PR86 Shaft Friction of CFA Piles in Chalk PDF

Project Report 86 London 2003
Shaft friction of CFA

piles in chalk
Prepared under contract to CIRIA by:
J A Lord of Ove Arup and Partners
T Hayward of Stent Foundations Limited (formerly of the University of Southampton)

© Copyright CIRIA 2017. No unauthorised copying or distribution permitted. For use by Central Piling only.
C R I Clayton of the University of Southampton
6 Storey's Gate, Westminster, London SW1P 3AU

TELEPHONE 020 7222 8891 FAX 020 7222 1708
EMAIL enquiries@ciria.org
WEBSITE www.ciria.org
Summary
This report contains the graphical plots and analyses of pile test data that were obtained
for a CIRIA research project to review some design aspects of continuous flight
augered piles in chalk.
The conclusions drawn from the analyses were that:

! the factor most influencing the shaft resistance of CFA piles in chalk is how readily
the chalk remoulds
! low penetration resistance appears to be a reasonable indicator of this
! low density chalk is much more prone to create remoulded material than
medium/high density chalk
! unless it can be demonstrated that medium density chalk is present over all the pile
shaft, the effective stress pile friction coefficient, β, should be taken as 0.45
throughout in deriving the shaft resistance of CFA piles in chalk
! if it can be demonstrated that medium density chalk is present over all the pile shaft,
categorised by N > 10 or qc > 4 MN/m², β, should be taken as 0.8 throughout in
deriving the shaft resistance of CFA piles in chalk.
Shaft friction of cfa piles in chalk

Construction Industry Research and Information Association
Project Report 86 © CIRIA 2003 RP 628 ISBN 0 86017 886 2
Keywords
Ground engineering, ground investigation and characterisation, piling, soil-structure
interaction
Reader interest Classification

Design, specification, construction, AVAILABILITY Unrestricted
managers, clients and supervising
engineers involved in civil and CONTENT Subject area review
geotechnical works STATUS Committee guided
USER Civil, geotechnical and

structural engineers,
engineering geologists
Published by CIRIA, 6 Storey’s Gate, Westminster London, SW1P 3AU. All rights reserved. No
part of this publication may be reproduced or transmitted in any form or by any means, including
photocopying and recording, without the written permission of the copyright holder, application
for which should be addressed to the publisher. Such written permission must also be obtained
before any part of this publication is stored in a retrieval system of any nature.
This publication is designed to provide accurate and authoritative information in regard to the
subject matter covered. It is sold and/or distributed with the understanding that neither the
author(s) nor the publisher is thereby engaged in rendering a specific legal or any other
professional service. While every effort has been made to ensure the accuracy and completeness
of the publication, no warranty or fitness is provided or implied, and the author(s) and publisher
shall have neither liability nor responsibility to any person or entity with respect to any loss or
damage arising from its use.
2 CIRIA PR86
Acknowledgements
Research This report is the outcome of a research project 628. The report was prepared by Ove
contractor Arup and Partners, in conjunction with Stent Foundations Limited and the
University of Southampton.
Authors Dr J A Lord MA (Cantab) PhD CEng FICE FConsE AFPWI of Ove Arup and Partners
Dr T Hayward BEng MSc DIC PhD of Stent Foundations Limited (formerly of the
Universtiy of Southampton)
Professor C R I Clayton BSc (Hons) MSc DIC PhD CEng CGeol FGS of the University
of Southampton
Steering group Following CIRIA’s usual practice, the research project was guided by a Steering Group
which comprised:
Chair Professor B G Clarke University of Newcastle upon Tyne
Members Mr I C Blight Rail Link Engineering/Halcrow Group Limited
Mr N Brooks Stent Foundations Limited
Dr F Chow Geotechnical Consulting Group
(ICE representative)
Mr D Corke Bachy Soletanche Limited
Mr D Patterson Highways Agency
Mr T Spink Mott MacDonald Limited
Mr S Wade Stent Foundations Limited (formerly of Westpile Limited)
Dr J Wilson WS Atkins Consultants Limited
(DTI representative)
CIRIA manager CIRIA’s research manager for the project was Dr A J Pitchford.
Project funders The project was funded by:

Construction Directorate (Department of Trade and Industry),
Research and Development Enabling Fund of the Institution of Civil Engineers
CIRIA’s Core Programme.
Member firms of the Federation of Piling Specialists provided in-kind support through
the provision of pile test data.
Contributors CIRIA and the authors are grateful for the help given to this project by the funders, the
members of the Steering Group, and by the many individuals who were consulted and
provided data. In particular, acknowledgment is given to the following:
! the late Ken Fleming for freely making available not only the results but also his
analysis of numerous tests on CFA piles in chalk
! Dave Illingworth of Westpile Limited for valuable discussions on his assessment of
the extensive CFA test pile programme at the Oracle site, Reading
! Fin Jardine, latterly of CIRIA, for helping to set up this research project and for his
invaluable comments.
CIRIA PR86 3
Executive summary
This report makes recommendations for the design of shaft resistance of CFA piles in
low- and medium- density chalk. It is not appropriate to the design of piles constructed
by any other method.
CIRIA Project Report 11 (PR 11) Foundations in chalk (Lord et al, 1994) concluded
that the average ultimate shaft resistance (τsf) of CFA piles in low- and medium- density
chalk could be expressed in terms of the empirical relationship:
τ sf = βσ v′
where σ′v is the average vertical effective stress along the pile shaft in chalk.
On the basis of relatively few suitable pile tests, it was recommended that β should be
taken as 0.45 as a lower limit for safe design. The UK piling industry subsequently
expressed concern that this value of β is conservative, resulting in unnecessarily long or
large diameter piles. CIRIA and member firms of the Federation of Piling Specialists
(FPS) collaborated in instituting this research project, in order to:
! obtain and analyse new CFA pile test results in chalk
! re-examine the pile test data available at the time PR11 was written
! review whether the published value of β is conservative
! explore the factors that affect it
! refine design recommendations.
Where low values of β were found, the chalk tended to be either structureless, or of low
density, or of Standard Penetration Test (SPT) N value ≤ 10. It is inferred that these are
indicators of chalk that are readily or substantially remoulded by the CFA piling process.
There is also some slight evidence that the shaft resistance of CFA piles increases with
the time elapsed between pile installation and testing, but as the magnitude of the increase
cannot be reliably predicted from the present extremely limited data, it would be very
dangerous to extrapolate higher β values for working piles than proven from site specific
test piles.
As the SPT has been found to be operator- and drilling-method dependent, does not
provide a continuous profile in chalk, is adversely influenced by the presence of flints,
and often exhibits considerable scatter, the use of an ‘averaged’ SPT blowcount
multiplied by a correlation factor to determine the shaft resistance of a CFA pile in
chalk is to be strongly discouraged. But in the absence of direct measurements, the SPT
can be used as an indicator of chalk with a low unit shaft resistance when N ≤ 10. The
static Cone Penetration Test (CPT) has the advantage over the SPT in providing a
continuous resistance profile in chalk and greater speed of operation. It has been found
that low penetration resistance chalk would be expected to have a cone resistance of
between about 2 and 4 MN/m².
It is recommended that in deriving the shaft resistance of CFA piles in chalk, the
following should be adopted in the absence of preliminary pile tests.
! For low penetration resistance chalk over part of the pile shaft, categorised by
N ≤ 10 or qc ≤ 4 MN/m², β = 0.45 throughout.
4 CIRIA PR86
! For higher penetration resistance chalk over all the pile shaft, categorised by
N >10 (only if flints are absent) or qc > 4 MN/m², β = 0.8 throughout. But if the
ground investigation data are of poor quality (eg insufficient boreholes, inadequate
penetration into chalk, little or no in-situ or laboratory testing, inadequate soil
description), β should be taken as 0.45 for safe design. If flints are present the SPT
N value may not be relied upon to indicate higher penetration resistance chalk.
The allowable capacity of a CFA pile, Pa, should be the lesser of:
Psu P  P P 
either, Pa = + bu  P a = su + bu if settlement to be less than 10 mm 
1.0 3.5  1.5 3.5 
Psu + P bu
or, Pa =
2 .5
where Psu and Pbu are the calculated ultimate shaft and base loads respectively.
These factors of safety are related to the CIRIA PR11 design method. For longer piles
in medium/high-density chalk, the overall capacity of the pile may be dictated by
concrete strength and not by the chalk.
Significant economies, through pile shortening, can be achieved by proving higher β

values in preliminary pile tests. There is also a greater potential for cost savings by
undertaking one preliminary pile test to 3 x WL (working load) instead of two working
pile tests to 1.5 x WL. Clients and contractors should wherever feasible seek to
undertake preliminary pile test(s) to enable a reduction in the number of working pile
tests during the contract.
As the presence of dissolution features cannot be readily identified in the construction

of CFA piles, it is recommended that a CPT is undertaken at each pile location at sites
identified during desk studies to be prone to dissolution. Techniques for identifying
dissolution features are discussed in Lord et al (2002).
CIRIA PR86 5
6
CIRIA PR86
Contents
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Executive summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
List of figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
List of tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2 Shaft resistance of bored and cfa piles . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2 Analysis of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 New cfa pile test data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Pile-chalk interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 Locality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.4 Pile length, diameter and overburden thickness . . . . . . . . . . . . . . . . . . . . 24
2.5 Time elapsed between installation and testing . . . . . . . . . . . . . . . . . . . . . 26
2.6 Level of water table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.7 Nature of the chalk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.8 Construction techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.9 Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.10 Examination of cfa pile in chalk in situ and future research . . . . . . . . . . 34
3 Identification of chalk with a low unit shaft resistance . . . . . . . . . . . . . . . . 36

3.1 Determination of dry density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2 Standard penetration test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.3 Cone penetration test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3.1 Use of CPT to predict base resistance . . . . . . . . . . . . . . . . . . . . . 37
4 Pile testing and factors of safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Appendix 1 Examination of exposed shaft of CFA piles . . . . . . . . . . . . . . . . . . . . . . 47

Appendix 2 Site Investigation methods appropriate for CFA piling in chalk . . . . . . . 57
Appendix 3 CFA piles in chalk pile load test information check list . . . . . . . . . . . . . 61
CIRIA PR86 7
LIST OF FIGURES
Figure 2.1 Comparison of total shaft load for CFA piles analysed using
CEMSOLVE and CIRIA PR11 (1994) . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 2.2 Average ultimate shaft resistance in chalk v. average vertical effective
stress on shaft in chalk for CFA piles . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 2.3 Average ultimate shaft resistance/average vertical effective stress for
CFA piles v. overburden thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
CFA piles v. pile length in chalk./pile diameter . . . . . . . . . . . . . . . . . . 25
Figure 2.5a Average ultimate shaft resistance/average vertical effective stress for
CFA piles v. pile length in chalk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Figure 2.5b Average ultimate shaft resistance/average vertical effective stress for
CFA and bored piles v. pile length in chalk . . . . . . . . . . . . . . . . . . . . . 26
CFA piles v. time between pile installation and testing . . . . . . . . . . . . 27
stress on shaft in chalk for CFA piles showing impact of water table
level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
stress on shaft in chalk for CFA piles showing impact of low
penetration resistance in chalk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
stress on shaft in chalk for bored piles showing impact of chalk
density. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Figure 2.10 Logic flow diagram for determining shaft resistance of CFA piles in
chalk (after Spink, 2002b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Figure 2.11a Allowable loads for 600 mm diameter piles in low penetration
resistance chalk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33
Figure 2.11b Allowable loads for 600 mm diameter piles in medium to higher
penetration resistance chalk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Figure A1.1 Pile S6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Figure A1.2 Pile S5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Figure A1.3 Pile S4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Figure A1.4 Pile S3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Figure A1.5 Pile S1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Figure A1.6 Pile N1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure A1.7 Pile N2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Figure A1.8 Details of exhumed CFA pile on A41 Aston Clinton bypass . . . . . . . . 56
LIST OF TABLES
Table 1.1 Records of CFA piles in CIRIA Report PR11 (1994) . . . . . . . . . . . . . . 15

Table 2.1 Pile load test results supplied by the Federation of Piling Specialists
(FPS) and others . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Table 2.2 Records of CFA piles since CIRIA Report PR11 (1994) . . . . . . . . . . . 20
Table 2.3 Comparison of Total Shaft Load, Base Modulus and Base Load for
CFA piles analysed using both CEMSOLVE and CIRIA PR11 (1994)
methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
8 CIRIA PR86
NOTATION
D diameter of pile
Db diameter of pile base
Ds diameter of pile shaft
d depth below top of chalk in metres
Eb modulus of chalk at base of pile
Ec modulus of elasticity of concrete pile shaft
Ey ‘yield’ modulus of chalk
Fb factor of safety on base of pile
Fs factor of safety on shaft of pile
L length of pile
N SPT blowcount
P load on pile head
Pa allowable capacity of pile
Pb load on pile base
Pbu ultimate base load
Ps total shaft load
Psu ultimate shaft load
qb stress on base of pile
qc cone resistance in MN/m²
qu ultimate bearing capacity of chalk

qy ‘yield’ stress of chalk
β effective stress pile friction coefficient
γ unit weight of soil
γconc unit weight of concrete
γd intact dry density of chalk
ρ settlement of pile
ρb settlement of base of pile
ρmax maximum settlement of pile
ρs elastic shortening of pile shaft
σ′v average vertical effective stress along pile shaft in chalk
τsf ultimate shaft resistance of pile
τsf average ultimate shaft resistance of pile in chalk
CIRIA PR86 9
10
CIRIA PR86
1 Introduction
1.1 BACKGROUND
The satisfactory performance of piles in chalk was demonstrated at the Chalk

Symposium in 1965 and at the Behaviour of piles conference in 1970. However it was
not until the Géotechnique Symposium in Print in 1976 entitled Piles in Weak Rock
that, as a result of the published case histories, a clearer picture emerged of the ability
of different types of pile to sustain loads in chalk. These were brought together by
Hobbs and Healy in 1979 in the CIRIA Report PG6 Piling in Chalk.
The design of foundations in chalk was reviewed by Lord (1990) in his Keynote
Address at the International Chalk Symposium at Brighton in 1989. This Chalk
Symposium prompted a complete re-evaluation of chalk as a whole in a CIRIA research
project (RP458). The first stage of the research project was a project report (PR11) by
Lord, Twine and Yeow (1994) Foundations in Chalk. This report progressed the 1989
Keynote Address on foundations in chalk, as a result of a much extended data-base of
plate loading tests, pile tests and case histories; it dealt principally with shallow
foundations and piled foundations, and in that respect superseded CIRIA report PG6.
The findings of the CIRIA research project 458, referred to above, were published in
Engineering in Chalk by Lord, Clayton and Mortimore (2002), which incorporates the
key elements of Foundations in chalk as follows:
Chapter 3 Description and classification of chalk
Chapter 7 Shallow foundations
Chapter 8 Piled foundations
Chapters 3 and 8 are relevant to this present report.
Ward et al (1968) in seeking to relate the mass deformation properties of the chalk at
Mundford to its structure, in particular its hardness and jointing, had devised a grading
scale based on visual inspection. Despite emphasising that these grades were
specifically for the conditions and requirements of the particular site, they have been
widely and indiscriminately applied since, for the classification of chalk sites. CIRIA
project report 11 (Lord et al, 1994), recognising the need for a revised approach to the
classification of chalk in terms of its engineering properties, identified the following
factors most likely to influence the behaviour of the chalk mass as being:
! hardness of the intact chalk, which can be expressed in terms of its dry density
! discontinuity aperture
! bedding/discontinuity spacing and pattern
A new classification for chalk was devised based upon:
For structured chalk:

! intact dry density (see below)
! typical discontinuity aperture referred to as the
! typical discontinuity spacing CIRIA Grading Scheme
CIRIA PR86 11
For structureless chalk:
! engineering behaviour.
The most up-to-date details are given in Lord et al (2002), but for the purposes of this
report, the intact dry density scales and discontinuity aperture classification may be
summarised as:
Density scale Intact dry density (Mg/m³)
Low < 1.55
Medium 1.55 – 1.70
High 1.70 – 1.95
Very high > 1.95
Grade Discontinuity aperture

A closed
B < 3mm
C > 3mm
D structureless or remoulded mélange
For Grades A, B and C, the discontinuity spacing is denoted by a numerical suffix.
CIRIA report P11 (Lord et al, 1994), also analysed pile case histories. For these
purposes, the following categories were identified:
! bored piles using shells, augers or buckets
! continuous flight auger piles (CFA)
! driven cast-in-place piles
! driven preformed piles
! driven cast-in-place piles with expanded base.
It was found that the shaft resistance of each of the different pile categories was
dependent upon the pile type.
1.2 SHAFT RESISTANCE OF BORED AND CFA PILES
A conclusion of CIRIA PR 11 Foundations in Chalk (Lord et al, 1994), which largely

stemmed from Twine’s (1990) earlier work, was that the empirical relationship:
τsf = 0.8σ ′v (subsequently expressed as τ = β σ ′v )

sf
should be adopted for estimating the shaft resistance of bored piles in low and medium
density chalk, where:
τsf = ultimate average shaft resistance for the length of the pile in chalk
σ ′v = average vertical effective stress along pile shaft in chalk.
The contribution of made ground and fill should be omitted in determining σ ′v unless the
pile capacity was proved otherwise by a pile test.
It had been expected that the same empirical relationship would apply to CFA piles in
low and medium density chalk, but the results indicated a lower-bound relationship of:
τsf = 0.45σ ′v
12 CIRIA PR86
for the nine or so piles in which an ultimate shaft resistance could be derived, ie those
CFA pile case histories where there had been sufficient pile-head settlement to permit
analyses (see Table 1.1 based on Appendix C2 of Lord et al, 1994). Many other case
histories had to be ignored because so little pile-head movement occurred under loads
to no more than 1.5 x Design Load that it was not possible to determine whether the
load was carried on the base of the pile or on the shaft. None of the CFA piles had been
instrumented.
Subsequently Fleming (1997), who had served on the CIRIA PR11 Steering Group and
was well aware of the paucity of CFA pile test data, supplied the results of 14 additional
CFA piles in chalk which had been analysed by CEMSOLVE. As four of these results
indicated β values of about 0.45, with the remainder generally in the range 0.6 to 1.0, the
lower-bound relationship recommended by Lord et al (1994) appeared to be justified.
The Federation of Piling Specialists (FPS), concerned that the values of β quoted in PR11
for CFA piles were conservative and were resulting in unnecessarily long/large diameter
piles, published a position paper in April 2000 on the Design of CFA piles in chalk
(FPS, 2000). In it they compared the allowable loads for CFA piles in chalk based on CIRIA
Report PG6 (Hobbs and Healey, 1979) with CIRIA Report PR11 (Lord et al, 1994).
The FPS concluded that in “structureless and structured blocky chalk” (identified by a
low SPT blowcount profile), the PR11 approach should be adopted, whereas for
“structured solid chalk” (identified by a high SPT blowcount profile), the PG6 design
approach should be adopted. The paper also appealed for additional CFA pile test data
and encouraged FPS members to extract additional information from pile tests by
enhancing loads where suitable opportunities arose.
In the meantime, Illingworth and Chantler (1999) had published the results of a
programme on 11 CFA piles in chalk at the Oracle site, Reading, Berkshire. In five of
these tests of the pile head movement was large enough for back analysis to be
undertaken, from which they found that:
τsf = 0.45σ ′v (1 test)

τsf = 0.8σ ′v (2 tests)
τsf = 2.7 σ′v (2 tests)
The authors ascribed this variation to the degree of weathering of the chalk, the higher
values being in chalk that was “relatively unweathered”. Illingworth and Chantler
report that static cone penetration testing was better able to reveal “the extent of
weathered chalk” than with SPTs.
The conclusions reached in PR11 were based on analysis of a fairly small number of
suitable pile tests then available for the project team to assess. In the years since PR11
was published, and during which the use of continuous flight augering as a technique
for installing piles in chalk has become more widespread, additional pile test results
have become available.
In order to decide whether published values of β are conservative and to consider what
reasons would justify the use of different values, this collaborative research project was
initiated by CIRIA and the FPS. The project was to re-examine the pile test data
available at the time that PR11 was written, and to analyse and review data made
available by FPS members and others, which had been obtained from tests on CFA piles
in chalk since 1994.
CIRIA PR86 13
The results of these analyses are presented in this report (in graphical and tabular form).
Current thinking on appropriate β values and other design criteria that should be applied
when the shaft friction of CFA piles in chalk is being assessed has been summarised
and included in the Executive Summary.
14 CIRIA PR86
Table 1.1 Records of CFA piles in CIRIA Project Report 11
CIRIA PR86
Case History Pile Source of test Pile dimensions Max. load Max. Total Over- Av.vert. Av. shaft Base load Base Base Base
No. data Pmax settlement shaft burden effective friction in Pb stress modulus SPT
Shaft Length (m) kN ∆max load Ps friction stress in chalk τsf kN qb Eb MN/m² N
dia. Total/Sleeved(s) mm kN load chalk σ ′v kN/m² kN/m²
m O’burden(o)/Chalk kN kN/m²
C1 Dartford CFA1 Cementation 0.75 24.70|12.30o|12.40 7500 M 26 3384 640 185 94* 4116 9316 278 50
CFA3 Cementation 0.75 26.60|12.00o|14.60 7500 M 21.2 2720 612 192 61* 4780 10820 518 50
C2 Caversham Rd TP1 Stent 0.6 14.75|6.50o|8.25 1650 M 17.83 592 296 97! 19* 1058 3741 74 15-27
Reading
C3 Bridge St, PPT Stent 0.6 14.75|7.40o|7.35 2250 M 20.8 1856u 583 99! 92 U u 394 1394 23 45
Reading 831l 583 99! 18U l 1419 5017 155 45
C4 Duke St, 26 Stent 0.45 12.00|5.20o|6.80 1125 M 12.94 263 100 109 17* 862 5421 120 29-41
Reading
C5 Luton TP1 Stent 0.45 12.00|3.10o|8.90 2000 M 15.95 1176 106 129! 85U 824 5179 97 31
TP2 Stent 0.45 12.00|3.10o|8.90 2500 M 21.89 1320 106 129! 97U 1180 7418 101 31
TP3 Stent 0.45 12.00|3.10o|8.90 2250 M 16.4 1463 106 129! 108U 787 4949 91 31
C6 Gravesend, 42 May Gurney 0.45 12.00|5.30o|6.70 1050 M 11.94 720 190 31! 56U 330 2076 42 23
Kent
C7 Canterbury PTP1 May Gurney 0.6 18.80|5.40o|13.40 3740 M 10.48 1652 352 167 51* 2088 7384 634 25
(1) PTP2 May Gurney 0.6 17.50|4.00o|13.50 2966 M 35.68 1399 217 104! 46U 1567 5541 54 17-25
PTP3 May Gurney 0.6 21.00|4.00o|17.00 4000 M 15.64 1929 217 120! 53* 2071 7323 297 26-31
C8 Thorpe Park PTP1 May Gurney 0.4 16.00|2.50o|13.50 1900 M 12.09 1109l 46 112! 63 U l 791 6294 241 19
1377u 46 112! 78 U u 523 4163 131 19
PTP2 May Gurney 0.4 9.00|2.50o|6.50 962.5 M 11.34 639 46 79! 73 U 323 2571 48 1
C9 Maidenhead TP1 Keller 0.4 9.50|5.40o|4.10 1700 M 32 696 191 82! 98 U 1004 7988 66 25
C10 Durrington TP1 Keller 0.4 12.00|4.85o|7.15 3000 M 15.2 1648 126 33! 169 1352 10759 280 50
F = Failure value U = Ultimate value

! = made ground/recent fill discounted u = upper bound estimate
M = Maximum value l = lower bound estimate
15
(1) = pile about 5m away from borehole * = interpretation of data less reliable, often on account
which encountered solution feature of limited pile-head movement
16
CIRIA PR86
2 Analysis of results
2.1 NEW CFA PILE TEST DATA
The appeal to the FPS membership for additional CFA pile test data has yielded about
60 pile tests. These are listed in Table 2.1, which includes those provided by Fleming
(1997) and Illingworth and Chantler (1999). In about 30 of these tests the pile head
settled sufficiently under load to permit back-analysis – these results are presented in
Table 2.2 in a similar manner to those in Table 1.1. The pile tests have generally been
analysed to derive the shaft and base components using the method outlined in
Appendix D of Lord et al (1994), in which the diameter of the pile shaft has been taken
as the nominal diameter of the auger. This contrasts with Illingworth and Chantler
(1999) who adopted an effective pile diameter based on measurements taken to assess
the volume of concrete pumped. It may be possible with careful calibration and records
to adduce/determine the volume of the pile in the ground and hence calculate an
average diameter. It may even be possible to develop a volume/depth profile. Neither,
however, fully represents either the pile shape or the changing characteristics of the
surrounding (disturbed) chalk or superficial ground. (Exhumation of a section of a shaft
of a CFA pile, described in Appendix 1, showed that the surface of the shaft undulated
by at least 20 mm.) For this reason therefore, it is more appropriate in analysis and
design calculations to use the auger diameter as the nominal diameter of the pile.
Illingworth and Chantler’s (1999) test results have been adjusted accordingly.
A comparison is made in Table 2.3 between results obtained from the method of
analysis given in Appendix D of Lord et al (1994) with those obtained using
CEMSOLVE. (CEMSOLVE is a method of single pile settlement prediction and
analysis developed by Fleming (1992). A full description of the back-analysis method is
given by England (1999). This is illustrated in Figure 2.1 for total shaft load. It may be
seen that agreement in terms of the total shaft load is very good, the results differing by
some 10 to 15 per cent with the CIRIA PR11 results being generally slightly higher.
This arises because CEMSOLVE tends to predict a modulus, Eb, for the chalk at the
base of the pile that is slightly greater than that obtained from CIRIA PR11, and hence
more load is attracted to the pile base in CEMSOLVE. It is shown in Section 4 of Lord et al
(1994) that when ρ/D exceeds 0.2 per cent (ie base settlement ρ exceeds about 1 mm for
a 500 mm diameter pile), the yield modulus Ey tends to determine settlement behaviour
at stresses exceeding about 400 kN/m². For medium/high-density chalk, Grades B and
C, Ey varies between 35 and 80 MN/m², of similar magnitude to the CIRIA PR11
moduli given in Table 2.3. Consequently the total shaft load given by CIRIA PR11 has
been used in Table 2.2 wherever possible. Nevertheless as agreement is good, this has
permitted the inclusion of certain CEMSOLVE analyses in Table 2.2 where otherwise
the pile-head movement was too small to permit analysis using the CIRIA PR11 method.
(A pile head movement in excess of 10 mm is required to use the CIRIA PR11 method).
The results of these and previous CIRIA PR11 analyses are plotted in Figure 2.2 as the
average shaft resistance against the average vertical effective stress. In the broad scatter
of results, the majority lie between β = 0.45 and β = 1.5. Various potential causes of
this variation have been investigated:
! locality
! pile dimensions, ie diameter, length and depth of overburden
! time elapsed between installation and testing
CIRIA PR86 17
" analysed
Table 2.1 Pile load test results supplied by the FPS and others # not analysed
Date No. of
Project Name Supplied by supplied Comments
pile tests
Salisbury Kvaerner Cementation November 1997 1 "
(Maltings Lane) Foundations
Newbury Kvaerner Cementation November 1997 1 "
(Park Street) Foundations
(Strawberry Hill) Foundations
(Market Street) Foundations
Reading Kvaerner Cementation November 1997 1 "
(Abbey Square) Foundations
(Greyfriars) Foundations
Windsor Kvaerner Cementation November 1997 4 "
(Goswell Road) Foundations
Worthing Kvaerner Cementation November 1997 1 "
(TP 446) Foundations
Salisbury Kvaerner Cementation November 1997 1 "
(New Street) Foundations
(Kings Road) Foundations
Reading
Kvaerner Cementation November 1997 1
(Thames Valley "
Foundations
Park)
(Valpy Street) Foundations
Lewes Rock & Alluvium 1997 1 # Insufficient head movement
Dover Rock & Alluvium 1997 1 # Insufficient head movement
Reading Westpile/ Illingworth & 7 tests > 10mm head
Chantler (1999) 11 (7)
(Oracle Centre) Pell Frischmann movement

Mott MacDonald/ May 2001
Guildford 1 "
Rock & Alluvium
Canterbury Cementation Foundation June 2001 1 # Insufficient head movement

(Whitefriars) Skanska
High Wycombe June 2001

Stent Foundations 1 # Insufficient head movement
(Oxford Road)
Dartford July 2001
Simplex Foundations 1 "
(Admiral Crescent)
Maidenhead July 2001
Simplex Foundations 9 # Insufficient head movement
(Grenfell Island)
Norwich May Gurney/GCG September 2001 5 # Insufficient head movement
Technopolis September 2001 3
May Gurney/GCG # Insufficient head movement
Norwich
Dean Park Ind. Est. October 2001
Arup 1 "
Beverley
Hedley Avenue November 2001 5
May Gurney/GCG # Insufficient head movement
West Thurrock
Newbury May Gurney/GCG November 2001 4 "
Reading December 2001 1
Westpile "
(Forbury Square)
Total 60 30 used
18 CIRIA PR86
Figure 2.1 Comparison of total shaft load analysed using CEMSOLVE and CIRIA PR11 (1994)
τsf
τsf
Figure 2.2 Average ultimate shaft resistance in chalk v average vertical effective stress on shaft
in chalk for CFA piles
CIRIA PR86 19
20
Table 2.2 Records of CFA piles since CIRIA Report PR11
Case History Pile No. Source of test Pile dimensions Max. load Max. Total shaft Over- Av. vert. Av. shaft Base load Base stress Base Base
data Shaft Length P max settlement load Ps burden effective friction in Pb qb modulus SPT N
dia. Total/Sleeved(s) kN ∆max mm kN friction stress in chalk τ kN kN/m² Eb
sf
(m) O’burden(o) load kN chalk σ ′v MN/m²
/Chalk kN/m² kN/m²
C11 Salisbury - Cementation 0.6 19|6o|13 3300 M 33 1760 U 159 122 65 U 1540 M 5400 M 59 40
Maltings Lane 4253 UC 15000 UC
C12 Newbury - Cementation 0.45 10.5 |5.5o|5 1100 M 11 600 U 92 95 72 * 500 M 3150 M 75 25
Park Street 1143 UC 7200 UC
C13 Newbury - Cementation 0.45 10 |3o|7 1500 M 10 955 U 37 109 93 * 545 M 3400 M 96 25
Strawberry Hill 1072 UC 6800 UC
C14 Newbury - Cementation 0.45 13.5 |4o|9.5 800 M 3 850 UC 57 95 59 C - - 220 C 17
Market Street 2000 UC 12600 UC
C15 Reading - Cementation 0.6 12 |3o |9 3150 M 23 1280 U 49 120 73 U 1900 M 6800 M 132 30
Abbey Square 3467 UC 12200 UC
C16 Reading - Cementation 0.45 10.8 |6.5o|4.3 1710 M 19 976 U 110 140 142 * 734 M 4600 M 67* 20 to
Greyfriars 2027 UC 12700 UC 40
C17 Windsor 112 Cementation 0.6 12.7 |4 o|8.7 2930 M 20 1436 U 90 87 82 U 1494 M 5250 M 96 70
Goswell 4597 UC 16200 UC
227 Cementation 0.45 11.8 |6 o|5.8 1650 M 11 1043 UC 120 90 113 C 1840 UC 11600 UC 187 C 70
52 Cementation 0.6 11.2 | 4o|7.2 2100 M 9 900 UC 79 79 61 C 4000 UC 14100 UC 282 C 40
295 Cementation 0.6 11.4|5o|6.4 2100 M 8 1370 UC 110 84 104 C 2300 UC 8100 UC 247 C
C18 Worthing TP446 Cementation 0.4 13.5|3o|10.5 1060 M 12 454 UC 21 83 33 C 2000 UC 15900 UC 150 C 30
C19 Salisbury - Cementation 0.5 17.7|4.5o|13.2 3000 M 24 1016 U 100 120 44 U 1984 M 10100 M 198 17
New Street 3794 UC 19200 UC
C20 Reading - Cementation 0.75 16|3.4o|12.6 4220 M 17 2113 U 53 115 69 U 2087 M 4700 M 261 50
Kings Road 7400 UC 16700 UC
C21 Reading - Cementation 0.6 20|11o|9 2470 M 13 1366 U 484 157 52 * 1104 M 3900 M 179 60
Thames Valley 5000 UC 17600 UC
CIRIA PR86
Park
C22 Reading - Cementation 0.6 12|1o|11 5400 M 29 2350 U 8 73 113 U 3050 M 10800 M 184 40
Valpy Street 6200 UC 21800 UC
CIRIA PR86
Table 2.2 Records of CFA piles since CIRIA Report PR11 (continued)
Case History Pile No. Source of test Pile dimensions Max. load Max. Total shaft Over- Av. vert. Av. shaft Base load Base stress Base Base
data Shaft Length P max settlement load Ps burden effective friction in Pb qb modulus SPT N
dia. Total/Sleeved(s) kN ∆max mm kN friction stress in chalk τ kN kN/m² Eb
sf
(m) O’burden(o) load kN chalk σ ′v MN/m²
/Chalk kN/m² kN/m²
C23 Reading T3 Westpile 0.6 18|5.4o|12.6 3375 MT 31 3375 M 127 114 137 - - - 29
Oracle T4 Pell Frischmann 0.75 18.5|9.8o|8.7 7438 U 108 1996 U 487 141 74 U 5442 U 12200 U 95 11
T5 (Illingworth & 0.45 15.5|9.8o|5.7 2946 U 65 1350 U 292 126 131 U 1596 U 10000 U 87 11
T6 Chantler 1999) 0.45 15.1|9.8o|5.3 1200 UT 66 1200 U 292 124 121 U - - - 11
T9 0.4 13|4.8o|8.2 1220 MT 19 1220 M 78 94 110 - - - 25
T10 0.45 12.6|2.4o|10.2 4550 M 11 4100 UC 24 104 282 C 4000 UC 25100 UC 350 40
T11 0.45 7|2.4o|4.6 900 UT 39 900 U 24 77 135 U - - - 14
C24 Guildford - Mott MacDonald 0.45 10|0o|10 3324 M 14 1489 0 63 105 U 1835 M 11500 M 347 40
Rock & Alluvium
C25 Dartford - Simplex 0.6 14.7|8.5o|6.2 2400 M 15 1206 313 117 76 * 1194 M 4230 M 117 35
Admiral Crescent
C26 Beverley - Arup 0.45 11|6.8o|4.2 2400 M 27 1167 U 487 93 115 U 1233 M 7770 M 87 50
Dean Park Ind Est
C27 Newbury PTP1 May Gurney/GCG 0.45 11.9|2.1o|9.8 2500 M 16 977 32 109 68 U 1523 M 9580 M 208 30
Vodafone PTP2 0.45 12.6|2.1o|10.5 2500 M 18 1392 32 113 92 U 1108 M 6970 M 124 30
PTP3 0.45 12.1|2o|10.1 2500 M 15 1395 17 100 97 U 1105 M 6960 M 148 30
PTP4 0.45 10.9|2o|8.9 2500 M 17 1223 17 94 96 U 1277 M 8020 M 150 20
C28 Reading TP1 Westpile 0.45 14|2o|12 4650 M 27 2848 U 17 117 167 U 1802 M 11300 M 153 40
Forbury Square
F = Failure ! = made ground/recent fill discounted

M = Maximum value (1) = pile about 5m away from borehole which encountered solution feature
U = Ultimate value u = upper bound estimate
* = interpretation of data less reliable, often on account of limited l = lower bound estimate
pile-head movement
T = Tension C = CEMSOLVE
21
22
Table 2.3 Comparison of total shaft load, base modulus and base load for CFA piles analysed using both CEMSOLVE and CIRIA PR11 (1994) methods
Case History Pile No. Pile dimensions Total Shaft Load Base Modulus Eb Ult. Base Load
Shaft dia.(m) Length (m) CEMSOLVE CIRIA PR11 CEMSOLVE CIRIA PR11 CEMSOLVE
Total/Sleeved(s) kN kN MN/m² MN/m² kN
O’burden(o)/Chalk
Salisbury
C11 0.6 19|6o|13 1530 1760 115 59 4253
Maltings Lane
Newbury
C12 0.45 10.5|5.5o|5 525 600 191 75 1143
Park Street
Newbury
C13 0.45 10|3o|7 1100 955 138 96 1072
Strawberry Hill
Reading
C15 0.6 12|3o|9 1008 1280 313 132 3467
Abbey Square
Reading
C16 0.45 10.8|6.5o|4.3 797 976 157 67 2027
Greyfriars
Windsor
C17 112 0.6 12.7|4o|8.7 1350 1436 152 96 4597
Goswell
Salisbury
C19 0.5 17.7|4.5o|13.2 812 1016 400 198 3794
New Street
Reading
C20 0.75 16|3.4o|12.6 1600 2113 282 261 7400
Kings Road
Reading
C21 Thames Valley 0.6 20|11o|9 1440 1366 175 179 5000
Park
Reading
C22 0.6 12|1o|11 1970 2350 - 184 6200
CIRIA PR86
Valpy Street
Reading T4 0.75 18.5|9.8o|8.7 2475 1610 80 117 10500
C23
Oracle T5 0.45 15.5|9.8o|5.7 1600 1350 75 87 2650
! level of water table
! nature of the chalk, ie density, and/or chalk Grade
! construction techniques including power of rig (in relation to pile diameter and
length, and chalk density and Grade).
These are considered separately below. First, however, the nature of the pile-chalk interface
is discussed as a general factor affecting the shaft resistance of CFA piles in chalk.
2.2 PILE – CHALK INTERFACE
As noted in Section 9.1 of Lord et al (1994), during the boring of inspection shafts in
medium-density chalk at Mundford and in low-density chalk at Norwich using,
respectively bucket and flight augers, it was observed that a skin of remoulded chalk
about 20–30 mm thick had been formed on the sides of the shaft in Grades B, C and D
chalk. Although no measurements were made, the skin seemed to be thicker in low-
density and structureless medium-density chalks than in Grade B and Grade C1-C3
medium-density chalk. (For a definition of chalk density and chalk Grades see Section
1.1.) It was also observed at Mundford that in medium- to high-density Grade A chalk,
the auger tended either to cut the interlocking blocks of chalk or else to pluck
discontinuity-bounded fragments out of the side of the shaft with very little formation
of remoulded chalk. The act of concreting a bored pile does not appear to displace the
moist remoulded chalk from the sides of the shaft.
The act of boring a CFA pile in chalk will also remould the surrounding chalk (see
Figure 30(e) and Section 9.1.3 of Lord et al, 1994). Furthermore, even though CFA
piles are non-displacement (rather than driven, displacement) piles, their excavation
causes considerable displacement, particularly ahead of the bit and laterally, and shear.
(Shearing of the chalk can cause complete remoulding, which in low-density chalks,
has been found to produce a bulking factor of 90 per cent, ie a reduction in volume –
see Lord et al, 2002). The auger rotation permits removal of the chalk from within the
cut diameter, but the edge of the auger flight shears the adjacent chalk not only as it
rotates but also each time an edge moves vertically up or down past a point on the wall
of the hole. Thus chalk higher up the hole experiences shear from the edge of every
flight that goes below it on augering down and on removal, as well as any intermediate
raising and lowering of the auger. Chalk 10 metres above the pile base might experience
perhaps 10 times as many rotations of the flight edge as the chalk at the base of the
hole, but considerably more up and down edge shears. While the remoulded chalk thus
formed may be partly removed in the flights or on the edges on withdrawal, it is likely
that the zone of remoulded chalk will be thicker the nearer it is to the top of the hole,
particularly if the auger is rotating eccentrically or buckles slightly when ‘crowded’
when penetrating a flint band or hardgrounds and so overcuts. The thicker the zone of
remoulded chalk the longer is the drainage path. Also, there is not the same opportunity
for drainage from (and to) the wall of the bore than with an open bored pile, the flights
being reasonably full of impermeable spoil. Although remoulded chalk is likely to be
denser than the intact chalk, particularly for low-density chalk, the benefits of this will
not be realised unless recementing occurs.
Subsequent concreting of the pile will create a reconsolidated remoulded chalk annulus
around the pile shaft. Evidence from the exhumed section of CFA pile shaft described
in Appendix A indicates that with time the remoulded chalk annulus recements and
bonds to the pile concrete to a varying degree. (Similar behaviour of remoulded chalk
has been observed during earthworks – see Chapter 5 of Lord et al, 2002). Thus the
nature of the remoulded chalk annulus, which largely dictates the frictional behaviour
of the CFA pile, varies between a moist silt immediately after concreting or a
recemented state after a month or two.
CIRIA PR86 23
At a very late stage in the project, the opportunity arose to examine a small arc of each
of some 12 CFA piles in chalk exposed over a height of about 5 m as the abutments to
an overbridge. The findings are described in Appendix 1. The piles had been installed
in medium-density Grade C2/C3 chalk some eight months prior to examination. In all
piles any remoulded chalk was very thin (about 1–5 mm) but had recemented. It was
also observed that immediately adjacent to the piles the chalk blocks had been pushed
sideways, so that either Grade A2/A3 or Grade B2/B3 chalk was actually in contact
with the pile.
2.3 LOCALITY
Several of the case histories listed in Tables 1.1 and 2.2 are of CFA piles installed in
areas of Reading and Newbury. Locality, as a potential cause of variation has, however,
been discounted because Illingworth and Chantler (1999) found a wide variation in β
values on the same site (Oracle) in Reading.
2.4 PILE LENGTH, DIAMETER AND OVERBURDEN THICKNESS
The following factors have been considered as possibly affecting values of β, the ratio
of the average shaft resistance to average vertical effective stress for that length of the
pile in chalk, either singly or in combination:
! thickness of overburden
! pile diameter
! pile length in chalk.
β is plotted against thickness of overburden in Figure 2.3 but no apparent correlation is

evident. It has not been possible to identify whether the geological nature of the
overburden, ie whether it is a gravel, sand, silt or clay, has any impact on β.
τsf
τsf
Figure 2.3 Average ultimate shaft resistance/average vertical effective stress for CFA piles v
overburden thickness
24 CIRIA PR86
From inspection of results of 0.45 and 0.6 m diameter piles in Table 2.2, there is no
evidence of β values less than or greater than 0.8 being associated with any particular
diameter. Similarly if β is plotted against the ratio of pile length in chalk to pile
diameter (see Figure 2.4), again no apparent correlation is evident.
τsf
τsf
Figure 2.4 Average ultimate shaft resistance/average vertical effective stress for CFA piles v pile
length in chalk/pile diameter
If β is plotted against the length of the pile in chalk (see Figure 2.5a) there is evidence
of a slight trend of a decrease in β with an increase in pile length; this trend is more
pronounced if the results of bored piles (from Appendix C1 of Lord et al, 1994) are
included, as in Figure 2.5b. This trend is somewhat surprising, but could be a
consequence of increased working of the chalk (as a result of auger rotation) necessary
to bore the pile to the requisite depth, as noted in Section 2.2.
τsf
τsf
Figure 2.5a Average ultimate shaft resistance/average vertical effective stress for
CFA piles v pile length in chalk
CIRIA PR86 25
τsf
τsf 0.8
Figure 2.5b Average ultimate shaft resistance/average vertical effective stress for CFA piles and
bored piles v pile length in chalk
2.5 TIME ELAPSED BETWEEN INSTALLATION AND TESTING
Most preliminary trial piles are tested under load approximately 14 days after
installation when the concrete has achieved an adequate strength.
The two pile tests in which the greatest values of β (up to 2.7) were recorded, were
piles T10 and T11 at the Oracle site in Reading (Table 2.2). Although not recorded as
such by Illingworth and Chantler (1999), it is understood (Illingworth, 2001) that these
two piles were tested more than 80 days after their installation. β values determined for
the 7 No. piles at the Oracle site, Reading and a further pile test at Guildford have been
plotted in Figure 2.6 against the time between pile installation and testing for each pile.
These limited data tentatively indicate a trend of β increasing with time, the result,
perhaps, of recementing of the remoulded chalk annulus around the pile or increased
concrete strength at the pile shaft/chalk interface over time – see also Section 2.2. (An
increase in capacity of at least 15 per cent between 14 and 42 days after installation of a
driven cast-in-place pile is reported in Section 9.1.4 of Lord et al, 1994).
As the magnitude of an increase in shaft capacity with time (ie an increase in β) cannot
be reliably predicted from the present extremely limited data, it would be very
dangerous to extrapolate higher β values for working piles than proven from test piles.
If a pile test just fails to achieve the specified load, however, it might be worth waiting
and then re-testing the pile after a further few days. As a general point it is important
when interpreting the results of pile tests to record the time that has elapsed between
installation and testing.
26 CIRIA PR86
Figure 2.6 Average ultimate shaft resistance/average effective stress for CFA piles v time
between pile installation and testing
2.6 LEVEL OF WATER TABLE
As noted in Section 2.2, the act of boring a CFA pile in chalk will create a zone of
remoulded chalk, which subsequently will reconsolidate as a possibly slightly
recemented annulus bonded to the pile concrete. The presence of groundwater could
both encourage remoulding of the chalk and inhibit recementing. In the plot of average
shaft resistance v. average vertical effective stress for the CFA pile test results in Figure
2.7, those tests in which the level of the water table is above the surface of the chalk,
are differentiated from those where the water level surface is in the chalk. The results
are not conclusive, but there is a qualitative trend which indicates that β values tend to
be lower if the whole of the pile shaft in chalk is below the water table.
τsf
τsf 0.8
in chalk for CFA piles showing impact of water level
CIRIA PR86 27
2.7 NATURE OF THE CHALK
Extensive remoulding of the chalk occurs during the act of boring a CFA pile in chalk
which, on the basis of examination of the sides of the shafts of auger-bored piles (see
Section 2.2), tends to be more pronounced in:
! all low-density chalk
! medium-density chalk where the discontinuities are open more than 3 mm and their
spacing is less than 60 mm, ie Grades C4 and C5
! structureless chalk (Grade D)
as identified in Section 1.1.
A full description of the engineering classification of chalk is given in Section 2.2 of

Lord et al (1994), amended slightly in Chapter 3 of Lord et al (2002). In these books, it is
emphasised that the visual inspection of the chalk in situ is the best way to classify
chalk. Nevertheless such in-situ inspection is not feasible at depth (except in special
cases) and certainly not beneath the water table. In these circumstances large diameter,
high quality rotary cored boreholes are required. The measurement of intact dry density
is possible on chalk samples recovered from boreholes (see Appendix 2), but this is
only slowly gaining acceptance in standard site investigation practice. Consequently in
most of the site investigation reports made available to piling contractors, such
information is absent or limited – measurement of SPT blowcounts at typically 1.5 m
intervals is often all that is available. But as pointed out by Lord et al (1994), the use of
the SPT for pile design in chalk has many shortcomings.
! Several authors, eg Mallard and Ballantyne (1976); Lord and Smith (1976);
Montague (1990); Twine and Grose (1990a), have shown that the SPT blowcount,
as recorded, would appear to be operator-dependent.
! The SPT blowcount is unable to distinguish between a medium-density Grade Dc
chalk at Mundford (Wakeling, 1970) and a low-density Grade B4 chalk at Norwich
(Twine and Grose, 1990a) for both of which N is typically about 10.
! The SPT may give misleading high N values in flinty chalk.

! Clayton (1990 and 1995) notes that in chalk the dependence of penetration
resistance on a number of independent factors makes precise interpretation of the
test impossible.
! N values often exhibit considerable scatter over the length of the pile shaft.
! Ascribing characteristics to the pile-chalk interface on the basis of SPT values is
even more tenuous than the uncertain link between the nature of the intact chalk and
SPT blowcounts.
Nevertheless, as the SPT is so widely used, particularly in ground investigation for piling,
those tests in which all SPT blowcounts in the chalk exceed 10 have been differentiated
from those tests in which N ≤ 10 in the plot of average shaft resistance v. average
vertical effective stress (Figure 2.8). (The possible influence of flints on the N value has
not been taken into account on this plot.) It can be seen that where N > 10, β is
generally greater than or equal to 0.8 whereas when N ≤ 10, β is generally less than 0.8.
These latter results are for piles in Grade D structureless chalk, and in low-density chalks.
28 CIRIA PR86
τsf
τsf
in chalk for CFA piles showing impact of low penetration resistance in chalk
A similar approach was adopted by Illingworth and Chantler (1999) who, recognising
the shortcomings of SPT profiling, and in particular the considerable scatter of results,
used cone penetration testing (CPT) data to identify changes in strata and provide a
better correlation of the profile of penetration resistance across the site. In their case
they delineated the thickness of what they termed “softer or weathered chalk” (ie the
surface of the underlying intact load-bearing chalk) on the basis of whether the cone
resistance, qc, exceeded 2 MN/m² (which approximately corresponds to N > 5). The
recommended use of the CPT, rather than SPT, is discussed further in Section 3.3.
As a result of the difficulties in sampling chalk with N ≤ 10, it is not possible to be

certain of its precise nature. Wakeling (1970) has shown that is was either medium-
density Grade C4-C5 chalk or Grade Dc chalk at Mundford, while Twine and Grose
(1990a) have shown that it was low-density Grade B4 chalk at Norwich. But in low-
density Grade A4 and B3 chalk at Northfleet N = 15 to 25. Hence not all low-density
chalks have a low penetration resistance. It is tentatively concluded that low penetration
chalk has a greater propensity to remoulding.
2.8 CONSTRUCTION TECHNIQUES
This variability in the shaft resistance of CFA piles is not unique to chalk. Over the past
decade there have been several examples of sites where for certain piles the shaft
resistance has been found to be very low whereas for the remainder it closely matches
the soil design strengths. Such soils have included Folkestone Beds, Sandgate Beds,
Atherfield Clay, Bagshot/Bracklesham Beds and Mercia Mudstone (Ove Arup and
Partners Project Files). This would imply that construction techniques and
workmanship may influence pile behaviour more than is currently accepted, and that
this alone could produce low β values under certain conditions.
It is widely believed that the rate of pumping of concrete relative to the rate of auger
withdrawal is a major factor in controlling the quality of CFA pile construction. To
ensure that the auger is withdrawn sufficiently slowly to allow a large enough volume
of concrete to be pumped there is increasing use of CFA rig instrumentation and in-cab
display screens showing “real time” data during pile construction. But rig
instrumentation must be robust, and has therefore necessarily remained relatively basic.
It is by no means certain that even where this is available, the quality of data and
control are sufficient to guarantee concreting quality.
CIRIA PR86 29
A research programme studying the installation and behaviour of CFA piles in progress
at the University of Southampton has found that currently available CFA piling rigs
cannot deliver sufficient torque to bore a pile without flighting the soil and if they did
screw themselves into the ground, they could not pull the auger out. (Flighting is a
vertical movement of the soil on the auger relative to the soil on the wall of the bore.)
Thus the soil must be flighted to keep the auger free. As a result, augers are typically
advanced at only about 50 per cent of their pitch per revolution, and flighting leads to
large reductions in total stress on the pile bore. In unfavourable ground conditions (eg
clay soil containing silty or sandy fabric) this can be expected to produce significant
softening of the pile bore and reductions in pile shaft capacity.
It is considered that there should be little difficulty in forming satisfactory CFA piles in
better quality structured chalk, but that in chalks, characterised by low penetration
resistance, there may be problems of softening and bore instability, and perhaps
associated lack of pile integrity, particularly below the water table.
It is known that concrete mixes are varied by piling contractors on the basis of
workability and availability of materials. Mixes generally contain a filler comprising
either Pulverised Fuel Ash (PFA) or Ground Granulated Blast Furnace Slag (GGBFS).
There is no data concerning whether the concrete mix affects pile/chalk adhesion, but it
is suggested that further data may be collected to investigate this.
Nonetheless, construction techniques or workmanship have been used to explain

discrepancies in pile behaviour. But analysis of test data indicates that natural variations
in the chalk on a site, together with variations in the interval between pile installation and
testing, can explain some attributions of behavioural variations to workmanship. For
this reason construction techniques have been considered of lesser impact on pile behaviour.
Despite requests to the Federation of Piling Specialists (FPS) for information on

opportunities to either exhume or examine the pile shaft/chalk interface of CFA piles
during the course of this research project, the only such opportunity occurred at a very
late stage in the project when twelve CFA piles, which supported and formed the
abutment to an overbridge on the A41 Aston Clinton by-pass, were exposed over a
small sector of their perimeter. The findings are described in Appendix 1. As it was
only possible to examine in detail about one-quarter of the perimeter of the shaft over a
length of about 2 m, it was not possible to determine the shaft dimensions, while a
study of the pile concrete/chalk interface was very restricted. However, this
examination was sufficient to reveal that the interface was complex, varying between a
very smooth and a very rough surface, with a thin skin of remoulded chalk, which had
recemented with time. It had been hoped that as part of this research project, a CFA pile
in chalk could be installed, loaded to failure and subsequently exhumed – most
economically achieved if a pile was installed in an area where a cutting was
subsequently to be formed in the chalk. Exhumation of complete pile(s) would also
permit comparison of pile overbreak /reduction in volume of remoulded chalk with the
pumped concrete volume measurements and the diameter of the finished pile. Only by
such correlation will it be possible to determine whether small increases in pile
diameter during construction can be relied upon – see also Section 2.1.
2.9 ASSESSMENT
Of the factors that have been considered as influencing β for CFA piles in chalk, it
would appear that two have the greatest impact:
(a) presence of chalk with lower penetration resistance
(b) time elapsed between pile installation and testing.
30 CIRIA PR86
A further three have less impact:
(c) length of pile shaft in chalk
(d) level of water table
(e) construction techniques and workmanship.
Factors (c) and (e) are not independent of and may be linked to factor (a). For example,
the factor of workmanship is frequently used to explain discrepancies in pile behaviour
on the same site. But the above analysis of pile test data indicates that natural variations
in chalk density and Grading on a site, together with different times between pile
installation and testing, could be responsible for as great differences in pile behaviour
as those attributed to workmanship.
As discussed in Section 2.7, low penetration resistance in chalk, as measured by SPT or

preferably CPT, is indicative of a chalk that is prone to remoulding around the pile
annulus during boring. Any remoulding will be exacerbated if there is:
! ‘working’ of the chalk – for CFA piles, the upper section of the shaft in chalk will
be worked more by the auger for longer piles than for shorter piles, as discussed in
Section 2.2
! abundant free water – hence piles with the water table above the surface of the
chalk are likely to create more remoulded chalk, while longer piles are more likely
to encounter the water table than shorter piles
! excessive rotation of the auger per m penetration, particularly if the CFA piling rig
is underpowered.
As noted in Section 2.5, there is currently insufficient data to be sure if there is an

increase in shaft friction with time as a result of increased recementing of the
remoulded chalk. Nevertheless the time elapsed between pile installation and testing
should always be recorded so that the test results can be interpreted appropriately.
It appears that for low penetration resistance chalk, as categorised by N ≤ 10 or qc ≤ 4

MN/m², β should be taken as 0.45. For medium-density chalk, or chalk where N >10
(only if flints are absent) or qc > 4 MN/m², β should be taken as 0.8. But if the ground
investigation data is of poor quality, for example:
! insufficient boreholes
! boreholes (or CPTs) more than 25 m from a pile location
! inadequate penetration into chalk
! no SPTs, or SPTs at 5 m centres, which do not correlate with chalk descriptions
! sole reliance on SPTs
! inadequate soil descriptions
! no reference to flints or groundwater
! absence of intact dry density laboratory testing
it is considered that β should be taken as 0.45 for a safe design. If flints are present the
SPT N value may not be relied upon to indicate higher penetration resistance chalk.
It should be noted that for bored piles in chalk the lower-bound value of β is 0.8 (see
Figure 2.9) with a possible upper bound for β of about 1.35. The majority of the lower-
bound pile test results were obtained at Castle Mall, Norwich, where the chalk is of
low-density. It is evident that the shaft resistance of CFA piles is less than that for bored
piles for low-density chalk. It is considered that this is due to the bored pile auger
causing less ‘working’ (ie remoulding) of the chalk for the same length of pile than
CIRIA PR86 31
does a CFA pile auger – see Section 2.2. Consequently because the remoulded zone is
thinner for bored piles, it will not take so long to reconsolidate/recement as for CFA
piles. Nevertheless there is currently insufficient data to quantify the influence of the
elapsed time between pile installation and testing, and any time-dependent
improvement in β should not be relied upon unless it is supported by site-specific
preliminary pile tests. A logic flow diagram for determining the shaft resistance of CFA
piles in chalk has been presented by Spink (2002b) – see Figure 2.10.
τsf
in chalk for bored piles showing impact of chalk density
The implications of the above design recommendations are illustrated in Figures 2.11a
and 2.11b. In the FPS position paper (FPS, 2000), the differences between the allowable
loads for 600 mm diameter CFA piles calculated using the CIRIA PG6 method for
bored piles (1979) and CIRIA PR11 method for CFA piles (1994) were compared. The
design cases and basis of design were:
Design cases
! Low SPT (corresponding to low penetration resistance chalk)
N = 5 + d where d = depth below top of chalk in metres.
! High SPT (corresponding to higher penetration resistance chalk)
N = 20 +d where d = depth below top of chalk in metres.
! Water at either 2 m or 7 m depth below ground level.
! Overburden thickness of either 2 m or 7 m.
CIRIA Report PG6 (1979) design approach

! Ultimate shaft friction, τsf, related to SPT N value from Figure 14 of PG6 (although
when that report was published the use of CFA piles in the UK was extremely
limited and none of the test results used as a basis for Figure 14 was for a CFA pile)
! Ultimate base resistance qu = 0.24 N MN/m² for N < 30
qu = 0.2 N MN/m² for N > 40
Pro rata for values of N between 30 and 40
! Allowable pile capacity Pa is the lesser of
Psu + Pbu or, Psu + Pbu
1.5 10 2.5
32 CIRIA PR86
33 CIRIA PR86
Figure 2.10
Logic flow diagram for determine shaft resistance of CFA piles in chalk (after Spink, 2002b)
CIRIA Report PR11 (1994) approach
! Ultimate shaft friction = β σ ′v
! Ultimate base resistance qu = 0.2N MN/m²
! Allowable pile capacity Pa is the lesser of
Psu Pbu Psu + Pbu

Pa = + or,
1 .0 3.5 2.5
(see Chapter 4 of this report)
The allowable loads for 600 mm diameter piles in low penetration resistance chalk with
β = 0.45 are shown in Figure 2.11a for different pile penetrations into chalk. For the
examples with only 2 m of overburden, the results from the two design approaches are
similar, but where the chalk is overlain by 7 m or more of overburden the effective
stress method gives higher allowable loads.
With β = 0.8 for higher penetration resistance chalk, the allowable loads for 600 mm
diameter piles are as shown in Figure 2.11b for the two design approaches. In comparison
with a similar graph in FPS (2000), the effective stress method now gives pile lengths
which, although greater than those given by the PG6 design approach for the same
allowable load, are about 8 m shorter than given by the CIRIA PR11 (1994) design
recommendations. As the PG6 design approach is not supported by rigorous back-
analysis of pile test behaviour whereas the effective stress method is (see Figure 2.8), it
is recommended that the design approach with β = 0.8 is adopted for higher penetration
resistance chalk, unless pile load test data show a higher value of b to be appropriate.
2.10 EXAMINATION OF CFA PILE IN CHALK IN SITU AND FUTURE

RESEARCH
Uncertainties about CFA piles in chalk, such as:

! the as-built diameter of the pile (sections 2.1 and 2.3)

! the nature of the pile-chalk interface in chalks of varying density and Grade
(Section 2.7)
! possible reconsolidation/recementing of remoulded chalk with time (sections 2.2
and 2.5)
! effects of construction techniques (Section 2.8)
could be better understood if there was an opportunity to exhume CFA piles of different
lengths in chalks of different densities and Grades. This could be most economically
achieved if a pile was installed in an area where a cutting was subsequently to be
formed in the chalk. Although there was a very limited opportunity to examine part of
the shaft of a CFA pile during the course of this research project, the authors would
welcome the chance to examine CFA piles in situ if the opportunity arises in the future.
There is also still a need for data to permit greater refinement of pile design both in
terms of in situ testing, such as
! correlation of behaviour of CFA test piles against CPT records (see also Section 3.3)
! more data on preliminary CFA test piles loaded to 2.5 or 3 x WL (see also Chapter 4)
– information to be recorded when a pile test is undertaken is given in Appendix 3
34 CIRIA PR86
Figure 2.11a Allowable loads for 600 mm diameter CFA piles in low penetration resistance in
chalk
Figure 2.11b Allowable loads for 600 mm diameter CFA piles in higher penetration resistance
chalk
CIRIA PR86 35
and further laboratory-based research into some of the factors/uncertainties affecting
CFA pile behaviour discussed in Section 2.2. Prominent among these would be:
! the nature of low penetration resistance chalk
! the propensity of low- and medium-density chalks to remould to differing extents
dependent on their natural moisture content and the magnitude of the working or
shearing of the chalk
! the factors which affect the ability of remoulded chalk to reconsolidate/recement with
time and the rate at which this occurs, and hence the time-dependent variation of β.
36 CIRIA PR86
3 Identification of chalk with a low unit
shaft resistance
The propensity of low penetration resistance chalk to remould more readily than higher
penetration resistance chalk and so provide less unit shaft resistance on CFA piles has
been assessed in Section 3.7. In order to identify chalk likely to break down and
remould to low strengths, three techniques are available:
! directly from measurement of dry density of lumps of chalk recovered from boreholes,
either in the laboratory or by simple field tests (see Table 3.7 of Lord et al, 2002)
! indirectly from Standard Penetration Test (SPT) blowcounts in non-flinty chalk
! indirectly from Cone Penetration Test (CPT) cone resistance.
The suitability of various site investigation methods appropriate to CFA piling in chalk
are listed in Appendix 2 (after Spink, 2002b).
3.1 DETERMINATION OF DRY DENSITY
The measurement of the dry density of lumps of chalk is probably the most reliable
way of identifying low density chalk but, as noted in Section 2.7, this technique is only
slowly gaining acceptance in standard site investigation practice. It is hoped that with
the publication of Engineering in Chalk (Lord et al, 2002) in which the importance of
chalk density is heavily emphasised, measurements of chalk density – a relatively cheap
laboratory test – will become more commonplace. Until then reliance will still have to
be placed on indirect indications obtained by in situ testing.
As indicated in Section 2.7, penetration resistance, which is indicative of a chalk’s

ability to remould, is a function of both the dry density and Grade of the chalk. Low
penetration resistance occurs in some low-density chalks and medium-density chalk of

Grades C4, C5 and D. The chalk Grade is best determined by in situ logging in trial
pits, but for site investigations for CFA piling, large diameter, high quality rotary cored
boreholes are required (see Appendix 2).
3.2 STANDARD PENETRATION TEST
The shortcomings of the SPT for pile design in chalk are restated in Section 2.7. Of
these, the errors associated with flints and operator-dependence (often considerable –
see Mallard and Ballantyne (1976), Twine and Grose (1990a) and Clayton (1990 and
1995)) are most serious. Furthermore SPTs, commonly made at 1.5 m intervals, do not
provide a continuous profile in the chalk and often exhibit considerable scatter. For
these reasons, the use of an averaged SPT blowcount multiplied by a correlation factor
to determine the shaft resistance of a CFA pile in chalk should be strongly discouraged.
In the absence of other information, the SPT may be used as an indicator of readily
remoulded and structureless chalk when N ≤ 10. As suggested in Figure 2.8 and Section 2.7,
if all SPTs record N > 10 this is indicative of medium- to high-density chalk and β can
be taken as 0.8, but if there are some values of N ≤ 10, β should be taken as 0.45. Even
if the SPT is operator-dependent, it would be expected that in low-density chalk the
blowcounts would usually be low (or indicative of the presence of dissolution features).
However, as N can vary by 100 per cent due to operator-dependency (see Clayton,
1995), the divisional boundary between low penetration resistance and higher
penetration resistances could also vary by the same amount.
CIRIA PR86 37
3.3 CONE PENETRATION TEST
The CPT has the advantage over the SPT of providing a continuous resistance profile in
chalk and greater speed of operation, but with the disadvantage of penetration being
prevented if numerous or large flints are encountered. Apart from the recent work of
Smith (2001), the only studies on its use to determine the engineering properties of
chalk were by Power (1982) and Powell and Quarterman (1994), both reported in
Lunne et al (1997) and summarised in Section 9.7.2 of Lord et al (2002). Power (1982)
correlated the CPT and the SPT for a number of sites, obtaining the relationship:
qc = 0.4N (MN/m²)
There was a great deal of scatter of his results, reflecting both the natural variability of
the chalk and the uncertainties associated with the SPT. Illingworth and Chantler (1999)
derived the same relationship at Reading, based on cone tests adjacent to boreholes.
They also found that the surface of intact chalk corresponded to the cone resistance, qc,
exceeding 2 MN/m² (ie N > 5). A similar value was observed on the Channel Tunnel
Rail Link at the base of dissolution features (reported by Smith, 2001). On the basis of
the above relationship between qc and N, a readily remoulded low-density or structureless
chalk would be expected to have a cone resistance of between 2 and 4 MN/m². It is to
be hoped that the greater use of the CPT as a ground investigation tool in chalk will, in
a few years’ time, provide a database capable of being ‘chalk type/area’ specific.
Although CPTs are capable, with careful interpretation, of identifying the presence and
depth of dissolution features in the chalk (see Smith, 2001), correlation against
borehole(s) is still necessary; techniques for identifying dissolution features are
discussed in more detail in Sections 5.8.1, 9.4 and 9.7.2 of Lord et al (2002). As the
construction of CFA piles in effect necessitates drilling blind, it is recommended that a
CPT is undertaken at each pile location at sites known to be prone to dissolution.
Experience has shown that CPTs are probably the quickest and most cost-effective
means of identifying the presence and depth of dissolution features in the chalk.
3.3.1 Use of CPT to predict base resistance

The use of the CPT is also perceived as providing a better means of determining the
ultimate base resistance of CFA piles in chalk. Lord et al (1994) had suggested a lower-
bound relationship of:
qu = 200N (kN/m²)
Applying to this the relationship between qc and N, Lord et al (2002) suggested that in
terms of cone resistance:
qu = 0.5 qc (MN/m²)
On the basis of the ultimate base resistances of 20.3 and 13.9 MN/m² predicted by
CEMSOLVE for piles T4 and T5a at the Oracle site, Reading reported by Illingworth
and Chantler (1999) (see also Table 2.2) and the CPT results reported by Smith (2001)
for that site, the relationship:
qu = qc (MN/m²)
would appear appropriate. But such relationships cannot be relied upon until a greater
database is available.
38 CIRIA PR86
4 Pile testing and factors of safety
One of the objectives of this research project was to identify how savings might be
made by the piling industry and their clients in the design and construction of CFA piles
in chalk. The recommendations given in Section 2.9 for determining the shaft resistance
of CFA piles in high penetration resistance chalk will permit a reduction in the length of
the pile and hence allow savings in terms of programme and materials. There are,
however, potential savings in time (and hence cost) to be had through re-addressing the
sequence of pile installation and testing. At present a typical piling project comprises
the following stages:
1. Installation and testing of preliminary trial pile(s).
2. Installation of working piles.
3. Proof testing of a limited number of working piles to 1.5 x WL, often as required by
local authorities.
Frequently Stage 1 is omitted and the pile design is based on lower-bound design
recommendations such as given in Lord et al (1994) and this report. Consequently such
design recommendations should always be lower-bound design lines as they have to
take account of the lowest common denominator or worst-case scenario. Nevertheless
the opportunity has always existed in the design of CFA piles in chalk to adopt other
design lines (for example, higher β values) provided that they can be substantiated by
preliminary pile tests. This was the procedure followed by Illingworth and Chantler
(1999) – see below.
If Stage 1 is omitted, the only pile testing is in Stage 3 and is little more than an
expensive test of workmanship, providing no indication of whether the overall factor of
safety is as low as 1.6 or greater than 2.5. Such working pile tests frequently cause
delays on site, not only to the piling contractor, but also to the main contractor. These
delays arise as a result of needing to:

! wait up to 28 days while pile concrete achieves the requisite cube strength
! install tension piles and/or pile reaction
! clear away pile reaction system on completion of the test.
As a result, part of the site may be sterilised for a period of four to six weeks, possibly
delaying the main contractor in a programme-critical area.
Such considerations could have severely delayed construction on the Oracle site in
Reading. Consequently Pell Frischmann Consulting Engineers and Westpile Ltd
approached the local authority to seek their co-operation by agreeing to a programme of
preliminary pile testing (as described by Illingworth and Chantler, 1999) but omitting tests
on working piles. Not only were savings achieved as a result of being able to use shorter
piles, but also as a result of a reduction in the main contractor’s construction programme.
It is considered that there is greater potential for cost savings by undertaking one
preliminary pile test to 3 x WL instead of two working pile tests to 1.5 x WL.
Consequently clients, their main contractors and piling sub-contractors should always
seek to undertake preliminary pile tests to enable a reduction in the number of working
pile tests during the contract. Furthermore as discussed in Section 2.5, there is some
evidence that for CFA piles, β may increase with time. In situations where the piling
CIRIA PR86 39
programme is phased and the number of piles warrants it, consideration should be given
to installing additional preliminary trial piles prior to the installation of Phase 1 working
piles, but only testing those piles just before the installation of the Phase 2 working
piles, to see whether the pile lengths can be reduced as a result of an increase in β.
In Lord et al (2002), the factors of safety for pile design in chalk were modified so that
the allowable capacity of the pile, Pa , should be the lesser of:
Psu P  P P 
either, P a = + bu  P a = su + bu if settlement to be less than 10 mm 
1.0 3.5  1.5 3.5 
Psu + P bu
or, Pa =
2 .5
where Psu and Pbu are the calculated ultimate shaft and base loads respectively. By
adopting a load factor on the base of the pile (Fb) of 3.5, base resistances of 600 to 800
kN/m² for low-density chalk and 1000 to 1800 kN/m² for medium/high-density chalk,
are sufficiently constrained to ensure that ρ/D >1 to 2 per cent, are unlikely to be
exceeded. The above factors of safety are broadly commensurate with those adopted for
pile design in soils other than chalk, and relate to acceptable pile head deflection issues
rather than failure. In general it has been found that for CFA piles in chalk, the overall
factor of safety of 2.5 tends to dictate the allowable pile capacity. For longer piles in
medium/high-density chalk, the overall capacity of the pile may be dictated by concrete
strength and not by the chalk (see, for example, Figure 2.11b).
In high/very high-density Grade A chalk, the ultimate base load can be assessed in
terms of rock socket design (see Lord et al, 2002), but may be limited by the maximum
permitted concrete stress in the pile shaft.
40 CIRIA PR86
5 Conclusions and recommendations
1. The shaft resistance τsf of CFA piles in chalk can be expressed in terms of the
empirical relationship:
τsf = β σ′v
where σ′v = average vertical effective stress along pile shaft in chalk.
2. It appears that the factors most influencing β for CFA piles are:
(a) whether the penetration resistance of the chalk is low

(b) time elapsed between pile installation and testing.
The following factors, which may be linked to (a), have a less effect on β:
(c) length of pile shaft in chalk
(d) level of water table
(e) construction techniques.
3. Low penetration resistance in chalk, as measured by the SPT or preferably the

CPT, is indicative of:
! some low-density chalk of any Grade
! all structureless (Grade D) chalk of any density
! medium-density Grades C4 and C5 chalk
Such chalk is prone to create remoulded material around the pile during boring.
4. Three techniques are available for identifying readily remoulded chalk:

! measurement of dry density of lumps of chalk recovered from boreholes, either
in the laboratory or by simple field tests (see Table 3.7 of Lord et al, 2002)
together with determination of chalk Grade where possible
! Standard Penetration Test (SPT) blowcounts (where flints are absent)
! Cone Penetration Test (CPT) cone resistance.
As the first technique by direct measurement is only slowly gaining acceptance,

reliance currently is largely placed on in-situ indicators such as the SPT and CPT.
5. The SPT has been found to be operator-dependent, does not provide a continuous
profile in chalk, often exhibits considerable scatter (Clayton, 1990 and 1995), and
is adversely affected by the presence of flints. For these reasons, taking an
averaged SPT blowcount multiplied by a correlation factor should not be used to
determine the shaft resistance of a CFA pile in chalk. Nevertheless, in the absence
of any alternative information, the SPT can be used as an indicator of readily
remoulded chalk when N ≤ 10.
6. The CPT has the advantage over the SPT in providing a continuous resistance
profile in chalk and greater speed of operation, but with the disadvantage of
penetration being prevented and possible damage to equipment if numerous or
CIRIA PR86 41
large flints and hard ground are encountered. Illingworth and Chantler (1999)
have found on one site that the surface of competent chalk corresponded to the
cone resistance, qc, exceeding 2 MN/m². A low penetration resistance chalk
would be expected to have a cone resistance of between about 2 and 4 MN/m².
7. As illustrated in Figure 2.10, it is recommended that in deriving the shaft

resistance of CFA piles in chalk, the following should be adopted in the absence
of preliminary pile tests:
For low penetration resistance chalk over part of the pile shaft, categorised by
N ≤ 10 or qc ≤ 4 MN/m², β = 0.45 throughout.
For high penetration resistance chalk over all the pile shaft, categorised by N > 10
(if flints are absent) or qc > 4 MN/m², β = 0.8 throughout. But if the ground
investigation data are of poor quality (see Section 2.9), β should be taken as 0.45
for safe design. If flints are present the SPT N value may not be relied upon to
indicate high penetration resistance chalk.
8. Unless shown otherwise by pile testing, the ultimate base resistance of CFA piles
in chalk, qu, should be taken as
qu = 0.5 qc (MN/m²)
where qc is the cone resistance averaged over a distance of one to two pile
diameters beneath the base of the pile.
9. The presence of dissolution features cannot be readily identified in the

construction of CFA piles, it is recommended that a CPT is undertaken at each
pile location at sites considered to be prone to dissolution. Techniques for
identifying dissolution features are discussed in Lord et al (2002).
10. The allowable capacity of a CFA pile, Pa, should be the lesser of:
Psu P  P P 
either, P a = + bu  P a = su + bu if settlement to be less than 10 mm 
1.0 3.5  1.5 3.5 
Psu + P bu
or, Pa =
2 .5
where Psu and Pbu are the calculated ultimate shaft and base loads respectively.
These factors of safety are related to the CIRIA PR11 design method. For longer
piles in medium/high-density chalk, the overall capacity of the pile may be
dictated by concrete strength and not by the chalk.
11. Significant economies, through pile shortening, might be achieved by

demonstrating higher β values in preliminary pile tests. There is also a greater
potential for cost savings by undertaking one preliminary pile test to 3 x WL
instead of two working pile tests to 1.5 x WL. Where the number of working piles
are significant, preliminary pile tests should be planned to investigate the change
of pile capacity as a result of time elapsed between pile installation and testing.
The authors consider that, where sufficient working piles enable it to become
feasible, sufficient preliminary pile tests planned to investigate both β and the
effect of time on capacity may well enable a reduction in the number of working
piles, and hence overall economies, to be made. In situations where the piling
programme is phased and the number of piles warrants it, consideration should be
42 CIRIA PR86
given to installing additional preliminary trial pile(s) prior to the Phase 1 working
piles, but only testing that pile just before the installation of the Phase 2 working
piles, to see whether there is an increase of β with time after pile installation and
whether advantage could be taken of this in pile length reduction.
12. The present research project has permitted a provisional ranking of the parameters
most affecting variance of β and thereby has identified how savings might be
made by the piling industry, particularly in medium/high-density chalk. But there
is still a need for data to permit greater refinement of pile design, with
opportunities for further research, notably:
! examination of pile/chalk interface of exhumed CFA piles in chalks of varying
density and Grade
! the nature of low penetration resistance chalk
! the propensity of low- and medium-density chalks to remould to differing
extents dependent on their natural moisture content and the magnitude of the
‘working’ or shearing of the chalk
! the factors that affect the ability of remoulded chalk to reconsolidate/recement
with time and the rate at which this occurs, and hence the time-dependent
variation of β
! correlation of behaviour of CFA test piles against CPT records
! more data on preliminary CFA test piles loaded to 3 x WL – information to be
recorded when a pile test is undertaken is given in Appendix 3
! instrumented pile test.
CIRIA PR86 43
44
CIRIA PR86
References
BOWDEN A J, SPINK T W and MORTIMORE R N (2002) The engineering

description of chalk: its strength, hardness and density, Quarterly Journal of
Engineering Geology and Hydrology, vol 35, part 4, Nov
BRITISH STANDARDS INSTITUTION (1981) BS1377: Part 2: (1990) British

Standard Methods of test for Soils for civil engineering purposes, Part 2, Classification
tests, British Standards Institution, London
BRITISH STANDARDS INSTITUTION (1999) BS5930: Code of Practice for Site

Investigations, British Standards Institution, London
CLAYTON C R I (1990) The mechanical properties of the Chalk, In CHALK, Thomas

Telford, London, pp 213–232
CLAYTON C R I (1995) The Standard Penetration Test (SPT): Methods and Use,
CIRIA Report 143, CIRIA, London, pp 143
ENGLAND M K C R (1999) A pile Behaviour Model, Ph.D Thesis, Imperial College
FLEMING W G K (1992) A new method of single pile settlement prediction and

analysis, Geotechnique, Vol XLII, No.3, Sept 1992
FLEMING W G K (1997) Personal communication
FPS (2000) Design of CFA piles in chalk, Ground Engineering, April 2000, pp 31–32
HOBBS N B and HEALY P R (1979) Piling in Chalk, DOE/CIRIA Piling Development

Group Report PG6, CIRIA, London
ILLINGWORTH D R (2001) Personal communication
ILLINGWORTH D R and CHANTLER J P (1999) CFA piles in Chalk at the Oracle,

Reading, Proc. Tunnel Construction and Piling ‘99 Int. Symp. and Exhibition, Sept
1999, London, Brintex Ltd, London, pp 45–56
LORD E R F and SMITH W E (1976) The misuse of SPT N value correlations with
Upper Chalk grades, Géotechnique, Vol.26, pp 217–220
LORD J A, TWINE D and YEOW H (1994) Foundations in chalk, CIRIA Project

Report 11, CIRIA, London, 189 pp
LORD J A, CLAYTON C R I, and MORTIMORE R N (2002) Engineering in Chalk,

CIRIA Publication C574, London
LUNNE T, ROBERTSON P K and POWELL J J M (1997) Cone Penetration Testing,

Blackie Academic and Professional
CIRIA PR86 45
MALLARD D J and BALLANTYNE J L (1976) The behaviour of piles in Upper
Chalk at Littlebrook ‘D’ Power Station, Géotechnique 26, (1) pp 1115–132
MONTAGUE K N (1990) The SPT and pile performance in Upper Chalk, Int. Chalk
Symposium, Brighton, 1989, Thomas Telford, London, pp 269–276
POWELL J J M and QUARTERMAN R S T (1994) A reappraisal of CPT testing in

chalk, BRE Report No.G/GP/9412
POWER P T (1982) The use of the electric static cone penetrometer in the
determination of the engineering properties of chalk, Proc. 2nd European Symp.
Penetration Testing, Amsterdam, pp 769–774
SMITH A K C (2001) The interpretation of static cone penetration tests in Chalk,

Ground Engineering, September 2001, pp 30–35
SPINK T W (2002a) The CIRIA chalk description and classification scheme. The
Channel Tunnel Engineering Geology Symposium, 1995, Quarterly Journal of
Engineering geology and Hydrology, vol 35, part 4, Nov
SPINK T W (2002b) Personal communication
TWINE D (1990) Discussion: Session III. Proc. Int. Chalk Symposium, Brighton, 1989,
Thomas Telford, London, pp 415–416
TWINE D and GROSE W J (1990a) Discussion: Session 1. Proc. Int. Chalk

Symposium, Brighton, 1989, Thomas Telford, London, pp 209–210
WAKELING T R M (1970) A comparison of the results of standard site investigation

methods against the results of a detailed geotechnical investigation in Middle Chalk at
Mundford, Norfolk, Proc. Conf. on In-situ Investigations in Soils and Rock, BGS,
London, pp 17–22
46 CIRIA PR86
Appendix 1
A1.1 EXAMINATION OF EXPOSED SHAFT OF CFA PILES
The only opportunity to either exhume or examine the pile shaft/chalk interface of CFA
piles during the course of this research project, occurred at a very late stage in the project
when twelve CFA piles which supported and formed the abutment to an overbridge at
Tring roundabout on the A41 Ashton Clint by-pass were exposed over a small sector of
their perimeter. (The Ashton Clinton by-pass is a Highways Agency design-and-build
contract undertaken by Balfour Beatty Civil Engineering with Ove Arup & Partners as
their designer).
The twelve CFA piles which support the bridge deck and which, when faced, form the
north and south abutments comprised 900 mm diameter piles installed at 2900 mm
centres (see Figure A1.8(a)). The piles were constructed between 16 and 18 January
2002; the exposed sector of pile S5 was examined in some detail on 14 August 2002.
Prior to piling the chalk had been covered by a layer of hardcore to form a platform
800 mm above the underside of the pile cap – the hardcore was pulled back at each of
the pile locations before boring commenced.
For pile S5, a section of pile between 1.2 and 3.0 m below the underside of the pile cap
(ie between 2.0 and 3.8 m below piling platform level) was examined. After exposing
the face of the pile with the bucket of a machine excavator, the exposed perimeter of
the pile was exposed by hand over a width of about 500–600 mm, about one-quarter of
the perimeter of the pile shaft. This was carefully examined and photographed.
On 20 September 2002 when all piles had been exposed over a height of about 5 m
between the underside of the pile cap and the sub-grade level of the new carriageway,
albeit only over a perimeter of between 300 and 600 mm, a further examination was
made for general trends and observable differences, and the piles photographed (see
Figures A1.1– A1.7).
The following broad conclusions were drawn:

! the chalk in which the piles were constructed was a medium-dense Grade C2–C3 chalk
! although it was relatively easy to pull away the chalk from a vertical face about
600mm distant from the edge of the piles, it was much harder to remove chalk from
around the pile – each piece of intact chalk had to be broken up to be able to pull it
away from the relatively tightly interlocking framework of chalk lumps in which it
was held. In essence, the action of the auger forming the pile had been to push aside
chalk lumps (in addition to remoulding the chalk) thereby reducing the discontinuity
aperture of chalk immediately adjacent to the pile, so that it became a medium-dense
Grade B2–B3 chalk and in places Grade A2–A3.
! apart from closing the aperture of discontinuities, the auger caused very little
disturbance, eg to marl bands (see Figures A1.1, A1.3 and A1.4).
! a skin of remoulded chalk, between 1 and 5mm thick, but in places between 5 and
10 mm had been formed immediately adjacent to the shaft of the pile. Seven months
after the pile had been constructed, the chalk which formed this remoulded skin had
completely re-cemented.
CIRIA PR86 47
! in places, particularly where the surrounding chalk was Grade A and the skin of
remoulded chalk tended to be thinner, the surface of the pile shaft was quite smooth
having the appearance of the cortex on a flint; here the chalk could be pulled away
cleanly (see figures A1.1 and A1.2).
! in other places, either the remoulded chalk skin penetrated about 10–15mm into the
concrete in small pockets, or small (10–20 mm) lumps of chalk were embedded in
the surface of the concrete, so that the surface of the concrete was irregular and
pitted, which with the embedded chalk gave it a rough appearance (see figures A1.1
and A1.2).
! where the shaft of the pile was rough with adhering chalk, the bond between chalk
and concrete was so strong that, when struck with a geological pick, the pile
concrete would chip before the bond was broken.
! as far as can be judged, the shaft of the pile tended to be rough where the pile passed
through a band of flints, although this could apparently vary from pile to pile; piles
S1 and S2 (see Figures A1.8(a) and A1.5) were generally much rougher over a 5 m
exposed length than piles S3–S5, as illustrated in Figures A1.1– A1.4, although the
latter possessed rough areas.
! although it was not possible to measure variations in the pile diameter, it was
observed that the surface of the pile was undulating in a vertical elevation by up to
20 mm (see Figures A1.8(b) and A1.2 and A1.4).
It is therefore evident from this very limited exhumation and examination of 300–600 mm
wide sectors over 2 to 5m height of twelve piles in medium-dense chalk that the nature
of the pile/chalk interface can vary between very smooth and very rough, and that a
thin skin of remoulded chalk does form but that this re-cements with time. Examination
of the pile/chalk interface of exhumed CFA piles in chalks of varying density and Grade
is necessary to develop a better understanding.
48 CIRIA PR86
CIRIA PR86
Figure A1.1
Pile S6
49
50
Figure A1.2
Pile S5
CIRIA PR86
CIRIA PR86
Figure A1.3
Pile S4
51
52
Figure A1.4
Pile S3
CIRIA PR86
CIRIA PR86
Note
Figure A1.5
Pile S1
Pile shaft generally rougher than piles S3–S6 (figures A1.1–A1.4)
53
54
Figure A1.6
Pile N1
CIRIA PR86
CIRIA PR86
Figure A1.7
Pile N2
55
56
Figure A1.8 Details of exhumed CFA pile on A41 Aston Clinton bybass
CIRIA PR86
Appendix 2
SITE INVESTIGATION METHODS APPROPRIATE FOR CFA

PILING IN CHALK (AFTER SPINK, 2002B)
A2.1 SITE INVESTIGATION METHODS
The following table summarises the geological features of a chalk site that may affect
the performance of a CFA pile, and the suitability (or otherwise) of the common
intrusive site investigation techniques for investigating each type of feature.
Suitability of investigation by:

Geological
Feature Rotary cored Cable tool borehole with Window
CPT Trial pit
borehole U100’s and SPT’s sampling
Thickness and
Good, but Moderate, but
nature of Moderate Good Moderate
depth limited depth limited
overburden
Presence and
Good, but Moderate, but
nature of Moderate Moderate Moderate
depth limited depth limited
solution features
No, unless
Moderate if
Difficult, but good if piezocone or
Water table Good Moderate standpipe
piezometer installed push-in piezocone
installed
used
Moderate in low and medium
Good, if
Good if large diame- density and poorer grade
Chalk weathering logged in situ
ter high quality core chalk. Poor or not possible in No No
Grade but depth
obtained high and very high density or

limited
in very flinty chalk
Good in low and medium-
density chalk. Poor, or not
Chalk density/
possible to obtain suitable Good, but
hardness/ Good No No
samples in high -and very depth limited
strength
high-density or very flinty
chalk.
Good in low- and medium-
Chalk natural possible to obtain suitable Good, but
Good No or poor Poor
moisture content samples in high- and very depth limited
chalk
Good in low- and medium-

Moderate where
possible to obtain suitable Good, but
Chalk lithology Good No or poor flints are small but
samples in high- and very depth limited
depth limited
chalk
Good, if large Moderate where

diameter high quality flints are small and Good, but Moderate where
Nature and
core obtained, but Moderate infrequent. In very depth limited flints are small but
frequency of flints
difficult in very flinty flinty chalk CPT's depth limited
chalk may be
impossible
CIRIA PR86 57
A2.2 GEOLOGICAL FEATURES
A2.2.1 Overburden
The thickness and nature of the overburden is best determined by cable tool boring with
appropriate sampling and SPT testing. If the overburden is thin, then trial pits may be a
suitable alternative.
A2.2.2 Solution features
Solution features are problematic to locate. If pile testing is carried out on a site where
solution features are present the geology could vary markedly between adjacent piles,
and this could only be determined by ground investigation immediately adjacent to each
pile test.
A2.2.3 Chalk grade
There is sometimes a tendency to think that Grade and chalk density are related, or
even synonymous. They are not related, and vary independently. It is essential to
determine both to categorise the nature of the chalk.
The SPT and CPT tests both measure something that is a combination of the chalk
Grade, and the intact density/hardness/strength although it is considered that the intact
material properties tend to have the most influence.
A2.2.4 Chalk density/hardness/strength
Bowden et al (2002) consider that chalk density/hardness/strength, although different

properties are essentially all indicative of a common characteristic related to the
porosity of the chalk (other than for Grey (formerly Lower) Chalk). This characteristic
an be assessed by the following tests:
! intact dry density
! saturated moisture content

! porosity
! field index methods using by nail, hammer and breaking (Bowden et al, 2002)
! point load index
! unconfined compressive strength.
Intact dry density and saturated moisture content are determined by the same BS1377 test
and there is no point in quoting both; intact dry density should be the preferred parameter.
Of the above tests the intact dry density test is the easiest to do, it gives reliable
numbers, it can be made on irregular lumps of chalk and does not require high quality
core. But U100 samples may not be suitable for these determinations as they may cause
severe damage in good (Grade A) chalk, and in high and very high density chalks, the
disturbance may be so severe that it is not possible to obtain lumps big enough for
intact dry density testing. Also, in these materials the chalk fabric can be partially
collapsed in driven samples, and misleading density results may be obtained.
58 CIRIA PR86
A2.2.5 Natural moisture content
Natural moisture content determination, and hence degree of saturation, should be carried
out on all intact dry density test samples. The samples must be well wrapped to prevent
them sweating or drying out. The degree of saturation of chalk is normally close to 100
per cent, but this is not always the case. Bowden et al, (2002) indicate that the behaviour
of the chalk changes when the degree of saturation is less than about 90 per cent.
A2.2.6 Chalk lithology
In the context of this study the significant lithological variations are the increased clay
content in the Grey (formerly Lower) Chalk, and the recemented nature of the chalk
rocks and local hardgrounds.
A2.2.7 Flints
Flints have a significant influence on the success of the various site investigation
techniques. High quality coring becomes difficult in very flinty chalk, and any U100
samples that encounter flints are usually rendered useless. Flint content will probably
have an impact on the amount and nature of CFA piling overbreak and the remoulded
chalk/pile interface.
A2.3 SUGGESTED INVESTIGATION SEQUENCE
Where a ground investigation is to be associated with preliminary pile testing, an ideal

sequence would assist in extrapolating the pile test results across the site and so provide
correlations for use on other sites where preliminary pile tests have not been carried
out, as follows:
! Borehole to be within 5 m of the pile test location. Cable tool borehole through the
overburden with appropriate sampling and in-situ testing. In the chalk, rotary cored
borehole using triple tube core barrels producing 100 mm diameter core within a
disposable plastic liner. Diamond core bits. Flush either water polymer or air mist
polymer. If there are no cable tool boreholes on the site, carry out an SPT test at the
start of each core run ie, nominally at 1.5 m centres. Log the core in accordance
with Spink (2002a). Determine intact dry density and natural moisture content at 1 m
intervals.
! If the chalk is low or medium density and lacking in flints, also carry out CPT
sounding about 5 m from the pile test location.
! If the chalk is low or medium density and lacking in flints, also carry out a cable
tool borehole about 5 m from the test location. Boring to include U100s at 1.5 m
centres, with SPT’s through the base of each U100. All U100s to be extruded for
careful destructive logging in accordance with Spink (2002a) and the recovery of
lump samples for intact dry density and natural moisture content testing. All SPTs to
have the blows for each 75 mm increment recorded to identify flints.
If cost precludes rotary cored boreholes and the chalk is low or medium density and
lacking in flints, a suggested investigation is:
! Cable tool boring with U100s at 1.5 m centres, with SPTs through the base of each
U100. All U100s to be extruded for careful destructive logging in accordance with
Spink (2002a) and the recovery of lump samples for intact dry density and natural
moisture content testing. All SPTs to have the blows for each 75 mm increment
recorded to identify flints in accordance with BS5930 (1999)
CIRIA PR86 59
! CPT sounding.
Where the chalk is high or very high density and/or flinty, cable tool boring will only
give SPT blowcounts, and it is unlikely that it will be possible to obtain reliable Grades,
intact density tests or natural moisture content results. Similarly, it is unlikely that CPT
testing will achieve a useful depth.
CPT testing on its own, without visual observations, sampling and testing, is not
recommended.
60 CIRIA PR86
Appendix 3
A3.1 CFA PILES IN CHALK
PILE LOAD TEST INFORMATION CHECK LIST
Data from tests on CFA piles in chalk will continue to be collected at the University of
Southampton, where a database will be maintained. Data from such tests, including as
much of the above information as possible, should be sent to:
Professor CRI Clayton, University of Southampton, Highfield, Southampton SO17 1BJ
Would be
REQUIRED INFORMATION Must have
good
Site data
Site location "
Pile location "
Plan co-ordinates, elevation (GL) "
Site investigation
Borehole log for test pile or logs close to test pile including chalk Grade "
Borehole logs for the whole site including chalk Grade "
Ground water level/piezometer data "
Any in-situ test data (eg SPT, CPT) "
Measurements of intact dry density of chalk and possible variation with depth "
Photographs of chalk cores "
Any other relevant data (eg stiffness values of the chalk from plate load tests) "
Pile test data

Pile dimensions (nominal diameter and length) "
Total length of pile that is sleeved "
Total length of pile in overburden "
Level of surface of competent chalk "
Load displacement diagram/data for pile test (would prefer tests where there had been more
"
than 10 mm displacement or there is a marked change in the load settlement curve).
Method of load test "
Date the pile was constructed and date of load test "
Concrete mix (water and cement contents, type and proportion of filler) "
Any other relevant data (eg was the pile suspect for any reason?) "
Piling rig
Type of rig used. "
Maximum torque the rig can supply or torque against rpm graph. "
Any data from the rig instrumentation (ie amount of concrete, revolutions made by auger,
"
torque, etc).
Any other relevant data "
CIRIA PR86 61
62
CIRIA PR86

PR86 Shaft Friction of CFA Piles in Chalk PDF

Uploaded by

Copyright:

Available Formats

PR86 Shaft Friction of CFA Piles in Chalk 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

PR86 Shaft Friction of CFA Piles in Chalk PDF

Uploaded by

Copyright:

Available Formats

Project Report 86 London 2003

Shaft friction of CFA

Prepared under contract to CIRIA by:

J A Lord of Ove Arup and Partners

T Hayward of Stent Foundations Limited (formerly of the University of Southampton)

C R I Clayton of the University of Southampton

6 Storey's Gate, Westminster, London SW1P 3AU

The conclusions drawn from the analyses were that:

Shaft friction of cfa piles in chalk

Reader interest Classification

USER Civil, geotechnical and

Project funders The project was funded by:

Significant economies, through pile shortening, can be achieved by proving higher β

As the presence of dissolution features cannot be readily identified in the construction

3 Identification of chalk with a low unit shaft resistance . . . . . . . . . . . . . . . . 36

3.2 Standard penetration test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4 Pile testing and factors of safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Appendix 1 Examination of exposed shaft of CFA piles . . . . . . . . . . . . . . . . . . . . . . 47

penetration resistance chalk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Table 1.1 Records of CFA piles in CIRIA Report PR11 (1994) . . . . . . . . . . . . . . 15

Db diameter of pile base

Ds diameter of pile shaft

d depth below top of chalk in metres

Eb modulus of chalk at base of pile

Ec modulus of elasticity of concrete pile shaft

Ey ‘yield’ modulus of chalk

Fb factor of safety on base of pile

Fs factor of safety on shaft of pile

P load on pile head

Pa allowable capacity of pile

Pb load on pile base

Pbu ultimate base load

Ps total shaft load

Psu ultimate shaft load

qb stress on base of pile

qc cone resistance in MN/m²

qu ultimate bearing capacity of chalk

qy ‘yield’ stress of chalk

β effective stress pile friction coefficient

γ unit weight of soil

γconc unit weight of concrete

γd intact dry density of chalk

ρb settlement of base of pile

ρmax maximum settlement of pile

ρs elastic shortening of pile shaft

σ′v average vertical effective stress along pile shaft in chalk

τsf ultimate shaft resistance of pile

τsf average ultimate shaft resistance of pile in chalk

The satisfactory performance of piles in chalk was demonstrated at the Chalk

Chapters 3 and 8 are relevant to this present report.

A new classification for chalk was devised based upon:

For structured chalk:

Grade Discontinuity aperture

For Grades A, B and C, the discontinuity spacing is denoted by a numerical suffix.

1.2 SHAFT RESISTANCE OF BORED AND CFA PILES

A conclusion of CIRIA PR 11 Foundations in Chalk (Lord et al, 1994), which largely

τsf = 0.8σ ′v (subsequently expressed as τ = β σ ′v )

τsf = 0.45σ ′v (1 test)