1282

Bearing capacity of soils for crawler cranes

Xiteng Liu, Dave H. Chan, and Brian Gerbrandt

Abstract: Few studies have been carried out on the bearing capacity of soils for crawler cranes. Directly applying the
bearing capacity equations used for shallow foundations to cranes often leads to conservative design. The total settlement
is of less concern for cranes than for buildings, and cranes can normally tolerate larger differential settlements. Computer
simulation and field studies have been carried out to study the allowable bearing capacity of soils for cranes. Equations
modified from the traditional method to calculate the bearing capacity for shallow foundations have been proposed. In general, it was found that the bearing capacity for crawler cranes could be increased by as much as 50% from that for foundations. This depends on the soil type, crane, and mat configuration. A design procedure in evaluating foundation support
for crawler cranes is also proposed.
Key words: soil bearing capacity, crawler crane, crane track pressure.
Resume : Il y a eu peu detudes realisees sur la capacite portante des sols pour les grues a` chenilles. Lapplication directe
aux grues des equations de capacite portante utilisees pour des fondations superficielles conduit souvent a` une conception
conservatrice. Le tassement total presente moins de souci pour des grues que pour des batiments, et les grues peuvent normalement tolerer des tassements differentiels plus importants. Une simulation par ordinateur et des etudes sur le terrain
ont ete realisees pour etudier la capacite portante admissible des sols pour des grues. On a propose des equations modifiees
par rapport a` la methode traditionnelle pour calculer la capacite portante pour les fondations superficielles. En general, on
a trouve que la capacite portante pour les grues a` chenilles pouvait etre augmentee par autant que 50 % par rapport a` celle
pour les fondations. Ceci depend du type de sol et de grue, et de la configuration du tapis. On propose aussi une procedure
de conception pour evaluer la fondation portante pour des grues a` chenilles.
Mots-cles : capacite portante des sols, grue a` chenilles, pression sur la chenille de la grue.
[Traduit par la Redaction]

Introduction
Soil bearing capacity for shallow footing foundations has
been studied extensively from the 1920s to the 1970s. It is
generally accepted that the two criteria that govern the allowable bearing capacity of soil for shallow foundations are
the shear strength of the soil and the tolerable settlement of
the foundation.
In practice, the allowable bearing capacity of soil is determined using either the direct or the indirect methods. The
direct method is to calculate the ultimate bearing capacity
directly from the strength properties of the soil. Classical
bearing capacity equations proposed by Terzaghi (1943),
Meyerhof (1956), and Vesic (1973) are commonly used.
The ultimate bearing capacity is further divided by a factor
of safety (FS) to obtain the allowable bearing capacity. The
FS takes into account uncertainties including the variability
of soil resistance, limitations of the theory, deformation of
Received 9 June 2006. Accepted 23 February 2008. Published
on the NRC Research Press Web site at cgj.nrc.ca on 28 August
2008.
X. Liu. Geotechnical Engineer, AMEC Earth & Environmental,
4810-93 Street, Edmonton, AB T6E 5M4, Canada.
D.H. Chan.1 Department of Civil & Environmental Engineering,
3-038, Markin/CNRL Natural Resources Engineering Facility,
University of Alberta, Edmonton, AB T6G 2W2, Canada.
B. Gerbrandt. Sterling Crane, 2440-76 Avenue, Edmonton,
AB T6E 6R2, Canada.
1Corresponding

author (e-mail: dave.chan@ualberta.ca).

Can. Geotech. J. 45: 12821302 (2008)

the ground, etc. The typical FS range for shallow foundations is between 2.5 and 3.5. The settlement of the foundation should also be evaluated, and the allowable bearing
capacity may need to be further reduced if the estimated settlement exceeds the maximum allowable settlement.
The indirect method is to evaluate the soil bearing capacity from the in situ soil tests using empirical or semiempirical correlations. In North America, the allowable
bearing capacity is commonly estimated from the blow
counts of the standard penetration test (SPT) or the tip resistance of the cone penetration test (CPT). More discussion
on the bearing capacity of soils can be found in Gupta
(2002), Lee and Salgado (2005), and Liu (2005).
There is no specific method to evaluate the allowable
bearing capacity for cranes. Traditionally, the allowable
bearing capacity for shallow foundations was used directly
for cranes. This is not appropriate as there are differences
between foundations for buildings and cranes in load duration, allowable differential settlement, importance of the
project, etc. Some authors (Shapiro et al. 1999) and research
institute, CIRIA (1996), suggest using a smaller FS (between
1.5 and 2.0) to calculate the allowable bearing capacity for
cranes supported by outriggers. However, this cannot be
simply applied to crawler cranes because the pressure under
the track is not evenly distributed and the differential settlement of the ground that may cause overturning of a crane
can be significant.
The objective of this study is to improve the current
methods in evaluating the bearing capacity for crawler
cranes and to provide a design procedure to determine the

doi:10.1139/T08-056

2008 NRC Canada

Liu et al.

foundation supports. Since the allowable bearing capacity of

soil for cranes is also controlled by both the shear strength
and the settlement of the soil (similar to that for foundations), it is feasible to modify the existing bearing capacity
equations to make them suitable for cranes. Most studies
have been carried out using computer simulations. Several
case studies have been done to test the proposed method for
bearing capacity calculations for crawler cranes.

1283
Fig. 1. Crane track pressure distribution through timber mats. L,
track length; B, footing width; B, equivalent footing width.

Track loading of a crawler crane

Load rating of a crane
Load rating is the most basic aspect of a crane. It is the
maximum allowable load for a specific radius in a particular
configuration while operating under defined conditions. The
load rating of a crane can be limited by the loadhoist rope
strength, the available line pull at the winch, the structural
strength of the crane, and the stability against overturning.
In most cases, the load rating of a crane is controlled by
stability against overturning unless at very short radii the
load rating can be governed by the other factors. One hundred and seventy-six mobile crane accidents prior to 1978
were studied and more than 71% were found to have been
caused by overturning (Shapiro et al. 1999). In most countries, it is accepted that the stability based load rating is set
as a percentage of tipping load. For crawler cranes, this percentage is set to be 75% in the United States and Canada,
while it is 66.7% in Europe.
Track pressure determination of crawler cranes
The lifting capacity of the mobile crane increased dramatically in the 1970s with the use of high-strength fine-grained
steel. DEMAG broke the 800 t (1 t = 1000 kg) limit in 1978
and the maximum lift capacity of a crawler crane is about
2600 t today. With the increase in the crane lifting capacity,
the forces that the crane exerts on the ground increase significantly. The crane force is distributed through the two
tracks of a crawler crane. The actual track pressure depends
not only on the magnitude and location of the force, but also
on the relative stiffness of the track and the ground. However, most analyses assume that the crawler frame and car
body are absolutely rigid. If the overall magnitude of the
load and the location of the centre of gravity are known,
the track pressure can be easily calculated using the principle of equilibrium. Although the rigid assumption works
well in general, the calculated track pressure rarely equals
that of the actual track pressure. Furthermore, because of
the rigidity assumption, the two tracks will have the same
eccentricity in the direction along the crane tracks.
The essential feature of the track pressure is its eccentricity, which is the major cause of the differential settlement of
the ground. The maximum track pressure frequently occurs
when the total centre of gravity of the crane weight and the
load are swung out at right angles to the direction of travel
by approximately 508608 (Becker 2001).
Track pressure spreading through timber mats
The maximum track pressure for larger cranes can be as
high as 2000 kPa, which is far beyond the ground bearing
capacity for most soils. Timber mats are usually used to
spread the track pressure to a larger area in case the track

pressure is high and the ground is not able to sustain the

load. Figure 1 illustrates the track pressure distribution and
spreading through a mat.
Traditionally, the track pressure through mats is assumed
to be uniformly distributed, and the extent of the pressure
distribution (equivalent footing width B shown in Fig. 1) is
governed by the longitudinal shear strength of the mat. Considering the fact that the mat is a reverse cantilever beam
supported by the crane track in the middle and loaded by a
distributed pressure by the ground, the maximum shear
stress at midheight of the mat can be calculated for a given
pressure spreading width. It is then assumed that the track
pressure could only be spread to an extent that the maximum shear stress of the mat does not exceed its allowable
longitudinal shear strength.
Crane levelness and maximum tilt angle
The rated load of a crane is based on the assumption that
the machine is standing on a firm, uniformly supported level
surface. This means that the ratings given are appropriate to
use only if the tracks are properly supported so that throughout the operation the crane will remain level to within 1%
(ISO 1991). Some manufacturers such as DEMAG specify
that the levelness of site preparation should be within 0.5%.
Nevertheless, it is common practice to set the ground deformation criterion to be 0.5% or 0.38 throughout the crane
operation.
It should be mentioned that the tilting of the crane track
(or the differential settlement of the ground) is not the same
as the tilting of the crane superstructure. The crane body and
boom may deflect significantly under a bending moment. As
a consequence, the tilt angle of the superstructure can be
much higher than the ground surface.

Computer model for bearing capacity

calculation
The study of bearing capacity for cranes is carried out using both numerical modeling and field observations. The
#

2008 NRC Canada

1284

field measurements provide a powerful means of calibrating

the parameters used in computer models and in verifying the
modeling results.
Model size verification
Ideally, a three-dimensional (3-D) computer program
should be used to model the ground response under the eccentric crane loading. However, a 3-D program was not
readily accessible for this study; therefore a two-dimensional
(2-D) finite difference program FLAC4.0 was used for numerical modeling. To counteract some of the limitations involved in using a 2-D program to model a 3-D problem, the
following provisions were undertaken.
Modeling multiple sections
In addition to modeling the section in the crane track
width direction, the two sections in the track length direction
were also modeled in the crane lift study. The plane strain
configuration used in the 2-D program assumes that both
the soil and the footings are infinitely long in the direction
perpendicular to the section it modeled. This is more applicable to the section in the track width direction, whereas
modeling the two sections in the track length direction is
not as reliable. However, these two sections are more critical in simulating the ground reaction during the crane operation because differential settlements mainly occur in the
track length direction.
Using proper model size
Theoretically, the ground settlement will increase with
the increase of the depth in the calculation. In practice, the
depth used in settlement calculations is usually chosen to be
the depth where the vertical stress induced by net contact
pressure is less than 1/10th of the net contact pressure. In
homogeneous soil, the influence depth is about two times
the footing width (2B) for square footing and about four
times the footing width (4B) for strip footing.
For cranes sitting directly on the ground without mats, the
L/B ratio (where L is the track length) is generally between
4 and 6, and a model depth of 4B is considered reasonable.
For cranes sitting on mats with an L/B ratio (where B is the
equivalent footing width as discussed later in the section entitled Equivalent footing width) generally between 1.2
and 2.0, a model depth of 2B may be used.
Elastic solution for footings on a finite soil layer was used
in evaluating the errors caused by using a 2-D program with
limit model depth (2B) to simulate the settlement of the
crane track with mats. The results showed that for L/B ratios ranging between 1.2 and 2.0 and Poissons ratio ranging
between 0.2 and 0.4, the error was generally within 10%.
However, since soil is not homogeneous, this effect cannot
be easily evaluated. Modeling sections in the track length direction is not as accurate as modeling sections in the track
width direction.
The model width has less influence on the settlement
study unless it is too small. In this study, the lateral boundary of the model was set to be at a distance of 3B from
the footing centre.
Model setup
Various models were built in this study for different ap-

Can. Geotech. J. Vol. 45, 2008

plications. The two sections for a typical crane lift are illustrated in Fig. 2. In general, three major components: the soil
mass, the crane mats, and the crane track were modeled.
The soil mass was typically modeled as multilayers of
elasticplastic materials. The properties and thicknesses of
each soil layer were taken from field and laboratory results.
The crane mats were modeled as cross isotropic elastic material with different modulus of elasticity in the longitudinal
and transverse directions. The crane track was modeled as
an elastic block with a modulus of elasticity of about 1/30th
to 1/10th of the modulus of steel. The modulus reduction is
to account for the difference between the actual steel section
of the track and the footing width used in the model.
The contacts between the soil and crane mats, between
the crane mats and the crane track, and between layers of
crane mats were modeled by interface elements. The track
pressure was applied on the crane tracks to account for the
interaction between the track and the mats.

Assumptions in the analyses

The following assumptions are made in the analyses of
the track pressure of a crawler crane:

Soil behaves in a linear elastic manner under the crane
load Since the crane track load is short-term, finegrained soil is likely to deform in an undrained fashion.
Total stress analysis is appropriate to evaluate the undrained bearing capacity and the immediate settlement.

Crane tracks and mats act as two spread footings
When timber mats are used to spread the crane track
pressure, the mats are placed such that each track sits at
the centre of each row as shown in Fig. 1. Therefore, the
mats for the two tracks are not connected. Although
sometimes only one row of long mats might be enough
for small cranes, it is still valid to assume that the two
tracks and mats behave as two spread footings.

The settlement interference between two tracks is negligible If two footings are placed close enough to each
other, their zones of influence will interact and the failure
surface may change from that of an isolated footing. The
bearing capacity will in general increase in this case
(Stuart 1962). On the other hand, the adjacent footing
will cause additional stress in the soil and will therefore
induce more settlement than an isolated footing. However, the degree of influence of an adjacent footing is
highly dependent on the soil type, footing type, and distance between two footings. A simple 2-D computer
model was used to qualify the interference between the
two adjacent footings. In the model, two strip footings of
the same width at various spacings were placed on top of
an elastic medium (soil). A uniform pressure was then
applied to both footings. Results of the model showed
that the influence of an adjacent footing on the footing
settlement dropped dramatically with an increase in the
distance between the footings. For cranes sitting directly
on the ground, the distance between two tracks is much
greater than the width of the track. Therefore, there is
practically no interference between two tracks. For cranes
sitting on mats, the typical ranges for span to width ratio
S/B and length to width ratio L/B are about 1.51.8 and
1.22.0, respectively. Taking the Poissons ratio for the
#

2008 NRC Canada

Liu et al.

1285

Fig. 2. Typical model setup for crane lifts. S, distance between tracks.

soils as 0.3, the additional settlement caused by an adjacent footing is about 2% to 10% compared to the settlement caused by the footing itself.

Ignore the craneground interaction The actual track
pressure distribution is dependent not only on the load
and moment applied on the crane, but also on the relative
rigidity of the crane, mats, and soils. The craneground
interaction tends to reduce the differential settlement and
equalize the pressure distribution. This effect is very
complex and it is ignored in this study.

Convert the bearing capacity problem for

cranes to footing foundations
As there are some fundamental differences between the
bearing capacities for footing foundations and for cranes, it
is necessary to convert the problem of bearing capacity for
cranes to the traditional bearing capacity problem for footing
foundations. To achieve this, the following three steps have
been taken: (i) use the maximum allowable settlement to
represent the levelness criterion for cranes; (ii) use an equivalent uniformly distributed pressure to represent the triangular or trapezoidal distribution of the track pressure; and
(iii) determine the equivalent footing width for a crane with
mats.
Convert the levelness criterion to allowable settlement
The out of levelness of a crane should be within 0.5%
during its operation. It should be noted that the tilting of a
crane is not a one-dimensional (1-D) problem but rather is a

2-D problem. The degree of tilting varies during the superstructure of a crane skew with a constant load and radius. It
is not practical to calculate the soil bearing capacity by evaluating the tilting of the crane at various boom orientation
angles. As a result, only the maximum tilting angle is proposed in the analysis. The maximum tilting angle usually
occurs at the following three critical boom orientations: the
boom is parallel to the tracks, the boom is perpendicular to
the tracks, or the boom is over the corner of a track.
The key factor for converting the levelness criterion to allowable settlement is to find a representative tilting angle
that can be expressed as a function of the maximum settlement. Since the representative tilting angle is directly related
to the allowable levelness of the crane, it is ideal to select
the representative tilting angle as one that is slightly greater
than the maximum tilting angles at the three critical boom
orientations to ensure they are within the limits of allowable
levelness.
The representative tilting angle is selected by observing
and estimating and then verifying by theoretical analysis.
Consider a typical crane operation as shown in Fig. 3. Assume the total crane load of G is acting at a radius R from
the crane rotation centre and at an angle a off the centreline
perpendicular to the crane tracks. The crane track has a dimension of BL with a span S between the two tracks. For
most cranes, the maximum pressure occurs at an angle a =
508~608, and the track span S is often equal to or a little bit
greater than the track length L. The tilting angles at the three
boom orientations can be expressed by the following using
the elastic solution for settlement of rigid footing:
#

2008 NRC Canada

1286

Can. Geotech. J. Vol. 45, 2008

Fig. 3. Diagram for crane tilting calculation.

1 2 Gt

n 

1

l 

EL2
1 2 Gt
4ELB

Iw
Im

2 1 2 G

 

EL2

4tcosI w

Lt1tcos sin
4B

Im5

where qn, ql, and qa are the tilting angles in the directions
perpendicular, parallel, and at an angle a to the tracks, respectively; E and n are the elastic modulus and Poissons ratio of the soil, respectively; Iw is the influence factor for the
settlement of a rigid footing; Im is the influence factor for
the rotation of a rigid footing; a is equal to 508~608; and
load eccentricity t = 2R/S.
The representative tilting angle, qrep, is defined as the
maximum settlement, dmax, over the length of crane track L,
which is
2

max
L


1   2 G1 tcos 
Ltsin 
Im
Iw

2EL2
4B

rep

where a = 508~608.
To compare this representative tilting with those three angles at different boom locations, it is convenient to use the
ratios of the three angles to the representative angle as:
3

rn

n
;

rep

rl

l
;

rep

ra

a

rep

Figure 4 shows the variation of these three ratios rn, rl,

and ra with the length to width ratio L/B of the footing and
the degree of load eccentricity t = 2R/S.
Because the rated load of a crane is based on one criterion
that requires at least a factor of 1.33 against overturning
(Shapiro et al. 1999), the load eccentricity t is rarely greater
than 0.75 during normal crane operation, especially for
heavy lift cranes, which usually have superlift counterweights that can move the centre of gravity further away
from its fulcrum.
For cranes sitting on mats, the L/B ratio (use equivalent

footing width as described later in the section entitled

Equivalent footing width) is usually between 1.2 and 2.0
and is rarely over 2.5. Within this range, these three ratios
are all less than 1, which indicates that the representative
tilting angle is larger than the maximum tilting angle at the
three critical boom orientations. As a consequence, the allowable settlement for a crane with mats can be conservatively expressed as
4

max
rep L  0:5%L

L

200

where [d] is the allowable maximum settlement of the crane

during operation.
For cranes sitting directly on the ground, the L/B ratio is
generally between 4 and 6, and the maximum tilting angle
could be as high as 1.2 times that of the representative tilting angle. To account for this effect, a factor of 1.2 should
be applied and the allowable settlement for crane without
mats can be written as
5

max
rep L 

0:5%
L
L

1:2
240

The concept of equivalent pressure

Unlike a footing foundation, the track pressure of a crane
changes greatly with changes in load, radius, and boom orientation during its operation. Another major feature of the
track pressure is that it is rarely uniformly distributed along
the track length due to the eccentricity of the crane load. Assuming the crane frame and tracks are absolutely rigid, the
theoretical track pressure shape is either a triangular or trapezoidal shape. The maximum pressure is usually used as
the design pressure in the evaluation of the soil bearing capacity. This maximum pressure usually occurs when the
crane boom is over the corner of the crane tracks. The use
of the maximum track pressure to represent the triangular or
trapezoidal distributed pressure seems to be conservative. To
find a design pressure that can better represent this pressure
distribution leads to the concept of equivalent pressure. The
equivalent pressure is a uniformly distributed pressure along
the whole track length that will cause the same amount of
settlement at the point where the maximum pressure is taking place. The equivalent pressure is first derived from theo#

2008 NRC Canada

Liu et al.

1287

Fig. 4. Variation of tilting ratios.

retical analysis based on an elastic assumption and then

verified by computer simulation.
Consider a rectangular footing with a dimension BL sitting on the ground as shown on Fig. 5. A linearly varying
pressure q1 to q2 is applied along the track, and the pressure
is uniform in the direction along the track width. Using the
theory of superposition and the theoretical settlement solution for rigid footings, the total settlement at points M and
N are
6

L
M N l  tan

2

1   2 q1 q2 B
q1  q2 L

Iw 
Im
E
2
24

Now, assuming an equivalent uniformly distributed pressure q~ that causes the same amount of settlement at point
M, the settlement can be expressed in the form of q~ as
7

M q~

1   2 B
Iw
E

Hence, the equivalent pressure q~ can be found to be

q1 q2
q1  q2
k
2
2
Im L
k
12Iw B
q~

8

A plot of k as a function of L/B is made using the theoretical Im and Iw values for rigid footings (Bowles 1982) to

find an empirical correlation in Fig. 6. It shows a fairly

good linear correlation between k and the ratio L/B. Since
the ratio L/B (use equivalent footing width as described
later in the section entitled Equivalent footing width) for
cranes with mats varies from 1.2 to 2.0, within this range,
the maximum factor k is about 0.7. Therefore, the maximum
equivalent pressure for the crane with mats is
9

q~

q1 q2
q1  q2
0:7
0:85q1 0:15q2
2
2

For cranes sitting directly on the ground, the L/B ratio is

generally greater than 4, the equivalent pressure is no longer
valid, and the maximum pressure should be used in the design.
In traditional foundation design, Meyerhof (1956) introduced the equivalent bearing pressure to account for the
load eccentricity as:
10

qeqv

P
q1 q2 BL 31 s2

q1

B0 L0 2L  2eL B 41 2s

where qeqv is the equivalent pressure proposed by Meyerhof

and s = q2/q1 (0 s 1).
It is easy to prove that the equivalent pressure proposed in
eq. [9] is always larger than the equivalent bearing pressure
for foundations. This means that the proposed equivalent
track pressure provides a higher FS against bearing capacity
failure due to eccentric loading.
Numerical models built with FLAC4.0 have also been
used to verify the validity of the equivalent pressure. As
#

2008 NRC Canada

1288

Can. Geotech. J. Vol. 45, 2008

Fig. 5. Diagram for settlement calculation of rigid footings.

Fig. 6. Correlation of factor k with L/B ratio.

shown in Fig. 7, the footing used in the model was assumed

to have a length of L and a constant height of 1.5 m. The
width of the footing B (perpendicular to the paper) was assumed to be 6 m. The Youngs modulus of the footing was
assigned to be 1/15th of the steels modulus to account for
the difference between the steel section of the crane track
and the assumed footing width. The size of the soil model
was chosen according to the section entitled Model size
verification. The soil was assumed to be elastic with an arbitrary modulus of elasticity of 50 MPa. The maximum
pressure q1 applied on the footing was assumed to be
200 kPa in the study.
The model was run with the variation of the stress ratio q2/q1,
the shape factor L/B, and the Poissons ratio n. Figure 8 illustrates the ratio of the maximum settlement by eccentric loading (dact) to the settlement by the equivalent pressure (deqv)
from the computer simulation. In general, this ratio is less
than 100%, which means that the use of equivalent pressure
leads to a conservative design approach. However, the eccentric loading tends to yield larger settlement (4%) than
the equivalent pressure with high Poissons ratio (n ? 0.5),
low stress ratio (q2/q1 ? 0) and large L/B ratio (L/B = 2.5).
This may be due to the limitations of the 2-D simulation, the
shallow model depth for large L/B ratio, the change of relative rigidity between crane track and the soil, etc.

Equivalent footing width

The traditional way to determine the track pressure distribution through crane mats is based on the strength of the
timber mat. However, it is not only limited by the strength
of the mat, but it is also a function of the soil type, the elastic properties of the soil and mat, the strength of the soil and
mat, the geometry of the mat, the track width, and the stress
level etc.
It is very important to adequately estimate the spreading
areas this will provide two major parameters in footing design: the bearing pressure and the appropriate footing width.
Computer simulation was used to find the correlation between the factors listed above and the equivalent footing
width.
Two computer models were used with variations of soil
type, elastic parameters of soil (E and n), mat type, track
width B0, thickness of mat d, and stress level q. The first
computer model was constructed to simulate the crane sitting on timber mats on homogeneous ground. For the reason
of symmetry, only half of the track was modeled as shown
in Fig. 9a. The model was then loaded to 10, 20, 30, 40,
and 50 mm deformation at the centre of the track. Stress
and displacement at the ground surface, as well as the track
pressure, were recorded for each stage of loading. Only
stages with 3050 mm settlement are of interest as the al#

2008 NRC Canada

Liu et al.

1289

Fig. 7. Computer models for equivalent pressure verification.

lowable maximum settlement for most cranes is within this

range.
The second computer model was built without timber
mats, and a uniformly distributed pressure was directly applied on the ground surface, as shown in Fig. 9b. The width
and intensity of loading was adjusted by trial and error to
achieve similar stress and displacement at the ground surface as that from the first computer model. In this way, the
width of the load can be treated as the adequate spread
width of the track pressure, as it causes similar stress and
settlement.
Figure 10 illustrates the stress and settlement profile for a
typical case where a 1 m wide crane track is sitting on 6 m
long, 0.3 m thick fir mats. The soil used in the model was a
stiff clay with soil properties of E = 50 MPa and v = 0.3.
The solid lines in the figure represent the results from the
first computer model that gives settlements of 30, 40, and
50 mm, respectively. The dotted lines represent the results
from the second computer model by assuming that the pressures from the first computer model are equally distributed
over a certain width. If both the stress and settlement profiles from the second model are close to those from the first
model, the width is found to be appropriate to represent the
equivalent footing width. For this particular case, the equivalent footing width was found to be B = 4.0 m.
From the results of the simulation, it can be found that
soil type, mat length, and Poissons ratio have less impact
on the equivalent footing width. The major factors are the
track width, Youngs modulus of the mat and soil, and the

thickness of the mat, d. Figures 1114 present a series of

plots of the equivalent footing width versus the track
width, the thickness of the mat, and the ratio of Youngs
modulus of the mat and soil. From that, a regression expression is derived as
 0:29
Em
11
B0 B 2d
 Lm
Es
where
B is the equivalent footing width
B is the track width
d is the thickness of the timber mat
Em is the Youngs modulus of the mat (&11 GPa for Douglas
fir and &20 GPa for Mora)
Es is the Youngs modulus of the soil
Lm is the length of the timber mat

Results from computer models also show that the soil

strength has little influence on the spreading width unless
the soil fails; this usually can be avoided by using a FS not
less than 2.0 in practice. Although there is no track pressure
component in eq. [11], the equivalent footing width is a
function of the stress level, and the equation is based on the
track pressure required to cause 3050 mm settlement.
Equation [11] should incorporate the strength of the mat
to ensure that the mat is not overstressed. It should be
pointed out that the allowable stress for mats used now is
also conservative and can be increased by 1.5 due to the
short duration of the load and other advantageous factors.
#

2008 NRC Canada

1290

Can. Geotech. J. Vol. 45, 2008

Fig. 8. Comparison of settlement caused by actual and equivalent track pressure.

Typical soil study

To provide general formulas to calculate the allowable
bearing capacity for crawler cranes, three types of soil were
used in this study:

sand and gravel, which represents cohesionless soils;

soft to medium clay, which represents normally consolidated or slightly overconsolidated cohesive soils; and

stiff clay, which represents heavily overconsolidated soils
(e.g., a glacial till).
For each type of soil, the formulas for the allowable bearing capacity of soil are modified from the prevailing bearing
capacity formula for foundations so as to be easily under-

stood by geotechnical engineers. All the formulae proposed

are based on theoretical analyses and computer simulation
results.
Sand and gravels
The allowable bearing capacity of sand is usually governed by the settlement of foundations. Because it is difficult to obtain undisturbed samples of sand, the elastic
parameters of sand are often derived empirically from in
situ SPT or CPT test results. Equation [12] (from Meyerhof
1956) is one of the most widely used methods nowadays. It
is based on 25 mm settlement for homogeneous sand with a
ground water table below the failure zone of the footing.
#

2008 NRC Canada

Liu et al.

1291

Fig. 9. Models to simulate equivalent footing.

Fig. 10. Settlement and stress profile using equivalent footing width.

Fig. 11. Correlations between equivalent footing width and crane track. Reg, regression.

2008 NRC Canada

1292

Can. Geotech. J. Vol. 45, 2008

Fig. 12. Correlations between equivalent footing width and mat thickness. Reg, regression.

Fig. 13. Correlations between equivalent footing width and Em/Es for clays. Reg, regression.

12

qa 12Nkd
B  1:2 m
0
12
B 0:3A
qa 8@
Nkd
B > 1:2 m
B

where qa is the allowable bearing capacity (kPa); N is the

average SPT blow count within the zone of influence; and
kd = 1 + 0.33(D/B) 1.33.
Bowles (1982) proposed a set of equations similar to
Meyerhofs with a 50% increase in the calculated bearing
capacity.
Assuming linear elastic behavior of sand prior to failure,
the allowable bearing capacity for crawler cranes can be expressed in the same way as Meyerhofs equation with footing sitting on ground surface

qa 1:5
13


N
F

B  1:2 m

12
B

0:3
A  N
qa @
B
F

B > 1:2 m

where [d] is the allowable settlement (mm); and F is the

factor to be determined. For foundations, F = 3.125 from
Meyerhofs equation and F = 2.08 from Bowles equation.
The only difference between eq. [13] and eq. [12] is that
eq. [13] explicitly applies the allowable settlement. By using
an allowable settlement of 25 mm, it turns into eq. [12] with
F = 3.125.
To find adequate values of the factor F, two sets of com#

2008 NRC Canada

Liu et al.

1293

Fig. 14. Correlations between equivalent footing width and Em/Es for sands. Reg, regression.

puter models were constructed to generate the pressure versus settlement profile: one for cranes sitting on mats and the
other for cranes sitting directly on the ground. A typical
crane track width of B = 1.5 m was used in both models as
it was found to have little influence on the determination of
factor F. For cranes with mats, a standard mat length of 6 m
was assumed. The soil used in the model was assumed to be
a MohrCoulomb material with linear elastic behaviour
prior to failure.
Soil properties from 41 cases discussed in Burland and
Burbridges (1985) paper were used in the two computer
models. In each case, information on soil type, SPT blow
counts N, footing size, load, and the observed settlement
was provided. The elastic modulus of each soil was back
calculated using a computer model to match the calculated
settlement and the observed settlement.
Figure 15 illustrates the variation of factor F with SPT
blow count for cranes without mats on different types of
sandy soils. The factor F generally remains nearly constant
with the change of SPT N values. Two lines representing
the factor F used in Meyerhofs equation and Bowles equation are also shown in the plot. The F = 2.08 used in
Bowles equation seems to be more reasonable for cranes
sitting directly on the ground.
For cranes sitting on mats, Fig. 16 also shows that the
SPT N values have no effect on the factor F. A value of
F = 4.3 might be appropriate to estimate the allowable
bearing capacity for cranes with mats, which is much
larger than that for cranes without mats. This could be attributed to the flexibility of the mats and high load intensity right beneath the crane track.
Since the allowable settlement of the crane is related to
the crane track length L, eq. [13] can be rewritten as follows
using different values of factor F:
For cranes sitting on the ground without mats

14

qa 3LN
B  1:2 m
0
12
B 0:3A
LN
qa 2@
B

B > 1:2 m

For cranes with mats



B0 0:3 2
LN
15
qa 1:2
B0
Another aspect to be considered is the strength of the soil.
This can be significant if the footing width is small and is
most applicable for cases when the crane is sitting on the
ground without a mat. To prevent the crane from bearing capacity failure and limit the soil behaving within the linear
elastic zone, a FS of 2.0 may be proper for cranes. Using
Meyerhofs ultimate bearing capacity equation, the allowable bearing capacity of sand based on strength consideration is
16

qa 

qu 32:8BN

 16BN
FS
2

However, this equation only applies to cases without

mats. For cranes with mats, because the footing width is relatively large, settlement is always the controlling factor.
Because the submergence of cohesionless soils will decrease the unit weight by half, the ultimate bearing capacity
should be reduced by up to one-half for ground water tables
higher than the base of the footing. If the ground water table
is at an intermediate position, the ultimate bearing capacity
for cranes without mats can be interpolated linearly.
The allowable bearing capacity of granular soils can also
be estimated from CPT test results because an empirical correlation between the CPT cone penetration resistance qc and
#

2008 NRC Canada

1294

Can. Geotech. J. Vol. 45, 2008

Fig. 15. Factor F for cranes without mats.

Fig. 16. Factor F for cranes with mats.

SPT blow count N exists. It is convenient to convert the

CPT cone resistance to the equivalent SPT blow count N
and then perform the allowable bearing capacity evaluation
using the previously mentioned equations.
Soft and medium clay
The allowable bearing capacity of saturated clay for
cranes on a flat ground surface based on strength consideration can be expressed as
17

qa

qu cNc sc 5:14 B=L

Cu
FS
FS
FS

where Cu is the undrained shear strength.

Because only the immediate settlement is concerned in
evaluating the bearing capacity for crawler cranes, the allowable bearing capacity based on settlement consideration is

18

qa

Eu

1   2 BIw

where [d] is the allowable settlement of the crane, and Eu is

the undrained modulus of saturated clay.
The undrained modulus of saturated clay is usually related
to its undrained shear strength as l = Eu/Cu in practice. The
value of l varied from 40 to 3000 for normally or lightly
overconsolidated clays, as summarized by Simons (1974).
For heavily overconsolidated clays, Bulter (1974) reported
this ratio to be between 150 and 830. Both authors indicated
that the ratio l is highly dependent on testing methods. Values between 300 and 500 are usually recommended for design purposes. Theoretically, the Poissons ratio for saturated
clay is n = 0.5.
Because the strength of soft to medium clay is very low,
only situations for cranes with mats are considered here. Re#

2008 NRC Canada

Liu et al.

1295

calling eq. [4] and rewriting eq. [18] in terms of l and crane
track length L yields
19

qa

 L
Cu
150Iw B0

It can be found that the allowable bearing capacity for saturated clay from both the strength consideration and settlement consideration is a function of its undrained shear
strength and the L/B ratio of the footing. Figure 17 shows
the variation of the allowable bearing capacity with the L/B
ratio. The L/B ratio for cranes with mats is about 1.2~2.0. It
can be seen from the figure that within this range the allowable bearing capacity based on settlement criterion is generally greater than that based on strength criteria with a FS of
2.0 except for l < 300 and L/B < 1.5. As a result, eq. [17]
with a FS of 2.0 may be adequate to estimate the allowable
bearing capacity of saturated clay for crawler cranes unless
evidence shows a small Eu/Cu ratio for some problematic
soils with high compressibility.
The ratio of immediate settlement to total settlement of
normally consolidated clay is only about 0.2. The allowable
settlement is usually taken as 25 mm for the design of shallow foundations. This means that the allowable immediate
settlement for shallow foundations is usually less than
10 mm. It is far less than the allowable settlement for
crawler cranes. This again indicates that the allowable bearing capacity for cranes should be higher than that for buildings.
It has been pointed out by Davis and Poulos (1968) that
for normally consolidated clay yielding and deviation from
linear behaviour will first occur when the FS against a bearing capacity failure is between 4 and 8; for slightly overconsolidated clays, the corresponding FS at first yielding is
2 to 3. Therefore, it is not recommended to use a FS less
than 2.0 for soft and medium clay.
Plots of pressuresettlement curves from computer simulations are shown in Fig. 18 for each specific Cu. A straight
line representing the allowable bearing capacity using a FS
of 2.0 is also included in each plot. It can be seen from the
plots that the computer simulation results match the theoretical analyses very well. The typical allowable settlement for
crawler cranes is about 3550 mm. Within this range, the allowable bearing capacity from settlement criterion is generally higher than that from strength criterion using a FS of
2.0.
Stiff clay
A similar analysis for soft and medium clay is carried out
here. Both cases, for cranes with mats and without mats,
were studied. For cranes without mats, eq. [19] can be rewritten by using a different allowable settlement formula as
20

qa

 L
Cu
180Iw B

Figures 19 and 20 show the allowable bearing capacity of

stiff clay from both strength and settlement aspects for
cranes with mats and without mats, respectively. Similar to
soft to medium clay, the allowable bearing capacity for
cranes with mats is generally controlled by the strength of
soil with a FS of 2.0. The plot of allowable bearing capacity

for cranes without mats shows that the allowable bearing capacity from settlement is much greater than that from
strength aspect with a FS of 2.0. As a result, a FS of 2.0 is
adequate to estimate the allowable bearing capacity of stiff
clay for crawler cranes regardless of whether the crane is
sitting on the mats or not.
Figures 21 and 22 from computer simulation show a series of plots of pressuresettlement curves for each specific
undrained shear strength value for cranes with mats and
without mats, respectively. Again, these plots testify the results from theoretical analysis.
For cranes without mats, the allowable bearing capacity
based on settlement considerations is much higher than that
based on strength with a FS of 2.0 for saturated stiff clay.
The use of FS = 2.0 only represents a 10 mm settlement as
shown in Fig. 22.

Case studies
Besides theoretical analyses and computer simulations, six
lifts on three different sites have been studied. They are:
(i) a test lift using a 400 t DEMAG CC2000 in Brighton
Beach, Ontario, (ii) a replace and reinstall vessel G using
a 600 t DEMAG CC2800 in Fort Nelson, British Columbia,
and (iii) lifts of Fractionator Burner and Reactor using a
1250 t DEMAG CC8800 in Ft. McMurray, Alberta. Case
studies were used to calibrate the soil parameters used in
the computer simulations and to test the proposed equations
for allowable bearing capacity for the crawler crane.
Field observation
The objective of field observation is to obtain the ground
settlement and the corresponding track pressure. Knowing
these two sets of data, computer models can be built to
back analyze the soil parameters and predict the soil bearing
capacity for cranes.
A traditional surveying technique using levels to shoot
scaled targets was used in measuring settlements of the
cranes since it is the most practical method given the available time and resources.
Two levels are usually set on each side of the crane to
measure the vertical displacement of each track during a lifting operation. They were calibrated and can reach an accuracy of about 0.5 mm in general conditions. The preferred
locations for levels should be close to the targets and out of
the settlement influence zone of the crane. Ten to thirty
metres from the targets is the most desirable distance. A
benchmark or reference point is also used to check the
movement of the level and eliminate the influence of it.
Measuring tapes mounted on the crane track were used as
the leveling targets. Although the settlement of the ground is
of the most concern, the vertical displacement of the crane
track is much easier to measure and may well represent the
ground settlement because the compressibility of the timber
mat and crane track is relatively small. Because the settlement profile along the crane track is close to a straight line,
only two end points on each track needed to be monitored.
These two end points were chosen to be the points right
above the first and last rollers contacted to the ground. Figure 23 illustrates the typical layout of levels and targets for
the settlement observation. The most direct and precise way
#

2008 NRC Canada

1296

Can. Geotech. J. Vol. 45, 2008

Fig. 17. Bearing capacity of soft clay from theoretical analyses.

Fig. 18. Load versus immediate settlement curve for soft clay.

to measure the track pressure of a crane is to bury several

pressure cells beneath the crane track. However, this is also
the most expensive way. An alternative is to calculate the
track pressure rather than measure it using the manufacturers software and knowing detailed crane load information. All the track pressure information used in the case
studies is based on the latter approach.
The crane configuration, load, and lift radius can be obtained from the lift plan. However, more precise information
from the computer mounted on the crane is usually available
and this was used in the evaluation of track pressure.
During a lift, the displacement of targets mounted on

crane tracks was monitored and recorded at intervals of

about 5 min. The detailed crane load information including
the load, radius, and boom orientation angle were also recorded simultaneously. This recorded information was then
used to calibrate the elastic properties of soil and estimate
the ground bearing capacity for crawler cranes.
Soil properties determination
The soil properties of each site were first interpreted from
the site investigation reports and then calibrated by comparing the computer simulations using the observed settlement
and track pressure.
#

2008 NRC Canada

Liu et al.

1297

Fig. 19. Bearing capacity of stiff clay for cranes with mats.

Fig. 20. Bearing capacity of stiff clay for cranes without mats.

This calibration procedure involved using two sets of

computer models with one perpendicular to the crane tracks
and the other along the crane tracks. Timber mats were included in the models to represent the real world situation
and avoid error caused by simplification. To account for the
rigid effect of the crane tracks, steel blocks with the same
dimensions as the tracks were placed onto the timber mats
in the models and the track pressure was applied on the steel
block rather than directly on the timber mats. The modulus
of the steel block was reduced properly to yield equivalent
rigidity of the crane track.
Although modeling along the tracks is not very precise in
simulating the real case, it is the only choice to evaluate the
differential settlement of each crane track. Some measures
are applied to limit the errors caused by the model to an acceptable level. These measures include: (i) spread track
pressure over a wider area; (ii) use the proper model size;
and (iii) compare the simulation results with those from sections perpendicular to the crane tracks.
The soil parameters used in the models are then adjusted
by trial and error to yield a similar amount of settlement to

that observed for each critical situation. A summary of the

soil properties from the site investigation report and calibrated soil parameters used in the analyses for the three
cases are presented in Tables 14.
It should be noted that the simulated settlement cannot
perfectly match the measured settlement. The source of errors can be attributed to:

The error in track pressure determination. The track pressure was calculated using the rigid assumption rather than
measured in the field.

The error caused by the computer model, mainly the result of using a 2-D program to model a 3-D problem.

Ignoring crane soil interaction.

Field observation errors.

The defects in site preparation (i.e., gaps between layers
of crane mats).

The assumptions in soil behavior and material parameters.
In general, the settlements from the computer simulation
are comparable to those from the field observations. Another
important result is that the settlement from models for sec#

2008 NRC Canada

1298

Can. Geotech. J. Vol. 45, 2008

Fig. 21. Load versus immediate settlement curve of stiff clay for cranes with mats.

Fig. 22. Load versus immediate settlement curve of stiff clay for cranes without mats.

2008 NRC Canada

Liu et al.

1299

Fig. 23. Typical level and target layout for settlement observation.

Table 1. Soil properties from site investigation Brighton

Beach, Ontario.

Soil type
Thickness (m)
Unit weight, g (kN/m3)
Relative density, Dr (%)
Water content, w (%)
Plastic limit, PL (%)
Liquid limit, LL (%)
SPT blow count, N
Triaxial Cu (kPa)
In situ vane Cu (kPa)
Remold vane Cu (kPa)

Sand and
gravel
fill
0.88

Silty
sand
1.55
18.5
85
11

24

Firm
silty
clay
1.84
17.7

24

65
32

Soft
silty
clay
>5.78
16.7

43
23
50
1
31
25
9

Table 2. Soil properties from site investigation Fort Nelson,

British Columbia.
Soil type
Thickness (m)
Unit weight, g (kN/m3)
Moisture content, w (%)
Plastic limit, PL (%)
Liquid limit, LL (%)
Initial void ratio, e0
Compressive index, Cc
Recompressive index, Cr
Unconfined compressive strength, qu
(kPa)
SPT blow count, N

Weathered
till
1.7
19.8
1523
1620
3157
0.66
0.037
0.013
240

Unweathered
till
>3.8
19.8
1417
1620
3157
0.59
0.054
0.014
420

16

28

2008 NRC Canada

1300

Can. Geotech. J. Vol. 45, 2008

Table 3. Soil properties from site investigation Ft. McMurray,

Alberta.
Soil type
Thickness (m)
Unit weight, g (kN/m3)
Relative density, Dr (%)
Moisture content, w (%)
Plastic limit, PL (%)
Liquid limit, LL (%)
SPT blow count, N
Cu (kPa)

Sand
fill
0.6
17.3
60
13

20

Native
sand
2.2
18
82
22

40

Clay
till
1.2
21

11
1316
2036
46
240

Fig. 25. Allowable bearing capacity determination for the Fort Nelson case.

Oil
sand

20.4

13

>50

Table 4. Calibrated soil parameters used in the analyses.

Soil

Youngs
modulus,
E (MPa)

Brighton Beach, Ontario

Sand and gravel
150
fill
Silty sand
37.5
Firm silty clay
19.5
Soft silty clay
5

Poissons
ratio, n

Friction
angle,
f (8)

Cohesion,
c (kPa)

0.15

48

0.3
0.49
0.49

36
0
0

0
65
25

0
0

120
210

0.15

48

0.3
0.3
0.49
0.3

36
36
0
>50

0
0
240

Fort Nelson, British Columbia

Weathered till
24
0.2
Unweathered till 63
0.2
Ft. McMurray, Alberta
Sand and gravel
150
fill
Sand fill
33
Native sand
55
Clay till
120
Oil sand
185

Fig. 24. Allowable bearing capacity determination for the Brighton

Beach case.

Fig. 26. Allowable bearing capacity determination for the Ft.

McMurray case.

tions parallel to crane tracks matched that from models for

the section perpendicular to crane tracks very well. This
proves that the use of equivalent footing width B and the
model size are valid.
Evaluation of soil bearing capacity using different
approaches
The soil bearing capacity can be estimated if the soil
properties and the dimensions of the footing are known. It
is of interest to compare the allowable bearing capacity calculated from the conventional method used for shallow
foundations, the proposed formulas, and the extrapolation
from the computer simulation.
Comparisons of the bearing capacity for the three cases
using these three methods are presented in Figs. 2426. The
ultimate bearing capacity and the FS associated the proposed
formulas are also illustrated in the figures. It is seen from
the figures that the proposed formulas yield results close to
those from the computer simulation for these three cases.
Summary of case studies
All of these cases illustrate that the use of a 2-D program
in modeling the 3-D bearing capacity problem can yield rea#

2008 NRC Canada

Liu et al.

1301

Fig. 27. Procedure to determine allowable bearing capacity for crawler cranes.

sonable results if an appropriate model size is used. Determining the bearing capacity for crawler cranes by using a
settlement criterion typically used for building foundations
is usually too conservative. Simply using a FS of 2 or 3 can
either overestimate or underestimate the soil bearing capacity for cranes. Using more than two layers of timber
mats is not recommended, as it may cause more open gaps
within the mats. Computer simulation of the fractionator lift
shows that the use of 9 m long mats at the bottom layer
does not contribute much to the bearing capacity.

Conclusions

Directly applying the traditional bearing capacity calculation used for building foundations to crawler cranes was
shown to be inappropriate and it may lead to errors.

The out of levelness of a crawler crane can be converted
to a maximum allowable settlement of the ground.

The nonuniformly distributed crane track pressure can be
represented by an equivalent pressure in determining the
bearing capacity for cranes with mats.
#

2008 NRC Canada

1302

The equivalent footing width for a crane with a mat is

mainly a function of crane track width, mat thickness,
and the relative rigidity of the mat and the soil.

Back analyses of three study cases provided good support
on the use of an equivalent pressure, equivalent footing
width, and model size for computer simulations. It is
also important to test the validity of the method proposed
in the estimation of bearing capacity for crawler cranes.

A typical design procedure for estimating the bearing capacity for crawler cranes is described in the following
steps:
1. Obtain information on the effective track length L and
track width B of the crane.
2. Obtain information on the maximum track pressure qmax
and the corresponding minimum track pressure qmin of
the same track.
3. Determine the soil allowable bearing pressure qa based
on different soil types. For clayey soil, eq. [17] using a
FS of 2.0 can be applied. For sandy soil, the allowable
bearing pressure can be estimated using eq. [14] accompanied with a strength check.
4. Compare the maximum crane track pressure qmax with
the allowable bearing pressure qa of the soil. If the allowable bearing pressure is less than the maximum crane
track pressure, a crane mat is required.
5. Calculate equivalent track pressure q~ using eq. [9].
6. Calculate equivalent footing width B using eq. [11]. It
should be noted that the calculated equivalent footing
width should be limited by the strength of the mat
against bending and shear failure.
7. Evaluate soil allowable bearing pressure under the mats
using eq. [15] for sandy soil. For clayey soil, eq. [17] is
valid with a FS of 2.0, and the footing width should be
replaced by the calculated equivalent footing width.
8. Compare the allowable soil bearing pressure qa with the
equivalent track pressure q~ . If the allowable bearing
pressure is less than the equivalent track pressure, adjust
the mat size and configuration and repeat steps 6 to 8 until the allowable soil pressure is greater than the equivalent track pressure.
The procedure described above is illustrated with a flow
chart in Fig. 27.
These conclusions may not be applicable for problematic
soils such as sensitive clay, loose sand, loess, and organic
soils. Moisture content change of topsoil as a result of precipitation or other reasons must also be considered in evaluating the bearing capacity for cranes.

Acknowledgements

Can. Geotech. J. Vol. 45, 2008

Sciences and Engineering Council of Canada (NSERC).

This research was supported by Sterling Crane of Edmonton,
Alberta.

References
Becker, R. 2001. The great book of mobile cranes. Vol. 1. Handbook of mobile and crawler crane technology. KM Verlags
GmbH, Griesheim, Germany.
Bowles, J.E. 1982. Foundation analysis and design. 3rd ed.
McGraw-Hill, New York.
Bulter, F.G. 1974. Heavily over-consolidated clays. In Proceedings
of the conference on settlement of structures, Cambridge, April
1974, John Wiley & Sons, New York. pp. 531578.
Burland, J.B., and Burbridge, M.C. 1985. Settlement of foundations
on sand and gravel. In Proceedings of the Institution of Civil
Engineers, London, 78: 13251381.
CIRIA. 1996. Crane stability on site. CIRIA special publication
131, Construction Industry Research and Information Association, London, UK.
Davis, E.H., and Poulos, H.G. 1968. The use of elastic theory for
settlement prediction under three-dimensional conditions. Geotechnique, 18: 6791.
Gupta, R.C. 2002. Estimating bearing capacity factors and cone tip
resistance. Soils and Foundations, 42: 117128.
ISO. 1991. Mobile cranes determination of stability.
ISO4305:1991(E), International Organization for Standardization, Geneva, Switzerland.
Lee, J., and Salgado, R. 2005. Estimation of bearing capacity of
circular footings on sands based on cone penetration test. Journal of Geotechnical and Geoenvironmental Engineering, 131:
442452. doi:10.1061/(ASCE)1090-0241(2005)131:4(442).
Liu, X. 2005. Soil bearing capacity for crawler cranes. M.Sc. thesis, Department of Civil and Environmental Engineering, University of Alberta, Edmonton, Alta.
Meyerhof, G.G. 1956. Penetration tests and bearing capacity of cohesionless soils. Journal of the Soil Mechanics and Foundations
Division, 82(SM1): 119.
Shapiro, H.I., Shapiro, J.P., and Shapiro, L.K. 1999. Cranes and
derricks. 3rd ed. McGraw-Hill, New York.
Simons, N.E. 1974. Normally consolidated and lightly overconsolidated cohesive materials. In Proceedings of the conference on settlement of structures, Cambridge, April 1974, John
Willey & Sons, New York. pp. 500530.
Stuart, J.G. 1962. Interference between foundations, with special
reference to surface footings in sand. Geotechnique, 12: 1522.
Terzaghi, K. 1943. Theoretical soil mechanics. John Wiley & Sons,
New York.
Vesic, A.S. 1973. Analysis of ultimate loads of shallow foundations. Journal of the Soil Mechanics and Foundations Division,
99(SM1): 4573.

Funding for this research was provided by the Natural

2008 NRC Canada