8 Footings on Sand

CONTENTS

8.1 Allowable Bearing Capacity

8.1.1 Shear Failure
8.1.2 Relative Density
8.1.3 Penetration Resistance
8.1.4 Gradation
8.1.5 Meyerhof’s Analysis
8.2 Settlement of Footings
8.2.1 Footing Size and Settlement
8.2.2 Footing Depth and Settlement
8.2.3 Penetration Resistance and Settlement
8.2.4 Water Table and Settlement
8.3 Rational Design of Footing Foundation on Sand
8.3.1 Typical Design Example
References

The principle of the design of footings on sands is essentially the same as the design
of footing on clays. In soil mechanics, the definition of sand refers to cohesionless
soils with little or no fines. This includes gravely sands, silty sands, clean sands,
fairly clean sands, and gravel.
Engineers as well as the public generally have the conception that sandy soils
are good bearing soils and will not pose much of a foundation problem. In fact, in
Riyadh, Saudi Arabia, during the oil boom period, most structures were erected on
sandy soils without the benefit of soil investigations.
In fact, distress experienced on structures founded on sand is not uncommon,
especially for subsoils containing large amounts of cobbles. Excessive settlement or
sometimes even shear failure can take place when there is a sudden change of the
water table elevation.

8.1 ALLOWABLE BEARING CAPACITY

The criteria for designing a safe foundation on sand are the same as those for footings
on clay. That is, the possibility of footings breaking in the ground generally refers
to “shear failure” or “punching shear,” and settlement produced by the load should
be within a tolerable limit.
For structures founded on sandy soils, the settlement can take place almost imme-
diately. The settlement criteria generally determine the allowable bearing capacity. Still,
the possibility of shear failure cannot be ignored. Geotechnical engineers often found
that punching shear took place at narrow footings and was ignored in the design.

0-8493-????-?/97/$0.00+$.50
© 1997 by CRC Press LLC

©2000 CRC Press LLC

 FIGURE 8.1 Cross-section through long footing on sand (left side, after Peck).

Sometimes it is difficult to determine whether the structure failed due to excessive

settlement or due to shear. For a consulting engineer dealing with medium or loose
sand, an ample factor of safety should be used. The design criteria should depend on
the results obtained from in situ testing rather than from theoretical analysis.

8.1.1 SHEAR FAILURE

The concept of shear failure of footing on sands (Figure 8.1) was established first
by Prandtl and later extended by Terzaghi, Meyernof, Buisman, Casqot, De Beer,
and many others.
The general approaches of all studies are similar. They usually follow the basic
assumption that the soil is homogenous, from the surface to a depth that is at least
twice the width of the footings.
As explained by Peck, the wedge a’o’d cannot penetrate the soil because of the
roughness of the base. It moves down as a unit. As it moves, it displaces the adjacent
material. Consequently, the sand is subjected to severe shearing distortion and slides
outward and upward along the boundary’s o’bd. The movement is resisted by the
shearing strength of the sand along o’bd and the weight of the sand in the sliding masses.
The mechanics involving the ultimate bearing capacity under such a condition
is very complex. It involves the passive pressure exerted by the adjacent soils, further
complicated by the drained and undrained conditions. The result in which the
consultants are interested is that the ultimate bearing capacity may be expressed as

1
qd ¢ - Bg Ng + g Df Nq
2

and the net ultimate bearing capacity as

qd = qd ¢ - g Df

=
1
(
B N + g D f Nq - 1 .
2 g g
)
where Ng Nq are bearing capacity factors, their values can be evaluated by Figure 8.2.

©2000 CRC Press LLC

FIGURE 8.2 Relation between bearing capacity factors and angle of internal friction and
penetration resistance (after Peck).

8.1.2 RELATIVE DENSITY

The relative density of sand is defined by the equation:

Dr =
(eo - e)
(eo - emin )
in which eo = void ratio of sand in its loosest state
emin = void ratio of sand in its densest state, which can be obtained in the
laboratory
e = void ratio of sand in the field

Relative density can be determined when the maximum, the minimum, and the
actual field density of the sand are known. The more uniform the sand (SP), the
nearer its eo and emin will approach the values of equal spheres. For well-graded
sands (SW), both eo and emin values are small and as a result a higher relative density
value is expected.

©2000 CRC Press LLC

8.1.3 PENETRATION RESISTANCE
The accuracy in the determination of the in situ angle of internal friction for granular
soils depends on the quality of the undisturbed samples. The procedures involve an
experienced operator, costly equipment, and time-consuming activities. For soil
containing a large percentage of gravel and cobble, laboratory testing depends on
the use of a large-diameter triaxial cylinder. For small projects, such elaborate
sampling and testing are often not justified.
The simplest and least expensive procedure is to correlate internal friction value
with standard penetration test results, as shown in Figure 8.2. The standard penetra-
tion test result can often be deceiving, as discussed in Chapter 3. Data obtained from
the penetration resistance test, when carefully employed, still presents the most
convenient and inexpensive method for determining the bearing capacity of granular
soils.
Geotechnical consultants depend on the field penetration resistance value to
assign the bearing capacity of footings on sand. Figure 8.3 is based on depth of
surcharge D = 3 and a factor of safety of two. The choice of the factor of safety is
discussed under a separate heading.
For most projects, the width of the footing is less than 5 ft. The beginnings of
the curves actually control most construction.

8.1.4 GRADATION
The gradation of granular soils directly affects their relative density, and hence their
bearing capacity. When gravely soils are encountered, their bearing value as well as
their amount of settlement can be different from sands with the same angle of internal
friction. Both the maximum and the minimum densities increase with the higher
percentage of gravel, up to probably as much as 60%. The presence of cobbles also
greatly affects the bearing capacity. The larger the size of the aggregate, the higher
the maximum and the minimum densities.
If the soil contains more than 50% of gravel with a maximum size exceeding
2 in., the values of the bearing capacity assigned for sands as indicated by the
previous curves are usually very conservative.
Gradation analysis should be performed for every granular soil. Care should be
taken that the sample represents the average site subsoil. With the information
furnished from the gradation analysis, it is possible to have a better correlation
between the penetration resistance and the bearing capacity.

8.1.5 MEYERHOF’S ANALYSIS

Bearing capacity analysis made by G.G. Meyerhof assumed that the shear zone
extends above the foundation level. Consequently, he assigned a much higher bearing
value than that of Terzaghi. Meyerhof estimated the relationship between density,
penetration resistance, and angle of internal friction of cohesionless soil as in
Table 8.1.

©2000 CRC Press LLC

FIGURE 8.3 Width of footing versus bearing capacity for various penetration resistance
value.

©2000 CRC Press LLC

 TABLE 8.1
Relationship Between Density, Penetration Resistance,
and Angle of Internal Friction
State of Relative Standard Penetration Angle of Internal
Packing Density Resistance (blow/foot) Friction (degree)

Very loose 0.2 4 30

Loose 0.2–0.4 4–10 30–35
Compact 0.4–0.6 10–30 35–40
Dense 0.6–0.8 30–50 40–45
Very dense 0.8 50 45

Using the above relationship, the ultimate bearing capacity can be expressed as
follows:

È Dù
qd = NB Í1 + ú 200
Î ( B) û

where D is the depth of the surcharge, B is the width of the footings, and N is the
penetration resistance in blows per foot.
From Figure 8.4, Meyerhof’s values are considerably higher, even with the use
of a factor of safety of three. It appears that his solution is nearer to the observed
actual load test results.
In Terzaghi’s analysis, footings with a width less than 10 ft and founded on
sands with a penetration resistance less than ten blows per foot, there is the risk of
shear failure. In Meyerhof’s solution, even with a very narrow footing founded on
low blow count sands, shear failure will not take place.
It is believed that Meyerhof’s solution is more realistic and should be used at
least for the upper limit in the bearing capacity determination. In sands containing
more than 50% gravel and cobble, Meyerhof’s solution can be applied with confi-
dence. However, for fine uniform sand, Meyerhof’s values should be used with care.

8.2 SETTLEMENT OF FOOTINGS

The stress and strain relationships of sand cannot be approximated by a straight line.
Hence, the term modulus of elasticity of the sand mass cannot be applied. An elastic
theorem cannot be used for estimating the amount of settlement for footings on sand
under a static load. The factors affecting the settlement of footings on sand are as
follows:

1. Relative Density — The rigidity of a sand mass increases sharply with

the increase of its relative density
2. The shape and size of the sand grain

©2000 CRC Press LLC

FIGURE 8.4 Width of footing versus bearing capacity for various penetration resistance
values, as used by Meyerhof.

3. Unit Weight — Unit dry weight directly reflects the degree of compact-
ness of the sand mass
4. Water Table — Since the submerged unit weight of sand is only about
half that of moist or dry sand, the water table plays an important role in
settlement.

©2000 CRC Press LLC

FIGURE 8.5 Relation between the width of square footing and settlement under same load
per unit area (after Peck).

8.2.1 FOOTING SIZE AND SETTLEMENT

Terzaghi has shown that about 80% of the total settlement is due to the consolidation
of the soil mass within the pressure bulb, bounded by the line representing a vertical
pressure of one fifth of the applied load intensity. By similarity, it is apparent that the
settlement should be proportional to the width (Figure 8.5). However, as soils cannot
be considered as homogeneous material, especially for the cohesionless soil, the effect
of size on settlement cannot be determined from the theoretical considerations.
Peck stated in 1996 that at a given load per unit of area of a base of a footing,
the depth of the body of sand subject to intense compression and deformations
increases as the width of the footing increases. On the other hand, at very small
widths, the ultimate bearing capacity of a loaded area is very small; consequently
at very small widths, even at low soil pressures, the loaded area may sink into the
ground as shown in Figure 8.5. This is one basis for commonly requiring the test
plate in a load test be at least 4 ft2.

8.2.2 FOOTING DEPTH AND SETTLEMENT

For footings of a given size, the greater the depth below the original grade, the
greater may be the allowable bearing pressure for a given settlement. It is not the
depth that directly affects the results; the important factor is the ratio of the depth
to the width (D/B), which is termed the depth factor. It was shown that for a depth
equal to one half the width of the footings, the amount of settlement is only half of
that in the surface loading condition.
Adjustment of the depth effect on the allowable bearing pressure on sand is
seldom attempted. However, it is a general practice for geotechnical engineers to
assign a higher bearing pressure for piers bottomed on sand than spread footings
founded on the same material.
It must be emphasized that the depth effect is based on undisturbed, homogenous
sands; disturbance to the soil during construction may affect the depth effect. Terzaghi

©2000 CRC Press LLC

 FIGURE 8.6 Penetration resistance and allowable pressure.

has called attention to the fact that the loosening of soil during the excavation of a
deep shaft in sand may lead to settlements that are as large as those which would
occur under the same loading at ground surface.

8.2.3 PENETRATION RESISTANCE AND SETTLEMENT

The simplest and easiest method in evaluating the bearing capacity of footings
founded on granular soils is by correlating with the penetration resistance value
(Figure 8.6). The standard penetration test, when performed on medium-grain gravel
and sand, is reliable and easy to perform. As early as 1948, Terzaghi and Peck
proposed the correlation of bearing capacity and penetration resistance with the
following equation:

N
qd = ¥ 2000
8

where qd = The allowable bearing capacity in pounds per square foot

N = Penetration resistance in blows per foot

©2000 CRC Press LLC

 The above equation is based on the following assumptions:

1. Allowable pressure is based on a footing settlement of 1 in.

2. The water table is at a depth of at least a distance equal to the width of
the footing.
3. The equation established on the basis that the width of the footing is less
than four feet, which is the size of footings most commonly used.
4. The footings are placed on the surface of the sand with no consideration
of the depth effect.
5. The soil consists of sands with little or no gravel.

Meyerhof in 1965 stated that by using Peck’s figure, the estimated settlements
vary from approximately 1.5 to 3 times the observed value.
In a recent publication, Peck changed the bearing capacity versus penetration
resistance equation by N/5 instead of N/8. Still, the consultants found that the
relationship is conservative. For sands with some gravel and for the usual footing
depth, about 3 ft below ground surface, the allowable soil pressure can be greatly
increased. At an upper limit, the equation can be modified as:

qd = Ê ˆ ¥ 2000 psf
N
Ë 5¯

The above relationship is plotted as shown in Figure 8.6. In choosing the proper
allowable pressure, geotechnical consultants should rely on their judgment more
than charts and figures. The following should enter into consideration:

1. The percentage of gravel and cobbles in the deposit

2. The depth and width ratios of the footing
3. The amount of silt and clay in the deposit
4. The possibility of rise of the water table
5. The possibility of the development of a perched water condition

If all conditions are not favorable, use Peck’s value. These conditions include
the following: if the material does not contain a large amount of gravel; if the footings
are placed near ground surface; and if the water table is near the base of the footings.
On the other hand, if all conditions are favorable, and if neighboring structures have
not suffered damage, then there is no reason why the upper limit as shown in
Figure 8.6 cannot be economically used.
Using the above relationship, a convenient chart can be made for estimating the
allowable soil pressure for various sizes of footings based on the penetration resis-
tance data as shown in Figure 8.7. A similar chart can be prepared for the upper
limit used by Peck. It is assumed that the footing width between 1 and 3 ft has the
same allowable pressure. The chart is not applicable for a footing width of less than
1 ft.

©2000 CRC Press LLC

FIGURE 8.7 Relationship between N value and allowable pressure for maximum settlement
of 1 in.

8.2.4 WATER TABLE AND SETTLEMENT

The position of the water table plays an important role in the determination of the
stability of foundations on sandy soils. The water table elevation affects both the
bearing capacity and the settlement of the foundation.

©2000 CRC Press LLC

 On the issue of correction for the water table, Peck commented in 1996 as
follows: If the water table lies above the loaded depth of the footing, the confining
pressure of the sand is reduced. Hence, the settlement correspondingly increases as
compared to the values if the water tables were below the loaded depth. However,
the reductions in confining pressure also cause the reduction in the standard pene-
tration resistance value. The two effects largely compensate for each other. Therefore,
the presence of a high water table can appropriately be ignored, and no water table
correction is needed.
On the other hand, if the water tables were to rise into or above the loaded depth
after the penetration tests were conducted, the actual settlement can be totally
different. The confining pressure of sand beneath the footings and beside the footings
is proportional to the unit weight of the sands. Hence if the sand mass has changed
from the dry or moist state to a submerged state, the settlement of the footings is
likely to be increased by as much as twice the amount.
It is important that the geotechnical engineer check the possibility of shear failure
on narrow footings founded on loose sand, with the water table located near the
footing level. Another aspect is the possibility that with the high water table condi-
tion, the ultimate bearing capacity of granular soils may be reduced by liquefaction
due to shock or vibration. (Earthquake consideration is not within the realm of this
book.)
The water table may not pose a problem at the time of investigation. But it is
possible that due to local condition changes, the water table will rise or drop to such
an amount that the stability of the structure will be endangered. A thorough inves-
tigation of the site is essential to determine such a possibility.
Throughout the site of a large structure, the depth of the water table may not be
uniform. This is especially true when irrigation ditches or other water-carrying
structures are located in the vicinity of the structures. Part of the footings in the
structure can be affected by the high water condition while others can be free from
the effects of water. Differential settlement of the footings is important and must be
carefully studied.

8.3 RATIONAL DESIGN OF FOOTING FOUNDATION

ON SAND
As discussed above, the stability of footing foundations on sand depends on many
factors. Both from the standpoint of shear and settlement, these factors cannot be
determined with certainty. Since almost all soils existing in nature are not homog-
enous, at a building site the soils vary in both vertical and horizontal directions.
Only in rare cases can the theoretical analysis apply. The following details should
be considered.

1. The basis of most theoretical analyses hinges on the value of the in situ
penetration test value. Extensive research has been done in the past years
to refine, correct, and correlate the field data. A field engineer admits that

©2000 CRC Press LLC

 the blow count data obtained greatly depends on the skill of the operator,
the condition of the sampler, and the depth from which the tests are taken.
Penetration resistance data cannot be treated as a mathematical function
and applied to an equation as in treatment of steel or plastic. This is more
evident when the blow count is below 4 or above 50.
2. A field engineer realizes that the penetration resistance value obtained
can be totally different within a short distance of 10 ft at the same depth.
For a given project, the number of borings and the frequency of the
penetration tests taken are limited and the N value obtained at best can
only give a general idea of the average subsoil condition.
3. The most important factor to be considered by the geotechnical engineers
is the water table level. The groundwater level is important not only at
the time of investigation but also in the future. Fortunately, the perched
water conditions seldom exist in the granular soils.
4. Clean Sand (SW-SP) seldom exists. Most granular soils contain apprecia-
ble amounts of fines. A percentage of silt and clay, as much as 15%, is
commonly encountered. In such cases, the settlement of the subsoil should
be controlled by a consolidation test. Settlement estimates on silty sands
should be treated in the same manner as the consolidation of clays.
5. Homogenous sand strata extending to a great depth seldom exist in nature.
Pockets and layers of soft silt or clay can sometimes be present within
the loaded depth of the footings. Such layers can easily be missed by the
driller or noticed by the field engineer. Consolidation of such strata may
control the settlement of the structure, and calculations of the settlement
of sands become a minor item.
6. When a soft layer of clay is located below the sand strata, settlement of
such a layer can control the behavior of the structure. Compressible clay
layers may be located at a considerable distance below the footings.

8.3.1 TYPICAL DESIGN EXAMPLE

From the detailed analysis of the design criteria to be considered in the design of
foundation on sand, it is up to the geotechnical engineer to choose a rational and
economical recommendation for the project. This can best be illustrated by the
following typical case.

Project: A single-story warehouse structure with

a full live load.
Column load: Varies from 50 to 200 kips.
Footing Depth: 3 ft below ground (below frost depth).
Field data: Average penetration resistance N = 10
(within loaded depth).
Factor of safety: Between two and three.
Allowable Maximum Settlement: 1 in. (differential settlement .75 in.).

©2000 CRC Press LLC

 From the previous discussions, namely:

Penetration resistance and allowable pressure (upper and lower limit).

(Figure 8.6)
Width of footing versus bearing capacity (Meyerhof). (Figure 8.4)
Width of footing versus penetration resistance. (Figure 8.3)
Penetration resistance and allowable pressure (lower limit). (Figure 8.7)

By using N = 10 value, the relationship between width of footing versus allow-

able pressure can be established by using Figures 8.8 and 8.9. Also, for settlement
consideration the relationship between width of footing and bearing capacity can be
established by using Figures 8.6 and 8.7. This is shown in Figures 8.8 and 8.9.
In the upper limit with a footing width less than 2.5 ft and in the lower limit
with a footing width less than 4.5 ft, the design is controlled by the shear consider-
ation. The choice of the use of the upper or the lower limit for the final recommen-
dation depends on the content of fines in the deposit, the water table condition, and
the uniformity of the deposit. These are discussed as follows:

1. If the subsoil contains more than 15% of gravel and cobble and if the
water table is located below the loaded depth of the footings with no
possibility of rising, then the upper limits can be used with confi-
dence.With a column load of 150 kips, the size of the footing should be
on the order of 7 ¥ 7 ft.
2. If the subsoil contains essentially sand with no appreciable amount of
gravel and cobbles, and there is the possibility that the water table may
rise to within the loaded depth of the footing, then the lower limit should
be used. That is, the size of the footing should be on the order of 9 ¥ 9 ft.
3. If other conditions are the same as above, but there is the possibility that
the water table may rise to near the ground surface, then the footing size
given above should be increased by as much as 50%. As an alternative,
a reliable permanent dewatering system or drainage system can be
installed to lower the water table. In such a case, the size of the footings
can be reduced.

Irrespective of all the research and study done on the stability of the footing
foundation on sand, the geotechnical engineer should use his or her past experience,
keen observations, and common sense to achieve a logical and safe foundation
system.

©2000 CRC Press LLC

FIGURE 8.8 Width of footing versus allowable pressure. Upper and lower design limit for
N = 10.

©2000 CRC Press LLC

FIGURE 8.9 Allowable column load and appropriate footing size for upper and lower
limits, with N = 10.

REFERENCES
G. G. Meyerhof, Shallow Foundations, ASCE Journal Soil Mechanics 91, No. SM 2, 1065,
1951.
G. G. Meyerhof, The Ultimate Bearing Capacity of Foundations, Geotec 2, 1951.
K. Terzaghi, R. Peck, and G. Mesri, Soil Mechanics in Engineering Practice, John Wiley-
Interscience Publication, John Wiley & Sons, New York, 1996.

©2000 CRC Press LLC