Managing risk and achieving reliable geotechnical designs using Eurocode 7 - Risk and Reliability in Geotechnical Engineering - page 402

SPT blow count

0 20 40 60 80 100

10

5

0

-5

-10

-15

-20

-25

-30

-35

Figure 10.2 Results of Standard Penetration Tests on London and Lambeth clays, with engineers'
interpretations of the characteristic value. (After Bond A. and Harris A. 2008. Decoding
Eurocode 7 , Taylor & Francis, London.)

selected by the geotechnical engineers are presented in Figure 10.2 and show a wide spread,
which indicates that geotechnical engineers are not good at selecting characteristic values,
particularly when the available data are scattered. This demonstrates that there is a need for
more specific guidance in Eurocode 7 on the selection of characteristic values and hence it is
planned to include some specific guidance on this in the forthcoming revision of Eurocode
7, due to be published in 2020.

10.4.3.2 Aleatory variability and epistemic uncertainty
As noted in the previous section, one of the requirements when selecting the characteristic
value of a parameter in designs to Eurocode 7 is to take account of the variability of the
measured parameter. Manufactured materials prepared under controlled conditions have
properties that are random variables whose inherent variability can be readily represented
by a particular statistical distribution. Since a large number of tests can be carried out on
such materials, the form of the distribution can be identified and the parameters to describe
the distribution, for example, the mean value and COV in the case of a Gaussian distribu-
tion, can be determined. Such materials are termed aleatory and are described objectively
using experimentally determined parameters. Selection of the characteristic value of such
a material, defined as a specified fractile, normally 5%, of the assumed probability distri-
bution curve, is straightforward. Aleatory variability is an inherent feature of nominally
homogeneous manufactured materials. It is related to the inherent variability of a material
and is not reduced by carrying out more tests. However, carrying out more tests, as in the
case of an unlimited test series, will enable the variability to be known more precisely.