TM 5-818-7

TECHNICAL MANUAL

FOUNDATIONS
IN
EXPANSIVE SOILS

HEADQUARTERS,

DEPARTMENT

OF

THE ARMY
SEPTEMBER 1983

This manual has been prepared by or for the Government and, except to the extent indicated below, is public property and not subject to copyright.
Copyrighted material included in the manual has been used with the knowledge
and permission of the proprietors and is acknowledged as such at point of use.
Anyone wishing to make further use of any copyrighted material, by itself and
apart from this text, should seek necessary permission directly from the proprietors.
Reprints or republications of this manual should include a credit substantially as
follows: Department of the Army USA, Technical Manual TM 5-818-7, Foundations in Expansive Soils, 1 September 1983.
If the reprint or republication includes copyrighted material, the credit should
also state: Anyone wishing to make further use of copyrighted material, by itself
and apart from this text, should seek necessary permission directly from the proprietors.

By Order of the Secretary of the Army:

I

FOUNDATIONS IN EXPANSIVE SOILS

CHAPTER 1.

2.

3.

4.

5.

6.

7.

8.

9.

APPENDIX A
B
C
D

Figure

INTRODUCTION
Purpose, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Background, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Causes and patterns of heave .. . . . . . . . . . . . . . . . . . . . .
Elements of design. . . . . . . . . . . . . . . . . . . . . . .
RECOGNITION OF PROBLEM AREAS
Site selection, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Hazard maps, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FIELD EXPLORATION
Scope, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Surface examination... . . . . . . . . . . . . . . . . . . . .
Subsurface exploration . . . . . . . . . . . . . . . . . . . .
Groundwater . ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
LABORATORY INVESTIGATIONS
Identification of swelling soils... . . . . . . . . . . . . . . . . . . . .
Testing procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
METHODOLOGY FOR PREDICTION OF VOLUME
CHANGES
Application of heave predictions . . . . . . . . . . . . . . . . . . . . .
Factors influencing heave . . . . . . . . . . . . . . . . . . . . .
Direction of soil movement . . . . . . . . . . . . . . . . . . . . . . . . . .
Potential total vertical heave.., . . . . . . . . . . . . . . . . . . . . .
Potential differential heave...,. . . . . . . . . . . . . . . . . . . . .
Heave with time, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DESIGN OF FOUNDATIONS
Basic considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Shallow individual or continuous footings . . . . . . . . . . . . . .
Reinforced slab-on-grade foundations . . . . . . . . . . . . . . . . .
Deep foundations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
MINIMIZATION OF FOUNDATION MOVEMENT
Preparation for construction. . . . . . . . . . . . . . . . . . . . . .
Drainage techniques, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stabilization techniques. . . . . . . . . . . . . . . . . . . . . . . . .
CONSTRUCTION TECHNIQUES AND INSPECTION
Minimization of foundation problems from construction . . .
Stiffened slab foundations . . . . . . . . . . . . . . . . . . . . . . . . . .
Drilled shaft foundations . . . . . . . . . . . . . . . . . . . . .
REMEDIAL PROCEDURES
Basic considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Evaluation of information . . . . . . . . . . . . . . . . . . . . . . . . . .
Stiffened slab foundations,.... . . . . . . . . . . . . . . . . . . . . .
Drilled shaft foundations. . . . . . . . . . . . . . . . . . . . .
REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CHARACTERIZATION OF SWELL BEHAVIOR FROM SOIL SUCTION .
FRAME AND WALL CONSTRUCTION DETAILS . . . . . . . . . . . . . . . . . . .
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Title
1-1
1-2
1-3
1-4
2-1

TM 5-818-7
Title

Figure

Page

4-1
4-2
5-1
5-3
5-2
5-4
5-6
5-7

5-3
5-4
6-1

6-3
6-5
6-5
6-6

6-2
6-3
6-4
6-5

6-7
6-6
6-8
6-7
6-8
6-8
6-9
6-10
6-14
6-15
6-17
6-18
7-2
7-4
7-5
B-3
B-4
B-5

LIST OF TABLES
Table

Title
1-1
2-1
2-2
3-1
3-2
4-1
4-2
5-1
6-1
6-2
6-3

Page
1-4
2-2
2-4
3-2
3-4
4-1
4-3
5-1
6-2
6-4
6-11

Table

Title

CHAPTER 1
INTRODUCTION
1-1. Purpose

This manual presents guidance and information for

the geotechnical investigation necessary for the selection and design of foundations for heavy and light
military-type buildings constructed in expansive clay
soil areas. The information in this manual is generally
applicable to many types of structures such as residences, warehouses, and multistory buildings. Emphasis is given to the maintenance of an environment that
encourages constant moisture conditions in the
foundation soils during and following construction.
Special attention must always be given to specific requirements of the structure such as limitations on allowable differential movement.
a. The guidance and information provided in this
manual can significantly reduce the risk of undesirable
and severe damages to many structures for numerous
expansive soil conditions. However, complete solutions
for some expansive soil problems are not yet available;
e.g., the depth and amount of future soil moisture
changes may be difficult to predict.
b. This manual presents guidance for selecting eco-

nomical foundations on expansive soil to minimize

structural distress to within tolerable levels and guidance for minimizing problems that may occur in structures on expansive soils.
1-2. Scope
a. Guidelines of the geotechnical investigation and
analysis necessary for selection and design of militarytype buildings constructed in expansive clay soil areas,
as outlined in chapters 2 to 5, consist of methods for
the recognition of the relative magnitude of the swelling soil problem at the construction site, field exploration, laboratory investigations, and application of
methodology for prediction of volume changes in
swelling foundation soils. Chapter 6 presents guidance
for selection of the type of foundation with structural
details of design procedures provided for reference.
Chapters 7 to 9 discuss methods of minimizing foundation movement, construction techniques and inspection, and considerations for remedial repair of damaged structures.
b. Guidance is not specifically provided for design
of highways, canal or reservoir linings, retaining
walls, and hydraulic structures. However, much of the

basic information presented is broadly applicable to

the investigation and analysis of volume changes in
soils supporting these structures and methods for
minimizing potential soil volume changes. Guidance is
also not specifically provided for the design of structures in areas susceptible to soil volume changes from
frost heave and chemical reactions in the soil (e.g., oxidation of iron pyrite), although much of the information presented can be useful toward these designs.
1-3. Background
This manual is concerned with heave or settlement
caused by change in soil moisture in nonfrozen soils.
Foundation materials that exhibit volume change
from change in soil moisture are referred to as expansive or swelling clay soils. Characteristic expansive or
swelling materials are highly plastic clays and clay
shales that often contain colloidal clay minerals such
as the montmorillonites. Expansive soils as used in
this manual also include marls, clayey siltstones, sandstones, and saprolites.
a. Damages from differential movement. The differential movement caused by swell or shrinkage of expansive soils can increase the probability of damage to
the foundation and superstructure. Differential rather
than total movements of the foundation soils are generally responsible for the major structural damage.
Differential movements redistribute the structural
loads causing concentration of loads on portions of the
foundation and large changes in moments and shear
forces in the structure not previously accounted for in
standard design practice.
b. Occurrence of damages. Damages can occur within a few months following construction, may develop
slowly over a period of about 5 years, or may not appear for many years until some activity occurs to disturb the soil moisture. The probability of damages increases for structures on swelling foundation soils if
the climate and other field environment, effects of
construction, and effects of occupancy tend to promote
moisture changes in the soil.
c. Structures susceptible to damages. Types of
structures most often damaged from swelling soil
include foundations and walls of residential and light
(one- or two-story) buildings, highways, canal and
reservoir linings, and retaining walls. Lightly loaded
1-1

TM 5-818-7

one- or two-story buildings, warehouses, residences,

and pavements are especially vulnerable to damage because these structures are less able to suppress the differential heave of the swelling foundation soil than
heavy, multistory structures.
(1) Type of damages. Damages sustained by these
structures include: distortion and cracking of pavements and on-grade floor slabs; cracks in grade beams,
walls, and drilled shafts; jammed or misaligned doors
and windows; and failure of steel or concrete plinths
(or blocks) supporting grade beams. Lateral forces may
lead to buckling of basement and retaining walls, particularly in overconsolidated and nonfissured soils.
The magnitude of damages to structures can be extensive, impair the usefulness of the structure, and detract aesthetically from the environment. Maintenance and repair requirements can be extensive, and
the expenses can grossly exceed the original cost of the
foundation.
(2) Example of damages. Figure 1-1 illustrates
damages to a building constructed on expansive soil
with a deep water table in the wet, humid climate of
Clinton, Mississippi. These damages are typical of
buildings on expansive soils. The foundation consists
of grade beams on deep drilled shafts. Voids were not
provided beneath the grade beams above the expansive
foundation soil, and joints were not made in the walls
and grade beams. The floor slab was poured on-grade
with no provision to accommodate differential movement between the slab and grade beams. The heave of
the floor slab exceeded 6 inches. The differential soil
movement and lack of construction joints in the structure aggravated cracking.
14. Causes and patterns of heave
a. Causes. The leading cause of foundation heave or
settlement in susceptible soils is change in soil moisture, which is attributed to changes in the field environment from natural conditions, changes related to
construction, and usage effects on the moisture under
the structure (table 1-1). Differential heave may be
caused by nonuniform changes in soil moisture, variations in thickness and composition of the expansive
foundation soil, nonuniform structural loads, and the
geometry of the structure. Nonuniform moisture
changes occur from most of the items given in table
1-1.
b. Patterns of heave.
(1) Doming heave. Heave of foundations, although

often erratic, can occur with an upward, long-term,

dome-shaped movement that develops over many
years. Movement that follows a reduction of natural
evapotranspiration is commonly associated with a
doming pattern of greatest heave toward the center of
the structure. Evapotranspiration refers to the evaporation of moisture from the ground surface and trans1-2

piration of moisture from heavy vegetation into the atmosphere. Figure 1-2 schematically illustrates some
commonly observed exterior cracks in brick walls from
doming or edgedown patterns of heave. The pattern of
heave generally causes the external walls in the superstructure to lean outward, resulting in horizontal,
vertical, and diagonal fractures with larger cracks
near the top. The roof tends to restrain the rotation
from vertical differential movements leading to additional horizontal fractures near the roofline at the top
of the wall. Semiarid, hot, and dry climates and deep
water tables can be more conducive to severe and progressive foundation soil heaves if water become available.
(2) Cyclic heave. A cyclic expansion-contraction
related to drainage and the frequency and amount of
rainfall and evapotranspiration may be superimposed
on long-term heave near the perimeter of the structure. Localized heaving may occur near water leaks or
ponded areas. Downwarping from soil shrinkage (fig.
1-2) may develop beneath the perimeter during hot,
dry periods or from the desiccating influence of trees
and vegetation located adjacent to the structure. These
edge effects may extend inward as much as 8 to 10
feet. They become less significant on well-drained
land. Heavy rain periods may cause pending adjacent
to the structure with edge lift (fig. 1-3) and reversal of
the downwarping.
(3) Edge heave. Damaging edge or dish-shaped
heaving (fig. 1-3) of portions of the perimeter maybe
observed relatively soon after construction, particularly in semiarid climates on construction sites with
preconstruction vegetation and lack of topographic relief. The removal of vegetation leads to an increase in
soil moisture, while the absence of topographic relief
leads to ponding (table 1-1). A dish-shaped pattern can
also occur beneath foundations because of consolidation, drying out of surface soil from heat sources, or
sometimes lowering of the water table. Changes in the
water table level in uniform soils beneath uniformly
loaded structures may not contribute to differential
heave. However, structures on a deep foundation, such
as drilled shafts with a slab-on-grade, can be adversely
affected by a changing water table or changes in soil
moisture if the slab is not isolated from the perimeter
grade beams and if internal walls and equipment are
not designed to accommodate the slab movement.
(4) Lateral movement. Lateral movement may affect the integrity of the structure.
(a) Lateral thrust of expansive soil with a horizontal force up to the passive earth pressure can cause
bulging and fracture of basement walls. Basement
walls and walls supporting buildings usually cannot
tolerate the same amount of movement as free-standing retaining walls. Consequently, such walls must be
designed to a higher degree of stability.

a.

Vertical

cracks

1-5. Elements of design

the structure are simply supported on-grade or attached to the structure, they can contribute to future
maintenance problems.
(2) Potential problems that could eventually affect the performance of the structure are best determined during the predesign and preliminary design
phases when compromises can be made between the
structural, architectural, mechanical, and other aspects of the design without disrupting the design process. Changes during the detailed design phase or during construction will probably delay construction and
pose economic disadvantages.

1-5

TM 5-818-7

CHAPTER 2

RECOGNITION OF PROBLEM AREAS

2-1. Site selection
The choice of the construction site is often limited. It
is important to recognize the existence of swelling soils
on potential sites and to understand the problems that
can occur with these soils as early as possible. A surface examination of the potential site as discussed in
paragraph 3-2 should be conducted and available soil
data studied during the site selection.
a. Avoidance of potential problems. If practical,
the foundation should be located on uniform soils subject to the least swelling or volume change. Discontinuities or significant lateral variations in the soil
strata should be avoided. Swampy areas, backfilled
ponds, and areas near trees and other heavy vegetation
should be avoided, Special attention should be given to
adequate compaction of filled areas, types of fill, and
leveling of sloped sites (para 7-1).
(1) Undeveloped sites. Undeveloped sites generally have little or no subsurface soil information available and require subsurface exploration (para 3-3).
(a) Substantial differential heave may occur beneath structures constructed on previously undeveloped sites where trees and other heavy vegetation had
been removed prior to construction, Soil moisture will
tend to increase since loss of heavy vegetation reduces
the transpiration of moisture. Construction of the
foundation over the soil will tend to further increase
soil moisture because of reduced evaporation of moisture from the ground surface.
(b) Swampy or ponded areas may contain greater quantities of plastic fine particles with a greater
tendency to swell than other areas on the site.
(c) Future irrigation of landscaped areas and
leakage from future sewer and other water utility lines
following development of the site may substantially
increase soil moisture and cause a water table to rise or
to develop if one had not previously existed. Filled
areas may also settle if not properly compacted.
(2) Developed sites. Subsurface exploration
should be conducted if sufficient soil data from earlier
borings are not available for the site selection and/or
problems had occurred with previous structures. Some
subsurface exploration is always necessary for site selection of any structure of economic significance, particularly multistory buildings and structures with special requirements of limited differential distortion.
(a) An advantage of construction on developed

sites is the experience gained from previous construction and observation of successful or unsuccessful past
performance. Local builders should be consulted to obtain their experience in areas near the site. Existing
structures should be observed to provide hints of problem soil areas.
(b) The soil moisture may tend to be much closer
to an equilibrium profile than that of an undeveloped
site. Differential movement may not be a problem because previous irrigation, leaking underground water
lines, and previous foundations on the site may have
stabilized the soil moisture toward an equilibrium profile. Significant differential movement, however, is
still possible if new construction leads to changes in
soil moisture. For example, trees or shrubs planted too
close to the structure or trees removed from the site,
change in the previous irrigation pattern following
construction, lack of adequate drainage from the structure, and improper maintenance of drainage provisions may lead to localized changes in soil moisture
and differential heave. Edge movement of slab-ongrade foundations from seasonal changes in climate
may continue to be a problem and should be minimized
as discussed in chapter 7.
(3) Sidehill or sloped sites. Structures constructed on sites in which the topography relief is greater
than 5 degrees (9 percent gradient) may sustain damage from downhill creep of expansive clay surface soil.
Sidehill sites and sites requiring split-level construction can, therefore, be expected to complicate the design. See chapter 7 for details on minimization of foundation soil movement.
b. Soil surveys, Among the best methods available
for qualitatively recognizing the extent of the swelling
soil problem for the selected site is a careful examination of all available documented evidence on soil conditions near the vicinity of the site. Local geological records and publications and federal, state, and institutional surveys provide good sources of information on
subsurface soil features. Hazard maps described in
paragraph 2-2 document surveys available for estimating the extent of swelling soil problem areas.

2-2. Hazard maps

Hazard maps provide a useful first-order approximation of and guide to the distribution and relative expansiveness of problem soils. These maps should be
2-1

TM 5-818-7

used in conjunction with local experience and locally

available soil surveys and boring data. The maps discussed in a and b below are generally consistent with
each other and tend to delineate similar areas of
moderately or highly expansive soil.
a. Waterways Experiment Station (WES) Map. This
map, which was prepared for the Federal Highway Administration (FHWA), summarizes the areas of the
United States, except Alaska and Hawaii, where swelling soil problems are likely to occur (fig. 2-1). The basis for classification depends primarily on the estimated volume change of argillaceous or clayey materials within the geologic unit, the presence of montmorillonite, the geologic age, and reported problems

due to expansive materials. The stratigraphy and mineralogy are key elements in the classification.
(1) Classification. The soils are classified into
categories of High, Medium, Low, and Nonexpansive
as shown in figure 2-1. The distribution of expansive
materials is categorized by the geologic unit on the basis of the degree of expansiveness that relates to the
expected presence of montmorillonite and the frequency of occurrence that relates to the amount of clay
or shale. The amount refers most significantly to the
vertical thickness of the geologic unit, but the areal extent was also considered in the classification. The
premises in table 2-1 guide the categorization of soils.

(2) Physiographic provinces. Table 2-2 summarizes the potentially expansive geologic units on the basis of the 20 first-order physiographic provinces. Figure 2-1 shows the locations of the physiographic provinces.

(2) Assessment map of expansive soils within the

United States. The major categories for classification

b. Other maps.
(1) Area map of susceptible soil expansion problems. A hazard map was developed by M, W. Witczak

(Transportation Research Board, Report 132) on the

basis of the occurrence and distribution of expansive
soils and expansive geologic units, the pedologic analysis, and climatic data to delineate areas susceptible to
expansion problems. Some geologic units for which
engineering experiences were not available may have
been omitted, and the significance of pedological soil
on expansion was not shown on the map.
2-2

of the severity of the swelling soil problem presented

by J. P. Krohn and J. E. Slosson (American Society of
Civil Engineers, Proceedings of the Fourth International Conference on Expansive Soils, Volume 1 (see
app. A) correspond to the following modified shrinkswell categories of the Soil Conservation Service (SCS)
of the U. S. Department of Agriculture:
High:
Moderate:

Low:

Soils containing large amounts of montmorillonite

and clay (COLE >6 percent)
Soils containing moderate amounts of clay with

Soils containing some clay with the clay consisting mostly of kaolinite and/or other low swelling
clay minerals (COLE <3 percent).

These categories of classification use the coefficient of

linear extensibility (COLE), which is a measure of the
change in linear dimension from the dry to a moist
state, and it is related to the cube root of the volume
change. Premises guiding the categorization of the
Krohn and Slosson map include: degree of expansion
as a function of the amount of expandable clay; cover
of nonexpansive glacial deposits; and low-rated areas
with nonexpansive and small quantities of expansive
soils. Environmental factors, such as climatic effects,
vegetation, drainage, and effects of man, were not considered.
(3) Soil Conservation Service county soil surveys.
Survey maps by SCS provide the most detailed surficial soil maps available, but not all of the United
States is mapped. Soil surveys completed during the
1970s contain engineering test data, estimates of soil
engineering properties, and interpretations of properties for each of the major soil series within the given
county. The maps usually treat only the upper 30 to 60
inches of soil and, therefore, may not fully define the
foundation soil problem.
(4) U.S. and State Geological Survey maps. The
U.S. Geological Survey is currently preparing hazard
maps that will include expansive soils.
c. Application of hazard maps. Hazard maps provide
basic information indicative of the probable degree of
expansiveness and/or frequency of occurrence of swelling soils. These data lead to initial estimates for the location and relative magnitude of the swelling problem
to be expected from the foundation soils. The SCS

count y survey maps prepared after 1970, if available,

provide more detail on surface soils than do the other
maps discussed in b above. The other maps used in conjunction with the SCS maps provide a better basis for
election of the construction site.
(1) Recognition of the problem area at the construction site provides an aid for the planning of field exploration that will lead to the determination of the
areal extent of the swelling soil formations and samples for the positive identification and evaluation of
potential swell of the foundation soils and probable
soil movements beneath the structure.
(2) Problem areas that rate highly or moderately
expansive on any of the hazard maps should be explored to investigate the extent and nature of the
swelling soils. Structures in even low-rated areas of potential swell may also be susceptible to damages from
heaving soil depending on the ability of the structure
to tolerate differential foundation movement. These
low-rated areas can exhibit significant differential soil
heave if construction leads to sufficiently large
changes in soil moisture and uneven distribution of
loads. Also, low-rated areas on hazard maps may include some highly swelling soil that had been neglected.
(3) Figure 2-1 indicates that most problems with
swelling soils can be expected in the northern central,
central, and southern states of the continental United
States. The Aliamanu crater region of Fort Shafter,
Hawaii, is another example of a problem area.

TM 5-818-7

CHAPTER 3
FIELD EXPLORATION
3-1. Scope
The field study is used to determine the presence, extent, and nature of expansive soil and groundwater
conditions. The two major phases of field exploration
are surface examination and subsurface exploration.
The surface examination is conducted first since the
results help to determine the extent of the subsurface
exploration. In situ tests may also be helpful, particularly if a deep foundation, such as drilled shafts, is to
be used.
3-2. Surface examination
a. Site history. A study of the site history may reveal considerable qualitative data on the probable future behavior of the foundation soils. Maps of the proposed construction site should be examined to obtain
information on wooded areas, ponds and depressions,
water-courses, and existence of earlier buildings. Surface features, such as wooded areas, bushes, and other
deep-rooted vegetation in expansive soil areas, indicate potential heave from accumulation of moisture
following elimination of these sources of evapotranspiration. The growth of mesquite trees, such as found
in Texas, and other small trees may indicate subsurface soil with a high affinity for moisture, a characteristic of expansive soil. Ponds and depressions are often
filled with clayey, expansive sediments accumulated
from runoff. The existence of earlier structures on or
near the construction site has probably modified the
soil moisture profile and will influence the potential
for future heave beneath new structures.
b. Field reconnaissance. A thorough visual examina-

tion of the site by the geotechnical engineer is necessary (table 3-1). More extensive subsurface exploration is indicated if a potential for swelling soil is evident from damages observed in nearby structures. The
extent of desiccation cracks, plasticity, slickensides,
and textures of the surface soil can provide a relative
indication of the potential for damaging swell.
(1) Cracking in nearby structures. The appearance
of cracking in nearby structures should be especially
noted. The condition of on-site stucco facing, joints of
brick and stone structures, and interior plaster walls
can be a fair indication of the possible degree of swelling that has occurred. The differential heave that may
occur in the foundation soil beneath the proposed
structure. however, is not necessarily equal to the dif-

ferential heave of earlier or nearby structures. Differential heave depends on conditions such as variation of
soils beneath the structure, load distribution on the
foundation, foundation depth, and changes in groundwater since construction of the earlier structures.
(2) Soil gilgai. The surface soil at the site should
also be examined for gilgai. Soil gilgai are surface
mounds that form at locations where the subsurface
soil has a greater percentage of plastic fines and is
thus more expansive than the surface soil. Gilgai begin
to form at locations where vertical cracks penetrate
into the subsurface soil. Surface water enters and
swelling takes place around the cracks leaving fractured zones where plastic flow occurs. These mounds
usually have a higher pH than the adjacent low areas
or depressions and may indicate subsurface soil that
had extruded up the fractures.
(3) Site access and mobility. Indicators of site access and mobility (table 3-1) may also influence behavior of the completed structure. For example, nearby
water and sewer lines may alter the natural moisture
environment. Flat land with poor surface drainage, as
indicated by ponded water, may aggravate differential
heave of the completed structure if drainage is not corrected during construction. Construction on land with
slopes greater than 5 degrees may lead to structural
damage from creep of expansive clay surface soils.
Trees located within a distance of the proposed structure of 1 to 1.5 times the height of mature trees may
lead to shrinkage beneath the structure, particularly
during droughts.
c. Local design and construction experience. Local

experience is very helpful in indicating possible design

and construction problems and soil and groundwater
conditions at the site. Past successful methods of design and construction and recent innovations should
be examined to evaluate their usefulness for the proposed structure.
3-3. Subsurface exploration
Subsurface exploration provides representative samples for visual classification and laboratory tests. Classification tests are used to determine the lateral and
vertical distribution and types of foundation soils. Soil
swell, consolidation, and strength tests are needed to
evaluate the load/displacement behavior and bearing
capacity of the foundation in swelling soil. The struc3-1

ture interaction effects in swelling soil are complicated

by the foundation differential movement caused by
soil heave. Sufficient samples should be available to allow determination of the representative mean of the
swell and strength parameters of each distinctive soil
stratum. The lower limit of the scatter in strength
parameters should also be noted.
a. Sampling requirements. The design of lightly
loaded structures and residences can often be made
with minimal additional subsurface investigations and
soil testing if the site is developed, if subsurface features are generally known, and if the local practice has
consistently provided successful and economical designs of comparable structures. Additional subsurface
investigation is required for new undeveloped sites,
multistory or heavy buildings, structures with previously untested or new types of foundations, and special structures that require unusually limited differential movements of the foundation such as deflection/length ratios less than 1/1000. Where the local
practice has not consistently provided satisfactory designs, a careful review of the local practice is neces3-2

sary. Corrections to improve performance compared

with earlier structures may prove difficult to devise
and implement and may require evaluation of the behavior of the subsurface foundation soils and groundwater conditions.
b. Distribution and depth of borings. The distribution and depth of borings are chosen to determine the
soil profile and to obtain undisturbed samples required
to evaluate the potential total and differential heave of
the foundation soils from laboratory swell tests, as
well as to determine the bearing capacity and settlement. Consequently, greater quantities of undisturbed
samples may be required in swelling soils than normally needed for strength tests.
(1) Borings should be spaced to define the geology
and soil nonconformities. Spacings of 50 or 25 feet and
occasionally to even less distance may be required
when erratic subsurface conditions (e.g., soils of different swelling potential, bearing capacity, or settlement)
are encountered. Initial borings should be located close
to the corners of the foundation, and the number
should not be less than three unless subsurface condi-

tions are known to be uniform. Additional borings

should be made as required by the extent of the area,
the location of deep foundations such as drilled shafts,
and the encountered soil conditions.
(2) The depth of sampling should be at least as
deep as the probable depth to which moisture changes
and heave may occur. This depth is called the depth of
down about 10 to 20 feet below the base of the foundation or to the depth of shallow water tables, but it may
be deeper (para 5-4c). A shallow water table is defined
as less than 20 feet below the ground surface or below
the base of the proposed foundation. The entire thickness of intensely jointed or fissured clays and shales
should be sampled until the groundwater level is encountered because the entire zone could swell, provided swelling pressures are sufficiently high, when given
access to moisture. Continuous sampling is required
for the depth range within the active zone for heave.
(3) Sampling should extend well below the anticipated base of the foundation and into strata of adequate bearing capacity. In general, sampling should
continue down to depths of 1.5 times the minimum
width of slab foundations to a maximum of 100 feet
and a minimum of three base diameters beneath the
base of shaft foundations. The presence of a weak,
compressible, or expansive stratum within the stress
field exerted by the entire foundation should be detected and analyzed to avoid unexpected differential
movement caused by long-term volume changes in this
stratum. Sampling should continue at least 20 feet beneath the base of the proposed foundation. Determination of the shear strength and stress/strain behavior
of each soil stratum down to depths of approximately
100 feet below the foundation is useful if numerical
analysis by the finite element method is considered.
c. Time of sampling. Sampling may be done when
soil moisture is expected to be similar to that during
construction. However, a design that must be adequate
for severe changes in climate, such as exposure to periods of drought and heavy rainfall, should be based on
maximum levels of potential soil heave. Maximum potential heaves are determined from swell tests using
soils sampled near the end of the dry season, which often occurs toward the end of summer or early fall.
Heave of the foundation soil tends to be less if samples
are taken or the foundation is placed following the wet
season, which often occurs during spring.
d. Sampling techniques. The disturbed samples and

the relatively undisturbed samples that provide minimal disturbance suitable for certain laboratory soil
tests may be obtained by the methods described in table 3-2. Drilling equipment should be well maintained
during sampling to avoid equipment failures, which
cause delays and can contribute to sample disturbance.

Personnel should be well trained to expedite proper

sampling, sealing, and storage in sample containers.
(1) Disturbed sampling. Disturbed auger, pit, or
split spoon samplers may be useful to roughly identify
the soil for qualitative estimates of the potential for
soil volume change (para 4-1). The water content of
these samples should not be altered artificially during
boring, for example, by pouring water down the hole
during augering.
(2) Undisturbed sampling. Minimization of sample disturbance during and after drilling is important
to the usefulness of undisturbed samples. This fact is
particularly true for expansive soils since small
changes in water content or soil structure will significantly affect the measured swelling properties.
(a) The sample should be taken as soon as possible, after advancing the hole to the proper depth and
cleaning out the hole, to minimize swelling or plastic
deformation of the soil to be sampled.
(b) The samples should be obtained using a push
tube sampler without drilling fluid, if possible, to
minimize changes in the sample water content. Drilling fluids tend to increase the natural water content
near the perimeter of the soil sample, particularly for
fissured soil.
(c) A piston Denisen or other sampler with a
cutting edge that precedes the rotating outer tube into
the formation is preferred, if drilling fluid is necessary, to minimize contamination of the soil sample by
the fluid.
e. Storage of samples. Samples should be immediately processed and sealed following removal from the
boring hole to minimize changes in water content.
Each container should be clearly labeled and stored under conditions that minimize large temperature and
humidity variations. A humid room with relative
humidity greater than 95 percent is recommended for
storage since the relative humidity of most natural
soils exceeds 95 percent.
(1) Disturbed samples. Auger, pit, or other disturbed samples should be thoroughly sealed in waterproof containers so that the natural water content can
be accurately measured.
(2) Undisturbed samples. Undisturbed samples
may be stored in the sampling tubes or extruded and
preserved, then stored. Storage in the sampling tube is
not recommended for swelling soils even though stress
relief may be minimal, The influence of rust and penetration of drilling fluid or free water into the sample
during sampling may adversely influence the laboratory test results and reduce the indicated potential
heave. Iron diffusing from steel tubes into the soil
sample will combine with oxygen and water to form
rust. Slight changes in Atterberg limits, erosion resistance, water content, and other physical properties may
occur. In addition, the outer perimeter of a soil sample
3-3

TM 5-818-7

stored in the sampling tube cannot be scraped to remove soil contaminated by water that may have penetrated into the perimeter of the sample during sampling. The sample may also later adhere to the tube
wall because of rust. If samples are stored in tubes, the
tubes should be brass or lacquered inside to inhibit corrosion. An expanding packer with a rubber O-ring in
both ends of the tube should be used to minimize moisture loss. The following procedures should be followed
in the care and storage of extruded samples.
(a) Expansive soil samples that are to be extrubed and stored should be removed from the sampling tubes immediately after sampling and thoroughly sealed to minimize further stress relief and moisture
loss. The sample should be extruded from the sampling
tube in the same direction when sampled to minimize
further sample disturbance.
(b) Samples extruded from tubes that were obtained with slurry drilling techniques should be wiped
clean to remove drilling fluid adhering to the surface
of the sample prior to sealing in the storage containers. An outer layer of 1/8 to 1/4 inch should be
trimmed from the cylindrical surface of the samples so
that moisture from the slurry will not penetrate into
the sample and alter the soil swelling potential and
strength. Trimming will also remove some disturbance
at the perimeter due to sidewall friction. The outer
perimeter of the soil sample should also be trimmed
away during preparation of specimens for laboratory
tests.
(c) Containers for storage of extruded samples
may be either cardboard or metal and should be
approximately 1 inch greater in diameter and 1.5 to 2
inches greater in length than the sample to be encased.
Three-ply, wax-coated cardboard tubes with metal bottoms are available in various diameters and lengths
and may be cut to desired lengths.
(d) Soil samples preserved in cardboard tubes
should be completely sealed in wax. The wax and cardboard containers provide an excellent seal against
moisture loss and give sufficient confinement to minimize stress relief and particle reorientation. A good
wax for sealing expansive soils consists of a 1 to 1 mixture of paraffin and microcrystalline wax or 100 percent beeswax. These mixtures adequately seal the sample and do not become brittle when cold. The temperature of the wax should be approximately 20 degrees
Fahrenheit above the melting point when applied to
the soil sample, since wax that is too hot will penetrate
pores and cracks in the sample and render it useless, as
well as dry the sample. Aluminum foil or plastic wrap
may be placed around the sample to prevent penetration of molten wax into open fissures. A small amount
of wax (about 0.5-inch thickness) should be placed in
the bottom of the tube and allowed to partly congeal.
The sample should subsequently be placed in the tube,

completely immersed and covered with the molten

wax, and then allowed to cool before moving.
(e) When the samples are being transported,
they should be protected from rough rides and bumps
to minimize further sample disturbance.
f. Inspection. A competent inspector or engineer
should accurately and visually classify materials as
they are recovered from the boring. Adequate classification ensures the proper selection of samples for laboratory tests. A qualified engineering geologist or
foundation engineer should closely monitor the drill
crew so that timely adjustments can be made during
drilling to obtain the best and most representative
samples. The inspector should also see that all open
borehoes are filled and sealed with a proper grout,
such as a mixture of 12 percent bentonite and 88 percent cement, to minimize penetration of surface water
or water from a perched water table into deeper strata
that might include moisture deficient expansive clays.
3-4. Groundwater
Meaningful groundwater conditions and engineering
properties of subsurface materials can often best be
determined from in situ tests. In situ tests, however,
are not always amenable to simple interpretation. The
pore water conditions at the time of the test may differ
appreciably from those existing at the time of construction. A knowledge of groundwater and the negative pore water pressure are important in evaluating
the behavior of a foundation, particularly in expansive
soil. Every effort should be made to determine the position of the groundwater level, its seasonal variation,
and the effect of tides, adjacent rivers, or canals on it.
a. Measurement of groundwater level. The most reliable and frequently the only satisfactory method for
determining groundwater levels and positive pore
water pressures is by piezometers with tips installed at
different depths. Ceramic porous tube piezometers
with small diameters (3/8-inch) risers are usually adequate, and they are relatively simple, inexpensive, and
sufficient for soils of low permeability.
b. Measurement of in situ negative pore water pressure, Successful in situ measurements of negative pore

water pressure and soil suction have been performed

by such devices as tensiometers, negative pore pressure piezometers, gypsum blocks, and thermocouple
psychrometer. However, each of these devices has
certain limitations, The range of tensiometers and
negative pore pressure piezometers has been limited to
the cavitation stress of water under normal conditions,
which is near one atmosphere of negative pressure.
The fluid-filled tensiometer is restricted to shallow
soils less than 6 feet in depth. The useable range of the
tensiometer is reduced in proportion to the pressure
exerted by the column of fluid in the tensiometer. Gyp3-5

sum blocks require tedious calibration of electrical

resistivity for each soil and dissolved salts greatly influence the results. Thermocouple psychrometer cannot measure soil suctions reliably at negative pressures that are less than one atmosphere and require a
constant temperature environment. Psychrometer
also measure the total suction that includes an osmotic
component caused by soluble salts in the pore water, as
well as the matrix suction that is comparable with the

3-6

negative pore water pressure. Tensiometers require

constant maintenance, while gypsum blocks and psychrometer tend to deteriorate with time and may become inoperable within one year. A routine field measurement of soil suction is not presently recommended
because of the limitations associated with these devices. Alternatively, laboratory measurements of soil
suction can be easily performed (para 4-2a).

CHAPTER 4
LABORATORY INVESTIGATIONS
4-1. Identification of swelling soils
Soils susceptible to swelling can be identified by classification tests. These identification procedures were developed by correlations of classification test results
with results of one-dimensional swell tests performed
in consolidometers on undisturbed and compacted soil
specimens. Classification data most useful for identifying the relative swell potential include the liquid
limit (LL), the plasticity index (PI), the COLE (para
chemical tests. Several of the more simple and successful methods recommended for identifying swelling soil
from classification tests described below were developed from selected soils and locations combined with
the results of limited field observations of heave.
These procedures assume certain environmental conditions for surcharge pressure (e.g., 1 pound per square
inch) and changes in moisture from the initial water
content (e.g., to saturation or zero final pore water
pressure),

were performed on 20 undisturbed clays and clay

shales from the states of Mississippi, Louisiana, Texas,
Oklahoma, Arizona, Utah, Kansas, Colorado, Wyoming, Montana, and South Dakota. Results of these
tests for a change in moisture from natural water content to saturation at the estimated in situ overburden
pressure (pressures corresponding to depths from 1 to
8 feet) indicated the degrees of expansion and potensents the percent increase in the vertical dimension or
the percent potential vertical heave. The classification
may be used without knowing the natural soil suction,
but the accuracy and conservatism of the system are
reduced. Soils that rate low may not require further
swell tests, particularly if the LL is less than 40 percent and the PI is less than 15 percent. Soils with these
Atterberg limits or less are essentially nonexpansive.
However, swell tests may be required for soils of low
swelling potential if the foundation of the structure is
required to maintain small differential movements
less than 1 inch (para 4-2c).

a. WES classification. Consolidometer swell tests

b. Texas Department of Highways and Public

Transportation (TDHPT) method. This procedure,

which is known as Tex-124-E of the TDHPT Manual

of Testing Procedures, is based on the swell test results
of compacted soils from Texas. Field heaves of each
soil stratum in the profile are estimated from a family
of curves using the LL, PI, surcharge pressure on the
soil stratum, and initial water content. The initial water content is compared with maximum (0.47 LL + 2)
and minimum (0.2 LL + 9) water contents to evaluate
the percent volumetric change. The potential vertical
rise (PVR) of each stratum is found from a chart using
the percent volumetric change and the unit load bearing on the stratum. These PVRs for depths of as much
as 30 feet or more are summed to evaluate the total
PVR. This method may overestimate the heave of low
plasticity soils and underestimate the heave of high
plasticity soils.

c. Van Der Merwe method. This method evolved

from empirical relationships between the degree of expansion, the PI, the percent clay fraction, and the surcharge pressure, The total heave at the ground surface
is found from

where
AH =
D=

PE =

and 1 inch/foot for low, medium, high, and

. very high

levels, respectively, of potential expansiveness, defined in figure 4-1 as functions of the PI and the mi-

cal direction, and they require an estimate of the depth

Van Der Merwe methods do not require estimates of

meter swell test results and field observations. This

method does not consider variations in initial moisture
conditions.
d. Physiochemical tests. These tests include identification of the clay minerals, such as montmorillonite, illite, attapulgite, and kaolinite, with kaolinite
being relatively nonexpansive, cation exchange capacity (CEC), and dissolved salts in the pore water. The
CEC is a measure of the property of a clay mineral to
exchange ions for other anions or cations by treatment
in an aqueous solution. The relatively expansive montmorillonite minerals tend to have large CEC exceeding
80 milliequivalents per 100 grams of clay, whereas the
CEC of nonexpansive kaolinite is usually less than 15
milliequivalents. The presence of dissolved salts in the
pore water produces an osmotic component of soil suction that can influence soil heave if the concentration
of dissolved salts is altered. In most cases, the osmotic
suction will remain constant and not normally influence heave unless, for example, significant leaching of
the soil occurs.
e. Other methods. Other methods that have been
successful are presented in table 4-2. These methods

heave assuming that all swell is confined to the verti-

the computed heaves become negligible. The Van Der

Merwe, McKeen-Lytton, and Johnson methods tend to
give maximum values or may overestimate heave,
whereas the remaining methods tend to give minimum
values or may underestimate heave when compared
with the results of field observations at three WES
test sections.
f. Application. These identification tests along with
the surface examination of paragraph 3-2 can indicate
problem soils that should be tested further and can
provide a helpful first estimate of the expected in situ
heave.
(1) More than one identification test should be
used to provide rough estimates of the potential heave
because limits of applicability of these tests are not
known. In general, estimates of potential heave at the
ground surface of more than 1/2 inch may require further laboratory tests, particularly if local experience
suggests swelling soil problems. Soil strata in which
the degree of expansion is medium or high should also
be considered for further swell tests (para 2-2c).
(2) The McKeen-Lytton method of table 4-2 has
been applied to the prediction of potential differential
heave for average changes in moisture conditions by
the Post-Tensioning Institute (PTI) for design and con-

struction of stiffened slabs-on-grade in expansive soils.

The PTI structural design procedure is described in
paragraph 6-3b.
4-2. Testing procedures

Quantitative characterization of the expansive soil

from swell tests is necessary to predict the anticipated

potential soil heave devaluation of swell behavior and

predictions of total and differential heave are determined from the results of tests on undisturbed specimens. Strength tests may be performed to estimate
the bearing capacity of the foundation soil at the final
or equilibrium water content. A measure of shear
strength with depth is also needed to evaluate soil sup4-3

TM 5-818-7

port from adhesion along the perimeter of shaft

foundations or the uplift that develops on the shaft
when swelling occurs.
a. Swell tests. Laboratory methods recommended
for prediction of the anticipated volume change or potential in situ heave of foundation soils are consolidometer swell and soil suction tests, The WES expansive soil studies show that consolidometer swell tests
may underestimate heave, whereas soil suction tests
may overestimate heave compared with heaves measured in the field if a saturated final moisture profile is
assumed (chap 5). The economy and simplicity of soil
suction tests permit these tests to be performed at frequent intervals of depth from 1 to 2 feet.
(1) Consolidometer. Recommended consolidometer swell tests include swell and swell pressure tests
described in Appendix VIII of EM 1110-2-1906. The
swell test may be performed to predict vertical heave
AH of soil thickness H when the vertical overburden
and structural pressures on thickness H are known
prior to the test. The total vertical heave at the ground
surface is the sum of the AH for each thickness H in
the soil profile. Figure 5-4 illustrates the application
of swell test data. The swell pressure test is performed
quired for prediction of vertical heave by equation

(5-8) discussed in paragraph 5-4e. The confining preslittle is known about swell behavior or groundwater
conditions, an appropriate swell testis given in (a) and
(b) below.
(a) An initial loading pressure, simulating field
initial (preconstruction) vertical pressure &, should be
applied to determine the initial void ratio e., point 1 of
(i.e., the lowest possible load) prior to adding distilled
water, point 2. The specimen is allowed to expand at
the seating pressure until primary swell is complete,
point 3, before applying the consolidation pressures.
(b) The swell test of figure 4-2 can eliminate
the need for additional tests when behavior is different than that anticipated (e.g., the specimen consolidates rather than swells following addition of water at
loading pressures greater than the seating pressure).
The void ratio-log pressure curve for final effective
pressures, varying from the seating to the maximum
applied pressure, can be used to determine heave or
settlements will occur for final effective pressures ex-

with respect to the initial vertical pressure&.

TM 5-818-7

pressure that must be applied to the soil to reduce the

volume expansion down to the (approximated) in situ
in appendix VIII of EM 1110-2-1906 tend to provide
lower limits of the in situ swell pressure, while the
simple swell test, figure 4-2, tends to provide upper
limits. The maximum past pressure is often a useful
(2) Soil suction. Soil suction is a quantity that also
can be used to characterize the effect of moisture on
volume changes and, therefore, to determine the
anticipated foundation soil heave. The suction is a tensile stress exerted on the soil water by the soil mass
that pulls the mass together and thus contributes to
the apparent cohesion and undrained shear strength of
the soil. The thermocouple psychrometer and filter
paper methods, two of the simplest approaches for
evaluation of soil suction and characterization of swelling behavior, are described in appendix B. The suction
procedure, which is analogous to the procedure for
characterization of swell from consolidometer swell
tests, is relatively fast, and the results can increase
confidence in characterization of swell behavior.
b. Strength tests. The results of strength tests are
used to estimate the soil bearing capacity and load/deflection behavior of shaft or other foundations. The
critical time for bearing capacity in many cases is
immediately after completion of construction (first
loading) and prior to any significant soil consolidation
under the loads carried by the foundation. The longterm bearing capacity may also be critical in expansive
foundation soils because of reductions in strength
from wetting of the soil.
c. Application. Sufficient numbers of swell and

strength tests should be performed to characterize the

soil profiles. Swell tests may not be necessary on specimens taken at depths below permanent deep groundwater levels.
(1) The representative mean of the swell and
strength parameters (and lower limit of the scatter in
strength parameters) of each distinctive soil stratum
should be determined down to depths of 1.5 times the
minimum width of mat slabs to a maximum of 100
feet and to at least three base diameters beneath the
base of shaft foundations.
(2) One consolidometer swell and one strength
test should be performed on specimens from at least
five undisturbed samples at different depths within
the depth of the anticipated active zone (e.g., within 10
to 20 feet beneath the base of the foundation). Suction
tests may also be performed at relatively frequent
depth intervals (e.g., l-foot increments) to better characterize swell behavior and thereby increase confidence in prediction of potential heave discussed in
chapter 5.
(3) One consolidometer swell and one strength
test should be performed on specimens from each
undisturbed sample (or at intervals of 2.5 feet. for
continuous sampling) at depths above the base of deep
shaft foundations to permit evaluation of the adjacent
soil heave and uplift forces exerted on the shaft/soil
interface, Suction tests may also be performed to further characterize swell behavior and increase confidence in prediction of potential heave.
(4) Suction test results can characterize the pore
pressure profile by indicating depths of desiccation
and wetting, which are useful for minimizing potential
foundation problems from soil movement and for evaluating remedial measures to correct problems.

4-5

TM 5-818-7

CHAPTER 5
METHODOLOGY FOR PREDICTION OF VOLUME CHANGES
5-1. Application of heave predictions

5-2. Factors influencing heave

Reasonable estimates of the anticipated vertical and

horizontal heave and the differential heave are necessary for the following applications.
a. Determination of adequate designs of structures
that will accommodate the differential soil movement
without undue distress (chap 6). These predictions are
also needed to estimate upward drag from swelling
soils on portions of deep foundations such as drilled
shafts within the active zone of moisture change and
heave. Estimates of upward drag help determine an
optimum design of the deep foundation.
b. Determination of techniques to stabilize the foundation and to reduce the anticipated heave (chap 7).

Table 5-1 describes factors that significantly influence the magnitude and rate of foundation movement.
The difficulty of predicting potential heave is complicated beyond these factors by the effect of the type
and geometry of foundation, depth of footing, and distribution of load exerted by the footing on the magnitude of the swelling of expansive foundation soil.
Additional problems include estimating the exact location that swelling soils will heave or the point source
of water seeping into the swelling soil and the final or
equilibrium moisture profile in the areas of heaving
soil.

5-3. Direction of soil movement

The foundation soil may expand both vertically and
laterally. The vertical movement is usually of primary
interest, for it is the differential vertical movement
that causes most damages to overlying structures.
a. Vertical movement. Methodology for prediction
of the potential total vertical heave requires an assumption of the amount of volume change that occurs
in the vertical direction. The fraction of volumetric
swell N that occurs as heave in the vertical direction
depends on the soil fabric and anisotropy. Vertical
heave of intact soil with few fissures may account for
all of the volumetric swell such that N = 1, while
vertical heave of heavily fissured and isotropic soil
may be as low as N = 1/3 of the volumetric swell.
b. Lateral movement. Lateral movement is very important in the design of basements and retaining
walls. The problem of lateral expansion against basement walls is best managed by minimizing soil volume
change using procedures described in chapter 7. Otherwise, the basement wall should be designed to resist
lateral earth pressures that approach those given by
(5-1)
horizontal earth pressure, tons per square
root
lateral coefficient of earth pressure at rest
soil vertical or overburden pressure, tons
per square foot
coefficient of passive earth pressure
order of 1 to 2 in expansive soils and often no greater
than 1.3 to 1.6.
5-4. potential total vertical heave
Although considerable effort has been made to develop
methodology for reliable predictions within 20 percent
of the maximum in situ heave, this degree of accuracy
will probably not be consistently demonstrated, particularly in previously undeveloped and untested
areas. A desirable reliability is that the predicted potential total vertical heave should not be less than 80
percent of the maximum in situ heave that will eventually occur but should not exceed the maximum in
situ heave by more than 20 to 50 percent. Useful predictions of heave of this reliability can often be approached and can bound the in situ maximum levels of
heave using the results of both consolidometer swell
and soil suction tests described in paragraph 4-2a. The
fraction N (para 5-3a) should be 1 for consolidometer
swell test results and a minimum of 1/3 for soil suction
test results. The soil suction tests tend to provide an
upper estimate of the maximum in situ heave (N = 1)
in part because the soil suction tests are performed
5-2

without the horizontal restraint on soil swell that

exists in the field and during one-dimensional consolidometer swell tests.
a. Basis of calculation. The potential total vertical
heave at the bottom of the foundation, as shown in figure 5-1, is determined by
i= NEL
DELTA(i)
AH=
i= NBX
i= NEL
(5-2)
i= NBX
where
AH= potential vertical heave at the
bottom of the foundation, feet
N = fraction of volumetric swell that
occurs as heave in the vertical direction
DX = increment of depth, feet
NEL = total number of elements
NBX = number of nodal point at bottom
of the foundation
DELTA(i) = potential volumetric swell of soil
element i, fraction
final void ratio of element i
initial void ratio of element i
The AH is the potential vertical heave beneath a flexible, unrestrained foundation. The bottom nodal point
NNP = NEL + 1, and it is often set at the active depth
(1) The initial void ratio, which depends on geologic and stress history (e.g., maximum past pressure),
the soil properties, and environmental conditions
shown in table 5-1 may be measured on undisturbed
specimens using standard laboratory test procedures.
It may also be measured during the laboratory swell
tests as described in EM 1110-2-1906. The final void
ratio depends on changes in the foundation conditions
caused by construction of the structure.
(2) The effects of the field conditions listed in table 5-1 may be roughly simulated by a confinement
pressure due to soil and structural loads and an assumption of a particular final or equilibrium pore
water pressure profile within an active depth of heave
pore water pressure profiles are related to the final
void ratio by physical models. Two models based on results of consolidometer swell and soil suction tests are
used in this manual (para 4-2a).
b. Pore water pressure profiles. The magnitude of
swelling in expansive clay foundation soils depends on
the magnitude of change from the initial to the equilibrium or final pore water pressure profile that will be
observed to take place in a foundation soil because of

TM 5-818-7

the construction of the foundation.

(1) Initial profile. Figure 5-1 illustrates relative
initial dry and wet profiles. The wet initial profile is
probably appropriate following the wet season, which
tends to occur by spring, while the dry initial profile
tends to occur during late summer or early fall. The
initial pore water pressure profile does not need to be
known if the consolidometer swell model is used because the heave prediction is determined by the differratios (fig. 4-2). The initial void ratio is a function of
the initial pore water pressure in the soil. The initial
pore water pressure profile, which must be known if
the soil suction model is used, may be found by the
method described in appendix B.
(2) Equilibrium profile. The accuracy of the prediction of the potential total vertical heave in simulating the maximum in situ heave depends heavily on the
ability to properly estimate the equilibrium pore water
pressure profile. This profile is assumed to ultimately
occur beneath the central portion of the foundation.
The pore water pressure profile beneath the foundation perimeter will tend to cycle between dry and wet
extremes depending on the field environment and
availability of water. The three following assumptions
are proposed to estimate the equilibrium profile. A
fourth possibility, the assumption that the groundwater level rises to the ground surface, is most conservative and not normally recommended as being
realistic. The equilibrium profile may also be estimated by a moisture diffusion analysis for steady-state
flow, which was used to predict differential heave as
part of the procedure developed by the Post-Tensioning Institute (PTI) for design and construction of slabson-grade (para 6-3b). The results, which should be
roughly compatible with the hydrostatic profiles
discussed in (b) and (c) below, lead to predictions of
heave smaller than the saturated profile.
(a) Saturated. The saturated profile, Method 1

in figure 5-1, assumes that the in situ pore water preschange and heave
(5-3)
foot at any depth X in feet within the active zone. Although a pore water pressure profile of zero is not in
equilibrium, this profile is considered realistic for
most practical cases and includes residences and buildings exposed to watering of perimeter vegetation and
possible leaking underground water and sewer lines.
Water may also condense in a layer of permeable subgrade soil beneath foundation slabs by transfer of
water vapor from air flowing through the cooler subgrade. The accumulated water may penetrate into
underlying expansive soil unless drained or protected
by a moisture barrier. This profile should be used if
other information on the equilibrium pore water pressure profile is not available.
(b) Hydrostatic I. The hydrostatic I profile,
Method 2 in figure 5-la, assumes that the pore water
pressure becomes more negative with increasing vertical distance above the groundwater level in proportion
to the unit weight of water
(5-4)
cubic foot).
This profile is believed to be more realistic beneath
highways and pavements where drainage is good,
pending of surface water is avoided, and leaking underground water lines are not present. This assumption will lead to smaller predictions of heave than the
saturated profile of Method 1.
(c) Hydrostatic II. This profile, Method 3 in fig
ure 5-lb, is similar to the hydrostatic I profile except
that a shallow water table does not exist. The negative
pore water pressure of this profile also becomes more
negative with increasing vertical distance above the
5-3

TM 5-818-7

weight of water
(5-5)
(d) Example application. Figure 5-2 illustrates
how the saturated (Method 1) and hydrostatic II
(Method 3) profiles appear for a suction profile without a shallow water table at a sampling site near
Hayes, Kansas. The initial in situ soil suction or negative pore water pressure was calculated from the given
natural soil suction without confining pressure To by
T
(5-6)
where

mean normal confining pressure, tons per

square foot

the hydrostatic equilibrium profile is nearly vertical

with respect to the large magnitude of soil suction observed at this site. Heave will be predicted if the saturated profile occurs (Method 1 as in fig. 5-1), while
shrinkage will likely be predicted if the hydrostatic II
(Method 3) profile occurs. The availability of water to
the foundation soil is noted to have an enormous impact on the volume change behavior of the soils. Therefore, the methods of chapter 7 should be used as much
as practical to promote and maintain a constant moisture environment in the soil.
c. Depth of the active zone. The active zone depth
changes in water content and heave occur because of
climate and environmental changes after construction
of the foundation.
be assumed equal to the depth of the water table for
groundwater levels less than 20 feet in clay soil (fig.
ro for the hydrostatic I equilibrium profile in the presence of such a shallow water table.

was assumed to be unity. The initial in situ soil suction

deep groundwater levels may often be determined by
evaluating the initial pore water pressure or suction
with depth profile as described in appendix B, The
magnitude of u., is then determined after the depth
that of the corresponding negative pore water pressure
(a) If depths to groundwater exceed 20 feet beneath the foundation and if no other information is
10 feet (for moist profiles or soil suctions less than 4
tons per square foot) and 20 feet (for dry profiles or
soil suctions greater than 4 tons per square foot) below
the base of, the foundation (fig. 5-lb). However, the
the base diameter of a shaft foundation. Sources of
moisture that can cause this active zone include the
seepage of surface water down the soil-foundation interface, leaking underground water lines, and seepage
from nearby new construction.
(b) The pore water pressure or soil suction is often approximately constant with increasing depth bebelow which the water content/plastic limit ratio or
soil suction is constant.
(c) If the soil suction is not approximately constant with increasing depth below depths of 10 to 20
to 2 feet below the first major change in the magnitude of the soil suction, as shown in figure 5-2.
d. Edge effects. Predictions of seasonal variations in
vertical heave from changes in moisture between extreme wet and dry moisture conditions (fig. 5-1) are
for perimeter regions of shallow foundations. These

TM 5-818-7

calculations require a measure or estimate of both seasonal wet and dry pore water pressure or suction profiles. It should be noted from figure 5-lb that perimeter cyclic movement from extremes in climatic
changes can exceed the long-term heave beneath the
center of a structure.
(1) Soil-slab displacements. A slab constructed on
the ground surface of a wet site may in time lead to
downwarping at the edges after a long drought or
growth of a large tree near the structure (fig. 5-3a).
Edge uplift may occur following construction on an
initially dry site (fig. 5-3b). The AH in figure 5-3 is
representative of the maximum differential vertical
heave beneath the slab, excluding effects of restraint
from the slab stiffness, but does consider the slab
weight.
(2) Edge distance. The edge lift-off distance e of
lightly loaded thin slabs at the ground surface often
varies from 2 to 6 feet but can reach 8 to 10 feet.
(3) Deflection/length ratio. The deflection/length
ratio of the slab is A/L, where A is the slab deflection
and L is the slab length. The angular deflection/span
5-3).

(5-8)
thickness of expansive soil layer, feet
swell index, slope of the curve between
points 3 and 4, figure 4-2
swell pressure, tons per square foot
final vertical effective pressure, tons per
square foot
The final effective pressure is given by
(5-9)

4-2. A simple hand method and an example of predicting potential total vertical heave from consolidometer
swell tests assuming a saturated equilibrium profile,
equation (5-3), are given in TM 5-818-1 and in figure
5-4. However, hand calculations of potential heave
can become laborious, particularly in heterogeneous
profiles in which a variety of loading conditions need
to be evaluated for several different designs,
(2) Computer applications. Predictions of potential total heave or settlement can be made quickly with
the assistance of the computer program HEAVE available at the U. S. Army Corps of Engineers Waterways

Experiment Station. The program HEAVE is applicable to slab, long continuous, and circular shaft foundations. This program considers effects of loading and
soil overburden pressures on volume changes, heterogeneous soils, and saturated or hydrostatic equilibrium moisture profiles (equations (5-3) to (5-5)). Results
of HEAVE using the saturated profile, equation (5-3),
are comparable with results of manual computations
described in figure 5-4.
5-5. Potential differential heave
Differential heave results from edge effects beneath a
finite covered area, drainage patterns, lateral variations in thickness of the expansive foundation soil, and
effects of occupancy. The shape and geometry of the
structure also result in differential heave. Examples of
effects of occupancy include broken or leaking water
and sewer lines, watering of vegetation, and ponding
adjacent to the structure. Other causes of differential
heave include differences in the distribution of load
and the size of footings.
a. Unpredictability of variables. Reliable predictions of future potential differential heave are often
not possible because of many unpredictable variables
that include: future availability of moisture from
rainfall and other sources, uncertainty of the exact locations of heaving areas, and effects of human occupancy.
b. Magnitude of differential heave.

(1) Potential differential heave can vary from zero

to as much as the total heave. Differential heave is often equal to the estimated total heave for structures
supported on isolated spot footings or drilled shafts because some footings or portions of slab foundations often experience no movement. Eventually, differential
heave will approach the total heave for most practical
cases and should, therefore, be assumed equal to the
total potential heave, unless local experience or other
information dictates otherwise.
(2) The maximum differential heave beneath a
lightly loaded foundation slab may also be estimated
by the procedure based on the moisture diffusion theory and soil classification data developed by the PTI.
Heave predictions by this method will tend to be less
than by assuming that the differential heave is the total potential heave.
5-6. Heave with time
Predictions of heave with time are rarely reliable because the location and time when water is available to
the soil cannot be readily foreseen. Local experience
has shown that most heave (and the associated structural distress) occurs within 5 to 8 years following construction, but the effects of heave may also not be observed for many years until some change occurs in the
5-5

foundation conditions to disrupt the moisture regime.

Predictions of when heave occurs are of little engineering significance for permanent structures. The impor-

tant engineering problems are the determination of

the magnitude of heave and the development of ways
to minimize distress of the structure.

5-7

DESIGN OF FOUNDATIONS
6-1. Basic considerations
a. Planning. Swelling of expansive foundation soils
should be considered during the preliminary design
phase and the level of structural cracking that will be
acceptable to the user should be determined at this
time.
(1) The foundation of the structure should be designed to eliminate unacceptable foundation and structural distress. The selected foundation should also be
compatible with available building materials, construction skills, and construction equipment.
(2) The foundation should be designed and constructed to maintain or promote constant moisture in
the foundation soils. For example, the foundation
should be constructed following the wet season if possible. Drainage should be provided to eliminate ponded
water. Excavations should be protected from drying.
Chapter 7 describes the methods of minimizing soil
movement.
b. Bearing capacity. Foundation loading pressures
should exceed the soil swell pressures, if practical, but
should be sufficiently less than the bearing capacity to
maintain foundation displacements within tolerable
amounts, Present theoretical concepts and empirical
correlations permit reasonably reliable predictions of
ultimate capacity, but not differential movement of
the foundation. Factors of safety (FS) are therefore applied to the ultimate bearing capacity to determine
safe or allowable working loads consistent with tolerable settlements. Further details on bearing capacity
are presented in TM 5-818-1.
c. Foundation systems. An appropriate foundation
should economically contribute to satisfying the functional requirements of the structure and minimize differential movement of the various parts of the structure that could cause damages. The foundation should
be designed to transmit no more than the maximum
tolerable distortion to the superstructure. The amount
of distortion that can be tolerated depends on the design and purpose of the structure. Table 6-1 illustrates
foundation systems for different ranges of differential

selection of the foundation. Figure 6-1 explains the

ing soil, rock, or sand. Pervious sand strata may provide a path for moisture flow into nearby swelling soil.
(1) Shallow individual or continuous footings.
Shallow individual or long continuous footings are often used in low swelling soil areas where the predicted
footing angular deflection/span length ratios are on
the order of 1/600 to 1/1000 or 0.5 inch or less of
movement.
(2) Stiffened mats (slabs). Stiffened mat foundations are applicable in swelling soil areas where predicted differential movement AH may reach 4 inches.
The stiffening beams of these mats significantly reduce differential distortion. The range provided in table 6-1 for beam dimensions and spacings of stiffened
slabs for light structures normally provides an adequate design.
(3) Deep foundations. A pile or beam on a drilled
shaft foundation is applicable to a large range of foundation soil conditions and tends to eliminate effects of
heaving soil if properly designed and constructed (para
6-4). The type of superstructure and the differential
soil movement are usually not limited with properly
designed deep foundations. These foundations should
lead to shaft deflection/spacing ratios of less than
1/600.
d. Superstructure systems. The superstructure
should flex or deform compatibly with the foundation
such that the structure continues to perform its functions, contributes aesthetically to the environment,
and requires only minor maintenance. Frame construction, open floor plans, and truss roofs tend to minimize
damage from differential movement. Load bearing
walls tend to be more susceptible to damage from
shear than the relatively flexible frame construction.
Wood overhead beams of truss roof systems provide
structural tension members and minimize lateral
thrust on walls. Table 6-2 illustrates the relative flexibility provided by various superstructure systems.
(1) Tolerable angular deflection/length ratios. The
ability of a structure to tolerate deformation depends
on the brittleness of the building materials, length to
height ratio, relative stiffness of the structure in shear
and bending, and mode of deformation whether heave
(dome-shaped, fig. 1-2) or settlement (dish-shaped, fig.
that can be tolerated, therefore, varies considerably

not a satisfactory basis of design in situations such as a

5-foot layer of highly swelling soil overlying nonswell-

PI = 25

PI 5 0

PI =40

footings or about twice the A/L ratio of the slab (fig.

5-3). Only rough guidance of the range of tolerable
ferent framing systems.
(a) Propagation of cracks depends on the degree
of tensile restraint built into the structure and its
foundation. Thus, frame buildings with panel walls are
able to sustain larger relative deflections without severe damage than unreinforced load-bearing walls.
Structural damage is generally less where the dishshaped pattern develops than in the case of center
heaving or edge downwarping because the foundation
is usually better able to resist or respond to tension
forces than the walls.
avoid cracking in single and multistory structures.
Plaster, masonry or precast concrete blocks, and brick
1/600 to 1/1000. However, cracks may not appear in
these walls if the rate of distortion is sufficiently slow
to allow the foundation and frame to adjust to the new

distortions. The use of soft bricks and lean mortar also

tend to reduce cracking. Reinforced masonry, reinforced concrete walls and beams, and steel frames can
pear in the structure. Deflection ratios exceeding
1/250 are likely to be noticed in the structure and
1/150 usually lead to structural damage.
(2) Provisions for flexibility. The flexibility required to avoid undesirable distress may be provided
by joints and flexible connections. Joints should be
provided in walls as necessary, and walls should not be
tied into the ceiling. Slabs-on-grade should not be tied
into foundation walls and columns but isolated using
expansion joints or gaps filled with a flexible, impervious compound. Construction items, such as reinforced concrete walls, stud frames, paneling, and
gypsum board, are better able to resist distortions and
should be used instead of brick, masonry blocks, or
plaster walls. The foundation may be further reinforced by making the walls structural members capa6-3

TM 5-818-7

ble of resisting bending such as reinforced concrete

shear walls. Several examples of frame and wall construction are provided in appendix C.
6-2. Shallow individual or continuous
footings
a. Susceptibility y to damage. Structures supported
by shallow individual or continuous wall footings are
susceptible to damages from lateral and vertical movement of foundation soil if provisions are not made to
accommodate possible movement. Dishing or substantial settlement may occur in clays, especially in initially wet soil where a well-ventilated crawl space is constructed under the floor. The crawl space prevents
rainfall from entering the soil, but the evaporation of
moisture from the soil continues. Center heave or edge
downwarping (fig. 1-2) can occur if the top layer of
soil is permeable and site drainage is poor. Fractures
may appear in walls not designed for differential

movement exceeds about 0.5 inch.

b. Applications. Shallow footings may be used

where expansive strata are sufficiently thin to allow
location of the footing in a nonexpansive or low-swelling stratum (fig. 6-2).
(1) A structural floor slab should be suspended on
top of the footing (fig. 6-2a) or the slab-on-grade
should be isolated from the walls (fig. 6-2b). The slabon-grade should be expected to crack.
(2) Figure 6-3 illustrates examples of interior construction for a slab-on-grade. Interior walls may be
suspended from the ceiling or supported on the floor.
A flexible joint should be provided in the plenum between the furnace and the ceiling. Sewer lines and
other utilities through the floor slab should be permitted to slip freely.
(3) Swelling of deep expansive soil beneath a nonexpansive stratum may cause differential movement
of shallow footings if the moisture regime is altered in
the deep soil following construction (e.g., change in
groundwater level, or penetration of surface water
into deep desiccated soil). Excavations for crawl spaces

or basements decrease the vertical confining pressure

and pore water pressure, which can cause the underlying expansive foundation soil to heave from adjustment of the moisture regime back to the natural pore
water pressures.
c. Basements. Basements and long continuous footings constructed in excavations are subject to swell
pressures from underlying and adjacent expansive soil.
(1) Walls. Basement walls of reinforced concrete
can be constructed directly on the foundation soil
without footings provided foundation pressures are
less than the allowable bearing capacity (fig. 6-4a).
However, placing heavy loads on shallow footings may
not be effective in countering high swell pressures because of the relative small width of the footings. The
stress imposed on the soil is very low below a depth of
about twice the width of the footing and contributes
little to counter the swell pressure unless the expansive soil layer is thin.
(2) Voids. Voids can also be spaced at intervals beneath the walls to increase loading pressures on the
foundation soil and to minimize flexing or bowing of
the walls (fig. 6-4b). The voids may be made with removable wood forms, commercially available card-

INTERIOR

board, or other retaining forms that deteriorate and

collapse (para 6-4d).
(3) Joints. Joints should be provided in interior
walls and other interior construction if slab-on-ground
is used (fig. 6-3). The slab should be isolated from the
walls with a flexible impervious compound.
(4) Lateral earth pressure on wall. The coefficient
of lateral earth pressure can exceed one if the backfill
is heavily compacted and expansive, or the natural soil
adjacent to the wall is expansive. Controlled backfills
are recommended to minimize lateral pressures and increase the economy of the foundation (para 7-3a).
Steel reinforcement can provide the necessary restraint to horizontal earth pressures, Unreinforced
masonry brick and concrete blocks should not be used
to construct basement walls.
d. Design. Standard design procedures for foundations of buildings and other structures are presented in
TM5-818-1.

6-3. Reinforced slab-on-grade foundations

a. Application. The reinforced mat is often suitable
for small and lightly loaded structures, particularly if

TM 5-818-7

the expansive or unstable soil extends nearly continuously from the ground surface to depths that exclude
economical drilled shaft foundations. This mat is suitable for resisting subsoil heave from the wetting of
deep desiccated soil, a changing water table, laterally
discontinuous soil profiles, and downhill creep, which
results from the combination of swelling soils and the
presence of slopes greater than 5 degrees. A thick, reinforced mat is suitable for large, heavy structures.
The rigidity of thick mats minimizes distortion of the
superstructure from both horizontal and vertical
movements of the foundation soil.
(1) Effects of stiffening beams. Concrete slabs
without internal stiffening beams are much more susceptible to distortion or doming from heaving soil.
Stiffening beams and the action of the attached superstructure with the mat as an indeterminate structure
increase foundation stiffness and reduce differential
movement. Edge stiffening beams beneath reinforced
concrete slabs can also lessen soil moisture loss and reduce differential movement beneath the slab. However, the actual vertical soil pressures acting on stiffened slabs can become very nonuniform and cause localized consolidation of the foundation soil.
(2) Placement of nonswelling layer. Placement of
a nonswelling, 6-inch-(or more) thick layer of (preferably) impervious soil on top of the original ground surface before construction of lightly loaded slabs is recommended to increase the surcharge load on the
foundation soil, slightly reduce differential heave, and
permit the grading of a slope around the structure
leading down and away from it. This grading improves
drainage and minimizes the possibility that the layer
(if pervious) could be a conduit for moisture flow into
desiccated foundation expansive soils. The layer
should have some apparent cohesion to facilitate
trench construction for the stiffening beams.
6-6

b. Design of thin slabs for light structures. Stiffened slabs may be either conventionally reinforced or
posttensioned. The mat may be inverted (stiffening
beams on top of the slab) in cases where bearing capacity of the surface soil is inadequate or a supported first
floor is required. The Department of Housing and Urban Development, Region IV, San Antonio Area Office, has documented a series of successful conventionally reinforced and posttensioned slabs for the
southern central states. Successful local practice
should be consulted to help determine suitable designs.
(1) Conventionally reinforced. The conventional
reinforced concrete waffle type mat (table 6-1), which
is used for light structures, consists of 4- to 5-inchthick concrete slab. This slab contains temperature
steel and is stiffened with doubly reinforced concrete
crossbeams. Figure 6-5 illustrates an engineered rebar
slab built in Little Rock, Arkansas. Appendix C provides details of drawings of reinforced and stiffened
thin mats. The 4-inch slab transmits the self-weight
and first floor loading forces to the beams, which resist the moments and shears caused by differential
heave of the expansive soil. Exterior walls, roof, and
internal concentrated loads bear directly on the stiffening beams. Clearance between beams should be limited to 400 square feet or less. Beam spacings may be
varied between the limits shown in table 6-1 to allow
for concentrated and wall loads. Beam widths vary
from 8 to 12 or 13 inches but are often limited to a
minimum of 10 inches.
(a) Concrete and reinforcement. Concrete compressive strength f c should be at least 2500 psi and
preferably 3000 psi. Construction joints should be
placed at intervals of less than 150 ft and cold joints
less than 65 ft. About 0.5 percent reinforcing steel
should be used in the mat to resist shrinkage and temperature effects.

(b) Preliminary design, The three designs for reinforced and stiffened thin mats presented in table
6-1 differ in the beam depth and spacing depending

ings for each of the light, medium, and heavy slabs are

- -

servative in view of still undetermined fully acceptable

design criteria and relatively high repair cost of reinforced and stiffened slabs. Stirrups may be added, particularly in the perimeter beams, to account for concentrated and exterior wall loads.
(2) Post-tensioned. Figure 6-6 illustrates an example of a posttensioned slab. Properly designed and
constructed posttensioned mats are more resistant to
fracture than an equivalent section of a conventional
rebar slab and use less steel. However, post-tensioned
slabs should still be designed with adequate stiffening
beams to resist flexure or distortion from differential
heave of the foundation soil, Experienced personnel
are necessary to properly implement the posttensioning.
(3) Assumptions of design parameters. Design
parameters include effects of climate, center and edge
modes of differential swelling, perimeter and uniform
loads, and structural dimensions.
(a) The effects of climate and differential swelling are accounted for by predictions of the maximum
differential heave AH and the maximum edge lift-off

related with the Thornthwaite Moisture Index (TMI) in

figure 6-7. The TMI, a climate related parameter
roughly estimated from figure 6-8, represents the
overall availability of water in the soil. The TMI can
vary 10 to 20 or more (dimensionless) units from year

exceed the range given in figure 6-7, depending on the

activity of the soil or extreme changes in climatic conditions (e.g., long droughts and heavy rainfall), the

(b) The loading distribution depends on the

architectural arrangement of the building and often
cannot be significantly altered. Perimeter and concentrated loads should be supported directly on the stiffening beams.
(c) The length and width of the slab are usually
fixed by the functional requirement. Beam spacing depends on the slab geometry and varies between 10 and
20 feet. The depth of stiffening beams is controlled by
the moment and shear capacity. The beam depth is adjusted as needed to remain within the allowable limits.
The width of the stiffening beam is usually controlled
by the excavation equipment and soil bearing capacity.
(4) Structural design procedure, The design proce6-7

(Department of Housing and Urban

Development,
Region
IV)

Figure 6-6. Post-tensioned slab in Lubbock, Texas, for single-family, single-story, minimally loaded frame residence.

dure should provide adequate resistance to shear, moment, and deflections from the structural loading
forces, while overdesign is minimized. An economically competitive procedure that builds on the early work
of the Building Research Advisory Board of the National Academy of Sciences is that developed for the
Post-Tensioning Institute (PTI).
(a) The PTI procedure is applicable to both con-

CLIMATE

ventionally reinforced and posttensioned slabs up to

18 inches thick. It considers the previously discussed
assumptions of the design parameters.
of the unloaded soil determined by the PTI procedure
reflect average moisture conditions and may be exceeded if extreme changes in climate occur.
(c) Material parameters required by the PTI pro-

cedure are the compressive strength of concrete; allowable tensile and compressive stresses in concrete; type,
grade, and strength of the prestressing steel; grade
and strength of the mild steel reinforcement; and slab
subgrade friction coefficient, The amount of reinforcing steel recommended by this procedure should be
considered a minimum. The slab-subgrade coefficient
of friction should be 0.75 for concrete cast on polyethylene membranes and 1.00 if cast on-grade.
This ratio may be as large as 1/360 for center heave
criterion is recommended by the PTI because edge lift
is usually much less than center lift deflections and the
stems of the beams resisting the positive bending moment may be unreinforced.
c. Design of thick mats. The state of the art for estimating spatial variations in soil pressures on thick
mats is often not adequate. These mats tend to be
heavily overdesigned because of the uncertainty in the
loading and the relatively small extra investment of
some overdesign.
(1) Description. Concrete mats for heavy structures tend to be 3 feet or more in thickness with a continuous two-way reinforcement top and bottom. An 8foot-thick mat supporting a 52-story structure in
Houston, Texas, contains about 0.5 percent steel,
while the 3-foot-thick mat of the Wilford Hall Hospital
complex at Lackland Air Force Base in Texas also contains about 0.5 percent steel. The area of steel is 0.5
percent of the total area of the concrete distributed
equally each way both top and bottom. The steel is
overlapped near the concentrated loads, and a 3-inch

cover is provided over the steel. The depth of the excavation that the mats are placed in to achieve bearing
capacity and tolerable settlements eliminates seasonal
edge effects such that the edge lift-off distance is not
applicable.
(2) Procedure. The thick mat is designed to determine the shear, moment, and deflection behavior
using conventional practice, then modified to accommodate swell pressures and differential heave caused
by swelling soils. The analyses are usually performed
by the structural engineer with input on allowable soil
bearing pressures, uplift pressures (hydrostatic and
swell pressures from expansive soils) and estimates of
potential edge heave/shrinkage and center heave from
the foundation engineer. Computer programs are commonly used to determine the shear, moments, and deflections of the thick mat.
(a) Structural solutions. The structural solution
may be initiated with an estimate of the thickness of a
spread footing that resists punching shear and
bending moments for a given column load, concrete
compressive strength, and soil bearing capacity. Following an estimation of the initial thickness, hand
solutions of mat foundations for limited application
based on theory of beams on elastic foundations are
available from NAVFAC DM-7. More versatile solutions are available from computer programs based on
theory of beams on elastic foundations such as
BMCOL 2, which is available at the U.S. Army Corps
of Engineer Waterways Experiment Station, and finite element analysis.
(b) Foundation soil/structure solutions. The

BMCOL 2 soil-structure interaction program permits

nonlinear soil behavior. Finite element programs are

also available, but they are often burdened with hard

to explain local discontinuities in results, time-consuming programming of input data, and need of experienced personnel to operate the program. The finite
element program originally developed for analysis of
Port Allen and Old River Locks was applied to the
analysis of the Wilford Hall Hospital mat foundation
at Lackland Air Force Base in Texas. Figure 6-9 compares predicted with observed movement of the 3.5foot-thick mat at Wilford Hall. Foundation soils include the fissured, expansive Navarro and upper Midway clay shales. These computer programs help refine
the design of the mat and can lead to further cost reductions in the foundation.
6-4. Deep foundations
The deep foundation provides an economical method
for transfer of structural loads beyond (or below) unstable (weak, compressible, and expansive) to deeper
stable (firm, incompressible, and nonswelling) strata.
Usually, the deep foundation is a form of a pile foundation. Numerous types of pile foundations exist of
which the most common forms are given in table 6-3.
Occasionally when the firm-bearing stratum is too
deep for the pile to bear directly on a stable stratum,
the foundation is designed as friction or floating piles
and supported entirely from adhesion with the surrounding soil and/or end bearing on underreamed footings.

a. General applications. Each of the types of piling

is appropriate depending on the location and type of
structure, ground conditions (see table 3-1 for examples), and durability. The displacement pile is usually
appropriate for marine structures. Any of the piles in
table 6-3 may be considered for land applications. Of
these types the bored and cast in situ concrete drilled
shaft is generally more economical to construct than
driven piles.
b. Application of drilled shafts. Table 6-4 describes
detailed applications of drilled shaft foundations including advantages and disadvantages. Detailed discussion of drilled shaft foundations is presented below
because these have been most applicable to the solution of foundation design and construction on expansive clay soils.
(1) A drilled shaft foundation maybe preferred to
a mat foundation if excavating toward an adequate
bearing stratum is difficult or the excavation causes
settlement or loss of ground of adjacent property.
(2) A drilled shaft foundation 20 to 25 feet deep
tends to be economically competitive with a ribbed
mat foundation,
(3) Drilled shafts may be preferred to mat founda-

ratios exceed 1/250, Mat foundations under such conditions may tilt excessively leading to intolerable distortion or cracking.

Figure 6-9. Settlement and deflection of a mat foundation.

pared with traditional strip footings, particularly in

open construction areas and with shaft lengths less
than 10 to 13 feet, or if the active zone is deep, such as
within areas influenced by tree roots.
c. General considerations.
(1) Causes of distress. The design and construction of drilled shaft foundations must be closely controlled to avoid distress and damage. Most problems
have been caused by defects in construction and by inadequate design considerations for effects of swelling
soil (table 6-5). The defects attributed to construction
techniques are discontinuities in the shaft, which may
occur from the segregation of concrete, failure to complete concreting before the concrete sets, and early set
of concrete, caving of soils, and distortion of the steel
reinforcement. The distress resulting from inadequate
design considerations are usually caused by wetting of
subsoil beneath the base, uplift forces, lack of an air
gap beneath grade beams, and lateral movement from
downhill creep of expansive clay.
(2) Location of base. The base of shafts should be
located below the depth of the active zone, such as below the groundwater level and within nonexpansive
soil. The base should not normally be located within
three base diameters of an underlying unstable stratum.
(a) Slabs-on-grade isolated from grade beams
and walls are often used in light structures, such as
residences and warehouses, rather than the more costly structural slabs supported by grade beams and
shafts. These slabs-on-grade will move with the expansive soil and should be expected to crack.
(b) To avoid fall-in of material from the granu-

lar stratum during underreaming of a bell, the base

may be placed beneath swelling soil near the top of a
granular stratum.
(3) Underreams. Underreams are often used to increase anchorage to resist uplift forces (fig. 6-10). The
belled diameter is usually 2 to 2.5 times the shaft
45- or 60-degree bells may be used, but the 45-degree
bell is often preferred because concrete and construction time requirements are less. Although the 45-degree bell may be slightly weaker than the 60-degree
bell, no difference has been observed in practice. The
following considerations are necessary in comparing
underreamed shafts with straight shafts.
(a) Straight shafts may be more economical
than underreams if the bearing stratum is hard or if
subsoils are fissured and friable. Soil above the underream may be loose and increase the upward movement
needed to develop the bell resistance.
(b) The shaft can often be lengthened to eliminate the need for an underream, particularly in soils
where underreams are very difficult to construct. Friction resistance increases rapidly in comparison with
end bearing resistance as a function of the relative
shaft-soil vertical movement.
(c) Underreams reduce the contact bearing pressure on potentially expansive soil and restrict the minimum diameter that may be used.
(4) Uplift forces. If bells or underreams are not
feasible, uplift forces (table 6-5) may be controlled by
the following methods:
(a) The shaft diameter should be the minimum
required for downloads and construction procedures
and control.
6-11

(b) The shaft length may be extended further

into stable, nonswelling soil to depths of twice the

The point n in figure 6-11 is the neutral point. The

(c) Widely spaced shafts may be constructed

exceeds the maximum uplift thrust (fig. 6-11) ex-

equal to the soil allowable bearing capacity. Wide

spans between shafts also reduce angular rotation of
the structural members. The minimum spacing of
shafts should be 12 feet or 8 times the shaft diameter
(whichever is smaller) to minimize effects of adjacent
shafts.
(d) The upper portion of the shaft should be
kept vertically plumb (maximum variation of 1 inch in
6 feet shown in fig. 6-10) and smooth to reduce adhesion between the swelling soil and the shaft. Friction
reducing material, such as roofing felt, bitumen slip
layers, polyvinyl chloride (PVC), or polyethylene
sleeves, may be placed around the upper shaft to re6-13

TM 5-018-7

duce both uplift and downdrag forces. Vermiculite,

pea gravel, or other pervious materials that will allow
access of water to the expansive material should be
avoided.
d. Design. The heave or settlement of the foundation usually controls the design and should not exceed
specified limits set by usage requirements and tolerances of the structure. The design of drilled shafts, in
addition to bearing capacity, should consider the method of construction, skin resistance, and uplift forces.
The computer program HEAVE (WES Miscellaneous
Paper GL-82-7) may be used to help determine the
movement of drilled shafts for different lengths and
diameters of the shaft, and the diameter of the underream for different loading forces.
(1) Skin resistance. Skin resistance develops from
small relative displacements between the shaft and the
adjacent soil. Positive (upward directed) skin friction,
which helps support structural loads, develops when
the shaft moves down relative to the soil. Uplift of adjacent swelling soils also transfers load to the shaft

foundation by positive skin friction and can cause

large tensile stresses to develop in the shaft. Negative
skin friction, which adds to the structural loads and increases the end bearing force, develops when the surrounding soil moves down relative to the shaft, Negative skin friction is associated with the settling of the
adjacent fill, loading of surrounding compressible
soils, or lowering of the groundwater level.
ated by the equation
(6-2)
adhesion, tons per square foot
ratio of horizontal to vertical effective
stress
vertical effective stress, tons per square
foot
angle of friction between the soil and
shaft, degrees

TM 5-818-7

soil against concrete. The skin resistance, which is a

function of the type of soil (sand, clay, and silt), is
usually fully mobilized with a downward displacement
of 1/2 inch or less or about 1 to 3 percent of the shaft
diameter. These displacements are much less than
those required to fully mobilize end bearing resistance.
(b) The fully mobilized skin resistance has been
compared with the undrained, undisturbed shear

bottom of the shafts in cohesive soils, reaching a maxi-

(6-3)

attributed to soil shrinkage during droughts and low

lateral pressure, while the reduction at the bottom is
attributed to interaction of stress between end bearing
and skin resistance.
(c) Skin resistance may also be evaluated in
terms of effective stress from results of drained direct
shear tests
(6-4)

found to vary between 0.25 and 1.25 depending on the

type of shaft and soil conditions. The reduction factor
is the ratio of mobilized shearing resistance to the un-

angle of internal friction. The effective cohesion is assumed zero in practical applications and eliminated
from equation (6-4). Most of the available field data

be independent of soil strength. Also, the in situ reduction factor may appear greater than one depending on
the mechanism of load transfer. For example, the
shaft load may be transferred over some thickness of
soil such that the effective diameter of the shaft is
greater than the shaft diameter D,. The reduction factor concept, although commonly used, is not fully

mally consolidated soils, while it is about 0.8 for over-

of 0.25 is recommended if little is known about the soil

or if slurry construction is used.
The reduction factor approaches zero near the top and

be calculated for normally consolidated soils by

(6-5a)
and in overconsolidated soils by
(6-5b)

cohesion is often assumed to be zero.

(2) Uplift forces. Uplift forces, which area direct
function of swell pressures, will develop against sur6-15

faces of shaft foundations when wetting of surrounding expansive soil occurs. Side friction resulting in uplift forces should be assumed to act along the entire
depth of the active zone since wetting of swelling soil
causes volumetric expansion and increased pressure
against the shaft. As the shaft tends to be pulled upward, tensile stresses and possible fracture of concrete
in the shaft are induced, as well as possible upward displacement of the entire shaft.
(a) The tension force T (a negative quantity)
may be estimated by
(6-6)
includes the weight of the shaft. Limited observations

and (6-4)) varies between 1 and 2 in cohesive soils for

shafts subject to uplift forces. The same swelling responsible for uplift also increases the lateral earth
pressure on the shaft. Larger K values increase the
computed tension force.
(b) The shaft should be of proper diameter,
length, and underreaming, adequately loaded, and
contain sufficient reinforcing steel to avoid both tensile fractures and upward displacement of the shaft.
ASTM A 615 Grade 60 reinforcing steel with a miniThe minimum percent steel required if ASTM A 615
Grade 60 steel is used is given approximately by
T
(6-7)
where T is the tension force in tons and the shaft diam-

more may be required. The reinforcing steel should be

hooked into any existing bell as shown in figure 6-10,
and it may also be hooked into a concrete grade beam.

Maximum concrete aggregate size should be one third

of the openings in the reinforcement cage.
d. Grade beams. Grade beams spanning between
shafts are designed to support wall loads imposed vertically downward. These grade beams should be isolated from the underlying swelling soil with a void
space beneath the beams of 6 to 12 inches or 2 times
the predicted total heave of soil located above the base
of the shaft foundation (whichever is larger). Steel is
recommended in only the bottom of the grade beam if
grade beams are supported by drilled shafts above the
void space. Grade beams resting on the soil without
void spaces are subject to distortion from uplift pressure of swelling foundation soil and are not recommended.
(1) Preparation of void space. Construction of
grade beams with void spaces beneath the beams may
require overexcavation of soil in the bottom of the
grade beam trench between shafts. The void space may
be constructed by use of sand that must later be blown
away at least 7 days after concrete placement, or by
use of commercially available cardboard or other retainer forms that will support the concrete. The cardboard forms should deteriorate and collapse before
swell pressures in underlying soil can deflect or damage the grade beams. The resulting voids should be
protected by soil retainer planks and spacer blocks.
Figure 6-12 illustrates some void details.
(2) Loading. Interior and exterior walls and concentrated loads should be mounted on grade beams.
Floors may be suspended from grade beams at least 6
inches above the ground surface, or they maybe placed
directly on the ground if the floor slab is isolated from
the walls. Support of grade beams, walls, and suspended floors from supports other than the shaft foundation should be avoided. Figure 6-13 illustrates typical
exterior and interior grade beams.

b Thickness

Figure 6-13. Typical exterior and interior grade beams.

TM 5-818-7

CHAPTER 7
MINIMIZATION OF FOUNDATION MOVEMENT
7-1. Preparation for construction
The foundation should always be provided with adequate drainage, and the soil properly prepared to minimize changes in soil moisture and differential movement.
a. Removal of vegetation. Existing trees and other
heavy vegetation should be removed. New plantings of
like items installed during postconstruction landscaping should not be located within a distance away from
the structure ranging from 1 to 1.5 times the height of
the mature tree.
b. Leveling of site. Natural soil fills compacted at
the natural water content and the natural density of
the in situ adjacent soil minimize differential movement between cut and fill areas of sloping ground,
trenches, or holes caused by removal of vegetation.
The volume of cut portions should be kept to a minimum. Cut areas reduce the overburden pressure on
underlying swelling soil and lead to time-dependent
heave.
c. Excavation.
(1) Construction in new excavations (within a few
years of excavating) without replacement of a surcharge pressure equal to the original soil overburden
pressure should be avoided where possible because the
reduction in effective stress leads to an instantaneous
elastic rebound plus a time-dependent heave. The reduction in overburden pressure results in a reduction
of the pore water pressure in soil beneath the excavation. These pore pressures tend to increase with time
toward the original or equilibrium pore pressure profile consistent with that of the surrounding soil and
can cause heave.
(2) Ground surfaces of new excavations, such as
for basements and thick mat foundations, should be
immediately coated with sprayed asphalt or other sealing compounds to prevent drying of or the seepage of
ponded water into the foundation soil during construction (fig. 7-1). Rapid-cure RC 70 or medium-cure MC
30 cutback asphalts are often used as sealing compounds, which penetrate into the soil following compaction of the surface soil and cure relatively quickly.

7-2. Drainage techniques

Drainage is provided by surface grading and subsurface drains.

a. Grading. The most commonly used technique is

grading of a positive slope away from the structure.
The slope should be adequate to promote rapid runoff
and to avoid collecting, near the structure, ponded
water, which could migrate down the foundation/soil
interface. These slopes should be, greater than 1 percent and preferably 5 percent within 10 feet of the
foundation,
(1) Depressions or water catch basin areas should
be filled with compacted soil (para 7-3a) to have a
positive slope from the structure, or drains should be
provided to promote runoff from the water catch basin
areas. Six to twelve inches of compacted, impervious,
nonswelling soil placed on the site prior to construction of the foundation can ensure the necessary grade
and contribute additional uniform surcharge pressure
to reduce uneven swelling of underlying expansive
soil.
(2) Grading and drainage should be provided for
structures constructed on slopes, particularly for
slopes greater than 9 percent, to rapidly drain off
water from the cut areas and to avoid pending of water
in cuts or on the uphill side of the structure. This
drainage will also minimize seepage through backfills
into adjacent basement walls.
b. Subsurface drains. Subsurface drains (fig. 7-1)
may be used to control a rising water table, groundwater and underground streams, and surface water
penetrating through pervious or fissured and highly
permeable soil. Drains can help control the water table
before it rises but may not be successful in lowering
the water table in expansive soil. Furthermore, since
drains cannot stop the migration of moisture through
expansive soil beneath foundations, they will not prevent all of the long-term swelling.
(1) Location of subsurface drains, These drains
are usually 4- to 6-inch-diameter perforated pipes
placed adjacent to and slightly below the baseline of
the external wall to catch free water (fig. 7-1).
(a) An impervious membrane should be placed
beneath the drain in the trench to prevent migration
of surface moisture into deeper soil. The membrane adjacent to the foundation wall should be cemented to
the wall with a compatible joint sealant to prevent
seepage through the joint between the membrane and
the foundation.
(b) If a 6- to 12-inch layer of granular material

was provided beneath a slab-on-grade, the granular

material in the drain trench should be continuous with
the granular material beneath the floor slab. The perforated pipe should be placed at least 12 inches deeper
than the bottom of the granular layer. An impervious
membrane should also be placed on the bottom and
sides of the drain trench but should not inhibit flow of
moisture into the drain from beneath the floor slab.
Granular fills of high permeability should be avoided
where possible.
(c) Deep subsurface drains constructed to control arising water table should be located at least 5
feet below a slab-on-grade. An impervious membrane
should not be placed in the drain trench. These drains
are only partially effective in controlling soil heave
above the drain trench, and they are relatively expensive. A more economical solution may be to place a
temporary (or easily removable slab-on-grade) with a
permanent slab after the groundwater table has
reached equilibrium.
(2) Outlets. Drains should be provided with outlets or sumps to collect water and pumps to expel
water if gravity drainage away from the foundation is
not feasible. Sumps should be located well away from
the structure. Drainage should be adequate to prevent
any water from remaining in the drain (i.e., a slope of
at least 1/8 inch per foot of drain or 1 percent should be
provided).
(3) Drain trench material. The intrusion of fines
7-2

in drains maybe minimized by setting the pipe in filter

fabric and pea gravel/sand.
7-3. Stabilization techniques
Two effective and most commonly used soil stabilization techniques are controlled backfilling and
continuous maintenance involving drainage control
and limited watering of surface soil adjacent to the
structure during droughts. Other techniques, such as
moisture barriers and lime treatment, are not widely
used in minimizing differential heave of single and
multistory buildings. Presetting or pending for periods of a few months to a year prior to construction is
often effective but normally is not used because of
time requirements. Prewetting should not be used on
fissured clay shales because swelling from water seeping into fissures may not appear until a much later
date and delayed problems may result.
a. Controlled backfills. Removal of about 4 to 8 feet
of surface swelling soil and replacement with nonexpansive, low permeable backfill will reduce heave at
the ground surface. Backfills adjacent to foundation
walls should also be nonswelling, low permeable material. Nonswelling material minimizes the forces exerted on walls, while low permeable backfill minimizes
infiltration of surface water through the backfill into
the foundation soil. If only pervious, nonexpansive
(granular) backfill is available, a subsurface drain at
the bottom of the backfill is necessary to carry off in-

TM 5-018-7

filtrated water (fig. 7-1) and to minimize seepage of

water into deeper desiccated foundation expansive
soils.
(1) Backfill of natural soil. Backfill using natural
soil and compaction control has been satisfactory in
some cases if nonswelling backfill is not available.
However, this use of backfill should be a last resort,
(a) In general, the natural soil should be compacted to 90 percent of standard maximum density
and should be wet of optimum water content. Foundation loads on fills should be consistent with the allowable bearing capacity of the fill. Overcompaction
should be avoided to prevent aggravating potentially
swelling soil problems such as differential heave of the
fill. Compaction control of naturally swelling soil is usually difficult to accomplish in practice. Some soils become more susceptible to expansion following remolding, and addition of water to achieve water contents
necessary to control further swell may cause the soil to
be too wet to work in the field.
(b) As an alternative, backfills of lime-treated
natural soil compacted to 95 percent standard maximum density at optimum water content may be satisfactory if the soil is sufficiently reactive to the lime (d
below), Lime treatment may also increase soil strength
and trafficability on the construction site.
(2) Backfill adjacent to walls. A IV on lH slope
cut into the natural soil should dissipate lateral swell
pressures against basement or retaining walls exerted
by the natural swelling material. The nonswelling
backfill should be a weak material (sand fill with friction angle of 30 degrees or lessor cohesive fill with cohesion less than about 0.5 tons per square foot) to allow the fill to move upward when the expansive natural soil swells laterally. Restraining loads should not
be placed on the surface of the fill. A friction reducing
medium may be applied on the wall to minimize friction between the wall and the backfill, TM 5-818-4
discusses details on optimum slopes of the excavation
and other design criteria.
b. Maintenance. Maintenance programs are directed toward promoting uniform soil moisture beneath the foundation. A good program consists of the
following:
(1) Maintenance of a positive slope of about 5 percent around the structure for drainage and elimination
of water catch areas.
(2) Maintenance of original drainage channels and
installation of new channels as necessary.
(3) Maintenance of gutters around the roof and diversion of runoff away from the structure.
(4) Avoidance of curbs or other water traps
around flower beds.
(5) Elimination of heavy vegetation within 10 to
15 feet of the foundation or 1 to 1.5 times the height
of mature trees.

(6) Uniform limited watering around the structure during droughts to replace lost moisture.
c. Moisture barriers. The purpose of moisture barriers is to promote uniform soil moisture beneath the
foundation by minimizing the loss or gain of moisture
through the membrane and thus reducing cyclic edge
movement, Moisture may still increase beneath or
within areas surrounded by the moisture barriers leading to a steady but uniform heave of the foundation or
slab-on-grade.
(1) Types of barriers. These barriers consist of
horizontal and vertical plastic and asphalt membranes
and granular materials. Concrete is an ineffective
moisture barrier. Longlasting membranes include
chlorinated polyethylene sheets, preferably placed
over a layer of catalytically blown or sprayed asphalt.
All joints, seams, and punctures should be sealed by
plastic cements or concrete/asphalt joint sealants.
ASTM D 2521 (Part 15) describes use of asphalt in
canal, ditch, and pond linings (app A).
(2) Horizontal.
(a) An impervious membrane on the ground surface in a crawl space where rainfall does not enter may
help reduce shrinkage in clayey foundation soils with
deep groundwater levels by minimizing evaporation
from the soil. A vapor barrier should not be placed in
ventilated crawl spaces if there is a shallow water table or if site drainage is poor because heave maybe aggravated in these cases. Figure 7-2 illustrates a useful
application of horizontal membranes,
(b) Other applications include the use of horizontal moisture barriers around the perimeter of structures to reduce lateral variations in moisture changes
and differential heave in the foundation soil. Plastic or
other thin membranes around the perimeter should be
protected from the environment by a 6- to 12-inchthick layer of earth.
(c) A disadvantage of these barriers is that they
are not necessarily reliable and may be detrimental in
some cases. For example, most fabrics and plastic
membranes tend to deteriorate with time. Undetected
(and hence unrepaired) punctures that allow water to
get in, but not to get out, commonly occur in handling
on placement. Punctures may also occur during planting of vegetation. If the barrier is a concrete slab, the
concrete may act as a wick and pull water out of the
soil.
(3) Vertical.
(a) Plumbing or utility trenches passing
through the barrier may contribute to soil moisture beneath the foundation.
(b) The vertical barrier (fig. 7-3) should extend
to the depth of the active zone and should be placed a
minimum of 3 feet from the foundation to simplify
construction and to avoid disturbance of the foundation soil. The barrier may not be practical in prevent-

1-3

ing migration of moisture beneath the bottom edge for

active zones deeper than 8 to 10 feet. The granular barrier may also help reduce moisture changes during
droughts by providing a reservoir of moisture. The
placement of a filter fabric around the trench to keep
fine particles from entering the perforated pipe will
permit use of an open coarse aggregate instead of a
graded granular filter. In some cases, the perforated
pipe could be eliminated from the drain trench.
d. Lime treatment. This treatment is the most
widely used and most effective technique of chemical
alteration to minimize volume changes and to increase
the shear strength of foundation expansive soils.
(1) Applications. Lime treatment is applied to the
strengthening and minimization of volume change of
soil in railroad beds, pavement subgrades, and slopes.
When this treatment is applied to foundation soils of
single and multistory structures, it is not always successful because the usefulness depends on the reactiveness of the soil to lime treatment and the thoroughness of dispersion of lime mixed into the soil.
(a) Lime treatment is effective in the minimization of volume changes of natural soil for backfill.
However, this treatment increases the soil permeability and the soil strength. The soil permeability should
be kept low to restrict seepage of surface water
through the backfill. The backfill strength should be as
low as possible compatible with economical design to
minimize the transfer of lateral swell pressures from

the natural in situ soil through the backfill to the basement and retaining walls.
(b) Lime treatment may be used to stabilize a 6to 12-inch layer of natural expansive soil compacted on
the surface of the construction site to provide a positive slope for runoff of water from the structure and a
layer to reduce differential heave beneath the floor
slab.
(c) Lime treatment may be applied to minimize
downhill soil creep of slopes greater than 5 degrees (9
percent) by increasing the stiffness and strength of the
soil mass through filling fractures in the surface soils.
If lime slurry pressure injection (LPSI) can cause a
lime slurry to penetrate the fissures in the soil mass to
a sufficient depth (usually 8 to 10 feet), then the limefilled seams will help control the soil water content, reduce volumetric changes, and increase the soil
strength. However, LSPI will probably not be satisfactory in an expansive clay soil that does not contain an
extensive network of fissures because the lime will not
penetrate into the relatively impervious soil to any appreciable distance from the injection hole to form a
continuous lime seam moisture barrier.
(d) LSPI may be useful for minimization of
movement of fissured foundation expansive soils down
to the depth of the active zone for heave or at least 10
ft. The lime slurry is pressure injected on 3- to 5-foot
center to depths of 10 to 16 feet around the perimeter
of the structure 3 to 5 feet from the structure.

(2) Soil mixture preparation. Lime should be thoroughly and intimately mixed into the soil to a sufficient depth to be effective. For stabilization of expansive clay soils for foundations of structures, mixing
should be done down to depths of active zone for
heave. In practice, mixing with lime is rarely done
deeper than 1 to 2 feet. Therefore, lime treatment is
normally not useful for foundations on expansive soil
except in the above applications. Moreover, poor mixing may cause the soil to break up into clods from normal exposure to the seasonal wetting/drying cycles.
The overall soil permeability is increased and provides
paths for moisture flow that require rapid drainage
from this soil. Lime treatment should be performed by
experienced personnel.
(3) Lime modification optimum content (LMO).
The LMO corresponds to the percent of lime that maximizes the reduction in the soil plasticity or the PI. The
reduction in plasticity also effectively minimizes the
volume change behavior from changes in water content and increases the soil shear strength.
(a) A decision to use lime should depend on the
degree of soil stabilization caused by the lime. Lime

treatment is recommended if a 50 percent reduction in

the PI is obtained at the LMO content (table 7- 1). The
PI should be determined for the natural soil, LMO,
LMO+ 2, and LMO - 2 percent content.
(b) The increase in strength of the lime-treated
soil should be similar for soil allowed to cure at least 2
or more days following mixing and prior to compaction
to similar densities.
(c) The amount of lime needed to cause the optimum reduction in the PI usually varies from 2 to 8 percent of the dry soil weight.
e. Cement treatment. Cement may be added to the
soil to minimize volume changes and to increase the
shear strength of the foundation expansive soil if the
degree of soil stabilization achieved by lime alone is
not sufficient. The amount of cement required will
probably range between 10 to 20 percent of the dry
soil weight. A combination of lime-cement or lime-cement-fly ash may be the best overall additive, but the
best combination can only be determined by a laboratory study. TM 5-822-4 presents details on soil stabilization with cement and cement-lime combinations.

TM 5-818-7

CHAPTER 8
CONSTRUCTION TECHNIQUES AND INSPECTION
8-1. Minimization of foundation problems from construction
Many problems and substandard performance of foundations observed in structures on expansive soils occur
from poor quality control and faulty construction practice. Much of the construction equipment and procedures that are used depends on the foundation soil
characteristics and soil profiles. Careful inspection
during construction is necessary to ensure that the
structure is built according to the specifications.
a.

Important elements of construction techniques.

Construction techniques should be used that promote

a constant moisture regime in the foundation soils during and following construction. The following elements of construction are important in obtaining adequate foundation performance in expansive soils.
(1) Excavations. The excavation should be completed as quickly as possible to the design depth and
protected from drying. An impervious moisture barrier should be applied on the newly exposed surfaces of
the excavation to prevent drying of the foundation
soils immediately after excavating to the design depth.
Sides of the excavation should be constructed on a lV
on lH slope or an appropriate angle that will not transmit intolerable swelling pressures from the expansive
soil to the foundation. The foundation should be constructed in the excavation as quickly as practical.
(2) Selection of materials. Selected materials
should conform to design requirements.
(a) Backfills should be nonswelling materials.
(b) Concrete should be of adequate strength and
workability.
(c) Reinforcing steel should be of adequate size
and strength.
(d) Moisture barriers should be durable and impervious.
(3) Placement of materials. All structural materials should be positioned in the proper location of the
foundation.
(4) Compaction of backfills. Backfills of natural
expansive soil should be compacted to minimize effects
of volume changes in the fill on performance of the
foundation. Backfills should not transmit intolerable
swell pressures from the natural expansive foundation
soil to basement or retaining walls.
(5) Drainage during construction. The site should
be prepared to avoid ponding of water in low areas.
Consideration should be given to compaction of 6 to 12

inches or more of impervious nonswelling soil on the

site prior to construction of the foundation to promote
drainage and trafficability on the site. Dehydrated
lime may also be sprinkled on the surface of expansive
soil to promote trafficability. Sumps and pumps
should be provided at the bottom of excavations if necessary to remove rainwater or subsurface drainage entering the excavation. Provision for after normal duty
operation of the pumps should also be made.
(6) Permanent drainage. Grades of at least 1 percent and preferably 5 percent, to promote drainage of
water away from the structure, should be provided
around the perimeter of the structure. Low areas
should be filled with compacted backfill. Runoff from
roofs should be directed away from the structure by
surface channels or drains. Subsurface drains should
be constructed to collect seepage of water through pervious backfills placed adjacent to the foundation.
b. Considerations of construction inspection. Table
8-1 lists major considerations of construction inspection. Inspections related to concrete reinforced slab
and drilled shaft foundations, the two most commonly
used foundations in expansive soil areas, are discussed
below.
8-2. Stiffened slab foundations
Items in table 8-2 should be checked to minimize defective slab foundations.
a. The inspector should check for proper site prep
aration and placement of the moisture barrier, steel,
and concrete. All drainage systems should be inspected
for proper grade and connections to an outlet.
b. Posttensioned slabs require trained personnel
and careful inspection to properly apply the posttensioning procedure. For example, anchors for the steel
tendons should be placed at the specified depth (lower
than the depth of the tensioning rods) to avoid pullout
during tensioning. Tendons should be stressed 3 to 18
days following the concrete placement (to eliminate
much of the shrinkage cracking) such that the minimum compressive stress in the concrete exceeds 50
pounds per square inch. Stressing should be completed
before structural loads are applied to the slabs.
8-3. Drilled shaft foundations
Items in table 8-3 should be checked to minimize defective shaft foundations. The foundation engineer

TM 5-818-7

should visit the construction site during boring of the

first shaft holes to verify the assumptions regarding
the subsurface soil profile, e.g., the nature and location
of the subsoils. Periodically, he or she should also
check the need for the designer to consider modifications in the design.
a. Location of shaft base. The base of the shaft is located in the foundation soils to maintain shaft movements within tolerable limits. This depth depends on
the location and thickness of the expansive, compressible or other unstable soil, sand lenses or thin permeable zones, depth to groundwater, and depth to foundation soil of adequate bearing capacity. The design
depth may require modification to relocate the base in
the proper soil formation of adequate bearing capacity
and below the active zone of heave. The purpose of locating the base of the shaft in the proper soil formation should be emphasized to the inspector during the
first boring of the drilled shaft foundation. Underreams may be bored in at least l.5-foot-diameter (preferably 2.5-foot) dry or cases holes where inspections
are possible to ensure cleanliness of the bottom.
b. Minimization of problems, Long experience has
shown that drilled shaft foundations are reliable and
economical. Nevertheless, many problems are associated with these foundations and can occur from inadequate understanding of the actual soil profile and
groundwater conditions, mistakes made while drilling,
inadequate flow of concrete, and improper reinforcement.
(1) Inadequate information.
(a) Site conditions should be known to permit
optimum selection of equipment with the required mobility.
(b) Subsurface conditions should be known to
permit selection of equipment with adequate boring
capacity.
(c) Type of soil (e.g., caving and pervious strata)
may require slurry drilling. Specifications should permit sufficient flexibility to use slurry for those soil
conditions where it maybe needed.
(d) Previously unnoticed sand lenses or thin
permeable zones in otherwise impervious clay may
cause problems during construction of drilled shafts.
Seepage through permeable zones may require casing
or slurry and may render construction of an underream nearly impossible.
(e) Overbreak or the loss of material outside of
the nominal diameter of the shaft due to caving soil is
a serious problem that can cause local cavities or defects in the shaft. The construction procedure (boring
dry, with casing, or using slurry) should be chosen to
minimize overbreak.
(2) Problems with the dry method. Caving,
squeezing soil, and seepage are the most common prob8-2

lems of this method. Stiff or very stiff cohesive soils

with no joints or slickensides are usually needed. Underreams are vulnerable to caving and should be constructed as quickly as possible.
(3) Problems with the casing method. Slurry
should be used while drilling through caving soil prior
to placement of the casing and sealing of the casing in
an impervious layer. An impervious layer is necessary
to install the bottom end of the casing.
(a) Casing should not be pulled until the head of
concrete is sufficient to balance the water head external to the casing; otherwise, groundwater may mix
with the concrete.
(b) Squeezing or localized reduction in the borehole diameter on removal of the casing can be minimized by using a relatively high slump concrete with a
sufficient pressure head.
(c) Casing sometimes tends to stick in place during concrete placement. If the concrete appears to be
setting up, attempts to shake the casing loose should
be abandoned and the casing left in place to avoid the
formation of voids in the shaft when the casing is
pulled.
(d) Steel reinforcement should be full length to
avoid problems in downdrag of the reinforcement
while the casing is pulled. The reinforcement cage
should also be full length if uplift forces are expected
on the drilled shaft from swelling soil.
(4) Problems with the slurry method. Slurry of
sufficient viscosity is used to avoid problems with caving soils. A rough guide to appropriate slurry viscosities is given by a Marsh cone funnel test time of about
30 seconds for sandy silts and sandy clays to 50 seconds for sands and gravels. The Marsh cone test time
is the time in seconds required to pour 1 quart of
slurry through the funnel. The workability of the
slurry should also be adequate to allow complete displacement of the slurry by the concrete from the
perimeter of the borehole and steel of the rebar cage.
(a) Slurries should be of sufficient viscosity to
eliminate settling of cuttings. Loose cuttings adhering
to the perimeter of the hole can cause inclusions and a
defective shaft.
(b) The tremie sometimes becomes plugged,
stopping the flow of concrete into the borehole. The
tremie should not be pulled above the concrete level in
the shaft before the concrete placement is completed,
otherwise inclusions may occur in the shaft following
reinsertion of the tremie into the concrete.
(c) The reinforcement cage may move up if the
tremie is too deep in the concrete or the concrete is
placed too rapidly.
c. Placement of concrete. Concrete strength of at
least 3,000 pounds per square inch should be used and
placed as soon as possible on the same day as drilling
the hole. Concrete slumps of 4 to 6 inches and limited

TM 5-818-7

aggregate size of one third of the rebar spacing are recommended to facilitate flow of concrete through the
reinforcement cage and to eliminate cavities in the
shaft. Care should be exercised while placing the concrete to ensure the following:
(1) Continuity while pulling the casing.

(2) Tip of tremie always below the column of

freshly placed concrete in wet construction; no segregation in a dry hole.
(3) Adequate strength of the rebar cage to minimize distortion and buckling.

Construction

CHAPTER 9
REMEDIAL PROCEDURES
9-1. Basic considerations
Remedial work for damaged structures is usually difficult to determine because the cause of the problem
(e.g., location of source or loss of soil moisture, and
swelling or settling/shrinking soil) may not be readily
apparent, A plan to fix the problem is often difficult to
execute, and the work may have to be repeated because
of failure to isolate the cause of the moisture changes
in the foundation soil, An effective remedial procedure
may not be found until several attempts have been
made to eliminate the differential movement. Requirements for minimizing moisture changes (chap. 7) are
therefore essential. The foundation should have sufficient capacity to maintain all distortion within tolerable limits acceptable to the superstructure. This distortion occurs from differential heave for the most severe
climates and changes in the field environment.
a. Specialized effort. Investigation and repair are
therefore specialized procedures that usually require
much expertise and experience. Cost of repair work
can easily exceed the original cost of the foundation.
The amount of damage that requires repair also depends on the attitudes of the owner and occupants to
tolerate distortion as well as damage that actually impairs the usefulness and safety of the structure.

be allowed to adjust, following completion of remedial

measures, for at least a year before cosmetic work is
done. The structure is seldom rebuilt to its original
condition, and in some instances, remedial measures
have not been successful.

fects of swelling soil tends to be cosmetic rather than

structural, and repairs are usually more economical
than rebuilding as long as the structure remains
sound. At-early signs of distress, remedial action to
minimize future distortion should be undertaken and
should be given a greater priority than the cosmetic repairs as this action will minimize maintenance work
over the long term. Maintenance expenses and frequency of repairs tend to be greatest in lightly loaded
structures and residences about 3 to 4 years following
the original construction. Overall maintenance can be
minimized by taking remedial action to minimize future distortion before extensive repairs are required
(e.g., breaking out and replacing sections of walls).

well as the time movements first became noticeable,

should be determined, Most cracks caused by differential heave are wider at the top than at the bottom.
Nearly all lateral separation results from differential
heave. Diagonal cracks can indicate footing or drilled
shaft movement, or lateral thrust from the doming
pattern of heaving concrete slabs. Fractures in slabson-grade a few feet from and parallel with the perimeter walls also indicate heaving of underlying soils. Level surveys can be used to determine the trend of movement when prior survey records and reliable benchmarks are available. Excavations may be necessary to
study damage to deep foundations, such as cracks in
shafts from uplift forces.
c. Sources of moisture. The source of soil moisture
that led to the differential heave should be determined
to evaluate the cause of damage. Location of deeprooted vegetation, such as shrubs and trees, location and
frequency of watering, inadequate slopes and pending,
seepage into foundation soil from surface or perched
water, and defects in drain, water, and sewer lines can

c. Examples of remedial procedures. The choice of

remedial measures is influenced by the results of site
and soil investigations as well as by the type of original construction. Table 9-1 illustrates common remedial measures that can be taken. Only one remedial
procedure should be attempted at a time so as to determine its effect on the structure. The structure should

9-2. Evaluation of information

All existing information on the foundation soils and
design of the foundation and superstructure should be
studied before proceeding with new soil investigations.
a. Foundation conditions. The initial soil moisture
at time of construction, types of soil, soil swell potentials, depth to groundwater, type of foundation and
superstructure, and drainage system should be determined. The current soil moisture profile should also be
determined. Details of the foundation, such as actual
bearing pressures, size and length of footings, and slab
and shaft reinforcing, should also be collected. Drilling
logs made during construction of shaft foundations
may be used to establish soil and groundwater conditions and details of shaft foundations. Actual construction should be checked against the plans to identify
any variances.

9-1

TM 5-818-7

make important changes in soil moisture and can lead

to differential heave.
9-3. Stiffened slab foundations
Most slab foundations that experience some distress
are not damaged sufficiently to warrant repairs. Damage is often localized by settlement or heave of one side
of the slab. The cause of the soil movement, whether
settlement or heave, should first be determined and
then corrected.
a. Stabilization of soil moisture. Drainage improvements and a program to control soil moisture at the
perimeter of the slab are recommended (chap 7) for all
damaged slab foundations.
b. Remedial procedures. Remedial work on slabs depends on the type of movement, Repair of a settled
area requires raising of that area, while repair of a
heaved area often requires raising the entire unheaved
portion of the slab up to the level of the heaved portion. Repair costs are consequently usually greater for
heaving than settling cases.
(1) Repair of a damaged slab consists of a combination of underpinning and mudjacking using a
cement grout. Mudjacking using a cement grout is required simultaneously with underpinning to fill voids
during leveling of the slab. Fractured slabs are usually
easier to repair than unfractured slabs that have been
distorted by differential movement because usually
only the fractured portion of the slab requires treatment. The distortion of unfractured slabs can also
cause considerable damage to the superstructure and
inconvenience to the occupants.
(2) Underpinning and mudjacking are applied simultaneously and usually clockwise around the slab

until all parts of the foundation are at the same elevation. If a heaved area is lowered to the same elevation
as the rest of the foundation, such as to repair a mushroomed or dome-shaped heave pattern, the slab is first
supported before digging out the soil to prevent the
slab from creeping down on the work crew during the
digging. Attempts at leveling dome-shaped distortion
by raising the perimeter may be unsuccessful because
mudjacking usually causes the entire slab to rise.
9-4. Drilled shaft foundations
Most damage to structures with shaft foundations consists of fractured slabs-on-grade. The shaft may contribute to the damage caused by migration of moisture
down the shaft/soil interface into swelling soil beneath
the shaft footing. The fracture pattern of open cracks
in the floor slab parallel to and several feet from the
wall often shows that the slab had not been free to
move near the walls. Damage to drilled shafts is often
caused by upward movement of the shaft from swelling soil beneath its base and by uplift forces on the
shaft perimeter from adjacent swelling soil.
ments and a program to control soil moisture around
the perimeter of the foundation are recommended
(chap 7).
removal of the slab and underlying wet soil, replacement with nonswelling soil, and placement of a new
slab isolated from the perimeter walls. Repair of
drilled shafts consists of cutting down the top of the
shaft and releveling the foundation. The tops of the
drilled shafts are cut to the elevation of the top of the
lowest shaft where possible.

Table 9-1. Remedial Measures

Measure

9-2

Description

Drainage

Slope ground surface (positive drainage) from structure; add drains

for downspouts and outdoor faucets in areas of poor drainage,
and discharge away from foundation soil; provide subdrains if
perched water tables or free flow of subsurface water are problems; provide flexible, watertight utility connections.

Moisture stabilization
(maintenance of
constant moisture
whether at high or
low levels)

Remove natural swelling soil and recompact with impervious, nonswelling backfill; install vertical and/or horizontal membranes
around the perimeter; locate deep-rooted vegetation outside of
moisture barriers; avoid automatic sprinkling systems in areas
protected with moisture barriers; provide a constant source of
moisture if a combination of swelling/shrinking soils is occurring; thoroughly mix 4 to 8 percent lime into soil to reduce
potential for swell or pressure-inject line slurry around
the perimeter of the structure.

Superstructure
adjustments

Free slabs from foundation by cutting along foundation walls; provide slip joints in interior walls and door frames; reinforce masonry and concrete block walls with horizontal and vertical tie
bars or reinforced concrete beams; provide fanlights over doors
extended to the ceiling.

TM 5-818-7
Table 9-1. Remedial MeasuresContinued
Measure

Description

Spread footings and

deep foundation
adjustments

Decrease footing size; underpin with deep shafts; mudjack using a

cement grout; reconstruct void beneath grade beams; eliminate
mushrooms at top of shafts; adjust elevation by cutting the top
of the shaft or by adding shims; increase footing or shaft
spacing to concentrate loading forces and to reduce angular distortion from differential heave between adjacent footings and
shafts.

Continuous wall
foundation
adjustments

Provide voids beneath portions of wall foundation; posttension; reinforce with horizontal and vertical tie bars or reinforced concrete
beams.

Reinforced and
stiffened slab-ongrade adjustments

Mudjack using a cement grout; underpin with spread footings or

shafts to jack up the edge of slabs,

TM 5-818-7

APPENDIX A
REFERENCES
Government

Publications

Department of the Army

Technical Manuals
TM 5-818-1
TM5-818-4
TM5-822-4

Procedures for Foundation Design of Buildings and Other Structures (Except

Hydraulic Structures)
Backfill for Subsurface Structures
Soil Stabilization for Roads and Streets

Department of the Navy

Change

A-1

TM 5-818-7

APPENDIX B
CHARACTERIZATION OF SWELL BEHAVIOR FROM SOIL SUCTION
B-1. Introduction
Soil suction is a quantity that can be used to characterize the effect of moisture on volume, and it is a measure of the energy or stress that holds the soil water in
the pores or a measure of the pulling stress exerted on
the pore water by the soil mass. The total soil suction
is expressed as a positive quantity and is defined as the

to the geometrical configuration of the soil and structure, capillary tension in the pore water, and water
sorption forces of the clay particles. This suction is also pressure-dependent and assumed to be related to
(B-1)
(B-2)
matrix soil suction, tons per square foot
compressibility factor, dimensionless
total mean normal confining pressure,
tons per square foot
ratio of total horizontal to vertical stress
in situ
total vertical pressure, tons per square
foot
The exponent
out confining pressure except atmospheric pressure.
Experimental results show that the in situ matrix sucity factor is determined by the procedure in paragraph
B-3d.
by the concentration of soluble salts in the pore water,
and it is pressure-independent. The effect of the osmotic suction on swell is not well known, but an osmotic effect may be observed if the concentration of soluble salts in the pore water differs from that of the externally available water. For example, swell may occur
in the specimen if the external water contains less
soluble salts than the pore water. The effect of the osmotic suction on swell behavior is assumed small compared with the effect of the matrix suction. The osmotic suction should not significantly affect heave if the
salt concentration is not altered.

B-2. Methods of measurement

Two methods are recommended for determining the
total soil suction: thermocouple psychrometer and filter paper. The suction range of thermocouple psychrometers usually is from 1 to 80 tons per square foot
while the range of filter paper is from 0.1 to more than
1,000 tons per square foot. Two to seven days are required to reach moisture equilibrium for thermocouple
psychrometer, while 7 days are required for filter
paper. The thermocouple psychrometer method is simple and can be more accurate than filter paper after
the equipment has been calibrated and the operating
procedure established. The principal disadvantage is
that the suction range is much more limited than the
filter paper method, The filter paper method is technically less complicated than the thermocouple psychrometer method; however, the weighing procedure
required for filter paper is critical and vulnerable to
large error.
a. Calibration. The total soil suction is given on the
(B-3)
pressure
tons per
universal gas constant, 86.81 cubic centimetres-tons per square foot/mole-Kelvin
absolute temperature, Kelvin
volume of a mole of liquid water, 18.02
cubic centimetres/mole
relative humidity
pressure of water vapor, tons per
square foot
pressure of saturated water vapor, tons
per square foot
Equation (B-3) shows that the soil suction is related to
the relative humidity in the soil. Both thermocouple
psychrometer and filter paper techniques require calibration curves to evaluate the soil relative humidity
from which the soil suction may be calculated using
equation (B-3). Calibration is usually performed with
salt solutions of various known molality (moles of salt
per 1,000 grams of water) that produce a given relative humidity. Table B-1 shows the modalities reB-1

TM 5-818-7
Table B-1. Calibration Salt Solutions
Suction, tsf
for cited molality of sodium chloride solution

Measured
temperature
15
20
25
30

0.053

0.100

0.157

0.273

0.411

0.550

1.000

3.05
3.10
3.15
3.22

4.67
4.74
4.82
4.91

7.27
7.39
7.52
7.64

12.56
12.75
13.01
13.22

18.88
19.22
19.55
19.90

25.29
25.76
26.23
26.71

46,55
47.50
48.44
49.37

quired for sodium chloride salt solutions to provide the

soil suctions given as a function of temperature.
b. Thermocouple psychrometer technique. The thermocouple psychrometer measures relative humidity in
soil by a technique called Peltier cooling. By causing a
current to flow through a single thermocouple junction
in the proper direction, that particular junction will
cool, then water will condense on it when the dewpoint temperature is reached. Condensation of this
water inhibits further cooling of the junction. Evaporation of condensed water from the junction after the
cooling current is removed tends to maintain a difference in temperature between the thermocouple and
the reference junctions. The microvoltage developed
between the thermocouple and the reference junctions
is measured by the proper readout equipment and related to the soil suction by a calibration curve.
(1) Apparatus. Laboratory measurements to evaluate total soil suction may be made with the apparatus
illustrated in figure B-1. The monitoring system includes a cooling circuit with the capability of immediate switching to the voltage readout circuit on termination of the current (fig. B-2). The microvoltmeter
(item 1, fig. B-2) should have a maximum range of at
least 30 microvolt and allow readings to within 0.01
microvolt. The 12-position rotary selector switch
(item 2) allows up to 12 simultaneous psychrometer
connections. The 0-25 millimeter (item 3), two
1.5-volt dry cell batteries (item 4), and the variable
potentiometer (item 5) form the cooling circuit, Equipment is available commercially to perform these measurements of soil suction.
(2) Procedure.
(a) Thermocouple psychrometer are inserted
into 1-pint-capacity metal containers with the soil
specimens, and the assembly is sealed with No. 13-1/2
rubber stoppers. The assembly is inserted into a 1- by
1- by 1.25-foot chest capable of holding six l-pint
containers and insulated with 1.5 inches of foamed
polystyrene, Cables from the psychrometer are
passed through a 0.5-inch-diameter hole centered in
the chest cover, The insides of the metal containers are
coated with melted wax to inhibit corrosion of the containers.
(b) The apparatus is left alone until equilibrium
is attained. Temperature equilibrium is attained within a few hours after placing the chest cover. Time to
reach equilibrium of the relative humidity in the air

measured by the psychrometer and the relative humidity in the soil specimen depends on the volume and initial relative humidity in the container. Equilibrium
time may require up to 7 days, but may be reduced to 2
or 3 days by repeated testing of soils with similar suctions.
(c) After equilibrium is attained, the microvoltmeter is set on the 10- or 30-microvolt range and
zeroed by using a zeroing suppression or offset control.
The cooling current of approximately 8 millimeters is
applied for 15 seconds and then switched to the microvoltmeter circuit using the switch of item 6 in figure
B-2, The maximum reading on the microvoltmeter is
recorded. The cooling currents and times should be
identical to those used to determine the calibration
curves.
(d) The readings can be taken at room temperature, preferably from 20 to 25 degrees Centigrade, and
corrected to a temperature of 25 degrees Centigrade
by the equation
(B-4)
where
microvolt at 25 degrees Centigrade
microvolt at t degrees Centigrade
Placement of the apparatus in a constant temperature
room will increase the accuracy of the readings.
(3) Calibration, The psychrometer are calibrated
by placing approximately 50 millilitres of the salt solutions of known molality (table B-1) in the metal containers and following the procedure in b(2) above to determine the microvolt output. Equilibration time may
be reduced to 2 or 3 days by surrounding the psychrometer with filter paper soaked with solution. The
suctions given for the known modalities are plotted
versus the microvolt output for a temperature of 25
degrees Centigrade. The calibration curves of 12 commercial psychrometer using the equipment of figure
B-1 were within 5 percent and could be expressed by
the equation
(B-5)
To
foot. The calibration curves using other equipment
may be somewhat different.
c. Filter paper technique. This method involves en-
closing filter paper with a soil specimen in an airtight
container until complete moisture equilibrium is

reached. The water content in percent of the dry

weight is subsequently determined, and the soil suction is found from a calibration curve.
(1) Apparatus. Materials consist of 2-inch-diameter filter paper, 2-inch-diameter tares, and a gravimetric scale accurate to 0.001 g. A filter paper is enclosed in an airtight container with the soil specimen.
(2) Procedure.
(a) The filter paper disc is pretreated with 3 percent reagent grade pentachlorophenol in ethanol (to inhibit bacteria and deterioration) and allowed to air
dry. Reagent grade pentachlorophenol is required because impurities in the treatment solution influence
the calibration curve. Care is required to keep the filter paper from becoming contaminated with soil from
the specimen, free water, or other contaminant (e.g.,
the filter paper should not touch the soil specimen,
particularly wetted specimens).
(b) Seven days are required to reach moisture
equilibrium in the airtight container. At the end of 7
days, the filter paper is transferred to a 2-inch-cliameter covered tare and weighed immediately on a
gravimetric scale accurate to 0.001 g. The number of

filter papers and tares weighed at one time should be

kept small (nine or less) to minimize error caused by
water evaporating from the filter paper.
(c) The tare is opened and placed in an oven for
5 degrees Centigrade. The ovendry weight of the filter
paper is then determined, and the water content as a
percent of the dry weight is compared with a calibration curve to determine the soil suction.
(3) Calibration. The ovendry water content of the
filter paper is dependent on the time lapse following
removal from the drying oven before weighing.
(a) The calibration curves shown in figure B-3
were determined for various elapsed times following
removal from the oven. The calibrations are given for
Fisherbrand filter paper, Catalog Number 9-790A, enclosed with salt solutions of various molality for 7
days. Calibration curve No. 1 resulted from weighing
the filter paper 5 seconds following removal from the
oven. Time lapses of 15 minutes and 4 hours lead to a
similar calibration curve (No. 3) of significantly smaller water contents than the 5-second curve for identical suctions. Calibration curve No. 2 was determined

TM 5-818-7

by removing 12 specimens from the oven, waiting 30

seconds to cool, then weighing as soon as possible and
within 15 minutes.
100

(b) Calibration curves based on the method used

to determine curve No, 3 with a waiting time between
15 and 30 minutes are recommended if the suctions of

B-6

TM 5-018-7

(B-9)

if the equilibrium moisture profiles of figure 5-1 (para

5-4b) are used.
b. Initial matrix suction. The initial matrix suction

surcharge pressure may be evaluated using

the soil suction test procedure on undisturbed specimens or may be calculated from equation (B-7) and
the natural (initial) water content.
T&without

without surcharge pressure may be calculated from

the assumption
(B-8)
coefficient of effective lateral earth pressure
final vertical effective pressure, tons per
square foot or from equation (B-1) setting
a
The final vertical effective pressure may be found
from

(5-3), (5-4), or (5-5).

d. Compressibility factor. The compressibility factor a is the ratio of the change in volume for a corresponding change in water content, i.e., the slope of the

density. The value of a for highly plastic soils is close

to 1, and much less than 1 for sandy and low plasticity
soils. High compressibility y factors can indicate highly
swelling soils; however, soils with all voids filled with
water also have a equal to 1.
(1) Figure B-5 illustrates the compressibility factor calculated from laboratory data for a silty clay
taken from a field test section near Clinton, Mississippi. Extrapolating the line to zero water content, as
shown in the figure, provides an estimate of l/R with

B-7

TM 5-818-7

heave for this case will be

CT=

(0.93) (2.79)
(100) (0,046)

= 0.564

vantage of this latter approach is that the equilibrium

matrix suction or pore water pressure profile is not
known, except that the final matrix suction will be
small and probably close to the saturated profile (equation (5-3)). The program HEAVE will compute the potential heave for this case as well as those shown in figure 5-1.

TM 5-818-7

TM 5-818-7

APPENDIX C
FRAME AND WALL CONSTRUCTION DETAILS
Figures C-1 through C-10 illustrate types of construction for expansive foundation soils. These figures were taken from U.S. Army Corps of Engineers Construction Engineering and Research Laboratory Technical Report M-81. The figures show practical wall ties to concrete and steel beams, wall connections with
control joints, details of interior partitions, bar joist first floor framing with grade
beams, and stiffened mat foundations.

c-s

TM 5-810-7

TM 5-818-7

C-7

TM 5-818-7

APPENDIX D

D-1

TM 5-818-7

D-2

TM 5-818-7

D-3

TM 5-818-7

The proponent agency of this publication is the Office of the Chief

of Engineers, United States Army. Users are invited to send comments and suggested improvements on DA Form 2028 (Recommended Changes to Publications and Blank Forms) direct to HQDA
(DAEN-ECE-G), WASH DC 20314.

By Order of the Secretary of the Army:

JOHN A. WICKHAM, JR.

General, United States Army
Chief of Staff

. Official:

ROBERT M. JOYCE
Major General, United States Army
The Adjutant General

DISTRIBUTION:
To be distributed in accordance with DA Form 12-34B, requirements for TM
5-800 Series: Engineering and Design for Real Property Facilities,