Documentos de Académico
Documentos de Profesional
Documentos de Cultura
ABSTRACT: Geotechnical engineers must be competent. They must be able to construct a geotechnical mod-
el, use total and effective stress analyses correctly, carry out basic analyses, determine design parameters and
complete safe and economic designs. They acquire these skills through education, training and experience.
This paper considers some of the basic things which geotechnical engineers should be able to do and how they
acquire these skills as children, at school, at university and at work. The paper also considers some of the ob-
stacles to effective education and training of geotechnical engineers.
1 INTRODUCTION
Civil engineers have to design and supervise con- 2 WHAT SHOULD GEOTECHNICAL
struction of the infrastructure necessary for modern ENGINEERS BE ABLE TO DO?
societies to operate. Geotechnical engineers design
and supervise construction of works in the ground 2.1 Construct a geotechnical model
and using natural geological materials. Their skills A common reason for of claims in construction con-
are both in engineering and in geology. tracts is “unforeseen ground conditions”. But the
Engineers are required to be competent. This question arises whether the ground conditions were
means they can design and construct works which really unforeseeable or whether the ground investi-
are both safe and economical, fit for purpose, aes- gation was inadequate.
thetically pleasing and fulfill a number of other re- Geotechnical engineers should be able to con-
quirements. Competence is acquired through educa- struct a reasonable geotechnical model. A geotech-
tion, training and experience at school, at university nical model is the basis of an interpretive report. It is
and at work. constructed from all the available investigation data
However geotechnical engineering is a risky contained in the factual report. This is the framework
business and there is much that can and does go upon which the design will be based.
wrong. It is said that the single most common cause The geotechnical model consists of: a list of the
of failure in construction, including delays and addi- principal strata; cross-sections of the site showing
tional costs is in the ground. Moreover, most ge- the locations of these strata; a list or table giving
otechnical structures move considerably less than their engineering properties; a summary of the
predicted and so were over-designed - over-design is groundwater conditions. Sometimes the geotechnical
a form of failure. The relatively high incidence of model is represented as a three-dimensional block
failures in geotechnical engineering means that ge- diagram (Fookes, 1997).
otechnical engineers are not as competent as they Because the ground conditions have to be interpo-
should be. lated between observation, usually from boreholes,
In this paper I will consider the basic competen- there will be uncertainties and these should be clear-
cies which geotechnical engineers should have – ly marked.
what they should be able to do safely and economi- Soils and rocks arrived in the ground by natural
cally - and how they acquire these competencies. I geological processes and so the geological model
will discuss the purposes of education and training must be geologically possible. The interpretive re-
and how these should be divided between school, port should include a geological history of the
university and employers. ground. This should include descriptions of the geo-
Currently the procedures for educating and train- logical processes which operated when all the strata
ing geotechnical engineers differ in detail across the in the model were deposited and what happened af-
European Community and World-wide. Nevertheless terwards. Sometimes geotechnical models are shown
the end result should be the same everywhere and in to be incorrect because there is no known process
this paper I will consider general principles rather which would have resulted in the sequence of strata
than specific practices. in the model or the groundwater conditions are not
hydraulically possible.
My students at City University had a ground
model project to do. They were given Figure 1(a),
some basic walk-over observations, some borehole engineers must have a clear knowledge of it. They
logs and some basic laboratory test results and they must understand the difference between drained
were asked to construct the geotechnical model for loading, undrained loading and consolidation. They
the site. must know when to carry out analyses in terms of to-
tal stress and when in terms of effective stress.
2.2 Correctly distinguish between total and 2.3 Carry out basic analyses by hand.
effective stress
Nowadays many geotechnical engineering analyses
The principle of effective stress is absolutely funda- are done using routine methods, spreadsheets and
mental to geotechnical engineering and geotechnical
computer programs. These analyses must always be on their mineralogy and history so these must be
checked by simple hand calculations. compatible. For example there are well defined rela-
There are some basic formulae which geotech- tionships between undrained strength and liquidity
nical engineers should be able to derive from first index and between large strain (critical state) friction
principles. The ones which I think are important are angle and grain characteristics.
as follows. (The analyses are given in Atkinson, As discussed in section 2.2 above geotechnical
2007.) engineers must be able to distinguish between
The long term limiting angle i of a slope in which drained construction which requires effective stress
there is steady state seepage parallel with the slope is analyses with effective stress parameters and un-
1 drained construction which requires total stress anal-
i ' (1) yses and total stress parameters. Undrained construc-
2
tion, either loading a foundation or unloading by
where ’ is the effective stress friction angle. This
excavation, will be followed by consolidation. Un-
should be derived using a polygon of forces.
drained loading will be followed by increase of
The ultimate bearing capacity qc of a long founda- strength and stiffness while unloading will be fol-
tion at the surface of soil which is undrained is lowed by reduction in strength and stiffness. Ge-
q 2 s (2) otechnical engineers must understand these process-
c u
where su is the undrained strength of the soil. This es clearly.
should be derived using simple upper and lower A geotechnical design must satisfy either an ulti-
bound calculations. mate limit state or a serviceability limit state or both.
The total horizontal stress on the active side of a To determine the ultimate limit sate analyses require
input of a soil strength together with a factor of safe-
smooth wall supporting drained soil is
ty to ensure that the ultimate state is not reached. To
K ( u ) u (3)
h a v determine a serviceability limit state analyses can ei-
where u is the pore pressure and ther apply a load factor to an ultimate limit state or
2 ' determine movements from soil stiffnesses.
K tan (4) Some of the basic analyses in geotechnical engi-
a 4 2
neering are in equations (1) to (6). These require in-
and ’ is the effective stress friction angle. This put of basic soil parameters for strength, stiffness
should also be derived using simple upper and lower and permeability (or coefficient of consolidation).
bound calculations. Soils have peak, critical state and residual
The consolidation settlement of a wide founda- strengths for both drained and undrained loading and
tion is the same as the settlement when the soil is they have stiffnesses which vary substantially with
drained and is both stress and strain (Atkinson, 2000). Parameters
1 measured in laboratory and in situ tests will have a
Dq (5)
M' natural scatter often represented by a distribution
where M’ is the one-dimensional compression mod- curve characterised by mean, moderately conserva-
ulus (= 1/mv). D is the depth of the compressible soil tive and worst credible values.
and q is the bearing pressure of the foundation. For slope stability analyses, factors of safety for
The time for one-dimensional consolidation to be ultimate limit states are usually in the region of 1.2
complete is approximately while for foundations load factors for serviceability
H2 limit states are in the region of 3.0.
t (6) Geotechnical engineers should be able to select
c
v strength and stiffness parameters and factors appro-
where H is the drainage path length and cv is the co- priate to the construction and to the limit state being
efficient of consolidation. This can be derived from considered. The issues are discussed by Atkinson
simple definitions of terms taking Tv = 1 at Ut = 1. (2007).
The point here is not so much that geotechnical
engineers can or cannot derive these formulae, they
are given in text-books, but whether they can do 2.5 Design safe and economic groundworks.
simple analyses by hand to check more complex Geotechnical design requires a good geotechnical
analyses. model and correct use of analyses soil parameters
and factors. It also requires consideration of how the
works will be built – the construction processes. The
2.4 Determine parameters for analyses
final design should be safe but also economical.
Geotechnical analyses require input of soil parame- If things go wrong - construction does not go ac-
ters and factors. cording to plan or there are unexpected events or
The ground engineering properties of soils and movements are too large, or much smaller than ex-
rocks (strength, stiffness and permeability) depend pected – it means that the geotechnical engineers got
something wrong. They lacked competence in a cru- drained shearing of dense sand. The basic principles
cial component of the design. learned here apply also to large scale construction. It
Notice that there is nothing here about correct use is a pity that these principles have often been forgot-
of standards or codes of practice. Competence re- ten by the time civil engineers graduate from univer-
quires the engineer to get the basics right; it does not sity.
require the engineer to follow a code or standard.