ENCN353: Geotechnical Engineering

Module 3: Shallow foundations

1 Introduction
1.1 What are shallow foundations?
Rafts, pads, footings, mats… these are all names for types of shallow
foundation. Essentially, shallow foundations lie on and within the first
few metres of soil. They tend to be wide (or long) rather than deep.

By the end of this module you should be able to:

1. Outline how we typically design for SLS and ULS for shallow
foundations (and geo-structures in general),

2. Explain the conditions that must be satisfied for an upper bound

analysis, for a lower bound analysis, and how the upper and
lower bound analyses are used to estimate the collapse load,

3. Develop equations for and calculate shallow foundation

capacities for undrained conditions from ‘first principles’ using
limit analysis,

4. Define the bearing capacity factors Nc and Nq,

5. Explain the function of each of the various bearing capacity

enhancement factors, and use them in calculation of footing
capacity.

1.2 Examples
Some examples of simple shallow foundations (not an exhaustive list!)
are given below

1.2.1 Strip footings

A strip footing in one that is long compared to its width. They are
typically used to support masonry walls. Modern strip footings are
constructed by excavating a trench, placing a reinforcement cage and
pouring in concrete.

ETB/JJMH Page 1
ENCN353:: Geotechnical Engineering

1.2.2 Pad footings

A pad footing is used to support isolated loads, such as from columns
or from support of raised slabs or flooring. They are often
recommended forms of foundation for sandy sites or those that may be
subjected to wind or water erosion.
erosion. They are generally constructed at
least 300mm below the ground surface to prevent undermining. In cold
climates they are placed below the frost penetration depth to prevent
the effects of frost heave. They are constructed in a similar method to
strip footings.

ETB/JJMH Page 2
ENCN353:: Geotechnical Engineering

1.2.3 Raft or mat foundations

Raft or mat foundations can be used in cases where pad footings
would need to be very large to support the loading, or in areas of
swelling soils (e.g. clays which expand and contract with moisture
content and hence, with the seasons). They are often reinforced
concrete slabs – and for tall buildings, can be several metres in
thickness. At a smaller scale, rafts may be stiffened at the loading
point by making the slab locally thicker. Ribs or beam-and-slab
beam
construction are used to make the raft slab stiffer.

ETB/JJMH Page 3
ENCN353:: Geotechnical Engineering

2 Deformation, failure & limit state design

Civil engineers must design foundations so that they are both safe in
terms of failure and serviceable in the sense that they will not deform
excessively under working loads
loads resulting in (for example) undesirable
cracking or impaired durability.
durability. These two states are sometimes
referred to as the Ultimate Limit State and Serviceability Limit
State.. Design for both these criteria is called Limit State Design.
Design
This course begins with the Failure of foundations
f ations (ULS) followed by
the Settlement of foundations (SLS).
Whereas, ULS requirements are generally fairly obvious (we would like
our geotechnical structures to remain standing!), SLS requirements
refer to the avoidance of excessive
excessive deformation. “Excessive”
deformation is that which tends to lead to damage to structures, which
are often relatively stiff and brittle. The definition of excessive will
therefore depend on the nature of the structure and load transfer. For
example, very ry little ground deformation may be acceptable in the
design of a concrete bridge foundation, whereas for a reinforced soil
foundation, much greater deformation would be required to impair its
structural integrity.

ETB/JJMH Page 4
ENCN353: Geotechnical Engineering

3 SLS & loading distributions beneath foundations

3.1 Worst case scenario?
When we consider deformation (SLS – Serviceability Limit State),
need to think about the average value and the variation in “settlement”
across the loaded area or structure.
Damage to a structure is more likely to occur as a result of differential
settlement across the structure than the ultimate uniform settlement of
it.

This means that, for example, not only do we need to know how far a
footing might settle overall but also how much the middle might settle
compared with the edge.
How might these differences arise? Possibilities include:
1.

2.

3.

4.

5.

The first four points require a careful consideration of site conditions

before considering the foundation itself. The last point requires an
understanding of how the structure will behave relative to the soil.

ETB/JJMH Page 5
ENCN353:: Geotechnical Engineering

3.2 Rigid vs Flexible foundations

It has been shown that rigid and flexible foundations result in different
stress distributions beneath them, i.e.:

Tomlinson (2004)
Rigid foundations can be thought of as settling uniformly
u but resulting
in a non-uniform
uniform stress distribution beneath the foundation.
foundation Mostly in
this course we will examine
xamine flexible foundation loading and response.
response

ETB/JJMH Page 6
ENCN353:: Geotechnical Engineering

4 How do we design & analyse for ULS and SLS?

SLS
1 For ULS design (i.e. against failure), the mostmost straightforward
method is to assume the soil is at the point of failure at every point
in the affected zone at the failure load.

For SLS design, onene (somewhat crude) method is simply to divide

the failure load found above by some “factor of safety” (FOS) to
obtain the working load. A typical FOS of 3 is sometimes used.

• Advantage: the he stress-strain

stress strain behaviour of the soil is not required
in detail for either ULS or SLS calculation.
calculation

• Disadvantage: previous knowledge and experience of the soil

and the particular
ticular circumstance are required. May lead to over-
over
design ( $$ ) or under-design
under ( ).

2 Another method commonly used for SLS is to use the stress-strain

response of the soil at working loads to determine the expected
elastic deformations and hence,
he settlements directly.

• Advantage: elastic methods of analysis can give a reasonable

approximation
proximation to soil deformation

• Disadvantage: full stress-strain

strain relationships for soil are rather
complicated – they are certainly not actually elastic,
elastic so a good
understanding
tanding of soil theoretical behaviour is required.
required
3 A third method makes use of the full stress-strain
train response of the
soil (often incorporating
rporating the stiffness of the structure), but with
more complicated constitutive soil models in place of elasticity. This
Thi
type of analysis is typically carried out as part of the discipline of
geotechnical numerical modelling.
modelling

• Advantage: a fully modelled soil-structure

soil structure interactive response

• Disadvantage: extensive knowledge

knowledge of soil parameters is
required and the analysis effort
effort can be relatively time-
time
consuming and costly.

ETB/JJMH Page 7
ENCN353: Geotechnical Engineering

This course:
Examine failure loads in soil for shallow and deep foundations (as in
#1 above for ULS).

Examine stress-strain response, elastic deformations and hence

settlement behaviour beneath shallow foundations and point loads (as
in #2 above for SLS).

ENCI453 (optional) course next year:

Examine stress-strain response, elastic deformations and hence
settlement behaviour for piled foundations.

Design shallow and deep foundations for both ULS and SLS.

ETB/JJMH Page 8
ENCN353: Geotechnical Engineering

4 ULS – Failure in soils

We have two possible failure criteria that we could apply to a soil that
is failing:
• Mohr-Coulomb τf = c’ + σ’ tan φ’ for drained conditions
• Tresca τf = cU for undrained conditions

For each of these criteria what are the principal stresses at failure?

For Mohr-Coulomb failure criterion:

We can relate the maximum and minimum principal stresses using

trigonometric relations:

Where Kp is called the “passive earth pressure coefficient “

Where Ka is called the “active earth pressure coefficient”

ETB/JJMH Page 9
ENCN353: Geotechnical Engineering

(Note on K0)
The “At rest earth pressure coefficient” K0, simply relates the in situ
horizontal effective soil stress to that of the vertical soil stress:

K0 takes a range of values depending on the formation and stress

history of the soil. It cannot be less than Ka and cannot be more than
Kp .

K0 = (freshly deposited loose sand & normally-

consolidated clay)

K0 = (overconsolidated / pre-compressed soils)

Empirically:

For NC soils:
(Jaky, 1944)

For OC soils:
(Mayne & Kulwhawy, 1982)

Where OCR is the overconsolidation ratio:

For Tresca (undrained) failure criterion:

ETB/JJMH Page 10
ENCN353: Geotechnical Engineering

5 Plasticity in soil mechanics

5.1 What is plasticity? Why do we use it?
When a shear stress is applied to soil, some small movement may
take place which we have previously modelled as elastic. If we
increase this stress further, we may find that progressively greater
movement takes place and the soil appears to offer less resistance
than before. This process is sometimes called “yield” and it is similar to
that seen in structural materials. If we remove the stress we find that
there is some permanent deformation – this is “plastic” behaviour (as
opposed to elastic – recoverable – behaviour). After some large
deformations we may find either that the soil body is flowing or that it
has fractured – this behaviour will depend on the stresses applied to
the soil and how dense it is.

Failure in metals (Case, Chilver & Ross, 1998)

Failure in timber (Case, Chilver & Ross, 1998)

ETB/JJMH Page 11
ENCN353: Geotechnical Engineering

Failure in undrained triaxial compression on OC clay (La Rochelle, 1988)

Failure in drained triaxial compression on sand (Desrues, 2004)

ETB/JJMH Page 12
ENCN353: Geotechnical Engineering

The failure behaviour of soil post-yield may be described in terms of

plasticity theory – i.e. the material is assumed to behave in an
irrecoverable plastic manner. This contrasts with SLS design, where
we will use elasticity theory to look at settlements and small
deformations. Here, we’re not so interested in the deformations
(because they’re already too great to be acceptable) but instead we
are interested in the load or stresses to cause outright failure.
Recalling that soil has a complex response to shear strain:

We have to make some simplifying assumptions. One of the greatest

assumptions is that the soil behaves like this:

I.e. we assume there is no strain “softening” (i.e. brittle, dense or

overconsolidated, OC, behaviour) or “hardening” (ductile, loose,
normally consolidated, NC, behaviour). We can see that by using
critical state strengths (φ’crit) however, we will get the best
approximation to the final state– hence, peak strengths (φ’peak) which
are followed by strain-softening should only be used with great
caution!

ETB/JJMH Page 13
ENCN353: Geotechnical Engineering

6 Upper and lower bounds – introduction

6.1 Comparison of elasticity and plasticity based solutions
We can solve engineering mechanics problems by use of either
elasticity-based or plasticity-based methods. For elasticity methods
we need to know about the constitutive (stress-strain) relationship for a
soil, which can depend on geology and stress-history. To complicate
matters, for many foundation problems different parts of the soil will be
experiencing different stresses, so they will be on different stress-
strain paths.

Hence, the advantages of using plasticity methods in soil mechanics

are that:
•

The main disadvantage is that

Rules that need to be satisfied

In the case of elasticity-based methods, we use:

a)

b)

c)

ETB/JJMH Page 14
ENCN353: Geotechnical Engineering

Analogously, in the case of plasticity-based solutions we use:

1) A lower bound solution which satisfies:

a)

b)

2) An upper bound solution involving:

c)

6.2 Definitions used in plasticity

6.2.1 The lower bound “safe” theorem

The lower bound theorem states that: “If a system of stresses with the
soil mass can be found which is in equilibrium with the external loads
and body forces (e.g. self weight), and which nowhere violates the
failure criterion for the soil, then those loads and forces represent a
lower bound to those which will actually cause collapse”.

6.2.2 The upper bound “unsafe” theorem

The upper bound theorem states that: “If a mechanism can be found
such that the work done by the external loads and body forces is equal
to the energy dissipated within the soil mass as it deforms, then those
loads and forces represent an upper bound to those which will cause
collapse”.

ETB/JJMH Page 15
ENCN353: Geotechnical Engineering

These theorems strictly apply to (elastic) perfectly plastic materials

where upon yield, neither strain-softening nor strain-hardening take
place. (Strictly speaking, the material must also behave such that it
strains at a constant rate with an associated flow rule1. More about this
is 3rd Pro and beyond!).

6.2.3 Finding a solution using lower and upper bounds

Upper and lower bound solutions which do not coincide will form a
bracket to those which may occur. If we can find upper and lower
bound solutions that give the same answer, we have the exact failure
solution (i.e. load to cause collapse). This is very rare. However, we
can get quite close even if we do not find the perfect analytical
solution…

That is:

1
Associated flow is where the vector of plastic strain increment is normal to the yield surface – a
condition also known as “normality”.

ETB/JJMH Page 16
ENCN353: Geotechnical Engineering

7 Shallow strip footings – Analysis using upper and

lower bounds
To start with, we will look at the 2-dimensional case of an infinitely long
strip footing. We can later add in some factors to account for 3-
dimensional behaviour.
Thhe figure below on the left shows the cross-section through a typical
shallow strip foundation. In order to analyse the problem we can
idealise the footing and surrounding soil as shown on the right.

The foundation load is modelled as an applied stress of σf, acting on

the founding plane beneath the footing. The adjacent soil is modelled
as a surcharge of σ0 = γD where D is depth of burial of the footing and
γ is the soil unit weight.
Water pressures may also act on the footing, with pore pressure u.
Hence, in terms of effective stress, the effective surcharge due to the
soil adjacent to the foundation σ’0 will be σ’0 = γD-u.
We can analyse this foundation in terms of either effective stress
(Mohr-Coulomb, drained, long-term behaviour for all soils) or total
stress (Tresca, undrained, short-term behaviour – most relevant for
short term behaviour of clays and fine silts).

ETB/JJMH Page 17
ENCN353: Geotechnical Engineering

7.1 Total stress analyses – undrained (short term)

We will start with the undrained analysis of foundations using total
stress analysis, simply because it is easier to visualise and construct.

7.1.1 Lower bound total stress, undrained analysis – Attempt 1

The most simple total stress analysis for the short-term collapse load
of a strip footing consists of an idealized stress field with active and
passive zones, separated by a discontinuity or cut / wound in the soil
• Undrained shear strength of the soil is cU
• The “discontinuity” or wound is assumed to be strength-less and
frictionless (i.e. does not resist shearing).

Zone A
It is assumed that the soil directly beneath the foundation is in a state
of active failure – i.e. the vertical stress σvA is the major principal equal
to σf, while the horizontal stress σhA is the minor principal stress equal
to (σf -2cu).

ETB/JJMH Page 18
ENCN353: Geotechnical Engineering

Zone B
The soil directly next to the foundation is assumed to be in a state of
passive failure – i.e. the vertical stress σvB is minor and is equal to σ0,
while the horizontal stress is major and is equal to (σ0+2cu).

Equilibrium across discontinuity

In order to have equilibrium across the strengthless wound, the
horizontal stresses must be equal, σhA=σhB.
Hence we can determine the ratio of surface stresses (on the founding
plane) at failure:

This stress field with only a single discontinuity on each side of the
foundation has a principal stress rotation of 90º across the
discontinuity (180º on the Mohr circle).
The value of NC=4 is not the optimal lower bound solution. That is, the
stress field shown in the above figure may be easily improved by
involving a more gradual rotation of stresses through larger numbers
of discontinuities and by removing the imposition that the wound has
no strength.

7.1.2 Lower bound Tresca, undrained analysis – Better attempts

To improve our lower bound estimate we may allow some soil shear
strength along the discontinuities. However, we still need to achieve a
rotation of 90° overall from the active state beneath the foundation, to
the passive state next to the foundation. This suggests that Mohr’s
circles representing the discontinuities must overlap (to allow for shear
strength), so we will also need more discontinuities.

ETB/JJMH Page 19
ENCN353: Geotechnical Engineering

Mohr’s circles and discontinuities - review

Take a general stress field with a discontinuity. On each side of the
discontinuity, we have an element of soil represented as:

Direct stresses normal to the discontinuity in zones A and B are

denoted σNA and σNB, while direct stresses tangential to the
discontinuity are denoted σTA and σTB. Shear stresses (which are
“complementary” – i.e. the same on both faces) are denoted τA and τB,
respectively.
Note that, across the discontinuity:
σNA = σNB
τA = τB but: σTA <> σTB

If we look at the same stress field in terms of principal stresses, we

may get this:

ETB/JJMH Page 20
ENCN353: Geotechnical Engineering

On the right we are using the longer arrows to denote the major
principal stress, σ1, and shorter arrows to denote the minor principal
stress, σ3 (I will subsequently omit the minor principal stress in later
diagrams and represent the major principal stress as a single arrow).
If we now draw the Mohr circles for the stresses shown, we can
determine a number of important relationships between the rotation of
principal stress and the discontinuity.
This is done below in a step-wise manner:
1. As stated previously, we find there is a common stress state,
(σ,τ) across the discontinuity, which can exist. This can be
shown as a common point on two intersecting Mohr’s circles.
2. Next we define the “Origin of Planes” (also known as the “Pole
for planes”) in zone A, OPA and that in zone B, OPB on the Mohr
circles. These points may be found by projecting a line from the
common stress point in the direction of the plane.

3. Regarding our discontinuity in real space:

a. We define θ as the angle of stress rotation across the
discontinuity.
b. We happen to find that it is easy to find this on the Mohr
circle!

ETB/JJMH Page 21
ENCN353: Geotechnical Engineering

ETB/JJMH Page 22
ENCN353: Geotechnical Engineering

7.1.3 Lower bound Tresca, undrained analysis – Attempt 2

Returning to the problem of finding a better (higher) lower bound, we
could try using 30º discontinuities, three on each side of the foundation
(we need to rotate by 90º, hence we need 90º /30º = 3 discontinuities
each side).

1. We know σ0, so we start by plotting the circle (radius cU) for zone
A on the Mohr plot. Given that σ0 acts on a horizontal plane, we
can draw a line from σ0 to find the origin of planes OPA for zone A
on the Mohr plot.

2. Recalling that we can plot the θ=30º stress rotation on the Mohr
circle for zone A (i.e. the point of the previous section), this will
give us the common stress point, N, for zones A and B. We can
then draw the Mohr circle for zone B (radius cU).

3. We can draw a line through OPA and N to give us the angle of the
discontinuity between zones A and B, and hence, the origin of
planes for zone B, OPB.

4. Similarly to step 2, we can plot the θ=30º stress rotation on the

Mohr circle for zone B to give us a common stress point, M, for
zones B and C. We can then draw the Mohr circle for zone C
(radius cU).

ETB/JJMH Page 23
ENCN353: Geotechnical Engineering

5. We draw a line through OPB and M to give us the angle of the

discontinuity between zones B and C, and hence, the origin of
planes for zone C, OPC. In this case we find it is the same point
as OPB.

6. We continue the process up to zone D…

7. A completely valid lower bound solution requires consideration of
stress equilibrium beyond the immediate failure area. If two
discontinuities from both corners of a loaded surface are
extended to infinity then where they cross they may form other
zones, such as that of Zone E. We can check the stresses in
zone E to ensure that the stress equilibrium and failure criterion
are not violated, by projecting a line from OPC parallel to the
discontinuity separating Zones C and E, to give the interaction of
the E circle with the C circle. This shows that the stress circle for
Zone E is the same as that for Zone B.

Hence, for this case we find that:

We could go on improving the solution by making the discontinuities

ever smaller. This is fully covered in 3rd Pro, ENCI452.
For now we will leave it at this value and move on to the upper bound
solution.

7.1.4 Upper bound Tresca, undrained analysis – Attempt 1

We can try to find an upper bound for the same problem as given in
section 4.1.1 by assuming a mechanism of collapse.
Kinematically admissible mechanisms for plastic materials (where
τmax=cu) are made up of straight lines or circles. We consider one of
the most simple mechanisms as consisting of a semi-circular slip, as
shown below:

ETB/JJMH Page 24
ENCN353: Geotechnical Engineering

Consider the rate of work done by the external forces and energy
dissipated along the slip, as it rotates about the origin O with angular
velocity ω.
Recall that work is: Force × Distance

Hence, work rate (power) is: Force × Velocity

Rate of work done by the foundation load, σf along AO, as it falls with
an average velocity of ωB/2:

Rate of work done against the surface load σ0 along OC, as it rises
with an average velocity of ωB/2:

Hence net rate of work by the external loads is:

Rate of energy dissipated along the slip surface:

Equating external work done and internal work dissipated:

ETB/JJMH Page 25
ENCN353: Geotechnical Engineering

So

Where, Nc is (again) the bearing capacity factor.

The value of NC=6.28 is not the optimal upper bound solution. That is,
the mechanism may be easily improved by searching for a more
critical slip circle or combination of wedges and circular parts.
We will try this next…

7.1.5 Upper bound Tresca, undrained analysis – Attempt 2

We can try to find a better upper bound for the same problem by
looking at 5 slide wedges beneath the footing. For example:

(Note that, for clarity here, we will dispense with the load due to the soil weight σ0,
and simply take σf as our applied foundation load – we can substitute (σf -σ0) for σf
at the end to include the effect of soil surcharge above the founding plane, if we
wish.)

Once we have decided upon the mechanism, we can draw a velocity

diagram or “hodograph” to solve the problem. Note that O denotes the
area outside the mechanism which is not moving – hence the lines o-
b, o-c, o-a, etc. show absolute movement, while the lines b-a, e-d etc.
show relative movement between those respective wedges.

ETB/JJMH Page 26
ENCN353: Geotechnical Engineering

From the velocity-diagram or “hodograph”, we can determine the

velocities of the sliding wedges relative to the downward velocity of the
foundation (which we will denote δv). Therefore, we can write a work
equation as before, equating external work with work dissipated along
the sliding planes.

External rate of work:

Work dissipated:

Hence, equating work:

With NC=5.77, this is better (i.e. lower) than our previous upper bound
attempt, but is still not the best solution…

ETB/JJMH Page 27
ENCN353: Geotechnical Engineering

7.1.6 Strip footing with vertical load on uniform, undrained soil

(Tresca) – Exact solution
We will not analyse in detail the exact solution to the previous set of
examples (there are many others). However, it is worth knowing that
an exact solution to this case does exist. It was first published by
Ludwig Prandtl in 1921, and is therefore sometimes called the “Prandtl
solution”
For the upper bound, the Prandtl solution consists of active and
passive zones, below and adjacent to the foundation, with a fan of
infinitesimally small discontinuities in between, which rotate in total
through 90º.
For the lower bound, the solution consists of three 45º sliding wedges,
with a 90º slip fan in between. Work is dissipated between the wedges
and within the fan.
The Prandtl solution looks like this:

Yielding an equal upper and lower bound solution of:

ETB/JJMH Page 28
ENCN353: Geotechnical Engineering

7.2 Mohr-Coulomb – drained analyses

We will go over the drained analysis of foundations using Mohr-
Coulomb in less detail here, simply because it is rather more
geometrically complicated (it is covered more in the 3rd Pro
Geomechanics course).

7.2.1 Lower bound Mohr-Coulomb – Attempt 1

The most simple effective stress analysis for the collapse load of a
strip footing consists of an idealized stress field with active and
passive zones separated by a frictionless wound or “discontinuity”.
This is shown below.

The Mohr circles for the active and passive zones are given below.

ETB/JJMH Page 29
ENCN353: Geotechnical Engineering

Zone 1
Under the foundation directly below the founding plane the vertical
effective stress σ’v1 is σ’f. At failure, this is an active zone (with σ’f the
major principal effective stress). So the horizontal effective stress σ’h1
is equal to Kaσ’f.

Zone 2
The vertical effective stress, σ’v2 at the founding level in the soil
adjacent to the founding plane is σ’0. At failure, this is a passive zone
(with σ’0 the minor principal effective stress). So, the horizontal
effective stress, σ’h2 is equal to Kpσ’0.

Equilibrium across the discontinuity

In order to have equilibrium across the frictionless wound, the
horizontal stresses must be the same for each zone (σ’h1=σ’h2).
Therefore:

Given that

We can determine the ratio of surfaces stresses (on the founding

plane) at failure:

Where, Nq is also called a bearing capacity factor. (But note it has a

different subscript from the bearing capacity factor used for undrained
analysis and is defined as the ratio between σ’f and σ’0 rather than the
difference!)

ETB/JJMH Page 30
ENCN353:: Geotechnical Engineering

Now, this result is very conservative,, as a result of introducing a single

frictionless discontinuity on either side,
side, and can be improved upon. We
can improve by allowing more gradual rotation of principal stresses
with more discontinuities and by allowing friction to act on the
discontinuities. This means the Mohr circles will overlap. In other
words, the method to determine better lower bound solutions is similar
to the total stress analysis case for undrained soil.
We will not go into details on improvements here. Suffice to say that
the least
east conservative lower bound turns out to be:
be

7.2.2 Upper bound Mohr-Coulomb

Mohr
If we assume the condition of normality (associated flow) applies to a
soil which is failing in a drained manner, with (τ/σ’)max = tan φ’, we find
that at failure the soil must dilate along long the failure surface with a
dilation angle ψ = φ’, ’, in order for the mechanism to be kinematically
admissible.
To
o allow for this movement, any
a curved slip surface must always alwa be at
an angle φ’ to the direction
ction of slip. This results in the slip being defined
as a logarithmic spiral.

ETB/JJMH Page 31
ENCN353: Geotechnical Engineering

We can analyse the failure surface in the same way that we analysed
the solutions used for the undrained case, however, we have this
additional dilative component to be aware of. We will not cover this
analysis in this course, except to state the result for interest, i.e.:

(covered in 3rd Pro)

Now, like the simple lower bound solution, we can improve on this
case. We will not go into detail, except to say that the lowest upper
bound solution is found to be:

(covered in 3rd Pro)

This means that an exact solution exists for this particular loading
case, since we have the same solution for lower and upper bounds.

Sensitivity to φ’
Note that, due to the natural log term (e), the exact solution is very
sensitive to the friction angle φ’ specified.

At φ’=20º (typical for a clay):

For φ’=30º (typical for a sand):

ETB/JJMH Page 32
ENCN353: Geotechnical Engineering

7.3 Bearing capacity enhancement factors

Bearing capacity enhancement factors are used to account for
differences in foundation shape, depth and soil weight from the
idealised case of a strip footing, as we have just covered.
We use different enhancement factors for short term undrained and
long term drained analyses.

7.3.1 Enhancement factors for undrained total stress analysis

The analysis using the Tresca (undrained) failure criterion is
conventionally modified using an equation of the form:

(σ f − σ 0 ) = (N C ⋅s C ⋅d C ) ⋅ cU
Where NC
sC

dC

An alternative version proposed by Brinch Hansen has a slightly

different form:

(σ f (
− σ 0 ) = N C ⋅ 1 + s c + d c ⋅ cU
* *
)
The shape factor sc (or sc*) allows for the fact that a footing is, in
reality, not infinitely long, while the depth factor dc (or dc*) takes
account of the fact that, in reality, the soil above the founding plane
has some strength and does more than simply act as a surcharge σ0.

Different values have been suggested for these factors, based on

physical model tests, numerical models and so forth. The values
quoted below are much used in industry.

ETB/JJMH Page 33
ENCN353: Geotechnical Engineering

Parameter Skempton Meyerhof Brinch Hansen

(1951) (1963) (1970)
Shape factor 1+0.2(B/L) 1+0.2(B/L) 0.2B/L
sc or sc*
Depth factor 1+0.23√ (D/B) 1+0.2(D/B) 0.4k where
dc or dc* up to max. of k=D/B if D/B≤ 1;
1.46 (D/B=4) k=tan-1(D/B) in
radian if D/B >1
Bearing capacity factors for Tresca failure criterion, for foundation of length L,
breadth B and depth D (Skempton, 1951; Meyerhof, 1963; Brinch Hansen, 1970).

7.3.2 Enhancement factors for drained effective stress analysis

The analysis using the Mohr-Coulomb (drained) failure criterion is
conventionally modified using an equation of the form:

  γB 
σ ' f = [N q ⋅ s q ⋅ d q ]⋅ σ ' 0 +  N γ ⋅ sγ ⋅ d γ ⋅ rγ ⋅  − ∆u  
  2 

Where for the left-hand part of the expression:

Nq

sq
dq

The purpose of the left-hand terms is the same for those used in the
undrained analysis (previous section).
The right-hand part of the expression is a little more complex to
explain. Basically, if we were to rely on the left hand part of the
expression, which uses σ’0 (the stress at the founding level), to
calculate the bearing capacity, it would be a significant underestimate.
This is because, with effective stress (Mohr-Coulomb) analysis, the
strength of the soil increases with applied mean stress (i.e. is
frictional). This means that the soil will have greater capacity at some
depth, due to the self-weight of the soil. (Note, this is not the case for
total stress analysis!).

ETB/JJMH Page 34
ENCN353: Geotechnical Engineering

So we need to account for the fact that the failure occurs over
approximately one footing depth B, which experiences an average
vertical effective stress of σ’0+(γB/2-∆u) – including the effects of pore
pressure.
So, the right-hand part of the expression covers the additional
influence of soil stress at depth to enhance bearing capacity, while the
left-hand part covers bearing capacity as if at the founding plane
alone.
This means for the right-hand part of the expression:
Nγ

sγ

dγ

rγ

Numerical values for the factors used in this expression as given by

Meyerhof (1963) and Brinch Hanson (1970) are given below. There
are other empirically derived factors that are also often quoted.

Parameter Meyerhof (1963) Brinch Hansen (1970)

Shape factor sq 1+0.1Kp(B/L) 1+[(B/L)tanφ’]

Depth factor dq 1+0.1√ Kp (D/B) 1+2tanφ’(1-sinφ’)k;

k=D/B if D/B≤ 1;
k=tan-1(D/B) in radian if
D/B >1
Nγ (Nq -1) tan(1.4φ’) 1.5(Nq -1) tanφ’

Shape factor sγ sq 1-0.4(B/L)

Depth factor dγ dq 1
Bearing capacity factors for Mohr-Coulomb failure criterion, for foundation of
length L, breadth B and depth D (Meyerhof, 1963; Brinch Hansen, 1970)

ETB/JJMH Page 35
ENCN353: Geotechnical Engineering

Notes on table:
Meyerhof’s expressions only apply for φ’>10º
Kp= (1+sinφ’)/(1-sinφ’)
rγ = 1 for B<2m; rγ = 1- 0.25 log10(B/2) for B≥ 2m (Bowles, 1998)

7.3.3 Improvement to enhancement factors

Some recent research suggests improvements on the enhancement
factors used in both drained and undrained analyses, which can be
somewhat conservative.
Other factors may also account for:

However, analytical solutions can also be found for most of these

things (many of which directly influence NC) without resorting to
empirical fudge-factors. There is extensive recent literature2 on the
subject.

For example animations of kinematic solutions (no hodographs

unfortunately!) see:
http://www.limitstate.com/verification/Undrained_footings.html3

More on shallow foundation design is covered in ENCI452.

2
See papers by Martin, CM and Smith, CC.
3
This website also has kinematic solutions for drained analyses

ETB/JJMH Page 36