Вы находитесь на странице: 1из 16

AN INVESTIGATION OF THE BEARING CAPACITY OF SOME

BORED AND DRIVEN PILES IN LONDON CLAY

G. G. MEYERHOF, M.Sc., Ph.D., and L. J. MURDOCK, MSc., Ph.D.

SYNOPSIS
The article describes a soil investigation and L’article d&it un examen du sol et des essais de
loading tests which were made on bored and driven charge faits sur des pieux de beton, for&s et enfonccs,
dans l’argile de Londres, afin d’en examiner la
concrete piles in London clay to study their bearing
charge limite et le tassement. On a trouvc que la
capacity and settlement. It was found that the resistance au cisaillement de l’argile fissurce apres
shearing strength of the fissured clay decreased le prelbvement d’echantillons diminue rapidement
rapidly with time after sampling, due to opening en fonction du temps, par suite de l’ouverture
of the fissures which could not be closed sufficiently de fissures qu’on n’a pu suffisamment refermer
moyennant une grande pression triaxiale. L’eau
by large triaxial pressures.
provenant du beton des pieux for.% dans un
Water from concrete of bored piles installed in sondage non-cuveli! ramollit l’argile, et l’on n’a pu
an unlined auger borehole softens the clay and Cviter ce ramollissement qu’en employant un
this softening could not be avoided except by an melange trop set pour en permettre l’utilisation.
Alors que la resistance a un point donne peut &tre
unworkable dry mix. Whilst the point resistance
bade sur la r&sistance specifique au cisaillement, le
can be based on the natural shearing strength, the frottement superficiel de ces pieux for&s est donnc
skin friction of these bored piles is closely given by avec une bonne approximation par la resistance
the fully softened strength. On the other hand, dans l’etat ramolli. Par centre, le frottement
the skin friction of concrete piles driven into stiff superflciel de pieux de b&ton enfoncbs dans une
argile terrace ou dure semble avoir une limite sup&i-
to hard clay seems to have an upper limit of 2,000
eure de 2,000 livres par pied car& ,Avec un facteur
lb. per square foot. At a factor of safety of 3
de &cur& de 3 centre une rupture par cisaillement,
against shearing failure, the settlement of widely l’affaissement de pieux largement &art&s, fores ou
spaced bored and driven piles in London clay is enfonces dans l’argile de Londres, ne dcpassera
unlikely to be greater than 1 inch. probablement pas un pouce.

INTRODUCTION
Most of the pile foundations constructed in Great Britain during the past few decades
consist of concrete piles, which are either precast and driven, or cast-m-place in a permanent
or temporary casing. In sufficiently firm clay soils, a casing may be omitted by installing
the pile in an unlined auger borehole, which has been done extensively for short piles for house
foundations in shrinkable clays. In order to ascertain how far that method could be used
in connexion with long piles for civil engineering structures, and how such piles compare
with similar driven piles, a soil investigation and series of loading tests were made on two sites
in the London area, at Southall (Middlesex) and Barnet (Hertfordshire). Economical methods
of installing piles in stiff clays were studied, and the factors contributing to the point resistance
and skin friction in London clay were investigated.

DESCRIPTION OF SITES AND SOIL PROPERTIES


Both sites are underlain by heavily preconsolidated, insensitive fissured London clay, and
during the summer when the investigation was made, no ground-water was struck. At
Southall, about 25 feet of stiff brown clay overlies hard blue clay of great depth, whilst at
Barnet about 2 feet of soft brown clay over a variable layer of clayey gravel up to 4 feet
thick were followed by stiff to hard brown clay throughout. In spite of the relatively uniform
index properties within the depth investigated, the shearing strength of the material increased
rapidly with depth, and in a similar way at both sites (see Figs 1). The strength was deter-
mined by unconfined and undrained triaxial compression tests, which gave substantially
267
268 MEYERHOF AND MURDOCK

u
C 4 I \
I \
f xl .
aI \ \
I I I \
i \
_r: 4 I
\
I I I \
.
I ,
.I0 I I
\

(a) Index and strength properties at Southall

(3) Index and etxength propertiee at Barnet


BEARING CAPACITY OF PILES IN L6NDON CLAY 269

the same result when testing was carried out immediately after sampling. However, when
tests were made on samples stored in tubes for several days, a significant decrease in the com-
pressive strength was observed even at lateral pressures considerably exceeding the over-
burden pressure. This decrease in strength, which amounted up to about one-half of the
original strength, was particularly noticeable for the stiffer and harder material (Figs l),
and may be explained by the gradual opening of fissures of the clay after sampling.* Since
all the later tests were made on specimens from 4-inch-diameter sample tubes, the amount of
opening of the fissures may have been influenced by the release provided in the tubes. For
tests made immediately after sampling, the residual shearing strength was up to 50 per cent
less than the maximum strength recorded in the tests.
Full remoulding of the material at the same water content resulted, generally, in a slight
increase of the shearing strength when compared with the shearing strength measured im-
mediately after sampling, by destroying the fissured structure of the soil. Full softening
of the clay caused a large drop in shearing strength to about 500 lb. per square foot throughout
the depth at both sites. The latter strerigth was generally obtained from undrained triaxial
compression tests on l&inch-diameter by 3-inch-high cylinders, which had been covered
by filter paper and softened in free water for 3 days (longer periods had shown no significant
further decrease of strength). Similar results were obtained by extrapolating to zero lateral
pressure, the shearing strength of consolidated-undrained triaxial tests, or extrapolating
to zero normal pressure of corresponding shear-box tests. At the fully softened strength
the average water content of the samples was from 5 to 8 per cent (of dry weight of soil)
greater than the original amount.
Since for stiff and hard clays the adhesion between concrete and soil may be less than the
shearing strength of the clay, some exploratory adhesion tests were ca.rried out in a shear
box. The lower frame was filled with a coarse stone (with a texture like smooth dense
concrete), and the upper frame contained undisturbed clay of shearing strengths between
2,000 and 6,000 lb. per square foot. The adhesion (skin friction) was found to increase with
normal pressure at a decreasing rate. For a pressure of less than about 1 ton per square
foot, the relationships between adhesion and pressure were similar and corresponded to a
coefficient of friction of about 0.8 for a dry stone and O-4 for a wet stone ; forlarge pressures,
the adhesion approached the shearing strength of the clay. The residual adhesion ranged
from 0.5 to O-8 times the maximum value.

METHOD OF INSTALLING BORED PILES


A schedule of the piles used in the investigation is given in Table 1. This includes eleven
bored piles and two driven piles. The loading frame used for testing the piles was fixed
to a number of anchor piles constructed by similar methods to those used iri forming the test
piles themselves, except that reinforcement was included to take the tensile stresses. The
lay-out of the piles at Southall and Barnet is shown in Figs 2. It will be noted that the
minimum distance between adjacent surfaces of the anchor and test piles was 3 feet. At
this distance it was thought that there would be little interference between the test piles and
the anchor piles during loading.
Short bored piles had previously been formed in holes produced by a .post-hole auger.
This method was, however, impracticable for the test piles in the present programme, in view
of the depth to which they were to be bored and the stiffness of the clay. TWOmethods were
adopted for comparison :-
(1) Using a derrick-type rig, as normally used on test borings for site investigations.
(2) Using a lorry-mounted power-driven auger.
* In view of the importance of these results, a programme of work in checking is being carried out on
another site in London ; preliminary examination of the results obtained has confirmed the reduction in
strength of the clay.
270 MEYERHOF AND MURDOCK

Figs 2

Scale: t mch = I foot

12730indiuterdiameter/lenkh
SOVTHALL SITE

Lay-out of experimental piles

The derrick-type rig was used only for the three test piles ,Uos S.2, 5.5, and S.6, and
a number of anchor piles sunk at Southall ; the power-driven auger was used for all the re-
maining piles, including those at Barnet.
The derrick was of the four-legged type, 20 feet high in the erected position, and having
a friction winch attached to two of the legs, the winch being driven by a 9-horse-power
petrol engine. The holes were bored by a clay auger turned manually, the winch being used
solely for lifting the boring rods and auger from the borehole. Boring a 20-foot pile in the
stiff brown London clay and claystones took 1 to 2 days ; a 30-foot pile took 2 to 3 days,
and a 40-foot pile about 5 days. The greater part of these times was occupied, however,
in the frequent soil sampling necessary for research purposes.
It was necessary to use casing tubes in some of the holes as indicated in Table 1, in order
to exclude possible local seepages of ground-water, and to prevent a certain amount of caving-
in of the soil. This casing was provided with a cuttin g shoe having an overall diameter about
2 inches greater than the nominal pile diameter gil-en in Table 1. The tubes were also
connected with external screwed collars, having an over-all diameter of about 13k inches.
The external diameter of the casing was Q inch greater than the nominal diameter given
in Table 1.
Other methods of boring which might have been adopted include the use of a percussion
shell filled by continuously dropping it in the hole. \Vith this method of boring, however,
it is usually necessary to use at least a small amount of water to lubricate the passage of the soil
passed through the clack valve into the shell itself, and this would have been undesirable
for the purposes of the tests. The characteristics of the hole would also have been different
from that bored by the power auger.
Bored piles are more often sunk by three-legged derricks provided with an air-operated
hoist than by the four-legged type of derrick already described. For continuous working,
the former is faster and more economical, but in view of the cost of operating a compressor
under intermittent working conditions necessitated by the test procedure, it was not deemed
to be worth while in this case. The derrick was also used for operating the tamper which
consolidated the concrete.
BEARING CAPACITY OF PILES IN LONDON CLAY 271

Table I.-Details of test piles

I I I
- -
Depth
Water/
Pile of
Method of cement Method of Use of
No. Size pene-
installation ratio of tamping concrete ,-asing tubes
tration : c:oncrete
ft.
-~___ _-
12 20 Bored : power auger Falling weight : medium None
z:; $3 derrick rig 8:: None
9, 99 ,9
s.3 z. z ,I power auger 9, ., 9, First 10 feet
Southall s.4 40 2, ;:; ># 7, I, None
s.5 14 20 ,, de&ck ;ig 0.4 ,. II >, None
in. 30 ,* 9, .> ?? ,. I, None
s”:; dia. 40 II power auger 0”:; 8, 3, ,> First 10 feet
i
_-

Bored : power auger 0.2 Falling weight : heavy None


,I ,, 3, Hand : light None
,* >> >I 8:; Falling weight : heavy None
Bamet
II 0.4 9, >I medium None
Driven : dyop-h:m- - - -
mer
- - -
- -

Test piles Nos S.1, S.3, S.4, and S.7 at Southall, and all the in-situ piles at Barnet
were bored by a Suco ,power-driven auger (see Fig. 3). This comprises a rotary drill rig
mounted on a heavy-lorry chassis, a vertical mast at the rear of the lorry carrying a cross-
head, which is raised or lowered by power from the lorry engine. The rotation of a square
kelly located at the top in the cross-head is affected by a crown and pinion with a sliding
bush at the bottom of the mast. The height of the mast and the length of the kelly, which
also functions as the drilling rod, permits a nominal depth of 20 feet to be drilled below the
ground surface, without making any stops for additional lengths of drilling rod.
The 20-foot-deep test and anchor piles were drilled in times ranging from 45 minutes to
2 hours, depending on the time spent on soil sampling. The 30-foot-piles were sunk in 3 to
4 hours, the majority of this time again being spent on soil sampling and in connecting and
disconnecting the extension rods for the lower 10 feet of the pile.
This machine is obviously much faster in operation than derrick-type rigs, but suffers
from two disadvantages :-
(u) There is no provision on the machine to deal with casing tubes, should this become
necessary due to the seepage of ground-water or caving-in of soil.
(b) It cannot be made to operate the tamper used in consolidating the concrete in the
hole ; this necessitates moving the power auger out of the way and using the
derrick-type rig for this purpose. (The disadvantage would not be of importance
if a wet concrete mix were used, instead of the dry mix adopted in this test series.)
For operational use on a large scale it would be necessary to develop a modified form of
power auger to overcome these two disadvantages.
Previously (Davey, 1949), bored piles have been constructed using relatively wet I : 2 : 4-
mix concrete dropped into the borehole with only light tamping. Since it was desired, in
the present series of tests, to establish whether any wetting-up of the clay occurred as a result
of absorption of water from the concrete, and if so, whether this wetting-up could be prevented,
much drier and richer mixes were used. The mixes consisted of 1 part normal Portland
cement to 16 parts river sand to 3 parts of a/&-inch river gravel ; the river gravel was a.
I*
BEARING CAPACITY OF PILES IN LONDON CLAY 273

Table 2
Crushing strength of test cubes made from concrete used in bored piles

Compressive strength of concrete

0.4 - 1 4,17@---5,290 / 4,730 / 5.910-6,710 / 6,170


-

DRIVING OF PRECAST CONCRETE PILES


Two precast concrete piles, 30 feet by 12 inches by 12 inches, and 20 feet by 12 inches
by 12 inches, were driven at Barnet. The cost of transport of a complete piling frame was
not warranted for the driving of two piles, and the driving was therefore carried out with
30-foot guyed leaders and a friction winch. The general arrangement is shown in Fig. 4,
the 30-foot leaders for the piling hammer being guyed from the top by wire ropes attached
to earth anchors with suitable arrangements for tightening the guys. The friction winch
was of the same type as that adopted for the in-situ piles, and its use eliminated the necessity
for a trip gear on the leaders. This arrangement worked quite satisfactorily.

NUMBER OF BLOWS PER INCH TEMPORARY COMPRESSION : INC”


.60 4

l-i-7-r

Figs 5. Driving records of temt piles


274 MEYERHOF AND MURDOCK

A 30-cwt drop-hammer was used for both piles. A steel helmet with timber dolly, and
packing comprising several layers of sacking, was used on the piles. The weights of the
piles were roughly 2 tons and 1) tons for the 30-foot and 20-foot piles, respectively, so that
the weight of the hammer was well above the normally accepted lower limit of half of that
of the pile. The efficiency of the hammer, calculated in accordance with Section 602 of
the Draft Code of Practice for Foundations, Part 4-Pile Foundations, is 0.46 for pile T.A,
and 053 for pile T.B. This is considerably higher than the 30 per cent efficiency recom-
mended in Section 7.01 of the Code. Driving records for the two piles are given in Figs 5.
Pile No. T.A had been driven satisfactorily to within 3 feet of ground level when the top
of the pile began to break up. Driving was stopped and the head of the pile repaired using
1 : I?J: 3 concrete. Driving was recommenced 12 days later and satisfactorily completed,
but slight further damage to the pile had to be made good. The effective length of the pile
below ground level was 23 feet. The pile was kept wet when being driven.
Pile No. T.B was driven to within 5 feet of ground level, at which point the top of the pile
became considerably damaged. Driving was stopped and the top of the pile was repaired
with 1 : 14 : 3 concrete. The pile was kept wet when being driven. An attempt was made
to recommence driving after 11 days, but after driving a further 3 inches the top of the pile
was again damaged, and it was decided to cut the pile off at ground level and test as a pile
with an effective length of 14 feet 6 inches.
The breakage of the piles is attributable mainly to the hard driving encountered, since,
as has been already mentioned, the shear strength of the clay at Barnet increased with depth.
After carrying out the loading tests, specimens were cut from both piles and tested in
compression across the pile sides to determine the crushing strength. These tests indicated
that the compressive strength of the concrete was 5,406 lb. per square inch at 14 months,
which was equivalent to a column strength of 3,000 to 4,000 lb. per square inch at 3 months,
the probable age at the date of driving.
The driving stresses in the piles were assessed in accordance with the methods given in
Building Research Technical Paper No. 20, the basis being as follows :-
Weight of hammer .. .. .. .. = 3,360 lb.
Weight of helmet and dolly . . .. .. = 168 lb.
Modulus of elasticity of concrete .. .. = 4.5 x 106 lb. per square inch
Drop of hammer .. .. .. .. = 3 feet 4 inches
Embedded length of pile .. .. .. = 14 feet
Stiffness for a l-inch thickness of sackcloth, te = 23,000 lb. per sq. in. after 25 blows
and to = 40,000 lb. per sq. in. after 4,000 blows
Stiffness of the dolly, te .. .. .. = 100,000 lb. per sq. in.
Under these conditions the head stress was estimated to be between 3,000 and 4,000
lb. per square inch depending on the value taken for to.
A comparison of the driving stress so calculated, with the column strength of the concrete
given above, indicates that under the driving conditions appertaining, failure was to be
expected.

LOADING TESTS

The general arrangement for applying loads to the test piles is illustrated in Figs 6.
Steel beams, sufficient to resist the bending when loading was in progress, were strapped to
tension rods embedded in anchor piles. The arrangement was slightly varied in the case of
pile No. T..B, in that a cantilevered system of steel beams was used to jack against. The
tensile steel is shown in Figs 6 as 2&-inch by 2f-inch R.S. rails. In other cases, high-tensile
steel wire and screwed rods were used instead of the old steel rail, depending upon the avail-
ability of these materials.
BEARING CAPACITY OF PILES IN LONDON CLAY 275

Sole: t Inch I I fax

Figs 6. Arrangement of load beams for pile tests

The loading system comprised a hydraulic jack, proving ring, and the necessary packing
plates and cross-girders. The load was measured by the proving ring, the reading on the
pressure gauge in the hydraulic jacking system being used as a check. The load was applied
in small increments, and settlement readings were taken by a dial fixed to a long beam
supported 5 to 6 feet from the test pile. The rate of loading the piles was sufficiently fast
to avoid any appreciable consolidation, without, on the other hand, causing an increase in
the shear strength due to a rapid rate of loading (that is, tests generally complete within the
working day).

TEST RESULTS
Typical load/settlement curves are given in Figs 7 and 8. The results of the pile loading
tests are summarized in Table 3.
The piles were tested first about 1 month after installation and, in most cases, again after
about 6 months and 18 months, to investigate their subsequent behaviour. The load/
settlement curves show that there was no increase of the bearing capacity with time ; the
driven piles indicated, in fact, a considerable reduction of the bearing capacity, because on
reloading, only the previous residual bearing capacity was reached. The settlement on
reloading the piles was generally much less than on first loading, due to the re-compression
of the soils ; for the same reason, the driven piles settled less than the corresponding bored
piles (apart from softening effect).
One of the bored piles at Southall (S.2) failed at a surprisingly low load, and a deep test
pit was made alongside to investigate the cause. It was ascertained that at a depth of 5
feet below ground level, a short section of the pile had crushed.
276 MEYERHOF AND YURDOCK

Fig. 7. Typical results of loading Fig. 8. Results of loading tests


tests on bored piles on driven piles

Analysis of the concrete indicated only one-half the normal cement content (that is,
a 1 : 8 mix) probably due to segregation of the relatively dry material and insufficient tamping.
Horizontal soil samples were taken towards the pile at various depths to determine the
water content of the soil. There was no significant difference in the water content near the
strong and weak portions of the pile (see Figs 9). However, the water content increased
rapidly towards the pile within about 2 inches from the shaft, where it was up to 7 per cent
(of the dry weight of soil) greater than the original amount. This increase in water content
confirms the results of previous research at the Building Research Station (B.R.S.), made
at the suggestion of Dr L. F. Cooling, which indicated that it was probably due to water
from the concrete, the water/cement ratio of 0.4 in this pile being greater than the amount
theoretically required for hydration of the cement.
In an attempt to avoid the water migration and consequent softening of the soil (see
below), some bored piles with a water/cement ratio of .,only O-2 were made at Barnet. Ex-
amination of the soil near one of these piles showed that the water content adjacent to the
pile was of the same order as that at greater distance, so that. local softening had, in fact,
been avoided. On the other hand, every one of these piles failed by crushing of the concrete
near ground level (Table 3). The observed crushing strength governing the bearing capacity
ranged from about 150 lb. per square inch for light ramming to about 300 lb. per square inch
for heavy ramming, which is about one-quarter of the crushing strength of corresponding
test cubes. It is of interest to note that the value of 300 lb. per square inch was also reached
by the above pile, No. S.2, at Southall, when the small skin friction above the failing section
is deducted from the maximum load.
BEARING CAPACITY OF PILES IN LONDON CLAY 277
Table 3.-Results of pile loading tests
- - -

Pile Len-
Observed
ultimate
bearing cap-
I Estimated
bearing capacity
-
Piling
tests : Hiley’s
Site Dia: Remarks
No. gth: acity (after Point Skin Total observed Iormula
1 month) estimated
in. feet (t) 6 ;I, a
_- -- -- _- _.

s.l 12 20 10 12
5.2
s.3
12
12 :8
&
34.5
16
16 41
Concrete fail we

Observed skin
s.4 12 40 43 16 27
friction=3lt
s.5 14 31.5 14 14
S.6 14 :8
Southal s.7 14 40 tz 4: ;z
Plate
P.l 12 21.5 7.3 10 - 10
P.2 12 31.5 12.1 16 - 16
P.3 12 39 10.0 16 - 16
:asinl
c.1 12 31.5 -
c.2 12 39 f:; -
_- _- _-
14 [ a.51 16 LO.31 Concrete failure
T.i
T:3
14
14
WI
:z
:t r0.41 3. .,
g1 [;‘;I 9. 8,
Bamet T.4 14 20 4:

2 sq. 26 96 18 142 160 O-6 Driven piles


2 sq. 13.5 42 12 38 50 0.8 ,, ,,
- -

LOADING TESTS ON STEEL PLATE

Loading tests were also made on circular steel-plate footings at the bottom of a borehole,
in order to obtain measurements of the base resistance. A borehole was augered and con-
tinuously lined with British Standard borehole casing of 1Zinch nominal diameter. The
actual dimensions of this casing were :-
Internal diameter .. .. .. .. .. .. 12 inches
External diameter .. .. .. .. .. .. 129 inches
External diameter of steel cutting shoe .. .. .. 14 inches
Diameter over screwed couplings .. .. .. .. 139 inches
The hole was augered an inch or so beyond the end of the casing, and any loose material
removed and the bottom carefully trued up. This operation is, of course, somewhat difficult
at the bottom of a borehole, and it is likely that slight inaccuracies in bedding occurred.
Loading tests were made on the plate at various depths until the ultimate bearing capacity
was registered. A loading test was then made on the casing described above. After the
test at 39 feet, the steel-plate footing was raised 1 foot from the bottom of the borehole,
and suspended from the ground surface by rods. The casing was then withdrawn while
concrete was poured,down the borehole to form a concrete pile about 14 inches diameter, with
its base suspended above the bottom of the hole. When the concrete had hardened, a load-
ing test was made. In the first part of this test, the resistance was due entirely to skin
friction, but at a penetration of about 11 inches there was a sudden increase in resistance
as the base of the pile came into contact with the bottom of the borehole, and for the remainder
of the test the pile behaved as a normal bored pile.
The results of the tests made by loading the steel casing are interesting in that the low
278 MEYERHOF AND MURDOCK

Figs 9

Distribution of water content


in clay near bored pile No. S. 2
BEARING CAPACITY OF PILES IN LONDON CLAY 279
shaft friction suggests that the clay had not closed completely on the pipe after passage of
the cutting shoe and couplings. The casing at each stage was loaded within 24 hours after
being driven, and it is probable that the friction indicated by the tests is almost entirely that
between the cutting shoe, couplings, and the clay. This lack of full contact of the clay may,
possibly, have slightly reduced the ultimate load on the footing plate, compared with the
load obtainable on the base when the hole is filled completely.

ANALYSIS OF TEST RESULTS

Bearing Capacity
The static bearing capacity, Q, of piles which are sufficiently strong to withstand the
installation and loading stresses, can be estimated from bearing-capacity theory, and for
saturated clay it can be represented (Terzaghi, 1943) by :-
Q=(cNc+D). A+c,F
In this equation the first term on the right-hand side represents the point resistance,
where c denotes average cohesioqat pile point
fiC ,, bearing-capacity factor = density of soil
D 9, depth of pile point (at base)
A >> area of pile point
and the second term represents the skin friction
where JZ, denotes average adhesion along shaft
F surface area of pile shaft
PreviouL;heoretical and experimental evidence (Meyerhof, 1951, Skempton, 1951) indicates
that the factor NC is of the order of 9, and that the cohesion c is the undisturbed shearing
strength of the clay. A check of this, at least so far as the present bored piles are concerned,
can be obtained from the plate loading tests at various depths in the borehole of pile No.
S.4, at Southall (Rodin and Tomlinson, 1952), where NC was found to range between 8.0 and
12.0 with an average of 9.4. Since the weight of the piles is practically the same as the over-
burden load, the net point resistance 9cA has been estimated for the various pile depths and
is given in Table 3.
The average adhesion on the shaft of bored piles can readily be estimated from the fully
softened shearing strength, c,, of the clay under zero overburden, a method which was found
(Meyerhof, 1951, Ward and Green, 1952) to be in reasonably good agreement with the results
of loading tests on short auger-bored piles. The skin-friction tests on pile No. S.4 at Southal
confirms this method, which is further supported indirectly by the observation that the water
content at pile shaft No. S.2 was of the same order as that found in laboratory compression
tests on samples which have been allowed to soften under zero overburden (Figs 9). On
account of the arching of the unset concrete in the auger boreholes, the effective lateral
pressure, even of long piles, is not expected to exceed that corresponding to a depth of a few
pile diameters. The very smaI1 magnitude of this lateral pressure is also indicated by the
absence of noticeable increase in bearing capacity due to any re-consolidation of the softened
clay along the shaft even 18 months after installation of the piles. The skin friction has,
therefore, been estimated from c,F (see Table 3). In the case of the driven piles, the
adhesion corresponds to the fuIIy remouided shearing strength, which for the present clay
was found to be practically identical with the undisturbed strength, so that the theoretical
skin friction may be taken as CF.
Apart from the cases of concrete failure mentioned above, the observed and estimated
total bearing capacities of the bored piles compare reasonably well for practical purposes.
The observed bearing capacities of the driven piles are, however, less than estimated, the
difference increasing from 15 per cent for the short pile to 40 per cent for the long pile. Since
the difference increases with greater penetration into the hard clay, the adhesion between
2*
280 MEYERHOF AND MURDOCK

concrete and soil seems to be a limiting factor. After analysis of previous pile loading tests
it was suggested (Meyerhof, 1951) that the adhesion had an upper limit of about 1 ton per
square foot ; the present loading tests indicate a maximum adhesion of about 2,000 lb. per
square foot when some allowance for the relatively small skin friction in the thick gravel
layer is made. It is not possible to relate this adhesion to the laboratory adhesion tests except
on the tentative assumption that the skin friction increases in direct proportion to the over-
burden, when the average earth-pressure coefficient on the shaft is of the order of 3 to 4.
The residual bearing capacity of the driven piles was about 20 per cent less than the maxi-
mum, compared with the drop of 20 to 50 per cent in the adhesion determined by laboratory
adhesion tests. The dynamic bearing capacity on the basis of Hiley’s pile-driving formula
was up to 50 per cent less than the observed bearing capacity, and does not even give an idea
of the relative bearing capacities at the two test levels in the present case. It should be
noted also that the driven piles carried double the load of the bored piles of similar dimensions,
which would indicate the importance in stiff soils of avoiding softening of the clay by the
concrete.
It should be noted that this statement may be applicable only to bored piles in which
no casing tube is used, and in which the release of pressure in the hole after boring has probably
allowed the fissures to open sufficiently to result in the appreciable decrease in the shear
strength which has been measured. If the borings had been lined with a flush casing,
as is, in fact, quite common practice in many soils, and had the concrete been rammed out
through the bottom of the tube while withdrawing it, this release of pressure might not have
occurred, and a higher bearing capacity for the pile might have been expected.
However, it is clear that until more definite evidence is available, the possible advantages
of using a casing tube cannot be taken into account in designing bored piles. The choice,
then, is mainly a matter of cost, that is, between the cost of sinking and withdrawing casing
tube with a pile of somewhat similar diameter to a driven pile in the same soil, and the cost
of the extra concrete necessary in the larger-diameter pile required to give the same bearing
capacity when the casing tube is omitted.

Settlement
Since the plate and pile loading tests were carried out sufficiently fast to avoid consolida-
tion of the clay, the observed movement during the test represents immediate settlement.
An approximate estimate of this movement of a plate may be made using either elastic
theory (Skempton, 1951) or plastic theory (Meyerhof, 1951). Both methods indicate that the
immediate settlement is approximately given by :
S = 2eB
where B denotes width of plate
e strain in compression test at same percentage of ultimate load as for plate.
The laboiatory stress/strain curves of the present undisturbed clay between 20 and 40
feet depth were found to lie in the shaded zone of Fig. 10 (a), so that the theoretical settlement
should lie between the limits indicated in Fig. 10 (b). Up to about one-half of the ultimate
load, the observed plate settlements are of the same order of magnitude, but for greater loads
the movements are much larger than estimated.
As a first approximation this theoretical method may also be used to estimate the immed-
iate settlement of bored piles ; this assumes that the skin friction is relatively small and re-
quires only a small movement for full mobilization, as shown by the results of the composite
pile test. The observed settlements of the bored piles are again of the same order as estimated
up to about a half of the ultimate load (Fig. 10 (c)) ; for greater loads the settlement of most
piles is under-estimated in spite of the conservative assumption that the whole load is effect-
ively carried by the pile point.
With driven piles, however, most of the load is carried by skin friction, and on the
BEARING CAPACITY OF PILES IN LONDON CLAY 281

simplifying assumption of zero point resistance a theoretical approach similar to that above
shows that : S = e B approximately.
The stress/strain curves of the remoulded clay were generally within the limits shown in
Fig. 10 (a), so that the settlement of the driven piles should be within the range indicated in
Fig. 10 (c). Although the observed movements follow the shape of the theoretical curves,
they are of rather smaller magnitude.

COWRESSIVE STRESS
STRENGTH
To
a

(a) RANGE OF STRESS/STRAIN RELAT~ONSH~F~


OF TRIAXIAL COMPRESSION TESTS

(b) SElTLEMENT OF PLATES

Figs 10.
Settlement of
piles and plates

(C) SElTLEMENT OF BORED AND DRIVEN PILES

The settlement of the piles increases with time under load owing to the creep and con-
solidation of the soil. To study this movement for single piles, some of the short bored piles
at Bamet have been kept under a constant load of about one-third of the ultimate for about
a year so far, as part of the research programme of the B.R.S. In that period, practically
282 MEYERHOF AND MURDOCK

the whole of the movement occurred during the first 3 months and about one-half of total
observed settlement took place during application of the load. If a similar relationship on
long-term loading may be expected for the present long piles, a factor of safety of 3 on the
ultimate bearing capacity would limit the settlement to about 1 inch in the case of widely
spaced piles in similar stiff to hard London clay (see Fig. 10 (c)).

CONCLUSION
The present investigation indicates that long bored in-situ piles can be installed by a
power auger in stiff clay much faster and more economically than they can be driven, especially
where the cost of transport of a complete piling frame is not warranted. On the other hand,
the bearing capacity of a bored pile installed in an unlined hole in stiff clay is much less than
that of a similar driven pile, because of softening of the clay adjacent to the pile. The soften-
ing is caused by the migration of water from the unset concrete to the clay, and is limited
to the soil within about 2 inches of the pile. Although it has been found possible to avoid
the softening by using a very dry mix, the concrete cbuld not be sufficiently compacted, even
by heavy tamping, and the corresponding piles failed prematurely by crushing of the shaft. It
is possible that the softening could be reduced or avoided by installing bored piles in a driven
lining tube ; even if the tube is subsequently withdrawn, the prestressing of the soil might
reduce considerably the swelling and opening of fissures of the clay. However, this method
decreases the speed of installation, so that the choice of procedure is largely one of economics.
The-bearing capacity of structurally sound bored piles, with adequate control of mix
and placing in unlined boreholes, can readily be estimated from bearing-capacity theory
using the fully softened shearing strength at the point. In fissured clays, the shear strength
has to be ascertainedas soon as possible after sampling, since it has been found that after
a few days the strength may drop to about one-half of the original value due to gradual
opening of fissures. Triaxial tests on such material indicated that the fissures could not be
sufficiently closed by lateral pressures of as much as twice the overburden pressure.
The point resistance of driven piles can be estimated as for bored piles, and their skin
friction is based on the remoulded shearing strength (or undisturbed for insensitive clay)
along the shaft with an upper limit of the strength of 2,000 lb. per square foot. The signi-
ficance of this limit is not fully understood, and a tentative correlation with the results of
exploratory laboratory adhesion tests is suggested, at least for the present piles. At a factor
of safety of 3, the settlement of widely spaced bored and driven piles in stiff and hard clay
is not likely to exceed 1 inch, and approximate methods have been suggested for estimating
the immediate movement of such piles.

ACKNOWLEDGEMENTS
The Authors are indebted to their colleagues, particularly Mr M. J. Tomlinson and
Mr T. K. Chaplin, for assistance in the field and laboratory work.
The work was carried out jointly by the Central Laboratory of Messrs George Wimpey
& Co. Ltd, and the Building Research Station of the Department of Scientific and Industrial
Research. The Paper is published by permission of the Directors of the firm and by the Direc-
tor of Building Research.
Crown copyright of this article is reserved.

REFERENCES
DAVEY, N., 1949. Short concrete piles for foundations on shrinkable clays. J. Roy. Inst. Bit. A&.,
57: 1:24.
MEYERHOF, G. G., 1951. The ultimate bearing capacity of foundations. Gkotechnique, II : 4 : 301.
SKBMPTON, A. W., 1951. The bearing capacity of clays. Bldg Res. Congress (London), 1 : III : 180.
TERZAGHI, K., 1943. Theoretical soil mechanics. Wiley, New Yovk.
WARD, W. H. and GREEN, H., 1952. House foundations : The short-bored pile. Publ. Wks and Munic.
Serv. Congress (London), p. 373.

Вам также может понравиться