599
Diagnosis of earth-fills and reliability-based design
Diagnostic de remblais de terre et conception basée sur la fiabilité
Nishimura S., Shuku T.
Graduate School of Environmental Science, Okayama University
Suzuki M.
Institute of Technology, Shimizu Co.
ABSTRACT: This research deals with the maintenance strategy of geotechnical structures such as earth-fill dams and river
embankments. To determine the soil parameters, the standard penetration test (SPT)
N
-values are frequently used. Firstly, a
statistical model for the
N
-values is determined from sounding test results. In this research, Swedish Weight Sounding (SWS) tests,
simpler than SPT, are employed. Secondly, an indicator simulation is conducted to interpolate the spatial distribution of the
N
-values,
and the results are utilized to find degraded areas inside the embankments and to maintain the embankments. Based on the statistical
model for the
N
-values, the shear strength parameter is derived through the empirical relationships, and then a reliability analysis of
the embankments is conducted considering the variability of the internal friction angle. Finally, the effect of improving the
embankments is discussed, comparing the calculated risks of the original state with the improved and restored state.
RÉSUMÉ : Cette recherche porte sur la stratégie de maintenance des structures géotechniques comme les barrages en remblais de
terre et les digues fluviales. Les valeurs
N
du test de pénétration standard (SPT) sont fréquemment utilisées pour déterminer les
paramètres du sol. Premièrement, le modèle statistique de
N
valeurs est déterminé à partir des résultats des essais de sondage. Dans
cette recherche, on utilise le sondage par poids suédois (SWS), plus simple que le SPT. Deuxièmement, une simulation indicatrice est
effectuée pour interpoler la distribution spatiale de
N
valeurs, et les résultats sont utilisés pour trouver les zones dégradées à l'intérieur
des remblais, pour l'entretien des remblais. Basés sur le modèle statistique de
N
valeurs, les paramètres de résistance au cisaillement
sont déduits des relations empiriques, ensuite, l'analyse de fiabilité des remblais est effectuée en tenant compte de la variabilité de
l'angle de frottement interne. Finalement, l'effet de l'amélioration du remblai est discuté, en comparant l’analyse de risque calculée à
partir de l’état initial et de l’état des remblais améliorés et restaurés.
KEYWORDS: earth-fill dam reliability-based design, indicator simulation, statistical model of
N
-value
1 INTRODUCTION
There are many earth-fill dams for farm ponds in Japan. Some
of them are getting old and decrepit, and therefore, have
weakened. Making a diagnosis of the earth-fills is important for
increasing their lifetime, and an investigation of the strength
inside the embankments is required for this task. In the present
research, firstly, the spatial distribution of the strength
parameters of decrepit earth-fills is discussed, and an
identification method for the distribution is proposed. Although
the strength of earth-fills is generally predicted from the
standard penetration test (SPT)
N
-values, Swedish Weight
Sounding (SWS) tests are employed in this research as a simpler
method of obtaining the spatial distribution of the
N
-values.
SWS tests are advantageous in that they make short interval
exams possible, because of their simplicity.
To mitigate disasters, improvement works are conducted on
the most decrepit earth-fill dams. Since there is a recent
demand for low-cost improvements, the development of a
design method for optimum improvement works at a low cost is
the final objective of this research. A reliability-based design
method is introduced here in response to this demand.
Generally, the identification of the spatial correlation of soil
parameters is difficult, since the usual sampling intervals are
greater than the spatial correlation. Therefore, sounding tests
are convenient for determining the correlation lengths. Tang
(1979) determined the spatial correlation of a ground by cone
penetration tests (CPT). Cafaro and Cherubini (1990) also
evaluated the spatial correlation with CPT results. Uzielli, et al.
(2005) considered several types of correlation functions for CPT
results. Firstly, statistical models for the
N
-values are
determined from the SWS test results. Secondly, the
relationship between the SPT and the SWS
N
-values is
modeled, including the transformation error term. The
N
-value
distributions derived from SWS are spatially interpolated with
the indicator simulation (Journel and Huijbregts 1978), which is
one of the geostatistical methods. The simulated spatial
distribution of the
N
-values can be used for the health
monitoring of the inside of an embankment. To evaluate the
risk to earth-fill dams, due to the earthquakes, the circular slip
surface (CSS) method is used as the stability analysis method
along with the soil-water coupling finite element method. The
finite element method is used to estimate the normal and the
shear stress values on the slip surfaces. In this study, the Monte
Carlo method (MCM) is combined with the CSS method to
obtain the probability of failure. The procedure for the CSS
method, combined with the MCM, has also been conducted by
Shinoda,
et al.
(2006) and Yoshida,
et al
. (2005).
The strength parameter, namely, internal friction angle
,
derived from SWS tests, is considered to be the probabilistic
variable in this research. Additionally, two transformation error
terms, namely, the error terms from the SWS N-value to the
SPT
N
-value, and from the
N
-value to the internal friction angle,
are introduced to the MCM. Finally, the risk to an earth-fill
dam is calculated from the costs that would be incurred due to
embankment failure and probability failure. In this study, the
effect of improving an embankment is evaluated as a reduction
in risk between the original and the improved states.
i gnosis of earth-fills and relia ility-ba ed design
Diagnostic de remblais de terre et conception basée sur la fiabilité
