.

4 4

3.5 3.5

HYDROCONSOLIDATION (%)

HYDROCONSOLIDATION (%)
3 3
Hydroconsolidation as a function of
compaction moisture content and vertical
stress (Huder 64) 2.5 2.5

2 2
400kPa
1.5 1.5

1 1

0.5 0.5
100kPa

0 0
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 88 90 92 94 96 98 100 102
VARIATION FROM SOMC (%) STANDARD DENSITY RATIO (%)
200kPa - Weathered Shale 250kPa - Weathered Shale 200kPa - Sandston e 300kPa - Sandstone
300kPa - Weathered Shale 120kPa - Silty Sand 400kPa - Sandston e 200kPa - Weathered Shale
120kPa - Silty Sand 120kPa - Sandy Clay 250kPa - Weathered Sh ale 300kPa - Weathered Shale
160kPa - Sandy Clay 200kPa - Sandstone 80kPa - Sandy Clay 120kPa - Silty Sand
300kPa - Sandstone 400kPa - Sandstone 120kPa - Sandy Clay 160kPa - Sandy Clay

Figure 23: Hydroconsolidation versus Standard Figure 24: Hydroconsolidation.


Density Ratio.

Figure 23 also shows the trend lines for hydroconsolidation to compaction moisture content for compacted fill subjected
to vertical pressures of 100 kPa to 400 kPa reported in Huder (1964).
For the alluvium where less than ideal fill conditions were simulated (low level of compaction at moisture content dry
of Standard Optimum Moisture Content) measured hydroconsolidation rates of up to 3.4% are consistent with reported
values for non-engineered fill in Charles and Watts (2002). Hydroconsolidation rates for the alluvium compacted at
95% SMDD are lower (up to 0.85%), but could still be of concern for deep fills subject to rapid inundation.
The measured hydroconsolidation values of up to 0.16% for the sandstone and up to 0.3% for the weathered shale are
relatively low and are unlikely to be of concern in deep engineered fill subject to the gradual re-establishment of a water
table. If over a period of years a rise in groundwater table occurred, the settlement due to inundation may not be
distinguishable from creep settlement.

8 CONCLUSIONS
The settlement properties of common fill materials in the Sydney Region have been assessed based on laboratory testing
data. The short term settlement parameters (D, D, Eu and E) calculated for sandstone and shale based on laboratory
testing are relatively low, compared to published values obtained from field plate bearing tests, but are consistent with
published values from laboratory tests.
When considering settlement due to the self weight of the fill, the test results indicate that primary consolidation
settlement for sandstone and shale occurs relatively rapidly, and should be complete within say 3 months of the
placement of fill. For deep engineered fills, the time taken to place the fill will usually result in short term settlements
being near complete by the time bulk filling has been complete.
The laboratory testing indicates that creep settlement of fill is highly dependent on the material type. Use of sandstone
and fresh shale should produce settlements that are acceptable even for relatively deep fills. More clayey materials,
such as the weathered shale tested, are likely to experience considerable creep settlement, which would appear to justify
limits being placed on the plasticity of fill materials to be incorporated into deep engineered fill specifications.
With an appreciation of the likely magnitude of creep settlement, the potential for differential settlement of deep
engineered fills needs to be considered. The potential for adverse impacts on structures, services, roads and kerbs
should be considered where fill thicknesses vary over short distances or where highly variable materials are used within
deep engineered fill.
In the design of residential structures on deep engineered fills it is unlikely to be feasible to reclassify lots from P to a
classification that would allow the standard designs of AS2870 (1996) Residential Slabs and Footings to be adopted.
Foundations should be designed in accordance with engineering principles considering potential differential settlements

Australian Geomechanics Vol 40 No 4 December 2005 71


SETTLEMENT OF DEEP ENGINEERED FILLS WADDELL & WONG

resulting from fill variability, time dependent creep movements and shrink swell movements. It is important to bear in
mind that it is usually differential settlement rather than total settlement that affects the performance of foundations and
buildings, as well as buried services and surface drainage.
Hydroconsolidation settlement of engineered fills is of concern where fill is placed at relatively low compaction levels
and where compaction is carried out at moisture contents dry of standard optimum moisture content. Clay samples
(weathered shale and sandy clay alluvium) exhibited higher levels of hydroconsolidation than more sandy samples
(sandstone and sand alluvium). To reduce the risk of hydroconsolidation settlement, deep engineered fills should be
compacted to a minimum of 98% Standard Maximum Dry Density at moisture contents near Standard Optimum
Moisture Content.
With adequate control on fill placement to ensure uniform distribution of material types, compaction level and moisture
conditioning, it should be possible to adopt shallow foundations on deep fills to meet the footing performance
expectations similar to natural sites.

9 REFERENCES
AS2870 (1996) Australian Standard: Residential slabs and footings - Construction, Standards Australia.
Charles, J.A. and Skinner, H.D. (2001a) Compressibility of Foundation Fill, Geotechnical Engineering 149 July 2001
Issue 3, pp. 145-157, Proceedings of the Institution of Civil Engineers.
Charles, J.A. and Skinner, H.D. (2001b) The delineation of building exclusion zones over high walls, February 2001,
pp. 28-33. Ground Engineering
Charles, J.A. and Watts, K.S. (2002) Treated ground: engineering properties and performance CIRIA C572, CIRIA
Huder, J. (1964) Die Zusammendrckbarkeit des Bodens und deren Bestimmung Schweizerische Bauzeitung, Heft
41. Reproduced in Hausmann MR (1990) Engineering principles of ground modification pp 71 - 72,
McGraw Hill.
Jeffery, R.P. and Thorne, C.P. (around 1980) The Specification, Production and Properties of Quality Fills paper
presented to Sydney Geomechanics Society.
Penman, A.D.M., Burland, J.B. and Charles, J.A. (1971) Observed and predicted deformations in a large embankment
dam during construction. pp 1-21, Proceedings of the Institution of Civil Engineers, 49.
Sherard, J.L., Woodward, R.J., Gizienski, S.F. and Clevenger, W,A, (1963) Earth and Earth-Rock Dams Wiley.

72 Australian Geomechanics Vol 40 No 4 December 2005

You might also like