Chapter 5 Contents

Page
5.1 Overview 5-3

5.2 Influence of Existing and Future Conditions on Soil and Rock Properties 5-3

5.3 Methods of Determining Soil and Rock Properties 5-4

5.4 In-Situ Field Testing 5-4

5.4.1 Well Pumping Tests 5-6
5.4.2 Packer Permeability Tests 5-6
5.4.3 Seepage Tests 5-6
5.4.4 Slug Tests 5-7
5.4.5 Piezocone Tests 5-7
5.4.6 Flood Tests 5-7

5.5 Laboratory Testing of Soil and Rock 5-8

5.5.1 Quality Control for Laboratory Testing 5-8
5.5.2 Developing the Testing Plan 5-9

5.6 Engineering Properties of Soil 5-10

5.6.1 Laboratory Index Property Testing 5-10
5.6.2 Laboratory Performance Testing 5-10
5.6.3 Correlations to Estimate Engineering Properties of Soil 5-12

5.7 Engineering Properties of Rock 5-13

5.8 Final Selection of Design Values 5-13

5.8.1 Overview 5-13
5.8.2 Data Reliability and Variability 5-14
5.8.3 Final Property Selection 5-16
5.8.4 Development of the Subsurface Profile 5-16
5.8.5 Selection of Design Properties for Engineered Materials 5-17

5.9 Properties of Predominant Geologic Units in Washington 5-20

5.9.1 Loess 5-20
5.9.2 Peat/Organic Soils 5-21
5.9.3 Glacial Till and Glacial Advance Outwash 5-22
5.9.4 Colluvium/Talus 5-23
5.9.5 Columbia River Sand 5-23
5.9.6 Columbia Basin Basalts 5-24
5.9.7 Latah Formation 5-24
5.9.8 Seattle Clay 5-25

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-1
Engineering Properties of Soil and Rock

Page
5.9.9 Bellingham Glaciomarine Drift 5-27
5.9.10 Coastal Range Siltstone/Claystone 5-27
5.9.11 Troutdale Formation 5-28
5.9.12 Marine Basalts - Crescent Formation 5-28
5.9.13 Mélange Rocks on Olympic Peninsula 5-28

5.10 References 5-29

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-2 December 2006
Chapter 5 Engineering Properties of Soil and Rock
5.1 Overview
The purpose of this chapter is to identify, either by reference or explicitly herein, appropriate methods
of soil and rock property assessment, and how to use that soil and rock property data to establish the
final soil and rock parameters to be used for geotechnical design. The final properties to be used for
design should be based on the results from the field investigation, the field testing, and the laboratory
testing, used separately or in combination. Site performance data should also be used if available to help
determine the final geotechnical properties for design. The geotechnical designer’s responsibility is to
determine which parameters are critical to the design of the project and then determine those parameters
to an acceptable level of accuracy. See GDM Chapter 2, and the individual chapters that cover each
geotechnical design subject area, for further information on what information to obtain and how to plan
for obtaining that information.

The focus of geotechnical design property assessment and final selection shall be on the individual
geologic strata identified at the project site. A geologic stratum is characterized as having the same
geologic depositional history and stress history, and generally has similarities throughout the stratum
in terms of density, source material, stress history, and hydrogeology. It should be recognized that the
properties of a given geologic stratum at a project site are likely to vary significantly from point to point
within the stratum. In some cases, a measured property value may be closer in magnitude to the measured
property value in an adjacent geologic stratum than to the measured properties at another point within the
same stratum. However, soil and rock properties for design should not be averaged across multiple strata.
It should also be recognized that some properties (e.g., undrained shear strength in normally consolidated
clays) may vary as a predictable function of a stratum dimension (e.g., depth below the top of the
stratum). Where the property within the stratum varies in this manner, the design parameters should be
developed taking this variation into account, which may result in multiple values of the property within
the stratum as a function of a stratum dimension such as depth.

5.2 Influence of Existing and Future Conditions on Soil and

Rock Properties
Many soil properties used for design are not intrinsic to the soil type, but vary depending on conditions.
In-situ stresses, the presence of water, rate and direction of loading can all affect the behavior of soils.
Prior to evaluating the properties of a given soil, it is important to determine the existing conditions
as well as how conditions may change over the life of the project. Future construction such as new
embankments may place new surcharge loads on the soil profile or the groundwater table could be
raised or lowered. Often it is necessary to determine how subsurface conditions or even the materials
themselves will change over the design life of the project. Normally consolidated clays can gain strength
with increases in effective stress and overconsolidated clays may lose strength with time when exposed
in cuts. Some construction materials such as weak rock may loose strength due to weathering within the
design life of the embankment.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-3
Engineering Properties of Soil and Rock

5.3 Methods of Determining Soil and Rock Properties

Subsurface soil or rock properties are generally determined using one or more of the following methods:
• in-situ testing during the field exploration program,
• laboratory testing, and
• back analysis based on site performance data.

The two most common in-situ test methods for use in soil are the Standard Penetration Test, (SPT) and the
cone penetrometer test (CPT). The laboratory testing program generally consists of index tests to obtain
general information or to use with correlations to estimate design properties, and performance tests to
directly measure specific engineering properties. The observational method, or use of back analysis, to
determine engineering properties of soil or rock is often used with slope failures, embankment settlement
or excessive settlement of existing structures. With landslides or slope failures, the process generally
starts with determining the geometry of the failure and then determining the soil/rock parameters or
subsurface conditions that cause the safety factor to approach 1.0. Often the determination of the
properties is aided by correlations with index tests or experience on other projects. For embankment
settlement, a range of soil properties is generally determined based on laboratory performance testing
on undisturbed samples. Monitoring of fill settlement and pore pressure in the soil during construction
allows the soil properties and prediction of the rate of future settlement to be refined. For structures such
as bridges that experience unacceptable settlement or retaining walls that have excessive deflection, the
engineering properties of the soils can sometimes be determined if the magnitudes of the loads are known.
As with slope stability analysis, the geometry of the subsurface soil must be adequately known, including
the history of the groundwater level at the site.

The detailed measurement and interpretation of soil and rock properties shall be consistent with the
guidelines provided in FHWA-IF-02-034, Evaluation of Soil and Rock Properties, Geotechnical
Engineering Circular No. 5 (Sabatini, et al., 2002), except as specifically indicated herein.

5.4 In-Situ Field Testing

Standards and details regarding field tests such as the Standard Penetration Test (SPT), the Cone
Penetrometer Test (CPT), the vane shear test, and other tests and their use provided in Sabatini, et al.
(2002) should be followed, except as specifically noted herein. Regarding Standard Penetration Tests
(SPT), the N-values obtained are dependent on the equipment used and the skill of the operator, and
should be corrected for field procedures to standard N60 values (an efficiency of 60 percent is typical for
rope and cathead systems). This correction is necessary because many of the correlations developed to
determine soil properties are based on N60-values.

SPT N values should be corrected for hammer efficiency, if applicable to the design method or correlation
being used, using the following relationship.

N60 = (ER/60%) N (5-1)

Where:

N60 = SPT blow count corrected for hammer efficiency (blows/ft)

ER = Hammer efficiency expressed as percent of theoretical free fall energy delivered by the
hammer system actually used.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-4 December 2006
 Engineering Properties of Soil and Rock

The following values for ER may be assumed if hammer specific data are not available:

ER = 60% for conventional drop hammer using rope and cathead

ER = 80% for automatic trip hammer

Hammer efficiency (ER) for specific hammer systems used in local practice may be used in lieu of the
values provided. If used, specific hammer system efficiencies shall be developed in general accordance
with ASTM D-4945 for dynamic analysis of driven piles or other accepted procedure.

Corrections for rod length, hole size, and use of a liner may also be made if appropriate. In general, these
are only significant in unusual cases or where there is significant variation from standard procedures.
These corrections may be significant for evaluation of liquefaction. Information on these additional
corrections may be found in: “Proceedings of the NCEER Workshop on Evaluation of Liquefaction
Resistance of Soils”; Publication Number: MCEER-97-0022; T.L. Youd, I.M. Idriss (1997).

N-values are also affected by overburden pressure, and in general should be corrected for that effect,
if applicable to the design method or correlation being used������������������������������������������
. N values corrected for both overburden
and the efficiency of the field procedures used shall be designated as N160. The overburden correction
equation that should be used is:

N160 = CNN60 (5-2)

Where,

CN = [0.77 log10 (20/σ′v)], CN < 2.0 (5-3)

CN = correction factor for overburden
N60 = N-value corrected for energy efficiency
σ′v = vertical effective stress at the location of the SPT N-value (TSF)

In general, correlations between N-values and soil properties should only be used for cohesionless soils,
and sand in particular. Caution should be used when using N-values obtained in gravelly soil. Gravel
particles can plug the sampler, resulting in higher blow counts and estimates of friction angles than
actually exist. Caution should also be used when using N-values to determine silt or clay parameters due
to the dynamic nature of the test and resulting rapid changes in pore pressures and disturbance within
the deposit. Correlations of N-values with cohesive soil properties should generally be considered as
preliminary. N-values can also be used for liquefaction analysis. See WSDOT GDM Chapter 6 for
more information regarding the use of N-values for liquefaction analysis.

In general design practice, hydraulic conductivity is estimated based on grain size characteristics
of the soil strata (see Highway Runoff Manual M31-16, Section 4-5). In critical applications, the
hydraulic conductivity may be determined through in-situ testing. A discussion of field measurement of
permeability is presented in Sabatini, et al. (2002), and ASTM D 4043 presents a guide for the selection

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-5
Engineering Properties of Soil and Rock

of various field methods. If in-situ test methods are utilized to determine hydraulic conductivity, one or
more of the following methods should be used:
• Well pumping tests
• Packer permeability tests
• Seepage Tests
• Slug tests
• Piezocone tests
• Flood tests or Pit Infiltration Tests (PIT) – applies mainly to infiltration facility design – see
Section 4-5 of the WSDOT Highway Runoff Manual (2004)

5.4.1 Well Pumping Tests

Pump tests can be used to provide an estimate of the overall hydraulic conductivity of a geologic
formation, and since it is in essence a full scale test, directly accounts for the layering and directionality
of the hydraulic characteristics of the formation. The data provided can be used to determine the
requirements for construction dewatering systems for excavations. However, pump tests can be quite
expensive and can take a significant amount of time to complete. Furthermore, care must be exercised
when conducting this type of test, especially if potentially contaminated zones are present that could be
mobilized during pumping. This could also create problems with disposal of the pumped water. Impact
to adjacent facilities such as drinking wells and subsidence caused by dewatering should be evaluated
when planning this type of test. For this test, the method prescribed in ASTM D 4050 should be used.
Analysis of the results of pumping tests requires experience and a thorough knowledge of the actual
geologic conditions present at the test location. The time-drawdown response curves are unique to a
particular geologic condition. Therefore, knowledge of the actual geologic conditions present at the test
location is required in order to choose the correct analysis procedure, e.g., whether the aquifer is leaky,
unconfined, or bounded, etc.

5.4.2 Packer Permeability Tests

Packer permeability tests can be used to measure the hydraulic conductivity of a specific soil or rock
unit. The information obtained is used primarily in seepage studies. This test is conducted by inserting
the packer units to the desired test location after the boring has been properly cleaned out. The packers
are expanded to seal off the zone being tested, and water is injected into the borehole under constant
pressure. Measurements of the flow rate are taken at regular time intervals. Upon completion of testing
at a particular depth, the packers are lowered to a new test depth. Test depths should be determined from
cores and geophysical logs of the borehole, prior to hydraulic conductivity testing. Note that if the packer
test is performed in soil borings, casing must be installed.

5.4.3 Seepage Tests

Three types of seepage tests are commonly used: falling head, rising head and constant water level
methods. In general, either the rising or falling level methods should be used if the hydraulic conductivity
is low enough to permit accurate determination of the water level. In the falling head method, the
borehole or piezometer is filled with water that is allowed to seep into the soil. The rate of drop of the
water surface in the casing is monitored. The rising head method consists of bailing the water out of the
borehole and observing the rate of rise until the change becomes negligible. The constant water level
method is used if soil is too permeable to allow accurate measurement of the rising or falling water level.
General guidance on these types of tests are provided in Mayne, et al. (1997).

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-6 December 2006
 Engineering Properties of Soil and Rock

Boreholes (or in subsequently installed piezometers) in which seepage tests are to be performed should
be drilled using only clear water as the drilling fluid. This precludes the formation of a mud cake
on the walls of the hole or clogging of the soil pores with drilling mud. The tests can be performed
intermittently as the borehole is advanced. In general, the rising head test is preferred because there is
less chance of clogging soil pores with suspended sediment.

Data from seepage tests only reflect the hydraulic conditions near the borehole. In addition the actual
area of seepage at the base of the borehole may not be accurately known. During the rising head test,
there is the danger of the soil at the bottom of the borehole becoming loosened or “quick” if too great a
gradient is imposed. However, seepage tests can be used in soils with lower hydraulic conductivities than
is generally considered suitable for pumping tests and large volumes of water do not need to be disposed
of. Also note that if the test is conducted inside the piezometer, the hydraulic conductivity measured from
this could be influenced by the material placed inside the borehole around the screened pipe.

5.4.4 Slug Tests

These tests are easy to perform and can be performed in a borehole in which a screened pipe is installed.
Two types of slug tests are commonly used, falling head and rising head. Falling head slug tests are
conducted by lowering a solid object such as a weighted plastic cylinder into the borehole causing an
instantaneous water level rise. As the water level gradually returns to static, the rate is recorded. A rising
head slug test can then be performed by suddenly removing the slug, causing an instantaneous lowering of
the water level. By monitoring the rate of rise or fall of the water level in the borehole, an estimate of the
hydraulic conductivity can be determined. . For this test, the method prescribed in ASTM D 4044 should
be used. However slug tests are not very reliable and may underestimate hydraulic conductivity by one
or two orders of magnitude, particularly if the test well has been inadequately developed prior to testing.
The test data will not provide an indication of the accuracy of the computed value unless a pumping test
is done in conjunction with the slug test. Because the slug tests are short duration, they reflect hydraulic
properties of the soil immediately surrounding the well intake.

5.4.5 Piezocone Tests

Details of the equipment and methodology used to conduct the piezocone test are provided in Sabatini,
et al. (2002). Piezocone data can be useful to estimate the hydraulic conductivity of silts and clays from
interpretation of the coefficient of horizontal consolidation, ch, obtained from the piezocone measurments.
The procedure involves pushing the cone to the desired depth, followed by recording pore pressures
while the cone is held stationary. The test is usually run until 50 percent of the excess pore pressure
has dissipated (t50). This requires knowledge of the initial in situ pore pressure at the test location.
Dissipation tests are generally effective in silts and clays where large excess pore pressures are generated
during insertion of the cone. Hydraulic conductivity can be estimated using various correlations with t50
and coefficient of horizontal consolidation (ch), (see Lunne, et al. (1997), and Sabatini, et al. (2002)).
Estimation of hydraulic conductivity from CPT tests is subject to a large amount of uncertainty, and
should be used as a preliminary estimate of permeability only.

5.4.6 Flood Tests

Flood tests or pilot infiltration tests are not always feasible, and in general are only used where unusual
site conditions are encountered that are poorly modeled by correlation to soil gradation characteristics,
and there is plenty of water available to conduct the test. The key to the success of this type of test is the
estimate of the hydraulic gradient during the test, recognizing that the test hydraulic gradient could be
much higher than the hydraulic gradient that is likely in service for the facility being designed. For more
information, see the WSDOT Highway Runoff Manual (2004).

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-7
Engineering Properties of Soil and Rock

5.5 Laboratory Testing of Soil and Rock

Laboratory testing is a fundamental element of a geotechnical investigation. The ultimate purpose of
laboratory testing is to utilize repeatable procedures to refine the visual observations and field testing
conducted as part of the subsurface field exploration program, and to determine how the soil or rock
will behave under the imposed conditions. The ideal laboratory program will provide sufficient data to
complete an economical design without incurring excessive tests and costs. Depending on the project
issues, testing may range from simple soil classification testing to complex strength and deformation
testing.

5.5.1 Quality Control for Laboratory Testing

Improper storage, transportation and handling of samples can significantly alter the material properties
and result in misleading test results. The requirements provided in WSDOT GDM Chapter 3 regarding
these issues shall be followed.

Laboratories conducting geotechnical testing shall be either AASHTO accredited or fulfill the
requirements of AASHTO R18 for qualifying testers and calibrating/verifications of testing equipment for
those tests being performed. In addition, the following guidelines (Mayne, et al., 1997) for laboratory
testing of soils should be followed:
1. Protect samples to prevent moisture loss and structural disturbance.
2. Carefully handle samples during extrusion of samples; samples must be extruded properly and
supported upon their exit from the tube.
3. Avoid long term storage of soil samples in Shelby tubes.
4. Properly number and identify samples.
5. Store samples in properly controlled environments.
6. Visually examine and identify soil samples after removal of smear from the sample surface.
7. Use pocket penetrometer or miniature vane only for an indication of strength.
8. Carefully select “representative” specimens for testing.
9. Have a sufficient number of samples to select from.
10. Always consult the field logs for proper selection of specimens.
11. Recognize disturbances caused by sampling, the presence of cuttings, drilling mud or other foreign
matter and avoid during selection of specimens.
12. Do not depend solely on the visual identification of soils for classification.
13. Always perform organic content tests when classifying soils as peat or organic. Visual classifications
of organic soils may be very misleading.
14. Do not dry soils in overheated or underheated ovens.
15. Discard old worn-out equipment; old screens for example, particularly fine (< No. 40) mesh ones
need to be inspected and replaced often, worn compaction mold or compaction hammers (an error in
the volume of a compaction mold is amplified 30x when translated to unit volume) should be checked
and replaced if needed.
16. Performance of Atterberg Limits requires carefully adjusted drop height of the Liquid Limit machine
and proper rolling of Plastic Limit specimens.
17. Do not use tap water for tests where distilled water is specified.
18. Properly cure stabilization test specimens.
19. Never assume that all samples are saturated as received.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-8 December 2006
 Engineering Properties of Soil and Rock

20. Saturation must be performed using properly staged back pressures.

21. Use properly fitted o-rings, membranes, etc. in triaxial or permeability tests.
22. Evenly trim the ends and sides of undisturbed samples.
23. Be careful to identify slickensides and natural fissures. Report slickensides and natural fissures.
24. Also do not mistakenly identify failures due to slickensides as shear failures.
25. Do not use unconfined compression test results (stress-strain curves) to determine elastic modulus
values.
26. Incremental loading of consolidation tests should only be performed after the completion of each
primary stage.
27. Use proper loading rate for strength tests.
28. Do not guesstimate e-log p curves from accelerated, incomplete consolidation tests.
29. Avoid “Reconstructing” soil specimens, disturbed by sampling or handling, for undisturbed testing.
30. Correctly label laboratory test specimens.
31. Do not take shortcuts: using non-standard equipment or non-standard test procedures.
32. Periodically calibrate all testing equipment and maintain calibration records.
33. Always test a sufficient number of samples to obtain representative results in variable material.

5.5.2 Developing the Testing Plan

The amount of laboratory testing required for a project will vary depending on availability of preexisting
data, the character of the soils and the requirements of the project. Laboratory tests should be selected to
provide the desired and necessary data as economically as possible. Specific geotechnical information
requirements are provided in the WSDOT GDM chapters that address design of specific types of
geotechnical features. Laboratory testing should be performed on both representative and critical test
specimens obtained from geologic layers across the site. Critical areas correspond to locations where
the results of the laboratory tests could result in a significant change in the proposed design. In general,
a few carefully conducted tests on samples selected to cover the range of soil properties with the results
correlated by classification and index tests is the most efficient use of resources.

The following should be considered when developing a testing program:

• Project type (bridge, embankment, rehabilitation, buildings, etc.)
• Size of the project
• Loads to be imposed on the foundation soils
• Types of loads (i.e., static, dynamic, etc.)
• Whether long-term conditions or short-term conditions are in view
• Critical tolerances for the project (e.g., settlement limitations)
• Vertical and horizontal variations in the soil profile as determined from boring logs and visual
identification of soil types in the laboratory
• Known or suspected peculiarities of soils at the project location (i.e., swelling soils, collapsible soils,
organics, etc.)

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-9
Engineering Properties of Soil and Rock

• Presence of visually observed intrusions, slickensides, fissures, concretions, etc in sample –

how will it affect results
• Project schedules and budgets
• Input property data needed for specific design procedures

Details regarding specific types of laboratory tests and their use are provided in Sabatini, et al. (2002).

5.6 Engineering Properties of Soil

5.6.1 Laboratory Index Property Testing
Laboratory index property testing is mainly used to classify soils, though in some cases, they can also be
used with correlations to estimate specific soil design properties. Index tests include soil gradation and
plasticity indices. For soils with greater than 10 percent passing the No. 200 sieve, a decision will need to
be made regarding the full soil gradation curve is needed, requiring a hydrometer test in addition to sieve
testing of the coarser particles (AASHTO T88), or if a coarse gradation is enough (AASHTO T27). The
full gradation range (AASHTO T88) will be needed in the following situations:

• Lateral load analysis of deep foundations using strain wedge theory

• Liquefaction analysis
• Infiltration design, or other analyses that require the determination of hydraulic conductivities
• Other analyses that require a d10 size, coefficient of uniformity, etc.

Classification using the coarse sieving only (AASHTO T27) may be adequate for design of MSE walls,
general earthwork, footing foundations, gravity walls, and noise walls. These end use needs should be
considered when planning the laboratory investigation for a project.

5.6.2 Laboratory Performance Testing

Laboratory performance testing is mainly used to estimate strength, compressibility, and permeability
characteristics of soil and rock. For rock, the focus is typically on the shear strength of the intact rock,
or on the shear strength of discontinuities (i.e., joint/seam) within the rock mass. For soil, shear strength
may be determined on either undisturbed specimens of finer grained soil (undisturbed specimens of
granular soils are very difficult, if not impossible, to get), or disturbed or remolded specimens of fine or
coarse grained soil. There are a variety of shear strength tests that can be conducted, and the specific type
of test selected depends on the specific application. See Sabatini, et al. (2002) for specific guidance on
the types of shear strength tests needed for various applications, as well as the chapters in the WSDOT
GDM that cover specific geotechnical design topics.

Disturbed soil shear strength testing is less commonly performed, and is primarily used as supplementary
information when performing back-analysis of existing slopes, or for fill material and construction quality
assurance when a minimum shear strength is required. It is difficult to obtain very accurate shear strength
values through shear strength testing of disturbed (remolded) specimens since the in-situ density and
soil structure is quite difficult to accurately recreate, especially considering the specific in-situ density
may not be known. The accuracy of this technique in this case must be recognized when interpreting
the results. However, for estimating the shear strength of compacted backfill, more accurate results can
be obtained, since the soil placement method, as well as the in-situ density and moisture content, can be
recreated in the laboratory with some degree of confidence. The key in the latter case is the specimen size

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-10 December 2006
 Engineering Properties of Soil and Rock

allowed by the testing device, as in many cases, compacted fills have a significant percentage of gravel
sized particles, requiring fairly large test specimens (i.e., minimum 3 to 4 inch diameter, or narrowest
dimension specimens of 3 to 4 inches).

Typically, a disturbed sample of the granular backfill material (or native material in the case of obtaining
supplementary information for back-analysis of existing slopes) is sieved to remove particles that are too
large for the testing device and test standard, and is compacted into a mold to simulate the final density
and moisture condition of the material. The specimens may or may not be saturated after compacting
them and placing them in the shear testing device, depending on the condition that is to be simulated.
In general, a drained test is conducted, or if it is saturated, the pore pressure during shearing can be
measured (possible for triaxial testing; generally not possible for direct shear testing) to obtained drained
shear strength parameters. Otherwise, the test is run slow enough to be assured that the specimen is fully
drained during shearing (note that estimating the testing rate to assure drainage can be difficult). Multiple
specimens tested using at least three confining pressures should be tested to obtain a shear strength
envelope. See Sabatini, et al. (2002) for additional details.

Tests to evaluate compressibility or permeability of existing subsurface deposits must be conducted on

undisturbed specimens, and the less disturbance the better. See Sabatini, et al. (2002) for additional
requirements regarding these and other types of laboratory performance tests that should be followed.

The hydraulic conductivity of a soil is influenced by the particle size and gradation, the void ratio,
mineral composition, and soil fabric. In general the hydraulic conductivity, or permeability, increases
with increasing grain-size; however, the size and shape of the voids also has a significant influence. The
smaller the voids, the lower the permeability. Mineral composition and soil fabric have little effect on the
permeability of gravel, sand, and non-plastic silt, but are important for plastic silts and clays. Therefore,
relationships between particle size and permeability are available for coarse-grained materials, some
of which are presented in the Correlations subsection (WSDOT GDM Section 5.6.2). In general, for
clays, the lower the ion exchange capacity of the soil, the higher the permeability. Likewise, the more
flocculated (open) the structure, the higher the permeability.

The methods commonly used to determine the hydraulic conductivity in the laboratory include, the
constant head test, the falling head test, and direct or indirect methods during a consolidation test. The
laboratory tests for determining the hydraulic conductivity are generally considered quite unreliable.
Even with considerable attention to test procedures and equipment design, tests may only provide values
within an order of magnitude of actual conditions. Some of the factors for this are:
• The soil in-situ is generally stratified and this is difficult to duplicate in the laboratory.
• The horizontal value of k is usually needed, but testing is usually done on tube samples with vertical
values obtained.
• In sand, the horizontal and vertical values of k are significantly different, often on the order of
kh = 10 to 100kv.
• The small size of laboratory samples leads to boundary condition effects.
• Saturated steady state soil conditions are used for testing, but partially saturated soil water flow often
exists in the field.
• On low permeability soils, the time necessary to complete the tests causes evaporation and equipment
leaks to be significant factors.
• The hydraulic gradient in the laboratory is often 5 or more to reduce testing time, whereas in the field
it is more likely in the range of 0.1 to 2.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-11
Engineering Properties of Soil and Rock

The hydraulic conductivity is expected to vary across the site; however, it is important to differentiate
errors from actual field variations. When determining the hydraulic conductivity, the field and laboratory
values should be tabulated along with the other known data such as sample location, soil type, grain-size
distribution, Atterberg limits, water content, stress conditions, gradients, and test methods. Once this
table is constructed, it will be much easier to group like soil types and k values to delineate distinct areas
within the site, and eliminate potentially erroneous data.

5.6.3 Correlations to Estimate Engineering Properties of Soil

Correlations that relate in-situ index test results such as the SPT or CPT or laboratory soil index testing
may be used in lieu of or in conjunction with performance laboratory testing and back-analysis of site
performance data to estimate input parameters for the design of the geotechnical elements of a project.
Since properties estimated from correlations tend to have greater variability than measurement using
laboratory performance data (see Phoon, et al., 1995), properties estimated from correlation to in-situ
field index testing or laboratory index testing should be based on multiple measurements within each
significant geologic unit (if the geologic unit is large enough to obtain multiple measurements). A
minimum of 3 to 5 measurements should be obtained from each geologic unit as the basis for estimating
design properties.
The drained friction angle of granular deposits should be determined based on the correlation provided in
Table 5-1.

Table 5-1. Correlation of SPT N values to drained friction angle of granular soils (modified after
Bowles, 1977).

N160 from SPT φ

(blows/ft) (ο)
<4 25-30
4 27-32
10 30-35
30 35-40
50 38-43

The correlation used is modified after Bowles (1977). The correlation of Peck, Hanson and Thornburn
(1974) falls within the ranges specified. Experience should be used to select specific values within the
ranges. In general, finer materials or materials with significant silt-sized material will fall in the lower
portion of the range. Coarser materials with less then 5% fines will fall in the upper portion of the range.

Care should be exercised when using other correlations of SPT results to soil parameters. Some published
correlations are based on corrected values (N160) and some are based on uncorrected values (N). The
designer should ascertain the basis of the correlation and use either N160 or N as appropriate. Care
should also be exercised when using SPT blow counts to estimate soil shear strength if in soils with
coarse gravel, cobbles, or boulders. Large gravels, cobbles, or boulders could cause the SPT blow counts
to be unrealistically high.

Correlations for other soil properties (other than as specifically addressed above for the soil friction angle)
as provided in Sabatini, et al. (2002) may be used if the correlation is well established and if the accuracy

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-12 December 2006
 Engineering Properties of Soil and Rock

of the correlation is considered regarding its influence if the estimate obtained from the correlation in the
selection of the property value used for design. Local geologic formation-specific correlations may also
be used if well established by data comparing the prediction from the correlation to measured high quality
laboratory performance data, or back-analysis from full scale performance of geotechnical elements
affected by the geologic formation in question.

Regarding soil hydraulic conductivity, the correlation provided in the WSDOT Highway Runoff Manual,
Section 4-5 should be used.

5.7 Engineering Properties of Rock

Engineering properties of rock are generally controlled by the discontinuities within the rock mass and
not the properties of the intact material. Therefore, engineering properties for rock must account for the
properties of the intact pieces and for the properties of the rock mass as a whole, specifically considering
the discontinuities within the rock mass. A combination of laboratory testing of small samples, empirical
analysis, and field observations should be employed to determine the engineering properties of rock
masses, with greater emphasis placed on visual observations and quantitative descriptions of the rock
mass.

Rock properties can be divided into two categories: intact rock properties and rock mass properties. Intact
rock properties are determined from laboratory tests on small samples typically obtained from coring,
outcrops or exposures along existing cuts. Engineering properties typically obtained from laboratory
tests include specific gravity, unit weight, ultrasonic velocity, compressive strength, tensile strength, and
shear strength. Rock mass properties are determined by visual examination of discontinuities within the
rock mass, and how these discontinuities will affect the behavior of the rock mass when subjected to the
proposed construction.

The methodology and related considerations provided by Sabatini, et al. (2002) should be used to assess
the design properties for the intact rock and the rock mass as a whole. However, the portion of Sabatini,
et al. (2002) that addresses the determination of fractured rock mass shear strength parameters
(Hoek and Brown, 1988) is outdated. The original work by Hoek and Brown has been updated and
is described in Hoek, et al. (2002). The updated method uses a Geological Strength Index (GSI) to
characterize the rock mass for the purpose of estimating strength parameters, and has been developed
based on re-examination of hundreds of tunnel and slope stability analyses in which both the 1988 and
2002 criteria were used and compared to field results. While the 1988 method has been more widely
published in national (e.g., FHWA) design manuals than has the updated approach provided in
Hoek, et al. (2002), considering that the original developers of the method have recognized the
short-comings of the 1988 method and have reassessed it through comparison to actual rock slope
stability data, WSDOT considers the Hoek, et al. (2002) to be the most accurate methodology. Therefore
the Hoek, et al. (2002) method should be used for fractured rock mass shear strength determination. Note
that this method is only to be used for highly fractured rock masses in which the stability of the rock slope
is not structurally controlled. See WSDOT GDM Chapter 12 for additional requirements regarding the
assessment of rock mass properties.

5.8 Final Selection of Design Values

5.8.1 Overview
After the field and laboratory testing is completed, the geotechnical designer should review the quality
and consistency of the data, and should determine if the results are consistent with expectations. Once
the lab and field data have been collected, the process of final material property selection begins. At this

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-13
Engineering Properties of Soil and Rock

stage, the geotechnical designer generally has several sources of data consisting of that obtained in the
field, laboratory test results and correlations from index testing. In addition, the geotechnical designer
may have experience based on other projects in the area or in similar soil/rock conditions. Therefore, if
the results are not consistent with each other or previous experience, the reasons for the differences should
be evaluated, poor data eliminated and trends in data identified. At this stage it may be necessary to
conduct additional performance tests to try to resolve discrepancies.

As stated in WSDOT GDM Section 5.1, the focus of geotechnical design property assessment and
final selection is on the individual geologic strata identified at the project site. A geologic stratum is
characterized as having the same geologic depositional history and stress history, and generally has
similarities throughout the stratum in its density, source material, stress history, and hydrogeology.
All of the information that has been obtained up to this point including preliminary office and field
reconnaissance, boring logs, CPT soundings etc., and laboratory data are used to determine soil and
rock engineering properties of interest and develop a subsurface model of the site to be used for
design. Data from different sources of field and lab tests, from site geological characterization of the
site subsurface conditions, from visual observations obtained from the site reconnaissance, and from
historical experience with the subsurface conditions at or near the site must be combined to determine the
engineering properties for the various geologic units encountered throughout the site. However, soil and
rock properties for design should not be averaged across multiple strata, since the focus of this property
characterization is on the individual geologic stratum. Often, results from a single test
(e.g. SPT N-values) may show significant scatter across a site for a given soil/rock unit. Perhaps data
obtained from a particular soil unit for a specific property from two different tests (e.g. field vane
shear tests and lab UU tests) do not agree. Techniques should be employed to determine the validity
and reliability of the data and its usefulness in selecting final design parameters. After a review of
data reliability, a review of the variability of the selected parameters should be carried out. Variability
can manifest itself in two ways: 1) the inherent in-situ variability of a particular parameter due to the
variability of the soil unit itself, and 2) the variability associated with estimating the parameter from the
various testing methods. From this step final selection of design parameters can commence, and from
there completion of the subsurface profile.

5.8.2 Data Reliability and Variability

Inconsistencies in data should be examined to determine possible causes and assess any mitigation
procedures that may be warranted to correct, exclude, or downplay the significance of any suspect data.
The following procedures provide a step-by-step method for analyzing data and resolving inconsistencies
as outlined by Sabatini, et al. (2002):
1) Data Validation: Assess the field and the laboratory test results to determine whether the reported test
results are accurate and are recorded correctly for the appropriate material. For lab tests on
undisturbed samples consider the effects of sample disturbance on the quality of the data. For index
tests (e.g. grain size, compaction) make sure that the sample accurately represents the in situ
condition. Disregard or downplay potentially questionable results.
2) Historical Comparison: Assess results with respect to anticipated results based on site and/or
regional testing and geologic history. If the new results are inconsistent with other site or regional
data, it will be necessary to assess whether the new data is anomalous or whether the new site
conditions differ from those from which previous data was collected. For example, an alluvial
deposit might be expected to consist of medium dense silty sand with SPT blow counts less than
30. If much higher blow counts are recorded, the reason could be the deposit is actually dense (and
therefore higher friction angles can be assumed), or gravel may be present and is influencing the SPT
data. Most likely it is the second case, and the engineering properties should probably be adjusted to

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-14 December 2006
 Engineering Properties of Soil and Rock

account for this. But if consideration had not been given as to what to expect, values for properties
might be used that could result in an unconservative design.
3) Performance Comparison: Assess results with respect to historic performance of structures at the
site or within similar soils. Back analysis of previous landslides and retaining wall movement in the
same geologic units under consideration, if available, should be performed to estimate shear strength
parameters. Settlement data from existing embankments, if available, should be used to estimate
compressibility and settlement rates. Results can be compared to the properties determined from field
and lab testing for the project site. The newly collected data can be correlated with the parameters
determined from observation of performance and the field and lab tests performed for the previous
project.
4) Correlation Calibration: If feasible, develop site-specific correlations using the new field and lab
data. Assess whether this correlation is within the range of variability typically associated with the
correlation based on previous historic data used to develop the generic correlation.
5) Assess Influence of Test Complexity: Assess results from the perspective of the tests themselves.
Some tests may be easy to run and calibrate, but provide data of a “general” nature while other tests
are complex and subject to operator influence, yet provide “specific” test results. When comparing
results from different tests consider which tests have proven to give more accurate or reliable results
in the past, or more accurately approximate anticipated actual field conditions. For example, results
of field vane shear tests may be used to determine undrained shear strength for deep clays instead of
laboratory UU tests because of the differences in stress states between the field and lab samples. It
may be found that certain tests consistently provide high or low values compared to anticipated
results.

The result of these steps is to determine whether or not the data obtained for the particular tests in
question is valid. Where it is indicated that test results are invalid or questionable, the results should be
downplayed or thrown out. If the test results are proven to be valid, the conclusion can be drawn that the
soil unit itself and its corresponding engineering properties are variable (vertically, aerially, or both).

The next step is to determine the amount of variability that can be expected for a given engineering
property in a particular geologic unit, and how that variability should influence the selection of the final
design value. Sabatini, et al. (2002) list several techniques that can be used:
1) Experience: In some cases the geotechnical designer may have accumulated extensive experience
in the region such that it is possible to accurately select an average, typical or design value for the
selected property, as well as the appropriate variability for the property.
2) Statistics: If a geotechnical designer has extensive experience in a region, or there has been extensive
testing by others with published or available results, there may be sufficient data to formally establish
the average value and the variability (mean and standard deviation) for the specific property. See
Sabatini, et al. (2002) and Phoon, et al. (1995) for information on the variability associated with
various engineering properties.
3) Establish Best-Case and Worst-Case Scenarios: Based on the experience of the geotechnical
designer, it may be possible to establish upper and lower bounds along with the average for a given
property.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-15
Engineering Properties of Soil and Rock

5.8.3 Final Property Selection

The final step is to incorporate the results of the previous section into the selection of design values for
required design properties.

Recognizing the variability discussed in the previous section, depending on the amount of variability
estimated or measured, the potential impact of that variability (or uncertainty) on the level of safety in the
design should be assessed. If the impact of this uncertainty is likely to be significant, parametric analyses
should be conducted, or more data could be obtained to help reduce the uncertainty. Since the sources
of data that could be considered may include both measured laboratory data, field test data, performance
data, and other previous experience with the geologic unit(s) in question, it will not be possible to
statistically combine all this data together to determine the most likely property value. Engineering
judgment, combined with parametric analyses as needed, will be needed to make this final assessment
and design property determination. At that point, a decision must be made as to whether the final design
value selected should reflect the interpreted average value for the property, or a value that is somewhere
between the most likely average value and the most conservative estimate of the property. However, the
desire for design safety must be balanced with the cost effectiveness and constructability of the design. In
some cases, being too conservative with the design could result in an un-constructible design (e.g., the use
of very conservative design parameters could result in a pile foundation that must be driven deep into a
very dense soil unit that in reality is too dense to penetrate with available equipment).

Note that in WSDOT GDM Chapter 8, where reliability theory was used to establish load and
resistance factors, the factors were developed assuming that mean values for the design properties
are used. However, even in those cases, design values that are more conservative than the mean may
still be appropriate, especially if there is an unusual amount of uncertainty in the assessment of the
design properties due, for example to highly variable site conditions, lack of high quality data to assess
property values, or due to widely divergent property values from the different methods used to assess
properties within a given geologic unit. Depending on the availability of soil or rock property data and
the variability of the geologic strata under consideration, it may not be possible to reliably estimate the
average value of the properties needed for design. In such cases, the geotechnical designer may have
no choice but to use a more conservative selection of design parameters to mitigate the additional risks
created by potential variability or the paucity of relevant data. Note that for those resistance factors that
were determined based on calibration by fitting to allowable stress design, this property selection issue is
not relevant, and property selection should be based on the considerations discussed previously.

The process and examples to make the final determination of properties to be used for design provided by
Sabatini, et al. (2002) should be followed.

5.8.4 Development of the Subsurface Profile

While WSDOT GDM Section 5.8 generally follows a sequential order, it is important to understand that
the selection of design values and production of a subsurface profile is more of an iterative process. The
development of design property values should begin and end with the development of the subsurface
profile. Test results and boring logs will likely be revisited several times as the data is developed and
analyzed before the relation of the subsurface units to each other and their engineering properties are
finalized.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-16 December 2006
 Engineering Properties of Soil and Rock

The ultimate goal of a subsurface investigation is to develop a working model that depicts major
subsurface layers exhibiting distinct engineering characteristics. The end product is the subsurface
profile, a two dimensional depiction of the site stratigraphy. The following steps outline the creation of
the subsurface profile:
1) Complete the field and lab work and incorporate the data into the preliminary logs.
2) Lay out the logs relative to their respective field locations and compare and match up the different soil
and rock units at adjacent boring locations, if possible. However, caution should be exercised when
attempting to connect units in adjacent borings, as the geologic stratigraphy does not always fit into
nice neat layers. Field descriptions and engineering properties will aid in the comparisons.
3) Group the subsurface units based on engineering properties.
4) Create cross sections by plotting borings at their respective elevations and positions horizontal to one
another with appropriate scales. If appropriate, two cross sections should be developed that are at
right angles to each other so that lateral trends in stratigraphy can be evaluated when a site contains
both lateral and transverse extents (i.e. a building or large embankment).
5) Analyze the profile to see how it compares with expected results and knowledge of geologic
(depositional) history. Have anomalies and unexpected results encountered during exploration and
testing been adequately addressed during the process? Make sure that all of the subsurface features
and properties pertinent to design have been addressed.

5.8.5 Selection of Design Properties for Engineered Materials

This section provides guidelines for the selection of properties that are commonly used on WSDOT
projects such as engineered fills. The engineering properties are based primarily on gradation and
compaction requirements, with consideration of the geologic source of the fill material typical for the
specific project location. For materials such as common borrow where the gradation specification is fairly
broad, a wider range of properties will need to be considered.

Common Borrow. Per the WSDOT Standard Specifications, common borrow may be virtually any
soil or aggregate either naturally occurring or processed which is substantially free of organics or other
deleterious material, and is non-plastic. The specification allows for the use of more plastic common
borrow when approved by the engineer. On WSDOT projects this material will generally be placed
at 90 percent (Method B) or 95 percent (Method C) of Standard Proctor compaction. Because of the
variability of the materials that may be used as common borrow, the estimation of an internal friction
angle and unit weight should be based on the actual material used. A range of values for the different
material properties is given in Table 5-2. Lower range values should be used for finer grained materials
compacted to Method B specifications. In general during design, the specific source of borrow is not
known. Therefore, it is not prudent to select a design friction angle that is near or above the upper end
of the range unless the geotechnical designer has specific knowledge of the source(s) likely to be used,
or unless quality assurance shear strength testing is conducted during construction. Common borrow
will likely have a high enough fines content to be moderately to highly moisture sensitive. This moisture
sensitivity may affect the design property selection if it is likely that placement conditions are likely to be
marginal due to the timing of construction.

Select Borrow. The requirements for select borrow ensure that the mixture will be granular and contain
at least a minimal amount of gravel size material. The materials are likely to be poorly graded sand and
contain enough fines to be moderately moisture sensitive (the specification allows up to 10 percent fines).
Select Borrow is not an all weather material. Triaxial or direct shear strength testing on material that
meets Select Borrow gradation requirements indicates that drained friction angles of 38 to 45 degrees are

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-17
Engineering Properties of Soil and Rock

likely when the soil is well compacted. Even in it loosest state, shear strength testing of relatively clean
sands meeting Select Borrow requirements has indicated values of 30 to 35 degrees are likely. However,
these values are highly dependent on the geologic source of the material. Much of the granular soil in
Washington has been glacially derived, resulting in subangular to angular soil particles and hence, high
shear strength values. Windblown, beach, or alluvial sands that have been rounded through significant
transport could have significantly lower shear strength values. Left-overs from processed materials
(e.g., scalpings) could also have relative low friction angles depending on the uniformity of the material
and the degree of rounding in the soil particles. A range of values for shear strength and unit weight
based on previous experience for well compacted Select Borrow is provided in Table 5-2. In general
during design the specific source of borrow is not known. Therefore, it is not prudent to select a design
friction angle that is near or above the upper end of the range unless the geotechnical designer has
specific knowledge of the source(s) likely to be used or unless quality assurance shear strength testing is
conducted during construction. Select Borrow with significant fines content may sometimes be modeled
as having a temporary or apparent cohesion value from 50 to 200 psf. If a cohesion value is used, the
friction angle should be reduced so as not to increase the overall strength of the material. For long term
analysis, all the borrow material should be modeled with no cohesive strength.

Gravel Borrow – The gravel borrow specification should ensure a reasonably well graded sand and
gravel mix. Because the fines content is under 7 percent, the material is only slightly moisture sensitive.
However, in very wet conditions, material with lower fines content should be used. Larger diameter
triaxial shear strength testing performed on well graded mixtures of gravel with sand that meet the Gravel
Borrow specification indicate that very high internal angles of friction are possible, approaching 50
degrees, and that shear strength values less than 40 degrees are not likely. However, lower shear strength
values are possible for Gravel Borrow from naturally occurring materials obtained from non-glacially
derived sources such as wind blown or alluvial deposits. In many cases, processed materials are used
for Gravel Borrow, and in general, this processed material has been crushed, resulting in rather angular
particles and very high soil friction angles. Its unit weight can approach that of concrete if very well
graded. A range of values for shear strength and unit weight based on previous experience is provided
in Table 5-2. In general during design the specific source of borrow is not known. Therefore, it is
not prudent to select a design friction angle that is near or above the upper end of the range unless the
geotechnical designer has specific knowledge of the source(s) likely to be used or unless quality assurance
shear strength testing is conducted during construction.

Backfill for Walls. Gravel backfill for walls is a free draining material that is generally used to facilitate
drainage behind retaining walls. This material has similarities to Gravel Borrow, but generally contains
fewer fines and is freer draining. Gravel backfill for Walls is likely to be a processed material and if
crushed is likely to have a very high soil friction angle. A likely range of material properties is provided
in Table 5-2.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-18 December 2006
 Engineering Properties of Soil and Rock

Table 5-2. Presumptive design property ranges for borrow and other WSDOT Standard Specification
materials.

WSDOT
Soil Type (USCS Cohesion Total Unit
Material Standard
classification)
φ (degrees) (psf) Weight (pcf)
Specification
Common 9-03.14(3) ML, SM, GM 30 to 34 0 115 to 130
Borrow
Select Borrow 9-03.14(2) GP, GP-GM, SP, 34 to 38 0 120 to 135
SP-SM
Gravel Borrow 9-03.14(1) GW, GW-GM, SW, 36 to 40 0 130 to 145
SW-SM
Gravel Backfill 9-03.12(2) GW, GP, SW, SP 36 to 40 0 125 to 135
for Walls

Rock embankment. Embankment material is considered rock embankment if 25 percent of the material
is over 4 inches in diameter. Compactive effort is based on a method specification. Because of the nature
of the material, compaction testing is generally not feasible. The specification allows for a broad range of
material and properties such that the internal friction angle and unit weight can vary considerably based
on the amount and type of rock in the fill. Rock excavated from cuts consisting of siltstone, sandstone
and claystone may break down during the compaction process, resulting in less coarse material. Also,
if the rock is weak, failure may occur through the rock fragments rather than around them. In these
types of materials, the strength parameters may resemble those of earth embankments. For existing
embankments, the soft rock may continue to weather with time, if the embankment materials continue to
become wet. For sound rock embankments, the strength parameters may be much higher. For compacted
earth embankments with sound rock, internal friction angles of up to 45 degrees may be reasonable. Unit
weights for rock embankments generally range from 130 to 140 pcf.

Quarry Spalls and Rip Rap. Quarry spalls, light loose rip rap and heavy loose rip rap created from shot
rock are often used as fill material below the water table or in shear keys in slope stability and landslide
mitigation applications. WSDOT Standard Specification Section 9-13 provides minimum requirements
for degradation and specific gravity for these materials. Therefore sound rock must be used for these
applications. For design purposes, typical values of 120 to 130 pcf for the unit weight (this considers the
large amount of void space due to the coarse open gradation of this type of material) and internal angles
of friction of about 40 to 45 degrees should be used.

Wood Fiber. Wood fiber fills have been used by WSDOT for over 30 years in fill heights up to about
40 feet. The wood fiber has generally been used as light weight fill material over soft soil to improve
embankment stability. Wood fiber has also been used in emergency repair because rain and wet weather
does not affect the placement and compaction of the embankment. Only fresh wood fiber should be used
to prolong the life of the fill, and the maximum particle size should be 6 inches or less. The wood fiber
is generally compacted in lifts of about 12 inches with two passes of a track dozer. Presumptive design
values of 50 pcf for unit weight and an internal angle of friction of about 40 degrees may be used for the
design of the wood fiber fills (Allen et al., 1993).

To mitigate the effects of leachate, the amount of water entering the wood should be minimized.
Generally topsoil caps of about 2 feet in thickness are used. The pavement section should be a minimum

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-19
Engineering Properties of Soil and Rock

of 2 feet (a thicker section may be needed depending on the depth of wood fiber fill). Wood fiber fill will
experience creep settlement for several years and some pavement distress should be expected during that
period. Additional information on the properties and durability of wood fiber fill is provided in
Kilian and Ferry (1993).

Geofoam. Geofoam has been used as lightweight fill on WSDOT projects for about 10 years. Geofoam
ranges in unit weight from about 1 to 2 pcf. Geofoam constructed from expanded polystyrene (EPS) is
manufactured according to ASTM standards for minimum density (ASTM C 303), compressive strength
(ASTM D 1621) and water absorption (ASTM C 272). Type I and II are generally used in highway
applications. Bales of recycled industrial polystyrene waste are also available. These bales have been
used to construct temporary haul roads over soft soil. However, these bales should not be used in
permanent applications.

5.9 Properties of Predominant Geologic Units in Washington

This section contains a brief discussion of soil and rock types common to Washington state that have
specific engineering properties that need consideration.

5.9.1 Loess
Loess is a windblown (eolian) soil consisting mostly of silt with minor amounts of sand and clay
(Higgens et al., 1987). Due to its method of deposition, loess has an open (honeycomb) structure with
very high void ratios. The clay component of loess plays a pivotal role because it acts as a binder (along
with calcium carbonate in certain deposits) holding the structure together. However, upon wetting, either
the water soluble calcium carbonate bonds dissolve or the large negative pore pressures within the clay
that are holding the soil together are reduced and the soil can undergo shear failures and/or settlements.

Loess deposits encompass a large portion of southeastern Washington. Loess typically overlies portions
of the Columbia River Basalt Group and is usually most pronounced at the tops of low hills and plateaus
where erosion has been minimal (Joseph, 1990). Washington loess has been classified into four geologic
units: Palouse Loess, Walla Walla Loess, Ritzville Loess, and Nez Perce Loess. However, these
classifications hold little relevance to engineering behavior. For engineering purposes loess can generally
be classified into three categories based on grain size: clayey loess, silty loess, and sandy loess (see
WSDOT GDM Chapter 10).

Typical index and performance properties measured in loess are provided in Table 5-3, based on the
research results provided in Report WA-RD 145.2 (Higgins and Fragaszy, 1988). Density values
typically increase from west to east across the state with corresponding increase in clay content. Higgins
and Fragaszy observed that densities determined from Shelby tube samples in loess generally result in
artificially high values due to disturbance of the open soil structure and subsequent densification. Studies
of shear strength on loess have indicated that friction angles are usually fairly constant for a given deposit
and are typically within the range of 27 to 29 degrees using CU tests. These studies have also indicated
that cohesion values can be quite variable and depend on the degree of consolidation, moisture content
and amount of clay binder. Research has shown that at low confining pressures, loess can lose all shear
strength upon wetting.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-20 December 2006
 Engineering Properties of Soil and Rock

Table 5-3. Typical measured properties for loess deposits in Washington state.

Type of Liquid Plasticity Angle of Internal

Dry Density (pcf)
Loess Limit Index Friction (o)
Clayey 33 to 49 11 to 27
70 to 90, with maximum of up to 95 to 98 27 to 29 from CU
Silty 14 to 32 0 to 11
(generally increases with clay content) tests
Sandy Nonplastic Nonplastic

The possibility of wetting induced settlements should be considered for any structure supported on loess
by performing collapse tests. Collapse tests are usually performed as either single ring (ASTM D 5333)
or double ring tests. Double ring tests have the advantage in that potential collapse can be estimated for
any stress level. However, two identical samples must be obtained for testing. Single ring tests have the
advantage in that they more closely simulate actual collapse conditions and thus give a more accurate
estimate of collapse potential. However, collapse potential can only be estimated for a particular stress
level, so care must be taken to choose an appropriate stress level for sample inundation during a test.
When designing foundations in loess, it is important to consider long term conditions regarding possible
changes in moisture content throughout the design life of the project. Proper drainage design is crucial
to keeping as much water as possible from infiltrating into the soil around the structure. A possible
mitigation technique could include overexcavation and recompaction to reduce or eliminate the potential
for collapse settlement.

Loess typically has low values of permeability and infiltration rates. When designing stormwater
management facilities in loess, detention ponds should generally be designed for very low infiltration
rates.

Application of the properties of loess to cut slope stability is discussed in WSDOT GDM Chapter 10.

5.9.2 Peat/Organic Soils

Peats and organic soils are characterized by very low strength, very high compressibility (normally or
slightly under-consolidated) and having very important time-consolidation effects. Often associated
with wetlands, ponds and near the margins of shallow lakes, these soils pose special challenges for the
design of engineering transportation projects. Deep deposits (+100 FT in some cases) with very high
water content, highly compressibility, low strength and local high groundwater conditions require careful
consideration regarding settlement and stability of earth fill embankments, support for bridge foundations,
and locating culverts.

The internal structure of peat, either fibrous or granular, affects its capacity for retaining and releasing
water and influences its strength and performance. With natural water content often ranging from
200-600 percent (over 100 for organic silts and sands) and wet unit weight ranging from 70 to 90 pcf,
it can experience considerable shrinkage (>50%) when its dried out. Rewetting usually cannot restore
its original volume or moisture content. Under certain conditions, dried peat will oxidize and virtually
disappear. Undisturbed sampling for laboratory testing is difficult. Field vane testing is frequently used
to evaluate in place shear strength, though in very fibrous peats, reliable shear strength data is difficult to
obtain even with the field vane shear test. Initial undisturbed values of 100-400 PSF are not uncommon

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-21
Engineering Properties of Soil and Rock

but remolded (residual) strengths can be 30 to 50 % less (Schmertmann, 1967). Vane strength however,
is a function of both vane size and peat moisture content. Usually, the lower the moisture of the peat and
the greater its depth, the higher is its strength. Strength increases significantly when peat is consolidated,
and peak strength only develops after large deformation has taken place. Due to the large amount of
strain that can occur when embankment loads are placed on peats and organic soils, residual strengths
may control the design.

Vertical settlement is also a major concern for constructing on organic soils. The amount of foundation
settlement and the length of time for it to occur are usually estimated from conventional laboratory
consolidation tests. Secondary compression can be quite large for peats and must always be evaluated
when estimating long-term settlement. Based on experience in Washington state, compression index
values based on vertical strain (Ccε) typically range from 0.1 to 0.3 for organic silts and clays, and are
generally above 0.3 to 0.4 for peats. The coefficient of secondary compression (Cαε) is typically equal to
0.05Ccε to 0.06Ccε for organic silts and peats, respectively.

5.9.3 Glacial Till and Glacial Advance Outwash

Glacial till typically consists of non-stratified deposits of clay, silt, sand and gravel with cobbles and
occasional boulders. Although the matrix proportions of silt and clay vary from place to place, the matrix
generally consists of silty sand or sandy silt (Troost and Booth, 2003). The glacial till has been glacially
overridden, but the upper 2 to 5 feet is often weathered and is typically medium dense to dense. The
glacial till generally grades to dense to very dense below the weathered zone.

Glacial till is often found near the surface in the Puget Sound Lowland area. The Puget Sound Lowland is
a north-south trending trough bordered by the Cascade Mountains to the east and the Olympic Mountains
to the west. The most recent glaciation, the Vashon Stade of the Fraser Glaciation occurred between
roughly 18,000 to 13,000 years ago. Glacial till deposited by this glaciation extends as far south as the
Olympia area.

Glacial till generally provides good bearing resistance because of its dense nature. Typically values used
in slope stability evaluations range from 40 to 45 degrees for internal friction angle with cohesion values
of 100 to 1,000 psf. Unit weights used for design are typically in the range of 130 to 140 pcf. The dense
nature of glacial till tends to make excavation and sheet pile installation difficult. It is not uncommon to
have to rip glacial till with a dozer or utilize large excavation equipment. Permeability in glacial till is
relatively low because of the fines content and the density. However, localized pockets and seams of sand
with higher permeability are occasional encountered in glacial till units.

Wet weather construction in glacial till is often difficult because of the relatively high fines content of
glacial till soils. When the moisture content of these soils is more than a few percent above the optimum
moisture content, glacial till soils become muddy and unstable, and operation of equipment on these soils
can become difficult.

Glacial advance outwash is similar in nature to glacial till, but tends to be more coarse grained and cleaner
(fewer fines). Properties are similar, but cohesion is lower, causing this material to have greater difficulty
standing without raveling in a vertical cut, and in general can more easily cave in open excavations or
drilled holes. Since it contains less fines, it is more likely to have relatively high permeability and be
water bearing. In very clean deposits, non-displacement type piles (e.g., H-piles) can “run”.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-22 December 2006
 Engineering Properties of Soil and Rock

For both glacial till and glacial advance outwash, cobble and boulder sized material can be encountered
any time. Boulders in these deposits can range from a foot or two in diameter to the size of a bus. In
some areas they can also be nested together, making excavation very difficult.

5.9.4 Colluvium/Talus
Colluvium is a general term used to describe soil and rock material that has been transported and
deposited by gravitational forces. Colluvium is typified by poorly sorted mixtures of soil and rock
particles ranging in size from clay to large boulders. Talus is a special type of colluvium and refers to
a particular landform and the material that comprises it. Talus deposits are generally made up of rock
fragments of any size and shape (but usually coarse and angular) located at the base of cliffs and steep
slopes.

Colluvium is a very common deposit, encompassing upwards of 90 percent of the ground surface in
mountainous areas. Colluvial deposits are typically shallow (less than about 25 to 30 feet thick), with
thickness increasing towards the base of slopes. Colluvium generally overlies bedrock along the slopes of
hills and mountains and intermixes with alluvial material in stream bottoms.

Subsurface investigations in colluvium using drilling equipment are usually difficult because of the
heterogeneity of the deposit and possible presence of large boulders. In addition, site access and safety
issues also can pose problems. Test pits and trenches offer alternatives to conventional drilling that may
provide better results. Subsurface investigations in talus are usually even more difficult, if not impossible.
Engineering properties of talus are extremely difficult to determine in the lab or in situ. A useful
method for determining shear strength properties in both colluvium and talus is to analyze an existing
slope failure. For talus, this may be the only way to estimate shear strength parameters. Talus deposits
can be highly compressible because of the presence of large void spaces. Colluvium and talus slopes
are generally marginally stable. In fact, talus slopes are usually inclined at the angle of repose of the
constituent material. Cut slopes in colluvium often result in steepened slopes beyond the angle of repose,
resulting in instability. Slope instability is often manifested by individual rocks dislodging from the slope
face and rolling downslope. While the slope remains steeper than the angle of repose a continuous and
progressive failure will occur.

Construction in colluvium is usually difficult because of the typical heterogeneity of deposits and
corresponding unfavorable characteristics such as particle size and strength variations and large void
spaces. In addition, there is the possibility of long term creep movement. Large settlements are also
possible in talus. Foundations for structures in talus should extend through the deposit and bear on more
competent material. Slope failures in colluvium are most often caused by infiltration of water from
intense rainfall. Modifications to natural slopes in the form of cut slopes and construction of drainage
ditches are one way that water can infiltrate into a colluvial soil and initiate a slope failure. Careful
consideration must be given to the design of drainage facilities to prevent the increase of water in
colluvial soils.

5.9.5 Columbia River Sand

These sands are located in the Vancouver area and may have been deposited by backwaters from the
glacial Lake Missoula catastrophic floods. The sands are poorly graded and range from loose to medium
dense. The sand is susceptible to liquefaction if located below the water table. The sands do not provide
a significant amount of frictional resistance for piles, and non-displacement piles may tend to run. Based
on the observed stability of slopes in this formation, soil friction angles of 28o to 32o should be expected.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-23
Engineering Properties of Soil and Rock

5.9.6 Columbia Basin Basalts

The basalt flows that dominate the Columbia Basin were erupted into a structural and topographic low
between the northern Rocky Mountains and the rising Cascade Range. During periods between the
flows, erosion would take place and tuffs, sandstones, and conglomerates would be deposited on top
of basalt flows (Thorsen, 1989). In some areas lake beds formed. The resulting drainage systems and
lakes were responsible for the extensive layer of sediments between, interfingering with, and overlying
the basalt flows. These sediments are generally thicker in areas peripheral to the flows, especially in and
along the western part of the basin. During the interludes between flows, deep saprolites formed on some
flow surfaces. Present topographic relief on the basin has been provided largely by a series of east-west
trending anticlinal folds, by the cutting of catastrophic glacial meltwater floods and by the Columbia
River system.

The most obvious evidence of bedrock slope failures in the basin is the presence of basalt talus slopes
fringing the river canyons and abandoned channels. Such slopes made up of clasts of nearly the same size
and are standing at near the angle of repose.

Bedrock failures are most commonly in the form of very large ancient slumps or slump flows. Block
glides may be locally important and probably result from failures along interbeds or palagonite zones at
flow contacts. Most of these ancient failures occur in areas of regional tilting or are associated with folds.
The final triggering, in many cases, appears to have been oversteepening of slopes or removal of toe
support.

Along I-82 on the west edge of the province and in a structural basin near Pasco, layers of sediments
interfinger with basalt flows. Some of these sediments are compact enough to be considered siltstone
or sandstone and are rich in montmorillionite. Slumps and translation failures are common in some
places along planes sloping as little as 8 degrees. Most landslides are associated with pre-existing failure
surfaces developed by folding and or ancient landslides. In the Spokane and Grande Ronde areas thick
sections of sediments make up a major part of the landslide complexes.

5.9.7 Latah Formation

Much of eastern Washington is underlain with thick sequences of basaltic flow rock. These flows spread
out over a vast area that now comprises what is commonly known as the Columbia Plateau physiographic
province (see WSDOT GDM Section 5.9.6). Consisting of extrusive volcanic rocks, they make up the
Columbia River Basalt Group (Griggs, 1959). This geologic unit includes several basalt formations, each
of which includes several individual flows that are commonly separated from one another by sedimentary
lacustrine deposits (Smith et al., 1989). In the Spokane area, these sedimentary rock units are called the
Latah Formation.

Most of the sedimentary layers between the basalt flows range from claystone to fine grained sandstone in
which very finely laminated siltstone is predominant. The fresh rock ranges in color from various shades
of gray to almost white, tan and rust. Much of the finer grained strata contain leaf imprints and other
plant debris. Because of its generally poorly indurated state, the Latah rarely outcrops. It erodes rapidly
and therefore is usually covered with colluvium or in steeper terrain hidden under the rubble of overlying
basaltic rocks.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-24 December 2006
 Engineering Properties of Soil and Rock

The main engineering concern for the Latah Formation is its potential for rapid deterioration by softening
and eroding when exposed to water and cyclic wetting and drying (Hosterman, 1969). The landslide
potential of this geologic unit is also of great engineering concern. While its undisturbed state can
often justify relatively high bearing resistance, foundation bearing surfaces need to be protected from
precipitation and groundwater. Construction drainage is important and should be planned in advance of
excavating. Bearing surface protection measures often include mud slabs or gravel blankets.

In the Spokane area, landslide deposits fringe many of the buttes (Thorsen, 1989). Disoriented blocks
of basalt lie in a matrix of disturbed inter flow silts. The Latah Formation typically has low permeability.
The basalt above it is often highly fractured, and joints commonly fill with water. Although this source
of groundwater may be limited, when it is present, and the excavation extends through the Latah-basalt
contact, the Latah will often erode (pipe) back under the basalt causing potential instability. The Latah
is also susceptible to surface erosion if left exposed in steep cuts. Shotcrete is often used to provide a
protective coating for excavation surfaces. Fiber reinforced shotcrete and soil nailing are frequently used
to for temporary excavation shoring.

The Latah formation has been the cause of a number of landslides in northeast Washington and in Idaho.
Measured long-term shear strengths have been observed to be in the range of 14 to 17 degrees. It is
especially critical to consider the long-term strength of this formation when cutting into this formation or
adding load on this formation.

5.9.8 Seattle Clay

The Seattle clay consists of proglacial lacustrine deposits of silt and clay that were deposited during
the transition from the interglacial to glacial period. The Seattle clay has been glacially consolidated
and is typically very stiff to hard. The Seattle clay is encountered in the Puget Sound Lowland as a
discontinuous deposit. It is the primary deposit affecting engineering design in the drumoidial hills
located in Seattle along SR-5 and on Mercer Island. This deposit can be more than 50 meters thick in the
Seattle area (Troost and Booth, 2003).

As a result of the glacial consolidation, the Seattle clay unit generally has high locked in lateral stresses.
The locked in stresses have created problems in virgin excavations into this geologic unit. Fractures and
slickensides are commonly encountered in this unit. As excavations are completed, the unit experiences
a lateral elastic rebound which has led to slope instability and problems for shoring contractors if the
locked in stresses and rebound are not incorporated into the design. The failure mechanism consists of
joints and fractures opening up upon the elastic rebound response. Hydrostatic pressure buildup within
the joints and fractures can then function as a hydraulic jack to further displace blocks of the silt and clay.
Appreciable movement can drastically degrade the shear strength along the planes of movement leading
to progressive failures. Such instability occurred in the downtown Seattle area when cuts were made into
the Seattle Clays to construct Interstate 5.

Based on considerable experience, the long-term design of project geotechnical elements affected by this
geologic unit should be based on residual strength parameters. For Seattle clays, the relationship between
the residual friction angle and the plasticity index as reported in NAVFAC DM7 generally works well
for estimating the residual shear strength (see Figure 5-1). In practice, residual shear strength values that
have been estimated based on back-analysis of actual landslides is in the range of 13 to 17 degrees.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-25
Engineering Properties of Soil and Rock

Figure 5-1. Correlation between residual shear strength of overconsolidated

clays and plasticity index (after NAVFAC, 1971).

It is important to note that residual strength of this formation is often achieved with a shear strain of 5%
of less as measured in triaxial or direct shear devices. For plane strain conditions, which typically govern
in slope stability and retaining wall applications, this critical shear strain could be significantly less than
this.

Another feature of this deposit is that it tends to be laminated with fine, water bearing sands. This
situation often results in instability, not only in open cuts, but also in the form of caving in relatively small
diameter shaft excavations.

Along the sides of these drumoidial hills containing Seattle clay are deposits from landslides that occurred
between glacial periods as the glaciers temporarily receded. Subsequent glacial advances partially
reconsolidated these ancient landslide deposits, but in the process left a highly fractured and jumbled
deposit of tilted clay blocks and displaced soil. These deposits can be highly unstable, and undrained or
peak drained shear strength parameters should not be used even for temporary designs or for wall types
that do not allow displacement (i.e., cylinder pile or tieback walls that are designed for K0 conditions).

As with most fine grained soils, wet weather construction in Seattle clay is generally difficult. When
the moisture content of these soils is more than a few percent above the optimum moisture content, they
become muddy and unstable and operation of equipment on these soils can become difficult.

Even though this geologic deposit is a clay, due to the highly overconsolidated nature of this deposit,
settlement can generally be considered elastic in nature, and settlement, for the most part, occurs as the
load is applied. This makes placement of spread footings on this deposit feasible, provided the footing
is not placed on a slope that could allow an overall stability failure due to the footing load (see WSDOT
GDM Chapter 8).

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-26 December 2006
 Engineering Properties of Soil and Rock

5.9.9 Bellingham Glaciomarine Drift

Glaciaomarine drift typically consists of unsorted, unstratified silt and clay with varying amounts of
sand, gravel, cobbles and occasional boulders. Glaciomarine drift is derived from sediment melted out of
floating glacial ice that was deposited on the sea floor. This material locally contains shells and wood.

The Bellingham area glaciomarine drift is found west of the Cascade Mountains and from about the
Mount Vernon area north. It is typically found at the surface or below Holocene age deposits. The upper
portion of this unit, sometimes to about 15 feet of depth, can be quite stiff as a result of desiccation or
partial ice contact in upland areas. This material typically grades from medium stiff to very soft with
depth. The entire glaciomarine drift profile can be stiff when only a thin section of the drift mantles
bedrock at shallow depths. Conversely, the entire profile is typically soft in the Blaine area and can be
soft when in low, perennially saturated areas.

This geologic unit can be very deep (150 ft or more). The properties of this unit are extremely variable,
varying as a function of location, depth, loading history, saturation and other considerations. The
soft to medium stiff glaciomarine drift has very low shear strength, very low permeability and high
compressibility. Based on vane shear and laboratory testing of this unit, the soft portion of this unit below
the stiff crust typically has undrained shear strengths of approximately 500 to 1000 psf, and can be as
low as 200 to 300 psf. The upper stiff crust is typically considerably stronger, and may be capable of
supporting lightly loaded footing supported structures. Atterberg limits testing will typically classify the
softer material as a low plasticity clay; although, it can range to high plasticity. Consolidation parameters
are variable, with the compression index (CC) in the range of 0.06 to over 0.2. Time rates of consolidation
can also be quite variable.

Wet weather construction in glaciomarine drift is very difficult because of the relatively high clay content
of these soils. When the moisture content of these soils is more than a few percent above the optimum
moisture content, glaciomarine drift soils become muddy and unstable, and operation of equipment on
these soils can become very difficult. Localized sandy and gravelly layers in the drift can be saturated
and are capable of producing significant amounts of water in cuts. Glacial erratics, such as cobbles and
boulders, can also be present.

5.9.10 Coastal Range Siltstone/Claystone

The Coast Range, or Willapa Hills, are situated between the Olympic Mountains to the north and
the Columbia River to the south. Thick sequences of Tertiary sedimentary and volcanic rocks are
present. The rocks are not intensely deformed and were not subjected to tectonism or the associated
metamorphism (Lasmanis, 1991). The Willapa Hills have rounded topography and deep weathering
profiles. The interbedded sandstone and fine grained sedimentary formations are encountered in highway
cuts. The material from these cuts has been used in embankments. Some of the rock excavated from
these cuts will slake when exposed to air and water and cause settlement of the embankment, instability
and pavement distortion.

Locally thick clay soils are present and extensive areas are underlain by sedimentary and volcanic rocks
that are inherently weak. Tuffaceous siltstone and tilted sedimentary rocks with weak interbeds are
common. The volcanic units are generally altered and or mechanically weak as a result of brecciation.
The dominant form of landslide in this area is the earthflow or slump-flow (Thorsen, 1989). Many of
these are made up of both soil and bedrock. Reactivation of landslide in some areas can be traced to
stream cutting along the toe of a slide.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-27
Engineering Properties of Soil and Rock

5.9.11 Troutdale Formation

The Troutdale Formation consists of poorly to moderately consolidated silt, sand and gravel deposited
by the Columbia River. These deposits can be divided into two general parts; a lower gravel section
containing cobbles, and upper section that contains volcanic glass sands. The formation is typically a
terrestrial deposit found in the flood plane of the Columbia River and the Portland Basin. The granular
components of the formation are typically well-rounded as a result of the depositional environment and
are occasionally weakly cemented. Occasional boulders have been found in this formation. Excavation
for drilled shafts and soldier piles in these soils can be very difficult because of the boulders and cemented
sands.

Slope stability issues have been observed in the Troutdale Formation. Significant landslides have
occurred in this unit in the Kelso area. Wet weather construction can be difficult if the soils have
significant fines content. As described above, when the moisture content of soil with relatively high fines
content rises a few percent above optimum, the soils become muddy and unstable. Permeability in this
soil unit varies based on the fines content or presence of lenses or layers of fine grained material.

5.9.12 Marine Basalts - Crescent Formation

The Crescent Formation basalts were erupted close to the North American shoreline in a marine setting
during Eocene time (Lasmanis, 1991). The formation consists mostly of thick submarine basalt
flows such as pillow lavas. The Crescent Formation was deposited upon continentally derived marine
sediments and is locally interbedded with sedimentary rocks. The Crescent Formation extends from the
Willapa Hills area to the Olympic Peninsula. During the middle Eocene, the Crescent Formation was
deformed due to accretion to North America. The pillow basalts have extensive zones of palagonite.
Along the Olympic Peninsula the basalts are generally highly fractured and are often moderately
weathered to decomposed.

The properties of the marine basalts are variable and depend on the amount of fracturing, alteration and
weathering. Borrow from cut sections is generally suitable for use in embankments; however, it may not
be suitable for use as riprap or quarry spalls because of degradation. All marine basalts should be tested
for degradation before use as riprap or quarry spalls in permanent applications.

5.9.13 Mélange Rocks on Olympic Peninsula

During the middle Miocene, convergence of the Juan de Fuca plate with the North American plate
accelerated to the point that sedimentary, volcanic, and metamorphic rocks along the west flank of
the Olympics were broken, jumbled, and chaotically mixed to form a mélange (Thorsen, 1989). This
formation is known as the Hoh rock assemblage. Hoh mélange rocks are exposed along 45 miles of
the western coast. Successive accretionary packages of sediments within the core of the mountains are
composed of folded and faulted Hoh and Ozette mélange rocks. Typical of mélange mixtures, which have
been broken, sheared and jumbled together by tectonic collision, the Hoh includes a wide range of rocks.
Extensively exposed in headlands and terraces along the Olympic coast consisting of resistant sandstone
and conglomerated sequence. The mélange rocks may consist of pillow basalt, deep ocean clay and
submarine fans. Slopes in tilted sedimentary rocks that have been extensively altered and/or contain weak
interbeds have been undercut by wave action in places along the Strait of Juan de Fuca. Slump flows or
bedding plane block glides form along the interbeds.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-28 December 2006
 Engineering Properties of Soil and Rock

Because of the variability of the mélange rocks and the potential for failure planes, caution should be
used when designing cuts. An adequate exploration program is essential to determine the geometry and
properties of the soil and rock layers.

5.10 References
AASHTO, 1988, Manual on Subsurface Investigations.

ASTM, 2004, Annual Book of ASTM Standards.

Allen, T. M., Kilian, A. P., 1993, “Use of Wood Fiber and Geotextile Reinforcement to Build
Embankment Across Soft Ground,” Transportation Research Board Record 1422.

Bowles, J. E., 1979, Physical and Geotechnical Properties of Soils, McGraw-Hill, Inc.

Dunn, I. S., Anderson, L. R., Kiefer, F. W., 1980, Fundamentals of Geotechnical Analysis,
John Wiley & Sons, Inc.

Griggs, A.B., 1976, U.S. Geological Service Bulletin 1413, “The Columbia River Basalt Group in the
Spokane Quadrangle, Washington, Idaho, and Montana”.

Higgins, J. D., Fragaszy, R. J., Martin, T. L., 1987, Engineering Design in Loess Soils of Southeastern
Washington, WA-RD 145.1.
Higgins, J. D., Fragaszy, R. J.,1988, Design Guide for Cut Slopes in Loess of Southeastern Washington,
WA-RD 145.2.

Hoek, E., and Brown, E.T. 1988. “The Hoek-Brown Failure Criterion – a 1988 Update.”
Proceedings, 15th Canadian Rock Mechanics Symposium, Toronto, Canada.

Hoek, E., Carranza-Torres, C., and Corkum, B., 2002, “Hoek-Brown Criterion – 2002 Edition,”
Proceedings NARMS-TAC Conference, Toronto, 2002, 1, pp. 267-273.

Holtz, R. D. & Kovacs, W. D., 1981, An Introduction to Geotechnical Engineering, Prentice-Hall, Inc.

Hosterman, John W., 1969, U.S. Geological Survey Bulletin 1270, “Clay Deposits of Spokane Co. WA”.

Joseph, N. L., 1990, Geologic Map of the Spokane 1:100,000 Quadrangle, Washington – Idaho,
Washington Division of Geology and Earth Resources, Open File Report 90-17.

Kilian, A. P., Ferry, C. D., 1993, Long Term Performance of Wood Fiber Fills, Transportation Research
Board Record 1422.

Lasmanis, R., 1991, The Geology of Washington: Rocks and Minerals, v. 66, No. 4.

Geotechnical Design Manual M 46-03 Engineering Properties of Soil and Rock

December 2006 Chapter 5-29
Engineering Properties of Soil and Rock

Lunne, et al., 1997, Cone Penetration Testing in Geotechnical Practice, E & FN Spon, London.

Mayne, P. W., Christopher, B. R., DeJong, J., 1997, FHWA-HI-97-021, Subsurface Investigations, NHI
course manual #132031.

Meyerhoff, G. G. , Journal of Soil Mechanics and Foundation Division, American Society of Civil
Engineers, January, 1956.

NAVFAC, 1971, Design Manual: Soil Mechanics, Foundations, and Earth Structures, DM-7.

Peck, R. B., Hanson, W. E. and Thornburn, T. H., 1974, Fooundation Engineering, 2nd Edition,
John Wiley & Sons, New York.

Phoon, K.-K., Kulhawy, F. H., Grigoriu, M. D., 1995, “Reliability-Based Design of Foundations for
Transmission Line Structures,” Report TR-105000, Electric Power Research Institute, Palo Alto, CA.

Sabatini, P. J., Bachus, R. C., Mayne, P. W., Schneider, T. E., Zettler, T. E., FHWA-IF-02-034, 2002,
Evaluation of soil and rock properties, Geotechnical Engineering Circular No. 5.

Schmertmann, J. H., 1967, Research Bulletin No. 121A, Florida Department of Transportation;
University of Florida.

Smith, G. A., Bjornstad, B.N., Fecht, K.R., 1989, Geologic Society of America Special Paper 239,
“Neogene Terrestrial Sedimentation On and Adjacent to the Columbia Plateau, WA, OR, and ID”.

Thorsen, G. W., 1989, Landslide Provinces in Washington, Engineering Geology in Washington,

Washington Division of Geology and Earth Resources, Bulletin 78

Troost, K.G. and Booth D.B. (2003), Quaternary and Engineering Geology of the Central and Southern
Puget Sound Lowland. Professional Engineering Practices Liasion Program, University of Washington,
May 1-3, 2003.

WSDOT Highway Runoff Manual M 31-16, March 2004.

WSDOT Standard Specifications for Road, Bridge, and Municipal Construction M 41-10, 2004.

Youd, T.L. and I.M. Idriss. 1997. Proceedings of the NCEER Workshop on Evaluation of Liquefaction
Resistance of Soils; Publication No. MCEER-97-0022.

Engineering Properties of Soil and Rock Geotechnical Design Manual M 46-03

Chapter 5-30 December 2006