The effect of targeted field investigation on the reliability of axially loaded piles: A random field approach
Journal
Geosciences
Date Issued
May 2020
DOI
10.3390/geosciences10050160
Abstract
This work deals with the e ect of targeted field investigation on the reliability of axially
loaded piles, aiming at an optimal serviceability and ultimate limit state design. This is done in
a Random Finite Element Method (RFEM) framework properly considering sampling in the analysis;
the RFEM method combines finite element analysis with the random field theory. In this respect,
the freely available program called RPILE1D has been modified by the authors as to consider sampling
of both soil and pile properties. In each RFEM realization, failure is considered to have occurred when
the calculated shaft resistance of pile considering spatially uniform properties (average of sampled
values from the soil and pile random fields) is greater than the respective “actual” one considering
spatially random properties for both soil and pile. The necessary numerical demonstration of the
proposed methodology is done by considering two sampling strategies: a) sampling from a single
point and b) sampling from a domain, both along the pile, whilst the various parameters governing
the statistical uncertainty of the problem are examined; 5383 di erent cases were considered in total.
As shown, a targeted field investigation may minimize or even eliminate the statistical error inserted
in the design. The error is quantified by the di erence in the probability of failure comparing di erent
sampling scenarios. Another main finding is that the optimal horizontal sampling location occurs
where the pile is going to be constructed. In addition, it is shown that the benefit of a targeted field
investigation is much greater than the benefit gained using characteristic soil property values.
loaded piles, aiming at an optimal serviceability and ultimate limit state design. This is done in
a Random Finite Element Method (RFEM) framework properly considering sampling in the analysis;
the RFEM method combines finite element analysis with the random field theory. In this respect,
the freely available program called RPILE1D has been modified by the authors as to consider sampling
of both soil and pile properties. In each RFEM realization, failure is considered to have occurred when
the calculated shaft resistance of pile considering spatially uniform properties (average of sampled
values from the soil and pile random fields) is greater than the respective “actual” one considering
spatially random properties for both soil and pile. The necessary numerical demonstration of the
proposed methodology is done by considering two sampling strategies: a) sampling from a single
point and b) sampling from a domain, both along the pile, whilst the various parameters governing
the statistical uncertainty of the problem are examined; 5383 di erent cases were considered in total.
As shown, a targeted field investigation may minimize or even eliminate the statistical error inserted
in the design. The error is quantified by the di erence in the probability of failure comparing di erent
sampling scenarios. Another main finding is that the optimal horizontal sampling location occurs
where the pile is going to be constructed. In addition, it is shown that the benefit of a targeted field
investigation is much greater than the benefit gained using characteristic soil property values.
File(s)![Thumbnail Image]()
Name
geosciences-10-00160-v2.pdf
Size
7.2 MB
Format
Adobe PDF
Checksum (MD5)
9c8dea68ec5c7b4fd683b1b72d59cff8