ICDF Vol 1 Conference Program - Complete Book PDF

COVER 4 COVER 1
3 BOLIVIAN
rd
3rd BOLIVIAN INTERNATIONAL CONFERENCE ON DEEP FOUNDATIONS

INTERNATIONAL
CONFERENCE ON
DEEP FOUNDATIONS
April 27 – 29, 2017

Santa Cruz de la Sierra, Bolivia
PROCEEDINGS
VOLUME 1
VOLUME 1
64427 ICDF Vol 1 Conference Cover_rb.indd 1 4/4/2017 10:06:35 AM

PROCEEDINGS
of the
3rd BOLIVIAN INTERNATIONAL
CONFERENCE ON DEEP FOUNDATIONS
April 27 – 29, 2017

VOLUME 1
Invited Lectures
Edited by
Bengt H. Fellenius
K. Rainer Massarsch
Alessandro Mandolini
Mario Terceros Herrera
Design, execution, monitoring
and interpretation of deep foundation methods
ORGANIZING COMMITTEE
Conference Chairman:
Mario Terceros Herrera (Bolivia)
Conference Advisory Committee:

Bengt H. Fellenius (Canada)
Alessandro Mandolini (Italy)
K. Rainer Massarsch (Sweden)
Chairman of B.E.S.T. Prediction Event:

Bengt H. Fellenius (Canada)
© 2017, 3° C.F.P.B.
Copyright Information
Manuscripts are published according to an exclusive publication
agreement between the author and the conference organizer. Authors
retain copyright to their works.
Printed by Omnipress, Madison, WI USA

Table of Contents
Volume 1
Preface ........................................................................................................................v
Conference Program ............................................................................................... vii
Invited Lectures..........................................................................................................1
Electrical Resistance Strain Gages in Instrumentation of Deep Foundations ............................................... 1

Albuquerque, P.J.R., Brazil
Pile Integrity Testing: History, Present Situation and Future Agenda ........................................................ 17
Amir, J.M., Israel
Advances in the Use of Drilled Foundations .............................................................................................. 33

Brown, D., USA
Best Practice for Performing Static Loading Tests. Examples of Test Results with
Relevance to Design ................................................................................................................................... 63
Fellenius, B.H., Canada
Conventional methods and recent developments in retaining walls ........................................................... 75

Gerresen, F.W., Germany
Recent Developments in Helical Piles ........................................................................................................ 89

Lutenegger, A.J., USA
Simple Approach to Static and Seismic Design of Piled Rafts ................................................................. 107
Mandolini, A, Di Laora, R. and Iodice, C., Italy
Recent developments in vibratory driving and soil compaction ............................................................... 125

Massarsch, K.R., Sweden
Recent Developments and Applications in Geotechnical Field Investigations for

Deep Foundations ..................................................................................................................................... 141
Mayne, P.W. and Niazi, F.S., USA
Effects of Installation Processes on the Axial Capacity of Pile Foundations in Sand .............................. 161
Prezzi, M., USA and Basu, P., India
Dynamic Loading Tests: A State of the Art of Prevention and Detection of

Deep Foundation Failures ......................................................................................................................... 177
Rausche, F., Alvarez, C. and Likins, G.E., USA
3rd Bolivian International Conference on Deep Foundations—Volume 1 iii

Deep Foundation Concepts in Energy Geo-engineering. Special Cases................................................... 199
Santamarina, J.C. and Shin, H., Saudi Arabia
Expander Body and Toe-Box: Expansion Devices for Deep Foundations Enhancement ........................ 209
Terceros A. M. and Terceros H. M.A., Bolivia
Publishing Policy ...................................................................................................229
iv 3rd Bolivian International Conference on Deep Foundations—Volume 1

PREFACE
The 3rd International Conference on Deep Foundations is held April 27 – 29, 2017 in Santa Cruz
de la Sierra, Bolivia. It follows two successful conferences held in 2013 and 2015. The conference
is organized with the support of INCOTEC SA in association with the Society of Engineers of
Bolivia, the Bolivian Society of Soil Mechanics and Geotechnical Engineering and the Chamber
of Construction of Santa Cruz. It is held at the UPSA Campus (Universidad Privada de Santa
Cruz), the main private university of the city and arranged with the support of the International
Society of Soil Mechanics and Geotechnical Engineering (ISSMGE), Technical Committee 212,
“Deep Foundations”.
The principal objective of the conference is to bring together local engineers and international
experts in order to facilitate the exchange of experience and to introduce to the region new design
concepts, methods and equipment for the application to deep foundations. The conference program
is composed of invited lectures, discussions, a field demonstration, and a pile testing prediction
event where international experts have been invited to predict the load-movement response of piles
in static loading carried out prior to the conference.
During the first two days of the conference, speakers of international repute have been invited to
present papers on specific topics, covering different aspects of deep foundations. The third day of
the conference is devoted to the presentation and discussion of tests a comprehensive pile testing
program. The Bolivian Experimental Site for Testing Piles (B.E.S.T.) was adopted by ISSMGE
TC 212 as a reference site for investigations on piles and pile groups. B.E.S.T. offers a unique
possibility to enhance the understanding of the performance of different pile types and pile groups
when subjected to load. The geotechnical conditions at the B.E.S.T. site have been documented
by detailed investigations, using state-of-the art testing and interpretation methods. The results of
the field testing programme, including interpretation of in-situ methods and results of the pile
loading tests will be presented during the third day of the conference.
Volume 1 of the proceedings comprises the papers presented at the conference. All papers have
been reviewed by at least two members of the Review Committee. The dedicated work by the
reviewers and their valuable contributions is gratefully acknowledged.
Volume 2 contains a description of the geological setting and the results of comprehensive
geotechnical investigations carried out at the B.E.S.T. site. It is the intention of the Conference
Organizers to make available all data from the B.E.S.T. site investigations and pile tests in digital
format at the conference web platform for use in future investigations, in cooperation with
ISSMGE TC 212.
Volume 3 includes a description of the test piles and the loading test programme. The predictions
as well as a presentation of test results will be published in a Volume 3 after the conference.
3rd Bolivian International Conference on Deep Foundations—Volume 1 v

vi 3rd Bolivian International Conference on Deep Foundations—Volume 1
3rd Bolivian International Conference on Deep Foundations
Conference Program
April 27th 2017
8:00- Registration
8:30-8:45 OPENING SESSION

Official opening of conference
8:45-9:45 SESSION 1
Paul Mayne (United States of America)
Recent developments and applications in geotechnical field Investigations - Session
Chairman: Jorge Alva (Perú)
9:45 -10:15 COFFEE BREAK
10:15-11:15 SESSION 2 – “Bengt H. Fellenius Lecture”

Frank Rausche (United States of America). Dynamic Loading Tests: A
State of the Art of Preventing and Detecting Deep Foundation Failures
Session Chairman: Bengt H. Fellenius
11:15-12:15 Presentation of Invited Papers 1 and 2

Carlos Santamarina (Argentina): Casos interesantes de interacción
suelo-estructura (Interesting cases of soil-structure interaction).
Tomás Murillo (Spain): Diseño y ejecución de pilotes con celda de

inyección en punta. Aplicación en ríos amazónicos (Design and execution of piles with
pile toe injection. Application in Amazonian rivers).
Session Chairman: Paulo Albuquerque (Brazil)
12:15-12:45 Opening and overview of Exhibition
12:45-14:00 LUNCH BREAK AND DIGITAL POSTER SESSION

Alan Lutenegger (United States of America): Recent development in helical piles.
Carlos Medeiros (Brazil): Energy measurement for CFA piles capacity
estimation.
Session Chairman: Morgan Nesmith (United States of America)
15:00-16:00 SESSION 3.
Bengt Fellenius (Canada): Discussion on best practice for performing static loading tests.
Examples of test results and relevance to design.
Session Chairman: Luciano Decourt (Brazil)
3rd Bolivian International Conference on Deep Foundations—Volume 1 vii

16:00-16:30 COFFEE BREAK
16:30-17:30 SESSION 4
Joram Amir (Israel): Testing pile integrity - past present and future.
Session Chairman: Victor Hugo Alvarez (Bolivia)
17:30-18:30 SESSION 5 – “K. Rainer Massarsch Lecture”

Oscar Vardé (Argentina): Subway stations retaining walls: Case
histories in soft and hard soils.
Session Chairman: K. Rainer Massarsch (Sweden)
April 28th 2017
8:30-9:30 SESSION 6.
Dan Brown (United States of America): State of art and state of practice in deep foundations.
Session Chairman: Oscar Vardé (Argentina).
10:30-11:30 SESSION 7.
Alessandro Mandolini (Italy): Design options for piled rafts - An overview.
Session Chairman: Juan Carlos Rojas V (Bolivia)

Mónica Prezzi (United States of America): Effects of installation processes on axial pile capacity
Mario Terceros Arce (Bolivia): Practical application of new expansion devices for pile
improvement in sandy soils.
Session Chairman: Walter Paniagua (Mexico)
14:30-15:30 SESSION 8.
Franz Werner Gerresen (Germany): Conventional methods and recent developments in retaining
walls.
Session Chairman: Carlos Medeiros (Brazil)
15:30-16:30 SESSION 9.
K. Rainer Massarsch (Sweden): Recent developments in vibratory driving and soil compaction.
Session Chairman: Bengt H. Fellenius (Canada).
viii 3rd Bolivian International Conference on Deep Foundations—Volume 1

17:00-18:00 SESSION 10.
Jarbas Milititsky (Brazil): Deep foundations pathologies.
Session Chairman: Renato Cunha (Brazil)
20:00 Gala Dinner
April 29th 2017
8:30-9:00 B.E.S.T. Presentation. Alesandro Mandolini (Italy), Chairman of TC212

Bengt H. Fellenius (Canada), Mario Terceros (Bolivia), Paulo Albuquerque (Brazil) and Peter K.
Robertson (Canada).
Session Chairman: Roger Frank (France), ISSMGE President
9:00-10:15 Session 1 – Soil Investigation Results. - Memorial Session in Honour of Prof. Silvano Marchetti
SCPT: Peter Robertson (Canada)
SDMT: Diego Marchetti (Italy)
SPT-T and DPSH: Luciano Decourt (Brazil)
PMT: Roger Frank (France)
GEOPHYSICS: K. Rainer Massarsch (Sweden)
Session Chairman: Paul Mayne (United States of America)
10:35-12:05 SESSION 2: Pile Installation and Instrumentation

EB and Toe Box: Mario Terceros A. (Bolivia)
FDP CFA and Drilled Piles: Morgan Nesmith (United States of America)
Micropiles: Mario A. Terceros H. (Bolivia)
Helical Piles: Alan Lutenegger (United States of America)
Session Chairman: Oscar Vardé (Argentina)
14:00-16:00 SESSION 3: Experimental Results and Predictions

Prediction Outcomes: Bengt H. Fellenius (Canada)
Pile Group and Piled Raft: Alessandro Mandolini (Italy)
Single Piles: Luciano Decourt (Brazil)
Session Chairman: Dan Brown (United States of America)
16:00-17:00 Mario Terceros Banzer Lecture

Luciano Decourt, (Brazil):Fifty years designing foundations. A retrospective
Session Chairman: Mario Terceros B (Bolivia)
17:00-17:20 CLOSING SESSION
3rd Bolivian International Conference on Deep Foundations—Volume 1 ix

x 3rd Bolivian International Conference on Deep Foundations—Volume 1
Electrical Resistance Strain Gages in
Instrumentation of Deep Foundations
Albuquerque, P.J.R.(1)
(1)
University of Campinas - Unicamp, Campinas, São Paulo, Brazil <pjra@fec.unicamp.br>
ABSTRACT. The purpose of this paper is to describe some instrumentation techniques for steel
piles and for cast-in-place piles by means of strain gages. A brief history of strain gages is
presented as well as the theoretical principles that the technique involves, showing the types of
connections used in pile instrumentation, indicating the advantages and disadvantages of the use
of each type. Determining the load distribution of a pile is the key objective of instrumentation.
Therefore, three ways are proposed to analyze the stress data collected. Also procedures for
instrumentation and installation of sensors in depth are presented. Instrumentation via strain gage
is a reliable technique commonly used in all areas of knowledge. However, it is necessary to adapt
the parameters one wishes to obtain to the technique to be used.
1. INTRODUCTION
The use of instrumentation to get parameters of the behavior of geotechnical structures is very
frequent nowadays in works such as: dams, slopes, foundations, containments, etc. For each type
of work and purpose of the data to be collected and analyzed, a different type of instrumentation
is used. According to Dunnicliff (1993), there are two general categories of measuring instruments.
The first category covers instruments used for in-situ determination of existing conditions of rock
and soil, such as resistance, compressibility, and permeability to be used in projects for which
purpose cone tests, vane tests, pressuremeter, etc, are used. The second category is used for
monitoring changing conditions during the construction phase and operation, such as water
pressure, total stresses, displacements, loads, and deformations.
Over the last decades, instrumentation equipment manufacturers have developed a variety of
materials and instruments for geotechnical monitoring. However, it is important to point out that
users must be aware of the desired parameters and of the techniques involved in each type of
instrument to get reliable and accurate information.
Several types of instruments may be used. In general, the most commonly used are based on
change of strain, employing vibrating wire gages, electrical resistance gages, and optical fiber
gages. Foundation engineering is constantly trying to learn about the behavior of deep foundations
in terms of load transfer. To this end, among the aforementioned techniques, strain gages are used.
This article is a study that aims to describe the technique of instrumentation of deep
foundations with the use of strain gages. The article will include both a theoretical approach and
practical experience.
2. ELECTRICAL RESISTANCE STRAIN GAGES
Electrical extensometers, also called strain gages, are not recent tools. In the 1930's, Edward
Simmons (California Institute of Technology, Pasadena, CA, USA) and Arthur Ruge
(Massachusetts Institute of Technology Cambridge, MA, USA) working separately, were the first
3rd Bolivian International Conference on Deep Foundations—Volume 1 1

engineers to use metal wires glued to the surface of a specimen to measure strain. The strain gages
included measuring the change of electrical resistance between with two terminals, mounted on a
support, having the purpose as an insulator.
Electrical extensometers are sensitive instruments that transform tiny strains in equivalent
variations of their electrical resistance. Using electrical extensometers is a means to measure and
record the phenomenon of strain as an electrical quantity. When an element is submitted to a
determined stress, the ensuing strain is transformed into an electrical signal that is captured,
decoded, and then translated into a numerical value.
There are several types of extensometers: inductance and capacitance extensometers, and also
piezoelectric extensometers. These showed later to be less successful as resistance extensometers,
however, they contributed considerably to the development of measurement of strains.
Current electrical resistance extensometers are wire extensometers and blade extensometers,
comprising an electricity conductive metal with large sensitivity to strain. Different from carbon
extensometers, these are not sensitive to variations in temperature and humidity. Electrical-type
extensometers are used to measure strain in different structures, such as bridges, machines,
locomotives, ships etc.
Awareness of internal stress is of great importance, since it is through stress that structures are
sized. The stress vs. strain characteristic of structural materials, such as steel, is sufficiently well
understood nowadays for most practical applications.
Strain depends on the stress applied and the limitations of the specimen. The characteristics of
the specimen material are resistance threshold, proportionality threshold, and yield point.
Strains caused by the action of loads or other external influences or by own weight, can be
measured with extensometers, same as distortion, which is an angular variation, although with a
certain complexity in comparison to measurement of linear strain. As most calculations are based
on stress, it is necessary to transform the effect of strains so they can be measured as stress. The
total strain in any direction comprises three portions: strain caused by variation in temperature;
strain due to Poisson effect, and primary strain which is stress in the direction of the imposed load.
As the device is glued to the surface and the part deforms, the value of its electrical resistance
changes proportionally; then the key problem here is humidity. However, gluing and sealing
techniques protect these sensors against humidity and protects against zero drift.
Sensitivity to temperature variations is one of the most important factors to be considered when
using strain gages due to two key factors: the difference in elongation among the part, the grid
support and the grid itself, and the variation in resistivity depending on temperature.
Results may vary as the temperature changes while strains are measured. The change in
temperature influences the material's linear expansion and that of the extensometer wire, besides
the specific resistance of the wire, as shown in Eq. (1) which will serve both for free expansion of
the extensometer and for the material:
∆𝐿𝐿 = 𝐿𝐿0 𝛼𝛼∆𝑡𝑡 (1)
where:
∆𝐿𝐿 = variation in length both of the extensometer wire and of the specimen
𝐿𝐿0 = initial length
𝛼𝛼 = material expansion coefficient, or resistivity coefficient
∆𝑡𝑡 = variation in temperature
2 3rd Bolivian International Conference on Deep Foundations—Volume 1

To minimize the effects of temperature variation regarding of environment, different types of
assemblies of strain gages can be used in resistive bridges, such as: 3-cable assembly (1/4 bridge),
2-cable assembly (half bridge), and full bridge.
The strain measured by the extensometer must be the same as the strain experienced by the
element studied. The imposed strain is proportional to the change of the measured change of
resistance via a gage factor as indicated in Eq. 2.
∆𝐿𝐿 ∆𝑅𝑅⁄𝑅𝑅
𝜀𝜀 = = (2)
𝐿𝐿 𝐾𝐾
where: 𝜀𝜀 = total strain

∆𝐿𝐿 = variation in length
𝐿𝐿 = initial length
∆𝑅𝑅 = variation in resistance
𝑅𝑅 = resistance
𝐾𝐾 = strain gage-factor, which ranges from 1 to 200 (commonest value is 2.0),
R = 60 – 10,000 Ω, (commonest value is 120 Ω or 350 Ω).
The variance of resistance caused by a change of strain is determined by a Wheatstone bridge.

A device having four resistors connected two by two in parallel as shown in Figure 1.
Fig. 1. Wheatstone bridge.
Using Eq. 2 and the principle of the Wheatstone bridge, we get to the generic Eq. 3:
Δ𝑅𝑅
= 𝐾𝐾 (𝜀𝜀1 − 𝜀𝜀2 + 𝜀𝜀3 − 𝜀𝜀4 ) (3)
𝑅𝑅
where: ε1 = strain due gage one

ε2 = strain due gage two
ε3 = strain due gage three
ε3 = strain due gage four
Being:
εi= εt + εf + εn (4)

Where: εt = strain due to temperature
εf = strain due to bending
εf = strain due to normal stress
Among the characteristics of electrical resistance strain gages are high measuring accuracy,
excellent dynamic response, and excellent linearity; they can be used immersed in water or in
corrosive gas atmospheres provided that the suitable treatment is made; remote measurements can
be made, etc. Because of such characteristics, this sensor has many applications in experimental
studies.
2.1 Types of circuits

Several types of circuits can be used in various areas. The type to be used will depend on the
property that one desires to obtain (momentum, resistance, torsion etc). Figures 2, 3 and 4 shows
the most commonly used types of circuits.
A) Half-bridge circuit with one active branch (Eq. 5):

VK VK
∆Vd = (ε 1 ) = (ε t + ε f + ε n )
4 4 (5)
Fig. 2. ¼ bridge with one active branch.
• ¼ bridge with 2 wires: In this type of connection, the strain gage signal could be affected
by variations in resistance of the cable because of temperature and change in length.
Depending on the cable length, the instrument will not be able to balance the bridge.
Calibration will be incorrect: the more incorrect, the greater the cable resistance. The
advantage of this type of connection is the use of a smaller number of cables. The
application of this connection is indicated for places with controlled temperature, short
cables with a large diameter, and short-term tests. For this type of circuit, absence of
bending and temperature change are required.

• ¼ bridge with 3 wires: The bridge balance is independent of cable length, but the cable
must have constant length and cross section; also the wires must be twisted with each other.
When the installation gets irradiation heat and the three cables are exposed, they must have
the same color, because the objects that absorb more colors within the spectrum create
more heat, if the wires have different colors they can affect the bridge balance. It must be
pointed out that the variation in cable resistance as a function of temperature does not
impact the measurements. The triple wires can cross places with high thermal variation
without impacting the measurements. This type of connection is very commonly used in
cases of long cables and large temperature variations along its length. It is indicated for
long-term tests. In this type of connection, the effect of temperature is eliminated but the
effect of bending is not.
B) Half-bridge circuit with two active branches (Eq. 6).

VK VK
∆Vd = (ε 1 − ε 2 ) = (2ε f )
4 4 (6)
Fig. 3. ½ Bridge with two active branches (adjacent arms).
This type of connection is suitable to situations of long-term tests with large thermal variation.
In this type of connection, the effect of temperatures and axial strength is eliminated. The readings
of strains will be due to bending only. The signal is amplified two times.

C) Full-bridge circuit with four active branches (Eq. 7).
VK VK
∆Vd = (ε 1 − ε 2 + ε 3 − ε 4 ) = ε n 2(1 + ν )
4 4 (7)
where: ν = Poisson ratio
Fig. 4. Full bridge with 4 active branches.
• This type of connection is the most suitable one to be used in situations of long-term tests
with large thermal variation of strain gages where it is necessary to eliminate the bending
effect. In this type of connection, the temperature effect is also eliminated.
2.2 Principles of analysis

The following examples will clarify how to apply the technique where a full bridge connection
with four active extensometers was used, so effects of temperature and bending strains were
eliminated and only strains from normal resistance were obtained.
The operational principle is simple: strain (ε) was measured for a pile element with a cross
section area, A, due to an applied increase of load was measured. Using Hooke's Law, we get Eq.
8.
P = AE c ε
(8)
where: P = load in the cross section

A = area of the pile cross section
Ep = pile modulus
ε = strain measured
By placing gages at selected depths in the pile, we obtain load at various depths and the load
distribution. Using Hooke's Law, we get the Young modulus in the pile's reference section and,
from that value on, the loads at each instrumented level are obtained. The slope of the load versus
strain diagram shown in Figure 5 is the pile axial stiffness, EpA. N.B., the stiffness is obtained
without input of the pile cross section, which is a variable for cast-in-place piles.

Fig. 5. Load vs. microstrain graph.
For the example of Fig. 5 in which product EpA is equal to 3,142,917 kN, the effective
diameter of the pile (bored with mechanical auger) was 0.44 m in average (A = 0.152 m2). This
way, the Young modulus is about 20.7 GPa.
Another way to analyze the instrumentation data is given by Fellenius (2016), using the
method of tangent modulus as indicated in Eq. 9.
 dσ 
Mt =   = aε + b
 dε  (9)
By integrating Eq. 9, we get to Eq. (10)

a (10)
σ =  ε 2 + bε
2
Eq. 11 shows the basic equation to calculate stress
σ = E s .ε (11)
Equaling Eqs. 10 and 11, we get Eq. 12.

E s . = 0,5aε + b
(12)
where: Mt = tangent modulus of composite pile material

Es = secant modulus of composite pile material
σ = stress of the pile cross section
a = slope of the tangent modulus line
ε = measured strain
b = ordinate intercept of the tangent modulus line

Figure 6 shows an example of application of the tangent modulus. The data used to get this figure
were the same as those used in Fig. 5.
Fig. 6. Tangent Modulus vs. microstrain graph
From the results shown in Fig. 6, a constant modulus of the order of 24.8 GPa (average) was
obtained.
As the strain gage level lies close to the pile, it is possible for determining both secant and
tangent modulus. The Fig. 7 shows that the tangent plot a bit of scatter, and the secant modulus is
less sensitive to such variations with a smoother curve, even though requires a well-established
zero level (Fellenius, 2016).
Fig. 7. Comparison between stiffness determined from tangent and secant modulus approaches

The analysis of the strain records produced the load distributions shown in Figure 8.
Fig. 8. Load transfer in depth (Albuquerque et al. 2011).
3. INSTRUMENTATION APPLIED TO PILE FOUNDATIONS
The experience is vast in the use of instrumentation via strain gages in pile foundations. In general
this technique involves instrumenting a steel bar that may or not be part of the pile cage in the case
of pre-cast or cast-in-place piles. For steel piles, the instrumentation must be made directly on the
metal profile. Below is a simplified presentation of the instrumentation processes.
3.1 Concrete Piles
As a general rule, the technique of installing instrumented bars inside the pile is used for this type
of piles, the difference being the executive procedure of the pile. In general, the instrumentation is
carried out with construction steel bars with 12.7 mm or 19.0 mm diameter, depending on the
diameter of the cross section of the pile. It is not appropriate using large cross section steel bars in
piles with small cross section, because the installation process can be arduous.
The manufacturing sequence of full-bridge bars is shown below. When preparing the bars, it
is important to know at what levels the bars will be placed so as to have appropriate cable lengths.
The cables to be used have 4 paths (full bridge connection), and must have a minimum cross
section area of 0.5 mm2 and an outer protective coating (PVC) and an inner protective lining with
twisted tin copper (Figures 9 - 14).

Fig. 9. Preparing the steel bar surface Fig. 10. Marking the place to glue the strain gage
Fig. 11. Gluing the strain gage Fig 12. Wheatstone bridge connection
Fig 13. Checking the bar on a tensile testing Fig 14. Instrumented bar guide
press

Once the instrumentation steps are over, the procedures for field installation start. In cases of
piles where the concreting is executed after insertion of the reinforcing cage, it is possible to fasten
the instrumentation on the cage (Fig. 15). Details of the installation technique have been presented
by Freitas Neto (2013), Polido et al. (2014), and Garcia (2015).
Fig. 15. Fastening the instrumented bar in the pile reinforcing cage
Gages can be installed in continuous auger flight piles or full displacement piles by inserting
a steel tube (smooth or corrugated; with external diameter of approximately 50 mm). For
continuous auger flight piles and full displacement piles, the tube is inserted before the pile is
drilled. Figure 16 shows the tube to be placed into the shaft of the continuous flight auger pile.
Figure 17 shows the tube extending above the pile head after completed installation (Albuquerque
2001).
Fig. 16. Tube being placed in the pile Fig 17. Tube after the pile is constructed

After the pile is completed, the instrumented bars are installed at the predetermined levels
inside the tube and the tube is grouted. The 500 mm bars are connected and placed inside a
galvanized tube in a predefined position to form a continuous bar. To make it possible to splice
the bars, a threading system of the ends is used with coupling of sleeves of the same material (Fig
18). After placing the bars inside the tube, it is filled with cement slurry with water / cement factor
equal to 0.5. Fig 19 shows the finished process.
Fig 18. Bar splicing Fig 19. Finished installation
Once the bar installation is completed, the step of forming and casting of the pile cap starts.
Fig. 20 shows the details. We can see the black side tube used as exit for the instrumentations
cables. Fig. 21 shows the cables coming out on the side of the pile crowning block at the time of
the loading test.
Fig. 20. Preparing the crowning block. Fig. 21. Completed block

3.2 Steel piles
For this steel piles, the instrumentation must be carried out on site by gluing the strain gages to the
pile section (Albuquerque and Melo 2014; Albuquerque et al. 2016). These spots must be arranged
so as to supply information from different depths as required.
The sensors are directly adhered (by cyanocrylate) to the core, after treating the surface with
a degreasing product. Strain gages are guarded against humidity and mechanical shocks by means
of a silicone resin, adhesive tape, and electrical guarding resin, besides epoxy resins and metal
fishplates in the section made of metal profiles.
The metal profile must be installed in several steps:
 Welding of metal fishplates at the joint between the base and the core of the metal profile in
order to provide protection against mechanical shock when the metal profile is driven (figure
22) and cables are passed through;
 Cleaning and treating the surface of the metal profile to be fastened to the extensometer later
on;
 Fastening the sensors (Figure 23) and applying resin to protect strain gages from humidity and
mechanical shock;
 Closing the instrumentation levels between the metal fishplates with epoxy resin for protection
against mechanical shock when driving the piles;
 Fastening electrical cables, which are connected to the strain gages along the instrumentation
levels of the metal segment in the upper part of the metal profile.
 Lifting and driving the metal profile must be performed with utmost care so as not to break the
cables.
Fig 22. Guarding fishplates. Fig 23. Strain gage fastened to the surface

While the pile is being driven, it is possible to get the strains resulting from each instrumented
level. To do so, a data acquisition system must be used to collect data with a minimum frequency
of 10 Hz. Fig. 24 shows the variation in strain obtained for each blow of the hydraulic hammer at
the level of the pile head when the pile driving is completed.
µm/m
16.03.13 17:24:46 100,00 ms (10 Hz)
445
112
223
334
445
556
667
778
889
1000
1111
1222
1333
1444
1555
1666
1777
1888
1999
2110
2221
2332
2443
2554
2665
2776
2887
2998
3109
3220
3331
3442
3553
3664
3775
3886
3997
4108
4219
4330
4441
4552
4663
4774
4885
4996
5107
5218
5329
5440
5551
5662
5773
5884
1
450
455
460
465
470
475
Fig. 24. Strain of each blow of the hydraulic hammer.
With this type of piles, the strains are obtained during and after driving. The value of residual
load can be obtained prior to the execution of the slow maintained load test.
4. FINAL CONSIDERATIONS
• The instrumentation technique via strain gages is widely used from medicine through
engineering, which proves that the technique is reliable. However, lack of theoretical and
practical knowledge may lead to unreliable results;
• Using strain gages to get strains and loads in deep foundations is a reliable technique that
can be applied to practically all types of piles;
• It is important to stress that the type of connection to be used (full bridge, ½ bridge or ¼
bridge) must be carefully considered so as to get highly reliable data;
• The instrumentation in cast-in-place piles or pre-cast piles (steel or concrete) is more
effective when the bars have been previously instrumented at the laboratory, and are
installed on site. In the case of continuous flight auger piles or another similar type of
concreting, a different installation technique should be used;
• In the case of steel piles, direct instrumentation in the profile proved to be appropriate;
however the procedure must be carried out on site and very carefully since preparation of
the surface, cleaning and gluing the sensors are key considerations for success. Another
factor that requires attention is during pile driving, since cables may get damaged and
jeopardize the entire process.

5. REFERENCES
Albuquerque, P. J. R., 2001. Estacas escavadas, hélice contínua e ômega: estudo do

comportamento à compressão em solo residual de diabásio, através de provas de carga
instrumentadas em profundidade (Bored, continuous flight auger and omega piles: study of
compressive behavior in diabase residual soil, through depth instrumented load tests). PhD
Thesis. Escola Politécnica, Universidade de São Paulo, São Paulo. 272 p.
Albuquerque, P. J. R., Massad, F., Viana da Fonseca, F., Carvalho, D., Santos, J., Esteves, E.C.,
2011. Construction Technique Influence on Piles Performance: Evaluation by Instrumentation
Part 1: University of Campinas - Unicamp - Experimental Field. Soils and Rocks. 24(1).pp-
25-50.
Albuquerque, P.J.R., Melo, E.O., 2014. Emprego de extensômetros elétricos de resistência para
instrumentação de estacas metálicas (Use of electrical resistance strain gages for
instrumentation of steel piles). Proceedings of the 17th COBRAMSEG, Goiânia, 6p.
Albuquerque, P.J.R., Melo, E.O., Melo, A.C., 2016. Análise de previsão de capacidade de carga
de estaca mista pré-moldada e metálica em solo sedimentar da cidade do Recife/PE. (Analysis
of bearing capacity of precast and steel mixed pile in sedimentary soil of the city of Recife /
PE). Proceedings of the 18th COBRAMSEG, Belo Horizonte, 8p.
Dunnicliff, J., 1993. Geotechnical instrumentation for monitoring field performance. John Wiley
and Sons, New York. 608p.
Fellenius, B.H., 2016. Basics on foundations design. Electronic Edition. 454 p. www.fellenius.net.
Freitas Neto, O. 2013. Avaliação experimental e numérica de radiers estaqueados com estacas
defeituosas em solo tropical do Brasil (Experimental and numerical evaluation of pile rafts
with defective piles in tropical Brazil soil) . PhD. Thesis, UnB, Brasília, G.TD-088/2013. 253.
Garcia, J. R., 2015. Análise experimental e numérica de radiers estaqueados executados em solo
da região de Campinas/SP (Experimental and numerical analysis of pile rafts executed in
Campinas/SP soil). PhD. Thesis. Faculdade de Engenharia Civil, Arquitetura e Urbanismo da
Universidade Estadual de Campinas. 359 p.
Polido, U.F., França, H.F., Albuquerque, P.J.R., Felix, M., Koehler, T., 2014. Estaca tipo
ECOPILE - prova de carga à compressão instrumentada (ECOPILE - instrumented
compression load test). Proceedings of the 17th COBRAMSEG, Goiânia, 8p.

Pile Integrity Testing:
History, Present Situation and Future Agenda
Amir, J.M.(1)
(1)
Piletest.com Ltd., Herzlia, Israel <jmamir@piletest.com>
ABSTRACT. Deep foundations have served humanity for the last few millennia but really fast
progress in piling systems and equipment had to wait until the 20th century. Today, piles can be
produced in all soil and rock formations, in diameters reaching four meters and depths of 150 m
or more. Moreover, it has become apparent that even the most advanced piling technology cannot
assure perfect products. As a results, the 1960's triggered the new discipline of integrity testing of
deep foundations. Present methods of integrity testing are either not-intrusive (mainly acoustic) or
intrusive, the latter requiring the installation of access ducts during casting. Currently, the analysis
of test results is invariably an inverse problem, thus, a unique accurate solution is still unavailable.
Future research will have to concentrate on the integration of all testing methods with common
interface and on advanced digital analysis methods.
1. HISTORY
1.1 Piling Technology
Since prehistoric times, humankind has looked for lakeside and riverside dwelling sites that offered
both ample water and protection from attack. To support their dwellings above high water level in
the muddy soil, driven timber piles were the obvious solution. Modern radiocarbon dating
technology established that timber piles recently discovered in London (Figure 1a) were more than
6,000 years old and are still in reasonable shape (Milne et al. 2010). This foundation method was
successfully practiced over the millennia in cities like Venice and Amsterdam. A painting by
Maximilien Luce (Figure 1b) proves that manual piledrivers were used to drive timber piles as late
as the twentieth century even in a developed city like Paris.
a b
Fig. 1: Wood piles - (a) London 4,600 BC, (b) Paris1905.

The demand for supporting heavier loads led to the use of stone-filled well foundations, such
as those supporting the Taj Mahal. The 19th century heralded new construction materials that the
piling community was quick to adapt: Portland cement (1824), steel tubes (1825), steel I-beams
(1849) and reinforced concrete (1867). The hand-dug Chicago Caisson, filled with reinforced
concrete, was introduced in 1893 to successfully support heavy high-rise buildings (White 1962).
During the 20th century, bored piles became prevalent due to the development of larger and
stronger drilling rigs. Currently equipment is available for constructing piles of practically any
diameter and length in all soil and rock formations, above and below water. Due to the proliferation
of piling equipment and dropping prices, piles have largely replaced labor-intensive spread
footings (Amir 1983). On the other hand, piles are sensitive to both soil conditions and standard
of workmanship. The first published report of defective piles (Hobbs 1957) reported unusual
necking of piles in South Africa cast in loose sand under artesian water. Szechy (1961) describes
a project in Budapest where piles cast under water were flawed and unable to support the design
loads. Since then the issue of pile integrity gradually began to play a growing part in quality
assurance programs of construction projects. Pile integrity may be defined as meeting the project
requirements, specifically physical dimensions, material properties and verticality. Any failure to
meet the above requirements is defined a flaw.
1.2. Flaw Occurrence

The process of constructing bored piles is essentially invisible. Therefore, especially when piling
in difficult soil conditions and in the presence of ground water, it is reasonable to expect a certain
percentage of flawed piles. Fleming et al. (1992) describe many situation that can lead to the
creation of flaws in both driven and bored piles. In a survey of 49,000 piles tested in five countries
(Amir and Amir 2008) showed that 1.85% of the piles had identified flaws, in addition to 6% that
were too short by 20% or more—a total flaw rate of close to 8%. Much higher flaw rates, up to
76%, were found on several sites where workmanship was sub-standard. Such high flaw rates are
evidently unacceptable, and can be avoided only by applying an integrity testing program as early
as possible.
1.3. Pile Integrity Fiascoes

1.31 Chicago 1966
The John Hancock Center in Chicago, when completed in 1969, was one of the tallest buildings
of the world. The construction of its foundations, however, was inflicted with trouble. The tower
was designed to rest on 57 concrete caissons bored by a massive drill rig to bedrock, 60 m below
grade. The caissons were cased in stages: A length of casing was lowered into the hole and filled
with concrete. Once the concrete began to harden, the casing was pulled up until its bottom was
barely embedded in concrete. This process was repeated until the caisson was concreted to ground
level. Once all the caissons were completed, steel erection for the superstructure began. The alarm
was sounded when one of the caissons started to settle excessively under the first floor column—
a mere 120-kN load. The engineers immediately stopped the project and embarked on an extensive
testing program that included coring and a number of novel non-destructive testing techniques
(Baker and Khan 1971). Apparently, some of the concrete adhered to the casing when it was pulled
out since 26 of the caissons were found defective, including one with a 4.3 m long inclusion of
mud. Around the clock repair work lasted four months at a total cost of 11 million dollars (82
million is present value) before work was resumed.

1.32 Tel Aviv 1996
A high-rise residential tower with three basement levels was built on a lot of 6,000 m2. To protect
neighboring streets and properties, a diaphragm wall was constructed around the site. The wall
was excavated with the use of polymer slurry that, unknown to the contractor, was of doubtful
origin. To save money, no integrity testing was specified for the wall. The true picture (Figure 2)
became clear only when the wall was completed and the contractor began to excavate the site.
Extensive repairs were necessary, at a cost of several million dollar.
Fig. 2: A flawed diaphragm wall in Tel Aviv.
1.33 Hong Kong 1999

The Hong Kong Housing Department (HD) hired the lowest bidder, Zen Pacific Ltd, to construct the
pile foundations for five residential buildings, 41 stories high, in the Yuen Chau Kok site. The large
diameter bored piles were supposed to penetrate through unstable layers of fill, marine deposits and
decomposed granite, and be belled out on solid granite rock (ICAC 2000). To support the boreholes
against caving, HD instructed the contractor to use continuous steel casing. Zen Pacific
subcontracted the work to Hui Hon Ltd. that had neither suitable drill rigs nor sufficient casing
material. Hui Hon soon confronted massive borehole collapses and decided, without the owner's
approval, to replace the casings with Super Mud. This however had no effect and as a result,
numerous piles had to be abandoned before reaching bedrock. At this stage, Hui Hon, on the brink
of bankruptcy, devised an extensive cover-up program: Working at night when HD staff were absent,
falsifying the site records, diverting the excess concrete amounts to other projects, blocking the
access tubes installed for cross hole testing and replacing the test with a different test that produced
no useful information. In addition, Hui Hon doctored the tape measure used to check the depths and
replaced defective cores with good ones taken from other piles. All this failed to save the buildings:
by the time the authorities became aware of the situation the two first buildings, already more than
33 stories tall, showed excessive settlements and had to be demolished. The cost? 650 million HK
dollars, two Hui Hon directors sentenced to 12 years in jail and the site agent to five.
The serious consequences of such events served as an incentive to the development of pile
integrity testing.

1.4. The development of pile integrity testing
Integrity testing aims at detecting zones of reduced cross section (necking) and/or inferior material
properties. At an early stage, this was done through direct methods, such as excavation or core
drilling. Excavation enables thorough visual inspection of the outer surface of the pile, but is
generally limited to the upper few meters. Core drilling, on the other hand, provides full
information about concrete quality to large depths, albeit for a small fraction of the pile's cross
section. The resulting core hole, however, enables further studies such as caliper and ultrasonic
testing or video photography.
The sixties and seventies of the 20th century saw the development of various indirect (imaging)
test methods, mostly using analog electronics with recorders or oscilloscopes to plot the data. Once
fast microprocessors became available in the late seventies, purpose-built computers and digital
signal processing were quickly incorporated into testing apparati. The next logical step came
during the nineties, when ruggedized computers became an affordable commodity and were turned
into virtual instruments just by hooking them with suitable transducers. Figure 3 is a chronological
representation of these developments.
Fig. 3: History of integrity testing techniques.
2. PROPAGATION OF ACOUSTIC WAVES IN PILES

After Smith (1960) introduced the one-dimensional wave equation to the analysis of pile driving,
it was soon adopted by most integrity testing methods. A basic familiarity with the principles
governing acoustic waves is therefore necessary to understand how they function. First, we have
to distinguish between the one-dimensional (long wave) case and the three-dimensional (short
wave) case.
2.1. One-Dimensional Wave Propagation

The mathematics involved are rather rudimentary, provided we first make a few reasonable
assumptions:

• The pile is prismatic rod with a constant cross-section A, elastic with Young’s Modulus, E,
and homogeneous with mass density, ρ.
• The wavelength is equal to or larger than the pile diameter.
• Cross sections remain plane, parallel, and uniformly stressed.
• Lateral inertia effects are negligible.
When we hit the pile head with a hammer, we create a compressive wave that travels
downwards along the pile. Βy combining Newton's second law with Hooke's law (Vincke and van
Nieuwenburg 1987) we get the one-dimensional wave equation:
𝜕𝜕2 𝑢𝑢 𝜕𝜕 2 𝑢𝑢
= 𝑐𝑐 2 (1)
𝜕𝜕𝑡𝑡 2 𝜕𝜕𝑥𝑥 2
Where u is the displacement and 𝑐𝑐 = √(𝐸𝐸/𝜌𝜌) is the wave speed in the rod. In concrete piles, for
instance, the harder the concrete the faster the waves. The general relation between concrete
compressive strength fc and wave speed is given by Amir (1988):
𝑐𝑐 = 𝐾𝐾𝑓𝑓𝑐𝑐 1/6 (2)
Clearly, c is determined by the physical properties of the pile and does not depend on the
strength of the blow. Actually, it is exactly the same for a large pile driver and a small plastic
hammer. For concrete grades used in piling it will typically vary between 3,600 and 4,400 m/s.
Particle velocity v, on the other hand, is a function of the stress, σ, applied to the head. The ratio
between the two, v/c, is equal to σ/E. Therefore, the force P in a given section is given by:
P = σA = (EA/c)v = Zv (3)
The factor Z is defined as the acoustic impedance.

When a compressive wave reaches a flaw (reduced impedance, e.g., a "neck" in the pile), it
separates into two parts: One continues to travel down as a compressive wave, while the other is
reflected upwards as a tensile wave. A particular instance is that of a stress-free toe (approximated
by a toe located in soft soil), where the wave is fully reflected upwards as a tensile wave until it
reaches the pile head and "pulls" the head down. In the opposite case, when a compressive wave
reaches a bulb (increased impedance) part of it continues downwards and the other part is reflected
back. In this case, however, both waves are compressive.
On its way along the pile the wave energy is reduced by skin friction. In homogeneous soil,
the total attenuation of the toe reflection is given by the following equation (Paquet 1992):
4𝐿𝐿 𝜌𝜌𝑠𝑠 𝑣𝑣𝑠𝑠

𝐴𝐴(𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁) = (4)
𝐷𝐷 𝜌𝜌𝑐𝑐 𝑐𝑐
Where L is the pile length, D is the pile diameter, 𝜌𝜌𝑠𝑠 and 𝜌𝜌𝑐𝑐 respectively are the soil and pile
densities, vs is the shear wave speed in the soil and c the wave speed in the pile. For a pile with a
typical slenderness ratio L/D of 25 in a soil with a shear wave speed of 250 m/s, the total attenuation
A is 4.7 Neper or e4.7 = 110. For modern digital instruments, it is a simple matter to amplify the
reflections by this amount, but at the same time, any noise present will also be amplified.

2.2. Three-Dimensional Wave Propagation
While in a rod we encounter mainly longitudinal waves (compressive or tensile), in an infinite
space, a dynamic impact may create several wave types that radiate from the point of application,
mainly:
• P-Waves – longitudinal wave in which particle motion coincides with the direction of
propagation
• S-Waves – transverse wave in which particle motion is perpendicular to the direction of
propagation.
• Rayleigh waves – P-Waves that advance close to a free surface in which the particles move
is both longitudinal and transverse directions.
For good-quality concrete, P-waves move at a speed of 4,470 m/s. S-Waves and Rayleigh
Waves are typically slower by 42% and 48%, respectively. In comparison, wave speed in a rod of
the same concrete material is 4,080 m/s.
3. INTEGRITY TESTING METHODS

Integrity testing methods are either non-intrusive or intrusive. The former need minimal
preparation of the pile head while the latter require special access ducts to be installed in the pile
during construction.
3.1. Non-intrusive methods

Non-intrusive methods are invariably based on the theory of one-dimensional wave propagation.
In the following, the Impact Echo Method will be described in some detail since all other non-
intrusive methods are actually derived from it.
3.1.1. Impact Echo Method

This method, also called Sonic, or Low Strain Impact was the first to be implemented in practice
(Beylich 1963) and today is the most widespread method in the world for testing the integrity of
all pile types (ASTM 2016a). In this test method, the pile head is hit with a small plastic hammer
that sends a compressive acoustic wave down the pile. A suitable transducer (usually
accelerometer) that is pressed against the pile head is triggered by the blow and monitors the
reflected waves (Figure 4). The accelerometer output is digitized and then integrated to provide
the graphic time history of the pile head particle velocity. The system then shifts and rotates this
graph to fit the horizontal axis and transforms the horizontal axis from time units t to length units
L by the equation L = c.t/2 . The user can then apply exponential amplification to the graph (to
compensate for skin friction) and pass it through a suitable digital filter. The resulting reflectogram
(Figure 5) can be instantly inspected to provide information about the length of the pile and its
continuity. For a reflectogram with anomalies, an iterative signal-matching technique can plot a
probable profile of the pile.

Fig. 4: Impact echo testing.
The Impact Echo test is fast and inexpensive, with typically less than one minute needed to
test a given pile (Amir & Amir 2008). The main drawbacks of the Impact Echo method are:
• The wavelength produced by the plastic hammer is in the order of 3-4 meters. Therefore,
the information it provides regarding the part of the pile close to the head is limited.
• The lengths reported are a function of the assumed wave speed that is usually unknown.
• It is influenced by skin friction, thus its effective penetration depth varies between 10
diameters in very hard soils and 60 diameters in very soft soils.
• The pile head must be accessible.
Fig. 5: A typical reflectogram.
The testing methods described in the following sections have tried to address some of these
shortcomings.
3.1.2. Frequency response

The Steady-State Frequency Response test was developed in France by Paquet (1968), specifically
for handling relatively shallow discontinuities. Paquet placed an electrodynamic shaker on top of
the pile, gradually increasing the excitation frequency while keeping the force constant. For each
frequency, the measure pile head velocity v and the applied force F were recorded. Paquet plotted
the ratio v/F, defined as the Mobility M, versus the frequency. The difference, ∆f, between
successive peaks (resonant frequencies) indicated the depth, z, of the uppermost discontinuity:
z = c/(2∆f) (6)
In addition, the inverse of the slope of the initial part of the mobility plot is equal to the
apparent low strain rigidity of the pile head. This method did not gain popularity because of the
bulky vibrator and the elaborate preparations it required. This situation has changed in the 1970's

when compact force transducers and Fast Fourier Transform became available. The Impact
Frequency Response method (Higgs and Robertson 1979) replaced the vibrator with an
instrumented handheld hammer incorporating a force transducer. Both force and velocity records
were transformed into the frequency domain and the mobility presented against the frequency.
In spite of its sophistication, the frequency response may be beneficial only in testing the upper
part of the pile; it is more difficult to interpret than the impact echo method and less capable of
dealing with slender piles in hard soils. Paquet (1992) realized that plotting the pile geometry is
viable only if the head excitation data is combined with the surrounding soil data, namely shear
wave speeds in the various soil layers. The methodology he developed, the Impedance Log
method, showed some success in controlled tests.
The PileInspect project was launched in 2013 by the European Union in order to improve and
standardize the procedures for pile integrity testing and evaluation. The consortium formed
employs no less than ten partners: Universities, research institutes, professional organizations and
private laboratories. The PileInspect hardware comprises a controllable vibrator, capable of
exciting the pile in the axial direction, an accelerometer mounted on the vibrator, additional
accelerometers mounted on the pile head and a control and processing unit. The whole setup is
almost identical to that produced by Paquet (1968), albeit with improved electronics. The aim this
time is to employ sophisticated signal processing techniques to power an autonomous system that
will decide if the pile is “damaged" or "undamaged” pile and estimate the damage severity. At the
time this paper was written there were still no positive results reported from this project.
3.1.3. Multiple Sensors

The main obstacle to accurate determination of the pile length is the imperfect knowledge of the
wave speed. (Paquet 1968) suggested embedding a second sensor at the bottom of the pile during
concreting. Although this approach provides the mean wave speed in the pile, it is impractical for
routine testing. Johnson and Rausche (1996) analyzed a model pile with two accelerometers
located 3.0 and 4.6 m, respectively, below pile head. They claim that the wave speed can be
calculated by dividing the sensor spacing by the travel time between them. This, however, can be
in error since the wave speed in the pile is not necessary uniform (Amir et al. 2014).
Niederleithinger et al. (2015) experimented with a pile equipped with five sensors attached to the
pile sidewall exposed by excavation. Evidently, this setup is unsuitable for standard quality
assurance of piles.
3.1.4. Parallel Seismic test

The Parallel Seismic test (Hurtado 1979), that can replace the impact echo method where there is
no access to the pile head, is based on wave transmission rather than on wave reflection. Its primary
purpose is to establish the depth of foundations supporting existing structures where no records
exist. In preparation to the test, a water-filled vertical access duct is installed in the ground, at a
typical distance of 0.5-1.5 m from the foundation element and to a depth exceeding the assumed
depth of the tested element by at least five meters. Where the ground is soft and water table
shallow, the duct can be replaced by inserting a suitable piezo cone into the ground. Above the
ground water table, the access duct must be grouted in the hole to assure proper wave transmission.
The superstructure above the tested element is hit repetitively with a hammer equipped with
an impact switch while a hydrophone is continuously lowered into the duct. The recorded pulses
are plotted vs. depth and, if the test is performed properly, a break in the first arrival plot denotes

the level of the pile tip. If the access duct is indeed parallel to the tested pile, the slope of the upper
branch is equal to the wave speed in the pile and indicates the pile material.
The Parallel Seismic test has proved its effectiveness in determining the pile length. However,
attempts to use it to detect flaws in piles (Niederleithinger 2012, DeGroot 2014) were inconclusive.
3.2. Intrusive methods

The main weakness of all non-intrusive methods lies in the fact that the input (hammer blow) is
applied to the head, while potential flaws may be situated many meters below. The intrusive
methods described below, while more expensive, offer improved flaw detectability by embedding
access tubes or sensors in the pile during the concreting stage.
3.2.1. Ultrasonic testing

The Ultrasonic Cross Hole method (Levy 1970) was developed in France in the late 1960's. In this
method, several equally spaced access ducts, typically 50 mm in diameter, are attached to the
inside of the reinforcement cage. The ducts, either steel or plastic, are filled with water to facilitate
wave propagation. Once the concrete has hardened, two ultrasonic transducers, an emitter and a
receiver, are in turn lowered to the bottom of each pair of ducts (profile). As the transducers are
raised in unison, the emitter sends short duration pulses at predetermined intervals. The First
Arrival Time (FAT) and Relative Energy (RE) of all pulses intercepted by the receiver are recorded
and plotted on a computer screen. An anomaly in either plot can point to a zone of inferior material.
This test has been widely accepted and standardized (ASTM 2016b) and is most probably the
leading method in the world for testing large-diameter piles.
The test as described above provides only one-dimensional information: the respective levels
of the bottom and top of a flaw. The tomographic option (Paquet and Briard 1976) has been
developed for establishing also the lateral dimensions of a flaw. Wherever a flaw is suspected, the
zone is investigated with one transducer stationary at a time and the other moving. The oblique
pulses are combined with the horizontal ones to draw the shape of the flaw. When this procedure
is repeated for all profiles the results can further be analyzed to produce a full three-dimensional
representation of the pile, including zones with different wave speeds (Figure 6). On the downside,
the method provides little, if any, information about the pile outside the area bounded by the access
tubes.
Another variation on the ultrasonic method is the single-hole test that is especially useful for
small-diameter piles where there is no room for multiple access ducts. In this case, the duct must
be of plastic material (Amir 2002) and both transducers are lowered into it, typically 600 mm
apart. This method can also be used in holes produced by core drilling to extend the investigated
range (Baker and Khan 1971).

a b c
Fig. 6: Results of three-dimensional ultrasonic testing: 3D interactive model
(a), vertical cross-section (b) and horizontal cross-section (c).
3.2.2. Gamma-gamma (radioactive) logging

In the radioactive method (Preiss 1971) the cylindrical probe consists of a weak radioactive source
(usually Cesium 137) and a photon counter, separated by a lead shield. The source emits gamma
radiation in all directions, and the photons are partly absorbed by the surrounding concrete and
partly backscattered and recorded by the counter. A high photon count means low concrete density
and vice versa. Since concrete density changes little with time, the radioactive method can be
applied soon after casting. With suitable calibration of the instrument, the photon count readings
can be readily transformed to density. On the downside, the readings are strongly dependent on
the proximity of rebars and the typical range of the probe is less than 100 mm. the method is
presently rarely used, mainly because of regulatory restrictions on the handling of radioactive
material.
3.2.3. Thermal Logging

Thermal Logging (Mullins and Kranc 2004) is based on the phenomenon that concrete hardening
is exothermic. Shortly after casting, the temperature of the fresh concrete begins to rise, the rate
mainly depending on the amount and composition of the cement, the W/C ratio and the aggregates
used. 24 to 48 hrs. after casting, maximum temperatures may exceed 700C, but at the same time
the outer surface of the pile (air or soil) starts to cool down according to Newton's Law of Cooling.
A temperature gradient develops and the heat flows out. The access ducts for this test must be dry,
and the temperatures measured by an infrared thermometer at regular intervals. If the access ducts
are equidistant from the center, all the temperature readings are supposed to be equal. A lower

temperature measured in any access duct may mean reduced concrete cover near that duct or an
adjacent inclusion of foreign material that does not produce heat. The main advantage of the
method is the ability to test the pile shortly after casting and to provide indication about the amount
of concrete cover. At the same time, analysis of thermal logging results is still an inverse problem
in which mass distribution is inferred from discrete temperature readings close to the boundary.
Some more shortcomings are:
• The method is sensitive to inhomogeneity in the surrounding soil, such as saturated sand
or ground water flow.
• The method is sensitive to variations in the amount of retarder used, especially in large
piles cast by several truck mixers.
• The method may totally miss a discontinuity of small vertical dimension.
• The test must be performed once the pile is hot enough, and before it had time to cool
down too much. Once this time windows was missed for any reason, there is no second
chance. This limitation can be overcome by using embedded strings of thermistors
connected to an automatic data logger, but this option is obviously more costly.
3.2.4. Optical means

Trying to actually look into six experimental piles led Sarhan et al. (2002) to construct them with
transparent polycarbonate tubes with inner diameters of 15.9 mm, enough for lowering a compact
video camera. The results, however, were not reported.
Habel and Krebber (2011) inserted fiber-optic sensors in two model piles, and managed to
monitor the strain paths in both models under low stress pulses. The flaw in one of the piles showed
very clearly.
3.2.5. Pile Verticality

If we define pile integrity as meeting the project requirements, it certainly includes pile verticality.
Practically all piling specifications limit the allowable deviation from the vertical, as excessive
deviation may lead to serious consequences: overloading of foundation piles, loss of water-
tightness in piled retaining walls, loss of parking space in deep basements etc.
Conscientious contractors check the verticality of the drill mast before they start drilling and
repeat this check during the operation. This, however, may not be enough since in deep holes, the
Kelly bar becomes flexible and the holes may deviate from the vertical. Several systems on the
market designed to measure the profile of open holes using ultrasonic waves can also check their
verticality, but their use is limited to slurry-filled holes. Systems based on inclinometers combined
with gyros (Amir and Amir, 2012) can measure the deviations of open holes using the drilling
bucket as a centralizer. With suitable adaptors, they can measure the deviation of finished piles
and diaphragm walls inside 50 mm access ducts.
4. PRESENT SITUATION (CONCLUSIONS)

After more than fifty years of evolution, integrity testing of deep foundations is now universally
accepted. At the same time, the industry is still searching for the Holy Grail: a practical, cost-
effective testing method that is able to accurately plot both internal and external geometry of the
pile and determine its material characteristics. Unfortunately, pile integrity testing is not rising to

the challenge, as evidenced by the statistics of nine Stress-Wave conferences (Figure 7). These
events – arguably the main forum on pile testing - show a consistent increase in the number of
papers dealing with integrity testing until 1996, with a marked output slowdown afterwards –
clearly a symptom of a stagnant discipline.
The analysis of all present non-intrusive methods is still a classic inverse problem, with no
unique solution. To achieve a probable solution, one must start with some reasonable assumptions
and check them with a suitable forward model. Given a reflectogram obtained by the impact echo
method, for example, the four basic assumptions we commonly make are:
• The tested element is a pile

• The one-dimensional wave equation is valid
• The wave speed in the pile is X m/s
• The skin friction distribution is known (or neglected).
Fig. 7: Numer of papers presented to stress-wave conferences.
The first assumption is trivial, but the second one is not. While for driven steel piles the one-
dimensional wave equation may be applicable, for the purpose of integrity testing, all the
assumptions on which it is based are more or less wrong. A pile with flaws, for instance, is neither
prismatic nor homogeneous, and the soil profile for a specific pile is known only approximately.
Under the circumstances, the best we can sometimes offer is a reasonable estimate of the pile
length. Evidently, the wave equation approach does not deliver and the most sophisticated
analyses will produce little more than an axisymmetric pile profile with no direct bearing on the
structural properties of the pile (Figure 8).

Fig. 8: Assumed pile profile from Impedance Log testing.
By definition, intrusive methods are much more informative than non-intrusive ones but they
too have their deficiencies. The engineer specifying the test has therefore to make a compromise
based on type of the piles, equipment availability and budget. In case of uncertainty, the prudent
engineer will muster some redundancy: hire another laboratory to repeat the test (a practice adopted
in Hong Kong after the 1999 fraud) or test with other methods. The cross-hole ultrasonic test, for
example, can be supplemented by single-hole testing to investigate the external part of the pile.
Fig. 9 illustrates the case of a pile with a 15% flaw at 4 m depth tested by impact echo with no
anomaly and by single hole ultrasonic clearly showing increased FAT and attenuation. Of course,
redundancy may produce conflicting results, where sound engineering judgement is unavoidable.
Fig. 9: A pile with a 15% flaw at 4 m – impact echo (left) and single hole ultrasonic (right).
5. FUTURE AGENDA (RECOMMENDATIONS)

Forecasting the future is inherently risky. However, even with all the advancement in piling
technology we may safely assume that flaws in piles will not disappear, ensuring continuing
demand for advanced integrity testing. Research and development will have to adopt revolutionary
thinking to meet this demand. The main effort should concentrate on integrating all the existing
testing and analysis methods into a comprehensive system with a common interface.
We shall start with integrating the intrusive and the non-intrusive methods by embedding
optical fibers in the piles during casting. This means will add a wealth of internal data (such as
temperature, stress and strain distribution) to the results of external mechanical excitation. At the

analysis stage the exterior pile profile will be obtained from methods such as CFA monitoring or
bored pile calipering. This will be augmented with interior mapping by ultrasonic tomography and
optical fiber readings to serve as the first iteration of pile geometry in a standard CAD format. If
the pile supports a superstructure such as pile cap it can be added to the model, together with data
from the closest borehole log. The software will discretize the system into elements and apply the
input load at the appropriate point. The calculated displacement time-history will be compared to
the measured one at the same point and the pile geometry adjusted to obtain the next iteration. A
suitable evolutionary algorithm will serve to find the pile geometry that will provide the best fit.
A pre-requisite for the success of the integrated system is to revamp our analytical approach.
Although intrinsically inapplicable to integrity testing, the one-dimensional wave equation (Smith
1960) has always been the backbone of the industry since it managed with the modest computing
resources available in the 1960's. However, assuming that Moore's Law is valid, computer power
has since increased more than1010 fold so it is not too early to divorce the time-honored wave
equation and move over to the Finite Element Method (FEM). Following is a short list of the
advantages of FEM:
• It can model piles with any shape and physical properties (Fig. 10)
• It can accurately model the surrounding soil profile, even if irregular, with proper
geotechnical parameters
• It can model the superstructure in cases where there is no access to the pile head
• The input force can be static or dynamic, transient or steady state, axial or lateral, concentrated
or distributed.
• The output displacements can be monitored at any convenient location.
By a rough estimate, this project may take at least ten years and several millions of Dollars
to complete. Initially the system could be developed in two dimensions (axisymmetry) and later
in full three dimensions. Once complete, the system will be capable of positively detecting at least
90% of important flaws. This will certainly lead piling technology towards better, safer and more
economical foundations.
Fig. 10: Piles with bulges (left) and finite element simulation (right).

6. REFERENCES
Amir, J.M., 1983. Piling in rock - construction aspects. Proc. 6th Asian Conf. SMFE, Haifa,
Vol. 1, pp. 231-234.
Amir, J.M., 1988. Wave velocity in young concrete, Proc. 3rd Conf. Application of Stress-Wave
to Piles, Ottawa, pp. 911-912.
Amir, J. M., 2002. Single-Tube Ultrasonic Testing of Pile Integrity, ASCE Deep Foundation
Congress, Orlando, Vol. 1, pp. 836-850.
Amir, E.I. and Amir, J.M., 2008. Statistical Analysis of a Large Number of PEM Tests on Piles,
Proc. 3rd Conf. Application of Stress-Wave to Piling, Lisbon, pp. 671-675.
Amir, J.M. and Amir, E.I, 2012. Testing of Bored Pile Inclination. Proc. 9th Intl. Testing and
Design Methods Deep Foundations, Kanazawa, pp. 233-236.
ASTM, 2016. Standard test method for low strain impact integrity testing of deep foundations
D5882-16, W. Conshohocken, Vol. 4.09.
Baker, C.N. and Khan, F., 1971. Caisson Construction Problems and Correction in Chicago,
ASCE Journal for Soil Mechanics and Foundation Engineering, 97(2) 417-440.
Beylich, M., 1963. Types des Fondations Classiques - Controle des Pieux. Compte Rendu des
Journees des Fondations. LCPC.
De Groot, P.H., 2014. Evaluation of the Parallel Seismic detection of defects in pile foundations,
M.Sc. Thesis, Delft Technical University, 192 p.
Fleming, W.G.K., Weltman, A.J., Randolph, M.F. and Elson, W.K., (1992). Piling Engineering
(2nd ed.), Blackie, Glasgow, pp. 251-281.
Habel, W.R. and Krebber, K., 2011. Fiber-Optic Sensor Applications in Civil and Geotechnical
Engineering, Photonic Sensors, Springer, 1(3) 268-280.
Hobbs, N.B., 1957. Unusual Necking of Cast-in-situ Concrete Piles, Proc. 4th ICSMFE, London,
Vol.2, pp. 40-42.
Hurtado, J., 1979. Mesure de la profondeur des fondations par micro sismique transparence. Rev.
Francaise de Geotechnique, No. 6, pp. 65-69.
ICAC Hong Kong, 2000. Yuen Chau Kok Development Project.
http://www.icac.org.hk/new_icac/eng/cases/piling/p03a.html.
Johnson, M. and Rausche, F., 1996. Low strain testing of piles utilizing two acceleration signals,
Proc 5th Intl. Conf on Application of Stress Wave Theory to Piling, Orlando, pp. 859-869.
Levy, J.F., 1970. Sonic pulse method of testing cast-in-site concrete piles. Ground Engineering,
London, 3(3), pp. 17-19.
Milne, G., Cohen, N. and Cotton, J., 2010. London's Top Secret, London Archaeologist,
12(11) 287-289.
Mullins, A.G. and Kranc, S.C., 2004. Method of testing integrity of concrete shafts, US Patent
6,783,273 B1.
Niederleithinger, E., Ertel, J.P. and Grohman, M., 2015. Low-Strain-Pfahlintegritatsprufung
reloaded Geht nicht doch ein bisschen mehr?, "Pfahlsymposium” 2015, Braunschweig.
Paquet, J. 1968. Etude Vibratoire des Pieus en Beton, Reponse Harmonique et Impulsionelle.
Application au Controle, Annales ITBTP, Vol. 245, pp. 788-803.
Paquet, J., 1992. Pile integrity testing - the CEBTP reflectogram. Piling - European practice and
worldwide trends (M.J. Sands, ed.), ICE, London, pp. 206-216.
Paquet, J. and Briard, M., 1976. Controle non destructif des pieux en beton, Annales de
L'institute Technique du Batiment 337, Paris, pp. 50-79.

Preiss, K., 1971. Checking of cast in place concrete piles by nuclear radiation methods, British J.
NDT Vol 13, pp. 70-76.
Sarhan, H.A., O'Neill, M.W. and Hassan, K.M., 2002. Flexural Performance of Drilled Shafts
with Minor Flaws in Stiff Clay, ASCE Journal Geotech. and Geoenviro. Engineering,
128(12) 974-985.
Szechy, C., 1961. Foundation Failures. Concrete Publications, London, p. 107.
Smith, E.A.L.,1960. Pile-Driving Analysis by the Wave Equation. ASCE Journal Engineering
Mechanics Div. 86(EM 4) pp. 35-61.
Vincke, J. and van Nieuwenburg, D., 1987. Theorie van de Dynamische Proeven (in Flemish),
Studiedagen Nov.87, Belgian Group ISSMFE, pp. 2-49.
White, R.E., 1962. Caissons and cofferdams. (Ch. 10 in Foundation Engineering, Leonards, G.A.
ed.), Mc. Graw-Hill, New York, pp. 908 ff.

Advances in the Use of Drilled Foundations
Brown, D.(1)
(1)
President, Dan Brown and Associates, Sequatchie, TN, USA <dbrown@dba.world>
ABSTRACT. The last few decades have seen significant advances in construction practices for
drilled foundations, and these developments have greatly affected foundation engineering
regarding the selection and use of deep foundations. Advances in construction have come in
response to greater challenges of 21st century infrastructure development, leading to technological
advances in equipment, materials, and testing. This paper describes some of the more significant
developments reflected in the author’s experiences in North America and is intended to provide
foundation engineers with increased awareness of modern capabilities and potential applications.
1. INTRODUCTION
The modern-day world presents an increasingly complex situation for foundation engineering. The
many challenges include congestion in the urban environment, with replacement or enhancement
of existing infrastructure while maintaining service. New structures are planned for increasingly
difficult geologic settings, since the “easy sites” are already occupied. Modern foundation designs
are subject to increased complexity and demands from seismic and other extreme event loadings.
Environmental constraints present further challenges to construction, as the public demands that
we minimize noise, vibration, and impacts to the natural environment. Reduced public tolerance
for pile driving activity has encouraged greater use of drilled foundations. Project delivery has
modified contractual relationships and financing aspects of projects, with the inevitable pressure
on foundation engineers and specialty contractors to take on risks and deliver faster.
Industry has responded to the market pressures associated with these challenges with
significant advances in construction technology, and thus foundation engineers have an incredible
variety of drilled foundation options at their disposal. This situation is both a blessing and a curse;
it presents opportunities for innovation and creativity, but also presents challenges in identifying
the optimal solution for a given situation. Constructors are more often engaged in the foundation
type selection and must therefore understand design issues and similarly designers are required to
have greater understanding of construction techniques and capabilities.
This paper describes some of the more significant advances in the construction of drilled
foundations. It is not intended to be an exhaustive “state-of-the-practice” paper, but rather a
reflection of the more notable advances. Brief descriptions of selected case histories from the
author’s experience are provided to illustrate the implementation of modern drilled foundation
technology. The author’s experience derives primarily from his practice in North America, and it
is recognized that the worldwide implementation of these technologies varies.
Although the definition of different technologies varies around the world, for purposes of this
discussion drilled foundations will be generally described using four categories in accordance with
the drilling techniques used. Micropiles are small diameter drilled piles installed using lightweight
drilling equipment in which the cuttings are generally flushed to the surface with a circulating
fluid. Continuous Flight Auger Piles (CFA piles) are installed by advancing the continuous auger
into the soil and then placing the concrete through the hollow stem as the augers are withdrawn.
Reinforcement is subsequently placed into the fluid concrete. Drilled displacement piles are

constructed using a technique similar to CFA piles, but using tooling and more powerful equipment
such that the drill tool is advanced while displacing the soil to form the hole rather than extracting
the soil. Drilled Shafts (commonly called “bored piles” in much of the world) are most often 1 m
or more in diameter and constructed by excavating and stabilizing a hole into the bearing formation
(often with drilling fluid and/or steel casing) followed by placement of reinforcement and then
concrete.
It is worth noting that the distinctions between drilled foundation technologies cited above are
becoming less distinct as drilling technology advances.
2. MICROPILES
Micropiles are most often 300 mm or less in diameter and often selected for use because of the
advantages provided by the lightweight and maneuverable equipment available to install these
piles and overcome site access difficulties. The use of these foundations has grown dramatically
during this period, and the popularity of this type of drilled foundation is largely the result of the
versatility of the equipment (as shown in Figure 1) used to install them. Micropiles are used for
applications including underpinning and seismic retrofitting and in locations and ground
conditions where more conventional deep foundations would be difficult or impossible to
construct. The use of micropiles has evolved from small, lightly loaded “root piles” into the current
practice where micropiles are designed to support substantial individual pile axial forces in the
same way as other types of piles.
a) Foothills Bridge, East Tennessee b) World Trade Center, New York City
Fig. 1. Micropile Drill Rigs in Restricted Access Locations.
Besides the ability to overcome difficult site access, micropiles can be drilled to provide good
foundation support into materials which are impossible to penetrate with driven piling or which
represent extremely difficult drilling conditions with larger diameter drilled foundations.
Examples include piles through boulders, fills including rubble or other hard debris, and karstic
formations in hard limestone. The micropiles may be advanced through porous layers with casing

until a substantial thickness of sound material is penetrated, and then the casing partially
withdrawn to form a permanently casing through the pervious strata and allow the pile to be
grouted into the sound layer.
The distinguishing feature from a design perspective is that the pile itself is typically designed
as a steel member such as a single bar which is bonded to the bearing stratum with a cement grout
below an unbonded upper section comprised of a grout-filled steel tube, as illustrated in Figures 2
and 3. The micropile grout is typically a mixture of water and Portland cement that may be simply
tremie-placed under gravity only, or may be pressure grouted during or after installation.
Micropiles are most effectively used where the bearing materials allow effective utilization of
the high strength of the steel. Granular soils or rock often provide suitable bearing formations and
micropiles are often used in rock or in highly variable conditions where difficult drilling may be
encountered.
Fig. 2. Typical micropile details from a building project.

Fig. 3. Completed micropile and connection to a footing for a pedestrian bridge.
In a typical application, the structural loads dictate the size of the steel element and then the
embedded length is determined to provide the geotechnical resistance necessary for the transfer of
load from the steel through the grout to the soil or rock. The transfer of axial load is typically
accomplished through side resistance in the portion of the pile below the casing (the bond zone),
with no reliance upon side resistance in the permanently cased zone or in end bearing.
The unit side resistance in the bond zone is not only affected by the type of soil or rock, but
may be strongly affected by the type of construction practice used for drilling and for grouting.
Because of the strong influence of construction technique, the average nominal unit grout-to-
ground bond strength is usually estimated empirically and verified through site-specific load
testing, with the final micropile geotechnical design performed by the specialty contractor. By
working in this way with either performance-based specifications or a design-build type of
contract, the contractor has the ability and responsibility to select the most appropriate and cost-
effective drilling and grouting techniques. The constructor thus typically has some design
responsibility and incentive to improve geotechnical performance, and static load tests are
routinely employed to provide verification of axial resistance.
A major factor in the broad acceptance and increased use of micropiles within the last 15 years
has been the work of industry groups to develop and promote standards, share knowledge and
expertise, and transform the technology into a more universally accepted foundation option. The
practice in the U.S. has been shaped by the joint micropile committee of the Deep Foundations
Institute and ADSC: The International Association of Foundation Drilling, whose members have
worked with the Federal Highway Administration and the National Highway Institute to produce
design and construction guidelines and training materials. An FHWA reference manual was
published in 2000 (Armour, et al. 2000) and subsequently updated (Sabatini, et al. 2005), which
provides a widely-used reference for the design and construction of micropiles. Micropiles were
only recently incorporated into the AASHTO Bridge Design Specifications in 2007 and the
International Building Code (IBC) in 2006.

2.1 Example Application of Micropile Foundations – Foothills Bridge, East Tennessee
The micropile design for the Foothills Parkway Bridge shown in the photo in Figure 1a, was part
of a design-build project for the U.S. National Park Service. The bridge is constructed on a steep
mountainside location in an environmentally sensitive area, with natural slopes approaching 1:1.
Micropiles were used to support both a temporary work bridge and the piers for the permanent
structure, and were drilled through residual overburden of decomposed rock to bear in the
underlying metasandstone and metaconglomerate as illustrated in Figure 4. Anchors were
incorporated into the foundation to resist passive earth pressures from the overburden soils which
have marginal stability against downslope creep.
Pier
Residual Soil
Cased zone Anchor (to rock)
4.6 m (15 ft)

uncased bond
zone
Sound
Weathered
Rock
Rock
Fig. 4. Micropile Foundation Design for a Typical Pier, Foothills Parkway Bridge No. 2.
Analyses of the pile group foundation for bridge foundation loads from the pier were used to
determine shear, moment, and axial demands on the individual piles and the design completed in
accordance with the AASHTO 2007 LRFD guidelines. The shear and moment demand is resisted
by the grout-filled permanent casing, which is 244 mm (9-5/8 inch) diameter, 12 mm (0.472 inch)
wall thickness, 550 MPa (80 ksi) yield strength, and extends through the overburden soils and
weathered rock zone. The casing was also installed so that no joint was located within 2.4 m (8 ft)
of the top of the pile beneath the footing. The piles include a No. 18 center bar (57 mm, or 2-1/4
inch diameter) with 414 MPa (60 ksi) yield strength.
The maximum factored axial load demand of 1,380 kN (310 kips) is resisted by the 203 mm
(8 in.) diameter uncased portion of the pile which extends 4.6 m (15 ft) into the rock. This socket
is designed for a nominal unit side resistance of 690 kPa (100 psi) and a resistance factor of 0.7.
The axial resistance was confirmed by load tests.
The key factor in utilizing micropiles for the Foothills Bridge was the ability to position a
small rig into place to install piles (shown in Figure 1a) with a minimum impact on the rugged and

environmentally sensitive site. A tubular steel work trestle was installed atop the temporary
micropiles, allowing construction of the permanent foundations and the remainder of the precast
segmental bridge from above ground as shown in Figure 5. This type of solution requires a
collaborative effort from both the designer and the constructor, as is facilitated by the design-build
system for project delivery.
Work Trestle Permanent Pier
Fig. 5. Construction of the Foothills Parkway Bridge using Micropile-supported Trestle.
2.2 Example Application of Micropile Foundations – International Market, Honolulu,

Hawaii
A recent project has been completed with extremely deep micropiles which were used to overcome
challenges of extremely deep soft soil, artesian groundwater, and site access limitations. The
micropile details illustrated previously in Figure 2 were employed to support a new retail facility
in crowded Honolulu. The International Market Place has been an icon in the heart of the tourist
district of Waikiki, with a bustling, open-air hub of small vendors peddling an array of local items.
A project to modernize this area but retain the basic character required foundations suitable for the
challenging ground conditions but also constructed in a way that could preserve the land, starting
with a huge old Banyan tree (Figure 6) that dominated the entrance to the property.
Micropiles extended through underconsolidated lagoon deposits to achieve axial resistance in
underlying dense soils, coralline layers, or volcanic basalt at depths of up to 90m in some areas. A
recent advancement in micropile drilling technology was utilized to ensure that bearing was
achieved by monitoring drilling resistance in real time during construction to provide quality
assurance that the piles had been installed to the appropriate depth and bond length.

Fig. 6. Micropile Drilling under the Banyan Tree (photo courtesy of Hayward-Baker).
2.3 Example Application of Micropile Foundations – U.S. Hwy 53 Bridge, Northern

Minnesota
The advances in drilling equipment and techniques from recent years have allowed the
construction techniques typically used for micropiles to be employed for larger diameters and
depths. The foundations constructed for the U.S. Hwy 53 bridge in Minnesota represent an
extension of micropile construction techniques for piles that are 762 mm (30 inch) diameter and
over 50 m deep.
This bridge, shown in Figure 7, is constructed across a canyon created by mining operations
to extract iron ore known as the Rouchleau Mine. Support of the 52m tall piers within the base of
this canyon required piles to penetrate through 30 to 35m of unconsolidated mine waste fill that
contained granular soil and boulders. The boulders include extremely hard rock with compressive
strengths as high as 350 MPa.
The solution employed was to install fully cased piles with a full-face percussion drill using
reverse circulation to extract cuttings (Figure 8). These piles proved to be very successful in
penetrating the boulder fill and were drilled through about 10 to 15m of weathered rock to provide
high end bearing resistance in the sound rock below. Load tests demonstrated that these piles were
capable to support the full structural capacity of the pile cross section in end bearing.
Based upon the results of the testing program the piles were designed as concrete-filled steel
pipe for a nominal axial resistance of 22 MN per pile (based on a nominal base resistance of
48 MPa) and with only reinforcement at the pile head to facilitate the connection to the footing.
The largest pier (Pier 2) is supported on a 2 x 9 group of piles.

Fig. 7. Construction of the U.S. Hwy 53 Bridge over the Rouchleau Mine Pit.
Crossover
Sub-Seal
Crossover
Sub Port
Shock
Absorber
Hammer
Pilot Bit
Fig. 8. Reverse Circulation Percussion Drilled Piles.

3. CONTINUOUS FLIGHT AUGER (CFA) AND DRILLED DISPLACEMENT PILES
Continuous Flight Auger Piles (CFA piles) are typically 300 to 1,000 mm diameter and most often
selected for use because of the advantages provided by the speed and cost-effectiveness of the
installation method and equipment. Often called “augered cast-in-place (ACIP)” or “augercast”
piles in U.S. practice, the distinguishing feature of the construction of these piles is the fact that
the concrete (sometimes a sand-cement mix) is placed through the hollow center of the continuous
flight auger drill string as the augers are withdrawn and then the reinforcement is placed into the
wet fluid mix after the casting operation is complete. The pile is thus a reinforced concrete
structural element and designed accordingly. CFA piles are usually most cost effective when used
at lengths of 10 to 30 m and constructed entirely in soils, although occasionally these piles are used
in weak rocks. Because of the speed with which the pile can be drilled and completed, it is not
uncommon for a constructor to install several piles within a single hour of work.
Two types of drilling rigs are employed in North American practice; the fixed mast hydraulic-
powered rigs that are typical of European manufacturers and the crane attached drills that have
been used in U.S. practice since the 1970’s. The hydraulic rigs typically have greater torque and
crowd and are better able to control the rate of advance so as to maintain stability through soft or
unstable ground. The crane attached drills can be faster and more cost effective if ground
conditions permit their use, and can be equipped to install piles to much greater depths than
possible with fixed mast systems; some such piles have been installed to depths of 50m or more in
Miami. The photos in Figure 9 below illustrate examples of each type of rig.
Fig. 9. Drill Rigs for CFA Piling.

Besides the advancements in drilling capabilities, the last 20 years have seen significant
advancements in the use of electronic controls for monitoring and guiding the drilling and casting
process. These improvements provide greater quality control and quality assurance in the
construction of CFA piles.
Control of the drilling process is important; the drill is typically rotated at a constant rate and
if the drill is allowed or forced to penetrate too quickly it can "corkscrew" into the soil and become
“hung”, i.e., the torque required to continue advancing exceeds the torque capacity of the rig. For
this reason, the rate of advance must often be restrained to ensure that the soil is cut and loosened,
but not so much that the augers are not kept charged full of soil to provide stability to the hole.
The rate of advance of the drill must allow conveyance of enough material up the flights to allow
for the volume of the drill itself and the bulking action of the soil as it is cut and remolded by the
auger. If the soils have sufficient cohesion and/or arching action to stand vertically without the
lateral support of the soil-filled auger, then the rate of penetration can be restrained to ensure easy
drilling since conveyance of soils up the flights is not a significant issue. However, in loose sands
below the groundwater or soft clays, excessive rotation of the drill without advancement will
convey soils up the auger flights like a screw pump, the lateral stress around the pile is reduced,
and the soil around the hole will side-load the auger resulting in loosening and ground subsidence
around the pile. These effects have been described by Fleming (1995), and the effects on soil
disturbance and pile behavior measured and described by Van Weele (1988) and Mandolini, et al.
(2002) (Figure 10). Observations of ground subsidence around CFA pile construction in sands and
soft soils have been noted on numerous projects, for example as described by Esrig et al. (1994).
a) Side loading the auger due to excessive b) Soil loosening due to excessive rotation
rotation (from Fleming, 1995) measured by CPT (from Van Weele, 1988)
Fig. 10. Effects of Excessive Rotation of CFA Auger.

Control of the casting process is the second key component in the construction of CFA piling;
the control of the withdrawal of the auger during concrete or grout placement must be synchronized
with pumping so that:
a) positive pumping pressure is maintained at the point of discharge at the bottom of the
augers,
b) a structural defect or neck in the pile does not result from pulling the auger string too fast,
and
c) wasteful pumping of excess concrete or grout does not occur, particularly in soft soils
where overconsumption would provide little or no benefit.
Through much of the history of the use of CFA piles, the skill of an experienced drill rig
operator has been recognized as a critical component, because the “feel” of the operator was always
so important to both advance the drill effectively and withdraw the drill during concreting in the
correct way. The use of modern electronic controls, shown in the photos of Figure 11, provides
the operator with direct feedback measurements on the critical parameters and also the ability to
document that the pile has been constructed in accordance with good practices. These days, rig
operators, having grown up playing electronic games, are quite comfortable operating a joystick
and using a graphical electronic display. For the constructor, the monitoring can also provide a
measure of productivity, since some systems provide a minute-by-minute log of the activity of the
drill rig. Equipment maintenance requirements represent another common function that may be
included as a part of the on-board computer system.
a) Controls on a hydraulic system b) Controls on a crane-mounted rig

Fig. 11. Automated Monitoring Systems in Use with CFA Pile Construction.
The most important control parameters include the rate of penetration and sometimes the
applied torque and crowd (down force) on the tools, rotation rate, the concrete or grout pressures,
and the volume of concrete or grout pumped as a function of the elevation of the auger tip and the
theoretical volume required to that point. When these parameters are calibrated to site-specific
loading tests the use of automated monitoring provides a high level of quality control and quality
assurance. The monitored parameters are recorded and documented in a production log that
provides a record of the successful completion of each pile.

Although not yet common in North American practice, there are available capabilities for the
on-board computer to take over the casting process, automatically matching the rate of withdrawal
of the auger to the rate of delivery of grout as measured through an in-line flowmeter while
maintaining a specified delivery pressure in the pump line at the top of the auger string.
3.1 Drilled Displacement Piles

The availability and use of more powerful fixed-mast hydraulic-powered rigs with their greater
torque and crowd has facilitated the use of drilled-displacement piles. The drill tooling for these piles
includes a feature that displaces rather than extracts the soil, as illustrated by a few of the different
types of tools in use in the photos of Figure 12. These tools are characterized by a displacing body
which is typically around 1.5 to 2 m (5 to 7ft) above the tip of the auger, with sometimes occasional
reverse flights at various intervals above the displacing body. The short length of auger below the
displacing body helps advance the tool by screwing into the soil below the displacing body and
pulling it downward. Various types of cutting shoes on the bottom may be employed, depending on
the type of soil to be penetrated. The photo on the left is from the construction of the Georgia
Aquarium in Atlanta and shows a tool extracted from the soil upon completion of casting. The lack
of spoils associated with the construction of this pile points to one of the advantages of this technique,
i.e., spoil removal and the mess associated with CFA piles is avoided.
Fig. 12. Drilled Displacement Tools.
The torque and crowd required to construct a drilled displacement pile is substantial, and the
modern fixed-mast hydraulic drill rigs are typically used for these piles. Because the pile fully
displaces the soil, there are no issues related to over-rotation of the auger and potential loosening
of the soil as described for CFA pile construction. The energy required to install the pile is related
to the resistance of the soil to the displacement, and so the piles are often installed to a depth that
is controlled by the capabilities of the drilling rig. The potential effect of lateral displacement or
heaving on nearby structures may also be a consideration.
With the monitoring equipment capabilities described previously for CFA piles, it stands to
reason that the measure of torque and crowd as a function of penetration might logically be related
to the axial resistance of the completed pile in a manner similar to a CPT sounding. Variations in
stratigraphy are readily detected, i.e., the penetration into a denser stratum is immediately evident
by the measured torque and crowd required to maintain penetration. Although a broad

methodology is not yet in widespread use, work is ongoing to develop site-specific correlations
between “installation effort” (NeSmith and NeSmith 2009) and load test results. Each individual
drilled displacement pile is then installed to conform to a specific criterion based on the
measurements of torque and crowd using the automated monitoring system. These advances offer
improved efficiency as well as quality control and quality assurance.
Another advantage of the drilled displacement pile is the ground improvement that is naturally
accomplished as the displacement tool densifies cohesionless soils and increases the in-situ
stresses in the ground. In this way, the construction of multiple displacement piles in a group
actually enhances the capacity of nearby piles, as demonstrated by Brown and Drew (2000). The
ground improvement associated with the installation of displacement piles has been demonstrated
by CPT soundings before and after installation, reported by Siegel et al. (2007) for a number of
sites composed of sandy soils. Examples of these data are provided on Figure 13.
Heavy line
is post
installation,
light gray
line is pre-
installation
Fig. 13. Effect of Drilled Displacement Pile Installation on Cone Tip Resistance in a Sandy Soil
(from Siegel et al. 2007).
Because of the ground improvement benefits with drilled displacement pile equipment, these
piles are popular for construction of pile raft foundations. The delineation between what is to be
called a “pile” as opposed to a “rigid inclusion” or “column” in terms of ground improvement
technology has become obscured and the terminology used may often reflect the design approach
with respect to building code requirements. When used primarily as a ground improvement
technique, the structure may in fact be designed to bear on spread footings that are not connected

to the installed pile elements (or columns, since they are not really used as piles), and the columns
may even be constructed of lower strength, unreinforced concrete.
3.2 Example Application of Drilled Displacement Piles for Ground Improvement

An example of the use of drilled displacement piles to achieve ground improvement is described
by Siegel and NeSmith (2011) for a site composed of loose silty sand for a hospital in Kentucky.
The technique was used to provide liquefaction mitigation and to increase subgrade stiffness so
that the structure was founded on shallow footings bearing on the composite ground. The project
included load tests on 3 m by 3m (10 ft by 10 ft) test foundations to applied bearing pressures of
335 kPa (7 ksf) to verify the performance. Photos of the footing construction and testing are shown
on Figure 14, along with a plot of the results of the three load tests. The three tests were performed
on an identical column layout in three different areas of the site. The layout under the footing
included six 300 mm diameter by 9 m deep columns and two 400 mm by 15 m deep columns,
arranged on a triangular spacing. Instrumentation on the columns and on the subgrade between
columns suggests that, at the maximum bearing pressure, the load was distributed about 40% to
the columns and about 60% to the soil subgrade beneath the footing.
Average Bearing Pressure, kPa

0 100 200 300 400
0
-10
Displacemetn, mm
-20
-30 Test 1
-40 Test 2
Test 3
-50
-60
Fig. 14. Load Tests of Drilled Displacement Columns Supporting a Spread Foundation.

4. DRILLED SHAFTS
Drilled Shafts (commonly called “bored piles” in most of the world) are most often 1 m or more
in diameter and constructed by excavating and stabilizing a hole into the bearing formation (often
with drilling fluid and/or steel casing) followed by placement of reinforcement and then concrete.
In this way, large diameter foundations are constructed which can transfer forces to deep,
competent bearing strata and provide very large axial and lateral resistance. With the capabilities
of modern equipment to install large drilled shafts and the improved testing capabilities for
verification of structural integrity and geotechnical performance, the use of a single drilled shaft
to support a single column is often used to maximize the foundation capacity in the smallest
possible footprint.
The major advancements in the use of drilled shaft foundations come from improvements in
equipment and construction methods and improvements in testing and verification of performance.
Drilled shafts can now be used with greater diameters and depths than ever before, a trend which
opens opportunities for applications for foundation and geotechnical engineers to employ drilled
shafts in new and creative ways. Methods of construction such as the use of base grouting and the
use of polymer drilling fluids (instead of bentonite) have been shown to provide improvements in
performance of drilled shafts. Testing technology has also evolved to a point of routine use for
verification of structural integrity and measurement of axial and lateral resistance to extremely
large loads. Current practices for construction, design, and testing of drilled shaft foundations are
provided by Brown, et al. (2010).
4.1 Equipment for Larger and Deeper Drilled Shafts

The drilled shaft construction industry has evolved from a relatively small group of subcontractors
to a much broader industry with a wide array of specialized equipment used for construction.
Although it is still largely a craft performed by specialty subcontractors, a larger number of general
contractors are self-performing this work and many subcontractors are concentrating on
specialized types of drilled shaft and other specialty drilled construction techniques. The increased
availability of specialized equipment which is focused on a particular construction technique
contributes to this trend. On large or complex projects, drilled shafts have been employed with
diameters of up to 4 m (13 ft) and depths of up to 80 m (260 ft).
One trend in recent years is a much-increased use of oscillator or rotator equipment to install
full length segmental casing. This type of equipment has been used to construct drilled shafts with
diameter of up to 3.6 m (12 ft), and offers particular advantages in potentially caving ground
conditions. The machines use hydraulic-powered jaws to clamp onto and twist the casing, and also
pull or push the casing in the vertical direction. The oscillator machines twist the casing back and
forth through a range of about 25° whereas the rotator provides the ability to twist the casing
continuously through 360° and effectively use the casing as a full-length coring tool. Photos of
oscillator and rotator machines are provided in Figure 15.
One of the advantages of these machines is that drilled shafts can be installed in caving ground
conditions with improved ability to stabilize the hole during excavation and concrete placement.
The Benetia-Martinez bridge in the San Francisco Bay area is an example of a project with very
challenging ground conditions composed of steeply bedded siltstone and shale with interbedded
layers of soft and hard rock. After great difficulties with open-hole drilling into this formation, the
project was successfully completed using the rotator equipment shown in Figure 15b. The rock
was removed from within the casing using a drop chisel to break the rock and a hammer-grab to
extract it.

The installation of the casing by twisting it into place allows the casing to advance ahead of
the excavation without the vibrations associated with the use of a vibratory hammer. Therefore the
system allows installation of large diameter drilled shaft foundations in close proximity to existing
structures with minimal risk of impact on the existing structure. Oscillator equipment was used to
install 3.6 m (12ft) diameter drilled shafts for the Gilmerton Bridge in Chesapeake, Virginia in
close proximity to two bascule bridges that remained operational during construction. The lack of
vibrations adjacent to the drilled shaft construction was a primary feature in the selection of this
method, and the use of large diameter drilled shafts minimized the size of the foundation footprint
under each individual column for the new structure. Similar very large oscillator-installed drilled
shafts were used on the Doyle Drive approach structures to the Golden Gate Bridge in San
Francisco (Faust 2011).
a) Drill-Mounted Oscillator b) Rotator Attached to Crane

Fig. 15. Oscillator and Rotator Machines.
The presence of a fully cased hole also provides a reduced risk of soil caving during concrete
placement, and therefore improved reliability for construction in applications such as bridge
foundations where flexural demands require the use of large diameter drilled shafts. Where artesian
groundwater conditions are present, the water table inside the casing can be readily maintained at
an elevation well above the ground surface to provide sufficient head within the shaft excavation
to counterbalance the artesian condition.
Katzenbach et al. (2007) reviewed available load test information on drilled shafts constructed
using the oscillator and rotator segmental casing method and report that the results are comparable
and in some cases favorable to other installation techniques. One factor that favors the performance
of drilled shafts constructed using this method is the fact that the teeth that are used on the cutting
shoe at the bottom of the casing tend to produce a roughened surface at the concrete/soil interface
as the casing is extracted. An opportunity to examine the surface of drilled shafts constructed using
this construction method was provided recently at the Huey Long Bridge in New Orleans (Brown
et al. 2010). The 2.8 m (9 ft) diameter drilled shafts were exposed within the sheet pile cofferdam
after placement of the seal slab and during construction of the footing. These foundations were
constructed prior to excavation of the cofferdam, with a corrugated metal pipe used as a temporary

form above the top of the drilled shafts. The photo in Figure 16 shows the herringbone pattern left
at the surface of the drilled shaft concrete due to the action of the teeth on the soil as the casing is
extracted.
Another advantage is the control on verticality provided by the increased stiffness of the
drilling system. While verticality is not often a critical factor for foundations, this aspect is
important for applications in which drilled shafts are used near underground structures or to
construct secant or tangent pile walls. Typical specifications for verticality of drilled shaft
foundations using conventional construction techniques are 1.5% in soil and 2% in rock (Brown,
et al. 2010). However, recent experiences in a test installation for the TransBay Terminal in San
Francisco suggest that oscillator/rotator equipment can maintain a verticality of about 0.35% to
0.5% for foundations as deep as 73 m (240 ft).
Fig. 16. Exposed Texture on the Drilled Shaft Surface, Huey P. Long Bridge.
Reverse-circulation drilling is another technique that has been increasingly used in recent
years to construct drilled shafts to large diameters and depths. This drilling technique provides full
face rotary cutting at the base of the excavation with the drilling fluid used to remove cuttings via
air-lift pumping up through the center of the drill pipe. This closed system avoids the need to cycle
in and out of the hole to remove cuttings from an auger and can also be very effective in excavating
rock.
The photos on Figure 17 illustrate the equipment used with this technique. The system in
Figure 17a is mounted onto a casing that was installed with a rotator, and is working in the space
beneath an existing bridge on I-90 in Connecticut to install 2.8 m (9 ft) diameter drilled shaft
foundations into the bedrock for the replacement bridge structure prior to demolition of the old
one. The drill removes the cuttings by pumping the cuttings and fluid up through the center drill
pipe, through the swivel at the top, and on to a spoil container via the discharge hose in the
foreground. Drilling fluid is simultaneously pumped into the top of the excavation through a return

line. An example of a full-face rotary cutting tool is shown in Figure 17b. This tool was used on
the Walter F. George Dam in Alabama to construct a cutoff wall into limestone. The bottom of the
air-lift pipe is located slightly off-center so that this pipe moves around and suctions the cuttings
across the face as the tool rotates.
a) Restricted Headroom Drilling b) Full Face Drilling Tool

Fig. 17. Reverse Circulation Drilling.
The Wolf Creek Dam project in Kentucky is an example of the advancement of drilled shaft
equipment and technology to overcome challenges in a way that was not possible years ago.
Seepage through the underlying limestone bedrock below has threatened the stability of the earth
dam that retains Lake Cumberland, the largest reservoir east of the Mississippi. A previous cutoff
wall had been constructed into the bedrock in the late 1970’s using the best available technology
at that time, but the seepage problem was not successfully resolved by that effort. Seepage has
found new paths under and around the wall, leading to sinkholes and soft wet areas downstream
as well as high measured pore water pressures in the embankment. The Wolf Creek Dam was in
critical need of remediation to correct the problem.
The key component of the repair to the dam is the construction of a secant pile cutoff wall,
and the construction of this wall utilizes reverse circulation drilling to construct drilled shafts to
very great depths. After lowering the reservoir and completing an initial grouting program, the
cutoff wall is constructed through the dam from a bench on the upstream face, as shown in the
photo of Figure 18.
First, a 1.8 m (6 ft) wide concrete diaphragm wall is constructed through the embankment to
the top of rock at a depth of around 25 to 30 m (80 to 100 ft). The secant pile wall is then
constructed to depths of up to 84 m (275 ft) through the concrete diaphragm wall and the karstic
limestone and into a sound limestone layer. The secant pile excavation is started using
conventional drills with rock augers to open a hole to a depth of around 15 m (50 ft) into the
diaphragm wall, and then completed using reverse circulation drilling as illustrated in the photos
of Figure 16. In order to maintain the alignment on such deep drilled shafts and ensure that the
secant piles overlap to form a water-tight cutoff wall, a pilot hole is first installed using directional
drilling techniques. The reverse circulation drill is equipped with a “stinger” to follow the pilot
hole and maintain the alignment during drilling.

Fig. 18. Work Platform at Wolf Creek Dam.
Fig. 19. Reverse Circulation Drilling for Secant Pile Cutoff Wall.
4.2 Base Grouting

Base grouting to enhance the axial resistance of drilled shaft foundations is a technology that has
been around for decades, but research in Tampa by Mullins et al. (2000) has spawned a renewed
interest in this technology in North American practice. Base grouting is a form of compaction

grouting at the toe of a drilled shaft which compresses and preloads the soil below the toe, increases
the state of stress in the ground, and can significantly increase the axial base resistance of drilled
shafts which are founded in granular soils (Mullins, et al. 2006). There is also benefit from base
grouting in that the grouting mitigates the effect of any loose granular material which might remain
as a result of imperfect cleaning of the base of the drilled shaft excavation. The technique provides
relatively little benefit in rock, cohesive soils, or strongly cemented materials. Another limitation
to the improvements achieved with base grouting is that the available side resistance of the drilled
shaft limits the magnitude of the pressure which can be applied.
The photos in Figure 20 illustrate some aspects of base grouting. Figure 20a shows a typical
base grouting apparatus attached to the base of the reinforcement cage. The photo in Figure 20b
shows a 1 m (3.5 ft) diameter and 8 m (25 ft) long drilled shaft at the Auburn University National
Geotechnical Experimentation Site. This shaft was base grouted and subsequently exhumed to
reveal the effects of base grouting in a very silty and medium dense soil. A relatively large volume
of approximately 0.3 m3 (10 ft3) of grout was injected at the base of this shaft, representing a
volume equal to about 30 cm (1 ft) length of shaft. The grout can be seen to have produced a bulge
at the base and also to have migrated up along the side wall of the drilled shaft over the lower one
to two diameters. The drilled shaft in the background of this photo was not base grouted.
a) Sleeve Port System for Grouting b) Exhumed Base-Grouted Shaft

Fig. 20. Base Grouting for Drilled Shaft Foundations.
An example of the use of this technique on a major project is described by Dapp and Brown
(2010) for the John James Audubon Bridge over the Mississippi River in Louisiana, which utilized
drilled shafts founded in dense alluvial sand. Each of the two pylon foundations for the cable-
stayed bridge included 21 drilled shafts which were 2.3 m (7.5 ft) diameter and approximately 60
m (200 ft) deep. The base grouting was accomplished via a sleeve-port system (tubes-a-manchette)
that utilized the crosshole sonic logging tubes. The eight tubes were connected in pairs across the
base of the drilled shaft to form four separate U-shaped circuits, as shown in the photo of Figure
20a).
The project included load tests using the a bi-directional load cell at the base of both grouted
and ungrouted drilled shafts to provide a comparison of performance for full scale foundations.
Figure 21 illustrates the improvement in base resistance achieved by base grouting to a pressure
of approximately 5 MPa (750 psi). The data shown are measured load at the base of the drilled

shaft from base grouted shafts except the curve test T3. Shaft T3 was constructed in an identical
way to the others but was not grouted and did not include the base grouting system.
O-cell load (MN)

0 5 10 15 20 25 30
0 0.0
-1 -25.4
Displacement of O-cell plate (inches)
Displacement of O-cell Plate (mm)

-2 -50.8
-3 -76.2
-4 -101.6
-5 -127.0
T3 2.1m (7ft) dia;
(Not Grouted)
-6 all others are -152.4
2.3m (7.5ft) dia
-7 -177.8
0 500 1,000 1,500 2,000 2,500 3,000 3,500
O-cell load (tons)
Fig. 21. Measured Base Resistance from the Base-Grouted Drilled Shafts at the John James Audubon
Bridge, Louisiana.
4.3 Polymer Drilling Slurry

Although the versatility of drilled shaft construction emerged using mineral slurry (mostly
bentonite) for drilling fluids, in recent years the use of polymer-based drilling fluids has become
the prevalent practice where wet-hole techniques are used for construction. Polymers have several
advantages over conventional bentonite for constructors, because it is more easily mixed, de-
sanded, and disposed. The long-chain polymers (shown in Figure 22) mix easily with water and
increase the viscosity of the fluid, and increased viscosity reduces the fluid loss into the
surrounding soil and provides a stabilizing fluid pressure when a positive head is maintained within
the drilled shaft excavation.
Unlike bentonite, polymers do not create a filter cake on the borehole wall, and therefore fluid
loss tends to be a greater than with bentonite slurry. Also, the density of polymer slurry tends to
be lower than bentonite fluids, and cannot be easily increased as with the addition of barite to
bentonite fluids. In coarse or gravelly sand or areas with very high or artesian groundwater levels,
these aspects may result in less effective performance with respect to borehole stability. If the fluid
head in the shaft excavation is not actively maintained at a level higher than the groundwater level,
the fluid level in the excavation will eventually fall and the supporting pressure may be lost. On
the other hand, the lack of a filter cake mitigates one of the major design concerns associated with
the use of bentonite slurry for drilled shaft construction, namely that excessive filter cake buildup
will be detrimental to the bond at the soil/concrete interface and therefore reduce the available side
resistance of the foundation.

a) Scanning Electron Micrograph, 800x b) Polymer Slurry in Use
Fig. 22. Polymer Drilling Fluids
(photo at left courtesy of Likos, Loehr, and Akunuri, Univ. of Missouri).
Specifications for polymer slurry construction often have evolved from those used for
bentonite, but the differences in performance of polymers require several modifications.
Experiences with polymer slurry construction indicate that designers should be aware of several
factors that affect practice.
The upper limits on viscosity used for bentonite are too restrictive for polymer; there is a need
to limit the viscosity of bentonite to avoid excessive filter cake buildup, but polymers can utilize
significantly higher viscosity in order to provide effective stabilization.
Unlike bentonite slurry, the density of polymer slurry will not be much higher than that of
water, and so where groundwater levels are very near the top of the drilled shaft it is critical that a
positive head of 2 m or more is maintained at all times. Where groundwater levels are near or
above the ground surface, it will be necessary that the contractor extend the casing above grade to
provide this head, plus additional distance to allow for fluctuations and working freeboard within
the casing.
Because there is no bentonite filter cake, it is not necessary to limit the exposure time of the
soil to slurry and/or require agitation of the sidewall. There is also substantial evidence that drilled
shafts constructed using polymer slurry result in higher values of unit side resistance in sands and
silts than similar foundations constructed using bentonite (Brown et al. 2002; Brown 2002, Meyers
1996).
Even where natural clays or shales are encountered in the soil, the polymers as shown in Figure
22b tend to stabilize these soils and prevent mixing of the clay with the drilling fluid. Many
contractors like to employ a small amount of polymer slurry when drilling through clay because it
reduces the tendency for clay to stick to the auger. There is also evidence that polymer drilling
fluids may reduce wetting of some shales and thereby reduce the tendency for degradation of shale
when the excavation is open. This behavior can result in improved side resistance for drilled shafts
socketed into shale, especially for large drilled shafts where several days may be required to
complete construction.

Axtell et al. (2009) described a case history for a bridge project in Kansas City where 3.2 m
(10.5 ft) diameter rock sockets were constructed into a shale, and even though the polymer slurry
filled hole was open for four days, load test measurements determined that the unit side resistance
in the socket was a relatively favorable value 720 kPa (15 ksf). The polymer slurry was perceived
to provide benefits with respect to preserving the integrity of the shale and was used for
construction even though casing extended to the rock and slurry was not required to stabilize the
hole. Slake durability tests of the shale with both river water and polymer slurry are summarized
in Table 1. The higher slake durability index and durability rating of the rock specimens tested in
polymer slurry indicates that the shale was subject to less significant degradation in the presence
of polymer slurry compared to that observed when the rock was exposed to plain river water.
Table 1: Slake Durability Test Results.

Natural
Slake Durability Durability Rating Based
Moisture
Sample Index on Shear Strength Loss
Content
(%) Type Id(2) (%) Type DRs
River Water 8.3 II 72.2 Intermediate 61.9
Polymer Slurry 8.3 II 98.2 Hard, more durable 78.6
Where fine sands and silts are present, polymer slurry can present a challenge from the
standpoint of cleaning the slurry. These sands and silts will not stay in suspension and will tend to
settle out slowly after completion of excavation. The de-sanding units used with bentonite slurry
construction do not work with polymer because the polymer molecules would be destroyed by the
shearing process in the de-sander and polymers will also tend to clog the screens. De-sanding of
polymer is normally accomplished by adding flocculants to help promote the settling of solids, a
process that requires that the slurry be maintained in a calm environment so that the sands can
settle out. Flocculation can occur either in the borehole (followed by pumping from the base of the
hole to remove solids) or in a weir tank after removal and replacement of the fluid in the hole with
clean slurry. In a very deep drilled shaft excavation filled with sand-contaminated polymer, solids
can rain out of suspension for days and if left untreated could result in contamination of concrete
during placement. Although flocculants can be used in the borehole to accelerate the process, the
most reliable approach is to fully exchange the fluid by pumping slurry from the base of the shaft
to a holding tank while adding clean slurry into the top of the excavation.
4.4 Verification Testing for Integrity

Only within the last 20 years has integrity testing and load testing of drilled shaft foundations
become commonplace, and the availability and use of these technologies has greatly improved the
efficiency of designs and the reliability of the constructed foundations.
The vast majority of integrity testing performed in North America uses crosshole sonic
logging (CSL) to verify the integrity of the concrete within the drilled shaft. CSL testing relies
upon the measurement of compression waves between pairs of tubes that are typically attached to
the reinforcement. The tubes are filled with water so that acoustic transponders and receivers can
be used to perform measurements through the water, tube, and concrete between tube pairs at
intervals of every few inches. By using multiple tubes, measurements can be performed at various
angles and directions across the drilled shaft diameter and around the perimeter. There is some use

of gamma-gamma testing (largely by Caltrans) to measure density of the concrete in the vicinity
around embedded PVC tubes, although CSL testing is used as a backup if anomalies are detected.
Recently, a promising new method called Thermal Integrity Profiling has been developed by
Mullins (2010) based on thermal measurements; the heat of hydration is measured via downhole
tubes and correlated with the presence of good concrete. This thermal technique offers the promise
to verify integrity before the concrete has fully hardened.
CSL testing and other measures of integrity testing through downhole access tubes offer the
potential to detect even relatively minor inclusions of soil, laitance, low strength concrete, or other
deleterious material. Besides improving the reliability of the constructed foundation, the
accountability provided by these test measurements provide quantifiable verification of an
effective contractor’s work plan and quality control for concrete placement. Effective construction
methods are apparent because of the successful integrity test measurements; ineffective methods
or poor controls are quickly detected. This accountability has had the effect of significantly
improving the quality of construction on public works projects where CSL testing is routinely
employed.
On the other hand, there is a distinct need for engineers and designers to recognize that
perfection is not achievable in this challenging construction environment, and that drilled shaft
designs should be relatively tolerant of minor imperfections. An example is illustrated in Figure
23 from an experimental drilled shaft at Lumber River, SC that was exhumed as a part of research
(Brown et al. 2005). The source of an anomaly in the CSL measurement was exposed when the
exhumed shaft was saw-cut at precisely the elevation revealed by the anomaly. The flaw was a
small pocket of segregated concrete that was lodged against the CSL tube and was approximately
the size of a tennis ball.
An imperfection detected as a result of integrity testing does not necessarily constitute a
deficiency in the drilled shaft. The size of the flaw exposed in Figure 19 should not be of serious
concern because the structural design of the drilled shaft must include adequate tolerance for such
small imperfections. The magnitude of a potential flaw detected by CSL tests can usually be
quantified by close examination of various signal paths across the cross section of the shaft and
even by using tomography techniques if needed. If a potential imperfection is detected it may be
at a location where the drilled shaft is not subject to the maximum flexural demands, and so a
greater tolerance may exist. An engineering evaluation of the structural and geotechnical
performance requirements at the elevation in question is required to determine if a deficiency exists
in a drilled shaft with an imperfection.
Another consideration is that sometimes an anomaly in the signal occurs as a result of non-
uniform concrete curing, tube debonding, or other artifacts of the test measurements. For these
reasons, an anomaly in the integrity test measurement should typically be verified with coring or
other means before expensive corrective action is warranted. An independent evaluation of the
anomaly may be necessary to determine if a real imperfection exists and if any such imperfection
is sufficient to constitute a deficiency.
4.5 Load Testing for Axial Resistance

The majority of load testing performed on drilled shafts in North America utilizes bi-directional
embedded loading jacks, also known as the Osterberg Cell® (O-cell). The O-cell is embedded
within the drilled shaft to engage the portion above the cell as a reaction against the portion of the
shaft below the cell, with measured pressure in the cell calibrated to load and independent
measurements of displacement of the two separate portions of the drilled shaft. This load testing

technology has allowed the measurement of extremely large axial resistance because of the
inherent simplicity in the test and the lack of need for a reaction system.
CSL Tube
anomaly
Fig. 23. Exposure of an Imperfection and CSL Anomaly.
Conventional static top down load tests are still occasionally used with drilled shafts, and load
tests of up to 50 MN (11,000 kips) have been performed with static reaction systems. Other
advancements with drilled shaft load testing include the use of rapid load testing and high strain
dynamic testing with signal matching, as would be performed on a driven pile. The rapid load
testing method is most often employed using the Statnamic® device, which launches a reaction
mass upward with about 20g of acceleration resulting in a downward thrust onto the drilled shaft.
This test method offers a relatively economical means of verifying axial resistance from the top
down without the need for a reaction system, and can often be performed on production
foundations. The equipment available to perform rapid load testing is currently limited to a
maximum applied force of around 45 MN (10,000 kips), and the maximum static resistance which
can be mobilized is slightly lower due to inertial and rate-of-load effects.
The major implications of the advancements in testing for high capacity drilled shaft
foundations are:
• Designers now have the means to obtain measurements that will provide the feedback
necessary to improve design practices.
• Alternative forms of project delivery such as design-build can now include
performance measurements for verification, and the availability of such testing can
allow for performance-based specifications to be employed in the design-build
process.
An example of the use of load testing for verification in design-build is the Honolulu Transit
project currently under construction. The first phase of this project includes approximately 10 km
(6 miles) of elevated guideway to be constructed in a tight space within existing right-of-way. A
single drilled shaft foundation at each pier provides maximum support in the minimum footprint.

Eight load tests using the O-cell method have been performed along the alignment in order to
evaluate both the range of ground conditions encountered and the range of construction methods
used.
Another example is provided by the New Mississippi River Bridge project in St. Louis, where
load tests were used to verify a contractor-proposed “alternative technical concept” or ATC
(Brown et al. 2011). This project utilized a conventional bid-build contract, but bidders were
encouraged to submit confidential ATC’s for review and possible approval during the pre-bid
period. A bidder with an approved ATC could bid the project including the ATC in lieu of the base
design for that portion of the work. The winning bidder submitted an alternative foundation design
which included heavily loaded drilled shaft foundations and a plan for load testing to verify the
axial resistance. The use load testing in the ATC design allowed the use of higher resistance factors
in the LRFD design methodology and potential savings in foundation costs. The investment in load
testing and increased performance risk to the contractor was considered as economically
advantageous because of the potential savings.
Each of the two pylon foundations for this cable-stayed bridge is composed of a 2 x 3 group
of drilled shafts, with permanent casing extending to rock. Each shaft is supported entirely by the
limestone bearing stratum through a 3.4 m (11 ft) diameter rock socket with depths into rock
ranging from 5 m (16.5 ft) to 6.7 m (22 ft). The photos in Figure 24 show the excavation of the
very hard limestone, which typically had compressive strength of around 140 MPa (20,000 psi),
but included thin seams of weaker material. The load test successfully demonstrated a total axial
resistance of 320 MN (72,000 kips), a value which exceeded the requirement for the ATC design.
O-cells
Cored rock from

test shaft
Fig. 24. Load Test Shaft for the Mississippi River Bridge, St. Louis.

5. SUMMARY AND CONCLUSIONS
Today’s engineers have a phenomenal variety of drilled foundation alternatives at their disposal.
The tools that can be employed to solve deep foundation problems range from small diameter
micropiles that can be installed in tight spaces using lightweight and portable rigs, to large diameter
drilled shafts capable of supporting enormous loads. Reliability is enhanced by technology ranging
from on-board rig monitoring and controls as well as post-construction integrity and load testing.
Innovative applications of micropiles have been described which exploit the capabilities of
this technology, along with some new and innovative techniques for installing micropiles. The use
of these drilled foundations is now becoming mainstream in North American practice, with
published design guidelines by agencies such as FHWA and with micropiles now incorporated
into building codes such as IBC and AASHTO.
The construction of CFA piles has matured so that these piles are recognized and more widely
accepted, and the use and reliability of these economical piles is greatly improved by the use of
onboard computer monitoring and control.
The advances in drilling equipment have lead to increased use of drilled displacement piles, a
technology which offers advantages from the inherent ground improvement that is achieved.
Displacement during drilling provides increased axial resistance and reliability in granular soils
compared to CFA piles and the elimination of most excavated materials from the piling operations.
The capabilities of the equipment and methods for installing drilled shaft foundations has lead
to larger and deeper drilled shafts so that effective solutions are provided for projects like the Wolf
Creek Dam and the John James Audubon Bridge. Integrity and load testing provides reliability for
improved design and construction as well as accountability for innovative project delivery methods
such as design-build and the use of contractor-developed alternative technical concepts.
The capabilities of the equipment and drilling techniques present opportunities for engineers,
as well as challenges. The opportunities are present because the work is more sophisticated than
ever, and engineers who are in a position to utilize the available technology can provide value and
efficiency to complex foundation engineering projects. The challenges are posed by the
requirement to understand the complexities of new drilled foundation techniques and the impact
of construction on the performance.
6. ACKNOWLEDGMENTS
The author wishes to acknowledge the contributions of the following individuals with respect to
the information conveyed in this paper: John Turner, Robert Thompson, Steve Dapp, Paul Axtell,
and Tim Siegel of the author’s engineering firm; John Wolosick of Hayward-Baker; Erik Loehr of
the University of Missouri; Andy Burns of Intercoastal Foundations and Shoring; Willie NeSmith
of Berkel and Company; Heinrich Majewski and Peter Faust of Malcolm Drilling Company;
Brannin Beeks of Raito, Inc.; Jim Holtje of PCL Construction; Steve Saye and Luis Paiz of Peter
Kiewit and Sons; Fabio Santillan of Trevi Icos Corp.; and John Kelley of Massman Construction.

7. REFERENCES
American Association of State Highway and Transportation Officials, AASHTO 2008, AASHTO
LRFD Bridge Design Specifications, Customary U.S. Units, 4th Ed., Section 10,
‘Foundations’, Washington, D.C.
Armour, T., Groneck, P., Keeley, J., and Sharma, S. (2000). Micropile Design and Construction
Guidelines Implementation Manual Report No. FHWA-SA-97-070, June, 376 p.
Atlaee, A., Burns, A. and Shah, H. (2010). Hammer-Grout Piles at the Bronx-Whitestone Bridge,
Proceedings of the Deep Foundations Institute 35th Annual Conference, Hollywood, CA, 11p.
Axtell, P., Thompson, R., and Brown, D. 2009. Drilled Shaft Foundations for the kciCON Missouri
River Bridge. Proceedings of the Deep Foundation Inst. 34th Annual Meeting, Kansas City,
10p.
Brown, D. (2002). The Effect of Construction on Axial Capacity of Drilled Foundations in
Piedmont Soils, J. of Geotechnical and Geoenvironmental Engineering, 128(12), pp. 967-973.
Brown, D., Axtell, P., and Kelly, J. (2011). The Alternate Technical Concept Process for
Foundations at the New Mississippi River Bridge, St. Louis, Proceedings of the Deep
Foundation Institute 36th Annual Conference, Boston, pp. 169-178.
Brown, D., Bailey, J. and Schindler, A. (2005). The Use of Self-Consolidating Concrete for Drilled
Shaft Construction: Preliminary Observations from the Lumber River Bridge Field Trials,
Proceedings of the GEC3 Conference, Dallas, TX, pp. 437-448.
Brown, D. and Chancellor, K. (1997). Instrumentation, Monitoring and Analysis of the
Performance of a Type-A INSERT Wall – Littleville, Alabama, Final Report RP 930-335,
Highway Research Center, Auburn University, 105p.
Brown, D., Dapp, S., Thompson, R. and Lazarte, C. (2007). Design and Construction of
Continuous Flight Auger Piles, Geotechnical Engineering Circular No. 8, Federal Highway
Administration Office of Technology Application, Office of Engineering/Bridge Division,
294p.
Brown, D.A. and Drew, C. (2000). Axial Capacity of Augered Displacement Piles at Auburn
University. Geotechnical Special Publication No. 100, ASCE, pp. 397-403.
Brown, D., Faust, P. and Santos, J. (2010). Construction of the Drilled Shaft Foundations for the
Huey P. Long Mississippi River Bridge, New Orleans, Proceedings of the Deep Foundation
Inst. 35th Annual Meeting, Hollywood, CA, 8p.
Brown, D. and Loehr, E. (2007). A Simple Solution for Slope Stabilization Using Micropiles Proc.
of the 32nd Annual Conf. on Deep Foundations, Deep Foundations Inst. Oct., pp. 93-104.
Brown, D., Muchard, M., and Khouri, B. (2002). The Effect of Drilling Fluid on Axial Capacity,
Cape Fear River, NC. Proceedings of the Deep Foundation Inst. 27th Annual Meeting, San
Diego, 5p.
Brown, D., Turner, J., and Castelli, R. (2010). Drilled Shafts: Construction Procedures and LRFD
Design Methods, FHWA/NHI Publication 10-016, Reference Manual and Participants Guide
for National Highway Inst. Course 132014, 972p.
Dapp, S. and Brown, D. (2010). Evaluation of Base Grouted Drilled Shafts at the Audubon Bridge.
Geotechnical Special Publication 199, Advances in Analysis, Modeling and Design, ASCE,
10p.
Esrig, M.I., Leznicki, J.K.,and Gaibrois, R.G. (1994). Managing the Installation of Augered Cast-
in-Place Piles, Transportation Research Record 1447, Transportation Research Board, pp. 27-
29.

Faust, P. (2011). Doyle Drive – Extreme Pile Installation, Proceedings of the Deep Foundation
Institute 36th Annual Conference, Boston, pp. 179-188.
Fleming, W.G.K. (1995). The understanding of continuous flight auger piling, its monitoring and
control, Proceedings, Institution of Civil Engineers Geotechnical Engineering, Vol. 113, July,
pp. 157 – 165. Discussion by R. Smyth-Osbourne and reply, Vol. 119, Oct., 1996, p. 237.
Hasenkamp, R.N, and Turner J.P. (2000). Micropile and Ground Anchor Retaining Structures for
Stabilization of the Blue Trail Landslide, FHWA-WY-00/04F, Federal Highway
Administration and Wyoming Transportation Department, Cheyenne, WY, June, 97p.
International Building Code (2006). International Code Council, Inc.
Katzenbach, R., Hoffmann, H., Vogler, M., O’Neill, M. and Turner, J. (2007). Load Transfer and
Capacity of Drilled Shafts with Full-Depth Casing, Proc. of the 32nd Annual Conf. on Deep
Foundations, Deep Foundations Inst. Oct., pp. 165-175.
Loehr, E. and Brown, D.( 2008). A Method for Predicting Mobilization of Resistance for
Micropiles Used in Slope Stabilization Applications. Report to the joint ADSC/DFI Micropile
Committee, 69p.
Mandolini, A., Ramondini, M., Russo, G., and Viggiani, C. (2002). Full-Scale Loading Tests on
Instrumented Continuous Flight Auger (CFA) Piles, Geotechnical Special Publication No.
116, Ed. by M. W. O’Neill and F. C. Townsend, American Society of Civil Engineers,
February, Vol. 2, pp. 1088 - 1097.
Meyers, B. (1996). A Comparison of Two Shafts: Between Polymer and Bentonite Slurry
Construction and Between Conventional and Osterberg Cell Load Testing, Paper Presented at
the Southwest Regional FHWA Geotechnical Conf., Little Rock, April.
Mullins, G. (2010). Thermal Integrity Profiling of Drilled Shafts, Journal of the Deep Foundations
Institute 4(2), pp. 54-64.
Mullins, G., Dapp, S. and Lai, P. (2000). Pressure-Grouting Drilled Shaft Tips in Sand,
Geotechnical Special Publication No. 100, ASCE, pp. 1-17.
Mullins, A.G., Winters, D. and Dapp, S.D., (2006). Predicting End Bearing Capacity of Post-
Grouted Drilled Shaft in Cohesionless Soils J. Geotech. and Geoenvir. Engrg., Volume 132,
Issue 4, pp. 478-487.
NeSmith, W. M. and NeSmith, W.M. (2009). Advancements in Data Acquisition-Based Design
for Drilled Displacement Piles. Geotechnical Special Publication No. 185, Ed. by M.
Iskander, D.F. Laefer and M.H. Hussein, American Society of Civil Engineers, March, pp.
447-455.
Ryan, P. (2010). Prototype Test Program and Monitoring During Construction of Drilled Shafts,
Transbay Transit Center, Final Report from Arup, May.
Sabatini, P., Tanyu, B, Armour, T., Groneck, P. and Keeley, J. (2005). Micropile Design and
Construction Report No. FHWA-NHI-05-039, December, 436 p.
Siegel, T.C., NeSmith, W.M., NeSmith, W.M. and Cargill, P.E. (2007). Ground Improvement
Resulting from Installation of Drilled Displacement Piles. Proceedings of the Deep
Foundations Institute 32nd Annual Conference, Colorado Springs, CO, pp. 131-138.
Siegel, T.C. and NeSmith, W.M. (2011). Confirmation of Composite Ground Design Using Field
Plate Tests, Proceedings of the Deep Foundations Institute 36th Annual Conference, Boston,
pp. 301-312.
Szynakiewicz, T. and Boehm, D. (2008). Limited Access Underpinning and Re-Leveling of a Six-
Story Office Building using Jet Grouted Micropiles, Proceedings of the Deep Foundations
Institute 33rd Annual Conference, New York, pp. 131-142.

Van Weele, A.F. (1988). Cast-in-situ piles – Installation methods, soil disturbance and resulting
pile behaviour Proc., 1st Int’l Geotech Seminar on Deep Foundations on Bored and Auger
Piles, (Ghent, Belgium) Van Impe, ed, Balkema, Rotterdam, pp. 219-226.

Best practice for performing static loading tests. Examples of
test results with relevance to design
Fellenius, B.H.(1)
(1)
Consulting Engineer, Sidney, BC, Canada. <bengt@fellenius.net>
ABSTRACT. When determining "capacity" from the result of a static loading test, the
profession has no common definition of "capacity". Thus, a group of professionals will come up
with an array of capacity values from the same test results. To improve reliability, piles are
usually instrumented with strain-gages at selected depths. The analysis of the strain records apply
the axial secant stiffness of the pile, which is best determined applying the direct secant modulus
method to the records from a gage level near the pile head and applying the indirect method, the
incremental stiffness method, to records of gage levels down the pile. Examples demonstrate the
necessity that a test be carried out with no unloading-reloading events included and that all
increments be equal and the load levels be held constant for an equal duration not shorter than 10
minutes. If so, the analysis of the records will produce the load distribution along the pile and the
pile toe load-movement response. When combined with the calculated soil settlement around the
piles, the designer will be able to determine the settlement of the piled foundation. The approach
is far more reliable and appropriate for a piled foundation design than one based on some
perceived value of "capacity" with the settlement estimate just referenced directly to the load-
movement of the test.
1. INTRODUCTION
In countries with a stagnant and code-driven piled foundation system, loading tests are rarely
performed. In contrast, in countries with an advanced and variable foundation industry, testing is
a fundamental part of the engineering process. Sometimes, the tests are performed as a part of a
design effort, sometimes they are performed for proof testing during or after construction. Most
of the times, the tests are conventional head-down test with no instrumentation down the pile.
Testing an uninstrumented pile has little value for design, however. Moreover, including
unloading-reloading steps in a test or uneven length of load-holding will adversely affect the
strain records. The pile axial stiffness, EA, can be determined from the measured strains by
applying direct secant and incremental stiffness methods, as is illustrated in this paper. However,
determining the axial loads imposed by the test can be difficult even when the test is properly
planned and executed.
2 CAPACITY
The dominant approach to assessing the results of a static loading test is to determine a pile
capacity from the pile-head load-movement curve. The term "capacity" implies an ultimate
resistance and is, in its purest form, defined as a continued movement for no increase of load
once reached, i.e., plastic response after an initial linear ("elastic") response. Figure 1 shows the
test results of a 14 m long, 400 mm diameter, bored pile equipped with a telltale to the pile toe:
the pile-head load is plotted against the movements of the pile head and the pile toe. The profile
consists of about 4 m of sand on 7 m of clay on a thick layer of dense sand. No obvious "kink" in

the pile-head load-movement curve can be discerned that could have been used to characterize a
capacity. As indicated by Fellenius (1975; 2017), capacity assessed from a static loading test can
be defined by several different methods and, for the curve shown, it would typically range from a
low of about 1,500 kN to the 2,100-kN maximum load applied, indeed, even beyond the
maximum load.
2,500
Pile-head load vs.
Pile Shortening toe movement PILE HEAD
2,000
1,500
LOAD (kN)
1,000
500
0
0 5 10 15 20 25 30 35 40 45 50
MOVEMENT (mm)
Fig. 1 Pile-head load-movement curve
What definition to employ differs widely within the geotechnical profession, as illustrated
by the following example from a prediction event organized by the Universidade Federal do Rio
Grande do Sul in the Araquari Experimental Testing Site, Brazil in 2015. The event comprised
a 1,000-mm diameter, 24 mm long bored pile constructed in bentonite slurry a in sand deposit.
The premise of the prediction was that the test be carried to a capacity defined as the load that
resulted in a pile head movement determined as 10 % of the pile diameter (100-mm), a definition
of "capacity" taken from the EuroCode that has its root in a misconstrued recommendation by
Terzaghi (Likins et al. 2012). Moreover, the definition does not consider obvious aspects such as
whether the pile is driven, bored, advanced by CFA or by full displacement methods, installed in
open borehole, if the hole was maintained by means of slurry, the pile constructed by full-
displacement method, or if the soils are clay or sand.
The task was to predict the pile-head load-movement curve for the test pile until the 10-%
defined maximum load was reached. After the test results had been published, I contacted all
predictors and asked them to tell me, using their own definition, what capacity the actual test
curve demonstrated. Twenty-nine, about half of the total, replied and Figure 2 compiles the
capacities received. The values diverged considerably. Seven accepted the organizers' assertion
that the capacity was the load that gave a movement equal to 10 % of the pile diameter, whereas
the others indicated values that were as low as two-thirds of the maximum—with a 21-mm
movement, as opposed to the 100 mm value stipulated by the organizers. Compilations from
other cases have been published that show similar diversity (Fellenius 2016b, Fellenius and
Terceros 2014). It is obvious that to make use of results of static loading tests performed by
others, a researcher cannot take stated "capacities" at face value, but needs to re-assess the results
in order to develop a consistent data base.

Fig. 2. Test results and capacities assessed by 29 predictors for the Araquari prediction case.
An additional phenomenon affecting the shape of the pile-head load-movement curve and
the assessment of "capacity" is presence of residual force, which is an environmental axial force
introduced in the pile during or after installation or construction of a pile. It is almost always
present in a driven pile and, on occasion, also in a bored pile. For example, jacked-in piles have a
more or less fully developed residual force. The residual force is always zero at the pile head or
at the ground surface, it then increases due to accumulated negative skin friction to a maximum
value at a "neutral plane" where it is in equilibrium with the below this point developing positive
shaft resistance and, usually, some residual toe force.
Figure 3 compares the pile-head load-movement curves of two piles, one with fully
developed residual force and one with no residual force. The piles and soil are otherwise
identical in all respects, including using the same ultimate resistance and load-movement
response for the pile elements. The figure includes the pile "capacities" determined according to
the in North America common offset-limit method (Davisson 1972, Fellenius 2017). Other
definitions show a similar difference between the capacity values determined from the two
curves. Obviously, presence or not of residual force will affect the evaluation of the results of a
test in addition to the fact that the capacity evaluated from the pile-head load-movement curves
are very different. The example emphasizes the uncertainty of the concept of "capacity".
1,000
800
Residual
Load present
LOAD (KN)
600
No Residual
Load present
400
200
= OFFSET
LIMIT LOAD
0
0 5 10 15 20 25 30
MOVEMENT (mm)
Fig. 3. The effect of presence of residual load.

3 DETERMINING THE PILE STIFFNESS, EA, AND LOAD DISTRIBUTION
It is common to instrument the pile in order to also determine the distribution of axial load down
the pile. In the past, the instrumentation consisted of telltales, i.e., rods that measured shorting
between two points in the pile. The simplest arrangement being a telltale to the pile toe
measuring the total shortening of the pile and enabling the pile toe movement to be determined
by subtracting the total pile shortening from the pile head movement, as was illustrated in
Figure 1. A modern type of telltales is the Glostrext system (Hanifah and Lee 2006). The
shortening caused by the applied load divided by the telltale length is the induced strain. That
strain multiplied with the pile stiffness, EA, is the average load over the telltale length.
Sometimes telltale records are back-analyzed employing total stress method with a unit shaft
resistance assumed constant along the pile. The average load in the pile is then located at the
mid-point of the telltale distance. However, unit shaft resistance is proportional to the effective
overburden stress and it, thus, increases linearly with depth. For a linearly increasing shaft
resistance, the average load over the telltale length should be plotted at the point h/√2 = 0.70h
(Fellenius 2017). Plotting the average value at mid-point is a common error. It implies more
shaft resistance in the upper portion of a pile and less in the lower portion. The error has
contributed to the “critical depth” fallacy (Fellenius 1995; 2017).
For a telltale starting from the pile head, the applied load minus two times the difference to
the average load represents the load at the telltale foot regardless of the unit shaft resistance
being assumed constant or linearly increasing. Thus, a single telltale from the pile head will
indicate two load values in the pile: one for the average value and one at the telltale foot. Of
course, as the accuracy of telltale measured shortenings is often poor, the values are should be
considered to be approximations (strain-gage measurement usually provide more representative
loads). More important is that combined with the telltale-determined movement of the pile toe,
the toe telltale provides the pile-toe load-movement response, a very useful record for the
analysis of the pile response.
Other than a load cell that measures the applied load (relying on the pressure in the
hydraulic jack is not recommended), gages that measure axial load directly are rarely used. Most
instrumentation consists of vibrating wire or electric resistance gages that measure the strain
induced at certain levels in the pile. Thus, the pile axial stiffness (EA) is an integral part of the
measuring gage. The stiffness, however, is not a fixed entity. The two components, modulus (E)
and area (A), are only known accurately for steel piles with definite shape. Premanufactured
piles, such as prestressed concrete piles have constant cross section, but their modulus can vary
within a wide range depending on amount of reinforcement, of course, but also on the fact that
concrete, unlike steel, does not have an a-priori known modulus, and reduces with increasing
stress (or strain). The actual cross section of a bored pile can deviate quite widely from the
nominal size depending on construction method and soil layering.
The issue of uncertain modulus can be addressed by placing a gage level at a distance below
the pile head that is sufficient to ensure that the uneven stress distribution immediately below the
pile head has equalized and yet not so deep that shaft resistance will interfere with the axial load
in the pile. That distance is about 2 to 3 pile diameters. The records obtained from the gage level
can then be used to determine directly the secant stiffness, EsA, (N.B., modulus and cross section
together) of the pile by plotting load divided by strain (Q/ϵ) versus strain (ϵ), as indicated in
Figure 4. The first point in the figure is off the straight-line relation, which is probably due to
initial friction in the system. However, a friction or reading correction of the 200-kN first load
increment by subtracting a mere 21 kN would bring also the first point into line.

4,000
SECANT STIFFNESS, EA (MN)

3,500
EAsecant (MN) = 2,600 - 0.07µε

3,000
2,500
y = -0.071x + 2,617
2,000
1,500
0 100 200 300 400 500 600 700 800
STRAIN, με (---)
Fig. 4. Pile stiffness for a 400-mm CFA pile (after Fellenius 2012).
The direct secant modulus plot for the gage records indicates an initial pile axial stiffness
(EA) of 2,600 MN reducing with increasing strain as the load increases. The decrease with
increasing strain is minute for the case. The important observation is that the stiffness is obtained
directly from the measurements and does not depend on a frequently uncertain value of the pile
cross section and a modulus obtained from calculations or testing of a separate specimen.
The direct secant method depends very much on the accuracy of the load and strain records.
As mentioned, if the pile has been unloaded and reloaded—whether this was intentional or not
matters little, the strain readings will be adversely affected by the so-introduce hysteresis effect.
Figure 5 shows the secant analysis of strain-gage records from a gage level in a 600-mm
diameter, octagonal, 33 m long, prestressed concrete pile driven into a sandy silt. The gage level
is located about 1.5 m below the pile head and 1.0 m below the ground surface. The pile was
loaded in twenty-six 150-kN increments to 3,900 kN maximum load. The night before the test,
an unprescribed check of the test set-up was carried out involving loading the pile to about
1,000 kN without taking records. As seen, the incident resulted in a disturbance of the initial part
of the secant stiffness line.
30
SECANT STIFFNESS, EA (GN)
20
EAsecant (GN) = 11.2 - 0.005µε
y = -0.0053x + 11.231
10
Secant stiffness after adding

20µε to each strain value
0
0 100 200 300 400 500
STRAIN (µε)
Fig. 5. Pile stiffness for a 600-mm diameter prestressed pile (after Fellenius 2012).

An initial curvature of the direct stiffness line can often be removed by adding a small strain
value to all measured strains. In this case, the added value is small, only 20 µϵ. Also for this pile,
the deviation of the first point of the corrected stiffness line is probably due to friction in the
system and it could have been removed by adding 77 kN to the first increment of load. The point
of the example is to show that a preceding full unloading-reloading cycle has affected the gage
records. Indeed, unloading-reloading should never be a part of a static loading test as such events
will adversely affect the records and may cause them to be unacceptable for analysis.
The direct secant method will not work where the pile is subjected to shaft resistance along
the length above the gage level. For those records, the stiffness can be determined from an
incremental stiffness method (tangent stiffness method). Figure 6 show head-down test results
from a gage level at 11 m depth in a 1.0-m diameter bored pile constructed in a marine clay. The
first measurements are affected by shaft resistance and the linear trend only develops after the
shaft resistance is fully mobilized. Note, in strain-softening or strain-hardening soil, the slope of
the line will be steeper or flatter, respectively, than the true stiffness line.
40
INCREMENTAL STIFFNESS (GN)
P3 Stage L1
SGL11
35
Tangent Stiffness, EtA (GN) = 32.6 - 0.020 µε
gives the Secant Stiffness, EsA (GN) = 32.6 - 0.010 µε
30
y = -0.0195x + 32.587
R² = 0.7729
25
20
0 100 200 300 400 500
STRAIN (µε)
Fig. 6.Pile incremental stiffness for a gage level 12 m down in a 1,000-mm

diameter bored pile. (Data from Fellenius and Tan 2010).
The loads translated from the strain records apply the secant stiffness times the strain, Es A ϵ.
The y-intercept (stiffness at zero strain) is the same for the secant and incremental plots and the
slope of the straight line in the incremental plot is twice that of the secant line. Thus, for the
shown plot, the secant stiffness, EsA (GN), is 32.6 - 0.010µϵ. Because the incremental method is
based on differentiation, small inaccuracies in the records are exaggerated and an incremental
stiffness plot is always more scattered than a direct secant plot. Thus, provided that the tests is
carried far enough so the applied load has imposed sufficient strain in the pile for a meaningful
plot of the data, the records can be used to determine the stiffness of a test pile at the various
gage levels down the pile (Fellenius 1989).
Figure 7A shows the load-movement curve from a bidirectional test on a 1.85 m diameter,
65 m long bored pile, for which accidental hydraulic leak necessitated an unloading and
reloading cycle (Thurber Engineering Inc., Edmonton; personal communication 2016). Figure
7B shows the incremental stiffness curves from the initial and reloading records.

CELL DOWNWARD MOVEMENT (mm)
0 110
INCREMENTAL STIFFNESS, ∆Q/∆(µϵ)

SGL3 INITIAL
1L 100
SGL3 RELOADING
-20
Unloading
because of a 90
-40 hydraulic leak
1L ( ) 2L
80
(GN)
-60
2L
70 y = -0.0215x + 81.827
R² = 0.7967
-80
60
-100 50
0 10,000 20,000 30,000 40,000 50,000 60,000 0 100 200 300 400 500 600 700
A
LOAD (kN) B STRAIN (µϵ)
Figure 7. Effect on the incremental stiffness curve from an unloading-reloading cycle.
It is clear from the examples illustrated in Figures 5 and 7 that using the load and strain
records to determine the pile stiffness requires that the data must not be affected by unloading-
reloading cycles. Many practitioners incorporate intentional unloading-reloading cycles in a
static loading test and they frequently keep the applied load constant for a longer duration at one
load level or other. I have many times tried to have users of such actions explain to me what they
expect to gain from this extraneous imposition on the test procedure, but nobody had ever been
able to tell me anything else than "this is what we always do" or "this is what I think the code
says I must do". The fact is that nothing is gained by this and such deviation from the simple
direct incremental procedure will instead result in that the investment in the instrumentation is
wasted.
Note, translating strain measurements to load requires that the measured strains cover an
acceptable range. There is little sense in investing in instrumentation if the induced strain at the
maximum load are smaller than 200 µϵ. Designing the pile and test toward achieving strains in
excess of 500 µϵ is preferable. If a calculation check shows that the strain are small, and it is
important to determine the load distribution, it is much better to do the test on a smaller diameter
pile and cautiously "extrapolate" the result to the larger diameter pile. For example, a pile with
half the diameter will show four times larger strain for the same load. In the process, it might be
realized that the original pile design is too conservative and a smaller diameter pile will suffice.
Of course, the transfer of the analysis results for the smaller diameter pile to the larger diameter
pile must be carried out with care and with some conservatism.
Furthermore, to obtain records that can be used for determining the pile axial stiffness and
load distribution, requires that a static loading test, be it a head-down test or a bidirectional test,
be carried out by applying equal increments of load and time. At each level, the load must be
maintained (held constant) for an equal length of time. The load-holding time can be short or
long, but an interval shorter than 10 minutes or longer than 20 minutes is impractical. The
frequently applied 5-minute load-holding interval is not suitable when testing piles with strain-
gage or other instrumentation down the pile. The reason is that it takes a few minutes after
adding an increment for the pile to react, i.e., for the gages down the pile to register the load
change. Therefore, the applied load and the strain-gage records are best combined after at least
ten minutes of constant load at the pile head (or bidirectional cell). And, such "waits" must be the
same for all load levels.

These days, it should not be necessary to state that the pump supplying pressure to the
hydraulic jack must be able to maintain the pressure automatically. However, only too often is an
investment in a sophisticated test employing instrumentation severely affected by the variation in
the applied load level during the load-holding phases originating from a pump without automatic
load-holding means or, even, a manually operated pump. It must be realized that the load exerted
by a hydraulic jack is affected by friction between the two cylinders making up the jack and that
friction is different when pressure is increased and when it is let to relax or is reduced (because
the direction of the relative movement between the outer and inner jack cylinders are different).
Naturally, the load is monitored with a load cell so any load variation would seem to be
measured. However, load cell or not, the load will vary up and down as the pressure pump is
relaxed or engaged, and knowing the variation is not the main point. A pump with an automatic
pressure-holding device will minimize the effect of friction as well as maintain an even load.
The days of manual recording of the records are over and records are now obtained at every
30 seconds, or so, and stored in a data collector (data acquisition unit). However, the records
must be obtained using a single data collector. Do not use one collector for the load records and
pile movements and a separate one for the strain records. Marrying records using the time stamp
does not work. My experience is that ever so often a line shift is made and a "divorce" occurs
before the marriage is completed.
The objective of determining the stiffness is to find the distribution down the pile for the
applied loads. The pile for which the load-movement results were shown in Figure 1 was
instrumented with four levels of strain-gages. The pile stiffness was determined using the
methods described above and the loads at each gage were determined from the measured strains.
Figure 8 shows the so-obtained load distributions. In addition, it shows the calculated pile toe
load plotted versus the measured pile toe movement. These results together with the load
distributions are the key results from a static loading test for use toward a piled foundation
design.
LOAD (kN)
0 500 1,000 1,500 2,000 2,500
0
SGL 4
2
4 SGL 3
DEPTH (m)
8 SGL 2
10
12
SGL 1
14
0
Toe Movement
10 PILE TOE:
Measured movement
(mm)
20 and calculated load

30
0 500 1,000 1,500 2,000 2,500
LOAD (kN)
Figure 8 Load distribution determined from strain gage records.

4. EMPLOYING THE TEST EVALUATIONS IN PILED FOUNDATION DESIGN
Figure 9 shows the results of testing and analysis of a 25 m long, strain-gage instrumented
auger-cast pile constructed through sand and silty clay to bearing in a glacial till designed
according to the Unified Design Method (Fellenius 2004; 2016a; 2017). The spacing between the
piles was large enough for the piles to be acting as single foundation-supporting units. The
distributions of load and settlement are shown for a test pile, at the site. After construction, a fill
was placed over the site, which introduced soil settlement and, therefore, downdrag.
Consequently, the long-term load distribution will increase downward from the applied dead
load to a maximum at the force equilibrium—the neutral plane. This is where the dead load plus
the drag force due to accumulated negative skin friction are equal to the positive shaft resistance
and the toe resistance below. The latter depends on the magnitude of the imposed toe movement.
The neutral plane is also the settlement equilibrium, the location where there is no relative
movement between the pile and the soil. The figure demonstrates how the forces and soil
deformation interact and that the settlement of the pile head, i.e., of the piled foundation, is
governed by the settlement at the neutral plane. The location of the neutral plane is controlled by
the loop from toe force that governs the force equilibrium that determines the settlement
equilibrium that results in a toe movement that establishes the toe force. For the loop to close, the
final toe force must be equal to the starting toe force. There is only one neutral plane location
that satisfies this requirement. The final result is the pile cap settlement as indicated in the figure.
The pile head
LOAD (kN) load that gave SETTLEMENT (mm)
Qd the toe load, Rt Pile Cap Settlement
0 2,000 4,000 6,000 0 50 100 150 200
0 0
Bef ore Con- In the long
struction term Silt
5 Sand 5
DEPTH (m)
10 10
Af ter Con- Soil Settlement
DEPTH (m)
struction
15 Clay 15
Qn Neutral Plane
20 20
25 25
Till
Rt
30 30
0 0
LOAD (kN)
1,000 1,000 q-z relation

LOAD (kN)
2,000 2,000
3,000 3,000
4,000 4,000
0 1,000 2,000 3,000 4,000 5,000 0 50 100
LOAD (kN) PILE TOE PENETRATION (mm)
Figure 9. The unified pile design loop for determining settlement

(Fellenius and Ochoa 2009).

The results of the instrumented static loading test and strain-gage analyses provide the
necessary load distributions and pile toe movement to combine with the settlement analysis
(which must incorporate all aspects causing the soil to settle) in order to determine the piled
foundation settlement. If the settlement is too large, lowering the neutral plane will reduce it,
which can be achieved by either lengthening the pile or reducing the load. The main point is that
the proper testing procedure, analysis methods, and design process enable implementing a safe
design within acceptable settlement. "Capacity" needs not be a part of the picture.
5. CONCLUSIONS
The scatter of capacity values resulting from the various definitions in vogue in the profession
and the effect of residual force on the evaluation of capacity are addressed. It is shown that
including an unloading-reloading event in a static loading test will adversely affect the
calculation of the pile axial stiffness. The analysis of the strain-gage records aim to determine the
load distribution and pile toe-load-movement, which, when combined with the soil settlement in
an interactive analysis, will show the settlement of the piled foundation. Designing for settlement
makes the conventional "capacity approach" redundant and provides a more reliable design.
6. REFERENCES
Davisson, M.T., 1972. High capacity piles. Proc. of Lecture Series on Innovations in Foundation
Construction, ASCE Illinois Section, Chicago, March 22, pp. 81-112.
Fellenius, B.H., 1975. Test loading of piles—Methods, interpretation, and proof testing. ASCE
Journal of the Geotechnical Engineering Division 101(GT9) 855-869.
Fellenius, B.H. and Atlaee, A., 1995. The critical depth – How it came into being and why it
does not exist. Proceedings of the Institution of Civil Engineers, Geotechnical Engineering
Journal, London, 113(2) 107 111.
Fellenius, B.H., 2004. Unified design of piled foundations with emphasis on settlement analysis.
Honoring George G. Goble—Current Practice and Future Trends in Deep Foundations. Geo-
Institute Geo-TRANS Conference, Los Angeles, July 27-30, 2004, Edited by J.A. DiMaggio
and M.H. Hussein. ASCE Geotechnical Special Publication, GSP125, pp. 253-275.
Fellenius, B.H., 1989. Tangent modulus of piles determined from strain data. ASCE,
Geotechnical Engineering Division, the 1989 Foundation Congress, F.H. Kulhawy, Editor,
Vol. 1, pp. 500-510.
Fellenius, B.H. 2012. Critical assessment of pile modulus determination methods. Discussion.
Canadian Geotechnical Journal, 49(5) 614-621.
Fellenius, B.H., 2015. Field Test and Predictions. Segundo Congreso Internacional de
Fundaciones Profundas de Bolivia, Santa Cruz May 12-15, Lecture, 22 p.
Fellenius, B.H., 2016a. The unified design of piled foundations. The Sven Hansbo Lecture.
Geotechnics for Sustainable Infrastructure Development – Geotec Hanoi 2016, edited by
Phung Duc Long, Hanoi, November 23-25, pp. 1-26.
Fellenius, B.H., 2016b. Fallacies in piled foundation design. Geotechnics for Sustainable
Infrastructure Development – Geotec Hanoi 2016, edited by Phung Duc Long, Hanoi,
November 23-25, pp. 41-46.
Fellenius, B.H., 2017. Basics of foundation design—a textbook. Electronic Edition,
www.Fellenius.net, 451 p.

Fellenius, B.H. and Ochoa, M., 2009. Testing and design of a piled foundation project. A case
history. Journal of the Southeast Asian Geotechnical Society, Bangkok, 40(3) 129-137.
Fellenius, B.H. and Terceros, M.H. 2014. Response to load for four different bored piles.
Proceedings of the DFI-EFFC International Conference on Piling and Deep Foundations,
Stockholm, May 21-23, pp. 99-120.
Hanifah, A.A. and Lee S.K., 2006. Application of global strain extensometer (Glostrext) method
for instrumented bored piles in Malaysia. Proc. of the DFI-EFFC 10th Int. Conference on
Piling and Deep Foundations, May 31 - June 2, Amsterdam, 8 p.
Likins, G.E. Fellenius, B.H., and Holtz, R.D., 2012. Pile driving formulas. Pile Driver Magazine,
No. 2, pp. 60-67.

Conventional methods and
recent developments in retaining walls
Gerressen, F-W.(1)
(1)
BAUER Maschinen GmbH, Schrobenhausen, Germany <Franz-Werner.Gerressen@bauer.de>
ABSTRACT. A retaining wall is a structure designed and constructed to resist the lateral
pressure of soil when there is a desired change in ground elevation that exceeds the angle of
repose of the soil. The walls must resist the lateral pressures generated by loose/soft soils or, in
the case of existing ground water, water pressures as well. Very typical is the use not only for the
stabilization of slopes, but of course for the use of excavation pits. The paper gives a brief
overview of existing methods, but will then focus on recent developments which improve the
opportunities of using the retaining wall system.
1. INTRODUCTION
Retaining walls are very important structures in the field of special foundation business. Not only
for the use of stabilization of slopes, but mainly as supporting structure for excavation pits.
Using retaining walls for excavation pits safes a lot of space and allows the installation of
excavation pits also close to adjacent structures and under the ground water table. There are
several different systems of retaining wall types existing as shown in figure 1.
Fig. 1. Comparison of retaining wall systems.
All of these systems have advantages and limitation as described above in terms of
deflections, water tightness and durability, but this is at that stage only the comparison in terms
of their functionality. Obviously, there are advantages and limitations for the various systems in

terms of soil conditions, depth limitations and costs. All these factors must be taken into
consideration to achieve the best solution in terms of feasibility and cost efficiency.
Recent developments now moving the borderlines for the use of retaining wall systems, e.g.
in terms of cost efficiency, execution quality, depth and soil limitations.
In the following some of these developments will be described more in detail.
2. RECENT DEVELOPMENTS
2.1 Secant pile walls using Cased Continuous Flight Auger piles (CCFA)
2.1.1 Construction principle
The fundamental principle of double rotary drive systems is a continuous flight auger in
combination with an outer casing, that are drilled simultaneously, but in the opposite direction,
into the ground. These characteristics resulted to the common term Cased Continuous Flight
Auger (CCFA) pile.
Fig. 2. Construction Procedure of CCFA pile.
The installation process as shown in figure 2, can be described as follows:

• A continuous flight auger in combination with an outer casing is drilled simultaneously,
but in opposite directions into the ground.
• The spoil is transported upwards by the auger flights surrounded by the casing and exits
through openings at the top of the casing. The system allows relative vertical movement
between the casing and the auger of +/- 300 mm. Depending on the soil conditions the
auger can drill in advance or within the casing. Normally and particularly in non-cohesive
soils, the casing needs to be maintained in advance or at the same depth to stabilize the
surrounding soil.

• Once the founding level is reached, the concrete pump is activated and concrete is
pumped through the swan neck and swivel to fill the hollow stem of the auger.
• Extraction of the CCFA equipment now begins and its speed of extraction is controlled
relative to the flow of concrete. The speed of extraction is adjusted so as to maintain a
slightly positive pressure of concrete in the hollow stem at all times.
• In the CCFA method, contrary to other piling methods, the reinforcing cage is now
installed after the borehole has been concreted. Depending on its design the pile can be
reinforced over part or to its full depth.
2.1.2 Quality control system of CCFA method for secant pile walls
The CCFA method is mainly used for the installation of secant pile walls. The order of
installation reduces many of the sequence issues experienced with traditional methods. The
quality standard for pile walls is necessarily very high as the costs in case of repairs would be
significant and repair work time consuming. It is therefore essential to employ an accurate
quality surveillance system.
As primary requirements, a stable working platform with less than 3% inclination and a
guide wall as shown in figure 3 are highly recommended.
Fig. 3. Typical concrete cast guide wall.
The guide wall ensures the correct starting location of every pile and facilitates the set-up of
auger and casing. The verticality of the casing is checked manually with a water level. However,
the verticality of the mast is both measured and recorded by electronic sensors. Verticality
corrections are easily made in the x & y directions either manually or automatically, ensuring
that the mast and tools are kept vertical at all times. For quality control all drilling data and data
relative to the verticality of the mast are available in real time.
Drilling assistant CCFA
Many factors contribute to the efficient drilling and subsequent concreting of CCFA piles. To
assist and allow the operator to monitor all the various parameters, drill and extraction assistant
programs have been developed. During the drilling phase the optimal performance is affected by
the applied torque and crowd forces and rotation speed. These will need to be varied for each
particular soil conditions encountered. In addition to these rig-related factors the geometry of the

casing and auger type also has an influence. All these factors are used in the drill assistant
program to achieve the desired optimal performance, which is measured as the penetration per
revolution. Control of this measure is widely accepted as a major factor in avoiding problem in
CFA and CCFA piling. In particular there are the following benefits:
• avoids over-flighting (crowd speed too slow in combination with too many flight
rotations, therefore loosening of surrounding soil)
• avoids corkscrew of auger (Crowd speed too quick in combination with too little flight
rotations, cork screwing effect)
• optimizes the filling grade of the auger
• assists the operator.
Extraction assistant for CCFA
An efficient concreting process is achieved by monitoring a combination of the concrete volume,

concrete pressure and the extraction rate. There are two basic methods of measuring the volume
of concrete supplied by a concrete pump. The first is the direct method where the volume is
measured by a flow meter placed in the concrete pipe work. Another method is to measure each
stroke of the concrete pump. A pressure gauge is installed in the concrete delivery pipe at the
drilling rig. The advantage of this system is its simplicity. Any considerable mistake can be
encountered easily on site. Each stroke that is not counted or counted too much is shown as an
abrupt discontinuity of the concrete flow. Therefore it is important to maintain the speed of the
concrete pump.
Now the extraction assistant for the concreting process has sufficient information to guide
the equipment for the casting of the pile. The retraction speed is calculated by volume formulas.
A significant over consumption of concrete depending on ground conditions, is adjusted in the
program. The working screen (figure 4) of the equipment displays all important information of
the production process in real-time.
Fig. 4. working screen of the extraction assistant.

Verticality
The verticality of the pile depends on an accurate alignment of the casing, the casing guide and
the stiffness of the drilling tools. The relative flexural rigidity of a casing is min. 100 times
higher than the rigidity of a CFA auger. The opposed drilling direction of casing and auger
increases the stability additionally as any deflection is compensated and the drill string is
stabilized. The high grade of verticality and pile quality could be seen impressively on a jobsite
where standard CFA and CCFA piles were used next to each other:
Fig. 5. Secant pile wall, diameter 750 mm.
During the execution of the pile wall by the CFA method, the operator noticed already a
deflection in the verticality of the piles. Especially when drilling a secondary pile, the auger
drifted out of direction. Therefore the piling method was changed to CCFA. The result was an
imposing improvement of pile quality in the same soil conditions, which could be seen after the
excavation of the pit. With the CCFA method a 1 in 200 verticality tolerance could be achieved,
whereas the secondary CFA piles show significantly higher deviations.
2.1.3 Conclusion for CCFA
The major application for the CCFA system is the construction of secant pile walls. In
comparison to traditional systems like cased bored piles using Kelly drilling or standard CFA
piling, the CCFA system has advantages in costs and time against the cased bore piles and in
quality against the standard CFA system. Therefore, the CCFA system can be seen as a
significant improvement for the installation of secant pile walls.

2.2 Cutter Soil Mixing (CSM)
2.2.1 Construction principle
The Cutter Soil Mixing (CSM) System differs essentially from traditional soil mixing techniques
in that the in situ mixing of the existing soil with self-hardening slurry is performed by mixing
tools rotating around horizontal axis rather than the traditional vertical axis. The in-situ soil
mixed with self-hardening slurry produces a rectangular shaped panel element, which e.g. takes
on the role of a cut-off and/or structural retaining wall.
The cement and bentonite content and the water/cement ratios of the mixing slurry are
determined based on the strength and/or permeability requirements of the project and the
properties of the soil being mixed. In general, for a stronger wall, cement content is increased
and water/cement ratio is lowered. Typically, sandy soils will require a larger amount of
bentonite in the slurry than clays. At some clay sites where enhanced resistance to permeability
is not required, acceptable liquefaction of the soil can be achieved without the use of bentonite.
A typical construction sequence as shown in Fig. 6 consists of the following steps:
Fig. 6. Construction Procedure.
• Excavation of a guide trench for collecting surplus slurry

• Positioning of the cutter head in wall axis. The construction of a guide wall is not
required.
• The mixing tool is driven into the ground at a continuous rate. The soil matrix is broken
up by the cutting wheels and at the same time a fluid is pumped to the nozzles, set

between the cutting wheels, where it is mixed thoroughly with the loosened soil. Adding
a compressed airstream can improve the breaking and mixing process in the down-stroke
phase. The direction of rotation of the wheels can be changed at any time. The rotating
wheels and cutting teeth push the soil particles through vertically mounted shear plates
that have the effect of a compulsory mixer. Penetration speed of the cutter and the
volume of pumped fluid are adjusted by the operator to create a homogeneous, plastic
soil mass which permits easy penetration and extraction of the machine. Typical speed of
penetration is 20 – 60 cm/min.
• After reaching the design depth, the mixing tool is slowly extracted while cement slurry
is continuously added. Homogenization of the fluidified soil mixture with the fresh
cement slurry is ensured by the rotation of the wheels.
• Reinforcing elements required for structural purposes can be inserted into the completed
wall. A standard solution is the insertion of H-beams. In case of shallow depths these will
usually penetrate due to their own weight; otherwise a light vibrator can be used to assist
their installation. The distance of the beams and beam cross-sections are designed on the
basis of the applied loads and on the results of the characteristic strength of the soil.
A continuous wall is formed in a series of overlapping primary and secondary panels.
Overcutting into fresh adjacent panels is called „fresh-in-fresh method“. The cutter technique
also allows the “hard-in-hard method”, whereby secondary panels are cut into the already
hardened primary panels. The cutting and mixing procedure can be executed in two ways:
One-phase system
During the penetration (down-stroke) phase, cutting, mixing, liquefying and homogenising is
performed while pumping the binder slurry into the soil. Adding compressed air is recommended
for assisting the down-stroke phase. As a rule of thumb about 70 % of the total slurry volume is
pumped during this phase. The backflow of soil and binder slurry is collected in the pre-
excavated trench or stored in a settling pond to be removed later off the site. After reaching the
design depth air flow is stopped. In the upstroke phase the remaining volume of binder slurry is
blended into the soil. The speed of extraction can be high as the majority of the binder slurry has
already been mixed with the soil in the down-stroke phase.
Two-phase system
The soil is liquefied and homogenized in the down-stroke phase by pumping of bentonite slurry
into the soil. The mixing process can be supported by adding compressed air. The backflow of
soil and bentonite can be pumped to a de-sanding plant where the sand is separated from the
slurry which is then pumped back to slurry storage area. When the backflow becomes too heavy
for pumping, it can be removed by a backhoe or a scratching belt. A scratching belt is a belt
which is placed on top of the trench, immerging into the trench and conveying the bentonite/soil
mixture automatically up-wards into a dewatering screen Using a hose pump the fluid fraction is
then pumped to the de-sanding plant unit for further treatment. After reaching the design depth,
the flow of bentonite is stopped and replaced by cement slurry. On the upstroke movement
cement slurry is mixed thoroughly with the liquefied soil. The speed of extraction and flow of
binder are adjusted to ensure that the total calculated quantity of binder is blended with the soil.

2.2.1 Equipment
The most important elements of the CSM unit are the cutter gear drives. They are driven by
hydraulic motors which are located in a water-tight housing. The slurry is introduced into the soil
directly between the mixing wheels. During construction, the counter-rotating mixing wheels and
vertically mounted plates are effectively acting like a forced action mixer. Because of this
mixing principle, CSM is suitable for mixing cohesive soils. For loosening and mixing the soil
different types of mixing wheels were developed.
Fig. 6. Mixing wheels and housing for hydraulic gear drives.
The size of individual panels is determined by the type and size of equipment being
deployed. Panels can be constructed in lengths ranging from 2.4 m to 2.8 m and wall thicknesses
of 0.55 m to 1.2 m. The mixing unit (Fig. 6) is either mounted on a guided Kelly bar or on a wire
rope-suspended cutter frame equipped with special steering devices. The standard set-up is the
"Kelly-guided" setup, capable of reaching depths down to 53 m (Fig. 7). "Rope-suspended"
systems are particularly suited for construction of deep soil mix walls. The greatest depth to date
at which a wire rope-suspended unit has successfully installed a soil mix wall is 80 m, using a
compact unit (Fig. 7).

Fig. 7. Left: Kelly-guided CSM unit, USA; right: rope-suspended CSM unit, China.
Both systems must be accompanied by an intensive quality assurance programme. All

process-specific production and plant-specific operating data are visualised throughout the
construction phase and stored for subsequent documentation and evaluation. Information
presented includes penetration rates, alignment, and slurry injection rates and volumes (Fig. 8).
Fig. 8. Rig operator’s on-board computer screen.

2.2.2 Comparison with other techniques
The CSM process has significant advantages over conventional techniques, e.g. secant pile walls
or sheet piling walls. These are:
• The existing soil is utilised as construction material.
• Very little spoil material is generated; this renders the technique particularly suited for
work on contaminated sites.
• CSM is an ideal alternative to the traditional system, a soldier beam wall with timber
lagging ("Berlin wall") for the use in high groundwater conditions, or to sheet pile
walls in soil formations unsuitable for pile driving or in close proximity to vibration-
sensitive buildings.
• CSM is a vibration free method.
• No delivery of ready mixed concrete is necessary.
In comparison to traditional deep soil mixing methods CSM has the following advantages:
• A high degree of verticality of wall panels is achieved by the counter-rotating cutter
wheels.
• The cutter principle ensures construction of clean and trouble-free joints even between
wall panels of different construction age e.g. after weekend breaks or prolonged
stoppages on site.
• Harder soil formations can be easily penetrated, broken down and mixed by using the
cutter wheels as cutting and mixing tool.
• Homogenises the cohesive soils and self-harden slurry through horizontal mixing.
• In relation to small base units, high daily output and high panel depth may be achieved.
2.2.3 Applications
There are many possible applications for the CSM method. The main applications are:
• Cut-off walls
• Retaining walls (often with cut-off wall function)
• Foundations
• Slope stabilization
• Soil improvement / soil stabilization
• Liquefaction mitigation
As this paper focuses on the application of retaining walls, the following references will give
an overview of the use of CSM. The system is commonly used to construct water-tight soldier
beam retaining wall system for excavation pits especially near sensitive inner-city buildings
where vibration-free systems are essential. CSM provides an alternative solution to conventional
methods such as secant pile walls and/or sheet piles.
An example for an excavation pit shows a jobsite executed in Sydney, Australia. For the new
construction of some residential buildings a retaining wall was executed with CSM method. The
first project stage consists of a 13-storey building with one additional underground car park
floor; see Fig. 9.

Fig. 9. Computer simulated building of first stage.
The soil mix wall was used as retaining wall during construction of the building as well as
permanent external wall for the car park floors. The mixing depth was between 10 and 16 m and
the wall thickness 55 cm. The total wall area for this building was about 4,000 m². After
finishing the soil mixing wall the pit was excavated incrementally to a depth of about 6 m. To a
depth of 10.5 m, the wall acts as a retaining wall, below 10.5 m as a cut-off wall. The required
28-day unconfined compressive strength (UCS) was 5 MPa for the retaining wall function and 3
MPa for the cut-off wall function. The requirements on the water tightness of the wall were
relatively high. The client insisted that no leaks are visible on the wall after excavation. The
permeability of the wall was allowed to be maximum 10-8 m/s after 28 days.
One main reason for the decision to use the CSM method on this site was that it was
possible to produce soil mixing walls with a relatively smooth surface. Hence it is possible to use
the produced walls as permanent face concrete walls without costly wall surface cutting.
Soil conditions
The subsoil investigation indicated a subsurface profile generally comprising filling to depths of
0.6 m to 1.4 m over quaternary alluvial deposits, which typically comprise fine to medium
grained "marine" sands with podsols natural sand, clayey sand, sandy clay. The permeability of
the natural sands was between 2x10-4 and 5x10-5 m/s. These sediments overlie the Hawkesbury
Sandstone, the surface of which was found in depth between 10 and 19 m. The rock surface was
quite uneven. The determined rock strength varied between extremely low (UCS < 0.6 MPa) and
medium strong (UCS 6 < 20 MPa). For sealing the excavation pit it was necessary to socket the
wall a minimum of 1 m into the sandstone or 2 m into the clay layer.
The Hawkesbury Sandstone, which is also known as Sydney sandstone or Yellowblock,
forms the bedrock for the most areas of the region of Sydney. This sandstone is well-known for
its durability, which is caused by a high to very high quartz content. The quartz content varies
depending on the rock layer and can be up to 95 percent. Two exemplary borehole logs are
shown in Fig. 10.

Fig. 10. Exemplary borehole logs; left for socketing into clay (2 m); right socketing into
Sandstone (1 m).
Soil mix wall construction
During the first stage of this project about 4,000 m² were constructed. Because of the mixing
depth the 1-Phase method was used. The soil mix wall was executed by using the “hard-into-
hard” sequence with an overcut of minimum 20 cm.
The CSM wall was installed using a BCM 5 mixing head mounted on a RTG RG 19 T base
carrier equipped with a round shaped Kelly bar, see Fig 11.
Fig. 11. RTG RG 19 T and excavator for pre-excavation and backflow handling.

For the slurry supply a MAT SCC-20 colloidal batch mixer with a maximum mixing capacity of
20 m³ per hour was used. The mixer was assembled with a water tank and a cement silo. The
cement slurry was delivered to the CSM by an eccentric screw pump. This pump was remote
controlled by the rig operator. For the pre-excavation and backflow handling an excavator was
used. The pre-excavation was about one meter deep and also one meter wide. After finishing the
pre excavation the soil mixing process was started. During installation of the first panel it could
be seen that the used teeth were not the right choice for socketing into the sandstone. The
penetration speed into the sandstone was about 3 cm per minute. The reason was that the
sandstone was much harder than expected. Therefore it was decided to change to newly
developed teeth named BAUER SB 38 HR DC, shown in Fig. 12.
Fig. 12. Cutting tooth BAUER SB 38 HR DC.
This tooth has carbide cutting edges on both sides. It can cut the sandstone much more
aggressively and allows a change of direction of mixing wheel rotation without increased tooth
wear. By using the BAUER SB 38 HR DC teeth the penetration speed in the sand could be
improved by about 28 % and the penetration speed into the sandstone by about 143 %.
ø 7 cm/min
Sandstone 10.2 m
Sand;
ø 22 cm/min partly silty,
medium
ø 80 cm/min dense –
extremely
dense,
partly
cemented
0.0 m
Fig. 13. Exemplary depth-time-diagram with average mixing speeds and the associated soil
profile for a panel with 1 m socketing into rock.

Fig. 13 shows an example of a depth-time diagram for a panel with 1 m socketing into the
Sandstone after changing the tooth to BAUER SB 38 HR DC. The achieved average penetration
speed for panels with socketing into clay was about 23 cm/min. The average extraction speed
was about 65 cm/min at a mixing depth of 16 m. Up to 6 panels per working day (10 hours) were
executed. The injected cement amounted was about 450 kg/m³ mixed soil. Immediately after
finishing the mixing process, two steel beams were installed per panel.
For quality control, cores were drilled out of the CSM wall. The laboratory tests showed
unconfined compressive strengths in a range of 5-10 MPa (depending on the different soil
layers). The results for the permeability were between 1.5x 10-9 m/s and 7.7x10-10 m/s. No
leakages could be seen when the pit was excavated. Furthermore the wall surface was as smooth
as expected, thus no further treatment of the wall surface was required.
2.2.4 Conclusion for CSM Application
The CSM technique offers a great diversity of possible applications, such as cut-off walls,
structural retaining walls, foundation elements and numerous others. The capacity to reach big
depths offers an enormous potential for the construction of deep walls. The described example
demonstrated that this construction technique, which combines the advantages of the cutter and
the soil mixing technique, can be used in soil conditions where common mixing methods are at
its limit. Even in these soil conditions the CSM elements showed a very high quality. Thus, the
CSM method is a useful supplement to traditional soil mixing methods and an interesting
alternative for the use as retaining wall.
3. SUMMARY
Looking to recent developments it can be seen that there is an ongoing process to move the
borderlines for the use of retaining wall systems, e.g. in terms of cost efficiency, execution
quality, depth and soil limitations. Obviously, there are also limitations for these new systems,
but they are offering advantages and therefore alternative solutions resulting in time and cost
saving. Obviously, still all factors must be taken into consideration to achieve the best solution in
terms of feasibility and cost efficiency, but the new developments provide more alternatives in
the future.

Recent Developments in Helical Piles
Lutenegger, A.J.(1)
(1)
University of Massachusetts, Amherst, Ma., USA <lutenegg@ecs.umass.edu>
ABSTRACT. Helical piles and anchors are factory manufactured deep foundation and anchor
elements that are provided as segmental modular units and assembled and installed in the field.
The shape and size of the central shaft, the number, size and spacing of the helical plates vary
depending on the soil and load requirements of the project. Even though helical piles and anchors
were introduced over 170 years ago into the civil engineering profession, they have only been
used in the modern era for about the last twenty years to any significant degree. Currently, the
application and use of this technology is one of the fastest growing markets in deep foundation
and anchor work in the world. A summary of the recent technological developments related to
helical piles and helical anchors is presented.
1. INTRODUCTION
The 2015 International Building Code, Section 1802.1, defines a Helical Pile as:
“Manufactured steel deep foundation element consisting of a central shaft and

one or more helical bearing plates. A helical pile is installed by rotating it into
the ground. Each helical bearing plate is formed into a screw thread with a
uniform defined pitch.”
This technology is not new. In the mid 1800’s, helical piles (“screw-piles”) were introduced by a
blind, self-taught Irish engineer, Alexander Mitchell, as an effective method for mooring ships
and supporting iron lighthouses in shallow offshore environments (Mitchell 1849). As a result of
these successes, the use of helical foundations and anchors grew rapidly and were used
throughout the world on large-scale construction, especially to support oceanfront shipping and
pleasure piers and railway bridges (Lutenegger 2011a). Most consisted of a single helical plate
with a solid round shaft or were fabricated as a “screw-cylinder” with a shaft diameter of 2 to 4
ft. in diameter. A decline in the use of helical piles occurred around the beginning of the 20th
century coinciding with the development of other deep foundation methods and installation
equipment, especially pipe driving equipment.
During the 2nd World War and shortly after, helical piles and screw-cylinders were
“rediscovered” as a viable deep foundation technology for the rapid reconstruction of large
wharves damaged or destroyed during the war. In the 1950’s, the expansion of rural electrical
transmission lines in the U.S. resulted in extensive used of helical anchors to support
transmission towers and utility poles from overturning. The current use of helical piles and
anchors now has impacted nearly every segment of civil construction around the world. As with
all foundations, the design of helical piles and anchors must be based on sound principles of soil
mechanics and field observations and to some degree the experience of the Engineer with local
ground conditions. In modern times, helical piles and anchors have been “reinvented”; steel

sections vs. cast and wrought iron; installation using hydraulic torque heads vs. hand installation;
round-shaft and square-shaft single-helix and multi-helix. They represent one of the fastest
growing markets in the foundation industry. This paper presents the Author’s view on recent
developments that have occurred in the manufacture, installation and design of helical piles and
anchors in that past several years.
Figure 1 gives an organizational chart showing the current primary uses of helical piles
and anchors. The applications are about equally split between applications for compression and
tension loading. Figure 2 shows the different categories of helical piles/anchors based on the
various geometries that are currently being manufactured and used in practice. Clemence and
Lutenegger (2015) presented results of an industry survey conducted on the current state-of-
practice of helical piles and anchors. One of the primary results of the survey indicated that there
is often a lack of geotechnical information on sites where helical piles are proposed for use. The
results also indicated that there was increasing use of larger diameter round-shaft helical piles to
develop larger axial capacities.
Fig. 1. Common uses of helical piles and anchors.
Fig. 2. Categories of helical piles and anchors based on geometry.
Most helical piles consist of a lead section with helical plates and extension sections that
are usually just additional lengths of shaft to allow the installation to extend to any desired depth.
The length of lead sections and extension sections vary depending on the needs of the project and
the availability of the individual components. Lead sections with lengths from 5 to 20 ft. (1.5 to
6.1 m) are typical. Like other deep foundation systems, the behavior of the foundation depends
on the quality of the installation provided by the Contractor. Helical piles and anchors are a
manufactured modular foundation/anchor system with predetermined dimensions produced at the
factory.

Manufacturers of helical piles and anchors typically use a fixed pitch helix for their
individual products. The helix is to be formed as a continuous spiral or section of a true
Archimedean screw. When more than one helical plate is attached to the shaft, the spacing is set
as some multiple of the pitch, so that under ideal installation conditions, each helical plate will
follow in the same path as the preceding plate. This is called “tracking” and is considered the
best situation for installation. In this way, an ideal installation advances as a “corkscrew” rather
than as an “auger”. The helical plate-(s) of a helical pile or helical anchor serves two primary
purposes; 1) to facilitate the installation of the pile/anchor, i.e., installation by rotation rather than
by driving or vibration; and 2) provide a component of end bearing to add to the total capacity of
the pile/anchor in both compression and tension.
RECENT DEVELOPMENTS IN INSTALLATION
Traditionally, helical piles have been installed using conventional construction equipment, such
as track excavators and skid steers, equipped with a high torque, low speed hydraulic head. In
recent years, there has been a growing trend to use the same equipment but fitted with a fixed
mast to be able to more carefully control the installation, maintain alignment, and reduce wobble
of the shaft. This greatly improves the quality of the installation and provides better contact
between the shaft and soil for round shaft helical piles.
Simultaneously, there is movement to provide continuous installation monitoring of
helical piles, recording the depth, torque, rotation speed, crown force, and rate of advance. This
is a natural development in line with the technology currently used for monitoring ACIP and
ACIPD piles. Both of these changes have the potential to make significant improvements in the
use of helical piles and provide Engineers with additional confidence and quality control on the
installation and behavior.
2. RECENT DEVELOPMENTS IN DESIGN PROCEDURES
2.1 Recognition of Installation Disturbance
The installation of nearly all deep foundations produces some level of soil disturbance. The
degree of disturbance and influence on behavior depends on the type and geometry of the pile,
the method of installation and the soil conditions. Most manufacturer’s literature suggests that
helical piles produce little disturbance and generally does not account for disturbance in design
procedures. The degree of disturbance to the soil during installation may be especially important
for saturated fine-grained soils where the disturbance from installation may produce a reduction
of the undrained shear strength.

Fig. 3. Disturbed soil zones beneath a helical pile in tension and compression.
Lutenegger et al. (2014) presented direct evidence that the installation of Helical Piles in
clay produces disturbance to the clay and a reduction in shear strength. Field vane tests
conducted directly over the top of the path of the helical plate showed a reduction in undrained
shear strength as compared to the undisturbed soil. Moreover, the results showed that additional
helices along the shaft produced additional soil disturbance over a single-helix pile. The
measured undrained shear strength after installation was not the fully remolded value but was
lower than the peak value, closer to post-peak strength. Additional results also demonstrated that
poor installation by the Contractor could produce more disturbance than an ideal or “perfect”
installation in which the advance of the pile was carefully controlled according to the pitch of the
plate.
Figure 4 shows results of field vane tests were conducted over the top of two round shaft
(RS) open end Helical Piles and a square shaft (SS) Helical Pile with a single 14 in. (356 mm)
helical plate with a thickness on 3/8 in. (9.5 mm) and a pitch of 3 in. (76 mm) attached at the
end. The reduction in undrained shear strength is not uniform but on average is about 75% of the
undisturbed values. It can also be seen that there was a large reduction in strength indicated for
the 2.875 in. (73 mm) round shaft pile between depths of about 5 to 7 ft. (1.5 to 2.1 m) with the
undrained shear strength approaching the remolded value. This would indicate excessive
disturbance and almost complete remolding of the soil.

0
RS 2875
RS4.5
1 SS5
Depth (ft.)
5
10
0 1000 2000 3000 4000 5000 6000
Undrained Shear Strength (psf)
Fig. 4. Results of field vane tests over single-helix round shaft and square shaft helical piles.
(Note 1000 psf = 48 kPa)
0 0
RS2875
1 RS450 1
SS5
2 2
3 3
RS2875
4 4 RS450
Depth (ft.)
Depth (ft.)
SS5
5 5
6 6
7 7
8 8
9 9
10 10
0 500 1000 1500 2000 2500 3000 3 4 5 6 7 8 9
Torque (ft.-lbs.) No. of Revolutions/ft.
Fig.5. Installation torque and advance for single-helix round shaft and square shaft helical piles.
(Note: 1000 ft.-lbs. = 1356 N-m)

Figure 5 shows that the installation torque on the round shaft piles generally increases
with depth as more of the pipe shaft is engaged with the soil as the pile advances. The
installation torque is expected to be higher for round shaft piles as compared to a square shaft
piles with the same size helical plate. However, as noted in Figure 5, the 2.875 in. (73 mm)
round shaft pile shows an increase in the number of helical plate rotations needed to advance the
pile starting at a depth of about 5 ft. (1.5m). That is, the anchor is not advancing at a rate of 3
in.(76 mm) per revolution (equal to the pitch) but instead is requiring up to 8 revolutions per foot
to complete the installation. This means that the helical plate is “churning” the soil rather than
advancing. This increased rotation produces additional remolding of the soil and a corresponding
decrease in undrained shear strength.
The results presented in Figure 5 suggest that there is a link between the quality of the
installation and the reduction of undrained shear strength resulting from rotation and installation
of the piles. Even the best installation produces a reduction in shear strength. The degree of
reduction in strength is complex, but is clearly related to the advance rate and the soil conditions
through which the helix is passing, in this case Sensitivity of the clay.
Figure 6 shows results of vane tests conducted immediately above a single-helix, double-
helix and triple-helix pile with 12 in. (305 mm) diameter plates. These results indicate that
double-helix and triple-helix piles produce more disturbance than a single helix. This means that
in clays additional plates with likely produce more soil disturbance and a greater reduction in
strength, producing a pile which has a combined plate efficiency of less than 100%. Where the
soil strength between or above the plates governs capacity, the mobilized strength available to
develop axial capacity decreases progressively with additional plates.
6
Depth (ft.)
10
11
12 FV1
FV2
13 SS5-12
SS5-12/12
14
SS5-12/12/12
15
0 1000 2000 3000 4000 5000 6000
Undrained Shear Strength (psf)
Fig. 6. Undrained shear strength over single-helix and multi-helix piles.

(Note 1000 psf = 48 kPa)

Once it is recognized that installation of helical piles in clay produces disturbance and a
reduction of available strength, it is important to understand how installation disturbance may
produce a reduction in capacity. In order to illustrate this, the load test results from two 2.875 in.
(73 mm) round shaft helical piles were compared with an adjacent identical pile in which the
quality of installation was considered better. Figure 7 shows a comparison of the installation
torque and the advance for the two piles. Initially, the installation torque is the same, but after a
depth of about 5 ft. (1.5 m) it can be seen that the torque developed by the two piles start to
diverge. This is the result of the larger number of rotations required for the SCG pile as
compared to the P pile. As the number of rotations increases, the torque decreases as the soil is
remolded.
0 0
RS2875 SCG
1 RS2875 P
1
2 2
3 3
RS2875 SCG
4 4
Depth (ft.)
RS2875 P
Depth (ft.)
5 5
6 6
7 7
8 8
9 9
10 10
0 500 1000 1500 2000 2500 3000 3 4 5 6 7 8 9
No. Revolutions per ft.

Torque (ft.-lbs.)
Fig. 7. Installation torque and advance for two 2.875 in. (73 mm) round shaft helical piles.
(Note: 1000 ft.-lbs. = 1356 N-m)
Results of tension tests performed on these two piles are shown in Figure 8 and demonstrate the
influence of disturbance on capacity. The additional disturbance produces a lower stiffness and
lower capacity.

25000
20000
Load (lbs.)
15000
10000
5000 RS2875 SCG

RS2875 P
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Displacement (in.)
Fig. 8. Load-displacement curves between two 2.875 in. (73 mm) round shaft helical piles.
(Note: 1000 lbs = 4.5 kN; 1 in. = 25.4 mm)
In 1950 Skempton presented a valuable discussion to the paper of Wilson (1950) on the
behavior of screw-piles in clays (Skempton 1950) noting that for multi-helix screw-piles it was
important to recognize that the clay beneath the upper screws had been remolded by the passage
of the first screw. However, Skempton (1950) further noted that all of the clay contributing to the
bearing capacity of the upper screws would not be fully remolded and, as a rough approximation,
suggested that it might be reasonable to assume that the average shear strength of the clay would
be equal to:
cp2 = c – [½(c – cr)] (1)
where:
cp2 = operational undrained shear strength

c = peak undrained shear strength
cr = remolded undrained shear strength
This observation and suggestion by Skempton brings into importance the Sensitivity of the clay
in considering installation disturbance.
Lutenegger et al. (2014) defined the Installation Disturbance Factor, IDF, as the ratio of
actual measured installation to the ideal or “perfect” installation:
IDF = (R)/(A/P) (2)
where:
R = measured number of revolutions per unit of advance

A = ideal number of revolutions per unit of advance
P = Pitch of helical plate

For example, for ideal or “perfect” installation of helical plates with a pitch of 3 in. (76 mm) the
value of IDF would be equal to 1. That is, 4 revolutions to advance 1 ft. Values of IDF should be
close to 1 for a high quality installation but may be as high as 4 or 5 if poor quality is
experienced during the installation. Figure 9 shows the IDF for the two 2.875 in. (73 mm) round
shaft anchors previously presented in Figure 7.
Lutenegger et al. (2014) suggested that conceptually, one way to quantify the effect of
installation disturbance may be to consider the available undrained shear strength after
installation in the context of the initial undisturbed undrained shear strength. That is, what
percentage of the undisturbed shear strength is still available for developing axial capacity after
installation? Logically, as the Installation Disturbance Factor increases the soil becomes more
disturbed and remolded and less strength is available. Also, for two soils with different
Sensitivity (low vs. high) equal Installation Disturbance Factors will produce different degrees of
disturbance and reduction of strength.
3
RS2875 SCG
4 RS2875 P
Depth (ft.)
10
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Installation Disturbance Factor
Fig. 9. Installation Disturbance Factors for two round shaft Helical Piles.
2.2. Recognition of Reduction in Efficiency of Multi-Helix Piles
Most manufacturer’s literature suggests that the capacity of multi-helix piles/anchors may be
obtained as a direct multiplier of the capacity of a single-helix pile/anchor. Detailed test results in
both clay and sands now show that this is not true, That is, the Efficiency of co-linear multi-helix
piles/anchors is less than 100%, which is likely related to disturbance. Efficiency, E, of a multi-
helix pile/anchor may be defined as:
E = [QultMH/(n x QultSH)] x 100% (3)

where:
QultMH = capacity of a multi-helix pile/anchor

QultSH = capacity of a single-helix pile/anchor
n = number of helical plates
Indirect evidence of Efficiency can be seen by comparing the installation torque and the
load-displacement behavior of single-helix and multi-helix piles. The attachment of additional
helical plates behind a lead plate is used to increase the load capacity of the pile/anchor, however
most tests show that this increase is not directly proportional to the number of helices or the
increase plate area. That is, a double-helix pile does not provide twice the capacity of a single-
helix anchor with the same diameter helical plate nor does a triple-helix anchor provide three
times the capacity of a single helix anchor. Figure 10 shows a comparison of the installation
torque measured for single and multi-helix piles. One might expect that in a uniform soil the
torque may be directly proportional to the number of helices for a central shaft with a fixed
diameter. In this case it can be seen that the torque does not increase proportionally with the
increased number of helices and suggests Efficiency less than 100%.
10
Depth (ft.)
12
14
16
RS2875-12
18 RS2875-12/12
RS2875-12/12/12
20
22
24
0 1000 2000 3000 4000
Torque (ft.-lbs.)
Fig. 10. Installation torque of single-, double-, and triple-helix piles in clay.
(Note: 1000 ft.-lbs. = 1356 N-m)
So it might be argued that the degree of disturbance increases as each successive plate passes
through the same soil as previously indicated in Figure 3. That is, the capacity of each successive
plate is reduced progressively by the preceding disturbance so that while the 1st plate may have
100% Efficiency, the second may have only 60 to 80% Efficiency and the 3rd 40 to 60%
efficiency, etc. depending on the soil. Figure 11 presents results from five different sets of tests

in medium stiff clay with helix diameters ranging from 8 in. to 15 in. diameter showing the
reduction in Efficiency with additional helical plates. It should be noted that similar results have
been obtained in sands (e.g., Clemence et al. 1994, Lutenegger 2011c, Tshua et al. 2012).
100
SS5-12
90 SS5-12
RS2875-10
80 RS2875-12
RS350-12
Trend
Efficiency (%)
70
60
50
40
30
20
1 2 3 4
Number of Helices
Fig. 11. Efficiency of multi-helix anchors in clay.
2.3. Understanding Behavior of Shallow Round Shaft Helical Piles
For the last 10 years there has been a growing trend for using larger diameter (6 in. to 16 in. (152
to 406 mm) diameter) round shaft helical piles to support higher loads, such as from bridges.
These piles typically have either one or two large diameter helical plates, about 16 in. to 36 in.
(406 mm to 914 mm) in diameter. The behavior of these piles may be very different than the
behavior of smaller helical piles and, therefore, the design procedure must accordingly be
different. The pipe shaft is usually open-ended with a wall thickness and strength of steel
sufficiently large to withstand the torsional forces applied during installation. So, in some
respects, a round-shaft helical pile is simply a pipe pile with a helical plate attached to the base
that is installed by rotation rather than driving.
Depending on the length of the shaft and the type of connection between additional pipe
sections if any, the contribution of capacity between the shaft and helical plates may vary. In
some cases, it is possible that the helical plates only act as a construction technique to install
what is otherwise a traditional pipe pile, i.e., installation by rotation rather than installation by
driving or vibration. The installation progresses because the pitch of the helical plate advances
during rotation and “pulls” the pipe pile downward as it advances. When the shaft resistance
developed along the outside of the pipe exceeds the capacity of the helical plate, advance will
stop. In some cases, multiple helices are used to help advance the pile to deeper depths to
develop higher capacity. Because of the similarities, the design approach for helical pipe piles
should be similar to the traditional design approaches currently being used for driven piles in
either sand or clay.

Figure 12 shows the results of three helical pile tests conducted on different diameter
round shaft helical piles all equipped with the same diameter helical plate (12 in. (305 mm))
installed to a depth of 8 ft. (2.4 m) in sand. These data clearly show that the uplift capacity
increases as the pipe diameter increases and illustrate the importance of the pipe shaft to overall
behavior. The capacity increases even though the effective area of the helix decreases as the pipe
shaft increases. A comparison between a plain pipe and a single-helix pile embedded to the same
depth is shown in Figure 13. Note that in tension the failure of the plain pipe shaft is very abrupt
which was characteristic of all tests. By contrast, the load-displacement curve obtained from the
screw-pile shows a more gradual curve, even up to displacements of 20% of the helix diameter.
Using both load-displacement curves it is possible to approximate the load distribution or
contribution of the pipe and helix to the total capacity. As the diameter of the pipe shaft
increases for a constant diameter helix, it should be expected that the distribution of load
between the shaft and helix changes and more percentage of the load is carried by the pipe shaft
and less by the helix.
The total uplift Capacity, QT, is the sum of the capacity of the Helix, QUH, plus the
capacity of the shaft, QS. The contribution of the pipe and helix may be evaluated at different
relative load levels depending on the choice of definition of capacity. If the capacity is defined as
the load producing a displacement of 10% of the helix diameter, as suggested by the
International Code Council (ICC) AC 358, (in this case 1.2 in. (30.5 mm)) then the capacity is
17,750 lbs.(80 kN) and the load carried by the pipe shaft is 3,500 lbs. (15.6 kN) while the helix
now only carries 14,250 lbs. (63.4 kN).
30000
25000
Uplift Load (lbs)
20000
15000
10000
2.875 in. Pipe with 12 in. Helix

5000 4.5 in. Pipe with 12 in. Helix
6.625 in. Pipe with 12 in. Helix
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Displacement (in.)
Fig. 12. Results of tension tests on helical piles with different pipe sizes in sand.
(Note: 1000 lbs = 4.5 kN; 1 in. = 25.4 mm)

30000
4.5 in. x 8 ft. Plain Pipe

25000
4.5 in. x 8 ft. Pipe with 12 in. Helix
Uplift Load (lbs.) 20000
15000
10000
5000
0
0.0 0.5 1.0 1.5 2.0 2.5
Displacement (in.)
Fig. 13. Comparison between uplift behavior of plain pipe pile and helical pile in sand.
(Note: 1000 lbs = 4.5 kN; 1 in. = 25.4 mm)
If the load distribution is evaluated at an even lower load level, corresponding to a more
realistic working or serviceability load condition, it might be convenient to apply a global Factor
of Safety of 2 to the 10% load, in this case giving a safe working load of 8,675 lbs. (35.6 kN) and
the distribution of load between the shaft and helix now becomes 3,300 lbs. (14.7 kN) carried by
the pipe shaft and 5,375 lbs. (23.9 kN) carried by the helix; i.e., 38% shaft, 62% helix. As the
trend toward the use of larger diameter pipes continues, a smaller portion of the load is
transferred to the helix.
Figure 14 shows the results of three helical piles tests conducted on different diameter
round shaft piles with the same diameter helical plate (12 in. (305 mm)) installed in clay. These
results again clearly show that capacity increases as pipe diameter increases and illustrate the
importance of the pipe shaft to overall behavior. Results may also be expressed in terms of the
ratio of helix diameter to pipe diameter, DH/DP to illustrate this behavior.
Figure 15 shows a similar comparison to Figure 4 between a helical pile and plain pipe
pile in stiff clay. The results obtained for both sands and clays are similar. For the same
geometry piles installed to the same depth in different soils, the distribution of load varies
considerably and in this case the shaft diameter is more important in clay than in sand. Both the
pipe shaft and the helical plate contribute to capacity but in different ways.
3. CORRELATING INSTALLATION TORQUE TO LOAD CAPACITY
There is a long history in the profession of using the installation torque to estimate the axial
capacity of Helical Piles and Anchors, beginning as early as the late 1800s, through a Torque
Factor, KT. The basic premise behind such correlations is that both the axial capacity, Qult, and
installation torque, T, are a function of the specific geometry of the pile/anchor and soil
properties, i.e., soil strength and therefore:

Qult = TKT (4)
28000
24000
20000
Uplift Load (lbs.)
16000
12000
8000 2.875 in. x 8 ft. Pipe with 12 in. Helix

4.5 in. x 8 ft. Pipe with 12 in. Helix
4000 6.625 in. x 8 ft. Pipe with 12 in. Helix
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Displacment (in.)
Fig. 14. Tension tests on helical piles with different pipe sizes in clay.
(Note: 1000 lbs = 4.5 kN; 1 in. = 25.4 mm)
20000
16000
Uplift Load (lbs.)
12000
8000
4000 4.5 in. x 8 ft. Plain Pile

4.5 in. Pipe x 8 ft. with 12 in. Helix
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Displacement (in.)
Fig. 15. Comparison between uplift behavior of plain pipe pile and helical pile in clay.
(Note: 1000 lbs = 4.5 kN; 1 in. = 25.4 mm)

The expectation of the Engineer is that as installation torque increases, axial capacity increases
proportionally. Many reported comparisons between installation torque and axial capacity show
wide variations. For example, Hoyt and Clemence (1989) compared results of a large number of
field tension tests in different soils and found that the accuracy between observed and calculated
values (ratio of measured to computed capacity) ranged from about 0.3 to 4.5, suggesting
considerable scatter in the accuracy of any individual value of KT. Despite this high variability,
some of these values of KT still appear in manufacturer’s literature.
Unfortunately, both the installation torque and the axial capacity may be influenced by a
number of factors that can make any correlation between torque and capacity tenuous as
summarized by Lutenegger (2013) and given in Table 1. An important factor that can influence
the measured installation torque but have little to no effect on the axial capacity is the rate of
installation or rate of rotation. This suggests that installation rate should to be monitored on
projects. Also, at present, there is no single method being used to measure installation torque,
some of which give erroneous data.
Table 1. Factors influencing Torque-to-Capacity correlations of helical piles and anchors
Factors Affecting Installation Torque Factors Affecting Axial Capacity
Pile/Anchor Soil Factors Contractor Pile/Anchor Soil Factors Load Test

Factors Factors Factors Factors
Plate Diameter Soil Type Rotation Rate Plate Diameter Soil Type Loading Rate
Number of Soil Strength Advance Rate Number of Plates Soil Strength (Increment Duration
Plates Soil Stiffness Down Force Plate Spacing Soil Stiffness & No. of
Plate Thickness Sensitivity Inclination Shaft Shape Sensitivity Increments)
Plate Pitch Water Table Connection Style Changes in Loading Direction
Shaft Shape & Water Table Waiting Time
Diameter Between
Connection Installation &
Style Testing (Aging)
Surface Texture Load Test
(Roughness) Interpretation for
Qult
Manufacturers of helical piles and anchors often suggest different values of KT for
different size round shaft elements. A simple analysis of the undrained behavior of plain pipe
piles installed by rotation in uniform clay shows that the value of KT for a plain pipe is only a
function of pipe diameter. Naturally, the addition of a helical plate or multiple plates to the pipe
shaft and the direction of loading complicate this relationship. However, this approach simply
illustrates how one factor, shaft diameter, may affect a torque-to-capacity correlation. The actual
correlation will be very complex since both shaft diameter and helix diameter may vary and
different combinations of shaft and helix diameter will affect the true distribution of components
in developing load capacity in different soils.
A further complication may be that installation torque, T, is measured during installation
while Qult is measured at some time after installation. This delay time allows the soil to adjust
(pore water pressure to dissipate and soil to regain strength) to the installation effects. Some
variation in Torque correlations may also develop if non-uniform soils are present and the shaft
is embedded in one soil layer while the helix is embedded in another type of soil. It is clearly

expected that both T and Qult will increase with increasing size and number of helices, but not
necessarily at the same rate.
Another complicating factor is that different definitions of axial capacity have been used
in the literature in developing torque-to-capacity correlations. This can lead to variable
correlations. The Author recommends that a single method be used to interpret load test results
and that failure load be defined as the load corresponding to a gross displacement equal to 10%
of the average helix diameter to be consistent with ICC AC358.
4. DEVELOPMENT OF HELICAL CAST-IN-PLACE DISPLACEMENT PILES

(HCIPD)
An alternative to using a plain steel central shaft that is popular in many areas is to surround the
shaft with Portland cement grout as the installation progresses. This is especially attractive for
square shaft piles/anchors since a small void is created between the pile shaft and the soil during
rotation that the grout can fill. The addition of the grout column enhances the stiffness of the
shaft and increases axial stiffness under compression loading (Vickars and Clemence 2000).
However, an additional benefit of the grout column is to enhance the capacity by the shaft
resistance that develops between the grout column and the surrounding soil in much the same
way that an auger cast pile or grouted soil anchor works (Lutenegger 2011b).
The contribution of the shaft to the capacity depends on the geometry of the Helical Pile
(lead helix section + grout column), the ground conditions into which it is installed, and the
working load (Lutenegger 2015). Helical Piles with small diameter grouted shafts, generally less
than about 6 in (150 mm), are often referred to as Grouted Shaft Helical Micropiles (GSHM).
Helical piles with grouted shafts larger than about 6 in. (150 mm) are referred to as Helical Cast-
in-Place Displacement Piles (HCIPDP). Grouted shaft helical piles are a unique type of
foundation that are constructed using a lead single- or multi-helix pile section and soil
displacement plates attached along the central square shaft. A cylindrical cavity is created around
the central steel shaft by the soil displacement plates which displace the soil laterally during
rotation of the shaft and plates as the installation proceeds. This is similar to the installation of
Auger-Cast-In-Place Displacement Piles (ACIPD) but on a smaller scale. The cavity is
continuously filled with cement grout from a small grout reservoir located at the ground surface
as the lead helical section advances into the soil.
Like other types of cast-in-place displacement piles, no soil cuttings are brought to the
surface so that site cleanup is minimal. Since the installation torque is continuously monitored as
the lead section advances, an indication of the soil conditions is also obtained at every pile
location. In addition to recording the depth and torque, the Grout Take is also measured and
recorded during pile installation to monitor the volume of grout along each section.
5. SUMMARY
Helical Piles offer the Geotechnical Engineer a number of potential advantages for projects,
including: rapid installation; immediate load capacity; installation in high groundwater
conditions; installation with traditional equipment; no soil cuttings; minimal site disruption and
cleanup; installation in limited access & low headroom areas; installation monitoring;
removable/reusable; installation at any orientation; minimal disruption to adjacent structures;
pre-screening of soil conditions at each location; minimal installation noise and vibration; wide
range of soil applicability and rapid field modifications.

Recent developments relate mostly to more appropriate design methods recognizing
installation disturbance and reduced efficiency of multi-helix piles. Advances in more controlled
installation equipment and continuous installation monitoring will improve the installation
quality and provide Engineers with more confidence in use.
6. REFERENCES
Clemence, S.P., Crouch, L.K. and Stephenson, R.W., 1994. Prediction of uplift capacity for
helical anchors in sand. Proceedings of the 2nd Geotechnical Engineering Conference, Cairo
University, 332-343.
Clemence, S.P. and Lutenegger, A.J., 2015. Industry survey of state-of-practice for helical piles
and tiebacks. Journal of the Deep Foundations Institute, 9(1): 21-41.
Hoyt, R. and Clemence, S.P., 1989. Uplift capacity of helical anchors in soil. Proceedings of the
12th International Conference on Soil Mechanics and Foundation Engineering, 2: 1019-1022.
Lutenegger, A.J., 2011a. Historical development of iron screw-pile foundations: 1836–1900.
International Journal for the History of Engineering and Technology, 81(1):108-128.
Lutenegger, A.J., 2011b. Behavior of grouted shaft helical anchors in clay. Journal of the Deep
Foundations Institute, 5(1): 63-74.
Lutenegger, A.J., 2011c. Behavior of multi-helix screw anchors in sand. Proceedings of the 14th
Pan-American Conference on Geotechnical Engineering, Toronto, Canada.
Lutenegger, A.J., 2013. Factors affecting torque correlations for screw-piles and helical anchors.
Proceedings of the 1st International Geotechnical Symposium on Helical Foundations, 211-
224.
Lutenegger, A.J., 2015. Load tests on grouted shaft helical micropiles in some U.K. soils.
Proceedings of the 16th European Conference on Geotechnical Engineering.
Lutenegger, A.J., Erikson, J. and Williams, N., 2014. Evaluating installation disturbance of
helical anchors in clay from field vane tests. Proceedings of the 39th Annual Deep
Foundations Institute Conference.
Mitchell, A., 1849. On submarine foundations: Particularly screw-pile and moorings. Civil
Engineers and Architects Journal, 12: 35-40.
Skempton, A.W., 1950. discussion of The bearing capacity of screw piles and screwcrete
cylinders. Journal of the Institution of Civil Engineers – London, 34(5): 76-81.
Tshua, C.H.C., Aoki, N., Rault, G., Thorel, L. and Garnier, J., 2012. Evaluation of the
efficiencies of helical anchor plates in sand by centrifuge model tests. Canadian Geotechnical
Journal, 49: 1102-1114.
Vickars, R. and Clemence, S.P., Vickars, R.A. & Clemence, S.P., 2000. Performance of helical
piles with grouted shafts. New Technological and Design Developments in Deep
Foundations, ASCE, 327-341.
Wilson, G., 1950. The bearing capacity of screw piles and screwcrete cylinders. Journal of the
Institution of Civil Engineers – London, 34(5): 4-73.

Simple approach to static and seismic design of piled rafts
Mandolini, A (1), Di Laora, R.(2) and Iodice, C.(3)

(1)
Università degli Studi della Campania “Luigi Vanvitelli”, Italy
alessandro.mandolini@unina2.it
(2)
Università degli Studi della Campania “Luigi Vanvitelli”, Italy raffaele.dilaora@unina2.it
(3)
Università degli Studi della Campania “Luigi Vanvitelli”, Italy chiara.iodice@unina2.it
ABSTRACT. Current design of piled rafts often neglects a number of aspects well known in
scientific literature, sometimes leading to increased costs without appreciable increase in
performance. This work makes an attempt to partially fill this lack between State of Art and State
of Practice; to this end, the paper briefly recalls available simple design methods regarding both
geotechnical and seismic issues and proposes a novel simple procedure to estimate the load-
settlement curve of a piled raft as well as the load sharing between piles and raft for vertical
centered loads. The method has been validated through comparison with numerical and
experimental data and applied to a case history; it will be shown that the proposed procedure,
despite its simplicity, is able to account for the non-linearity in the soil behavior, the latter being
responsible of progressive variation of the load sharing with increasing applied top load. Further,
regarding seismic issues, simplified formulae from past works are recalled and discussed.
1. INTRODUCTION
From a conceptual standpoint, a piled raft consists of a group of piles surmounted by a raft in
direct contact with the ground. In order to describe the portion of the total load Qpr taken by the
raft (Fig. 1), it is possible to refer to the load sharing ratio αr defined as:
α r = Qr Qpr (1)
where Qr is the load taken by the raft. A load sharing ratio αr = 1 represents a shallow
foundation with no piles, while αr = 0 represents a pile group with a raft not in contact with
ground; piled raft foundations cover the range 0 < αr < 1.
When designing the structure foundations, engineers typically consider as a first option to
adopt a shallow foundation system such as a raft. If this choice does not guarantee a satisfactory
performance, the adoption of piles underneath the shallow foundation is certainly a common
design alternative. Although this rationale cannot be generalized as the design process is the
outcome of a number of aspects, some of them lying beyond the mere technical considerations, it
is indeed evident that piles are generally required to increase bearing capacity and/or reduce
settlement. However, once the decision of adopting piles has been made, it is common
assumption that the design loads are carried solely by the piles, thus neglecting the contribution
of the raft-soil contact. This traditional design approach is clearly conflicting with the collected
experimental evidence (see for example Mandolini et al. 2005). For pile groups with pile at
relatively small spacing/diameter ratio (say s/d = 3÷4) and covering the entire raft area AR
(AG/AR ∼ 1, with AG the area covered by pile group), the percentage of load taken by the raft is

not less than 20% approximately (Fig. 2). The latter value increases up to 60-70% for increasing
s/d or decreasing AG/AR.
Fig. 1. Definition of different foundation systems.
Fig. 2. Percentage of load carried by the raft: experimental data (Mandolini et al., 2005).
The common simplification of neglecting the raft contribution, which is conservative when
piles are required to increase the foundation bearing capacity, becomes totally erroneous when
the raft alone possesses adequate bearing capacity, often larger than the one provided by the pile
group, and piles have the only role of decreasing and/or regulating settlements.
In the last decades, many specialists focused on this field, either from theoretical or
experimental point of view (for a more comprehensive coverage, reference may be made to
Randolph 1994; Poulos et al. 2001; Mandolini 2003). It can be certainly stated that is nowadays

possible to install piles underneath a raft with different ‘geotechnical’ purposes: to increase the
resistance and the overall stiffness of the raft (Capacity and Settlement Based Design approach,
CSBD), to reduce average settlement (Settlement Based Design approach, SBD) or differential
settlement (Differential Settlement Based Design approach, DSBD). It follows that ‘ready for
use’ suggestions are by now available; they will be briefly summarized herein.
2. GEOTECHNICAL DESIGN ISSUES (CBD, CSBD, SBD, DSBD)

Mandolini (2003) proposed a schematic chart for choosing the foundation type and the proper
‘geotechnical’ design approach (Fig. 3). In the figure, the center, A, represents an ideal condition
or optimum for which, under a certain applied vertical load Qpr, an unpiled raft having width BR
and bearing capacity Qr,u, has a global factor of safety FSUR equal to some fixed minimum value
(arbitrarily assumed as 3 in the figure) and, at the same time, experiences an overall
displacement wUR equal to some admissible value (wUR/wadm =1). From a general point of view,
three design situations may occur:
a) both the estimated values of FSUR and wUR are acceptable (Location [1]): the adoption of
an unpiled raft is possible and therefore piles could be required to reduce the state of the
stress and deformation into the raft;
b) both the estimated values of FSUR and wUR are not acceptable (Locations [2] and [3]); piles
have to be added in order to increase the value of FSUR and to reduce the displacement wUR
(CSBD);
c) although the factor of safety is equal (Location [4]) or larger (Location [5]) than the
minimum admissible value, the displacement under working loads is greater than wadm.
Piles have thus to be added with the aim of reducing displacement (SBD and/or DSBD).
10,0
CBD or CSBD SBD o DSBD
[2]
[3]
[4]
[5]
[-]
adm [-]
adm
1,0
/w
UR/w
A (BR = Bopt ) RBD

wUR
w
unrealistic [1]
raft
0,1
0 1 2 3 4 5 6
FSUR
FS =Q
UR = Rr,u
UR//V
QPR [-]
pr [-]
Fig. 3. Chart for selection of the design approach (modified from Mandolini 2003).

3. THE “MISSING LINK” BETWEEN STATE OF ART AND STATE OF PRACTICE
Despite the accumulated theoretical and experimental knowledge on the subject, simplistic over-
conservative assumptions are still employed in design routine. One of the main reasons is
certainly that designers are often reluctant to perform complicated numerical analyses to
investigate the soil-pile-raft interaction problem, which clearly require expertise on the topic and
considerable extra computational cost.
With the aim of filling the lack between State Of Art and State Of Practice, emphasis is given
in this work to simplified yet reliable design procedures. A simplified method, which is able to
take into account also nonlinear behaviour of soil, is proposed to estimate the load-settlement
curve of a piled raft and predict the load sharing between piles and raft for vertical centred load.
Such a method may be easily extended to horizontal loads.
The proposed procedure is an extension of the PDR method (Poulos and Davis 1980;
Randolph 1994; Poulos 2000), which is of particular usefulness to introduce some basic aspects
of the ‘piled raft concept’ and is therefore summarized in the following.
3.1 The PDR method

In its original form, the PDR method is based on the following assumptions: (1) piles and raft
behave as linearly elastic systems until failure; (2) the raft is rigid and subjected to a vertical
central load, hence only a uniform vertical displacement can occur. Due to the hypotheses about
the raft rigidity, in principle, the method is applicable only to small piled rafts (BR/L < 1) for
which CSBD or SBD approaches are of relevance (Fig. 3) and relevant differential settlements
would not occur (Russo and Viggiani 1998). However, if the aim is the assessment of load
sharing between pile group and raft and the estimation of the load-average settlement curve, the
method can be still applied to furnish useful information for design purposes.
Starting from the knowledge of the behavior of pile group and raft considered as individual
systems, the PDR method combines them to predict the behavior of the piled raft system. From a
conceptual standpoint, it is possible to express the settlement of the pile group as the settlement that
piles would experience in absence of the raft plus an interaction settlement due to the presence of
the raft. The same applies for the behavior of the raft in the combined system. In matrix form:
⎡ 1 α pr ⎤
⎢ ⎥
⎡ wp ⎤ ⎢ K p K r ⎥ ⎡Q p ⎤
⎢ ⎥=⎢ ⋅⎢ ⎥ (2)
⎣ wr ⎦ ⎢ α rp 1 ⎥ ⎣Qr ⎦
⎥
⎢⎣ K p K r ⎥⎦
where Qi and Ki are load and stiffness of pile group (subscript “p”) and raft (subscript “r”)
considered as stand-alone systems, while wi is the settlement in the combined foundation; αpr
and αrp are interaction terms related to the influence of raft settlement on the settlement of the
pile group and vice versa, respectively.
By imposing the compatibility condition (wp = wr) and introducing the further hypothesis that
the terms on the secondary diagonal are equal (αrp/Kp = αpr/Kr), one obtains the expressions for
the stiffness of the piled raft and the load sharing:

Kr
1−
Kp
( 2α rp − 1)
K pr = Kp (3)
Kr 2
1− ⋅ α rp
Kp
K
Qp
( 1 − α rp ) r
Kp
βp = = (4)
Q pr K
1 − α rp r
Kp
Note that the last condition should be considered as a further hypothesis of the method, rather
than a consequence of reciprocity, which presents some difficulty to be applied with the aim of
proving the equality of the interaction terms. In addition, the validity of such hypothesis has been
proven numerically only for the capped pile (i.e. a circular raft equipped with an isolated pile) in
Randolph (1983). However, this hypothesis will be accepted also in the proposed method and its
validity for any piled raft will be not discussed further in this work.
Regarding the value of αrp, Clancy and Randolph (1993) performed a number of numerical
analyses, finding out that for square pile groups made by more than 16 piles (4x4 or bigger) at
moderate-to-large spacing (larger than 4 diameters) such a coefficient remains constant with a
value of about 0.8.
Equations (3) and (4) are valid until piles reach failure (Location A in Fig. 4). Beyond this
point and until raft failure, any load increment is transferred directly to the raft and therefore the
settlement is given by:
Q A Q pr - Q A
w= + (5)
K pr Kr
Fig. 4. The PDR method in its original form.
3.2 Proposed method

The method proposed herein employs arbitrary non-linear load settlement curves for both pile
group and raft. For example, tangent stiffnesses may be expressed as:

np
dQ p ⎛ Q ⎞
= K p = K p ( Q p ) = K p ,0 ⎜1 − p ⎟⎟
dwp ⎜ Q
⎝ p ,u ⎠ (6a,b)
nr
dQr ⎛ Q ⎞
= K r = K r ( Qr ) = K r ,0 ⎜1 − r ⎟
dwr ⎝ Qr ,u ⎠
where Kp and Kr are the tangent stiffnesses of pile group and raft (with Kp0 and Kr0 their initial
value), Qp,u and Qr,u the ultimate loads, np and nr real positive numbers regulating the shape of
the curves. It is easy to verify that np [nr] = 0 corresponds to the assumption of elastic behavior
of the piles [raft] until failure.
Equations (6a,b) may be readily integrated to get the load displacement curves:
ni
⎛ Q ⎞
(Qi ,u − Qi ) − Qi ,u ⎜1 − Q i ⎟
wi = ⎝ i ,u ⎠
with i = p, r (7)
ni
⎛ Qi ⎞
K i ,0 ( ni − 1) ⎜1 − ⎟
⎝ Qi ,u ⎠
The above expression is valid for ni ≠ 1. For ni = 1 the expression of the load-settlement curve
is obtained by calculating the limit, so that:
Qi ,u ⎛ Q ⎞
wi = − ln ⎜1 − i ⎟ with i = p, r (8)
Ki ,0 ⎝ Qi ,u ⎠
Note that for ni < 1 the settlement for Qi approaching Qi,u is finite and equal to:
Qi ,u
wi = with i = p, r (9)
Ki ,0 (1 − ni )
Once the load-settlement curves have been defined for both the pile group and the raft, the
proposed method assumes that each load increment on the piled raft is shared between piles and
raft depending on their current stiffnesses. On the contrary, interaction terms always depend on
the initial stiffnesses. This assumption, verified numerically and experimentally (Caputo and
Viggiani 1984) for pile-to-pile interaction, will be adopted for the pile group-raft interaction.
In matrix form, the method assumes:
⎡ 1 α rp ⎤
⎢ ⎥
⎡ dwp ⎤ ⎢ K p K p ,0 ⎥ ⎡ dQ p ⎤
⎢ ⎥=⎢ ⋅⎢ ⎥ (10)
dw
⎣ r ⎦ ⎢ α rp 1 ⎥ ⎣ dQr ⎦
⎥
⎣⎢ K p ,0 K r ⎦⎥

Forces equilibrium and displacements compatibility requires that:
⎧⎪dQpr = dQp + dQr

⎨ (11)
⎪⎩dwpr = dwp = dwr
The stiffness of the piled raft is then obtained as:
K pr =
{
K p ,0 ⋅ K p ⋅ ⎡⎣ K p − ( 2 ⋅α rp ⋅ K r )⎤⎦ + ( K p ,0 ⋅ K r ) } (12)
K p2 ,0 − ⎡⎣α rp2 ⋅ ( K p ⋅ K r )⎤⎦
while the raft/piles load ratio is given by:
dQr K r ⋅ ⎡⎣ K p ,0 − ( α rp ⋅ K p )⎤⎦
β= = (13)
dQp K p ⋅ ⎡ K p ,0 − ( α rp ⋅ K r )⎤
⎣ ⎦
The method may be readily employed in a spreadsheet. Note that although Eqs. (6)-(9) are
considered in this study, the proposed procedure allows to employ any kind of load-settlement
curve, including a piecewise function.
Applying a constant load increment ΔQpr, at the load increment i the procedure involves the
following steps:
1) Calculation of the tangent stiffness Kpr,i of the piled raft and of βi as:
K pr ,i =
{ }
K p ,0 ⋅ K p ,i −1 ⋅ ⎣⎡ K p ,0 − ( 2 ⋅α rp ⋅ K r ,i −1 )⎦⎤ + ( K p ,0 ⋅ K r ,i −1 )
(14)
K 2 p ,0 − ⎡⎣α 2 rp ⋅ ( K p ,i −1 ⋅ K r ,i −1 )⎤⎦
K r ,i −1 ⋅ ⎡⎣ K p ,0 − ( α rp ⋅ K p ,i −1 )⎤⎦
βi = (15)
K p ,i −1 ⋅ ⎡⎣ K p ,0 − ( α rp ⋅ K r ,i −1 )⎤⎦
2) Evaluation of settlement increment Δwpr,i, current settlement wpr,i and load increment
percentages on pile group and raft as:
ΔQ pr
Δw pr ,i = (16)
K pr ,i
w pr ,i = w pr ,i −1 + Δw pr ,i (17)

Δ Q p ,i 1 ΔQr ,i β
= and = i (18)
ΔQ pr 1+ βi ΔQpr 1 + βi
3) Evaluation of load increments ΔQp,i, ΔQr,i and total loads Qp,i, Qr,i on pile group and raft as:
1 β
ΔQ p ,i = ⋅ ΔQ pr and ΔQr ,i = ⋅ ΔQ pr (19)
1+ βi 1 + βi
Q p ,i = Q p ,i −1 + ΔQ p ,i and Qr ,i = Qr ,i −1 + ΔQr ,i (20)
4) Calculation of the total load Qpr,i on piled raft and of the total load percentages αp,i αr,i
carried by pile group and raft as:
Q pr ,i = Q p ,i + Qr ,i (21)
Q p ,i Qr ,i
α p ,i = and α r ,i = (22)
Q pr ,i Q pr ,i

5) Adjournment of current raft and pile group stiffnesses through the functions selected to
describe the evolution of stiffness with load. If these functions obey to Eqs. (6a,b), new
values of piles and raft stiffness are:

np nr
⎛ Qp,i ⎞ ⎛ Qr ,i ⎞
K p,i = K p,0 ⎜1 − ⎟ and K r ,i = K r ,0 ⎜ 1 − ⎟ (23)
⎜ Q ⎟
⎝ p,u ⎠ ⎝ Qr ,u ⎠
The above steps are repeated for a number of load increments equal to Qpr,u/ΔQpr.
3.3 Validation of the method

The method is here validated through comparison with numerical results obtained by two
different studies on piled rafts, namely the contributions by de Sanctis and Mandolini (2006) and
Comodromos et al. (2009), and with centrifuge results reported in Fioravante and Giretti (2010).
An application to a well-documented case history is also reported.

3.3.1 Comparison with de Sanctis and Mandolini (2006)
The authors performed a number of Finite Element Method analyses aimed at investigating the
behavior of piled rafts in clays. The case n = 9, s/d = 8, L/d = 20, BR/d = 20 is herein considered.
Fig. 5 reports the load-settlement curves of the raft and pile group alone according to the
numerical results. Such curves are then fitted through Eq. (7), where the ultimate loads are
evaluated by the method proposed by Chin (1970) for the piles (Qp,u = 25.7 MN) and through the
classical plasticity theory for the raft (Qr,u = 66.8 MN). Initial stiffnesses and coefficients np and
nr are calibrated to minimize the error between the curve and the discrete numerical data,

resulting in: Kp,0 = 2547 MN/m, Kr,0 = 716 MN/m, np = 1.48, nr = 1.91. To take into
consideration the decrease in the ultimate load of the raft due to the shadow effect provided by
piles, Qr,u in Eq. (6b) has been multiplied by a coefficient αUR = 1 – 3 (AG/AR) / (s/d) proposed
by the authors, in the case under examination equal to 0.76. The proposed procedure leads to the
results illustrated in Fig. 6, where the performance of the method is compared to authors’ FEM
results. As it can be noticed the method, despite its simplicity, furnishes results in very good
agreement with rigorous results both in terms of load-settlement curve and load sharing.
3.3.2 Comparison with Comodromos et al. (2009)

In a similar fashion, the method is applied to numerical results obtained through Finite
Difference Method analyses by Comodromos et al. (2009). The case examined in the paper refers
to a group of 3x3 piles of diameter 1.2 m and length 38 m, surmounted by a square raft having
width BR equal to 9 diameters, embedded in multi-layer subsoil consisting of an alternation of
silty sands and clays. The procedure described above is utilized to define the input load-
settlement curves for piles and raft alone (Qp,u = 75.6 MN, Qr,u = 239 MN, Kp,0 = 5642 MN/m,
Kr,0 = 725 MN/m, np = 1.28, nr = 4.45), whereas the value of αUR is set to 0.56 following de
Sanctis and Mandolini (2006). Fig. 7 reports the raw numerical data and the interpolation curves
used for applying the present method, whereas Fig. 8 reports the results, compared with ones
coming from the FDM analysis, in terms of load-settlement curves and load sharing. Again, the
good performance of the simplified method can hardly be overstated.
3.3.3 Comparison with Fioravante and Giretti (2010)

The authors report a number of tests carried out in sandy soil by means of geotechnical
centrifuge. Nevertheless, only the experimental data relative to a single pile is available. To
obtain the load-displacement curve of the pile group, interaction coefficients, applied by
Randolph and Wroth (1979), have been used to amplify, for each value of the applied load, the
linear component of the single pile displacement, given by the load divided by the initial
stiffness, in order to account for group effects. The elastic settlement of the group has then been
added to the nonlinear component of the single pile displacement to obtain the nonlinear load-
settlement curve of the whole group. To account for the increase in ultimate load due to
installation effects, a factor of 1.4 has been multiplied to the ultimate group load previously
determined, following the indications of Vesic (1969) for a 3x3 pile group at spacing s/d = 5.
The resulting curve is plotted in Fig. 9. Given that the group clearly approaches the failure load,
the last value has been used as ultimate load, while the first point has been utilized to derive the
initial stiffness Kp,0 = 1306 MN/m at the prototype scale. The least square procedure results in np
= 0.37, corresponding to an almost elastic-perfectly plastic behavior. The raft behaves as quasi-
linear across the investigated loads, so that resulting values for the interpolation curve are Qr,u =
91 MN, Kr,0 = 106 MN/m and nr = 0.03. Results of the proposed method are shown in Fig. 10
along with the ones from centrifuge test. Considering also that the input pile group curve has
been not derived directly from the centrifuge tests, the agreement between the predicted and
measured loads and settlements should be considered as very satisfactory. Note that in this case
the original PDR method leads to quite accurate results, given the almost linear behavior of the
piles until failure and the elastic response of the raft within the range of loads examined.

Fig. 5. Pile group and raft load-settlement curves from de Sanctis and Mandolini (2006) and
associated interpolation curves utilized for the proposed method.
Fig. 6. Comparison between results from the proposed method and rigorous FEM analyses
provided by de Sanctis and Mandolini (2006).

Fig. 7. Pile group and raft load-settlement curves from de Comodromos et al. (2009) and

Fig. 8. Comparison between results from the proposed method and rigorous FDM analyses
provided by Comodromos et al. (2009).

Fig. 9. Pile group and raft load-settlement curves from Fioravante and Giretti (2010) and

Fig. 10. Comparison with results from the proposed method and centrifuge test results by
Fioravante and Giretti (2010).

3.3.4 Application to a full scale case history
The above procedure is here applied to a well-documented case history related to the main pier
of the cable stayed bridge over the river Garigliano in Italy (Mandolini and Viggiani, 1992;
Mandolini et al. 2005; Viggiani et al. 2012) where a thick deposit of rather compressible silty
clay is found. The foundation of the main pier is made up by a 4-m thick, 10.6 x 19 m raft and
144 driven tubular steel piles, filled with concrete, having length of 48 m and an average
diameter of 0.381 m.
Reference is made here to long term conditions, so that the bearing capacities of the raft and
the single pile may be estimated by utilizing classical bearing capacity formulae assuming an
angle of shearing resistance ϕ = 35°. Under this assumption, bearing capacity factors for the raft
were estimated as Nq = 33.3, following Prandtl (1921), and N = 37.1 (Davis and Booker, 1971).
γ
Regarding the pile, a value of shaft coefficient β = 0.25 has been assumed, while for base
resistance a value of Nq = 20 has been considered (Berezantzev, 1961).
To account for group effects the ultimate axial capacity of the pile group has been taken as the
sum of the individual contributions of single piles multiplied by an efficiency factor of 0.8,
following the indications by De Mello (1969). Bearing capacities of raft and pile group were
found to be 540 and 480 MN, respectively.
The initial stiffness of the raft has been calculated by the formula proposed by Carrier and
Christian (1973), employing a low-strain stiffness of the soil varying according to the law G0(z*)
= 3.8 + 1.5z* (with shear modulus in MPa), where z* = 4 m is the depth of the raft base from soil
surface. The resulting initial stiffness of the raft-soil system has been estimated equal to Kr0 =
550 MN/m, thus leading to an estimated settlement of the unpiled raft not lower than 20 cm. The
unpiled raft stiffness has been considered as constant with the load (that is, nr = 0) as for the case
at hand the raft-soil contact is expected to be linear at least until the working loads, due to the
expected low load percentage carried by the raft.
For sake of completeness, it is worth mentioning that the raft bearing capacity has been
reduced by 88% following the indications by de Sanctis and Mandolini (2006); however, for the
hypothesis of linear behavior of the raft-soil system, this assumption has no effect on the
behavior of the piled raft under working loads.
According to Mandolini and Viggiani (1997), pile group stiffness has been assessed by
summing the elastic contribution due to the pile-to-pile interaction to the nonlinear component of
the settlement of the single pile. Initial stiffness of single pile was estimated by the classical
Randolph and Wroth (1978) formula and was calculated as 200 MN/m, therefore obtaining Kp0 =
2387 MN/m (Butterfield and Douglas 1981). Adopting the suggestion from the numerical studies
above for fine-grained soils, a factor np = 1.5 has been utilized to describe the nonlinear behavior
of the single pile. Finally, a raft-pile group interaction factor αrp = 0.7 has been adopted,
according to the graphs shown by Clancy and Randolph (1993).
The comparison between the results obtained by the proposed procedure and the experimental
load-settlement curve is reported in Fig. 11. Looking at the final measured settlement (that is,
after consolidation), the agreement between experimental data and predicted settlement is very
good. In particular, under the working load of 113 MN the estimated settlement is 53 mm as
opposed to the long-term measured settlement of 52 mm. Interestingly, the method predicts that
piles adsorb 88% of the total working load, almost identical to the measured value of 87% after
10 years from construction (see Mandolini et al. 2005; Viggiani et al. 2012).

Fig. 11. Comparison with results from a case history of the main pier of the cable stayed bridge
over the river Garigliano.
4. SEISMIC DESIGN
Modern seismic codes (e.g. Eurocode 8) require that piles shall be designed to withstand both
inertial forces (i.e. coming from the superstructure vibrations) and kinematic action (i.e. due to
the soil deformations arising from the passage of seismic waves). Due to seismic soil
deformations, even in an hypothetical absence of a superstructure, piles are subjected to bending
moments resulting from their deflection. Long fixed-head piles in homogeneous soil (i.e., with
stiffness constant with depth) are unable to offer significant opposition to the displacement that
soil tries to impose (at least at low frequencies, which are of major concern), so that pile
curvature at the top is equal to the curvature of soil at surface (Di Laora et al. 2013). Simplified
expressions for the assessment of kinematic bending are available for subsoil conditions other
than homogeneous. A generalized formulation for inhomogeneous soil has been recently
proposed by Di Laora and Rovithis (2015). The authors considered the problem of a flexible pile
embedded in a subsoil with stiffness varying continuously with depth according to the
generalized function (Fig. 12a):
n
⎡ z⎤
G s (z) = G sd ⎢a + (1 − a) ⎥ (24)
⎣ d⎦
where Gs is the shear modulus of the soil, Gsd being the corresponding value at the depth of
one diameter, z is the depth from ground surface while a and n are coefficients regulating the
distribution of stiffness with depth.

(a) (b)
Fig. 12. Pile-head kinematic bending moment (Di Laora and Rovithis, 2015).
It is straightforward to verify that n = 0 leads to a constant soil stiffness, while n = 1

corresponds to a linear variation of Gs with depth. In the latter case, a = 0 will provide the special
case of zero stiffness at surface (i.e., Gs proportional to z).
As reported in Fig. 12b, the authors found out that pile curvature at head (1/R)p is equal to an
“effective” soil curvature (1/R)s,eff defined as:
a s ρs
(1/ R )p ≅ (1/ R )s,eff = (25)
G s ( La / 2 )
where as and ρs are surface acceleration imposed by the earthquake and mass density of the
soil, supposed constant along depth, respectively. The authors also provide an iterative
expression to calculate pile active length in such inhomogeneous soil. Instead of the iterative
procedure, Karatzia and Mylonakis (2016) furnish the closed-form expression:
4
⎧ 1 4+ n ⎫
⎪ ⎡ ⎤ ⎪⎪
1 ⎪⎢ 4 n+4
5 ⎛ π Ep ⎞ ⎥4
La = d ⎨ a + ( n + 4 )(1 − a ) ⎜ ⎟ − a ⎬ (26)
1 − a ⎪⎢ 16 ⎝ 2 Esd ⎠ ⎥ ⎪
⎢ ⎦⎥
⎪⎩ ⎣ ⎪⎭
where Esd refers to soil Young’s modulus at the depth of one diameter. The bending moment at
head is then equal to:
πd4
M = Ep (1/ R )p (27)
64

Equation (25) clearly reveals that the kinematic bending moment at pile head induced from an
inhomogeneous soil is the one induced by an equivalent homogeneous soil of stiffness Gs =
Gs(La/2), which means, in practical terms, that the stiffness ruling the interaction phenomenon is
the one at the depth of 3-4 diameters. It is therefore evident that, for the same value of equivalent
shear wave velocity Vs,30, the inhomogeneous soil will always provide more severe kinematic
moments.
These considerations may also affect pile size, providing some limitations for the diameter.
Considering for simplicity constant stiffness, kinematic moments given by Eq. (27) are
proportional to the fourth power of pile diameter, while section capacity is proportional to the
third power. This means that there always be a maximum diameter beyond which pile cannot
withstand kinematic moment. Such diameter for soft soil and high seismicity may be of the order
of 1 meter or even lower (Di Laora et al., 2013).
On the other hand, kinematic interaction may be helpful for the structure, as seismic motion at
the foundation is generally of smaller magnitude compared to that in free-field conditions, i.e. in
absence of the foundation itself (Di Laora and de Sanctis 2013). To give an idea of the
magnitude of ‘filtering effect’ exerted by piles, Fig. 13 reports some numerical results obtained
by Rovithis et al. (2015), where the ratio of structure accelerations with and without piles is
depicted as function of the fundamental period of the structure for different values of Vs,30 and
different soil stiffness profiles. It is evident that filtering effect is pronounced for stiff/short
structures (low period of vibration), large pile diameters and increases with increasing degree of
inhomogeneity of soil.

Fig. 13. Reduction of structural acceleration due to pile-induced filtering effect (Rovithis et al.
2015).

5. CONCLUSIONS
In this paper a number of geotechnical and seismic issues in matter of design of piled rafts have
been discussed. Further, a novel method for the assessment of the load-settlement curve and load
sharing between piles and raft has been proposed for piled rafts subjected to static vertical
centered load. The method takes into account in a simplistic yet accurate manner nonlinear soil
behavior and provide results which compare well with both numerical and experimental studies
in the literature as well as a well-documented case history of a bridge pier.
Further, seismic action provides further design issues to be addressed. Simple formulae are
available in literature to assess bending moments due to the kinematic interaction between piles
and the surrounding soil which undergoes seismic shaking. This action is proven to be critical for
large-diameter piles in soft soil. However, under these circumstances piles may be very helpful
to reduce the seismic demand to the superstructure.
6. REFERENCES
Berezantzev, V.G., Khristoforov, V., and Golubkov, V., 1961. Local bearing capacity and
deformation of piled foundations. Proc. of the V Int. Con. Soil Mech. Found. Eng., Paris,
July 17-22, Vol. 2. pp. 11-15.
Butterﬁeld, R., and Douglas, R. A. 1981. Flexibility coefﬁcients for the design of piles and pile
groups. Construction Industry Research and Information Association, Technical Note, 108.
Caputo, V., and Viggiani, C., 1984. Pile foundation analysis: a simple approach to nonlinearity
effects. Rivista Italiana di Geotecnica, 18(2), 32-51.
Carrier, W. D., and Christian, J. T., 1973. Rigid circular plate resting on a non-homogenous
elastic half-space. Geotechnique, 23(1), 67-84.
Chin, F. K., 1970. Estimation of the ultimate load of piles not carried to failure. In “Proceedings
of the 2nd Southeast Asian Conference on Soil Engineering”, pp. 81-90.
Clancy, P., and Randolph, M. F., 1993. An approximate analysis procedure for piled raft
foundations. International Journal for Numerical and Analytical Methods in Geomechanics,
17(12), 849-869.
Comodromos, E. M., Papadopoulou, M. C., and Rentzeperis, I. K., 2009. Pile foundation
analysis and design using experimental data and 3-D numerical analysis. Computers and
Geotechnics, 36(5), 819-836.
Davis, E.H., and Booker, J.R., 1971. The bearing capacity of strip footings from the standpoint
of plasticity theory. Proc. I Australian-New Zeland Con. in Geomech., Melbourne, pp. 276-
282.
De Mello, V.F.B., 1969. Foundations of buildings on clay. State of the Art Report, Proc. VII
Int. Conf. Soil Mech. Found. Eng., Mexico City, Vol. 1, pp. 49-136.
de Sanctis, L., and Mandolini, A., 2006. Bearing capacity of piled rafts on soft clay soils. Journal
of Geotechnical and Geoenvironmental Engineering, 132 (12), 1600-1610.
Di Laora, R., and de Sanctis, L., 2013. Piles-induced filtering effect on the foundation input
motion. Soil Dynamics and Earthquake Engineering, 46, 52-63.
Di Laora, R., Mylonakis, G., and Mandolini, A., 2013. Pile-head kinematic bending in layered
soil. Earthquake Engineering & Structural Dynamics, 42(3), 319-337.
Di Laora, R., and Rovithis, E., 2015. Kinematic Bending of Fixed-Head Piles in
Nonhomogeneous Soil. Journal of Geotechnical and Geoenvironmental Engineering, 141(4),
04014126.

Fioravante, V., and Giretti, D., 2010. Contact versus noncontact piled raft foundations. Canadian
Geotechnical Journal, 47(11), 1271-1287.
Karatzia, X., and Mylonakis, G., 2016, Discussion of "Kinematic bending of fixed-head piles in
nonhomogeneous soil" by Raffaele Di Laora and Emmanouil Rovithis. Journal of
Geotechnical and Geoenvironmental Engineering, 142(4), 07015042.
Mandolini, A., 2003. Design of piled rafts foundations: practice and development. Proc. 4th Int,
Geotechnical Seminar on Deep Foundations on Bored and Auger Piles, Millpress,
Rotterdam, pp. 59-80.
Mandolini, A., Russo, G., and Viggiani, C., 2005. Pile foundations: experimental investigations,
analysis and design. State of the Art Report at XVI ICSMFE, Osaka, Japan, September 12-
16. Vol. 1, pp. 177-213.
Mandolini, A., and Viggiani, C., 1992. Terreni ed opere di fondazione di un viadotto sul fiume
Garigliano. Rivista Italiana di Geotecnica, 26(2), 95-113.
Mandolini, A. and Viggiani, C., 1997. Settlement of piled foundations.Géotechnique, 47(3), 791-
816.
Poulos, H.G., 2000. Practical design procedures for piled raft foundations. In “Design
applications of raft foundations”, Hemsley J.A. ed., Thomas Telford, pp. 425-467.
Poulos, H.G., Carter, J.P. and Small, J.C., 2001. Foundations and retaining structures – Research
and practice. Proc. XV Int. Conf. Soil Mechanics and Foundation Engineering, Istanbul,
Vol. 4, pp. 2527-2606.
Poulos, H.G., and Davis, E.H., 1980. Pile Foundations Analysis and Design. Wiley, New York.
Prandtl, L., 1921. Hauptaufsätze: Über die eindringungsfestigkeit (härte) plastischer baustoffe
und die festigkeit von schneiden. ZAMM-Journal of Applied Mathematics and
Mechanics/Zeitschrift für Angewandte Mathematik und Mechanik, 1(1), 15-20.
Randolph, M.F., 1983. Design of piled raft foundations. Proc. Int. Symp. on Recent
Developments in Laboratory and Field Tests and Analysis of Geotechnical Problems,
Bangkok, pp. 525-537.
Randolph, M.F., 1994. Design methods for pile groups and piled rafts. State of the Art Report,
XIII ICSMFE, New Delhi, January 5-10, Vol. 5, pp. 61-82.
Randolph, M.F., and Wroth, C.P., 1978. Analysis of vertical deformation of vertically loaded
piles. J. Geotech. Eng. Div. ASCE, 1978, 104(12), 1465-1488.
Randolph, M.F., and Wroth, C.P., 1979. An analysis of the vertical deformation of pile
groups. Geotechnique, 29(4), 423-439.
Rovithis, E., Di Laora, R., and de Sanctis, L., 2015. Foundation motion filtered by piles: effect of
soil inhomogeneity. XVI ECSMGE , DOI 10.1680/ecsmge.60678, Vol. 3, 165.
Russo, G., and Viggiani, C., 1998. Factors controlling soil-structure interaction for piled rafts.
Proc. Int. Conf. on Soil-Structure Interaction in Urban Civil Engineering, Darmstadt, pp.
297-322.
Vesić, A., 1969. Experiments with instrumented pile groups in sand. Proc. of Symposium on
Performance of Deep Foundations, San Francisco, June 1968, American Society for Testing
and Materials, ASTM, Special Technical Publication, STP 444, pp. 177-222.
Viggiani, C., Mandolini, A., and Russo, G., 2012. Piles and pile foundations. CRC Press.

Recent developments in vibratory driving and soil compaction
Massarsch, K.R.(1)
(1)
Geo Risk & Vibration AB, Bromma, Sweden, <rainer.massarsch@georisk.se>
ABSTRACT. During the past two decades, the capacity and performance of hydraulic vibrators
has undergone a rapid development. By the use of vibrators with variable frequency and eccentric
moment, the transmission of the vibration energy to the surroundings can be controlled. Methods
and empirical rules for estimating the drivability of sheet piles are presented. The application of
modern vibrators is not limited to the driving of piles and sheet piles. The efficiency of vibratory
compaction can be enhanced by adapting the vibrator to the resonance frequency of the vibrator-
pile-soil system. The optimization of the compaction process by electronic process control is
highlighted and illustrated. A new concept is proposed where the dynamic penetration resistance
is correlated to the pile penetration speed, measured in terms of the number of vibration cycles
during the driving. The validity of the concept is demonstrated by a case history. Finally, examples
of pile and ground improvement solutions are presented which take advantage of the fundamental
aspects of vibratory driving.
1. INTRODUCTION
Modern, hydraulic vibrators can be used for different construction processes, such as the
installation and/or extraction of piles and sheet piles and compaction . However, in spite of their
frequent use in the foundation industry, many aspects of vibratory driving are still based on overly
simplistic empirical rules. There is a need for a better understanding of the parameters of
importance for vibratory driving of piles and sheet piles. When using vibrators for pile and sheet
pile driving or vibratory compaction of granular soils, the following issues need to be considered:
• drivability: select appropriate vibrator type and capacity to assure efficient installation
• bearing capacity: verify that the required bearing capacity can be achieved
• vibratory compaction: assure that the compaction requirements can be achieved
• environmental impact: minimize emission of ground vibrations and/or noise.
The four issues are closely interrelated and one cannot be addressed without considering the other
three.
2. VIBRATOR PERFORMANCE
2.1 Fundamentals
The performance of modern vibrators can be controlled and adapted for optimal performance.
Several parameters govern the vibratory driving process and it is important to realize how they can
be varied to achieve efficient driving or compaction. Vibratory excitation affects a pile in a
different way than does impact driving. An impact hammer imparts a short duration blow to the
pile head, sending a stress wave down the pile. The initial hammer energy is reduced in passing
through the pile helmet and pile cushion. Each blow must be able to activate the inertia of the pile

and overcome the static soil resistance along the pile shaft and at the pile toe. In contrast, in the
case of vibratory driving, the pile is rigidly connected to the vibrator, resulting in minimal energy
loss. The vibration frequency is relatively slow, typically below 40 Hz (2,400 rpm). Thus, the wave
length is much longer than in the case of impact driving. The pile is kept in an oscillating motion
during the entire vibratory driving process. In the case of vibratory driving in coarse-grained soil,
the toe resistance is approximately similar to that during impact driving, while the shaft resistance
is usually lower (Westerberg et al. 1995). It is generally recognized that vibratory driving is most
effective in coarse-grained soil and less efficient in fine-grained (cohesive) soils.
The most important parameters of the vibrator that govern the drivability of piles and sheet
piles, are the eccentric (sometimes called “static”) moment, the vibration frequency, and the mass
ratio between the vibrator and pile. These parameters will be discussed briefly in this section.
Modern vibrators are hydraulically driven, which allows continuous variation of the vibrator
frequency during operation. The vertical oscillation of the vibrator is generated by counter-rotating
eccentric masses. The peak value of the centrifugal force, Fv, acting in the vertical direction,
depends on the eccentric (static) moment, Me, and on the circular frequency ϖ (2 π f), of the
rotating eccentric masses.
(1)
Many modern vibrators also enable the centrifugal force to be adjusted continuously during
operation, resulting in a variable displacement amplitude, s. This feature allows the machine
operator to start up and shut down the vibrator at zero vibration amplitude, thereby reducing the
risk of vibration amplification due to resonance effects. An additional important factor for the
vibrator performance is the displacement amplitude, which together with the centrifugal force, Fv,
determines the driving ability of the vibrator. For a vibrator suspended above the ground surface,
the axial displacement amplitude (single amplitude), s can be determined from the following
relationship.
𝑀𝑀𝑒𝑒
𝑠𝑠 = (2)
𝑀𝑀𝑑𝑑
The “total dynamic mass”, Md, is the sum of all masses, which need to be accelerated by the
vibrator. This includes, e.g., the mass of vibrator (but not the static mass separated by soft springs
from the vibrator), the pile, and the vibrator clamp. Note that most equipment manufacturers
express the displacement amplitude as peak-to-peak ("double") amplitude. From Eq. (2), it can be
appreciated that the free pile displacement amplitude, s, is independent of the vibration frequency,
f. In order to maximize the displacement amplitude, the total dynamic mass, Md, should be kept as
small as possible.
2.2 Vibrator Performance Characteristics
Before 1980, most hydraulic vibrators had fixed eccentric moment, with a typical operating
frequency between 20 and 30 Hz. These vibrators were used for basic construction works, such as
driving and extracting of sheet piles. Table 1 shows a typical range of performance parameters for
vibrators with fixed eccentric moment.

TABLE 1. Müller Vibrators with fixed eccentric moment (Müller-ThyssenKrupp 2004).
Vibrator Units MS-25 H2 MS-25 H3 MS-50 H2 MS-50 H3

Centrifugal force kN 774 774 1430 1430
Eccentric moment kgm 25 25 50 50
Speed rpm 1680 1680 1615 1615
Frequency Hz 28 28 26.9 26.9
Static pulling force kN 400 400 500 500
Dynamic mass1) kg 1930 2550 3340 3820
1)
Total mass kg 3200 3600 6300 6790
Vibration (double) mm 25.9 19.6 29.9 26.2
amplitude1)
1)
without clamping device/pile
Vibrators have experienced a rapid development in terms of power, range of operating

parameters (eccentric moment and frequency), and monitoring of the driving and extraction
process. An important advantage of hydraulic vibrators over traditional impact hammers (i.e., not
underwater hydraulic impact hammers) is that they can be operated under water. A major
development in vibrator design came with the introduction of a stepwise adaptation of the eccentric
moment according to specific driving requirements. Such change of the eccentric moment is made
manually by adding or removing weights, Figure 1. For example, if high frequency operation is
required, weights are removed on-site to increase to the desired frequencies with the same
centrifugal force.
Eccentric mass
Space for
eccentric mass
Fig. 1. Vibrator eccenter allowing for adding up to three additional masses.
As mentioned, the displacement amplitude given by vibrator manufacturers is usually in terms

of double-amplitude and usually for a freely suspended vibrator (without clamp and attached
pile/sheet pile). Typical performance characteristics of vibrators with stepwise variable eccentric
moment are shown in Table 2.

TABLE 2. Müller Vibrators with stepwise variable eccentric moment
(Müller-ThyssenKrupp 2004).
Vibrator Units MS-25 MS-50 HHF MS- MS-120 M-200

HHF 100 HHF HHF
HHF
Max. Centrifugal kN 750 1500 2500 3003 4000
force
Eccentric moment kgm 25 50 100 116 190
Steps2) kgm 12/15/ 24/30/ 48/60/ 80/94/ (98)/110/
20/25 40/50 80/100 110/116 150/190
Frequency steps Hz 39.9/35.2/ 39.9/35.2/3 36/32/ 30.9/28.3/ 30/26/ 22.9
30.5/27.3 0.5/27.3 27.8/2 26.2/25.6
5
Static pulling kN 280 500 600 1200 1200
force
Dynamic mass1) kg 2900 4500 7700 8900 11750
Total mass1) kg 3700 6100 10900 15500 15500
Vibration mm 17.2 22.2 26.0 26.1 32.4
(double)
amplitude1)
Vibration mm 8.3/10.3/ 10.7/13.3/1 12.5/1 18.0/21.1/ 16.7/18.7/
(double) 13.8/17.2 7.8/22.2 5.6/20. 24.7/26.1 25.5/32.4
1)
amplitude 8/26.0
(steps)
1)
without clamping device/pile; 2)available eccentric moment
In the 1990s, vibrators with variable frequency and variable amplitude were introduced with
resonance-free starting-up and shutting-down of vibratory driving. Such vibrators allow the
operating frequency and eccentric moment (and thus amplitude) to be varied according to specific
driving requirements and soil conditions. Typical performance parameters of vibrators with
variable frequency and variable eccentric moment (amplitude) are shown Table 3. It should be
mentioned that vibrator performance data of most modern manufacturers are similar to those
shown in Table 1 through 3, for instance vibrators manufactured in North America (APE, ICE) or
Europe (Dieseko, PTC, RTG).
2.3 Assessment of Vibrator Capacity
Vibrator manufacturers have provided empirical guidelines for the selection of vibrators. In Figure
2, the centrifugal force required for driving sheet piles is shown as a function of pile mass and
depth of penetration.

TABLE 3. Resonance-free Müller Vibrators with variable frequency and eccentric moment
(Müller-ThyssenKrupp 2004).
Vibrator Units MS-10 MS-16 MS-20 MS-28 MS-32 MS-48 MS-62

HFV HFV HFV HFV HFV HFV HFV
Centrifugal kN 610 968 1230 1473 1980 2960 2998
force
Eccentric kgm 0-10 0-16 0-19.5 0-28 0-32 0-48 0-62
moment
(variable)
Speed rpm 2358 2370 2400 2190 2375 2350 2100
Frequency Hz 39.3 39.5 40.0 36.5 39.6 39.0 35.0
Static pulling kN 180 300 300 500 600 600 800
force
Dynamic mass kg 1700 2565 2530 3120 4850 6520 6850
w/o clamping
device
Total mass1) kg 2300 3530 3600 5320 7250 9700 11165
Vibration mm 11.8 12.5 15.4 18.0 13.2 14.7 18.2
amplitude1)
1)
without clamping device/pile
Fig. 2. Empirical chart for estimation of required vibrator centrifugal force for driving of sheet
piles. The example (red lines): sheet pile mass: 3.6 ton maximum penetration depth: 20
m, driving into medium dense sand (III) requires vibrator with centrifugal force of at least
1,600 kN (based on Müller-ThyssenKrupp 2004).

The five soil categories depicted in Figure 2 are defined in Table 4.
Table 4. Drivability of sheet piles into granular soils, based on cone penetration tests, cf. Fig. 2.
I II III IV V
Test Type Very loose Loose Medium Dense Very Dense
Dense
SPT, N <4 4 - 10 10 - 30 30 - 35 > 50
(blows/0.3m)
CPT, qc <5 5 - 10 10 - 15 15 - 20 > 20
(MPa)
DPH1), N10 <5 5 - 10 10 - 15 15 - 20 > 20
(blows/10 cm)
1)
Heavy dynamic probing: mass 50 kg, height of fall 500 mm
Figure 2 and Table 4 are based on experience from sheet pile driving, compiled by one
particular vibrator manufacturer (Müller-ThyssenKrupp, 2004). Such guidelines must be treated
with caution. The guidelines assume that the user understands the limitations involved. In the
example shown in Fig. 2, a pile with mass 3.6 ton shall be driven to a depth of 20 m into medium
dense sand (III) (qc ≈12 MPa). According to Figure 2, a vibrator with a centrifugal force of about
1,600 kN will be required.
An important aspect of Figure 2 is that the diagram was based primarily on experience from
sheet pile projects, where the shaft resistance dominates. In the case of tubular piles, however, the
toe resistance contributes significantly to the total driving resistance. Westerberg et al. (1995)
suggested that the dynamic toe resistance during vibratory driving of piles is approximately equal
to the CPT cone stress, qc. Figure 3 shows the centrifugal force, required to drive a tubular pile
with closed toe into granular soils of variable density, Massarsch and Fellenius (2017). The soil
conditions are expressed according to the classification given in Table 4.
The application of Figure 3 is illustrated by the following example. The centrifugal force
required to drive a tubular pile with a toe diameter of 350 mm into a granular soil with a cone
resistance qc ≈ 10 MPa is approximately 950 kN. If the soil at the pile toe has a cone stress qc ≈ 15
MPa, the required centrifugal force will be approximately 1450 kN.
An additional parameter to be considered when assessing vibratory driving is the relative
displacement of the pile to the soil. The relative displacement is particularly important for
overcoming the shaft resistance in fine-grained (cohesive) soils. The larger the displacement
amplitude, the more effective the driving process will be. The displacement amplitude depends on
the total dynamic mass to be accelerated by the vibrator (dynamic mass) and the eccentric moment.
In the following example, a sheet pile shall be driven by a vibrator (MS-28) with a 3,250-kg
dynamic mass (vibrator and clamp). The mass of the 15 m long pile is 2,700 kg. The displacement
amplitude can now be estimated from Eq. (2). The double displacement amplitude (2s) of the
suspended pile, which is usually quoted by vibrator manufacturers, is 9 mm (Me = 28 kgm). During
sheet pile penetration, the vibration amplitude decreases, depending on the stiffness of the soil to
be penetrated.

Fig. 3. Centrifugal force required to drive tubular pile (closed toe). Soil classification according
to Table 4. The example illustrates the effect of cone stress on centrifugal force required
to drive a pile with 350 mm toe diameter, Massarsch and Fellenius (2017).
3. MONITORING OF VIBRATORY DRIVING

3.1 Monitoring System
An important advantage of vibratory driving is that many aspects of the installation process can be
monitored, controlled, and documented. With modern electronic measuring equipment, it is
possible to acquire, display, and record information from different sensors, which can be mounted
on the pile, the vibrator, the power pack, and the ground. Figure 4 shows the set-up of a vibratory
monitoring system, which initially was developed for deep vibratory compaction.
When monitoring vibratory driving of piles and sheet piles, it is desirable that a data logger
be used to record the following parameters:
• Position of pile/sheet pile

• Time of recording (at least one reading per second)
• Depth of sheet pile during penetration or extraction
• Operating frequency
• Acceleration of vibrator (pile top acceleration)
• Static force applied to the pile (pushing or lifting force)
• Hydraulic pressure of power supply
• Vibration velocity on or in ground (geophones).

Fig. 4. Monitoring components of vibratory driving. Detail of control unit is shown in Fig. 5.
Ground vibrations can be recorded using geophones, Figure 5a, and can be displayed to the
machine operator during vibratory driving, Figure 5b. Monitoring of information obtained during
the vibratory driving process and response of the ground and/or of adjacent structures is an
important aspect of modern vibratory works. For instance, in the case of vibratory driving in the
vicinity of vibration-sensitive buildings or installations, a computer-operated system can be used
to control the maximum vibration intensity in order to ensure that specified limiting values are not
exceeded. Moreover, when vibrators are used for deep vibratory compaction, the vibration
measurements can be used to guide the operation to ensure that maximum transfer of vibration
energy is obtained from the vibrator/probe system to the surrounding soil, for example when using
the resonance compaction system (Massarsch, 1991).
3.2 Optimization of Vibratory Driving and Compaction
From vibration monitoring, several important parameters can be derived such as: pile and soil
displacement amplitude, resonance frequency of vibrator-pile-soil system, pile/probe penetration
speed. These parameters can be displayed to the machine operator in real time and assist in the
optimization of the driving or compaction system. The displacement amplitude provides important
information regarding the dynamic mass to be accelerated by the vibrator. This information is
particularly important when driving piles or sheet piles into cohesive soils. Soil compaction is
enhanced when the vibrator is operated at the system resonance frequency, at which ground
vibrations are strongly amplified. However, pile or sheet pile penetration will be low at the
resonance frequency, at which the risk of vibration problems in surrounding buildings increases.

a) Ground vibration measurement b) Display of vibration parameters measured during pile
by tri-axial geophone penetration (System Gamperl und Hatlapa)
Fig. 5. Monitoring of ground vibrations and display of measurements to machine operator.
The pile penetration speed is another important parameter, which reflects the efficiency of
vibratory driving. However, the penetration speed is relevant only when the pile or sheet pile is
allowed to penetrate under the full weight of the vibrator.
3.3 Ground Vibrations during Resonance Compaction
The above described monitoring system has been used on a large number of projects, where
resonance was used to increase the compaction efficiency in coarse-grained soil. In the following
case history, a steel pipe pile (with closed toe) was vibrated into loose to medium dense sandy soil,
using a Müller MS100 variable-frequency vibrator, see Table 2. The dynamic ground was
monitored by a tri-axial geophone, placed 4 m away from the compaction point. The soil profile
consisted of medium to dense sand, which prior to compaction was loose to medium dense. During
the initial probe penetration phase, the compaction probe was vibrated at high frequency (35 to
39 Hz). During the compaction phase, the vibration frequency was gradually lowered until
resonance and, thus, maximum vibration amplification was achieved (about 13 - 16 Hz). The
response at the ground surface during vibratory driving is shown in Figure. 6. The vertical ground
vibration velocity during penetration varied typically between 0 and 4 mm/s (average 2 mm/s),
with peak values around 6 mm/s. At resonance (between 12 and 16 Hz), the vertical vibration
velocity increased significantly, with maximum values reaching 20 mm/s. The vibration
amplification was about 5 to 10. At high frequencies, the probe penetration velocity is usually
significantly higher (by a factor of up to 10) compared to driving at resonance frequency.

Fig. 6. Vertical vibration velocity measured at 4 m distance from the vertically oscillating
compaction probe during penetration/extraction (32 – 38 Hz) and compaction phase (12 –
18 Hz).
Field monitoring of ground vibrations can provide valuable information regarding the transfer
of vibration energy from the pile to the surrounding soil. It is apparent that efficient pile penetration
occurs when the vibration frequency is significantly higher than the resonance frequency of the
vibrator-pile-soil system. On the other hand, transfer of vibration energy and, therefore, the
compaction effect is enhanced when the vibrator is operated at the resonance frequency.
4. ASSESSMENT OF VIBRATORY DRIVING PERFORMANCE
When a pile or sheet pile is to be installed by a vibrator, the selection of the driving equipment and
the installation process should be carefully planned. The vibrator type and capacity as well as the
driving parameters (operating frequency and amplitude) should be chosen, based on geotechnical
information. Using an unsuitable vibrator will not only result in project delays and infer additional
costs but under unfavorable conditions, also produce damaging ground vibrations in buildings or
installations in the ground. The following paragraph proposes a concept which can be used to
estimate vibrator centrifugal force, based on CPT results. It is recommended to determine the
optimal vibratory driving process by field trials. The objective is to develop a correlation between
CPT penetration resistance and pile penetration speed for a given vibrator type and pile size.
Rational design of a vibratory driving project requires site information that includes a well-
established soil profile with soil description obtained from laboratory study of soil specimens. The
best additional direct information is a continuous record of soil layering and density, such as

provided by a CPTU sounding. The information obtained from SPT can also be valuable. In
Europe, continuous light dynamic probing (DPL) or heavy dynamic probing (DPH) is frequently
used for assessment of pile driving resistance. Unless information is available from vibratory
driving from similar geology and equipment, a field trial is the best way of designing an installation
of specific piles or sheet piles.
4.1 Predicting Driveability
The driveability of piles or sheet piles can be related to the penetration resistance from dynamic
probing. In the case of heavy dynamic probing (DPH), a mass of 50 kg with a height of fall of 500
mm impacts on a steel rod and the number of blows for a penetration of 10 cm (N10) is determined.
An approximate relationship between different penetration testing methods and strength of
granular soils is given in Table 4. It is possible to correlate the penetration resistance to the number
of vibration cycles required for a pile or sheet pile to penetrate into granular soil. The speed of
probe (pile or sheet pile) penetration can be measured (preferably recorded every second, but
reported as penetration per minute). During driving, it is important that the vibrator rests fully on
the probe and is not held back by the machine operator, as this will affect the probe penetration
speed. It is also important, that the vibration frequency is high (above the resonance frequency)
and kept as constant as possible. At a high vibration frequency, the vibration resistance will be
caused primarily by the toe resistance of the probe, similar to the case at dynamic penetration
testing.
In the following example, a sheet pile is driven by a vibrator operating at 40 Hz into a sand
deposit consisting of several layers with variable DPL penetration resistance, blows/10 cm, shown
in Figure 7.
Fig. 7. Typical dynamic penetration test, Light Dynamic Penetrometer (DPL), from Schönit
(2009). The standard for the DPL (10 kg, 500 mm height of fall) is to record the
penetration as blows per 10 cm penetration).

From the depth measurement during vibratory driving of the sheet pile, the penetration speed
can be determined as a function of depth. Figure 8 shows the penetration speed (m/s).
Fig. 8. Sheet pile penetration speed measured at three different vibration frequencies (two tests
for each frequency), after Schönit (2009).
At field trials, the key vibrator performance parameters to record are the eccentric moment,
the vibration frequency, the centrifugal force, and the displacement amplitude. It is important to
record the displacement amplitude of the vibrator-pile system prior to driving of the pile
penetration into the ground. The intensity of ground vibrations adjacent to the pile should also be
recorded, as this information can be used to determine the risk of vibration amplification should
the vibrator operating frequency be too close to the resonance frequency of the vibrator-pile-soil
system. As the vibrator operating frequency, f (Hz or rpm) is known, it is possible to convert the
measured pile penetration speed, v (cm/min) into an equivalent number of penetration cycles, ce
per depth interval (cycles/cm).
(3)
In the above example, the vibrator operating frequency was kept constant at 40 Hz (2,400
rpm). It is now possible to convert the penetration speed from Figure 8 into an equivalent number
of vibration cycles, as shown in Figure 9. It is important to appreciate that the above shown
correlation is soil-type specific and influenced by the type and capacity of vibrator, and the size
and geometry (toe area) of the sheet pile.

Fig. 9. Relationship between penetration resistance, N10 and number of vibration cycles at
operating frequency of 40 Hz, cf. Figures 8.
5. ADDITIONAL APPLICATIONS OF VIBRATORY DRIVING
Vibrators can be used for the efficient installation of piles and sheet piles in coarse-grained soil.
However, innovative solutions have been developed to broaden the range of vibrator applications
and to facilitate vibratory driving in difficult soils.
At present, it is difficult to assess the bearing capacity of preformed piles, installed by
vibratory driving. However, recent studies regarding the bearing capacity of cast-in situ piles have
been published by Zeilinger and Hudelmaier (2009). They investigated the bearing capacity of
cast-in-situ piles, installed by different driving methods (impact-driven, vibrated and vibro-jetted).
Static and dynamic loading tests were performed on these piles that suggest that vibratory driving
and water jetting does not have a detrimental effect on the static bearing capacity of in situ cast
piles.
One interesting new concepts in foundation design during the past decades is the use of
settlement-reducing elements (piled rafts). The foundation slab is supported by conical concrete
elements, installed by vibratory driving (cf. Figure 11). This foundation concept is similar to
ground improvement by stone columns or mixed in place columns. Prefabricated elements
(concrete, steel or timber) can be used to increase the stiffness of loose or soft soil layers. However,
vibrated conical elements do not suffer from the uncertainties associated with the installation of
stone columns in soft, compressible soils. Massarsch et al. (1997) described a design concept,
which is based on the load-sharing between the foundation slab foundation and reinforcing
elements.

Fig. 11. Installation of conical concrete elements by vibrator with variable frequency.
6. CONCLUSIONS
Vibratory driving of piles and sheet piles is a common method, especially for the installation of
foundation units and sheet piles. The most important vibrator parameters for vibratory driving are
the centrifugal force, the vibration frequency, and thus also of the eccentric moment.
The vibrator-pile-soil interaction is a function of the resonance vibration frequency. Results
of a theoretical analysis are presented which can be used to predict reliably the resonance of the
vibrator-pile-soil system. The most important parameters governing the resonance frequency are
the stiffness (shear wave speed) of the soil and the mass of the vibrator and the sheet pile. The
eccentric moment does not affect the resonance frequency.
Empirical rules can be used to assess the required vibrator capacity (centrifugal force) for
sheet pile driving in primarily coarse-grained soil. However, a more reliable concept can be used
to predict the drivability of sheet piles or piles. A method is described which makes it possible to
correlate the penetration resistance to the number of vibration cycles during sheet pile driving. A
case history is presented which demonstrates the concept and shows that a correlation exists in
coarse-grained soil between penetration resistance (dynamic penetrometer) and number of
vibration cycles per depth interval. Based on this information, which can be obtained either from
field trials or experience from past projects, it is possible to determine the penetration speed during

vibratory driving. This information can be valuable when selecting the required vibratory driving
equipment and predicting the duration of vibratory installation of piles or sheet piles.
Vibratory driving is very effective when installing conical ridged elements, which can be used
to improve loose soil deposits.
7. ACKNOWLEDGEMENT
The advise and assistance by Bengt H. Fellenius is gratefully acknowledged. Also, the astute
comments by the reviewers of the manuscript are appreciated.
8. REFERENCES
Massarsch, K.R. 1991. Deep Soil Compaction Using Vibratory Probes, ASTM Symposium on
Design, Construction and Testing of Deep Foundation Improvement: Stone Columns and
Related Techniques, Robert C. Bachus, Ed. ASTM Special Technical Publication STP 1089,
Philadelphia, 1990, pp. 297-319.
Massarsch, K.R., Westerberg, E., and Broms, B.B. 1997. Footings supported on settlement-
reducing vibrated soil nails. XIV, International Conference on Soil Mechanics and Foundation
Engineering, Hamburg 97, Proceedings, Vol. 3, pp 1533-1539.
Massarsch, K.R. and Fellenius, B.H. 2017. Grundlagen des vibrationsrammens von Pfählen und
Spundbohlen (Fundamentals of vibratory driving of piles and sheet piles). Geotechnik. Ernst
und Sohn – Wiley. Accepted for publication.
Müller-ThyssenKrupp. 2004. MÜLLER Vibrator technology. ThyssenKrupp GfT Bautechnik,
Machinery - Product range. 15 p.
Schönit, M. 2009. Online-Abschätzung der Rammguttragfähigkeit beim langsamen Vibrations-
rammen in nichtbindigen Böden (Online-estimation of bearing capacity of driven piles during
slow vibratory driving in non-cohesive soils). Doctoral thesis, Universität Karlsruhe (TH),
Fakultät für Bauingenieur-, Geo- und Umweltwissenschaften. Universitätsverlag Karlsruhe,
Reihe F, Heft 65, 200 p.
Westerberg, E., Massarsch, K. R., and Eriksson, K. 1995. Soil resistance during vibratory pile
driving. Proceedings: International Symposium on Cone Penetration Testing, CPT´95, Vol. 3,
pp. 241 – 250.
Zeilinger, H. and Hudelmaier, K. 2009. The bearing capacity of vibrated cast-in-place-piles. Pile
Driver, Q4, V. 6, p. 44 – 47.

Recent developments and applications in
geotechnical field investigations for deep foundations
Mayne, P.W.(1) and Niazi, F.S.(2)

(1)
Georgia Institute of Technology, Atlanta, Georgia, USA <paul.mayne@ce.gatech.edu>
(2)
Indiana University - Purdue University, Fort Wayne, Indiana, USA <niazif@ipfw.edu>
ABSTRACT. For routine site investigations, the use of hybrid tools such as the seismic piezocone
and seismic flat dilatometer offer superior efficiency and economy to provide sufficient profiling
of subsurface conditions and the evaluation of geoparameters, as compared with routine soil
boring, drilling, and sampling. A single sounding provides up to five separate measurements on
soil behavior that can be used to calculate capacity and displacements for axial and lateral pile
foundation capacity, as well as in applications involving quality control for ground modification
projects. Axial pile capacity can be assessed using traditional approaches that utilize limit plasticity
and static equilibrium, or alternatively via direct in-situ testing methods. Load tests from a bridge
project in Minnesota are presented to illustrate the applicability of seismic piezocone testing.
1. INTRODUCTION
Each and every civil & environmental project requires a geotechnical site investigation to
determine the makeup of the underlying ground at that particular location. This can be
accomplished today using a combination and assortment of geophysical methods, soil drilling &
sampling for laboratory testing, and in-situ field testing.
There are currently many different types of deep foundation systems being used commercially
in support of civil engineering works for building loads, bridge piers, ports, and tower structures
both onshore and offshore. Pile types can include driven steel H and pipe (open-end versus closed-
end), precast versus prestressed concrete, timber, composite, tapered, and monotubes, or systems
of bored or augered deep foundations, such as cased or uncased drilled piers, slurry shafts,
augercast pilings, and caissons, as well as specialty types (e.g., Fundex, Omega, Screw, Dewaal).
Consequently, the notion that a single test number such as the SPT-N value can suffice for all the
needs in the analysis and design of modern piling foundations must be replaced with a more
rational belief that multiple measurements are paramount.
Herein, we shall explore the utilization of more modern tests, such as the seismic cone and
seismic dilatometer, for the collection of several measurements concerning soil behavior. In
addition to assessing soil stratigraphy and geoparameters in an efficient and economic manner,
these tests permit an evaluation of the nonlinear axial load-displacement response of deep
foundations.
1.1 Conventional Site Investigation

For the past century, the conventional approach to geotechnical subsurface investigation has been
the use of rotary drilling, augering, and sampling methods to create boreholes. Most common, the
use of open drive samplers permits the collection of small disturbed cylindrical soil specimens via
the standard penetration test (SPT). This provides a crude index (N-value) that needs an important

correction for the energy inefficiency of various drop hammer systems in use since its debut in
1902, including: pinweight, donut, safety, and automatic hammers.
At the time that the geotechnical profession became aware of the energy efficiency issues
(Seed et al. 1985; Skempton 1986), the average energy efficiency of SPTs in practice was
approximately 60%, so this became the reference value for the correction, designated; N60, which
is obtained:
N60 = CE ∙ Nmeasured = (ER/60) ∙ Nmeasured
where CE = correction factor for energy efficiency

ER = energy ratio for the particular hammer system used (ASTM D 4633)
Nmeasured = number of hammer blows to drive a split-spoon sampler a vertical distance of
300 mm (or 1 foot), reported as blows/0.3 m.
An illustration showing the importance of this correction is depicted in Figure 1 using SPT
data from the National Geotechnical Experimentation Site (NGES) at Northwestern University
(Finno et al. 2000). Here, the upper soils are comprised of clean fine sands (0.15 mm < D50 < 0.30
mm) that were subjected to two sets of SPTs in soil test borings using a safety hammer and an
automatic hammer. Figure 1a shows the raw uncorrected N-values while Figure 1b presents the
energy-corrected values. The significance and importance of the energy corrections on the N-
values is quite evident.
Fig 1. SPT profiles at the national test site at Northwestern University: (a) uncorrected N-values;
(b) corrected N60 (data from Finno et al. 2000)
There is a general misconception by geoengineers that if the SPT is performed using an

autohammer, the results do not need to be corrected for energy content. Recent studies have made
compilations of energy measurements by different organizations at various sites in the USA, all
using automatic hammers (Davidson et al. 1999; Honeycutt et al. 2014; Sjoblom et al. 2002). These
ER data represent over 20,000 field measurements per ASTM D 4633 and indicate a documented
range of ER from 45% to 95% for autohammer efficiencies. Therefore, one cannot assume a value
of ER for a valid correction of energy on a particular system. Interestingly, the state-of-the-practice

Fig. 2. Variety and selected assortment of in-situ tests for geotechnical site investigations
using autohammers has now risen to a mean value ± one standard deviation of ER (ave) ≈ 82% ±
7% based on measurements taken in the past five years (Honeycutt et al. 2014).
1.2 In-Situ Test Methods

A variety of in-situ devices and probes have been devised for characterizing the subsurface
conditions and providing quantification on the geotechnical engineering properties for pile
foundation analysis and design. Figure 2 provides an overview that depicts the various instruments,
penetrometers, blades, vanes, innovative devices, and inventions that are available. Certain of these
tests are rather specialized towards assessment of a particular geoparameter (e.g., total stress cell
for K0 evaluation; ball penetrometer for assessing undrained shear strength in very soft soils).
Some new developments specific to geotechnical foundation design include:
(a) thermal conductivity dissipation tests (TCT) that use temperature-measuring
penetrometers to assess the thermal properties of the ground for energy pile design, as
detailed by Akrouch et al. (2016);
(b) cone load test (CLT) that derives modulus and stiffness data (i.e., t-z curves) for piling
foundation design (Ali et al. 2010);
(c) an in-situ scour erosion evaluation probe (ISEEP) that provides field measurements on
the potential for scour in sands;
(d) multi-piezo-friction penetrometer (MPFP) for measuring soil-pile roughness and
quantification of soil-structure interaction effects on a site-specific basis (Hebeler and
Frost 2006).

(e) parallel seismic method (PSM) for determining the unknown lengths of existing pile
foundations, often needed in bridge replacement or retrofit projects (Niederleithinger
2012).
In contrast to these specifically-focused devices, some in-situ tests are quite versatile and can
provide multiple measurements on soil engineering parameters during a single sounding, such as
the seismic piezocone test (SCPTu) and seismic dilatometer test (SDMT), as depicted in Figure 3.
These are both hybrid devices which combine downhole geophysics with a static penetrometer or
probe that collect four to five independent readings with depth (Mayne and Campanella 2005).
Fig. 3. Hybrid in-situ geotechnical tests: seismic piezocone and seismic dilatometer
1.3 Seismic Piezocone Test

The SCPTu obtains the following measurements: qt = cone tip resistance, fs = sleeve friction, u2 =
dynamic porewater pressures, and Vs = shear wave velocity. All of the data are collected in the
field by computer, either analog or digital format, and can be processed on-situ immediately and/or
transmitted by wireless to the engineering office. The first three penetrometer readings are useful
in assessing geostratigraphy, layering, and soil type, as well as provide estimated pile shaft friction
and end-bearing resistances. Of additional value, the shear wave velocity provides a fundamental
stiffness of the ground via:
G0 = Gmax = t ∙Vs2 (1)

where G0 = initial tangent shear modulus and t = t/ga = total soil mass density, t = soil unit
weight, and ga = 9.81 m/s2 = gravitational constant. This is particularly important in calculations
involving the evaluation of pile displacements and axial load transfer distributions along the pile
length.
A representative SCPTu sounding performed in coastal plain sediments of Norfolk, Virginia
is shown in Figure 4. The upper 9 m of soils are comprised of recent Holocene deposits with sandy
and silty layers in the first 5 meters and a soft clay layer in the depth intervals from 5 to 9 m. From
depths below 9 m, the sounding penetrates into the Yorktown formation which a Miocene age
marine deposit composed of calcareous sandy clays to sandy clays. Generally, Vs readings
obtained by standard downhole testing (DHT) are taken at 1-m depth intervals, as shown by the
square dots in Figure 4d. Also presented in this figure are the results from special continuous
interval Vs recordings using the GT autoseis source which can generate a wavelet every 1 s (Ku et
al. 2013). This offers the opportunity for the field performance of the SCPTu to be as fast,
productive, and efficient as regular cone penetration testing (CPT).
Fig. 4. Representative seismic piezocone sounding in Norfolk, Virginia, USA
1.4 Seismic Flat Dilatometer

In a similar manner, the seismic flat dilatometer (SDMT) can provide a comparable number of
readings from a single sounding: p0 = contact pressure, p1 = expansion pressure, Vp = compression
wave velocity, and Vs = shear wave velocity. If addition information is desired about the soil
permeability and time rate of consolidation, then the SCPTu can also obtain a fifth reading, i.e. t50
= time rate of porewater dissipation, and the SDMT can collect tflex = time rate of the p0 reading.
Example results from SDMT are presented by Marchetti et al. (2008) and Amoroso et al. (2014).

2. INTERPRETATION OF CONE PENETRATION TESTS
2.1 Geoparameter Evaluation from CPTu
The interpretation of cone and piezocone penetration tests can be made on the basis of theoretical,
analytical, numerical, empirical, and statistical relationships (Lunne et al. 1997; Mayne 2007;
Schnaid 2009). Efforts at calibration of the interpretative CPT procedures rely on one or more of
the following: (a) matching field results with benchmark values obtained from laboratory tests on
undisturbed samples, (b) backcalculating parameters from full-scale foundation performance, (c)
1-g model testing in chambers; and (d) centrifuge testing using mini- and micro-penetrometers.
Advantages and shortcomings are associated with each of these approaches.
For evaluating the axial capacity of deep foundations, the CPTu must evaluate the following
geoparameters: (a) soil type; (b) unit weight; (c) effective friction angle; (d) stress history; (e)
undrained shear strength; and (f) lateral stress coefficient. Moreover, in assessing pile
displacements and axial load distributions, the small strain shear modulus (G0 = Gmax) plays an
important role. Each of these are briefly discussed in subsequent sections, as restricted to
uncemented sands and clays of low to medium sensitivities.
2.2. Soil Behavioral Type
Since soil samples are not normally collected during CPT, indirect methods for soil classification
are often used. These generally include: (a) approximate "rules of thumb"; (b) soil behavioral type
charts; and (c) probabilistic methods. The approximate rules suggest that sands are identified when
qt > 5 MPa and u2 ≈ u0, whereas intact clays occur when qt < 5 MPa and u2 > u0 (Mayne et al.
2002). For fissured overconsolidated clays, u2 readings are often negative. Probability-based
methods are discussed by Tumay et al. (2013).
The most popular methods are based on soil behavioral charts with popular favoring of the 9-
zone classification system (Lunne et al. 1997; 2009) that uses normalized piezocone readings: (a)
normalized tip resistance: Q = (qt - vo)/vo', (b) normalized sleeve friction: F (%) = 100∙fs/(qt -
vo); and normalized porewater pressure: Bq = (u2 - u0)/(qt - vo). Any of these indirect CPT soil
classification approaches should be cross-checked and verified for a particular geologic setting
before routine use in practice.
The development of a CPT material index (Ic) has been found advantageous in the initial
screening of soil types and helps to organize the sounding into 9 different zones of similar soil
response. In this case, the CPT index is found from (Robertson 2009):
I c  (3.47  log Qtn ) 2  (1.22  log Fr ) 2 (2)
where Qtn = [(qt - vo)/atm]/(vo'/atm)n = stress-normalized net cone resistance

n = exponent that is soil-type dependent: n = 1 (clays); ≈ 0.75 (silts); ≈ 0.5 (sands)
atm = atmospheric pressure (1 atm ≈ 1 bar = 100 kPa)
Initially, an exponent n = 1 is used to calculate the starting value of Ic (i.e., Qtn = Q) and then the
exponent is upgraded to:
n = 0.381∙Ic + 0.05∙(vo'/atm) - 0.05 ≤ 1.0 (3)


Fig. 5. Soil behavioral type zones and algorithms for zone identification by CPT
Then the index Ic is recalculated. Iteration converges quickly and generally only 3 cycles are
needed to secure the operational Ic at each depth. The soil zones and associated Ic values are
detailed in Figure 5. The sensitive soils of zone 1 can be screened using the following expression:
Zone 1: Qtn < 12 exp ( -1.4 ∙ Fr) (4)

The stiff soils of zone 8 (1.5% < Fr < 4.5%) and zone 9 (Fr > 4.5%) can be identified from:
1
Zones 8 and 9: Qtn  (5)
0.005( Fr  1)  0.0003( Fr  1) 2  0.002
Then, the remaining soil types are identified by the CPT material index: Zone 2 (organic clayey
soils: Ic ≥ 3.60); Zone 3 (clays to silty clays: 2.95 ≤ Ic < 3.60); Zone 4 (silt mixtures: 2.60 ≤ Ic <
2.95); Zone 5 (sand mixtures: 2.05 ≤ Ic < 2.60); Zone 6 (clean sands: 1.31 ≤ Ic < 2.05); and Zone
7 (gravelly to dense sands: Ic ≤ 1.31). The red dashed line at Ic = 2.60 represents an approximate
boundary separating drained (Ic < 2.60) from undrained behavior (Ic > 2.60).

2.3 Soil Unit Weight
Soil unit weight can be estimated from the CPT sleeve friction resistance (Mayne 2015):
t/w = 1.22 + 0.15 ∙ ln (100*fs/atm+0.01) (6)
where w = unit weight of water.
2.4 Effective Stress Friction Angle

The effective stress friction angle (') is one of the most important soil properties as it governs the
strength of the geomaterial, as well as affects soil-pile interface and pile side friction. While an
effective cohesion intercept (c') can also be considered, this is usually reserved for cemented or
bonded geomaterials or unsaturated soils and may lose its magnitude with time, ageing, or with
prolonged environmental changes.
For clean quartz to silica sands where porewater pressures are essentially hydrostatic (Bq = 0),
the following expression has been calibrated with triaxial compression test results from
undisturbed sand samples and normalized cone resistances, as presented in Figure 6 (Mayne 2007;
Robertson & Cabal 2015):
' (deg) = 17.6° + 11.0° log (Qtn) (7)
Fig. 6. Evaluation of effective friction angle in undisturbed sands from CPT

Fig. 7. Evaluation of effective friction angle in clays and silts from CPTu
In the case of soft to firm intact clays and silty clays, the effective friction angles is determined
from the normalized cone resistance and porewater pressure parameters (Senneset et al. 1989;
Mayne 2016), as shown in Figure 7. The exact solution when the angle of plastification  = 0 is
given as:
tan 2 (45   ' / 2)  exp(  tan  ' )  1
Q 
1  6  tan  '(1  tan  ' )  Bq

which can be approximately inverted into the form (Mayne 2007):
' = 29.5°∙Bq0.121 ∙ [0.256 + 0.336∙Bq + log Q ] (9)
This algorithm is specifically applicable for the following ranges of porewater pressure parameter
(0.1 < Bq < 1) and effective stress friction angles (20° < ' < 45°).
2.5 Stress History

The stress history can be characterized by an apparent yield stress or preconsolidation stress (p'),
as well as by its normalized and dimensionless form, OCR = p'/vo' = overconsolidation ratio. A
generalized approach for soils using net cone resistance has been formulated (Mayne 2015):
p' = 0.33∙(qt - vo)m' ∙ (atm/100)1-m' (10)

where m' = exponent depends on soil type: m' = 1 (intact clays); ≈ 0.85 (silts); ≈ 0.72 (sands), as
presented in Figure 8.
Fig. 8. Evaluation of yield stress or preconsolidation stress in soils from CPT
The exponent m' has also been calibrated with CPT material index:
0.28
m'  1 
1  ( I c / 2.65) 25 (11)
2.6 Undrained Shear Strength

For undrained loading of clays at constant volume, a temporary and transient condition occurs
which is represented by the undrained shear strength (su). This can be related to the effective stress
strength envelope (c' = 0; ') and stress history (i.e., OCR) in the form:
su = (sin'/2) ∙ (OCR)∙ vo' (12)
which corresponds to a simple shear mode.
2.7 Lateral Stress Coefficient

The horizontal geostatic state of stress is represented by the lateral stress coefficient, K0 = ho'/vo',
commonly referred to as the at-rest condition. The magnitude of K0 for soils that have been loaded
and unloaded can be approximately estimated from:

K0 = (1 - sin') ∙ OCRsin' (13)
The value of K0 finds applicability in assessing the pile side friction via the beta method, whereby
as a first approximation:  = K0 ∙ tan'.
2.8 Additional GeoParameters

If desired, additional engineering properties can be determined during SCPTu (e.g., Robertson and
Cabal 2015). For instance, the effective cohesion intercept (c') can also be determined from
plotting (qt ‐ vo) versus vo' (Mayne 2016). If piezo-dissipation measurements are taken, the
coefficient of consolidation (cvh) and permeability (k) can be assessed (Mayne and Campanella
2005). This can be useful for pile driving projects in clays and silts that require an estimate of time
for equilibrium of excess porewater pressures caused during installation, as well as for ground
modification projects where these data are used for calculating time-rate-of-consolidation and wick
drain spacings.
.
3. RELEVANCE OF SCPTu in DEEP FOUNDATION DESIGN
The analysis of the axial load-displacement-capacity response of deep foundations is usually
separated into two analytical components: (a) capacity; and (b) displacements. The SCPTu
provides sufficient data input to handle the requirements using conventional calculations using
limit equilibrium and plasticity solutions, as well as direct methods that are based on statistical
analyses of large foundation load test databases, most recently funded by the offshore energy
industry, including oil, gas, and wind.
3.1 Traditional Methods for Evaluating Side Friction and Toe Resistance
In common practice, pile shaft friction (rs) is often calculated using alpha methods for clays and
beta methods for sands (Brown et. al. 2010). The beta method has also shown applicable for both
clays and sands (O'Neill 2001; Fellenius 2016). Using the information from Section 2.7, a
generalized expression for shaft friction can be given by:
rs = CM ∙ CK ∙ K0 ∙ vo' ∙ tan' (14)
where CM = pile material factor = 1 (rough cast-in-place concrete); 0.9 (prestressed concrete)
0.8 (timber); and 0.7 (steel);
CK = installation factor = 0.9 (bored or augered); 1.0 (low displacement, e.g. H-pile or
open end pipe); and 1.1 (driven solid, e.g. prestressed concrete, closed-end pipe).
The evaluation of toe resistance (rt) is often taken from limit plasticity solutions with associated
shape and depth factors (Brown et al. 2010). For undrained toe response, the full value may be
attained, thus:
undrained: rt = Nc ' ∙ su (15)
where Nc' = 9.33 for a circular pile.

For drained toe response, the situation is more complex as the pile toe resistance increases with
toe displacement (Fellenius 2016). Thus, from a practical standpoint, only a portion of the
theoretical resistance will be realized:

drained: rt = vo' ∙ Nq' ∙ fx' (16)
where Nq' ≈ 0.77 ∙ exp ('/7.5°) = approximate expression for bearing factor
fx' = strain incompatibility factor = 0.1 (bored piles); 0.2 (jacked); 0.3 (driven)
3.2 Direct CPT methods for Axial Pile Capacity

In lieu of assessing soil parameters (t', su, K0, OCR) from CPT results that are input into
theoretical formulae, considerable research efforts have centered on Direct CPT methods whereby
the measured CPT results are scaled straightaway to provide unit shaft (rs) and unit toe (rt)
resistances. At least 35 different Direct CPT methods are available (Niazi and Mayne 2013).
A particular interesting and versatile approach is the UniCone Method (Eslami and Fellenius
1997) as it employs all three piezocone readings (qt, fs, u2) in its assessment. The unit shaft and
toe resistances are obtained from the effective cone resistance: qE = (qt - u2) according to;
rs = Cse ∙ (qt - u2) (17)
rt = Cte ∙ (qt - u2) (18)
where Cse is a coefficient for shaft friction based on soil type and Cte = toe resistance coefficent.
Figure 9 provides a quick summary overview of the approach and full details on UniCone are
given elsewhere (Fellenius 2016).
Fig. 9. Overview of UniCone Method (after Eslami and Fellenius 1997; Fellenius 2016)

A modified UniCone approach has been devised that uses the CPT material index (Ic) to assign
values to the Cse and Cte coefficients (Niazi and Mayne 2016). For SBTn zone 1 (sensitive clays),
the Cse coefficient is obtained from:
SBTn Zone 1: Cse = 0.074 - 0.004 ∙ [Qtn - 12 ∙ exp(-1.4∙F)] (19)
and for the remaining soil types, the shaft coefficient may be estimated from the following
expression:
SBTn Zones 2 to 9: Cse = 1 ∙ 2 ∙ 3 ∙ 10 [0.732∙Ic - 3.605] (20)
where 1 = pile type factor: 0.84 (bored piles), 1.02 (jacked), 1.13 (driven); 2 = load direction
factor: 1.11 (compression) and 0.85 (tension), and 3 = loading rate factor (1.0 for soils with Ic <
2.6 and for Ic > 2.6: 0.97 (stepped load) and 1.09 (constant rate of penetration).
The toe coefficient may be estimated from:
All SBTn Zones: Cte = 10 [0.325∙Ic - 1.218] (21)
The total axial compression capacity (Qult) is the sum of shaft capacity (Rs) plus toe capacity (Rt):
Qult = Rs + Rt = ∫ (As ∙ rs) dz + At ∙ rt (22)
where As = circumferential area of the pile at depth z and At = toe area of the pile.
3.3 Pile Displacements and Axial Load Transfer

Elastic continuum theory provides closed-form expressions for calculating pile top displacements
under axial loading, as well as detail the magnitude of axial load transfer with depth. These
solutions require an evaluation of the ground stiffness that can be expressed either in terms of an
equivalent Young's modulus (E) or shear modulus (G), since E = 2∙G(1+), where  = Poisson's
ratio (i.e., drained ' = 0.2 and undrained u = 0.5).
The simple case of a rigid pile situated in a single soil layer is depicted in Figure 10 showing
the expressions for axial pile displacement and proportions of load transfer to the toe and shaft.
The soil modulus can be constant with depth (i.e., homogeneous with E = 1), pure Gibson (E =
0.5) or generalized Gibson condition (E = EsM/EsL). More complex solutions for compressible
pilings situated with toes bearing on stiff strata are also available (e.g., Brown et al. 2010).
3.4 Approximate Nonlinear Modulus

The equivalent elastic soil modulus is nonlinear with applied load level, especially from the
nondestructive small-strain to middle-strain to strength ranges (Jardine et al. 2004). One simple
algorithm that finds application to pile foundation loading is a modified hyperbola that provides a
reduction factor to the initial small-strain elastic modulus (Mayne 2006):
E/Emax = 1 - (Qt/QtULT)g' (23)
where g' = fitted exponent (≈ 0.3 ± 0.1 for uncemented and non-structured soils).

Fig. 10. Elastic continuum solution for rigid axial pile in soil layer
The seismic piezocone test (SCPTu) thus finds special application to deep foundation analysis
because it provides sufficient data (qt, fs, and u2) for axial pile capacity calculations, as well as
supplying the necessary soil stiffness (Emax) for the evaluation of displacements and load transfer.
To illustrate the usefulness of the SCPTu results in axial pile evaluation, a case study from load
tests for the Wakota Bridge are presented.
4. CASE STUDY - WAKOTA BRIDGE, MINNESOTA
The ten-lane Wakota Bridge is located southeast of Saint Paul, Minnesota and was completed in
2010. The bridge enables interstate loop 494 to cross the Mississippi River. During the design
phase, load tests on both driven open-end and closed end steel pipe piles were conducted with test
piles having an outer diameter d = 0.457 m, an embedded length L = 32 m, and wall thickness t =
12.5 mm (Dasenbrock 2006). Both piles were loaded in axial compression, then afterwards loaded
in tension, using a static reaction frame arrangement.
Soil conditions at the site can be assessed via the SCPTu sounding presented in Figure 11,
indicating primarily firm sands with a few interbedded clay layers found at depths of 1, 3.5, 7 - 12,
17, and 22 - 27 m. The post-processing of the SCPTu provided direct estimates of the soil type via

Fig. 11. Seismic piezocone results at I-494 Wakota Bridge site, Saint Paul, Minnesota
material index (Ic), unit weight (t), effective friction angle ('), preconsolidation stress (p'), and
K0 profiles for input into beta method for pile shaft resistance (ave. rs = 65 kPa), as well as by
direct CPT methods using Unicone (ave. rs = 70 kPa) and Modified Unicone (ave. rs = 75 kPa).
Evaluation of the toe resistance determined rte = 3593 kPa by the Modified Unicone expression.
The evaluation of the two tests in axial compression are presented in Figure 12, while Figure
13 shows the two tests in tension. Overall, the agreement between the load-displacement responses
Fig. 12. Axial compression load tests at Wakota site: (a) open-end pile; (b) closed-end pile

Fig. 13. Axial tension load tests at Wakota site: (a) open-end pile; (b) closed-end pile
are comparable between the measured field load tests and those generated using the elastic
continuum solutions that rely on SCPTu data for input. For reference, the Euro criterion for
"capacity" is taken as that load corresponding to (s/d) = 10% is shown on the measured load test
curve.
5. DEVELOPMENTS IN GEOPHYSICS
5.1 Electromagnetics vs. Mechanical Wave Geophysics

Corresponding series of advancements in geophysics have also been made in the past two decades
that are of direct use and benefit to site investigations concerning deep foundations. The broad
categories of geophysics include: (a) electromagnetic wave techniques, and (b) mechanical wave
methods. Electromagnetic wave methods include ground penetrating radar (GPR), surface
resistivity surveys (SRS), and electromagnetic conductivity (EM). Mechanical wave approaches
include crosshole testing (CHT), downhole testing (DHT), and Rayleigh wave methods, also
referred to as surface wave techniques. The geophysical tests can be either invasive in boreholes
or direct push probes, or noninvasive where the sensors (geophones, accelerometers, or electrodes)
are place in patterned arrays or configurations at the surface of the ground. An extensive review
of geophysical methods is detailed by Wightman et al. (2003).
5.2 Rayleigh Wave Mapping

Of special mention in deep foundations applications is the utilization of Rayleigh wave methods
to map the degree of heterogeneity of shear wave velocities in the ground and present these as
subsurface profiles that are cross-sections across the project site. For instance, Figure 14 presents
the results of a series of multi-channel analysis of surface wave (MASW) tests using streamers
(movable plastic sleds) that contain the geophones and the individual surveys were taken at 1-m
horizontal intervals across the site. Each MASW took less than 1 second using a sledge hammer.
The streamer was pulled to the next location and the entire series took about 3 hours.

Fig. 14. Results of combined results of 120 MASW surveys (courtesy of Illmar Weemees)
The MASW results were combined to form the cross section shown in Figure 14 with a clear
indication of poor ground (Vs < 150 m/s), normal soil conditions (Vs ≈ 200 m/s), and very strong
ground conditions (Vs > 300 m/s). Such information would be valuable prior to selection of
locations of pile load tests, production pile installation, and/or necessary pile length
determinations, as well as beneficial to applications in extent of ground modification and other
construction activities.
6. CONCLUSIONS
Site characterization is an important component for the proper selection of deep foundation
alternatives. Several advances in geotechnical site investigation have been made that quantify
needs in foundation performance, including specialized probes for evaluation of scour potential,
thermal soil properties, pile friction roughness, and soil stiffness, as well as a suite of noninvasive
geophysics techniques. Of particular benefit is the utilization of hybrid tests such as seismic
piezocone and seismic dilatometer that collect information at two ends of the stress-strain-strength
curves in soils, namely the small-strain stiffness (Gmax and Emax) from the shear wave velocity and
the shear strength (max or su, c', and ') that are both needed in the assessment of the axial load-
displacement-capacity of deep foundations.
7. ACKNOWLEDGMENTS
The author gives thanks to Derrick Dasenbrock for sharing the MnDOT SCPTu and load test
results. Appreciation is given to ConeTec of Richmond, BC for support of GT in-situ research
activities.

8. REFERENCES
Akrouch, G.A., Briaud, J-L., Sanchez, M. and Yilmaz, R. 2016. Thermal cone test to determine
soil thermal properties. J. Geotechnical & Geoenvironmental Engrg. 142 (3): 04014085.
Amoroso, S., Monaco, P., Lehane, B.M. and Marchetti, D. 2014. Examination of the potential of
the SDMT to estimate in-situ stiffness decay curves. Soils and Rocks 37 (3): 177-194.
Ali, H., Reiffsteck, P., Baguelin, F., van de Graaf, H., Bacconnet, C. and Gourvès, R. 2010.
Settlement of pile using cone loading test: load settlement curve approach. Proc. 2nd Intl.
Symp. on Cone Penetration Testing (CPT'10), Huntingdon Beach, California: Paper ID 3-24.
Brown, D.A., Turner, J.P. and Castelli, R.J. 2010. Drilled shafts: construction procedures and
LRFD design Methods. Report FHWA-NHI-10-016, National Highway Institute, Federal
Highway Administration, Washington, DC: 972 p.
Dasenbrock, D.D. 2006. Assessment of pile capacity by static and dynamic methods to reconcile
design predictions with observed performance. Proc. 54th Minnesota Geotechnical
Engineering Conference, Edited by Labuz, J., Univ. of Minnesota, St. Paul: 95-108.
Davidson, J.L., Maultsby, J.P. and Spoor, K.B. 1999. Standard penetration test energy calibrations.
Report Number WPI 0510859 by Univ. Florida - Gainesville submitted to Florida Dept.
Transportation, Tallahassee: 123 pages.
Eslami, A. and Fellenius, B.H. 1997. Pile capacity by direct CPT and CPTu methods applied to
102 case histories. Canadian Geotechnical Journal 34 (6): 886-904.
Fellenius, B.H, 2016. Basics of Foundation Design, Electronic Edition, Sidney, British Columbia,
453 pages: www.fellenius.net
Finno, R.J., Gassman, S.L. and Calvello, M. 2000. The NGES at Northwestern University.
National Geotechnical Experimentation Sites (GSP 93), ASCE, Reston, Virginia: 130-159.
Gabr, M., Caruso, C., Key, A. and Kayser, M. 2013. Assessment of in-situ scour profile in sand
using a jet probe. ASTM Geotechnical Testing J. 36 (2): 1-11.
Hebeler, G.L. and Frost, J.D. 2006. A multi-piezo-friction attachment for penetration testing.
Geotechnical Engineering in the Information Technology Age (Proc. GeoCongress, Atlanta),
ISBN 9781604234909; ASCE, Reston, Virginia: 1-6.
Honeycutt, J.N., Kiser, S.E. and Anderson, J.B. 2014. Database evaluation of energy transfer for
cme automatic hammer SPT. J. Geotechnical & Geoenv. Engrg. 140 (1): 194-200.
Jardine, R.J., Gens, A., Hight, D.W. and Coop, M.R. 2004. Developments in understanding soil
behavior. Advances in Geotechnical Engineering (Proc. Skempton Conference), Thomas
Telford, London: 103-206.
Ku, T., Mayne, P.W, and Cargill, E. 2013. Continuous-interval shear wave velocity profiling by
auto-source and seismic piezocone tests. Canadian Geotechnical J. 50 (1): 382–390.
Lunne, T., Robertson, P.K. and Powell, J.J.M. 1997. Cone Penetration Testing in Geotechnical
Practice. EF Spon/Blackie Academic, London, 352 p.
Marchetti, S., Monaco, P., Totani, G., and Marchetti, D. (2008). In-situ tests by seismic dilatometer
(SDMT). From Research to Practice in Geotechnical Engineering (Proc. GeoCongress 2008,
GSP 180, New Orleans), ASCE, Reston, Virginia: 292-311.
Mayne, P.W., Christopher, B., Berg, R., and DeJong, J. 2002. Subsurface Investigations -
Geotechnical Site Characterization. Publication No. FHWA-NHI-01-031, National Highway
Institute, Federal Highway Administration, Washington, D.C., 301 p.

Mayne, P.W. and Campanella, R.G. 2005. Versatile site characterization by seismic piezocone
tests. Proceedings, 16th International Conference on Soil Mechanics & Geotechnical
Engineering, (ICSMGE, Osaka), Vol. 2, Millpress, Rotterdam: 721-724.
Mayne, P.W. 2007. NCHRP Synthesis 368 on Cone Penetration Test. Transportation Research
Board, National Academies Press, Washington, D.C., 118 p. www.trb.org
Mayne, P.W. 2015. Keynote lecture: In-situ geocharacterization of soils in the year 2016 and
beyond. Advances in Soil Mechanics, Vol. 5: Geotechnical Synergy (Proc. 15th PCSMGE,
Buenos Aires), IOS Press, Amsterdam: 139-161.
Mayne, P.W. 2016. Evaluating effective stress parameters and undrained shear strengths of soft-
firm clays from CPTu and DMT. Australian Geomechanics J. 51 (4): 27-55.
Niazi, F.S. and Mayne, P.W. (2013). Cone penetration test based direct methods for evaluating
static axial capacity of single piles. Geot. and Geological Engrg. 31 (4), Springer: 979-1009.
Niazi, F.S. and Mayne, P.W. (2016). CPTu-based enhanced UniCone method for pile capacity.
Engineering Geology 212, Elsevier: 21-34.
Niederleithinger, E. 2012. Improvement and extension of the parallel seismic method for
foundation depth measurement. Soils and Foundations 52 (6): 1093-1101.
O'Neill, M.W. 2001. Side resistance in piles and drilled shafts (34th Terzaghi Lecture). J.
Geotechnical & Geoenvironmental Engineering 127 (1): 3-16.
Robertson, P.K. 2009. Cone penetration testing: a unified approach. Canadian Geotechnical J. 46
(11): 1337-1355.
Robertson, P.K. and Cabal, K.L. 2015. Guide to Cone Penetration Testing for Geotechnical
Engineering, 6th Edition, Gregg Drilling & Testing, Inc., Signal Hill, California: 142 p.
Schnaid, F. 2009. In-Situ Testing in Geomechanics: The Main Tests. Taylor & Francis Group,
London, 322 p.
Seed, H.B., Tokimatsu, K., Harder, L.F. and Chung, R.M. 1985. The influence of SPT procedures
in soil liquefaction resistance evaluations. J. Geotechnical Engineering 111 (12): 1425–1445.
Senneset, K., Sandven, R. & Janbu, N. 1989. Evaluation of soil parameters from piezocone tests.
In-Situ Testing of Soil Properties for Transportation, Transportation Research Record 1235.
National Research Council, Washington DC: 24-37.
Sjoblom, D., Bischoff, J., and Cox, K. 2002. SPT energy measurements with the PDA". Proc. 2nd
Intl. Conf. on the Application of Geophysical and NDT Methodologies to Transportation
Facilities & Infrastructure, Conference sponsored by FHWA, TRB, and CALTRANS.
Skempton, A.W. 1986. Standard penetration test procedures and effects in sands of overburden
pressure, relative density, particle size, ageing, and overconsolidation. Geotechnique 36 (3):
425-447.
Tumay, M.T., Hatipkarasulu, Y., Marx, E.R. and Cotton, B. 2013. CPT/PCPT-based organic
material profiling. Proc. 18th ICSMGE, Paris: 633-636.
Wightman, W.E., Jalinoos, F., Sirles, P. and Hanna, K. 2003. Application of Geophysical Methods
to Highway Related Problems. Contract Report No. DTFH68-02-P-00083, Federal Highway
Administration, Washington, D.C., 742 p.

Effects of Installation Processes on the
Axial Capacity of Pile Foundations in Sand
Prezzi, M.(1) and Basu, P.(2)

(1)
Purdue University, West Lafayette, IN, USA, <mprezzi@purdue.edu>
(2)
Indian Institute of Technology Bombay, Powai, MH, India, <pbasu@civil.iitb.ac.in>
ABSTRACT. The axial capacity of pile foundations in sand is affected by the method of
installation. Pile driving (or jacking) causes significant changes in soil state (the density of the
sand below and in the vicinity of the pile increases). However, at any given depth, the frictional
resistance of displacement piles degrades due to release of lateral stresses acting on the pile shaft.
Therefore, in addition to stress state and density, design methods for displacement piles in sand
have incorporated a shaft resistance degradation factor into the design equations. Accounting for
plug formation and evolution during driving and loading of partial-displacement piles (e.g.,
open-end pipe piles, H piles) are important. Accordingly, design methods have added the
incremental filling ratio and/or plug length ratio as input variables. This paper presents some of
the current pile design methods and discusses recent advances in research that have led to
significant improvement in pile design methods.
1. INTRODUCTION

The response of piles to axial loading depends, to a significant extent, on the installation
methods employed. Accordingly, piles are often classified according to the level of soil
displacement imposed on the in situ soil by the installation method. Non-displacement piles
(e.g., bored piles or drilled shafts), which are constructed by removing soil from the ground
down to the desired depth with partial augers and placing concrete and reinforcement into the
excavated hole, are on one end of the spectrum of pile response to loading. Full-displacement
piles (e.g., closed-end pipe piles or precast reinforced concrete piles) are on the other end of the
spectrum of pile response to loading. Figure 1 shows how the main types of piles are classified
according to the soil displacement induced by their installation processes.
Full-displacement piles are mostly installed by driving or jacking (monotonic or multi-stroke
jacking) them into the ground. The installation of full-displacement piles both displaces laterally
the sand in the vicinity of the shaft of the pile and preloads the soil at the base; both of these
effects lead to changes in void ratio and stress state. Installation of a displacement pile in sand
causes densification of the sand within an annular zone surrounding the pile with fast dissipation
of the excess pore pressure generated during installation. In addition, degradation of frictional
resistance along the pile shaft due to shear loading and unloading cycles (induced by hammer
blows or jacking strokes) is caused by the contractive tendency of the sand immediately adjacent
to the pile shaft. As shown in pile installation simulations, the sand in this zone is subjected to a
shear loading cycle that generates a dilative response, leading to an increase in void ratio and
thus setting up a tendency for contractive response in subsequent shearing cycles. Such
contractive tendency induces a release of the lateral stress acting on the pile shaft and a reduction
in shaft resistance (Figure 2). Friction degradation decreases with the installation depth and
needs to be considered in design. Note that pipe piles may be installed using vibratory hammers

(this is not part of the scope of this paper), thus inducing dynamic loading (to the ground)
different from that induced by jacking and impact driving.
Pile Types
Non-displacement Small or Partial Displacement Full-displacement
! Drilled Shafts ! H Piles ! Driven Piles

?
! Continuous
! Open-end
Pipe Piles
?
! Drilled
! Jacked Piles
Soil is removed from the
ground and the created void Flight Auger (in some soils) Displacement
is filled with concrete Piles (DD) Piles
Unit shaft or base resistance
Fig. 1. Effect of pile installation on soil displacement and unit shaft and base resistances.
Initial stage Just after pile tip passes element A (Expansion of a cavity)
Cavity limit Pressure σʹh1 : σʹh1 >> σʹh0

Dilation Contraction
σʹh0 A σʹ h1
(a) (b)
Contraction due Expansion leading to reduction of

to shearing radial stress/support/ stiffness
Normal Stress σʹh2< σʹh1
σʹ h2
τs Normal Stress σʹ h2 reduces with increasing
number of shear cycles: Friction degradation
Formation of
shear band (c)
Fig. 2. Mechanics of friction degradation along the shaft of a displacement pile: (a) initial state at
point A within the ground (b) pile tip passes point A – a cavity is created within the ground, and
(c) shearing associated with installation and/or loading.
Because of these installation differences, a displacement pile subjected to an axial load

shows a stiffer response than a non-displacement pile of the same geometry installed in the same
soil profile. Data from side-by-side axial load tests on instrumented piles show that the

measured unit shaft resistance at any given depth is greater for displacement piles than for non-
displacement piles, as expected.
Several types of auger piles are currently available in the market. Auger piles are installed by
using a continuous-, segmented- or partial-flight auger to drill into the ground, with the installation
process producing different levels of soil displacement depending on the geometry of the drilling
tool used. Continuous-flight-auger (CFA) or auger cast-in-place (ACIP) piles, which are installed
by introducing the CFA auger to the desired depth, pumping high slump concrete through the
hollow-stem of the auger all the way to the base of the pile, ascending the drilling tool from the
borehole slowly as concrete is placed, and then pushing down the reinforcement into the concrete,
are generally associated with small to no soil displacement and are often classified as non-
displacement piles. Another class of auger piles, commonly referred to as drilled displacement
(DD) piles in the U.S. and screw piles in Europe, are being frequently used in practice (Brown and
Drew 2000; Brown 2005; Prezzi and Basu 2005; Basu et al. 2010). DD piles are rotary
displacement piles installed by inserting a helical, partial-flight auger (segment of an auger) into
the ground with both a vertical force and a torque. In contrast to the CFA and ACIP piles, there is
lateral displacement (with minimal spoil) of the in situ soil during installation of DD piles. The
cylindrical cavity created by the installation process is filled with grout or concrete. The
installation of DD piles induces complex changes to the stress state of the soil and produces greater
soil displacement than those produced by drilled shafts, CFA or ACIP piles.
There are other types of piles (e.g., open-end pipe piles, H piles) that show behavior
intermediate between non-displacement and full-displacement piles. These piles are often called
partial-displacement piles. The level of soil displacement during the installation of open-end pipe
piles and H piles depends on plug formation and evolution during driving. If driven under fully-
plugged conditions, the load response of these piles may approach that of full-displacement piles;
however, if driven under fully-unplugged conditions, their response to loading is closer to that of
non-displacement piles. Many factors affect plug formation, such as pile diameter, soil type, void
ratio and stress state, and presence of very dense layers in the soil profile. Open-end pipe piles are
used whenever the soil profile conditions are such that it would be difficult to drive a closed-end
pipe pile to the required depth. Offshore, open-end large-diameter pipe piles, driven mostly under
unplugged conditions, are often used to carry the large axial and lateral loads of fixed oil platforms.
Given the wide range of soil displacement that is possible during pile foundation installation
depending on the type of pile and soil profile conditions, the challenge for geotechnical engineers
is to capture in simple design equations, that can be used easily in practice, all the factors that
affect the response of different piles to axial loading. In addition to the factors mentioned above
that directly affect pile resistances, it is also important to be able to estimate shaft resistance
degradation and pile setup; both of which are affected by the pile installation processes. In this
paper, we review design methods for piles in sand and discuss recent advances in the development
of design equations that improve the accuracy of axial pile resistance calculations.
2. DESIGN METHODS FOR SINGLE PILES

2.1 General Pile Design Framework
Figure 3 shows the axially-loaded design problem for a pile of diameter B and length L. The
ultimate pile resistance Qult is expressed as:

Qult = Qb,ult + QsL (1)

where Qb,ult = ultimate base resistance
QsL = limit shaft resistance.
Qult = QsL +Qb,ult
Pile head
Unit limit shaft

resistance due to
Pile friction or adhesion:
length L q
sL
Pile width or
B
diameter B
Pile base Ultimate unit base

(or tip) resistance: q
b,ult
Fig. 3. Axial capacity of pile foundations
The ultimate resistance Qult is associated with an ultimate load criterion typically defined in
many design methods as the load corresponding to a relative settlement w/B at the pile head of
10%. This has been done to account for the fact that a structure supported by piles will
experience the same settlement as the head, not the base, of the piles. Design methods are
developed based on a specific ultimate load criterion, but simulations have the advantage that
any settlement level may be considered (either with respect to the head or the base of the pile), so
different ultimate load criteria or serviceability limit states may also be checked in design.
The ultimate base and limit shaft resistances are given by:
n
QsL = ∑ qsLi Asi (2)
i =1
Qb,ult = qb,ult Ab (3)
where the subscript i represents a soil layer (i = 1, 2, 3, …) for which the shaft resistance is to be
calculated, n is the total number of layers in the soil profile that are crossed by the pile, qsLi and
qb,ult are the unit limit shaft and unit ultimate base resistances, Ab is the projected plan area of the
pile base and Asi is the surface area of the pile shaft within the ith layer.
The unit shaft and base resistances can be estimated using indirect and direct design
methods. Indirect design methods require that fundamental soil variables of interest (e.g., relative
density and critical-state friction angle for sands, and undrained shear strength and critical-state
and residual-state friction angles for clays) be estimated first either from laboratory or in situ test
data. Direct design methods correlate field test data (e.g., cone resistance qc or SPT blow count
number N) directly with the unit shaft and base resistances of different piles. The focus of this
paper is mainly on indirect and CPT-based design methods applicable to sand, as discussed next.

Effective stress analysis is used to calculate the unit shaft and base resistances in sand. The
fundamental property-based design equation for the unit limit shaft resistance of a single pile in
sand is:
qsL = Kσ vʹ0 tan δ = βσ vʹ0 (4)
where σʹv0 is the initial in situ vertical effective stress in the center of the layer where qsL is
calculated, δ is the sand-pile interface friction angle (it is usually expressed as a function of the
critical-state friction angle ϕc of the sand and pile surface roughness), K is the coefficient of
lateral earth pressure. The effective stress analysis expressed in equation (4) is referred to as the
β method ( β = K tan δ ) in the pile literature as well. The fundamental design equation for the
unit ultimate base resistance of a single pile in sand is:
qb ,ult = N q ,ultσ vʹ0

(5)
where Nq,ult is a bearing capacity factor that is linked to the ultimate load criterion.
2.2 Non-displacement Piles

One of the main advantages of drilled shafts is that they can be of large diameter (greater than
1m in diameter), and thus a single pile can carry a large axial column load as well as a lateral
load without the need of a pile cap. However, installation of drilled shafts may be difficult in
some soil profiles that are collapsible and require the use of drilling mud or other means to
maintain the borehole stable.
In the Purdue method (Loukidis and Salgado 2008; Han et al. 2017), design equations for the
unit resistances were developed based on results of 3D finite element analyses (FEAs) performed
using an advanced, bounding-surface, critical-state based sand model:
⎡ K DR ⎡ ⎛ σ v 0 ' ⎞⎤ ⎤
⎢1.5− 0.35ln ⎜ ⎟⎥
100 ⎝ p A ⎠⎦⎥ ⎥
qsL = ⎢ 0.3 K0 −0.4 0.67e ⎣⎢ σ vʹ0 tan φc (6)
⎢e 0 ⎥
⎣ ⎦
and
1.83 0.4
qb ,ult qb ,10% ⎛ D ⎞ ⎛ σ h' 0 ⎞
= = 62 ⎜ R ⎟ ⎜ ⎟ (7)
pA pA ⎝ 100% ⎠ ⎝ pA ⎠
where K0 = coefficient of earth pressure at-rest, pA = reference stress (= 100 kPa) and DR =
relative density of the sand (%) and σʹh0 is the initial in situ horizontal effective stress at the pile
base.
Eq. 6 shows the direct dependence of qsL on the sand relative density and stress state (K
increases with increasing sand density and decreasing initial vertical effective stress) and the
critical-state friction angle, which is an intrinsic variable of the sand (ϕc is used instead of δ in

Eq. 6 because shearing occurs within the sand through formation of a shear band for rough pile-
soil interfaces, as is the case for cast-in-place piles); this direct dependence is something that
other methods are unable to capture (some methods prescribe the use of a single value for K,
irrespective of sand density and stress state). The results of the FEAs showed that the unit
ultimate base resistance is independent of pile diameter when the pile slenderness ratio L/B is
less than 50 and increases with increasing sand density, as expressed in Eq. 7.
Alternatively, whenever site investigation includes CPTs and cone resistance is available at
the base of the pile, qb,ult can be obtained from (Salgado 2008; Salgado et al. 2011):
qb,ult = qb ,10% = 0.23e−0.0066 D qcb,avg

R
(8)
The cone resistance and relative density of the sand are further related by (Salgado and Prezzi
2007):
⎛q ⎞ ⎛σʹ ⎞
ln ⎜ c ⎟ − 0.4947 − 0.1041φc − 0.841ln ⎜ h 0 ⎟
p ⎝ p A ⎠ ≤ 100%
DR (%) = ⎝ A ⎠ (9)
⎛σʹ ⎞
0.0264 − 0.0002φc − 0.0047 ln ⎜ h 0 ⎟
⎝ pA ⎠

2.3 Displacement Piles
Simulation of the installation of displacement piles is difficult as it requires use of not only an
advanced constitutive model for the sand but also advanced numerical techniques that are able to
handle the large-deformations and rotations induced in the sand during driving or jacking.
Research on developing the capabilities for simulating large-deformation problems (cone
penetration and pile installation) using FEM and the Material Point Method (MPM) is underway.
Many design methods are available in the literature for driven closed-end pipe piles. In the
Purdue method, following the general expression proposed by Randolph (2003), the unit shaft
and base resistances are estimated as (Han et al. 2016; Salgado 2008; Salgado et al. 2011):
h
qsL = 0.2 + (0.02qc / σ vʹ0 − 0.2) exp(−0.05 )σ vʹ0 tan δ c (10)
B

and
qb ,ult = [1 − 0.0058DR ]qcb,avg (11)
where h is the distance from the center of the layer under consideration to the pile base, δ c is the
sand-pile interface friction angle, and qcb ,avg is the average cone resistance within a zone B above
and 1.5B below the base of the pile. Note that the exponential term in Eq. 10 is a frictional
resistance degradation term that is equal to 1 at the base of the pile.
The unit shaft and base resistances are estimated from the cone resistance qc in the ICP
method (Jardine et al. 2005) as:

qsL = (σ rcʹ + 2GΔr / R) tan δ c
0.13 −0.38
⎛σʹ ⎞ ⎛ ⎡ h ⎤⎞
σ rcʹ = 0.029 Fload qc ⎜ v 0 ⎟ ⎜ max ⎢ R ,8⎥ ⎟
⎝ pA ⎠ ⎝ ⎣ ⎦⎠
−1
G = qc ⎣⎡0.0203 + 0.00125η − 1.216 ×10−6η 2 ⎤⎦
−0.5
η = qc ( p Aσ vʹ0 ) (12)
and
qb,ult = max [0.3,1 − 0.5log( B / BCPT )] qcb,avg (13)
where R is the radius of the pile, Δr is equal to 0.02mm for lightly rusted steel piles, B is the
diameter of the pile and BCPT is the diameter of the cone. The factor Fload is equal to 1 for
compressive loading and 0.8 for tensile loading. In the ICP method, qcb,avg is the qc value
averaged within a zone 1.5B above and 1.5B below the pile base.
The UWA method (Lehane et al. 2005) has a formulation for the unit shaft resistance that is
somewhat similar to that of the ICP method and also includes a factor on the shaft resistance that
accounts for the difference in compressive and tensile loading:
f ⎛ 4GΔr ⎞
qsL = ⎜ σ rcʹ + ⎟ tan δ c
fc ⎝ B ⎠
−0.5
⎡ ⎛ h ⎞⎤
σ rcʹ = 0.03qc ⎢ max ⎜ , 2 ⎟ ⎥ (14)
⎣ ⎝ B ⎠⎦
−0.75
⎡ q /p ⎤
c A
G / qc = 185 ⎢ 0.5 ⎥
⎢⎣ (σ vʹ0 / p A ) ⎥⎦
and
qb ,ult = 0.6qcb ,avg (15)
where f / fc = 1 for compressive loading, f / fc = 0.75 for tensile loading and Δr = 0.02mm. In the
UWA method, qcb,avg is calculated following the Dutch averaging technique (Schmertmann 1978;
de Kuiter and Beringen 1979) in which the CPT cone resistance is averaged over a zone
extending from 8B above the pile base to 0.7B to 4B below the pile base, depending on whether
qc increases or decreases below the pile base. Both the ICP and UWA methods as well as the
Purdue method have a degradation factor that is a function of h/B.
For piles jacked in sand, shaft resistance degradation was considered by Basu et al. (2011)
with a term that is a function of the number of jacking strokes N. The following expressions were
obtained from the FEAs performed for monotonic installation of a jacked pile (N = 1):

⎛ K ⎞ ⎡ DR ⎧⎪ ⎛ σ vʹ0 ⎞ ⎫⎪⎤
⎜ ⎟ = 2.27 exp ⎢ ⎨2.5 − 0.55ln ⎜ ⎟ ⎬⎥ (16)
⎝ K 0 ⎠ N =1 ⎣⎢100 ⎩⎪ ⎝ p A ⎠ ⎭⎪⎦⎥
for isotropic sand fabric, and
⎛ K ⎞ ⎡ DR ⎧⎪ ⎛ σ vʹ0 ⎞ ⎫⎪⎤
⎜ ⎟ = 0.93exp ⎢ ⎨2.8 − 0.45ln ⎜ ⎟ ⎬⎥ (17)
⎝ K 0 ⎠ N =1 ⎣⎢100 ⎪⎩ ⎝ p A ⎠ ⎪⎭⎦⎥
for anisotropic sand fabric.
With (K/K0)N=1, the ratio K/K0 for a number of jacking strokes N can be calculated from:
⎛ K ⎞ ⎛ K ⎞
⎜ ⎟ = µN ⎜ ⎟ (18)
⎝ K0 ⎠ N ⎝ K 0 ⎠ N =1
The degradation coefficient µN in Eq. (18) is an exponential function of N:
{
µ N = A1 + (1 − A1 ) exp − A2 ( N − 1)
A3
} (19)
where A1, A2, and A3 are coefficients that for an initially isotropic sand fabric are given by:
A1 = 0.02
0.3
⎛ p ⎞
A2 = 0.2 ⎜ A ⎟ (20)
⎝ σ vʹ0 ⎠
1.1
⎛D ⎞
A3 = 1.0 − 0.4 ⎜ R ⎟
⎝ 100 ⎠
and, for an initially anisotropic sand fabric, these coefficients are:
A1 = 0.02
0.3
⎛ p ⎞
A2 = 0.25 ⎜ A ⎟ (21)
⎝ σ vʹ0 ⎠
2
⎛D ⎞
A3 = 1 − 0.3 ⎜ R ⎟
⎝ 100 ⎠

2.4 Partial-displacement Piles
2.4.1 Open-end Piles
Assessment of plug formation during driving is necessary in the design of open-end pipe piles.
Current methods include specific parameters, such as the incremental filling ratio IFR and the
L
plug length ratio PLR = plug , in the design equations. Figure 4 shows an open-end pile of outer
L
diameter Bo, inner diameter Bi, and a plug length of Lplug at the end of driving.
Fig. 4. Open-end pipe pile design problem.
In the Purdue method (Paik and Salgado 2003, Paik et al. 2003, Paik et al. 2004), which is
based on model pile tests performed in a calibration chamber, the unit limit shaft resistance qsL is
calculated as:
qsL = K0 β ( 7.2 − 4.8PLR )σ vʹ0 tan δ c (22)

where β = 1.0 for dense sands (75%<DR≤100%), 0.4 for medium sands (40%<DR≤75%), and
0.22 for loose sands (0%<DR≤40%), and δ = 0.85ϕc. As proposed by Lehane and Randolph
(2002), the ultimate base resistance Qb,ult can be calculated as the summation of the contribution
to the base resistance from the annular steel area As and the area of the soil plug Aplug:
Qb ,ult = As qc + Aplug qb ,ultND (23)
where qb,ultND is the unit ultimate base resistance of a non-displacement pile (Lee and Salgado
1999).
The unit limit shaft resistance in the ICP method for open-end pipe piles is obtained from:
qsL = (σ rcʹ + Δσ rdʹ ) tan δ c

0.13 −0.38
⎛σʹ ⎞ ⎛ ⎡ h ⎤⎞
σ rcʹ = 0.029 Fload qc ⎜ v 0 ⎟ ⎜ max ⎢ R* ,8⎥ ⎟
⎝ pA ⎠ ⎝ ⎣ ⎦⎠
R* = (R 2
o − Ri2 ) (24)
Δσ rdʹ = 2GΔr / Ro
−1
G = qc ⎣⎡0.0203 + 0.00125η − 1.216 ×10−6η 2 ⎤⎦
−0.5
η = qc ( p Aσ vʹ0 )
where R* is the equivalent radius of a closed-end pipe pile with the same steel area as the open-
end pipe pile with outer radius Ro and inner radius Ri. Similar to what was proposed for closed-
end pipe piles, the factor Fload is equal to 1 for compressive loading and 0.8 for tensile loading.
However, the ICP method recommends that qsL be further multiplied by 0.9 for open-end piles
loaded in tension. The unit ultimate base resistance depends on whether the open-end pipe pile
behaves as fully plugged or unplugged. For unplugged piles:
Qb,* ult = qcb,avg A* (25)
*
π ( B02 − Bi2 )
where A = . Open-end pipe piles are considered to be fully plugged when:
4
qc
Bi < 0.02 ( DR − 30 ) or Bi < 0.083 BCPT (26)
Pa
and
qb,ult = max [0.15, 0.5 − 0.25log( Bo / BCPT )] qcb,avg

⎡ π Bo2 * ⎤ (27)
Qb,ult = max ⎢ qb ,ult , Qb ,ult ⎥
⎣ 4 ⎦

According to the UWA method (Lehane et al. 2005; Xu et al. 2008), the unit resistances are
calculated as follows:
f ⎛ 4GΔr ⎞
qsL = ⎜ σ rcʹ + ⎟ tan δ c
fc ⎝ Bo ⎠
−0.5
* 0.3
⎡ ⎛ h ⎞⎤
σ rcʹ = 0.03qc ( A rs) ⎢ max ⎜ , 2 ⎟ ⎥
⎣ ⎝ Bo ⎠ ⎦
0.2
Ars* = 1 − IFR ( Bi2 Bo2 ) , IFRmean ≈ min ⎡1, ( Bi 1.5m ) ⎤
⎣ ⎦
−0.75
⎡ q /p ⎤
c A
G / qc = 185 ⎢ 0.5 ⎥
⎢⎣ (σ vʹ0 / p A ) ⎥⎦ (28)
and
qb,ult = ⎡⎣0.15 + 0.45 Arb* ⎤⎦ qcb,avg
0.2
(29)
Arb* = 1 − FFR ( Bi2 Bo2 ) , FFR ≈ min ⎡1, ( Bi 1.5m ) ⎤
⎣ ⎦
where f / fc = 1 for compression and 0.75 for tension, and FFR is average IFR for the last 3B
depth of pile penetration.
2.4.2 H Piles
Design methods for displacement piles are also used to design H piles but most of them do not
provide specific recommendations for H piles. The limit shaft capacity of H piles depends on
whether under loading a fully plugged mode is operative or not. If the soil in the space between
the flanges forms a plug and becomes an integral part of the pile during loading then the H pile
outside perimeter (the outer boundaries of the H pile cross section are used for calculation of the
shaft surface area) and the gross cross-sectional area (flange width × depth) have been proposed
for use in shaft and base resistance calculations, respectively. When a plug does not form, which
is most likely the case for H piles in sand under static loading condition, shaft resistance may be
mobilized along the entire surface area of the H pile (full H pile-soil interface contact perimeter),
with the actual cross-sectional steel area used in base capacity calculations. In the case of sandy
soils, the use of the total H pile-soil interface area is often recommended for shaft capacity
calculations, and the actual H pile cross-sectional area for base capacity calculations (Tomlinson
1987; Salgado 2008). However, plug formation in H piles and pipe piles needs to be further
studied to improve or develop design methods that are more accurate.
The ICP method provides specific design recommendations for H piles. The H pile cross-
sectional base area Ab is calculated as (see Figure 5):
Ab = As + 2 X p (D − 2t ) (30)
where As = steel area of the H pile, D = depth, t = thickness, Bf = width of flange and
Xp = Bf /8 if Bf/2 < (D – 2t) < Bf (31)

Xp = Bf 2/[16(D – 2t)] if (D – 2t) ≥ Bf (32)
As is the
area of the D
hatched
section
B
f
Fig. 5. Geometry for H pile design problem.
The unit ultimate base resistance is taken as the average cone resistance at the pile base
(qb,ult/qcb,avg = 1). For shaft resistance calculations, Jardine et al. (2005) recommended the use of
the H pile outer perimeter, with qsL calculated as:
0.13 −0.38
⎪⎧ ⎛ σ vʹ ⎞ ⎡ ⎛ h ⎞⎤ Δr ⎪⎫
qsL = ⎨0.029 Fload qc ⎜ ⎟ ⎢ max ⎜ * , 8 ⎟ ⎥ + 2G * ⎬ tan δ (33)
⎩⎪ ⎝ pA ⎠ ⎣ ⎝R ⎠⎦ R ⎪
⎭
where Fload = 0.8 for tension and 1.0 for compression, and R* = [Ab/π]0.5.
Seo et al. (2009) reported the results of a pile load test performed on a H pile installed in
multilayered profile and indicated that the use of reduction factors to the shaft capacity of H piles
using displacement pile methods seems appropriate when full H pile-soil interface contact
perimeter is assumed in calculations [note that reduction factors between 0.5 and 0.8 have been
suggested; e.g., Meyerhof 1976 and Fleming et al. 1992].
2.4.3 Drilled Displacement Piles

Many different types of DD piles are used throughout the world [Auger Pressure-Grouted
Displacement (APGD), Atlas, De Waal, Fundex, Olivier, Omega, SVB, and SVV]. The main
difference between DD piles and drilled shafts and CFA piles is that the installation of DD piles
is done using a drilling tool consisting of a partial-flight auger and a displacement body that is
advanced into the ground using both a vertical force and a torque.
Most of the design methods available for DD piles (e.g., the method by Bustamante and
Gianeselli 1993, 1998 is based on the results of 24 load tests performed on Atlas piles) used for
calculation of shaft capacity were developed based on pile load test data with only the load at the
pile head known (when pile load tests are performed on non-instrumented piles, it is not possible
to separate the shaft rand base resistances).
Recently, Basu et al. (2014) performed 1D quasi-axisymmetric FE simulations of DD pile
installation and loading in sand. The DD pile installation consisted of four stages (Figure 6a-d):
(i) cavity expansion, (ii) downward and torsional shearing along the drilling tool-soil interface
(simulating the insertion of the drilling tool into the ground), (iii) upward shear unloading along
the borehole (simulating the removal of the drilling tool), and (iv) shaft loading (simulating an

axial load test). Sand was modelled using the Loukidis and Salgado (2009) advanced, bounding-
surface plasticity model. The assumptions of the FEAs are that there is no caving in or bulging of
the borehole wall during concrete (or grout) placement and hardening, the pile shaft is perfectly
rigid (a reasonable assumption given the large concrete-to-soil stiffness ratio), and any lateral
expansion of the shaft due to any vertical loading is negligible.
Initial stage NOTE:

Torsional
Stage 1: Stage 2: Stage3: shear stress
Insertion of Removal of Limit loading acting on the
drilling tool drilling tool along the soil element
pile shaft in Stage 1 is
not shown in
the figure
uθ
σ′r0 A τ1 τ2 τ3
uz uz
σ′r1 σ′r2 σ′r3 tan η =
η uθ
(a) (b) (c) (d) (e)
Fig. 6. Idealization of DD pile installation and loading: (a) initial state, (b) insertion of the
drilling tool into the ground (cavity expansion followed by vertical downward and torsional
shearing), (c) removal of the drilling tool (upward shear unloading along the borehole wall), (d)
axial loading of the pile(simulating an axial load test), and (e) installation parameter η.
The following design equations for the lateral earth pressure coefficient K [which are to be
used for calculation of qsL following Eq. 4] that take into account the initial soil state and the rate
of penetration of the drilling tool into the ground during pile installation resulted from the FEAs:
0.11
K ⎛σʹ ⎞ ⎡⎛ D (%) ⎞ ⎧⎪ ⎛ pA ⎞ ⎫⎪ ⎤
= 0.33 ⎜ v0 ⎟ exp ⎢⎜ R ⎟ ⎨3.59 + 0.53ln ⎜ ʹ ⎟ ⎬ (1 − 0.11tan η )⎥ (34)
K0 ⎝ pA ⎠ ⎢⎣⎝ 100 ⎠ ⎪⎩ ⎝ σ v0 ⎠ ⎪⎭ ⎥⎦
for initially anisotropic sand fabric, and
−0.11
K ⎛σʹ ⎞ ⎡⎛ D (%) ⎞ ⎧⎪ ⎛ pA ⎞ ⎫⎪ ⎤
= 1.30 ⎜ v0 ⎟ exp ⎢⎜ R ⎟⎨ 2.91 + 0.38ln ⎜ ⎟⎬ (1 − 0.12 tan η ) ⎥ (35)
K0 ⎝ pA ⎠ ⎢⎣⎝ 100 ⎠ ⎩⎪ ʹ ⎠ ⎭⎪
⎝ σ v0 ⎦⎥
for initially isotropic sand fabric. η is a rate installation parameter expressed as:

⎛u ⎞
η = tan −1 ⎜⎜ z ⎟⎟ (36)
⎝ uθ ⎠
where uz and u are the vertical and rotational (torsional) displacements of the drilling tool used
θ
in pile installation (Figure 6e). The installation parameter η represents the ratio of the
advancement of the drilling tool into the ground and the rotation of the drilling tool. Lower
values of η imply slower rate of penetration of the drilling tool into the ground during drilling (as
expected for installation in dense sands), while higher values of η imply easier drilling
conditions (as expected for installation in loose sands). For a drilling condition in which the
auger rotates a single full rotation to penetrate a length equal to the pitch length of the auger, η is
the flight angle of the auger. Note that the value of K/K0 for sand with initially anisotropic sand
fabric is always smaller than of initially isotropic sand fabric.
3. CONCLUSIONS
This paper presents several pile design methods that directly or indirectly account for the effects
of pile installation methods on calculation of axial capacity of piles in sand. Degradation of
frictional resistance along the pile shaft with increasing number of installation-induced shear
cycles is recognized in all design methods for displacement piles discussed in this paper. The
use of cone resistance qc in the calculation of unit shaft and base resistances is widely adopted.
However, guidelines to arrive at a representative qc value at the pile base differ from one method
to the other. Although the capacity calculation methods discussed herein are based on analytical,
numerical and experimental studies, theoretical simulations of installation of displacement piles
and partial displacement piles are still not at a sufficient level of rigor and there is still the need
for field load test data on instrumented piles, with side-by-side comparison of different pile types
installed in the same profiles, in order to further improve the prediction capabilities of pile
capacity calculation methods.
4. REFERENCES
Basu, P., Prezzi, M. and Salgado, R. (2014). Modeling of installation and quantification of shaft
resistance of drilled-displacement piles in sand. International Journal of Geomechanics,
ASCE, 14(2), pp. 214–229.
Basu, P., Loukidis, D., Prezzi, M., and Salgado, R. (2011). Analysis of shaft resistance of jacked
piles in sand. International Journal for Numerical and Analytical Methods in
Geomechanics, John Wiley & Sons, 35(15), pp. 1605-1635.
Basu, P., Prezzi, M., and Basu, D. (2010). Drilled displacement piles – current practice and
design. DFI Journal: The Journal of the Deep Foundations Institute, 4(1), pp. 3-20.
Brown, D.A. and Drew, C. (2000). Axial capacity of augered displacement piles at Auburn
University, New Technological and Design Developments in Deep Foundations.
Proceedings of sessions of Geo- Denver 2000, Geotechnical Special Publication No. 100,
ASCE, pp. 397-403.
Brown, D.A. (2005). Practical considerations in the selection and use of continuous flight auger
and drilled displacement piles. Advances in auger pressure grouted piles: design,
construction and testing. Advances in Designing and Testing Deep Foundations,
Geotechnical Special Publication No. 129, ASCE, pp. 251-261.

Bustamante, M. and Gianeselli, L. (1993). Design of auger displacement piles from in-situ tests.
Deep Foundations on Bored and Auger Piles, BAP II, Balkema, Rotterdam, pp. 21-34.
Bustamante, M. and Gianeselli, L. (1998). Installation parameters and capacity of screwed piles.
Deep Foundations on Bored and Auger Piles, BAP III, Balkema, Rotterdam, pp. 95-108.
de Kuiter, J., and Beringen, F. L. (1979). Pile foundations for large North Sea structures. Marine
Geotechnology, 3(3), 267–314.
Fleming, W. G. K., Weltman, A. J., Randolph, M. F. and Elson, W. K. (1992). Piling
Engineering. Surrey University Press, Surrey.
Han, F., Prezzi, M., Salgado, R. and Zaheer, M. (2016). Axial resistance of closed-ended steel-
pipe piles driven in multilayered soil. Journal of Geotechnical and Geoenvironmental
Engineering, DOI: 10.1061/(ASCE)GT.1943-5606.0001589.
Han, F., Salgado, R., Prezzi, M., Lim, J. (2017). Shaft and base resistance of non-displacement
piles in sand. Computers and Geotechnics, 83, pp. 184–197.
Jardine R., Chow F., Overy R. and Standing J. (2005). ICP design methods for driven piles in
sands and clays, Thomas Telford Publishing, London, ISBN: 0 7277 3272 2.
Lee, C. J. and Salgado, R. (1999). Determination of pile base resistance in sands. Journal of
Geotechnical and Geoenvironmental Engineering, ASCE, 125(8), pp. 673-683.
Lehane, B. M. and Randolph, M. F. (2002). Evaluation of a minimum base resistance for driven
pipe piles in siliceous sand. Journal of Geotechnical and Geoenvironmental Engineering,
ASCE, 128(3), pp. 198-205.
Lehane, B. M., Schneider, J. a., and Xu, X. (2005). The UWA-05 method for prediction of axial
capacity of driven piles in sand. Proceedings of the International Symposium. on Frontiers
in Offshore Geotechnics (IS-FOG 2005), pp. 683–689.
Loukidis, D., and Salgado, R. (2008). “Analysis of the shaft resistance of non-displacement piles
in sand.” Géotechnique, 58(4), pp. 283–296.
Loukidis, D. and Salgado, R. (2009). Modeling sand response using two-surface plasticity.
Computers and Geotechnics, 36(1-2), pp. 166-186.
Meyerhof, G. G. (1976). Bearing capacity and settlement of pile foundations. Journal of
Geotechnical Engineering, Vol. 109 (6), pp. 797-806.
Paik, K., and Salgado, R. (2003). Determination of bearing capacity of open-ended piles in sand.
Journal of Geotechnical and Geoenvironmental Engineering, 129(1), pp. 46–57.
Paik, K., Salgado, R. and Lee, J. (2004). Design lessons from load tests on open- and closed-
ended pipe piles. Proceedings of the Fifth International Conference on Case Histories in
Geotechnical Engineering, New York, April, Paper No. 1.11.
Paik, K., Salgado, R. Lee, J and Kim, B. (2003). Behavior of open- and closed-ended piles
driven into sands. Journal of Geotechnical and Geoenvironmental Engineering, ASCE,
129(4), pp. 296-306.
Prezzi, M. and Basu, P. (2005). Overview of construction and design of auger cast-in-place and
drilled displacement piles. Proceedings of DFI’s 30th annual conference on deep
foundations, Chicago, pp. 497-512.
Randolph, M.F. (2003). Science and empiricism in pile foundation design. Géotechnique, 53(10),
pp. 847-875.
Salgado, R. (2008). The engineering of foundations. McGrawHill, New York.
Salgado, R., Prezzi, M. (2007). Computation of cavity expansion pressure and penetration
resistance in sands, International Journal of Geomechanics, 7 (4), pp. 251 – 265.

Salgado, R., S. I. Woo, and D. Kim. Development of Load and Resistance Factor Design for
Ultimate and Serviceability Limit States of Transportation Structure Foundations.
Publication FHWA/IN/JTRP-2011/03. Joint Transportation Research Program, Indiana
Department of Transportation and Purdue University, West Lafayette, Indiana, 2011. doi:
10.5703/1288284314618.
Schmertmann, J. H. (1978). Guidelines for cone test, performance, and design. U.S. Federal
Highway Administration, FHWATS-78209.
Seo, H., Yildirim, I. Z. and Prezzi, M. (2009). Assessment of the axial load response of an H
pile driven in multi-layered soil. Journal of Geotechnical and Geoenvironmental
Engineering, 135 (12), pp. 1789-1804.
Tomlinson, M. J. (1987). Pile design and construction practice. Viewpoint, London.
Xu, X., Schneider, J. A., and Lehane, B. M. (2008). Cone penetration test (CPT) methods for
end-bearing assessment of open- and closed-ended driven piles in siliceous sand. Canadian
Geotechnical Journal, 45(8), pp. 1130–1141.

Dynamic Loading Tests: A State of the Art of Prevention and
Detection of Deep Foundation Failures
Rausche, F.(1), Alvarez, C.(2) and Likins, G.E.(3)

(1)
Pile Dynamics, Inc., Cleveland, OH, USA <frausche@pile.com>
(2)
GRL Engineers, Inc., Los Angeles, CA, USA <calvarez@grlengineers.com>
(3)
Pile Dynamics, Inc., Cleveland, OH, USA <glikins@pile.com>
ABSTRACT. While observations during pile installation have been used for centuries to produce
a quality deep foundation, using electronic measurements routinely has only become possible with
modern sensors and computers. The beginning of this development took place more than 50 years
ago in the 1960s. While the traditional and one of the most important methods of deep foundation
quality assurance is the static loading test, it is now often partially or completely replaced by the
Dynamic or High Strain Dynamic Loading Test. After a brief review of all available Quality
Assurance and Quality Control methods, the paper describes the High Strain Test procedure, its
theoretical background, equipment, software and standards. Two examples, one for driven and one
for a cast-in-situ pile, demonstrate the benefits and limitations of the available methods.
1. INTRODUCTION
All deep foundation types, whether prefabricated and driven or cast-in-situ after drilling will be
called a “pile” in this paper. Deep foundations are needed where surficial soils have insufficient
strength to support a building, bridge or other structure. Because of their importance for adequate
performance and because of their cost, deep foundation specifications and building codes
increasingly demand thorough testing. As a reward for a thorough testing regime, modern codes
reward the project with reduced factors of safety (AASHTO, 2014). The economic impact from
the savings realized with reduced factors of safety can be very significant (Likins 2015).
1.1 Static Loading Methods
Tests were always performed, most notably the static top-down loading test, to verify adequate
foundation performance. The load carrying capacity of both driven piles and bored piles has
traditionally been evaluated by static load tests (SLT). Using either dead weights or reaction piles,
the test pile is loaded by a hydraulic jack pressing against a reaction frame either in compression
(ASTM D1143) or in tension (ASTM D3689). For larger test loads ASTM D1143 requires an
instrumented load cell.
Although the top load static test is the reference test against which all other tests methods are
judged, the test’s shortcomings are numerous: The test is expensive and time consuming, contains
many potential error sources and is non-unique as far as interpretation, as has been widely
documented in the literature (i.e., Fellenius, 1990). Moreover, when it comes to large loads, the
test can be very dangerous due to high “dead” loads (kentledge) or locked-in energies when elastic
reaction systems are in use.
For large cast-in-situ piles with high capacities the safer and usually more economical bi-
directional test has been developed (Elisio Da Silva, 1983 and Osterberg, 1984) and is being

performed with increasing frequency around the world. It is particularly useful for large cast-in-
situ piles. Rather than placing the jack at the top of the pile, the jack is attached to the reinforcing
cage and inserted in the pile. While the embedded jack is often placed near the bottom of the shaft,
it can also be placed at any location along the pile length. When pressurized it then exerts a
downward force below the jack (resisted by end bearing and any shaft resistance below the jack)
and an upward force above the jack (resisted by the pile weight and the shaft resistance above the
jack. The upward and downward force components are combined into an equivalent applied load
using strain compatibility principles. Advantages over top loading tests are first that very high
loads can be safely applied and secondly the more accurate determination of end bearing vs.
displacement compared to any other loading method. Disadvantages are that the jack (or jacks for
high loads) is not recovered, that no load cell can be reasonably utilized for improved accuracy,
that it must be assumed that the shaft resistance is the same under upward and downward
movements and that the separation of the upper and lower shaft sections near the toe cause a
different stress and strain regime than under top loading, and that the pile top is at zero load during
the testing (so structural integrity of the upper shaft is not evaluated as it would be in the top-down
test).
1.2 Dynamic Formulas
For driven piles, dynamic formulas have been widely accepted because of their simplicity; their
benefits, if any, and limitations have been described by Allin et al. (2015) and Likins et al. (2012a).
Dynamic formulas use generally only an assumed energy and an observed blow count to assess
pile bearing capacity. Their use is, therefore, limited to driven piles. Their simplicity is often given
as a reason for their use, however, that simplicity is also the reason for poor correlations with static
test results which a myriad of versions and factor adjustments could not overcome, thus requiring
uneconomically high factors of safety. In addition these formulas do not allow for stress
calculations.
1.3 Pile Dynamic Analysis and Simulation Methods
A numerical solution of the underlying wave equation became practical with the availability of the
electronic computer. Great interest among both construction and academic professionals was
generated by Smith (1960) in the United States. At about the same time in Europe, Fischer (1960),
for example, developed his graphical solution of the wave equation which he applied to a number
of pile driving problems and which was related to software developed by others such as Meunier
(1984). Any of these impact simulation programs proved to be much more accurate than the
traditional dynamic formulas and not only provided a relationship between the number of blows
per unit penetration (blow count) and bearing capacity, but also a reasonably accurate prediction
of dynamic pile stresses. Among the additional options included in these programs was the
driveability analysis which became particularly popular since it simulated not only what happens
under one hammer blow, but rather what can be expected during the complete pile installation
process using diesel, hydraulic, air, steam or even vibratory hammers (e.g., Rausche et al., 2009).

1.4 The Case Method of Dynamic Pile Testing
Beginning in the late 1950s measurements which could be routinely performed on construction
sites were developed at Case Western Reserve University (Goble and Rausche, 1970a); Hussein
and Goble (2004) briefly described these developments. Interpretation of the measured force and
velocity signals were based on closed-form solutions of the wave equation (Goble and Rausche,
1970b). This approach was so successful, particularly after being introduced in Sweden and from
there around the world, that the first Stress Wave Conference was held in Gothenburg in 1980 with
repeat conferences every 4 years. A wealth of papers describing the advancement of the dynamic
pile testing methods has been generated by these conferences. Similar development efforts were
also made in Europe, see for example Beringen et al. (1980). The Case Method is further described
below.
The sensors developed during the 12 years of research at Case Western Reserve University
(CWRU) included reusable strain transducers which would be bolted to the side of the pile. The
strain signal was then multiplied with the pile’s elastic modulus and cross sectional area to yield
the pile force. Alternatively a so-called “top transducer” was used (Goble and Rausche, 1970a)
which could be calibrated in a universal testing machine and avoided concrete modulus
uncertainties when cast-in-situ piles were tested. The top transducer was basically an instrumented,
short section of heavy-wall pipe. Velocity was measured using accelerometers whose signals were
integrated to yield velocity. While various changes in sensor technology improved significantly,
the quality of the signals recorded and evaluated, the basic measurement systems are still the same
as developed during the original research. ASTM D4945, “Standard Test Method for High Strain
Dynamic Testing of Deep Foundations”, specifies how the dynamic tests have to be performed.
Similar standards now exist in many other countries or regions such as Australia, Brazil, China,
and Europe.
For the calculation of pile bearing capacity two approaches were chosen; one was a simple
equation that could be solved by computer between hammer blows and one a signal matching
approach which would be more computer time extensive. The simple equation was first based on
a rigid pile model with resistance calculated at the time of zero velocity used (Goble and Rausche,
1970a). Today this approach is also called an unloading point method and is applied to the
interpretation of the Rapid Load Testing Method (Middendorp et al., 1992). However, it soon
became clear that a closed form solution to the wave equation such as has been described by
Timoshenko and Goodier, (1951) would be more accurate (Rausche et al., 1972); it was called the
“Case Method Equation”.
The signal matching procedure developed by the Case team, called CAPWAP®, relied first on
Smith’s numerical analysis approach (Rausche et al., 1972); it was later replaced by the
characteristics solution to the wave equation (Rausche, 1988) called CAPWAPC (today it is again
referred to as CAPWAP). Today a fast automatic signal matching program, iCAP (Likins, et al.,
2012b) is also available for uniform, driven piles. While not taking the place of the more accurate
and powerful CAPWAP program, iCAP can replace the Case Method with the advantage that it
does not require an estimated damping factor. Also, results from iCAP are unique, i.e., independent
of user experience. In the 1970s other formulas were also developed for the assessment of hammer
performance, pile stresses and pile integrity from pile top force and velocity as discussed below.

2. PREVENTION AND DETECTION OF DEEP FOUNDATION FAILURES
Two different deep foundation failures have to be distinguished Geotechnical and Structural
failures. The former is best detected or prevented by loading tests and, for driven piles, dynamic
monitoring. The latter is generally only measured and assessed by dedicated integrity tests. While
the static loading test is least suitable for detecting structural deficiencies unless they cause the
pile to catastrophically fail during the loading, the dynamic loading test has a better chance to
detect defects as will be discussed in the Section on dynamic monitoring. The following describes
the most common and recognized integrity test methods and lists additional integrity tests that are
less commonly used after mentioning a few construction monitoring methods which can help
prevent pile integrity problems.
2.1 The Pile Installation Recorder
This device measures, during the construction of augered-cast-in-place (ACIP) or Continuous

Flight Auger (CFA) piles, the pumped concrete volume as the auger is retracted. While volume
measurements sometimes only rely on the counting of pump strokes, the preferred and more
accurate means is uses the magnetic flow meter. The basic result of this monitoring is a concrete
volume vs. depth record, sometimes augmented by pump pressure and auger torque (Brown et al.,
2007) While the volume vs. depth record is an important document (Piscsalko et al, 2004) it is
even more important that the operator can view the measurements in real time and take corrective
action should the incremental pumped volume with depth fall below requirements.
2.2 Drilled Hole Inspection by Caliper
After completion of the drilling and prior to placing concrete in the bored pile, the shape or profile of the
hole can be measured with a so-called caliper. Assuming that no further changes in the hole occur after the
measurement and during concreting, the shape of the finished pile has been established. Some caliper
devices use mechanical arms while more modern devices use an ultrasonic technique.
2.3 The Shaft Inspection Device (SID)
For bored piles, when end bearing is considered in the design, the condition of the bottom of the drilled
hole is important and must be “clean”, meaning loose sediment removed, so that end bearing is activated at
a relatively small displacement rather than first compressing a weak debris layer. This is particularly
important in rock sockets. While not directly measuring pile capacity, the SID is being used with a thin
measuring rod penetrating the soft sediments. The rod penetration into the sediment is viewed with a remote
camera which also displays the cleanliness of the bottom surface.
2.4 The Quantitative Inspection Device (SQUID)
A more advanced inspection tool, the SQUID, actually measures the force and distance required to penetrate
any potential debris layer and also the resistance in the bearing layer using one or more instrumented cones
(Fig. 1). The device is conveniently attached to the Kelly bar or drill stem and quickly inserted in the water
or slurry filled or dry hole. The whole process typically takes less than 15 minutes. When the hole is
confirmed as clean and the required penetrometer force is adequately confirmed, the pile can be concreted
and the end bearing included in the design. Such a device can be particularly cost effective to minimize the
depth of a rock socket by determining when the rock is of sufficient strength.

2.5 Low Strain Integrity Testing, (Pulse Echo Testing, PIT)
One of the earliest methods, and one which has great similarities with the high strain method, is low strain
integrity testing (Rausche et al, 1988). The method of data collection is specified in ASTM D5882. The
method can be quickly applied and testing 100 piles in one day is not impossible. Higher testing rates may
lead to poor data quality.
Once concrete of the pile has hardened sufficiently, the pile top is struck by a hand-held hammer which
generates a force wave that travels down the pile shaft, reflects off the pile toe or other cross section changes
and then propagates back to the pile top. An accelerometer, attached to the pile top, measures both the input
and reflections. The acceleration is integrated to velocity and various enhancements to the velocity record
are made to facilitate data interpretation which is based on basic wave propagation theory (Likins and
Rausche, 2000). For example, Fig. 2 shows two records of pile top velocity vs. time enhanced by an
exponential amplification function to compensate for signal losses due to soil and pile material damping.
The top signal shows a strong positive reflection at a time corresponding to the designed 25 m length of the
pile. The bottom graph, on the other hand, displays a smaller reflection corresponding to a 15 m depth
suggesting an anomaly of pile size or concrete quality 10 m above the pile toe.
Occasionally a different type of low strain data interpretation, referred to as the Transient Response
Method, (Rausche et al., 1991), is employed. Actually, it requires that, in addition to the velocity, the impact
force of the hand held hammer is measured. It calculates by Fourier Transfer the Mobility of the pile in the
frequency domain thereby yielding, for example, a pile stiffness value.
Fig. 2. Presentation of PIT records (a) top intact

pile and (b) defective pile
Fig.1. The SQUID prior to insertion

in the drilled hole: (1) calibrated
force transducer; (2) displacement
measuring plate.
2.6 Cross Hole Sonic Logging (CSL)
Crosshole Sonic Logging (Likins et al, 2004), commonly called “CSL”, requires installation of at least two
water filled access tubes in the pile so that concrete quality can be assessed by measuring the wave travel
time in the concrete between a transmitter in one tube and a receiver in the other tube. Typically one tube
for each 300 mm of pile diameter is required. The method is standardized in ASTM D6760. From the “First
Arrival Time” (FAT) of the signal and the spacing between access tubes, the wave speed can be calculated.

The wave speed magnitude is considered a measure of the concrete quality. FAT decreases of more than
20 or 30% are generally considered an anomaly of concern. CSL is not sensitive to the surrounding soil and
is not limited by pile length. Unfortunately, it also gives no indication of the quality or quantity of the
concrete outside the reinforcing cage and sometimes may falsely indicate problems due to “debonding” of
the access tubes from the surrounding concrete which inhibits wave propagation. If low wave speeds are
detected in several scans at the same cross section, then a “tomography” analysis (Likins et al., 2007) can
be useful to estimate the extent of the anomaly (Fig.3).
2.7 Thermal Integrity Profiling
Cement produces heat during the curing process. This phenomenon is the basis for Thermal Integrity
Profiling of the entire cross section of bored piles (Piscsalko et al., 2013). During the curing of a bored pile,
the center of the pile has the highest temperature while the perimeter has the lowest temperature since it is
adjacent to the soil and the heat is flowing from the pile into the soil. The more cement content in a concrete
mix, the higher the temperature created and, therefore, the lower the temperature the lower the concrete
quality or pile size. The Thermal Integrity Profiling method procedures are governed by ASTM D7949.
While temperature can be sensed by infrared probes in access tubes, it is more convenient to measure the
temperature by attaching cables with thermal sensors to the reinforcing cage. One such instrumented cable
is installed equidistantly around the cage for each 300 mm of pile diameter. The average temperature of the
shaft can be correlated to the effective average shaft radius. Local deviations from the average shaft
temperature can then be related directly to deviations from the average shaft radius, allowing for a complete
evaluation of the entire cross section including the concrete cover outside the reinforcing cage. Fig. 4 shows
a 3D image calculated from temperature measurements. It should be noted that the test has to be performed
while the concrete cures which makes for a quick turnaround of results (often be completed within 24 hours
of casting concrete), saving valuable construction time.

2.8 Other Integrity Test Methods
There are other methods which have been reported on and which are occasionally used depending
on the purpose and or preference of the specifier. The following lists first three of more
successfully used methods and then two additional ones that are less frequently used.
1. The Gamma-Gamma Method helps identify concrete cover defects by measuring
concrete density around the inspection tubes and which is used in conjunction with CSL
testing.
2. The Length Inductive Test Method (LITE) which also requires a borehole next the pile to
be tested for length; in this case the pile has to be made of metal.
3. The Parallel Seismic Method which requires a borehole next to the pile, an accelerometer
in the borehole and a light hammer impact on the pile top for pile length determination.
4. The Bending Wave method which applies impacts to a pile (typically a timber pile under
a structure) perpendicular to its axis (there are similarities with the Low Strain Method).
5. The Vibration Method which applies a variable frequency vibration to the pile top and
which is evaluated in the frequency domain similar to the Transient Response Method.
3. DYNAMIC PILE ANALYSIS AND HIGH STRAIN TESTING

3.1 Wave Equation Analysis
Dynamic formulas do not consider pile type or dimensions, hammer configuration or actual
hammer efficiency (Allin et al., 2015), driving system components, or soil type and profile and
they cannot calculate stresses; these limitations gave rise to the development of the wave equation
approach by Smith (1960). Among several proprietary computer programs now in existence,
GRLWEAP is probably most widely used and is further described in Rausche et al., 2004 and
Hannigan et al., 2016.
In this “wave equation” approach, the hammer and pile are modeled by a series of masses and
springs and an initial impact velocity is assigned to the ram segments. The soil is modeled with
both springs, representing the static capacity, and dampers, representing viscous effects. Using
short time increments the resulting forces in the springs and motions of all pile segment masses
are computed as time progresses. Thus, the maximum stresses at every location in the pile can be
evaluated for any modeled situation, and the final net displacement calculated for an assumed or
calculated (by standard static methods) capacity. Generally several capacities are assumed and the
corresponding blow counts computed and the resulting relation of input capacities to computed
blow counts is known as a bearing graph. Given accurate soil information, capacity can be
estimated for any depth by static analysis and the installation analyzed at various depths of
penetration as a check for pile driveability, i.e., to prove that the hammer is capable of installing
the pile to the desired depth or capacity while keeping driving stresses within acceptable bounds.
While the resistance distribution is determined by standard static geotechnical methods, such
as the alpha or beta methods, additional dynamic soil resistance parameters are needed. They
include for the static resistance the quake (equal to the ultimate resistance divided by the soil
stiffness). A dynamic resistance parameter, called damping factor, also has to be specified.
Additionally, for the so-called driveability analysis, which determines an estimated blow
count vs. depth, the so-called soil setup factor has to be input for each soil layer. A driveability
method popular for monopole installations, called friction fatigue, requires an additional shape
factor which describes the resistance distribution during pile driving (Alm and Hamre, 2001).

These parameters as well as a nearly complete hammer and driving system input help is provided
by the GRLWEAP program.
The wave equation analysis has been widely accepted as a valuable tool in assessing
compatibility of the hammer with the pile for a specific soil profile, and to evaluate the driving
stresses in such scenarios to prevent pile structural damages. While not a “test” in itself, combining
the results of a wave equation analysis with the observed pile penetration per blow (set) or blow
count taken at the end of a pile installation or during a restrike test allows for assessing the bearing
capacity of the pile. An input screen of the GRLWEAP program is shown in Fig. 5 representing
data for a 1724x35.4 mm open ended pipe pile of 47.6 m length, driven through water and then
sand into calcarenite rock, reaching a final depth below mudline of 19.1 m. The hydraulic hammer,
an IHC 200S, was run at a reduced energy of 72% at the end of driving. A restrike test conducted
10 days later was run with an energy setting of 57%. For these two energy levels the wave equation
analysis with standard soil parameters for granular soils produced the bearing graphs shown in Fig.
6. Entering these bearing graphs with respective end-of-drive and restrike blow counts of,
respectively, 54 and 125 blows for 250 mm penetration (corresponding to 4.6 mm and 2 mm set
per blow) the wave equation would predict capacities of 12,000 and 15,700 kN. The bearing graphs
also indicate refusal capacities between 22,000 and 24,000 kN. The associated calculated
compressive stresses were 146 and 133 MPa. These results will be further discussed below.
While this program was developed for impact driven piles and has become an indispensable
tool for pile installation preparation and construction control, today it is also used to check the
driveability of vibratory driven piles (Rausche and Beim, 2012) and for the preparation of dynamic
loading tests even on cast-in-situ piles (Hussein et al., 1996).
Fig. 5. Example Bearing Graph Input. Fig. 6. Example Bearing Graph Output
3.2 High Strain Testing

3.2.1 Dynamic Monitoring
For impact driven piles, an installation log showing hammer energy setting and observed blow
count is still an indispensable quality control tool, however, in addition electronic measurements
are frequently taken for a quantification and more accurate assessment of pile driving equipment

performance, pile stresses, pile integrity and pile bearing capacity. Fig. 7 (left) shows photos of
strain and acceleration sensors attached to an H-pile; strain is converted to force by multiplication
with cross sectional area and elastic modulus. Velocity is calculated from acceleration by
integration over time. Fig. 7 (right) shows a force transducer on top of a drilled shaft; it provides
a direct measurement of force. The force and velocity data are analyzed in real time by closed-
form solutions of the wave equation, most notably the “Case Method” initially developed during
the CWRU research and later modified to some degree (Rausche et al, 1985). The closed-form
Case Method Equation allows for immediate analysis on site blow-by-blow to calculate the static
soil resistance component, R, using Equation 1.
R = (1-J)(F(t1) + Z V(t1))/2 + (1+J)(F(t2) - Z V(t2))/2 (1)
where
F(t1) is the measured force at a chosen time, t1,
V(t1) is the measured pile velocity at the same time t1
t2 = t1 + 2L/c
L is the length of the pile below the sensors
c is the pile material wave speed and
Z = EA/c is the pile impedance
E is the pile elastic modulus,
A is the pile cross sectional area and
J is the damping factor which is related to the soil type, typically ranging from 0.4 for coarse
grained soils to 1.0 for cohesive soils. It is normally determined by CAPWAP analysis of
one of the records obtained.
Equation (1) is evaluated for all times t1 between impact and the end of each of the acquired
records, i.e., of all monitored hammer blows. The maximum value of resistance, Rx, is considered
the best estimate of the ultimate soil resistance. The damping factor, J, has to be estimated based
on soil grain size or determined by correlation with either CAPWAP analyses or static load tests.

Fig. 7. Left: Accelerometers (right) and strain transducers (left) and data transmitter all
attached to an H-pile; Right: Instrumented, cushioned “top transducer” on bored pile.
Another convenient formula allows for calculation of the energy transferred to the pile:
E(t) = +F(t) V(t) dt (2)
The maximum value of E(t) is often referred to as EMX or ENTHRU and is a valuable hammer
performance indicator. The ratio of EMX to rated energy, Er, is called transfer ratio (also transfer
efficiency or global efficiency). Experience has shown that certain types of hammers such as
diesel, air/steam or hydraulically powered ones, have certain acceptable ranges of energy transfer
for certain types of piles.
Stresses at the pile top are directly obtained from the strain measurements. Using one-
dimensional wave propagation theory, the average compression at the pile toe and the maximum
tension at any location along the shaft can also be evaluated from the pile top measurements.
Keeping these stresses below the recommended limits based on structural material properties
reduces the possibility of pile damage (Hannigan et al., 2016).
The extent of damage and its location can be quantified using the so-called Beta Method
(Rausche and Goble (1979), Rausche et al., 1988 and Likins and Rausche (2014)). This method,
applicable to uniform piles, is based on the wave propagation based theory and basically calculates
an integrity indicator less than 100% if a tension wave reaches the pile top prior to the arrival of
the stress wave from the pile toe. Clearly, this method has a great similarity to damage detection
by the Low Strain method however, rather than investigating only the velocity record for
reflections, the force signal now serves as a reference line. Fig. 8 shows two records of a 41 m
long 380 mm square section concrete pile. The records consist of force and velocity times pile
impedance (Z=EA/c, see Eq. 1 above); the product of velocity and impedance has the units of
force. The vertical lines in these records indicate the onset of impact, to, and the onset of the wave
return from the pile toe, i.e., to plus 2 times the pile length below sensors divided by the wave
speed (2L/c).

The upper graph shows the force slightly increasing relative to velocity in the first 2L/c time
period and the difference F minus VZ is a measure of the shaft resistance. In contrast, the lower
graph shows a relative velocity increase about L/c after impact which can be interpreted as a signal
from mid-pile length. This is indicated by a dashed vertical line. Since only a reduction in pile
impedance can be responsible for such a record feature (velocity increasing relative to force) it
must be concluded that the pile is damaged at mid-length.
Fig. 8: Pile top force (solid) and velocity (dashed) records of a pile tested
before (top) and after damage occurred (bottom).
Dynamic monitoring results for each hammer impact are conveniently plotted either vs. pile
penetration depth or vs. blow number. These results are instructive and may be used, for example,
to optimize the driving process. Fig. 9 shows for the 1724x25.4 mm open ended pipe pile,
discussed above, pile top and bottom stresses (left), transferred energy and blows/minute (center)
and blow count plus Case Method resistance for a damping factor J=0.5 (right). This is a
comprehensive summary of results for every hammer blow and would, for example, indicate that
the hammer energy setting should be reduced for areas with high stresses or that a reduced pile
length is acceptable, if capacity is reached earlier than originally anticipated. Also, this information
can be conveniently used to performed so-called refined wave equation analyses, i.e., wave
equation analyses with hammer, driving system and soil parameter input values adjusted for
energy, stress and blow count results matching the field observations.

Fig. 9. Left: Pile driving monitoring results for the example case; Right: Wave equation
bearing graph using dynamic loading test results for refinement.
3.2.2 Dynamic Loading Tests
Each hammer blow applied to a pile loads the pile during a short time period and activates soil
resistance forces. The force and velocity data recorded under an impact loading can, therefore, be
interpreted as a very quick loading test. Both driven piles and cast-in-situ piles can be tested in that
manner (Rausche and Seidel, 1984 and Seidel and Rausche, 1984).
Since changes in the soil occur during pile driving, the end of driving blows may not be
indicating the soil resistance that a static loading test would or what would be present under long
term loading. Both soil resistance increases (soil setup) or decreases (relaxation) with time after
pile installation have been observed. While for monitored and impact driven piles the end-of–
driving (EOD) data is valuable for quality assurance and control purposes, it is, advisable to
perform a restrike test after a certain, soil-type dependent waiting time. The beginning of restrike
(BOR) data are usually closely related to the long term bearing capacity. During restrike testing it
is possible that soil setup is so high that the installation pile hammer may not enough energy to
move the pile under a single blow sufficiently (at least 2 mm) to mobilize the available ultimate
soil resistance. In that case only a lower bound of the available soil resistance, or “proof load”,
would be calculated from the pile top measurements. To remedy this problem a high energy drop
hammer is often mobilized as would also be necessary when cast-in-situ piles are tested. Another
remedy would be the so-called superposition, i.e., combining EOD end bearing with BOR shaft
resistance (Hussein et al., 2002).
The dynamic loading can be performed in two ways; the first often uses a pile driving hammer
whose driving system, in particular, the cushions have been designed for a quick and efficient
loading with cushioning just enough to protect the pile from excessive stresses. If a drop hammer
is used, the pile top cushion is designed so that a high energy transfer is possible. This test
configuration is covered by ASTM D4945.

The second dynamic loading method is referred to as Force Pulse or Rapid Test (ASTM
D7383); Fig. 10 shows different configurations of the test setup. Originally, this test was performed
by burning a combustive fuel in a chamber located above a load cell on the pile top and below a
heavy reaction mass. This system was called Statnamic testing (Bermingham and Janes, 1989).
The force pulse thus generated lasted for about 100 ms or about 5 to 10 times longer than typical
dynamic load tests. Also the rise time, i.e., the time it would take the force to reach its peak was
significantly longer that of a dynamic test. It was later realized that a similar force pulse could be
generated by a heavily cushioned drop hammer. For example, the Fundex Pseudo Static tester uses
large steel springs as per Schellingerhout and Revoort (1996). Alternatively using a dynamic
testing system with very thick plywood cushions as per Rausche et al. (2008) or thick sandwiches
of synthetic material described by Miyasaka et al. (2009) would produce long force pulses.
However, the rise time of the signals produced by the latter two loading systems is similar to those
of the dynamic test.
For the force pulse test, typical required drop weights or reaction masses are 5 to 10 percent
of the desired ultimate capacity while dynamic tests are usually done with 1% drop weights when
testing piles on rock and 2% otherwise. The advantage of a relatively long duration pulse is that
tension stresses in the pile are then of little concern. However, since the velocity of the pile under
the force pulse is at or above 1 m/s, dynamic resistance forces and inertia forces must be
considered, and unless the pile is further instrumented with strain gages, resistance distribution
cannot be deduced from this test.
There are widely differing opinions on how to deduce the equivalent static test from the basic
measurements, particularly in cohesive soils (e.g. Middendorp et al., 1992; Matsumoto, 1994;
Hajduk, 2000; Schmuker, 2005; Weaver, 2010; Brown, 2013). If there is a significant net
settlement (minimum 3 percent of pile diameter), then the soil resistance is considered as “fully
mobilized” and is perhaps considered more reliable (Miyasaka et al, 2009). In addition to axial
tests, the Statnamic device has been deployed to apply lateral impacts, which help model, for
example, ship impacts and earth quake loads.
While the Case Method is a good first indication of the mobilized soil resistance, for dynamic
load testing the force and velocity data are always analyzed in a rigorous manner by “signal
matching”. The most commonly used software is called CAPWAP which stands for the Case Pile
Wave Analysis Program. It is practically a wave equation program, but without the need to
modeling hammer and driving system since the hammer impact effect is known from the pile top
force and velocity records. (Rausche and Goble, 1972) and years of experience have demonstrated
good agreement with static loading test results (Likins and Rausche, 2004). The advantage over
the Case Method, besides finding the damping factors, is the ability to calculate quakes and soil
resistance distribution. The CAPWAP method can also be used when the rise time of the force
pulse is short as in the system described by Miyasaka et al., (2009).
The long rise time force pulse tests are not easily analyzed by CAPWAP because of the lack
of a clear wave front activating successively resistance along the pile. These measurements are
generally interpreted (a) by defining the dynamic soil resistance at the time when the pile top starts
to rebound or unload, which implies that then the velocity and therefore damping is zero, and (b)
by multiplying the unloading force by a rate effect correction factor. This factor varies between
1.0 for sands and approximately 0.5 for highly cohesive soils.

4. EXAMPLE: HIGH STRAIN TESTING OF AN OPEN ENDED PIPE PILE
As an example of a CAPWAP result, the beginning of restrike (BOR) test of the wave equation
and monitoring case discussed earlier is shown in Fig. 11. This example, the same that was used
for demonstrating the wave equation approach, is particularly interesting because of occasional
problems encountered with the dynamic testing of large diameter open ended pipe piles where full
end bearing may not be mobilized in dynamic test due to soil plug slipping caused by inertia effects
(Brown and Thompson, 2015). The CAPWAP summary page of Fig. 11 includes the record
analyzed (upper right), the force signal match (upper left) the calculated resistance distribution
(lower right) and the calculated load-displacement curve (lower left). The numerical values shown
indicate a calculated total capacity of 11,700 kN with 3,400 kN end bearing. The end-of-driving
(EOD) results, also analyzed by CAPWAP, indicated a lower capacity of 6,300 kN with 1,600 kN
end bearing demonstrating an almost doubling of capacity during the 10 day waiting time between
EOD and BOR.
Compared to the wave equation prediction shown above, obviously much lower capacity
values have been calculated by CAPWAP. This difference is primarily due higher dynamic
resistance factors: for the end-of-drive the CAPWAP quakes were 7.5 and 9.6 mm vs. the standard
2.5 mm used in the wave equation analysis. This is probably due to an increased pore water
pressure which made the soil “spongy” or “bouncy” and thus more energy absorbing. For the
restrike it was noticed that primarily the shaft damping was much higher than normally assumed
(1.0 vs 0.16 s/m). Repeating the analyses, as so-called Refined Wave Equation Analyses, with the
CAPWAP calculated quakes and damping for EOD and a strongly increased toe damping for BOR
yielded the bearing graphs shown in Fig. 9, right, with capacities of 6,600 and 11,500 kN vs. 6,300
and 11,800 kN CAPWAP. (Note that adjustments to the CAPWAP soil parameters are usually
needed to achieve blow count agreement, because of differences in the blow count calculation
approaches of the two programs). Refined Wave Equation Analyses are useful when analyzing
different situations at the same site based on a representative dynamic load test. Dynamic
measurements and/or local experience are obviously invaluable for assessing proper dynamic
resistance quantities and therefore better wave equation predictions.

2
000
0 2
000
0
F
orceM
sd F
orceM sd
kN kN
F
orceC
pt Velo
cityMsd
1
250
0 1
250
0
5
000 5
000
1
5 9
5ms 1
5 9
5ms
-2
500 8L
/c -2
500 8L
/c
-1
000
0 -1
000
0
P
ileIm
ped
ance
1
500
S
h aftRe sista
nce L
e ngthb .S enso rs 4
5.4m
L
oad(kN
)
P
ileTop D
istrib
ution E
m be d m e nt 1
9.1m
0 2
000 4
000 6
000 8
000 1
000
0 1
200
0 1
200
Botto
m T
o pA re a 133
4.4cm 2
0
.0 E
n dB e a ringA re a 2
.33m2
9
00 T
o pP e rim eter 5
.42m
kN/m
5
.0 T
o pE -M od ulus 2
06843MP a
R
U= 1
1659k
N 6
00 T
o pS p e c.W eigh t 7
7.3kN/m 3
1
0.0 S
F= 8259k
N T
o pW aveS pd. 5123m/s
E
B= 3400k
N 3
00 Overa llW .S. 5275m/s
Displacement (mm)
D
y= 20
.1mm
1
5.0 0
D
x= 22
.1mm MatchQu a lity 1.39
0
S
ET/B
l= 2
.0mm T
opC o
m p r.S tre ss 1
21.0MP
a
2
0.0 2
000 E
B
MaxC om pr.S tress 1
44.1MP
a
MaxT ensio nS tress -2
0.52MP
a
2
5.0 4
000
Avg.S
h aftQu
a ke 2.9m m
6
000
kN
3
0.0 T
o eQua ke 2.8m m
S
F Avg.S
h aftSmithDpg
. 1
.01s/m
8
000
3
5.0 T
o eSmithDam pin
g 0
.58s/m
1
000
0 P
ileF
orce
a
tR u
4
0.0 1
200
0
Fig. 11. CAPWAP summary for the 2nd restrike blow of the open ended pipe example.
The same open ended pipe pile was also subjected to an instrumented static loading test, 16 days
after the restrike test., allowing for a comparison of statically and dynamically determined pile top
load-displacement curves; the dynamic and static displacements are shown cumulatively in
chronological testing order in Fig. 12, left; obviously the EOD test showed a relative lack of
stiffness compared to the BOR and the static tests, apparently caused by the large quakes
(equivalent to a reduced soil stiffness). The static test was conducted by applying 5 loading cycles.
The envelope enclosing the five cycles indicated a 12,500 kN capacity for the Davisson failure
criterion (Fellenius, 1990) which suggests a small, additional soil setup during the 16 day waiting
period after the restrike test; in any event a good agreement with the BOR test was achieved.
Instrumentation attached to the pile along its length also allowed for a comparison of the forces in
the pile at the corresponding CAPWAP load level and the Davisson failure load. Obviously, the
resistance distribution calculated from the dynamic test is similar but not very close to the statically
determined one (Fig. 12, right).

5. EXAMPLE: STATIC AND DYNAMIC LOADING TESTS OF A CFA PILE
Continuous Flight Auger (CFA) or Augered-Cast-in-Place piles (ACIP) are produced by drilling
with an auger of the length of the pile and then pressure injecting grout from the bottom up through
the hollow stem of the auger while the auger is extracted. While the monitoring of this installation
process is valuable (see Pile Installation Recorder above) testing for geotechnical and structural
sufficiency is equally important. Statically testing a few piles in the beginning of production piling
may be good to proof the overall suitability of the pile for the project, however, usually only a few
static loading tests are performed even though the number of piles on a site may be very large.
Quick and relatively inexpensive dynamic testing is, therefore, frequently performed too, and
agreement between static and dynamic test results has generally been very good. Also in recent
years the use of embedded sensors has generated valuable comparison records which should, in
the future, lead to even better agreement between dynamically and statically determined resistance
distributions. Of course, for cast-in-situ piles, the question of elastic modulus and cross sectional
area below grade is always a problem and may lead to inaccurate calculation of force from strain.
The following example shows results from both static and dynamic loading tests on a 44.5 m
long CFA pile of 610 mm diameter which was auger-cast through deep soft cohesive soils into a
sedimentary rock. Dynamic acceleration and force were measured on pile and with a top
transducer, respectively. The pile had been instrumented with strain gaged sister bars attached to
the center reinforcement bar of the pile at six locations. After the grout had sufficiently hardened,
the dynamic load test was performed by dropping a 16-ton APPLE ram from various drop heights
onto the lightly cushioned top transducer. The permanent displacement occurring under the
dynamic impact analyzed was 3.2 mm. A few days later a static load test was performed.
For the interpretation of the static test on cast-in-situ piles, AASHTO specifies an offset
criterion of 3.8 mm + D/120 for a 610 mm (or less) diameter (D) pile and 3.8 mm + D/30 for a 914
mm pile (or greater), with linear interpolation. In the present case the offset elastic line would
intersect the slightly extrapolated static load test curve at a little less than 10,000 kN while the

dynamic test activated 9,200 kN resistance. Practically the same static capacity value would be
found if the Butler-Hoy criterion (Fellenius, 1990) were applied to the static load-displacement
curve. The Butler-Hoy criterion is preferred for the interpretation of cast-in-situ piles in the
building industry. In any event, the static and dynamic load-displacement curves agree quite well
(Fig. 13, left) although the dynamic curve has a steeper slope at the origin. This difference is also
evident in the resistance distriution curves from CAPWAP and static test (Fig. 13, right). It may
be explained by a progressive failure of the upper shaft resistance during the static loading.
6. SUMMARY
Ideally, the monitoring of pile installation, both by driving and drilling, would avoid the need for
additional quality assurance methods and both pile installation recording and dynamic monitoring
are a great help in these efforts. However, the complexity of soil behavior and construction
procedures require testing of the final deep foundation elements though good construction
monitoring may reduce the number of required tests.
A variety of structural (integrity) and geotechnical (loading) tests have been developed, all
with their specific benefits and limitations. Dynamic methods, based on wave propagation theory,
generally have the advantage of being the least costly and least time consuming.
With proper care and good loading equipment (drop weights), today’s dynamic loading tests
are of high quality and yield good agreement with static tests. However, differences due exist; they
can often be avoided by understanding the effects of the dynamic loading on the soil and potential
error sources in both static and dynamic test procedures and interpretations. Increased use of
instrumented dynamic and static testing will lead to further understanding of the load transfer in

the subsurface materials, promising greater acceptance and therefore improved economies of deep
foundations.
Dynamic testing checks both bearing capacity and pile integrity, features which make them
widely accepted and specified. Modern building codes have recognized this and included them as
an alternative to traditional methods. These codes tend to reward a comprehensive testing program
with lower factors of safety and therefore with improved economy (Likins, 2015).
7. ACKNOWLEDGEMENT
The authors wish to thank Berkel and Company Contractors as well as Six Construct for
contributing to this paper by releasing the data of the examples demonstrating the use of
instrumentation on steel piles as well as ACIP piles.
8. REFERENCES
AASHTO, (2014), LRFD Bridge Design Specifications, 7th Ed., Pub. Code LRFDUS-7, American
Assoc. of State Highway and Transportation Officials, Washington DC, 20001
Allin, R., Likins, G., Honeycutt, J., (2015), Pile Driving Formulas Revisited. Proc. of the Intl.
Foundations Congress and Equipment Expo 2015: San Antonio, TX. American Society of
Civil Engineers: Reston, VA
Alm, T., and Hamre, I. (2001), Soil Model for Pile Driveability Predictions Based on CPT
Interpretation, Proc. of the 15th Int. Conf. on Soil Mechanics and Geotechnical Engineering,
2, Istanbul, Turkey, 1297-1302.
ASTM D1143, “Standard Test Methods for Deep Foundations Under Static Axial Compressive
Load”, American Society for Testing and Materials, West Conshohocken, PA.
ASTM D3689, “Standard Test Methods for Deep Foundations Under Static Axial Tension Load”,
American Society for Testing and Materials, West Conshohocken, PA.
ASTM D 4945, “Standard Test Method for High Strain Dynamic Testing of Deep Foundations”,
ASTM D5882, Standard Test Method for Low Strain Integrity Testing of Deep Foundations,
ASTM D6760, Standard Test Method for Integrity Testing of Concrete Deep Foundations by
Ultrasonic Crosshole Testing, American Society for Testing and Materials, West
Conshohocken, PA.
ASTM D7383, Standard test Methods for Axial Compressive Force Pulse (Rapid) Testing of Deep
Foundations, American Society for Testing and Materials, West Conshohocken, PA.
ASTM D7949, Standard Test Method for Thermal Integrity Profiling of Concrete Deep
Foundations, American Society for Testing and Materials, West Conshohocken, PA.
Beringen, F.l., van Hooydonk, W.R., and Schaap L.H.J., (1980) Dynamic Pile Testing: an Aid in
Analyzing Pile Driving Behavior. Proc. of the Intl. Seminar on the Appl. of Stress wave
Theory on Piles, Stockholm, A.A. Balkema, Rotterdam.
Bermingham, P., Janes, M., (1989), An Innovative Approach to Load Testing of High Capacity
Piles, Int’l Conf. on Piling and Deep Foundations, London.
Brown, D., Dapp, S., Thompson, R., Lazarte, C., (2007), Geotechnical Engineering Circular NO.
8 Design and Construction of Continuous Flight Auger (CFA) Piles, Report No. FHWA-HIF-
07-03, Federal Highway Administration, Washington, D.C.

Brown, D. and Thompson, , W.,R., (2015), Design and Load Testing of Large Diameter Open-
Ended Driven Piles, A Synthesis of Highway Practice, Transportation Research Board,
NCHRP Synthesis 478, Washington, D.C.
Brown, M.J., Powel, J.J.M. (2013) Comparison of rapid load test analysis techniques in clay soils.
ASCE Journal of Geotechnical & Geoenvironmental Engg. 139, 152-161
Elisio Da Silva, P., (1983), Hidrodynamic Expansive Cell, a New Way to Perform Loading Tests,
Arcos, Belo Horizonte, Brazil. (Also in Portugese at VIII COBRAMSEG - Brazilian Congress
on Soil Mechanics, 1986)
Fellenius, B.H., (1990), Guidelines for the Interpretation and Analysis of the Static Loading Test,
Deep Foundations Institute, Sparta, NJ, USA, www.dfi.org..
Fischer, H.C., (1960), New Graphodynamical Pulse Method of Computing Pile-Driving Processes,
Appl. Sci. Res. A9, 93-139.
Goble, G.G., and Rausche, F., (1970a), Pile Load Test by Impact Driving, Highway Research
Record No. 333, Pile Foundations, Washington, D.C., 123-129.
Goble, G., G., and Rausche, F., (1970b), Dynamic Measurements of Pile Behavior. Proc. of the
Conf. on Design and Installation of Pile Foundations and Cellular Structures, Lehigh
University, Lehigh, PA. 165-172.
Hannigan, P.J., Rausche, F., Likins, G.E., Robinson B.R., and Becker, M.L., (2016), Geotechnical
Circular No.12, Design and Construction of Driven Pile Foundations, Volumes I and II,
Report FHWA-NHI-16-010, US Department of Transportation, FHWA, Washington DC,
20590
Hajduk, E., Paikowsky, S., Mullins, G., Lewis, C., Ealy, C., Hourani, N., (2000), The Behavior of
Piles in Clay during Statnamic and Different Static Load Testing Procedures, Second
International Statnamic Seminar, Tokyo, Publisher: Balkema, 59-73.
Hussein, M., Goble, G., (2004), A Brief History of the Application of STRESS-WAVE Theory to
Piles. Current Practices and Future Trends in Deep Foundations, Geotechnical Special
Publication No. 125.
Hussein, M., Likins, G., and Rausche, F., (1996), Selection of a Hammer for High Strain Dynamic
Testing of Cast-in-Place Shafts, Proc. Of the Fifth Int. Conf. on the Application of Stress-
wave Theory to Piles, Orlando, FL, 759-772.
Hussein, M., Sharp, M., and Knight, W.F., (2002), The Use of Superposition for Evaluating Pile
Capacity, Deep Foundations, an International Perspective on Theory, Design, Construction
and Performance, Geot. Special Publ. No. 116, American Society of Civil Engineers, 6-21.
Likins, G., (2015), Pile Testing – State of the Art, 8th Seminar on Special Foundations Engineering
and Geotechnics, Sao Paulo, Brazil, June 2015
Likins, G., Fellenius, B., Holtz, R., (2012a), Pile Driving Formulas: Past and Present. Full-Scale
Testing and Foundation Design: ASCE Geo-Institute Geotechnical Special Publication No.
227; 737-753.
Likins, G., Liang, L., Hyatt, T., (2012b), Development of Automatic Signal Matching Procedure -
iCAP®. Proceedings from Testing and Design Methods for Deep Foundations; IS-Kanazawa:
Kanazawa, Japan; 97-104.
Likins, G., Rausche, F., (2004), Correlation of CAPWAP with Static Load Tests. Proceedings of
the Seventh Int’l Conf. on Application of Stresswave Theory to Piles: Petaling Jaya, Selangor,
Malaysia; 153-165.

Likins, G., Rausche, F., (2000), Recent Advances and Proper Use of PDI Low Strain Pile Integrity
Testing. Proceedings of the Sixth Int. Conf. on the Appl. of Stress-wave Theory to Piles 2000:
São Paulo, Brazil; 211-218.
Likins, G., Rausche, F., (2014), Pile Damage Prevention and Assessment Using Dynamic
Monitoring and the Beta Method. From Soil Behavior Fundamentals to Innovations in
Geotechnical Engineering. ASCE GeoInstitute Geotl. Special Publication No. 233: Reston,
VA; 428-442.
Likins, G.E., Rausche, F., Webster, K., Klesney, A., 2007. Defect Analysis for CSL Testing. Geot.
Special Publ. No. 158 Contemporary Issues in Deep Foundations; Proceedings from Geo-
Denver 2007 New Peaks in Geotechnics: Denver, CO. (CD-ROM)
Likins, G., Webster, S., Saavedra, M., (2004), Evaluations of Defects and Tomography for CSL.
Proceedings of the Seventh Int. Conf. on the Appl. of Stresswave Theory to Piles: Petaling
Jaya, Selangor, Malaysia; 381-386.
Matsumoto, T., Tsuzuki, M., Michi, Y., (1994), Comparative Study of Static Loading Test and
Statnamic on a Steel Pipe Pile Driven to Soft Rock, Proc. 5th Int’l. Conference and Exhibition
on Piling and Deep Foundations, Bruges, Belgium.
Meunier, J., (1984), Laws of Soil-Pile Interaction in a Pile Driving Simulation Program, Second
Int’l Conf. on the Application of Stress Wave Theory on Piles: Stockholm, Sweden; 326-333.
Middendorp, P., Bermingham, P. & Kuiper, B., (1992), Statnamic load testing of foundation piles,
4th Int. Conf. on the Application of Stresswave Theory to Piles, The Hague 21–24 September
1992, pp. 265–272.
Miyasaka, T., Likins, G., Kuwabara, F., Rausche, F., Hyodo, M., (2009), Improved Methods for
Rapid Load Tests of Deep Foundations. Contemporary Topics in Deep Foundations; 2009
International Foundation Congress and Equipment Expo,
Osterberg, J., (1984), A New Simplified Method for Load Testing Drilled Shafts, Foundation
Drilling, Association of Drilled Shaft Contractors, Dallas TX.
Piscsalko, G., Likins, G. E., (2004), Automated Inspection Control of Augercast Piles. Proceedings
from the M. O'Neill Auger Cast-in-Place Pile Sessions: Recent Experiences & Advancements
in the U.S. and Abroad on the Use of Auger Cast-in-Places Piles, 83rd Annual Transportation
Research Board Meeting: Washington, D.C.
Piscsalko, G., Likins, G., White, B., (2013), Non-Destructive Testing of Drilled Shafts – Current
Practice and New Method. Proceedings from the International Bridge Conference: Pittsburg,
PA.
Rausche, F., and Beim, J., (2012), Analyzing and Interpreting Dynamic Measurements Taken
During Vibratory Pile Driving, Proc. from the Int. Conf. on Testing and Design Methods for
Deep Foundations, IS Kanazawa, Japan, 123-1313.
Rausche, F., (1988) CAPWAPC: Developments and Recommendations. Notes from the PDA
Users’ Seminar, Cleveland, OH, USA.
Rausche, F., Moses, F., Goble, G., (1972). “Soil Resistance Predictions from Pile Dynamics”,
Journal of the Soil Mechanics and Foundations Division, American Society of Civil
Engineers. (Reprinted in Geotechnical Special Publication No. 125, August, 2004. American
Society of Civil Engineers: Reston, VA; 418-440).
Rausche, F., Goble, G., (1979), Determination of Pile Damage by Top Measurements. American
Society for Testing and Materials, Special Technical Publication 670: Philadelphia, PA; 500-
506.

Rausche, F., Liang, L., Allin, R., and Rancman, D., 2004. Applications and Correlations of the
Wave Equation Analysis Program GRLWEAP, Proceedings of the Seventh Int’l Conf. on the
Application of Stresswave Theory to Piles 2004: Petaling Jaya, Selangor, Malaysia; 107–123.
Rausche, F., Nagy, M., Webster, S., and Liang, L., 2009. CAPWAP and Refined Wave Equation
Analyses for Driveability Predictions and Capacity Assessment of Offshore Pile Installations.
Proc. Of the ASME 28th Int, Conf. on Ocean, Offshore and Arctic Engg.; Honolulu, Hawaii,
USA, 1-9.
Rausche, F., Seidel, J., (1984), Design and Performance of Dynamic Tests of Large Diameter
Drilled Shafts. Second Int’l Conf. on the Application of Stress Wave Theory on Piles:
Stockholm, Sweden; 9-16.
Rausche, F., Goble, G., Likins, G., (1985), “Dynamic Determination of Pile Capacity”, ASCE
Journal of the Geotechnical Engineering Division, Vol 111, No 3, 367-383.
Rausche, F., Likins, G., Hussein, M., (1988), Pile Integrity by Low and High Strain Impacts. Third
Int’l Conf. on the Application of Stress-Wave Theory to Piles: Ottawa, Canada; 44-55.
Rausche, F., Likins, G., Miyasaka, T., Bullock, P., (2008), The Effect of Ram Mass on Pile Stresses
and Pile Penetration. Proc. of the 8th Int’l Conf. on Application of Stress Wave Theory to
Piles 2008: Lisbon, Portugal.
Rausche F. and Shen R.K., and Likins, G., (1991), A Comparison of Pulse Echo and Transient
Response Pile Integrity Test Methods, Proc. Of the Transportation Research Board Annual
Meeting, Washington, D.C.
Schellingerhout, A., Revoort, E., (1996), Pseudo Static Pile Load Tester, Fifth Int’l Conf. on the
Application of Stress Wave Theory on Piles, Orlando, FL.
Schmuker, C., (2005), Comparison of Static Load Tests and Statnamic Load Tests, (in German),
Thesis Biberach University, Germany. (Method Shown in English in Paper by Middendorp et
al, (2008), Verification of Statnamic Load Testing with Static Load Testing in a Cohesive Soil
Type in Germany, Eighth Intl. Conf. on the Application of Stress Wave Theory on Piles,
Lisbon, Portugal, IOS Press.)
Seidel, J., Rausche, F., (1984), Correlation of Static and Dynamic Pile Tests on Large Diameter
Drilled Shafts. Second Inter’l Conf. on the Application of Stress Wave Theory on Piles:
Stockholm, Sweden; 313-318.
Smith, E.A.L., (1960), Pile Driving Analysis by the Wave Equation, American Society of Civil
Engineers Journal of Soil Mechanics and Foundations Division, 86(4), 35-61.
Timoshenko, S. and Goodier, J.N., 1951, Theory of Elasticity, 2nd Ed., McGraw-Hill Book
Company, Chapter 15.
Weaver, T., Rollins, K., (2010), Reduction Factor for the Unloading Point Method at Clay Soil
Sites, Journal of Geotechnical and Geoenvironmental Engineering, American Society of Civil
Eng’rs, Reston VA, 643-646.

Deep Foundation Concepts in Energy Geo-engineering.
Special Cases
Santamarina, J.C.(1), and Shin, H.(2)

(1)
KAUST, Saudi Arabia <carlos.santamarina@kaust.edu.sa>
(2)
University of Ulsan <shingeo@ulsan.ac.kr>
ABSTRACT. Challenges within the field of energy geo-engineering require careful

reassessment of deep foundation concepts. We investigate three special cases: deep foundations
in fractured rock, thermo-mechanical ratcheting in energy piles, and the analysis of production
wells where negative skin friction may cause the structural collapse of the well. These problems
require alternative analyses, new constitutive models and adequate numerical simulation
schemes. Commonly used soil-structure interaction solutions may significantly misrepresent the
underlying interaction problem and lead to inadequate results.
1. INTRODUCTION
Energy is required to sustain our lives. The world’s population could reach 9 billion by the year
2040. Population growth and increasing energy demands from developing nations combine to
cause an anticipated 50% increase in the power demand over the next two decades. Fossil fuels
(petroleum, coal and natural gas) account for about 85 percent of the primary energy consumed
worldwide. Fossil fuel reserves exceed the needs for the next century, however, there are
increased concerns with rising CO2 levels in the atmosphere and climate change implications.
The situation is aggravated by the disparity in time scales between the time required for national-
scale decisions (e.g., the 2-to 4-year political cycle), and the time scales for phenomena that
affect energy infrastructure and the environment (e.g., the 50-year life of energy infrastructure
and the 100,000 year half-life of some radioactive isotopes in high-level nuclear waste).
The geotechnical discipline is central to energy. Energy geo-engineering involves site
characterization, infrastructure development, resource recovery, energy storage,
decommissioning, and waste geo-storage. Facilities and infrastructure often require deep
foundations, including wind turbines, tanks, buildings, and towers. However, unique conditions
develop in the context of energy applications.
This manuscript explores three special cases. The first case centers on deep foundations in
fractured rock in compression and tension. The second study addresses the addition of heat
exchangers for underground heat storage, and the potential for thermo-mechanical ratcheting.
The third study extends soil-structure interaction concepts developed for deep foundation to the
analysis of wells used for dewatering and resource recovery, in particular, the generation of
negative skin friction which may cause the structural collapse of the well (the case of sand drains
used to accelerate the compaction of fine-grained layers has similar hydro-mechanical coupled
features). These problems require alternative analyses, new constitutive models, and adequate
numerical simulation schemes.

2. CAISSONS IN FRACTURED ROCK
Fractures exhibit much higher compliance and fluid conductivity than the intact rock matrix,
govern the geo-plumbing of the formation and define planes for localized mechanical
deformation. The prevalent effect of fractures on the hydro-mechanical behavior of rock masses
hinders the application of isotropic elastic models or solutions based on intact rock strength ((see
early efforts in Bray (1977) from Goodman (1989)). Indeed, the orientation of fractures affects
the stress and displacement fields, the settlement and capacity of deep-foundation systems, and
their interaction with nearby structures, such as tunnels and excavations.
The hydro-mechanical simulation of fractured rocks involves either discrete or continuum
approaches. Discrete models include the distinct element method DEM, the discontinuous
deformation analysis DDA (Cundall 1980, Shi 1988, Jing, 2003), however, these fundamentally
robust explicit methods are limited by data reliability and problem size (Gan and Elsworth 2016,
Gutierrez and Youn 2015). Continuum models use fully coupled governing equations to
represent the fractured rock mass, either as an explicit joint-continuum, an implicit joint-
continuum, or an equivalent continuum (Kim 2007). The explicit joint-continuum model
introduces interfaces or joints to represent each discontinuity (Goodman 1968, Desai et al.,
1984). The implicit joint-continuum model incorporates the effect of joint sets into the
constitutive relations for the equivalent rock mass (Amadei and Goodman 1981, Wang and
Huang, 2009). Finally, the equivalent continuum models apply to highly jointed rock masses and
large scales problems (Sitharam et al. 2001). The strong hydro-mechanical coupling between
fluid pressure and stress-deformation-strength in jointed rock mass is a critical model
requirement. Continuum models facilitate the implementation of coupled hydro-mechanical
formulation through the use of force equilibrium and macroscopic mass balance equations.
2.1 Implicit Model – Caisson Analysis

We developed a three-dimensional implicit joint-continuum model for a rock mass with
persistent joint sets of arbitrary orientation and spacing (model details, verification, and
implementation in Shin and Santamarina 2017). We used the model to explore two cases in deep
foundations: (a) an end-bearing caisson and (b) a caisson anchored in a fractured rock mass and
subjected to pull-out. Figures 1a and b compare the mean stress for an isotropic elastic medium,
an inclined transverse-isotropic medium, and a fractured rock simulated with an implicit joint-
continuum model. It can be observed that
• Mean stress contours are severely distorted within the jointed rock mass as compared to the
isotropic medium, in agreement with the orientation of joints.
• The transverse-isotropy model predicts unrealistically deep and biased stress contours (this
results from stress continuity across the joint and the enforced Poisson’s ratio normal to the
isotropic plane for a positive definite stiffness tensor).
These results highlight the need for proper simulation of jointed rock masses, and show the
inadequacy of transverse-isotropy as a surrogate model for the analysis of fractured rocks.

Fig. 1. Caisson in fractured rock: Mean stress field normalized by the stress applied at
the head. (a) Top row: end-bearing. (b) Bottom row: pull-out. Cases: (left) isotropic
medium, (center) transverse-isotropic model inclined at β=45°, (right) implicit joint-
continuum model at β=45°. Caisson diameter 1m, length in rock 5 m. Mechanical
properties of intact rock and joints: normal stiffness ratio Ec/(s⋅Kn)=4, shear stiffness
ratio Ec/(s⋅Ks)=20 (in terms of joint spacing s, intact rock Young’s modulus Ec, normal
and shear stiffness of joints Kn and Ks).
3. DEEP FOUNDATIONS SUBJECTED TO REPETITIVE THERMAL LOADS

The long pile-soil contact length makes piles effective heat exchangers allowing the use of the
soil mass as a heat capacitor to satisfy low-grade energy needs, such as oscillating heating and
air conditioning during winter and summer months (Brandl 2006). Thermal changes alters the
pile-soil interaction due to thermally-driven soil consolidation and creep, changes in radial
effective stress, and the pile thermo-elastic strains (Amatya et al. 2012, Bourne-Webb et al.
2012, Knellwolf et al. 2011, Laloui et al. 2006, McCartney 2011, Ng et al. 2016, and Rotta Loria
and Laloui 2016). Let’s explore the long-term response of thermal-active piles subjected to static
vertical load and thermal cycles.
3.1 Analysis: Model

We adopt a one-dimensional pile-soil load-transfer method modified to account for thermo-
elastic effects in the pile (Coyle and Reese 1966, Poulos and Davis 1980, Knellwolf et al. 2011,
details in Pasten and Santamarina 2014).

Equilibrium. Consider the pile i-th slice (diameter D and slice length Lo). The axial force on the
ith-slice upper interface Qi [N] equals the sum of the axial force on its lower interface Qi+1 [N]
and the mobilized shaft resistance
Qi = Qi +1 + si ⋅ π ⋅ D ⋅ L0 equilibrium (1)
where the shaft resistance is the side friction acting on the ith-slice si [Pa] times the element
contact area.
Side friction si [Pa]. The pile-soil shear strength siult = σ′vi·K0·µ reflects the vertical effective
stress σ′vi [Pa], the coefficient of horizontal stress K0 [-], and the pile-soil interface friction
coefficient µ [-]. The mobilized side friction has an elasto-plastic displacement response:
⎧ ⎛ δ i + δ i +1 ⎞
⎪ − siult if ⎜ ⎟ ≤ −δ s*
⎪ ⎝ 2 ⎠
⎪ ⎛ δ + δ i +1 ⎞ ⎛ δ + δ i +1 ⎞
−δ * < ⎜ i ⎟ < δs , (2)
*
si = si (δ i , δ i +1 ) = ⎨k i ⋅ ⎜ i ⎟ if
⎪ ⎝ 2 ⎠ ⎝ 2 ⎠
⎪ ⎛ δ + δ i +1 ⎞
⎪ siult if δ s* ≤ ⎜ i ⎟
⎩ ⎝ 2 ⎠
where δi [m] and δi+1 [m] are the relative pile-soil displacements at the element upper and lower
interfaces, ki= siult/δs* [Pa/m] is the shaft stiffness, and δs* [m] is the critical relative displacement
to mobilize the pile-soil shear strength siult [Pa]. This value includes the elastic displacement field
away from the pile and it is pile-size dependent (δs* = 0.005·D to 0.02·D - Hirayama 1990, Reese
1978).
Compatibility. The change in the element length Δi = δi+1 − δi [m] combines thermal ΔTi [m] and
stress-dependent effects Δσi [m]:
Qi + Qi +1 L0
Δi = δ i +1 − δ i = ΔTi − Δσi = α ΔT L0 - (3)
2 A⋅ E
in terms of the thermal expansion coefficient α [°C-1], the temperature change ΔT [°C], the pile
cross-section A = π · D2/4 [m2] and Young’s modulus E [Pa].
Pile tip resistance Qtip [N]. We adopt an elasto-plastic tip response, with a constant critical
relative displacement δtip* [m] required to mobilize the ultimate tip resistance Qtipult [N]:
ult
⎧⎛ Qtip ⎞
⎪⎜ * ⎟⎟δ tip
⎜ if 0 < δ tip < δ tip*
Qtip = Qtip (δ tip ) = ⎨⎝ δ tip ⎠ (4)
⎪ Q ult if *
δ < δ tip
⎩ tip tip
where critical relative displacement is in the order of δtip* = 0.1·D.

Normalized Depth
ment i Static
z/L
Ultimate pile capacity Qult [N]. It is the sum of the shaft and the tip resistance Qult = S ult + Qtip
ult
.
head ult head
Then, the factor of safety for an applied load Q is FS = Q /Q .
Temperature. We 0 consider a constant0 amplitude ΔT temperature

0 change that is synchronous
along the pile length (as observed in instrumented piles Bourne-Webb et al. 2009, Laloui et al.
Normalized Depth
2006). Furthermore, we assume that the soil temperature at depth remains relatively unaffected
by daily and seasonal weather changes.
0.5 0.5 0.5 Cycle 1
z/L
3.2 Results
The mobilized shaft friction and axial load changes with the number of cycles. Fig. 2 shows the
initial condition and after 50 cycles
1 1 1
0 0 1000 2000 0 0 30 60 0 0 5 10
Normalized Depth z/L
Normalized Depth
0.5 0.5 0.5 Cycle 50

z/L
1 1 1
0 1000 2000 0 30 60 0 5 10
(a) Axial force
Axial Force Side Friction
(b) Side friction Rel. Displ.
(c) Rel. Displace.
Q [kN]
Q [kN] ss [kPa]
[kPa] δδ [mm]
[mm]
(a) Fig. 2. Thermally active piles. Initial condition

(b) (continuous line) and after 50 thermal
cycles. (a) axial force, (b) side friction, and (c) relative displacement. (from Pasten
and Santamarina 2014).
Heating expands the pile in both directions: up in the upper part and down in the lower part.
Thermal contraction upon cooling partially reverses this trend, and plastic displacements
accumulate. The following cases help clarify trends:
• End-bearing pile with no shaft resistance (Sult/Qtult 0). The pile elongates and contracts with
thermal cycles, but the tip load is constant and there is no thermally induced settlement at the
tip.
• No static load applied Qhead =0 (FS = Qult / Qhead ∞). Shaft and tip resistance become
established after the first thermal cycle and there is no accumulation of permanent
displacements. As a corollary, we can conclude that settlement accumulation during thermal
cycles requires the bias effect of an applied pile load Qhead
• With static load applied Qhead>0. Thermo-mechanical ratcheting (continuous settlement) can
only happen if the applied load Qhead and the shaft resistance Sult exceed the tip resistance
Qhead + Sult > Qtipult. Otherwise, the accumulation of permanent displacement will tend
towards an asymptotic value (shakedown condition).

The asymptotic accumulation of permanent displacement triggered by thermal cycles increases
(1) with a higher applied “bias” load Qhead or decreasing factor of safety Qhead/Qult, and (2) with a
higher ratio between the free thermal elongation ∆T and the displacement required to yield the
shaft resistance δs*, that is ∆T/δs* = (α·∆T·L) / δs*. In fact, large-diameter and low slenderness
piles (smaller L/D) are less susceptible to thermal effects due to the larger critical displacement
δ* required to mobilize shaft resistance and tip capacity, and the relatively lower free thermal
elongation ∆T.
The number of cycles to reach asymptotic conditions increases with the applied load Qhead
and the shaft resistance Sult relative to the tip resistance (Qhead + Sult)/Qtipult. It is also affected by
the selected relative shaft-to-tip critical displacements δs*/δtip*. The pile head displacement δhead
follows an exponential function with the number of thermal cycles n (Pasten and Santamarina
2014)
δ head n→∞ − δ head −n

= e N* . (5)
δ head n→∞
Settlement accumulation reaches 63% of the asymptotic pile settlement when the number of
cycles n equals the characteristic number N*. Most cases analyzed as part of this study show that
the characteristic number of cycles is less than 10-or-20 for common field conditions where the
factor of safety exceeds FS=3-or-4.
4. NEGATIVE SKIN FRICTION: WELL-SOIL INTERACTION

Hydro-mechanical coupling may trigger complex soil-structure interaction problems. For
example, water, oil, and gas recovery involve depressurization, ensuing changes in effective
stress cause reservoir compaction, trigger reversal of shear forces along the well, increase the
radial stress acting on the production screen, and may cause the well failure in buckling due to
excessive axial load, shear or collapse modes. A proper understanding of well-sediment
interaction is required to anticipate the consequences of depressurization and to develop optimal
production strategies (Moridis et al. 2011). The situation applies to sand columns as well.
From a deep foundation perspective, this is a classical problem of drag force causing
excessive axial stress. But there are a few caveats. In particular, the effect may be either “local”
or “global” depending on the well separation, w, and thickness, h, of the strata-bound reservoir
that compacts between two low-permeability layers.
Global: small w/h ratio. In this case, depressurization causes an extensive compression of the
reservoir, all layers above the production layer will move downwards and will tend to mobilize
negative skin friction (typically, this is the case of sand column patterns used in field
compaction). The solution is based on standard soil-pile interaction analyses, where equilibrium
conditions relate the change in the casing axial force Pz [N] at depth z [m] to the mobilized shear
resistance against the casing τz [kPa] at the same depth (Poulos and Davis 1980)
∂Qz
dz = − π d well τ z dz (6)
∂z

This equation can be readily solved in finite differences
Qi − Qi+1 = − (π d well Δz )τ i (7)
This analysis of sediment-well interaction assumes that the sediment column above the
production layer is a perfectly rigid body that settles a prescribed amount across the production
horizon. The solution is sensitive to values selected for yield displacement δy (different values
than those assumed in 3D FEM in order to account for sediment deformation), settlement of the
production horizon, and tip stiffness and bearing capacity.
Typical results in Fig. 3 show the mobilization of the negative skin friction above the
production horizon, high peak axial loads near the top of the production horizon, and the critical
role of tip stiffness on the mobilized peak load. This last observation suggests the need for
engineering well completion using soft-end conditions. Note that large axial loads may force the
well to penetrate into the lower layer (i.e., mobilize tip resistance) or cause the longitudinal
failure of the well.
(a) Shaft Resist. fs [kPa] (b) Pile Load Pz [kN] (c)
-500 0 500 -2000 0 2000 4000
0 0 12,000
20 20 10,000
controlled by
40 40 shaft capacity
Max Axial Load [kN]
8,000
60 60
6,000
Depth z [m]
Depth z [m]
80 80
4,000
100 100
2,000
120 controlled by production
120
induced settlement
0
140 140 0.01 0.1 1 10 100 1000 10000 100000
Tip Stiffness k [MN/m]
160 160
Fig. 3. Production well under “global” settlement condition. (a) mobilized shaft
resistance, (b) axial load, (c) maximum axial load as a function of tip stiffness. Case:
155m long well, casing OD=128mm, production horizon between z=140 and z=155m,.
Local: large w/h ratio. The analysis of single wells, or wells with separation greater than the
layer thickness s/h>>1 cannot be conducted with an equivalent 1D formulation, in fact, we used
a fully coupled hydro-mechanical finite element model to explore the consequences of
depressurization. In particular, the analysis must carefully capture the evolution of hydraulic
conductivity k [m/s] as a function of the void ratio e: we used a power equation k/k0=(e/e0)b
where k0 is the hydraulic conductivity at the reference void ratio e0, and the b-exponent captures
the sensitivity of hydraulic conductivity to changes in the void ratio. For fine grained sediments
the exponent can be b=4 and larger (Ren and Santamarina 2017). The coupled hydro-mechanical
FEM model shows the complex nature of sediment-well interaction during depressurization (Fig.
4). Trends from the parametric study show:

• The depressurization field is limited to a narrow region around the well, i.e., “local”, a
compression bulb forms around the well, and settlements are much smaller than 1D-settlements
computed for the same depressurization. Upwards displacement is often observed from layers
beneath the production horizon.
• Tension develops in the well above the production horizon.
• The maximum compression takes place within the production horizon, and it can reach the
longitudinal failure load for the casing during high depressurization. Axial loads are smaller or
equal to those computed under “global” layer compaction conditions.
• Higher b-exponents cause narrower depressurization fields and more “local” sediment-well
interaction (Fig. 4). Conversely, the assumption of a constant hydraulic conductivity k leads to
a gross overestimation of the axial force in the well.
Casing axial force Q [MN]
Casing axial force [MN]
b=0.0
b=0.0 b=2.0
b=2.0 -2 0 2 4 6 8 10
0
-20 b=0
b=1
-40 b=2
b=3
-60
Depth [m] b=4
Depth [m]
-80 b=5
b=4.0
b=4.0 b=6.0
b=6.0
b=6
-100
-120
-140
-160
Fig. 4. Production well under “local” settlement conditions. (left) Pore water pressure
around the production zone for various b-exponents in the constitutive model for
hydraulic conductivity. (right) Axial force distribution along the casing.
5. CONCLUSIONS
Fracture characteristics and orientation greatly affect the stress field, deformation and strength of
deep foundation systems in fractured rock masses. Special caution is required when analyzing
caisson-caisson and caisson-boundary interaction such as in the vicinity of excavations and
tunnels. The situation is aggravated when changes in hydrological conditions are anticipated.
Energy piles subjected to thermal cycles experience thermally-induced pile displacements
that can result in the accumulation of plastic displacements with the number of thermal cycles,
even though, the ultimate pile capacity remains constant. Ratcheting is unlikely in most
applications. Instead, the plastic displacements accumulate towards an asymptotic settlement that
is proportional to the free thermal expansion of the pile and inversely proportional to the factor
of safety. Preliminary results suggest that most of the thermally-induced plastic displacements
take place in the first few cycles (typically less than 10 to 20 cycles for standard applications).

Drag forces triggered by hydro-mechanical coupling controls the design of deep production
wells as well as sand-drains. Global settlement conditions prevail when nearby wells cause the
compaction of the produced layer in an extended area, and can be analyzed as a 1D soil-pile
interaction problem. Conversely, local conditions and a narrow compaction bulb develop in
isolated wells and thin production horizons. Settlement, axial loads and strain fields are very
distinct in these two cases.
Modeling has gained increased acceptance in deep foundation analyses. However, analysts
must remain vigilant of underlying assumptions. Examples presented here highlight the
consequences of inadequate analyses: (1) equivalent continuum analyses using transverse-
isotropy as a surrogate models for fractured rock (in addition, the implicit formulation is most
beneficial when hydro-mechanical coupling is anticipated). (2) 1D soil-pile interaction analyses
can be misleading when conditions are local or layer bound, as demonstrated in the case of
production wells.
6. ACKNOWLEDGMENTS
Support for this research was provided by the KAUST Endowment at King Abdullah University
of Science and Technology. G. Abelskamp edited the manuscript.
7. REFERENCES
Amadei, B. and Goodman, R.E. (1981). A 3-D constitutive relation for fractured rock masses,
Proceedings of the international symposium on mechanical behaviour of structured media,
pp. 249-268.
Amatya, B. L., Soga, K., Bourne-Webb, P. J., Amis, T., and Laloui, L. (2012). Thermo-
mechanical behaviour of energy piles. Géotechnique, 62(6), pp. 503-519.
Bourne-Webb, P. J., Amatya, B., and Soga, K. (2012). A framework for understanding energy
pile behaviour. Proceedings of the ICE - Geotechnical Engineering. pp.?
Bourne-Webb, P. J., Amatya, B., Soga, K., Amis, T., Davidson, C., and Payne, P. (2009). Energy
pile test at Lambeth College, London: geotechnical and thermodynamic aspects of pile
response to heat cycles. Géotechnique, 59(3), pp. 237-248.
Brandl, H. (2006). Energy foundations and other thermo-active ground structures. Geotechnique,
56(2), pp. 81-122.
Cundall, P.A. (1980). UDEC-a generalized distinct element program for modelling jointed rock.
Report PCAR-1-80, Peter Cundall Associates, European Research Office, US Army Corps
of Engineers- pp.??
Coyle, H. M., and Reese, L. C. (1966). Load transfer for axially loaded piles in clay. Journal of
Soil Mechanics and Foundations Div, 92(2), pp. 1-26.
Desai, C.S., Zamman, M.M., Lightner, J.G., and Siriwardane, H.J. (1984). Thin layer element for
interfaces and joints, Int J Numer Anal Methods Geomech, 8, pp. 19-43.
Gan, Q., and Elsworth, D. (2016). A continuum model for coupled stress and fluid flow in
discrete fracture networks, Geomech Geophys Geo-Energy Geo-Resour, 2(1), pp. 43-61
Goodman, R.E., Taylor, R.L., and Brekke, T.L. (1968). A model for the mechanics of jointed
rock, J Soil Mech Div ASCE 94, SM3. pp. 637-59.
Goodman, R. E. (1989). Introduction to Rock Mechanics, Wiley, New York

Gutierrez, M. and Youn, D.J. (2015), Effects of fracture distribution and length scale on the
equivalent continuum elastic compliance of fractured rock masses, Journal of Rock
Mechanics and Geotechnical Engineering, 7, pp. 626-637
Hirayama, H. (1990). Load-settlement analysis for bored piles using hyperbolic transfer
functions. Soils and Foundations, 30(1), pp. 55-64.
Jing, L. (2003). A review of techniques, advances and outstanding issues in numerical modelling
for rock mechanics and rock engineering, Int J Rock Mech Min Sci Geomech, 40, 283-353.
Kim, J.M. (2007). Hydraulic conductivity and mechanical stiffness tensors for variably saturated
true anisotropic intact rock matrices, joints, joint sets, and jointed rock masses,
Geosciences Journal, 11, pp. 387-396.
Knellwolf, C., Peron, H., and Laloui, L. (2011). Geotechnical analysis of heat exchanger piles.
Journal of Geotechnical and Geoenvironmental Engineering, 137(10), pp. 890-902.
Laloui, L., Nuth, M., and Vulliet, L. (2006). Experimental and numerical investigations of the
behaviour of a heat exchanger pile. International Journal for Numerical and Analytical
Methods in Geomechanics, 30(8), pp. 763-781.
McCartney, J. (2011). Engineering performance of energy foundations. Pan-Am CGS
Geotechnical Conference, Toronto, Canada.
Moridis G.J., Collett T.S., Pooladi-Darvish M., Pooladi-Darvish M., Santamarina C., Boswell R.,
Kneafsey T.J., Rutqvist J., Reagan M.T., Sloan E.D., Sum A., and Koh, C. (2011).
Challenges, Uncertainties, and Issues Facing Gas Production from Gas-Hydrate Deposits.
SPE Res Eval and Eng 14(1): 76-112. SPE-131792-PA. doi: 10.2118/131792-PA.
Ng, C. W. W., Gunawan, A., Shi, C., Ma, Q.J. and Liu, H. L. (2016). Centrifuge modelling of
displacement and replacement energy piles constructed in saturated sand: a comparative
study. Géotechnique Letters. Vol. 6, pp. 34-38.
Pasten, Cesar, and J. Carlos Santamarina. (2014), Thermally Induced Long-Term Displacement
of Thermoactive Piles. Journal of Geotechnical and Geoenvironmental Engineering, vol.
140(5).
Poulos, H. G., and Davis, E. H. (1980). Pile foundation analysis and design, Wiley and Sons,
New York.
Reese, L. C. (1978). Design and construction of drilled shafts. Journal of the Geotechnical
Engineering Division, 104(1), pp. 91-116.
Ren, X-W., and Santamarina, J.C. (2017), The Hydraulic Conductivity in Sediments,
Geophysical Journal International, under review.
Rotta Loria A.F., Laloui, L. (2016), The interaction factor method for energy pile groups,
Computers and Geotechnics, vol. 80, pp. 121-137.
Shi, G. (1988), Discontinuous deformation analysis-a new numerical model for the statics,
dynamics of block systems. PhD thesis, University of California, Berkeley, USA.
Shin, H. and Santamarina, J.C. (2017), Implicit joint-continuum model for hydro-mechanical
analysis in jointed rock mass, under review.
Sitharam, T.G., Sridevi, J. and Shimizu, N. (2001). Practical equivalent continuum
characterization of jointed rock masses, Int J Rock Mech Min Sci, 38: pp. 437-48.
Wang, T.T., and Huang, T.H. (2009). A constitutive model for the deformation of a rock mass
containing sets of ubiquitous joints, Int. J. Rock Mech. Min. Sci., 46(3), pp. 521-530.

Expander Body and Toe-Box: Expansion Devices for Deep
Foundations Enhancement
Terceros A.M.(1), Terceros H.M.A.(1)
(1)
Incotec S.A., Santa Cruz de la Sierra, Bolivia <mta@incotec.cc>, <math@incotec.cc>
ABSTRACT. The paper presents two technologies for deep foundations expansion devices for
increasing stiffness and reducing uncertainty of foundation: the Expander Body (EB) and the Toe
Box (TB). The common characteristic of both devices is their ability to stiffen the toe response by
pre-compressing the soil at the toe level. This is achieved by expanding the devices by pressure
grouting, the EB expands horizontally and the TB vertically. Both devices compress the soil, which
increases its strength and stiffness. During grouting, pressure and volume are measured, providing
important information of the initial and final state of stress of the soil. The measured pressure-
volume relationship can be correlated to geotechnical parameters, similar to a pressuremeter test.
The measured pressure-volume parameters, combined with the soil characterization, allow
estimating the toe response for every pile. The EB and TB devices can be used as quality control
of the performance of each pile, as it becomes possible to identify unexpected soft deposits and to
take remedial measures during the construction phase.
1 THE EXPANDER BODY

1.1 Expander Body Technology
The Expander Body (EB) is made of a thin steel sheet and folded into a cylindrical body that is
installed at the toe of a pile or as a soil anchor (cf. Berggren et al. 1988; Massarsch 1994). The
technology was developed in Sweden during the 1980s. Since then, several thousands of EB have
been installed worldwide (e.g. in Sweden, Norway, Germany, Japan, Korea).
An important aspect of the system is that inflation of the EB at the toe of a pile compresses
the soil laterally, improving the pile toe response (Massarsch and Wetterling 1993). In granular
soils, the expansion process compacts the soil and increases horizontal stresses. In fine-grained
soils, such as clays and silts, expansion results in an increase of total stress and pore water pressure.
When the excess pore water pressure dissipates, reconsolidation leads to higher effective stresses
and thus increased shear strength and stiffness. In soft rocks and very stiff clays, the high grouting
pressure creates very high friction between the soil and the EB, even at very small deformations.
The EB was initially conceived mainly for sandy and silty soils, considering those soils as
optimal for densification. During the past 20 years, the company Incotec has redesigned the
expander body (EBI) and added new features, such as an improved steel folding process and a
novel post-grouting device, which makes it possible to further improve the soil below the EB.
Around 20.000 EBIs have since been installed in Bolivia and elsewhere (Argentina, Brazil,
Paraguay, Peru, México, USA and Canada). As will be demonstrated below, the enhanced EBI
system can be installed in a wide range of soils, from soft to very stiff clay, silts and sands as well
as gravely soil and even soft rock (soft sandstone).
The EB is placed at the toe of a bored or driven pile and subsequently expanded by injection
of cement grout, creating a bulb. Figure 1 shows the gradual expansion of an EB.

Fig. 1. Steps of EB expansion and computer-monitoring of pressure and grout volume.
Table 1 shows the size of EB models, prior to and after full expansion. Note that as a result
of inflation, the length of the EB is reduced. As this effect can result in some soils in
decompression below the pile toe, the above-described post grouting device can be used to
increase the pressure below the EB, thereby enhancing the strength and stiffness of the soil
below the EB.
Table 1. EB models, sizes and their applications.
1.2 Description of the EB System

The EB resembles the bulb of a Franki pile or other bulb piles, created by compaction of concrete
at the pile toe. However, an important advantage of the EB system is that grouting is a controlled
and fully monitored process. As the grout volume and the applied pressure are measured for each
pile, it is possible to estimate the soil conditions around the pile toe in a similar way as the Menard

pressuremeter test (PMT). The monitoring system consists of a pressure gauge and an ultrasonic
flow meter. Both instruments send the signal to a data logger where the values are processed in
order to get three basic records: flow vs. time; pressure vs. time and pressure vs. volume. The
software developed by Incotec displays a pressure vs. volume curve, which is required for the
interpretation of the EB behavior. As the grout is surrounded by an impermeable steel body, grout
does not permeate into the soil or cause fracturing. The toe area of the partially or fully expanded
body can be estimated based on the measured grout volume. Each injection record provides a
volume-pressure curve, cf. Figure 2. The expansion process can be divided into three phases: initial
phase (recompression of disturbed soil), elastic phase (compression of soil) and stabilizing phase
(soil in a plastic state – similar to limit pressure in PMT).
Fig. 2. Different phases of the expansion process.
Inspection of the pressure-volume curve provides valuable information regarding the strength
and deformation properties of the soil at the pile toe. An important advantage is that soil properties
are obtained at the pile toe after installation of the pile, thereby taking into account the effects of
pile installation process. This aspect is illustrated by the following example. Figure 3 shows the
effect of pile installation method on the pressure-volume behavior of an EB. Piles TP3 and TP4
were located a few meters apart and founded on a dense sandy soil with SPT value N60>20. TP3
was installed to a depth of 9.6m and TP4 to a depth of 17.5m.
Fig. 3. Effect of the pile installation method on the pressure-volume response. TP3: full
displacement auger pile, 9.6m long, TP4: slurry drilled pile, 17.5m long.

The EB of pile TP3 was installed as a Full Displacement auger pile (FDP) to 9.6m depth while
TP4 was installed as a slurry drilled pile to 17.5m depth. The slope of the pressure grout volume
curve of TP3 was at the beginning of expansion essentially vertical, indicating that the soil
surrounding the EB is stiff. In the case of pile TP4, expansion of the EB required larger lateral
displacement to reach the limit pressure, indicating that the soil had been loosened as a result of
the drilling process using bentonite. The difference in the limit pressure between pile TP3 and pile
TP4 is due to the significantly higher stress level at 17.5 m depth (TP4).
Variable soil conditions are common in quaternary sedimentary deposits where monitoring of
the grouting process makes it possibility to detect softer layers, not detected during soil
investigation. Figure 4a (pile A) and 4b (pile B) present the injection curves from two different
piles at the same project. Both piles are 16m long and were to be founded on a dense sandy layer.
Nevertheless, during the grouting of pile B, the limit pressure was only 36 % of the limit pressure
of pile A, indicating an unexpected softer soil at the pile toe. Based on this information it was
possible to reinforce the foundation, while still in this phase of construction, avoiding the risk of
settlement problems.
a) Pile A - injection diagram in dense sandy soil
b) Pile B - injection diagram in softer soil

Fig. 4. Examples of EB grouting diagrams of two piles on same project site.

Figure 5 shows the injection curves of piles with EB 612 founded in soft sandstone, in
Asunción Paraguay. The piles were drilled under bentonite in a 400 mm diameter shaft. The initial
stiffness of the pressure-volume curve is clearly visible.
Fig. 5. Pressure vs. injection curves of piles with EB 612, founded in weathered sandstone in
Asunción-Paraguay.
1.3 Installation Methods

As the initial diameter of the EB is small (cf. Table 1), it is possible to install EBs by a variety of
methods, from drilled and augered piles to jacked or vibrated piles, cf. Figure 5a and b. Figure 5c
shows the EB application for underpinning and reinforcement of foundations in areas with limited
head room, using small-diameter bored piles. The main factor, which governs the required EB size
is, beside the soil type, the diameter of the pile shaft. The EB diameter after expansion should
always be at least 10 cm larger than the pile shaft diameter. In practice, this value is 15 cm but,
when used in combination with micropiles, this value can be as large as 60 cm (20 cm shaft, 80
cm EB). The sequence of grouting EB piles within a pile group is similar to that of installing driven
piles in sandy soils. At first, the central piles of the groups should be grouted first and thereafter
the surrounding piles. At the end of the injection of all EBs in a pile group, the density of the soil
between the piles is maximized.

a) Auger pile with EB b) Vibrated EB pile
c) Small-diameter bored pile

Fig. 6. Examples of EB-pile installation methods.
In the case of auger piles, the EB can be grouted shortly after installation and up to a period
of several weeks. The maximum time period of grouting after installation of auger piles was 7
months. As the concrete cover of EB auger and drilled piles under bentonite is relatively thin and
expansion of the EB produces tension stresses, the grout cover breaks easily during expansion of
the EB. In a test performed in a concrete cylinder of 60 cm in diameter, the needed pressure to
break the grout cover was only 200 kPa. Thus, grouting of EBs in a pile group can be performed
in time intervals of one or two weeks.
1.4 Post Grouting

During expansion, an EB shortens typically by about 10 %. Two situations can occur. If the shaft
resistance of the pile at the time of grouting is not yet fully developed, the pile shaft will be pulled
down. Alternatively, the pile shaft keeps its position and due to shortening of the EB, a zone of
partially decompressed soil is created below the EB. To eliminate the potential problem of reduced
stiffness below the EB, a device has been developed, which makes post-grouting below the inflated
EB (EBI) possible. An example of a post-grouted EBI is shown in Figure 7.

Fig. 7. Exhumed EBI with post grouting, showing the final shape of the body, where the
diameter of the upper part is some bigger than in the lower one. This EB was installed in dense
sandy soil.
Post-grouting of the soil can be used to increase the stress below the inflated EB, cf. Figure 8.
Figure 8a illustrates the stress field developed during EB expansion (red arrows) and the
subsequent post-grouting (yellow arrow). A special valve was developed for injecting grout under
controlled conditions below the EB. During the post-grouting process, grout volume and injection
pressure are measured, thus providing additional information about the response of the soil below
the pile toe. Figure 8b shows a typical post-grouting diagram. Note that this diagram was obtained
after EB expansion had been carried out.
Fig. 8a. Illustration of stress field due to grouting Fig. 8b. Post-grouting diagram

1.5 Quality Control
Quality control (QC) is an important aspect of the EB system as it allows the verification of each
EB after installation. The QC process comprises three steps: a) determination of injection curve
(pressure vs. volume) of all EBs, b) evaluation of the service load of each pile based on past
experience, see Table 2, and c) development of a database, where the grouting pressure vs volume
relationship can be compared with results from pile loading tests and site-specific geotechnical
information (e.g. CPTU or SPT).
Figure 9a, 9b and 9c show the limit pressure from EB injection measurements at three projects:
a) medium dense sandy soils, b) clayey gravely soil and c) very stiff clay. The three cases provide
important statistical data. The dispersion demonstrates the typical heterogeneity of sedimentary
soils. In these cases, the gravely soil is less homogeneous. Unexpected variations obtained from
volume-pressure measurements are immediately detected after EB injection. Thus, if necessary,
the pile groups can be redesigned and provided with additional EBs. Alternatively, additional
post-grouting can be carried out as described above. In the below shown cases, design could be
based on lower boundary values of N (SPT) or qc (CPT) for the defined depth of installation, thus
making the foundation solution more economical and still providing full control of pile quality.
TABLE 2. Typical report of EB test results, providing design information for the EB database.

Pressure (MPa) Volume (l)
AVERAGE 3.66 116.2
STANDARD DEVIATION 0.68 20.1
Fig. 9a. Variation of grouting results in a project with EBs embedded into a medium dense sand
layer.
Pressure (MPa) Volume (l)

AVERAGE 3.06 185.09
Fig. 9b. Variation of grouting results in a project with EBs embedded into a gravely clay layer.

Pressure (Mpa) Volume (l)
AVERAGE 4.00 124.98
Fig. 9c. Variation of grouting results in a project with EBs embedded into a very stiff clay layer.
Extensive pile loading tests have been performed on piles with and without EBs. In all cases
the pile toe response with EB shows a remarkable increase of service load. For similar geometry
(pile diameter ranging between 350mm and 450 mm, lengths between 12 m and 18 m) and similar
soils (mostly silty sandy soils, medium dense, with clayey layers), the piles with EB have service
loads 2.5 to 4 times higher than identical piles without EB, for the same deformations measured at
the pile head. This behavior is consistent with the increase of the toe response, (cf. Fellenius et al,
2014, Terceros et al. 2014, Rosales et al. 2015). Table 3 shows statistical information about average
values and standard deviation values of injection pressure from different projects.
TABLE 3. Examples of injection pressures and their variation from projects in different
geotechnical conditions.
STANDARD % of SD REFERRED
AVERAGE GROUT
PROJECT DEVIATION TO AVERAGE
PRESSURE (MPa)
(MPa) VALUE
SILOS SOFIA 1.97 0.400 20.263
TRES CARABELAS 4.54 0.680 14.978
SANTA ELENA 4.06 0.622 15.320
FLORENCIA 3.81 0.721 18.924
ITAGUAZÚ 4.44 0.784 17.659
UAGRM 3.89 0.432 11.112
PARQUEO ALAS 3.63 0.330 9.091
ITATI 4.96 0.612 12.342

STANDARD % of SD REFERRED
AVERAGE GROUT
PROJECT DEVIATION TO AVERAGE
PRESSURE (MPa)
(MPa) VALUE
FERROTODO 4.53 0.448 9.897
DOMANI 3.66 0.611 16.700
CLÍNICA NUCLEAR 1.76 0.330 18.768
TORRE DUO 3.30 0.400 12.121
EDIFICIO ZURITA 1 2.48 0.364 14.679
ALIANZA 5.22 0.760 14.560
ARTEMISA 2.99 0.561 18.760
TORRE SALTO 3.14 0.555 17.690
TORRE EQUIPETROL 4 3.43 0.397 11.567
2 THE TOE BOX

2.1 Brief description of the technology
The concept of expanded-base piles has been around since the 1930s and toe grouting has been
used for more than 50 years (Bolognesi and Moretto 1973, Bruce 1986). In some instances,
problems can develop controlling results of the grouting, e.g., the grout can flow up along the pile
shaft or outward away from the area below the pile toe.
These problems are avoided by the herein presented method of equipping the pile with a "toe
box", which essentially is a telescopic arrangement of two cylinders that are moved apart, during
a pressure-grouting process. The box is wrapped with a non-woven geotextile that acts as a filter
containing the grout when, at the end of the grouting, some leakage may occur. Figures 10a and
10b show photographs of a toe-box in its final stages of production and Figure 10c shows a toe-
box attached to the end of a reinforcing cage ready for installation in the pile.
The box outer diameter is 15 % smaller than the nominal pile diameter. The toe box has been
installed in over 200 piles in 4 different Bolivian bridge projects on sedimentary soils. All of the
bridge piles had diameters ranging between 1,200 mm and 1,500 mm and lengths ranging between
21 and 35 m.
Fig. 10a. Toe Box. Fig. 10b. Toe Box with non-woven
geotextile.

Fig. 10c. Toe Box attached to the reinforcing cage.
2.2 Installation and behavior of the Toe-Box

After the drilling of pile, the Toe Box is lowered into the hole together with the reinforcement
cage, as illustrated in Figure 11. It is very important that the grouting pipes and the telltales are
secured to the cage before the installation in order to ensure proper grouting results and reliable
telltale records.
Fig.11. Toe Box installation in a large diameter pile.
About one week after concreting the pile, the toe-box is expanded axially by pressure-grouting
with the shaft resistance serving as reaction. The expansion of the toe-box (mostly the pile toe
displacement) is measured by two telltales. All associated pile compression and pile head
movements are also measured during the process. The grouting process is slow, taking at least an
hour to complete, in order to prevent build-up of pore pressure. The grout pressure and grout
volume is monitored at the pump and the records are saved and graphically displayed.

The grouting is terminated based on the measured grout pressure or volume, or upward
movement, if any, of the pile. Usually, grouting is terminated when the Toe-Box has expanded
completely and the grouting pressure exceeded 1.5 MPa. If the pressure is smaller, indicating a
soft soil, the grout volume is increased to 200 % of the Toe-Box nominal volume, i.e., significant
leakage is accepted. The increase is carried out in steps with 10 to 20 minutes pauses. If the pile
head moves upward more than 5 mm, grouting is terminated, as this shows that the pile shaft
resistance is close to be exceeded and, therefore, further expansion of the toe-box does not produce
any further benefit.
The grout pressure in the toe-box is the pressure at the grout pump. The additional pressure at
the Toe Box due to the column of grout between the grout pump and the toe-box is equal to or
marginally larger than the friction in the grout tube, i.e., the pump reading represents the toe-box
grout pressure.
The injection causes the toe-box to expand axially, thus, compressing the soil under the pile
toe, which preloads the soil and, therefore, increases the stiffness of the pile toe response.
Although, theoretically, a 300-mm axial expansion is possible, the actual axial expansion is limited
to about 250 mm, which amounts to a theoretical grout volume of 400 L. This limitation is imposed
by the size (height) of the Toe Box.
Figure 12 shows a typical grouting report, where the maximum grouting pressures, grout
volume, and pile head and pile toe displacements are shown; pressure/volume diagram and force
displacement diagram are also shown, illustrating the densification achieved during the grouting
of the Toe-box. As the Toe-Box is expanded, the stiffness of the toe increases with the
displacement, to a point (approximately at 42 mm) where a higher stiffness increase occurs and
practically no more downwards movement is recorded. Coincidently, at this point, the pile head
starts to move upwards. The densification produced by the toe-box expansion, which increased the
toe stiffness, made it possible to partially overcome the shaft resistance, allowing the complete pile
to have an 8 mm upwards movement. Notice that the shaft resistance has not been fully mobilized,
as the grouting pressure and it corresponding load, has not reached a final maximum stabilized
level.

Pile Information
CAP Nº: H5 Diameter (mm): 1200
PILE Nº: C3 Lenght (m): 30
Injection Data
Max. Surface Pressure (MPa): 5.66
Total Volume (l): 539.70
Total movilized force (kN): 8803.6
Total upward movement (mm): 8
Load at the beginning of the uplift (kN): 2000
DIAGRAM UNIDIRECTIONAL FORCE VS. MOVEMENT
Force vs. Movement

Fuerza (kN)
0.0 1000.0 2000.0 3000.0 4000.0 5000.0
-20
Head, upward movement
-10
0
Movement (mm)
10
20
30
40
50 Toe, downward movement
60
INJECTION DIAGRAM PRESSURE VS. VOLUME

7
6
Pressure vs. volume
5
Pressure (MPa)
0
0 100 200 300 400 500 600
Volume (L)
Fig. 12. Typical Toe-Box report.
2.3 Soil type detection

Figure 13a shows the pressure/volume behavior of two piles (C6N3 and C9N3) located in two
different piers 120 m apart from each other. The cast-in-situ piles were 1,500 mm in diameter and
drilled with temporary casing instead of slurry. The piles were constructed to a depth of 21.5 m.
The expansion proceeded similarly for the both piles. The occasional low-pressure spikes occurred
due to the staged grouting procedure, i.e., the grouting is interrupted every 50 L, and resumed after
a wait time of 10 to 20 minutes.

Figure 13b shows the measured grout pressures converted to load at the toe-box level versus
the movements of the toe-box’s bottom (pile toe) and pile head—downward for the pile toe and
upward for the pile head. These force-movement curves are presented in a format similar to that
resulting from a bidirectional pile test. Note that the ordinate axis has different scales for upward
and downward movement.
Fig. 13a. Injection Diagrams for piles C6N3 AND C9N3.
Fig. 13b. Load-Movement diagrams.

The piles responded differently during the grouting, as reflected in the load-movement curves.
According to the soil profile (Figure 14), Pile C6N3 was embedded in a dense sandy layer with
13% of fines content, while Pile C9N3 was embedded in a layer of high plasticity clay with 93 %
fines content. The grout pressure (Figure 13a) reflected the type of soil in which the toe-box was
expanded. Pile C9N3 showed a low final grouting pressure and no further increase in pressure
occurred for continued pumping of grout, which is typical of expansion in a clayey layer. In
contrast, when grouting Pile C6N3, the pressure kept increasing together with the continued
expansion. The approximate pile toe stiffness calculated from the grouting diagram is consistent
with this observation. While the toe-box from Pile C6N3 was not able to expand further than 37
mm due to the toe stiffness response from the compressed sand, in Pile C9N3, the Toe Box
expanded to its limits with no increase in toe stiffness. The pile toe apparently failed.
Comparing the results shown in Figure 12 with the piles explained in Figure 13a and 13b, it
is noted that the grouting pressure measured for Pile H5C3 (Figure 12) is very similar to that of
Pile C6N3 in Figure 13a, which was embedded in a dense sand layer. Thus, it is possible to
conclude that the toe of Pile H5C3 is also embedded in a coarse-grained soil.
SOIL PROFILE - PILE C6N3 SOIL PROFILE - PILE C9N3

DEPTH SOIL % OF DEPTH SOIL % OF
1500mm pile, L=21.5m (M) DESCRIPTION FINES (M) DESCRIPTION FINES
0.0 0.0
LOOSE TO
MEDIUM DENSE 3
G.W.T. LOOSE TO G.W.T.
SAND
MEDIUM DENSE 10
SAND 4.1
5.1
GRAVELLY SAND
13
(5mm gravel)
GRAVELLY SAND
6 8.3
(5mm gravel)
HIGH PLASTICITY
68
9.5 CLAY
10.2
17
DENSE SAND 12 11.4
13.4
GRAVELLY SAND (8mm
7
gravel) DENSE SAND 17
14.9
16.5
HIGH PLASTICITY
77 GRAVELLY SAND
CLAY 21
(8mm gravel)
18.3
19.9
DENSE SAND 22
20.5
TOE-LEVEL TOE-LEVEL
HIGH PLASTICITY
93
CLAY
DENSE SAND 13
23.5
LOW PLASTICITY
71
CLAY
25.0 25.0
Fig. 14. Soil profiles for piles C6N3 and C9N3.

2.4 Toe stiffener
High-strain dynamic tests were performed on two different Toe Box piles at another bridge project:
The piles, Piles P1N2 and P1N3, were 1,200mm diameter, 25 m long, cased, drilled shafts
constructed into a dense sand layer. For Pile P1N2, the construction procedure went smoothly and
the installation of the Toe-Box was flawless. Pile P1N3 was tested without the Toe-Box being
expanded.
Figure 15a shows the grouting diagram and its correspondent load-movement diagram for Pile
P1N2. The pressure behavior during the grouting corresponds to a dense sandy soil, which was
confirmed during the perforation of the pile. Figure 15b shows the load movement diagram
obtained with the grouting information. The Toe Response obtained with the CAPWAP analysis
was superimposed in order to compare the toe stiffness measured with Toe-Box to that determined
in the CAPWAP analysis. The slope of the CAPWAP-determined static toe response is
commensurable with static response from a reloading of the pile.
Fig. 15a. Grout pressure vs. grout volume Pile P1N2.
Fig. 15b. Load-Movement diagrams for Pile P1N2.

The Toe Box system is specifically interesting for large diameter piles (600 mm and larger)
drilled with slurry support or with temporary casing. These drilling methods often leave loosened
soil and/or accumulation of debris at the pile toe, resulting in a soft pile toe response. Furthermore,
the benefits in the toe-stiffness and pile mobilization are increased together with the diameter, i.e.,
wider pile and toe-box diameters mean higher achieved force with the same grouting pressures,
same pump.
3 CONCLUDING REMARKS
Both of expansion devices, the Expander Body and the Toe Box, are presented in the paper as
stiffness improvement devices and as quality control devices, giving the possibility to verify the
desired pile toe response. Should the pile response show to be less than expected, the verification
gives the possibility to reinforce the foundation if necessary, while the project is still in its
construction phase, avoiding any possible further problem.
Furthermore, both the Expander Base and the Toe Box can be installed with a wide variety of
drilling systems. The main decision whether to use one or the other rests with the pile diameter.
While the EB is recommended to be installed in a pile of up to 600 mm, because its maximum
practical expansion diameter is 800 mm, the Toe Box starts to be practical with pile diameters
ranging from 600 mm and above.
Loading tests have been performed in piles with both of the systems, demonstrating that, when
the devices are used, the supported working loads of the piles can be considerably increased for
the same level of limiting deformation.
4 REFERNCES
Berggren, B., Sellgren, E. and Wetterling, S. 1988. Expanderkroppar. Anvisningar för

dimensionering, utförande och kontroll (Expander Body. Instructions for design, installation
and control). Swedish Commission on Pile Research, Report 79, 54 p.
Bolognesi, A.J.L. and Moretto O. 1973. Stage grouting preloading of large piles in sand,
Proceedings of the Eighth International Conference, ISSMFE, Moscow, August 12-19, Vol.
2.1, pp. 19-25.
Bruce, D.A., May 1986. Enhancing the performance of large diameter piles by grouting. Ground
Engineering, 19(4) 9-15.
Fellenius, B.H. and Terceros, Herrera. M. 2014. Response to Load for four different types of piles.
Proceedings, International Conference on Piling and Deep Foundations, Stockholm, May
21-23, pp. 99-120.
Fellenius, B.H., 2017. Basics of foundation design. Electronic Edition. www.Fellenius.net. 464p.
Massarsch, K.R. and Wetterling, S., 1993. Improvement of augercast pile performance by
Expander Body system. 2nd International Seminar on Deep Foundations on Bored and
Auger Piles, Ghent, June 1 - 4, pp. 417-428.
Massarsch, K. R. 1994. Execution, supervision and quality control of anchors. Panel Discussion,
Section 3.3, Construction, Instrumentation and Real Time Management, XIII. Proceedings,
International Conference on Soil Mechanics and Foundation Engineering, New Delhi,
January 5-10, Vol. 5, pp. 317 - 319.
Rosales, O., 2015. Efecto de procesos constructivos de gran desplazamiento en el comportamiento
de pilotes perforados en arenas. Proceedings, 15th Panamerican Conference on Soil

Mechanics and Geotechnical Engineering. Buenos Aires, November 15-18, vol. 2, pp. 1544-
1551.
Terceros Herrera M., And Massarsch K. R. 2014. Cast in situ piles with expander body in sandy
soils – Superpile - Cambridge, Ma. June 18-20. Presentation.
Terceros Arce, M., and Terceros H., M. The use of the Expander Body with Full Displacement
Piles In Medium Dense To Dense Sandy Soils. 15th Panamerican Conference on Soil
Mechanics and Geotechnical Engineering. Buenos Aires, November 15-18, vol. 2, pp. 1702
- 1712.

Publishing Policy and Terms of Use
1 General
It is the intention that readers have access to the widest possible range of content and
that this content can be shared to enable the most effective research, study, teaching,
and practice of geotechnical and geo-environmental engineering and associated fields of
application. Efficient dissemination of published information is in the interest of the
geo-engineering profession.
The three main players have – sometimes – conflicting interest in the publication and
information dissemination process. The primary interests are briefly outlined below:
Author • Retain full copyright to published material

• Publication by reputed publisher, giving exclusive publishing
right
• Review of manuscript prior to publication
• Publication in print and in electronic format
• Permission to post his/her manuscript on own web site
• Partial re-use of his/her published information in subsequent
publications
• Ability for others to use excerpts of published information
(figures, tables, text excerpts) at no cost and without
administrative burden
Publisher • Obtain complete and exclusive copyright to publication

• Limit cost for producing publications
• Restrict free access to information (commercialization)
• Sale of publications (books, journals etc.) in printed and -
increasingly also - electronic format
• Avoid duplicate publication
Geo-profession • High quality of publication assured by review process

• Facilitate dissemination of information
• Limit cost for producing publications
• Limit cost for acquiring publications
• Ability to reproduce published material
• Avoid duplicate publication
The author(s) of a manuscript is usually not aware of the rights and limitations when
submitting manuscripts to publishers.
This document outlines the Publishing Policy endorsed by ISSMGE and intended to
provide a compromise for producing future publications.

2 Rights of Author
The rights of Author(s) are intended to help in their academic and professional work
and are subject to the following conditions:
a. these rights may not be used for commercial purposes;
b. these rights should not be used in a way that involves duplicate publication
that will compete with the Publisher’s own publications; and
c. each use must contain an acknowledgement to the journal/book as the

original source of publication in the form: “[Journal/Book]
[Volume/Publication Year] [URL of the item as published by Publisher]
[Year].”
Ownership of the copyright contained in the Manuscript (“the Material”) remains with
the Author(s) who retain the following rights:
1. to post copy of their submitted Manuscript (Pre-print) on their own Web

site, an institutional repository, or their funding body's designated archive;
2. to post copy of their accepted Manuscript (Post-print) on their own Web

site, an institutional repository, or their funding body's designated archive.
Authors who archive or self-archive accepted Publications must provide a
hyperlink from the manuscript to the Journal's Web site;
3. Authors and any academic institution where they work at the time may
reproduce their Manuscript for the purpose of course teaching;
4. to reuse – as stated in paragraph 3 - part of the Manuscript in other works

created by the Author(s), provided the original publication is acknowledged
through a note or citation in a format acceptable to the Publisher;
5. to prepare derivative works from the publication.
Permission must be sought for uses other than those defined above. Other than the use
outlined above no publication may be captured or downloaded electronically into any
format without the Publisher's and Author’s specific written permission.

3 Publishing Conditions
The Author(s) retain copyright but assigns the Publisher the exclusive right to publish
the Material.
The substantial investment made by the Publisher in protecting and enhancing the
quality of the content is recognised. It is necessary, therefore, to impose the following
conditions on the use of published Material.
3.1 General Rights

Corresponding Author and contributing Authors grant full and exclusive publishing
rights of their Material to the Publisher. In assigning exclusive publishing rights, the
Author is not forfeiting the rights to use the contribution as stated in paragraph 2.
3.2 Patent and Trademark Rights

The right to be identified as the Author(s) (i.e. by asserting Authors’ rights under
copyright law).
3.3 Re-use of Material by Author(s)

In assigning publication permission, the right of the Author to use parts of his or her
Material elsewhere are retained. Figures, tables, and other limited extracts may be used
without permission as long as they represent less than 20 % of the new work. However,
republishing the whole article, or a substantial part of it, requires the permission of the
Publisher.
The Publisher shall maintain a firm policy against duplicate publication unless there are
exceptional circumstances.
3.4 Re-use of Material by Others

The re-use of three figures, tables or text extracts less than 100 words by third parties is
free of charge and permitted without the Author’s or Publishers permission. In all case
proper credit must be clearly given as stated above.
All requests to reproduce or re-use the whole or a substantial part of the Material in
another publication will be conditional upon the Authors’ consent and subject to the
Publisher’s approval.
The party seeking permission will be instructed to write to the Publisher and to obtain
the Authors’ consent.

COVER 4 COVER 1
3 BOLIVIAN
rd
3rd BOLIVIAN INTERNATIONAL CONFERENCE ON DEEP FOUNDATIONS

INTERNATIONAL
CONFERENCE ON
DEEP FOUNDATIONS
April 27 – 29, 2017

PROCEEDINGS
VOLUME 1
VOLUME 1
64427 ICDF Vol 1 Conference Cover_rb.indd 1 4/4/2017 10:06:35 AM

ICDF Vol 1 Conference Program - Complete Book PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ICDF Vol 1 Conference Program - Complete Book PDF

Uploaded by

Copyright:

Available Formats

COVER 4 COVER 1

3rd BOLIVIAN INTERNATIONAL CONFERENCE ON DEEP FOUNDATIONS

April 27 – 29, 2017

64427 ICDF Vol 1 Conference Cover_rb.indd 1 4/4/2017 10:06:35 AM

April 27 – 29, 2017

Conference Advisory Committee:

Chairman of B.E.S.T. Prediction Event:

Printed by Omnipress, Madison, WI USA

Conference Program ............................................................................................... vii

Electrical Resistance Strain Gages in Instrumentation of Deep Foundations ............................................... 1

Advances in the Use of Drilled Foundations .............................................................................................. 33

Conventional methods and recent developments in retaining walls ........................................................... 75

Recent Developments in Helical Piles ........................................................................................................ 89

Recent developments in vibratory driving and soil compaction ............................................................... 125

Recent Developments and Applications in Geotechnical Field Investigations for

Dynamic Loading Tests: A State of the Art of Prevention and Detection of

3rd Bolivian International Conference on Deep Foundations—Volume 1 iii

Publishing Policy ...................................................................................................229

iv 3rd Bolivian International Conference on Deep Foundations—Volume 1

3rd Bolivian International Conference on Deep Foundations—Volume 1 v

8:30-8:45 OPENING SESSION

9:45 -10:15 COFFEE BREAK

10:15-11:15 SESSION 2 – “Bengt H. Fellenius Lecture”

11:15-12:15 Presentation of Invited Papers 1 and 2

Tomás Murillo (Spain): Diseño y ejecución de pilotes con celda de

12:15-12:45 Opening and overview of Exhibition

12:45-14:00 LUNCH BREAK AND DIGITAL POSTER SESSION

14:00-15:00 Presentation of Invited Papers 3 and 4

3rd Bolivian International Conference on Deep Foundations—Volume 1 vii

17:30-18:30 SESSION 5 – “K. Rainer Massarsch Lecture”

April 28th 2017

9:30-10:30 COFFEE BREAK

11:30-12:30 Presentation of Invited Papers 5 and 6

12:30-14:30 LUNCH BREAK AND DIGITAL POSTER SESSION

16:30-17:00 COFFEE BREAK

viii 3rd Bolivian International Conference on Deep Foundations—Volume 1

20:00 Gala Dinner

April 29th 2017

8:30-9:00 B.E.S.T. Presentation. Alesandro Mandolini (Italy), Chairman of TC212

10:15-10:35 COFFEE BREAK

10:35-12:05 SESSION 2: Pile Installation and Instrumentation

12:05-14:00 LUNCH BREAK AND DIGITAL POSTER SESSION

14:00-16:00 SESSION 3: Experimental Results and Predictions

16:00-17:00 Mario Terceros Banzer Lecture

17:00-17:20 CLOSING SESSION

3rd Bolivian International Conference on Deep Foundations—Volume 1 ix

2. ELECTRICAL RESISTANCE STRAIN GAGES

3rd Bolivian International Conference on Deep Foundations—Volume 1 1

∆𝐿𝐿 = 𝐿𝐿0 𝛼𝛼∆𝑡𝑡 (1)

2 3rd Bolivian International Conference on Deep Foundations—Volume 1

where: 𝜀𝜀 = total strain

The variance of resistance caused by a change of strain is determined by a Wheatstone bridge.

Fig. 1. Wheatstone bridge.

where: ε1 = strain due gage one

3rd Bolivian International Conference on Deep Foundations—Volume 1 3

2.1 Types of circuits

A) Half-bridge circuit with one active branch (Eq. 5):

Fig. 2. ¼ bridge with one active branch.

4 3rd Bolivian International Conference on Deep Foundations—Volume 1

B) Half-bridge circuit with two active branches (Eq. 6).

Fig. 3. ½ Bridge with two active branches (adjacent arms).

3rd Bolivian International Conference on Deep Foundations—Volume 1 5

where: ν = Poisson ratio