Soil conditioning for pipejacking and tunnelling: properties of sand/foam

mixtures

G.T. Houlsby and S. Psomas
Oxford University, UK

Abstract

Soil conditioning techniques are used to alter the properties of soils to make them more suitable for
excavation by tunnel boring machines. Conditioning agents include foams, polymers and bentonite
slurry, either alone or in combination. Conditioning agents may have additional benefits in providing
lubrication for the advancing tunnel shield, or in the case of pipejacking for the entire tunnel shaft.

In spite of the now widespread use of conditioning agents in tunnelling practice, little basic
information is known about how they affect the fundamental properties of soils. The purpose of this
paper is to report a series of laboratory tests on mixtures of sands and conditioning agents. The
compressibility and permeability characteristics of sand/foam mixtures were studied using a Rowe
Cell. The shear strength characteristics of the mixtures were studied using shear box tests. The paper
concentrates principally on the use of foams as conditioning agents, but also includes some results
involving use of polymers and bentonite.

The principal conclusions, backed up by experimental data, are:
Foam-soil mixtures are stable at remarkably high voids ratios, even when subjected to
significant stress levels (approximately 200kPa).
Mixtures at such high voids ratios exhibit extremely low angles of friction.

Introduction

Soil conditioning agents are now fairly widespread in tunnelling practice, where they have a number
of applications. At the cutter head they may be used to improve excavation rates, reduce wear, reduce
power requirements and improve the control of water ingress. In the chamber and screw conveyor of
an earth pressure balance (EPB) machine they may improve the flow characteristics of the soils, and
allow the pressure to be more easily controlled. If a pumped spoil removal system is used they may
again be necessary to improve the flow characteristics. Depending on the conditioning agents used,
there may then be a necessity for a separation process before spoil disposal.

Conditioning agents may also interact with the process of lubrication of the advancing tunnel shield, or
in the case of pipejacking the lubrication of the entire pipe string. In the latter application, effective
lubrication can reduce the jacking forces, and hence reduce the need for expensive interjack stations.

In a typical modern tunnelling operation it is likely that a variety of different lubrication/conditioning
agents may be employed at different points in the construction process. Although much empirical
information has been gathered on the effectiveness of different agents in particular applications, most
of this information is highly site and application specific. It is often also related to specific proprietary
brands of materials, and may be held as confidential by commercial parties. There is therefore a need
for generic information on the ways that different soil conditioning agents interact with a variety of
soils.

In clays the desired purpose of the agents will often not be to mix intimately with the parent clay
material, but instead the object will be to obtain relatively large cuttings of clay suspended within a
slurry formed by the conditioning agent. In coarser grained materials (sands and gravels), however, it
is anticipated that none of the original structure of the soil would be preserved, and there would be an
intimate mixing of the soil with the conditioning agent. The purpose of this paper is to present data on
the properties of such mixtures.

In testing soils one would first examine their most basic characteristics with simple tests, and this
approach is taken here for the mixtures studied. Tests were used to determine the compressibility,
shear strength and permeability (although little information was obtained on the last). The tests used
were relatively small scale, using both standard and non-standard equipment. The materials most
commonly used as conditioning agents are bentonite, polymers and foams. When using these agents,
the interaction between foam and soil is the least well understood, so the investigations concentrated
on foam-soil mixtures, although the other materials also received some attention. The results of the
tests can be used to assess their suitability for the various applications described above.

Equipment and procedures

Foam generator

Figure 1 shows the design of the foam generator. The fluid consisting of water and the foaming agent
is pre-mixed to the correct proportion and placed in a 7 litre capacity cylinder. Air pressure is used to
drive the fluid into an inverter, in which the fluid and air phases mix. Compressed air at a higher
pressure than the fluid also enters the inverter, entraining the fluid as it exits. The fluid/air mix then
passes through a foam conditioner, which is a cylinder packed with small plastic rings. This serves
to divide the bubbles and form the structure of the foam. Finally the foam exits, with the device giving
a flow of about 2 to 4 litres/minute at an expansion ratio (total volume/volume of fluid) of about 15.

Pressure regulator
Relief valve
On/off valve
Air @ 7 bar
Foam
(2 - 4 litres/minute)
Pressure guage
Foam conditioner
Flow
meter
Air
Fluid
Level
indicator
Inverter
Foam
Fluid
Air

Figure 1: Foam generator

Mixing equipment

The sand/foam mixtures were mixed in an industrial quality food mixer of approximately 5 litre
capacity before transferring to the testing equipment. A careful audit was kept of the various
components in the mix (sand/liquid/air), and these values compared with those deduced from
measurements of density from sampling of the specimens tested. Cases where there was any
significant discrepancy (e.g. due to some loss of fluid in transfer of the mixture to the testing cell)
were rejected.

Rowe Cell

Compression tests were carried out in a standard 75mm diameter Rowe Cell (see Figure 2). Stiff
porous discs were placed on the top and bottom of the sample so that the boundary conditions are
fixed strain rather than free strain. It was not possible to measure separately the flow of liquid and air
from the sample, so the degree of saturation at intermediate stages of the test is unknown. It is
possible, however, to determine the degree of saturation at the beginning and end of the test. The tests
were carried out with load increments representing a multiplication of the previous load by
approximately 1.414 (i.e. doubling in two load steps).

soil/foam mixture
filters


Figure 2: Rowe cell (after Head, 1986
1
)

Shear Box
Shear testing was carried out in a standard Casagrande type shear box, 80mm square in plan, as
shown in Figure 3. Vertical load was applied by dead weights. Horizontal displacements were applied
at a constant rate and the corresponding shear load measured by a load cell. Both vertical and
horizontal displacements were measured.



Figure 3: Shear box (from Head, 1994
2
)

Materials Studied
Sands

Two different types of sand were used, and both were relatively uniform. The coarse sand was
Leighton Buzzard silica 14/25 sand (yellow). This has a coefficient of uniformity CU = 1.3, a sub-
angular grain shape, specific gravity G
s
= 2.65 and mean particle diameter d
50
= 0.85 mm. Minimum
and maximum voids ratio for the coarse sand were 0.49 and 0.79 respectively. The fine sand was
Leighton Buzzard silica DA 81DF. It is also a very uniform material, with CU = 1.4, G
s
= 2.65 and
d
50
= 0.165 mm. Minimum and maximum voids ratio for fine sand were measured as 0.61 and 0.91
respectively.
Foaming agents

Five different types of foam agent were tested. The foam generator was able to operate with all of
them, producing acceptable quality micro-foam. Early tests indicated only small differences in results
on foams of comparable quality, so a standard foaming agent was used for the tests reported here. The
material used was CETCO Versa VSX, which is a synthetic polymeric foam agent.

It is found that the addition of small quantities of further additives can improve the quality of the
foam, and this was found to be particularly beneficial in connexion with the coarse sand. After some
trials, the final foam solution consisted of 3% by volume VSX Versa foaming agent, and 0.7% by
volume of a mixture of VPC oil with Instapac425 and SC200 (this polymer mixture is referred to as
FOP). The role of FOP is to act as a booster enhancing the bubble production. This mix produced
foam with a stable bubble size. From a microscopic inspection, the size of the foam was in the range
of 0.1 mm to 1 mm, and the expansion ration was approximately 15.
Bentonite and polymer

The bentonite used was the CETCO Hydraul-EZ, a sodium montmorillonite bentonite able to swell to
about 10 times its original volume. The proportion of bentonite powder to water is critical for the
rheological characteristics of the slurry. After some trials, a bentonite:water ratio of 5:95 by weight
was used. In tunnelling applications, the bentonite slurry dosage should be greater than the calculated
void space, to create the impermeable 'filter cake'. The theoretical porosity ( )
max max max
1 e e n + = of
the sand in its loosest state was 0.476 for the fine sand and 0.441 for the coarse sand. The volume of
the bentonite slurry added was then defined by
sand slurry
V n V
max
= . Where is a measure of the ratio
of the slurry volume to the maximum voids volume for the sand. Values of for the tests ranged
from 0.5 to 1.9.

It is convenient to use the bentonite slurry voids ratio e
bs
, which can be defined as the ratio of the
water used (V
w
) over the volume of the bentonite V
bs
, so that e
bs
= V
w
/ V
bs.
For the compressibility tests
the bentonite slurry voids ratio varied between 17 and 31.

Lyon (1997
3
, p. 18) states that the ability of bentonite to swell is due to the presence of
montmorillonite, which provides greater surface area upon which water molecules may be absorbed.
The presence of calcium in the water reduces the effectiveness of bentonite because calcium ions have
a higher charge valence, and therefore hold the crystal lattices of sheets more tightly, allowing less
dispersion in water. The water used for the mixing was tap water with a measured pH between 6.5 and
6.8. However, better mixing was achieved when the water had a pH between 7 and 8. The presence of
calcium was detected by adding a tiny proportion of ammonium oxylate. The water was treated with
caustic soda (NaOH) or soda ash (Na2SO4) so that the pH reached 8.

In all cases with coarse sand and in some with fine sand, a small proportion of a polymer mixture
WOP (Water:VCP Oil:Polymer) was added. In the case of coarse sands, the addition of WOP
helped to produce a more homogenous foamed soil by creating a higher viscosity fluid matrix. The
initial proportions were (4:1:1) respectively, and the dosage from 0.01ml to 0.04ml of WOP mixture
per g of dry sand. The addition of this mixture was necessary in the cases where the water content of
the sand was more than 27-35%. WOP was added during mixing in the soil mixer bowl as a pre-
conditioner, prior to the addition of foam. The performance of the 'WOP' mixture improved when the
proportion of oil:polymer changed from 4:1 to 2:3. The final 'WOP' dosages used were 25ml for the
fine sand and 50ml for the coarse sand. In the case of coarse sand, this quantity was the minimum
required to achieve a homogenous mixture in the mixer bowl so that a representative slurry sample
could be tested. However, when bentonite slurry was added, the effectiveness of the 'WOP' mixture
decreased.

A measure of the viscosity of the slurry may be made with the Marsh funnel, which has a diameter of
150mm at the top, tapering over a distance of 300mm to a smooth bore tube 50mm long with an inside
diameter 4.8 mm. Over half of the top opening is a wire screen with apertures of 1.6mm. The time (in
seconds) required for a certain quantity of slurry (1500ml) to pass through the exit tube is measured.
The measurement in the case of fine sand with bentonite was about 45 seconds whereas in the case of
coarse sand with bentonite was about 59 seconds.

Testing Programme

Compression Tests

Each compression test was carried out incrementally, with a factor of 1.414 between successive stress
levels. During each load stage the drainage lines were at first closed, so that an undrained
compression is measured. Such a compression would be negligible for a saturated specimen, but for
the sand/foam mixtures there is a significant compression of the bubbles in the foam. Then the
drainage line was opened and fluid allowed to escape from the sample. Unfortunately, it was not
possible to make independent measurements of the amounts of air and liquid expelled. A typical plot
of displacement against time, showing both the undrained and drained phases is given in Figure 4.

-7.00
-6.75
-6.50
-6.25
-6.00
-5.75
1850 1900 1950 2000 2050 2100 2150
time [s]
c
o
m
p
r
e
s
s
i
o
n

[
m
m
]undrained stage
drained stage


Figure 4: Displacement v. time plot for load increment at 28.3 kPa for foam/fine sand test

Only the end-points for each load increment are plotted in the following figures, i.e. the voids ratio
achieved at the end of the drained stage. Figure 5 shows results from tests on mixtures of foam and
fine sand. The lowermost curve is for sand alone, initially prepared in a loose state. At low stress level
the voids ratio lies just below the maximum voids ratio for the sand. As the stress increases the voids
ratio reduces by a small amount, but the sand remains in a loose to medium density (demonstrating
that for a sand compaction to high densities by application of stress alone is not possible).

The upper curves are for two different tests on foam/sand mixtures. The differences between the initial
voids ratios at low stress levels are entirely due to the amounts of foam added. At low stress level it is
possible to make a sand/foam mixture at a voids ratio much higher than that which can be achieved for
sand alone. In fact in terms of the conventional Relative Density, the initial voids ratios correspond to
a Relative Density in the region of 200%. As the stress level is increased the two curves converge,
indicating that at higher stresses the behaviour is not so sensitive to the initial amount of foam added.
Presumably and extra foam is expelled from the mixture, leaving an amount that corresponds to a
stable structure at that particular stress level.

The tests were continued to a stress level of 226kPa, which is typical of the stress levels that may be
encountered in, for instance, pipejacking operations. At this stage the voids ratio of the sand/foam
mixture is still well above the voids ratio at 0% Relative Density, in fact the voids ratio corresponds to
a Relative Density of between 80% and 100%. This is a remarkable finding. It was unsurprising that
sand/foam mixtures could be made at high voids ratios, but it was quite unexpected that such high
voids ratios could be sustained at a remarkably high stress level. Note that the sand/foam mixture has a
truly composite action which is much more than the sum of the component parts: sand by itself would
have been compacted to a much lower density, and the foam by itself would have been crushed at such
a stress level, but the sand/foam mixture is stable in a remarkably loose state. This may have
fundamental implications for tunnelling operations.

Figure 6 shows results from using other soil conditioning agents. The lowermost curve is again for
sand alone. The next is for a sand/bentonite mix, with a relatively low bentonite dosage just sufficient
to achieve an initial voids ratio just above
max
e . The compressibility of the sand/bentonite mix is very
comparable to that of the sand alone. The next curve is for a sand/bentonite/polymer mix, again at a
relatively low dosage. The principal effect of the polymer is to increase the effective viscosity of the
pore fluid. The sample does show, however, a somewhat higher compressibility, but this may be
related to the slightly higher initial voids ratio, as towards the end of the test the voids ratio of very
comparable to that of the sand/bentonite mix.

0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
10 100 1000
log
v (kPa)
v
o
i
d
s

r
a
t
i
o
e
saturated fine sand
fine sand+foam 30%
fine sand+foam 50%
dry sand-emax,emin


Figure 5: Consolidation of fine sand and foam mixtures


0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
10 100 1000
log v (kPa)
v
o
i
d
s

r
a
t
i
o

e
fine sand +foam+bentonite+polymer
fine sand+bentonite+polymer
fine sand+bentonite+polymer
fine sand + bentonite
fine sand
dry sand-emax,emin


Figure 6: compression curves for sand/bentonite mixes

The next curve demonstrates how the compressibility is affected by the dosage of the bentonite. With
a much higher dosage (sufficient to more than fill the voids even at the loosest state) a very high voids
ratio mix is obtained. This is sustained until quite high stress levels, at which there is quite a sudden
change of stiffness. It is likely that this observed change of stiffness may be highly dependent on the
particular characteristics of the sand and the bentonite, so no very definite conclusions should be
drawn. However, it may be assumed that at high bentonite dosages relatively high voids ratios will be
sustained up to a certain stress level, but at sufficiently high stress levels the density would approach
that of the sand alone.

The uppermost curve is for a sand/foam/bentonite mixture. It starts at an exceptionally high initial
voids ratio. Although the compressibility is very high, at the end of the test the voids ratio is still well
above
max
e , and is comparable to the voids ratio achieved in the other foamed tests. A particular
feature of this test is the unloading curve, where it can be seen that the elastic rebound i s quite large.
This must be principally due to re-expansion of highly pressurised bubbles. It appears that the
bentonite is effective in trapping such bubbles, and that the combination of the use of bentonite and
foam can produce a mix which has not only ahigh compressibility, but also exhibits a high rebound.

Clearly the effect of time is important, as it may be expected that the foam would degrade with time
(as it does in air), and thus become less effective. Figure 7 shows the influence of time on a particular
sand/foam/polymer mix. The uppermost curve is for a relatively rapid test (1 hour per increment) and
the lower curve for a slower test (10 hours per increment). The curves are broadly similar, but it is
999999clear that with elapsed time the foam does become less effective. However, even for the slower
test the voids ratio remains remarkably high up to stresses of about 80kPa. With more detailed testing
it may be possible to establish a set of contours of equilibrium voids ratio with effective stress for
given elapsed times since foam formation.

0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
10 100 1000
log v (kPa)
v
o
i
d
s

r
a
t
i
o

e
fine sand+foam+polymer (1 hour)
fine sand+foam+polymer (10 hours)
fine sand
dry sand-emax,emin


Figure 7: effect of time on compressibility of sand/foam mixtures

Shear Box Tests

Conventional direct shear box tests were carried out on sand and sand/foam mixtures. Results of
typical tests at low stress level are given in Figure 8. The solid curves show the result for a
medium/loose sand. The stress/displacement curve shows a modest peak, which is associated with the
small amount of dilation exhibited by the sample. The stress then levels off at a value equivalent to a
constant volume angle of friction, consistent with the volume changes, which level off too.

The dotted curve shows an equivalent test on a sand/foam mixture. A much lower shear stress is
observed. As would be expected the sand/foam mixture compresses during shearing. However, even
though the compression approximately levels off towards the end of the test, the stress ratio still lies
well below the value obtained in the sand test. Thus a sand/foam mixture exhibits a much lower
constant volume angle of friction than the sand by itself. This is also a remarkable result, as it is often
assumed that the constant volume angle of friction for a sand represents a lower bound, which can be
relied on irrespective of the state of the sand.

Figure 9 shows an equivalent pair of tests at a much higher stress level (226kPa). The principal
difference lies in the test on the sand alone this time the stress level is high enough to suppress the
dilatancy, and no peak in the shear stress curve is observed. The response exhibited by the sand/foam
mixture is broadly similar to that at low stress level, with again a much lower angle of friction.

-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00
shear displacement (mm)
s
h
e
a
r
/
n
o
r
m
a
l

s
t
r
e
s
s

(
k
P
a
)
-1.6
-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.6
v
e
r
t
i
c
a
l

d
i
s
p
l
a
c
e
m
e
n
t

(
m
m
)
Figure 8: Shear box tests on fine sand at low stress level (28.3 kPa)

-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
shear displacement (mm)
s
h
e
a
r
/
n
o
r
m
a
l

s
t
r
e
s
s

(
k
P
a
)
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
v
e
r
t
i
c
a
l

d
i
s
p
l
a
c
e
m
e
n
t

(
m
m
)

Figure 9: Shear box tests on fine sand at high stress level

How can the two observations (a) that sand/foam mixtures can exist at remarkably high voids ratios
and (b) that such mixtures exhibit remarkably low angles of friction, be drawn together? The obvious
framework to use is one that addresses the influence of voids ratio on angle of friction, and this topic
was treated in some detail by Bolton (1986
4
, 1987
5
). However, he was only addressing conventional
sand, which could not exist at the voids ratios encountered for the sand/foam mixtures.

Figure 10 shows a plot of angle of friction, from the direct shear box tests, plotted against voids ratio
at the start of the test. The solid dots show results for fine sand. Note that samples were deliberately
prepared in as loose as possible condition, and several in fact plot as looser than the normally accepted
minimum density. The solid line shows Boltons correlation for variation of angle of friction with
relative density, assuming an angle of friction at constant volume of 31
o
. Boltons correlation assumes
that the angle of friction falls with increasing voids ratio, but that this reduction does not continue
below
cv
, which is assumed to be attained at about a Relative Density of 20%. The evidence from the
sand tests is either that the angle of friction at constant volume is in the region of 27
o
(unlikely for a
silica sand) or that for very loose specimens (which may still be compressing at the end of the test) a
lower angle of friction than
cv
is obtained in the direct shear tests with rather limited displacements.

emin=0.61 emax=0.90
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00
voids ratio e
a
n
g
l
e

o
f

f
r
i
c
t
i
o
n

(
d
e
g
r
e
e
s
)
fine sand
foamed fine sand
Bolton's correlation

Figure 10: variation of angle of friction with voids ratio: fine sand

The open circles show the data for the sand/foam mixtures. Although there is a considerable scatter,
they in fact confirm that the broad trend predicted by Boltons correlation is continued at the voids
ratios looser than those that can be obtained for sand alone. The angle of friction continues to drop as
the voids ratio increases, falling to the astonishingly low value of 7
o
.

A mechanistic explanation is that the sand/foam mixture may be thought of a s consisting of two types
of particles, firstly sand grains, and secondly bubbles, which may be thought of as frictionless and
deformable particles. The more bubbles, the lower the angle of friction. One would therefore expect
that there would be a correlation with the degree of saturation too. This is indeed observed to be the
case (the lower angles of friction within the scatter band correspond to lower degrees of saturation),
but the quantitative evidence is limited.

Figure 11 shows the equivalent data for coarse sand, confirming that the above result is not simply an
unusual feature of the fine sand tested. Exactly the same trends as described above are observed. Note,
however, that both the sands tested are relatively single-sized, and it is necessary to extend the result
to poorly sorted materials too.

emax=0.79
emin=0.49
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00
voids ratio e
a
n
g
l
e

o
f

f
r
i
c
t
i
o
n

(
d
e
g
r
e
e
s
)
foamed coarse sand
coarse sand
Bolton's correlation
Figure 11: variation of angle of friction with voids ratio: coarse sand

Permeability Tests

Permeability tests were carried out in the Rowe cell under constant head conditions. The results were
only partially satisfactory, as the head losses in the filters either side of the sample were significant,
and the necessary corrections to the measurements were therefore large. However, a clear indication
that the presence of foam decreased the permeability of the soil to water flow was obtained. Typically
the permeability was reduced about tenfold.

Discussion

These tests have demonstrated two main findings for sand/foam mixtures:

(a) Mixtures of sand and foam are stable at remarkably high voids ratios for significant periods of time
at stress levels up to 200kPa (and possibly higher, as this was the highest stress level tested). The voids
ratios on initial mixing of sand and foam at low stress level were many times higher than that in the
loosest state for a normal sand, and depend largely on the chosen foam injection rate. The voids ratios
on compression to 200kPa are, however, also well above that in the loosest state normally
encountered. The foam/soil mixture has (unsurprisingly) a high compressibility at stresses up to
200kPa, but shows a much stiffer response on unloading. It had of course been anticipated that the
mixtures would have a very high initial voids ratio and compressibility: the remarkable finding from
the tests was that high voids ratios could still be sustained at relatively large stresses.

(b) At high voids ratios (only sustainable through the addition of foam) sands exhibit remarkably low
angles of friction, with the angle of friction reducing with increasing voids ratio in a way that is
consistent with previously obtained data for sands at different densities. The remarkable result here is
the extension of Boltons correlation to voids ratios considerably higher than that at the loosest
normal state. Boltons correlation suggests that loose sand exhibit an angle of friction
cv
, and that
denser materials show higher angles of friction due to dilatancy. In the experiments described here
foamed sands at states much looser than the normal loosest state were found to exhibit angles of
friction much lower than
cv
, which had perhaps been previously envisaged as a lower bound to the
friction angle. Angles of friction as low as 6
o
were measured.

These findings are of fundamental significance for the tunnelling industry. Firstly the finding that
foam/soil mixtures can be stable at relatively high stresses is very valuable. It means that foams
injected, say, at the cutting face will not simply be destroyed under high stress, but that a stable foam-
soil mixture could be sustained in the chamber of an EPBM. Furthermore the relatively high
compressibility of this mixture will ease control of pressure within the chamber.

The low angle of friction of the foam soil/mixture has implications for reduced power consumption
and reduced wear throughout the tunnelling process. Paradoxically it may also be a problem. Effective
control of pressure in the chamber requires controlled removal of spoil through the screw conveyor,
and this can only be achieved if there is sufficient resistance within the conveyor that the spoil does
not simply flow through it. Control of foam quantities to ensure that sufficient friction is maintained
will be necessary. However, once through the conveyor a low angle of friction would be very
beneficial for pumped removal of spoil, which would otherwise be very difficult for coarse grained
materials.

The control of properties at different points in the tunnelling process suggests that addition of foams at
multiple points in the excavation and spoil handling process may be desirable. Although this paper has
not been directed towards the important issue of disposal of spoil, a further attraction of foams is the
relatively small amounts of potential pollutants that are added to the soil for a large volume of foam.

Conclusions

Simple laboratory tests on sand/foam mixtures have demonstrated that they can exist at remarkably
high voids ratios at stresses up to about 200kPa (and possibly higher). At these high voids ratios the
sands also exhibit extremely low angles of friction. Both the high compressibility and low frictional
strength of these mixtures have important implications for the use of foams in tunnelling operations.
Foams also serve to reduce the permeability of a sand. Use of other soil conditioning agents (bentonite
and polymers) has also been examined. Although when used alone they have a less dramatic influence
on compressibility and strength characteristics, they can play an important role in stabilising and
modifying a sand/foam mixture.

Acknowledgements

The work described here was supported by the Pipe Jacking Research Group, a consortium supported
by the Pipe Jacking Association, a number of water supply companies and industrial partners.

References

1. Head K.H., 1986. Manual of soil laboratory testing. Effective stress tests. Vol. 3, Pentech press,
London.
2. Head K.H., 1994 Manual of soil laboratory testing. Compressibility, shear strength and
permeability. 2
nd
edition, Vol. 2, Pentech press, London.
3. Lyon J., 1997. Drilling manual. CETCO Europe.
4. Bolton M.D., 1986. The strength and dilatancy of sands. Geotechnique Vol 36, No 1, pp. 65-78.
5. Bolton M.D., 1987. The strength and dilatancy of sands. Discussion Geotechnique Vol 37, No 2,
pp. .219-236.