SHALLOW FOUNDATIONS

M. Hesham El Naggar, Ph.D., P. Eng

Professor of Geotechnical Engineering

Faculty of Engineering

The University of Western Ontario

1
 II.1 Ultimate Bearing Capacity of Shallow Foundations
1. Introduction
2. Terzaghi's Bearing Capacity Theory
3. General Bearing Capacity Theory
a. Effect of footing shape
b. Effect of footing depth
c. Footing with inclined load
d. Footing with base inclination
e. Footing with ground inclination (near slopes)
f. Effect of water table on bearing capacity

4. Bearing Capacity of Footing Subjected to Eccentric or Moment Loading

5. Bearing Capacity of Foundations in Layered Soils
6. Bearing Capacity from SPT and CPT Testing
7. Safety Factors in Foundation Design and Allowable Bearing Capacity
II.2 Foundation Settlements
1. Types of Foundation Settlements
2. Methods of Computing Immediate (Elastic) Settlements
3. Consolidation Settlements
4. Reliability of Settlement Computations
5. Structural Tolerance to Total Settlement and Differential Settlements
II.3 Mat Foundation
1. Types of Mat Foundation
2. Bearing Capacity of Mat Foundation
3. Settlement of Mat Foundation

2
CRITERIA FOR FOUNDATION DESIGN

The proper design of a shallow foundation has to ensure that the structure load is

carried safely, the serviceability requirement is met and the feasibility of the

foundation solution is assured. This is ensured by:

1.Bearing capacity: The foundation, including the underlying soil and rock, must

be adequately safe against failure or the structure could fail.

2.Settlement: Total and differential settlement must remain within tolerable limits

so as not to affect the integrity and function of the structure.

3.Feasibility: The foundation must be feasible both technically and economically,

and practical to construct.

3
II.1 Ultimate Bearing Capacity of Shallow Foundations
1. Introduction
Shallow foundations are constructed on the ground surface or at a shallow depth
beneath the ground surface. The purpose of a shallow foundation is to transmit the
structural loads to the ground over a larger area than the area of the structural
component supported. The size of the footing can be selected so that the contact
pressure between the footing and the soil is small enough to provide an acceptable
factor of safety against bearing failure and to ensure that the settlement is less than
a given tolerance. Shallow foundations include spread footing foundations and
mat foundations. The analysis of spread footing will be discussed first and the
design of mat foundations will follow.

Some of the shapes of spread footing and their respective dimensions are shown in
Fig. 1.

Figure 1 Shapes and dimensions of spread footings

4
Ultimate bearing capacity: is the critical applied pressure at which the soil mass
beneath the foundation will theoretically collapse.
Allowable bearing capacity: is the applied pressure for which the foundation
should be designed (applying a factor of safety qall = qult / F).

2. Terzaghi's Bearing Capacity Theory

Terzaghi developed a general formula for ultimate bearing capacity of spread
footing foundations under the following assumptions:
• The depth of the footing is less than or equal to its width ( D ≤ B ).
• The foundation is rigid and has a rough bottom.
• The soil beneath the footing is a homogeneous semi-infinite mass.
• Strip foundation with a horizontal base and level ground surface under vertical
loads.
• The general shear mode of failure governs and no consolidation of the soil
occurs (settlement is due only to shearing and lateral movement of the soil).
• The shear strength of the soil is described by s = c + σ′ tan φ.

Figure 2 Geometry of failure surface for Terzaghi’s bearing capacity theory a)

assumed b) actual

5
Under the above mentioned assumptions, the bearing capacity is given by

qult = c′ Nc + γ′ D Nq + 0.5 γ′ B Nγ (1)

where
Nc , Nq, Nγ are bearing capacity factors and are functions of φ′
c′, φ′, γ′ = effective cohesion, angle of friction and unit weight
B, D = width and depth of the foundation

Equation 1 consists of three terms: the first represents the contribution from the
cohesive strength of the soil; the second represents the contribution from the
surcharge applied at the foundation level; and the third represents the contribution
from the self-weight of the soil beneath the foundation.

Equation 1 represents the case of strip footing. Terzaghi extended this solution to
square and circular foundations by adding empirical coefficients, i.e.

For square footing

qult = 1.3 c′ Nc + γ′ D Nq + 0.4 γ′ B Nγ (2)
and for circular foundations
qult = 1.3 c′ Nc + γ′ D Nq + 0.3 γ′ B Nγ (3)

Because of the shape of the failure surface, the soil parameters used in the first and
third terms of Eqs. 1-3 should be representative of the soil between the bottom of
the footing and a depth B below the foundation level. The unit weight of the soil
in the second term represents the soil above the foundation level.

6
3. General Bearing Capacity Theory
Equations 1 to3 do not address several aspects of actual foundation configurations.
For example, foundations can be rectangular, deeper than one width below the
ground surface or subjected to an inclined load. Several researchers considered
different deviations from Terzaghi’s assumptions and proposed some correction
factors to be applied for different situations. Vesic (1973, 1975) proposed a
general bearing capacity formula that accounts for most practical situations. He
retained Terzaghi’s basic formula for the ultimate bearing capacity and added
correction factors to be implemented as appropriate as will be discussed. Thus, it
can be applied to a much broader range of loading and geometry conditions. The
general bearing capacity formula is given by

q ult = c ′N c s c d c ic bc g c + γ ′DN q s q d q i q bq g q + 0.5γ ′BN γ sγ d γ iγ bγ g γ (4)

where
sc, sq, sγ = shape factors
dc, dq, dγ = depth factors
ic, iq, iγ = load inclination factors
bc, bq, bγ = base inclination factors
gc, gq, gγ = ground inclination factors

Vesic recommended the following formula for the bearing capacity factors:
Nq −1
Nc = for φ′ > 0 and Nc = 5.14 for φ′ = 0 (5a)
tan φ ′

N q = e π tan φ ′ tan 2 (45 + φ ′ / 2) (5b)

Nγ = 2(Nq + 1) tanφ′ (5c)

7
It should be noted that the general capacity equation is written in terms of the
effective stress parameters (drained conditions) but can also be applied for total
stress (undrained conditions) as well.

The correction factors are evaluated as discussed in the following sections.

3.a. Effect of footing shape

Shape Factors: the equation is originally for a strip foundation. Other shapes are
accounted for, using shape factors.

⎛ B ′ ⎞⎛ N q ⎞⎟
S c = 1 + ⎜ ⎟⎜⎜ (6a)
⎝ L ′ ⎠⎝ N c ⎟⎠
⎛ B′ ⎞
S q = 1 + ⎜ ⎟ tan φ ′ (6b)
⎝ L′ ⎠
⎛ B′ ⎞
S γ = 1 − 0.4⎜ ⎟ (6c)
⎝ L′ ⎠

where B ′ and L ′ are effective width and length of the footing, and equal to B and L
for centrically loaded footings. For footings subjected to eccentric loading or
moment, B ′ and L ′ should reflect the pressure distribution beneath the footing as
will be discussed later.

For strip footing, B/L approaches 0, so Sc, Sq and Sγ become equal to 1.

8
3.b. Effect of footing depth
Terzaghi’s theory limits the depth of the footing to one B, and ignores the shear
resistance within the soil above the footing level. Vesic removed this limitation by
introducing depth factors. The depth of the footing is considered using the
following depth factors.

d c = 1 + 0.4k (7a)

d q = 1 + 2k tan φ ′(1 − sinφ ′) 2 (7b)

dγ = 1 (7c)

where k = D/B, for D/B ≤ 1 and k = tan-1(D/B) for D/B > 1.

9
3.c Footing with inclined load
Load inclination factors are used to calculate the capacity of a footing subjected to
centric loads that do not act perpendicular to the base of the footing. The load
inclination factors are:

mH
ic = 1 − ≥0 (8a)
B ′L ′c ′N c
m
⎡ H ⎤
iq = ⎢1 − ⎥ ≥0 (8b)
⎣ V + B ′L ′c ′cotφ ′ ⎦
m +1
⎡ H ⎤
iγ = ⎢1 − ⎥ ≥0 (8c)
⎣ V + B ′L ′c ′ cotφ ′ ⎦

where m = (2+x)/(1+x); x = B/L for H parallel to B; x = L/B for H parallel to L, and

V and H are the vertical and horizontal components of the inclined load.

The load inclination factors are equal to 1 for loads acting perpendicular to the
base of the foundation. Both iq and iγ = 1 for φ = 0.

10
3.d Footing with base inclination
In some cases the load acting on the footing is inclined at a large angle from the
vertical. In this situation, it may be better to incline the base of the footing at the
same angle. The base inclination factors are:
α
bc = 1 − (9a)
147°
⎛ α tan φ ′ ⎞
bq = bγ = ⎜1 − ⎟ (9b)
⎝ 57° ⎠

3.e. Footing with ground inclination (near slopes)

A footing located near the crest of a slope has a lower bearing capacity than that on
level ground. The ground inclination factors are:

β
gc = 1 − (10a)
147°
g q = g γ = (1 − tan β )
2
(10b)

If β = 0, the ground inclination factors become equal to 1.

11
3.f. Effect of rate of loading and level of water table on bearing capacity
The ultimate bearing capacity of a footing depends on the type of soil, rate of
loading and the location of the ground water table (GWT) relative to the
foundation level.

The bearing capacity is calculated for two cases of rate of loading. The first case is
rapid loading (short term conditions), in which undrained conditions are assumed
and the total stress analysis is considered when evaluating the footing capacity.
The second case is slow loading (long term conditions), in which drained
conditions are assumed and the effective stress analysis is considered when
evaluating the footing capacity.

Undrained Conditions: For clay soils loaded rapidly, undrained conditions are
assumed (i.e. c = cu, φ = 0° and γ = γtotal or γbulk). In this case (for φu = 0), Nc =
5.14, Nq = 1, Nγ = 0 and the bearing capacity becomes:

qult = 5.14cu s c d c ic bc g c + γDs q d q iq bq g q (11)

The undrained shear strength, cu, should be representative of the soil within a
distance 2/3-1 B below the foundation level. The soil unit weight is γbulk = the bulk
unit weight of the soil above the foundation level.

Drained Conditions: For clay soils loaded slowly and for sands, drained conditions
are assumed (c = c′, φ = φ′ and γ = γ′). For sands and normally consolidated clays,
c′ = 0. Equation 4 is used to calculate the ultimate bearing capacity. The depth of
the GWT influences the choice of the soil unit weight in the equation as follows.

12
 i. If the GWT is at a depth B or lower below the foundation level, use γbulk.
ii. If the GWT is at the foundation level, use γbulk for the second term of Eq. 4
and γsub for the third term.
iii. If the GWT is at a level between these two cases, use γbulk for the second
term of Eq. 4 and an averaged unit weight (of γbulk and γsub) for the third
term.
iv. If the GWT is above the foundation level, use γsub for the third term and the
averaged unit weight (of γbulk and γsub) for the second term.
v. If the GWT is at the ground surface, use γsub for both terms.

13
4. Bearing Capacity of Footing Subjected to Eccentric or Moment Loading
If the load applied to the footing is eccentric or the footing is subjected to a
bending moment in addition to the vertical load, it is assumed that the load acts at
the centre of a footing of reduced size (effective size) as shown in Fig. 3.

The effective dimensions are computed as follows:

B′ = B –2 eB (12a)
L′ = L – 2eL (12b)

For a footing subjected to a vertical load V, moment MB and a moment ML, the
eccentricities eB = MB/V and eL = ML/V.

Figure 3 Footing subjected to eccentric vertical load

14
5. Bearing Capacity of Foundations in Layered Soils
The general bearing capacity equation assumes that the footing is supported by a
homogeneous soil deposit that extends to a considerable depth. In practice, layered
soil profiles are often encountered. The soil within one B below the footing level
influences the capacty of the footing. Therefore, the soil parameters used in Eq. 4
should be representative of this soil mass.

For a footing that rests on two distinct clay layers (Fig. 4), Meyerhof and Hanna
proposed the following solutions.

D
BxL
B
H
Layer I cu1, γ1

Layer II cu2, γ2

Figure 4 Footing on layered soil

Case I: cu1/cu2 > 1 (top layer is stronger)

⎡ ⎛ B ⎞⎤ ⎛ B ⎞⎛ 2 c H ⎞
q ult = ⎢1 + 0.2⎜ ⎟ ⎥ cu 2 N c + ⎜1 + ⎟⎜ a ⎟ + γ 1 D
⎣ ⎝ L ⎠⎦ ⎝ L ⎠⎝ B ⎠
(13)
⎡ ⎛ B ⎞⎤
≤ ⎢1 + 0.2⎜ ⎟ ⎥ cu1 N c + γ 1 D
⎣ ⎝ L ⎠⎦

where ca is adhesion at the interface between the two layers and is given in Fig. 5.

15
Figure 5 Adhesion ca for the case of strong clay layer underlain by a weak clay
layer (after Meyerhof and Hanna, 1978)

Case II: cu1/cu2 < 1 (top layer is weaker)

2
⎛ H⎞
qult = qt + (qb − qt )⎜1 − ⎟ ≥ qt (14a)
⎝ B⎠

⎡ ⎛ B ⎞⎤
qt = ⎢1 + 0.2⎜ ⎟⎥ cu1 N c + γ 1 D (14b)
⎣ ⎝ L ⎠⎦

⎡ ⎛ B ⎞⎤
qb = ⎢1 + 0.2⎜ ⎟⎥ cu 2 N c + γ 2 D (14c)
⎣ ⎝ L ⎠⎦

Some engineers consider Eq. 14a to be non-conservative. Hence, proper

engineering judgment should be exercized when using it.

16
6. Bearing Capacity from SPT and CPT Testing
The ultimate bearing capacity of a footing supported on sand may be estimated
from the SPT value, N, using the correlation proposed by Parry (1977)

⎛ D + 0.73B ⎞
qult ( MPa) = 0.24 N F ⎜ ⎟ (15)
⎝ D + 0.75 B ⎠

where NF = N value obtained from the field (SPT) at a depth of 0.75B below the
foundation level, D and B are in metres. For D/B < 1, qult = 0.24NF.

The ultimate bearing capacity of a footing (with D/B ≤ 1.5) may be estimated from
CPT measurements, qc, using correlations proposed by Schmertmann (1978).

Footing on sand
qult (ton / ft 2 ) = 28 − 0.0052(300 − q c )1.5 (for strip footing) (16a)

qult (ton / ft 2 ) = 48 − 0.009(300 − q c )1.5 (for square footing) (16b)

Footing on clay
qult (ton / ft 2 ) = 2 + 0.28q c (for strip footing) (16c)

qult (ton / ft 2 ) = 5 + 0.34q c (for square footing) (16d)

It should be noted that these relationships are approximate.

17
7. Safety Factors in Foundation Design and Allowable Bearing Capacity
Most practising geotechnical engineers use the allowable stress design (ASD)
method when designing shallow foundations. In this case, the allowable bearing
capacity is given by:
qult
q all = (17)
F
The foundation is then designed so that the applied bearing pressure, qapp, does not
exceed the allowable pressure, qall, i.e.,
qapp ≤ qall (18)
The value of the factor of safety depends on:
• soil type;
• site investigation;
• soil variability; and
• importance of the structure and consequences of a failure.

Limit States Design

The limit states design has begun to gain popularity among geotechnical engineers.

In the limit states design the bearing capacity of the footing is considered as part of

the Ultimate Limit State (ULS). The safety of the foundation is satisfied in the

ULS design by using partial safety factors for the load and strength parameters.

This is done as follows:

1. Factored loads are calculated by multiplyng specified loads and forces by load

factors obtained from the respective codes (e.g. NBCC or OHBDC). The

18
 factored applied pressure is obtained by dividing the factored load by the base

area of the footing.

2. Design shear strength parameters are calculated by multiplying the shear

strength parameters of the soil by resistance factors. The design capacity of the

foundation is calculated using the design shear strength parameters.

3. To satisfy the ULS, the design capacity must be equal to or greater than the

factored applied pressure.

The load factors are 1.25 for Dead Load (DL), 1.5 for Live Load (LL), Wind Load

(WL) and Seismic Load (SL). The resistance factors are 0.8 for angle of internal

friction (fφ) and 0.5 to 0.7 for cohesion (fc).

19
Example
A square footing 1.5 x 1.5m in plan, is founded at a depth of 1m in a deep layer of
clay. The footing is loaded by an inclined load and moment loadings such that the
eccentricity in both directions is 0.3m. The water table is at foundation level and
the saturated unit weight of the clay is γsat = 17 kN/m3. The same value can be used
for γbulk of the clay above the GWT. The representative properties of the soil are:
cu = 60kPa, φu = 0°, c′ = 0 and φ′ = 28°. The horizontal component of the inclined
load acts along a diagonal of the footing. If the magnitude of the inclined load is
100kN and it is inclined at 60° to the horizontal, compute the bearing capacity of
the footing for
a) Short term conditions (rapid loading); and
b) Long term conditions (after dissipation of excess pore pressure).

20
21
II.2 Foundation Settlements
A successful design of shallow foundations must include settlement analysis as
well as the bearing capacity analysis. In many cases, settlement controls the design
of shallow foundations, especially when the foundation width is large, and the
bearing capacity analysis is secondary. The allowable bearing capacity (design
bearing capacity) is the smaller of the values:

⎧ qult
⎪
q all = ⎨ FS (20)
⎪q all ( settlement )
⎩

1. Types of foundation settlements

The components of settlement of a foundation are:
1. Si = immediate settlement
2. Sc = consolidation settlement
3. Ss = secondary compression ( creep )

These components are shown on a typical settlement-time relationship of a

foundation in the figure.

22
The settlement increase with time can be described by

S(t) = Si + U Sc + Ss (21)

where U = average degree of consolidation. Generally, the final settlement of a

foundation is of interest, and U is considered equal to 1 (i.e. 100% consolidation)
as it represents the largest settlement that will occur. Thus, the final (total)
settlement is:
Sf = S i + S c + Ss (22)

NOTES:
• Ss is usually neglected in settlement computations. However, this is an
important component in organic soils, peat and some clays.
• For sand and gravel, water dissipates immediately and Sc ≈ 0 ⇒ Sf = Si.
• Si is calculated from elastic theory, U from 1-D theory of consolidation, and Sc
from 1-D consolidation theory and sometimes with Skempton and Bjerrum
correction.

2. Methods of Computing Immediate (Elastic) Settlements

Footing on homogeneous soil:
The immediate (or undrained) settlement, Si, is obtained from elastic theory
assuming semi-infinite soil mass

Si = Cs q B (1 - ν2) / Eu (23)

23
where q is the increase in stress due to the foundation, B is the foundation width,
and ν and Eu are the undrained Poisson’s ratio and elastic modulus of the soil,
respectively. Finally, Cs is the influence factor which is a function of the shape
and rigidity of the foundation. The values of Cs are given in Table 1. The cases
for a flexible and a rigid foundation are compared in this table.

Notes:
For footings on sand, Eu ≈ Es′ and ν = ν′, and Sf = Si. For footings in clay, νu =
0.5. Values of Eu obtained from conventional unconfined compression tests on
tube samples of clays are usually too low. Values determined from block samples
or back-figured from settlement observations of nearby structures on the same soil
deposit are much more representative.

Table 1 Values of shape and rigidity factor Cs

Shape Centre Corner Average
Circle 1.00 0.64 0.85
Circle (rigid) 0.79 0.79 0.79
Square 1.12 0.56 0.95
Square (rigid) 0.99 0.99 0.99
Rectangle 1.5 1.36 0.67 1.15
(L/B) 2 1.52 0.76 1.30
3 1.78 0.88 1.52
5 2.1 1.05 1.83
10 2.53 1.26 2.25
100 4.00 2.00 3.70

24
Footing on soil layer of finite thickness:
For rigid foundations (rectangular or square) in a layer underlain by bedrock, the
settlement is calculated as

Si = ic q B / Eu (24)

where ic = influence coefficient, as given in Fig. 6 (Fig. 12.2 CFEM). For sand and
gravel, Eu = Es′.

Figure 6 Nomogram for calculating settlement at the characteristic point (after

Kany, 1959)

25
Christian and Carrier (1978) presented a similar solution for flexible footings
(rectangular, square or circular), but it only applies to the undrained case (νu = 0.5).
The immediate settlement is given by:

qB
S i = µ 0 µ1 (25)
Eu

where µ0 and µ1 are depth and shape factors obtained from Fig. 7.

Figure 7 Depth and shape factors for settlement evaluation of flexible footings
(after Christian and Carrier, 1978)

26
Note: Values of Eu obtained from conventional unconfined compression tests on
tube samples of clays are usually too low. Values determined from block samples
or back-figured from settlement observations of nearby structures on the same soil
deposit are much more representative.

3. Consolidation Settlements
The consolidation settlement of cohesive soils is normally computed based on
laboratory tests. Consolidation tests are performed on undisturbed samples of the
soil and the test results are used to establish the parameters that characterize the
deformation of the soil. These parameters depend on the stress history of the soil
formation and its permeability. These parameters are used to calculate the
foundation settlement using the void-ratio approach, i.e.,

H ⎡ ⎛ σ ′p ⎞ ⎛ σ ′ + ∆σ ′ ⎞⎤
Sc = ⎢C cr log⎜⎜ ⎟⎟+C c log⎜ v 0 ⎟⎥ (26)
⎜ ⎟
1 + e0 ⎣⎢ ⎝ σ v′0 ⎠ ⎝ σ ′p ⎠⎦⎥

where H = layer thickness, e0 = soil void ratio, ∆σv = stress increase at the centre of
′
the layer due to the foundation load, σ v 0 = initial effective vertical stress at the
centre of the soil layer, σp′ = preconsolidation pressure of the soil, Cc =
compression index and Ccr = recompression index. The parameters e0, σp′, Cc and
Ccr are to be determined from consolidation tests in the lab. For normally
′
consolidated soil, σ ′p = σ v 0 and Eq. 26 reduces to:

H ⎡ ⎛ σ v′0 + ∆σ ′ ⎞⎤
Sc = ⎢ c ⎜⎜
C log ⎟⎟⎥ (27)
1 + e0 ⎣ ⎝ σ ′
v0 ⎠⎦

27
Alternatively, the consolidation settlement can be calculated from:

H
S c = ∫ mV ∆σ V dz (28)
0

where mv = volume coefficient of compressibility, obtained from lab tests.

Modified 1-D Theory of Skempton and Bjerrum

The consolidation settlement calculations given by Eqs. 26-28 are based on 1-D
laboratory consolidation tests. In the field, the state of stress is 3-dimensional and
the calculated settlement should be corrected for that effect. Skempton and
Bjerrum (1957) proposed a settlement correction factor to calculate the
consolidation settlement accounting for the effects of the limited area of loading
(3-D effects). This factor is used to modify the calculated 1-D settlement as
follows:
1. Use Eq. 26, 27 or 28 (as appropriate) to obtain the 1-D consolidation settlement.
2. Determine the pore water pressure parameter, A, from triaxial tests
u −σ3
(A= ). The ranges of A are: sandy clays (0.5 to 0.7); normally
σ1 −σ 3
consolidated clays (0.5 to 1); and overconsolidated clays (-0.5 to 0).
3. Determine the correction factor (settlement ratio), µ, from Fig. 8 for a given
value of Hc/B, where Hc = thickness of clay layer.
4. Calculate the modified 3-D consolidation settlement from

Sc (3-D) = µ Sc (1-D) (29)

28
Observations of settlement showed fair agreement between predicted settlement
using Eq. 29 and measured settlement.

The final settlement is then obtained as the sum of the immediate settlement
(undrained conditions) and the consolidation settlement (after dissipation of excess
pore water pressure).

Fig. 8 Correction factor for 3-D effects (settlement ratio)

29
NOTES:
1. The bearing capacity of granular material is usually high. In this case, limiting
settlement usually controls the design of shallow foundations. Allowable soil
pressure is limited by settlement of 25 mm.
2. The determination of the consolidation settlement requires calculating the stress
increase at the mid point of the clay layer. There are different approaches to
calculate the stress increase with an elastic medium (Boussinesq solution and its
variations, Westergaard solution and its variations and the 2v:1h approximate
approach. In practice, the2v:1h approach is usually used to evaluate the stress
increase at the mid point of the clay layer.
3. For thick clay layers, the layer is subdivided into a number of sublayers. The
consoildation settlement is then calculated as the sum of the settlement of the
sublayers.

4. Reliability of Settlement Calculations

The engineer has to make several simplifying assumptions when calculating the
settlement for actual field conditions. These assumptions include: the soil modulus
(undrained and drained), the Poisson’s ratio, the compression index, coefficient of
volume change, preconsolidation pressure, drainage conditions, and thickness of
clay layer. Field performance may not be similar to predictions made based on
these assumptions (e.g. soil layering is not uniform, in situ properties of the soil are
different from values established in the laboratory). It is important to monitor field
performance, and some adjustments may be needed during construction. For
important projects, field tests such as plate loading tests are usually performed for
the evaluation of bearing capacity and settlement.

30
5. Structural Tolerance to Total Settlement and Differential Settlements

The total and deferential settlement of a structure are kept within certain limits to
avoid:
a) overstressing the structure
b) excessive maintenance or unacceptable aesthetic.

The tolerable settlements of different structures vary considerably e.g., for road
embankments, storage silos and tanks a settlement of 300mm - 600mm may be
acceptable, but for machine foundations the settlement may be limited to 5mm -
30mm.

Large settlement of various components of a structure may cause damage to the

structure and/or render it unfunctional (serviceability limit state). In many cases,
the differential settlement is more damaging than the total settlement. To reduce
differential settlement, the designer may limit the total settlement and use the
following equation for the calculation of the differential settlement:

(Sdiff)Max = 1/2 Sf (31)

The differential settlement may also be evaluated in terms of the angular distortion
given by:
∆
Sdiff = (30)
Li
where ∆ = relative settlement between two points and Li is the lateral dimension
between the two points.

31
Based on a large number of settlement observations and performance of structures,
the suggested limits for tolerable differential settlements are shown in Table 2
(Table 12.4, CFEM).

Table 2 Angular distortion limits for structures

Angular
Type of limit and structure
distortion
1/100 Danger limit for statically determinate structures and retaining walls
- Safe limit for statically determinate structures and retaining walls
1/150 - Danger limit for open steel and reinforced concrete frames, steel
storage tanks, and tilt of high rigid structures
- Safe limit for open steel and reinforced concrete frames, steel
storage tanks, and tilt of high rigid structures
1/250 - Danger limit for panel walls of frame buildings and tilt of bridge
abutments
- Tilting of high buildings may become visible
1/300 Limit when difficulties with overhead cranes are to be expected
- Safe limit for panel walls of frame buildings and tilt of bridge
1/500 abutments
- Danger limit for sagging of unreinforced load-bearing walls
Limit when difficulties with machinery sensitive to settlement are to
1/750
be expected
- Safe limit for sagging of unreinforced load-bearing walls
1/1000
- Danger limit for hogging of unreinforced load-bearing walls
1/2000 Safe limit for hogging of unreinforced load-bearing walls

32
II.3 Mat Foundation
A mat (raft) foundation is a large spread footing that provides support for the entire
structure. Mat foundations are used when spread footings for a building cover
more than one third of its footprint area or when the soil profile is erratic (i.e.
spatially non-uniform) and prone to excessive differential settlements. They are
also used when uplift loads are larger than the capacity of spread footings (e.g.
soils with heave problems) or when the bottom of the structure is located below the
ground table and waterproofing becomes represents an important concern.

1. Types of Mat Foundation

Some of the common types of mat foundations are: flat plate with uniform
thickness; flat plate with localized thickness increase under columns (not
common); flat plate with beams running both directions and the columns are
located at beam intersections; and slab with basement walls. Mats may also be
supported by piles.

33
2. Bearing Capacity of Mat Foundation
Various methods have been used to design mat foundations. The different methods
idealize the mat in general either as a rigid block (same as spread footings) or a
flexible plate.

Rigid assumption
The rigid block assumption allows the bearing capacity of the mat foundation to be
evaluated using the same equations employed for spread footings (e.g. Eq. 4). In
this case, the pressure distribution beneath the mat is either uniform (if it is loaded
centrically) or varies linearly (if eccentric or moment loads are present). This
assumption, however, may be non-conservative for the structural design as it may
underestimate the shears, moments and deformations of the mat.

Flexible assumption
The flexible assumption is more accurate in modeling the pressure distribution
beneath the mat and in evaluating the shears, moments and deformations of the
mat. However, it is more difficult to implement.

There are different approaches that are based on the flexible assumption. These
approaches include: the subgrade reaction approach, the Winkler approach, the
pseudo-coupled approach and the finite element approach.

Subgrade reaction approach: The soil resistance is modeled using the “modulus of
sugrade reaction” which is given by:

q
ks = (32)
δ

34
where ks = coefficient (modulus) of subgrade reaction, q = applied pressure, and δ
= settlement. The soil is then represented as a bed of springs each with a stiffness
ks per unit area. The coefficient of subgrade modulus can be obtained from plate
load tests.

Winkler model: It is based on the subgrade modulus approach. A set of springs is

used to characterize the soil resistance. The stiffness constant can be linear or
equivalent linear to account for the nonlinearity of the soil. The disadvantage of
the Winkler model is that the settlement under the mat would be uniform if the
applied load is uniform. The actual performance of mats shows that the central
zone of the mat experiences higher contact pressure and higher settlement than the
exterior zone of the mat.

Pseudo-coupled method: It is similar to the Winkler approach but the constant of

the springs in the central zone of the mat would be less than the constant of the
springs along the exterior of the mat. This is done to account for the fact that the
central zone of the foundation experiences higher contact pressure and higher
settlement. With weaker springs within the central zone of the mat, the pattern of
predicted settlement would be similar the observed behaviour.

Finite element method: The mat and the supporting soil are modeled using the
finite element approach. It allows modeling the flexibility of the mat, nonlinearity
and inhomogeneity of the soil. It yields stresses and deformations within the mat
and the supporting soil.

35
36