1435
Liquefaction impact revisited
L'impact de la liquéfaction revisité
Barends F.B.J.
1, 2
, Meijers P.
1
, Schenkeveld F.M.
1
, Weijers J.B.A.
2, 3
1
Deltares,
2
TU Delft and
3
Rijkswaterstaat, The Netherlands
ABSTRACT: Two aspects of liquefaction of carefully prepared partly loosely packed sand are tested: the intensity of a liquefaction
trigger and the extent of induced excess pore water pressures when partly liquefaction occurs. The results of several 1D and 3D tests
are shown. It appears that hardly any liquefaction happens when the trigger intensity is small. When yet partly liquefaction occurs,
significant excess pore pressures arise also in the non-liquefied surroundings. These excess pore pressures slowly diminish during the
sedimentation of the liquefied mass. The related densification itself, the porosity decrease, is however so small that the sand remains
loose and the likelihood of a new liquefaction to occur due to subsequent new triggers remains. So, if the intensity of a trigger is small
and the corresponding liquefied zone is limited, rigorous measures at great expenses are not required to guarantee e.g. stability of
buildings and dikes. However, in case a trigger is intense and liquefaction zone is large, its destructive impact can be far reaching
particularly due to the related excess pore pressures in a much wider area.
RÉSUMÉ : Deux aspects de la liquéfaction des sables partiellement lâches sont testés: l’intensité du facteur déclenchant d’une
liquéfaction et l’étendue des surpressions d’eau induites lorsqu’une liquéfaction partielle se produit. Les résultats de plusieurs essais
1D et 3D sont présentés. Il semble que la liquéfaction est quasi inexistante lorsque l’intensité du facteur déclenchant est faible.
Lorsque cependant une liquéfaction partielle se produit, d'importantes surpressions interstitielles apparaissent aussi dans les environs
non liquéfiés. Ces excès de surpressions interstitielles diminuent relativement lentement au cours de la sédimentation de la masse
liquéfiée. La densification elle-même, et la diminution de la porosité sont cependant si petites que le sable reste lâche et la probabilité
d'une nouvelle liquéfaction demeure. Ainsi, si l’intensité du facteur déclenchant est petite et la zone de liquéfaction correspondante est
limités, les mesures rigoureuses et coûteuses ne sont pas nécessaires pour garantir par exemple la stabilité des bâtiments et des digues.
Cependent, dans le cas où le facteur déclenchant est fort et la zone de liquéfaction est grande, son impact destructeur peut être de
grande envergure en particulier en raison des excès de surpressions interstitielles dans une zone bien plus large.
KEYWORDS: sand, mud, density, liquefaction, diffusion, sedimentation, trigger, pore pressure.
1 LIQUEFACTION.
Sand deposits in lowlands and delta areas are usually loosely
packed. Structures built on top may suffer from weakened
support if these sands liquefy. The liquefaction of loosely
packed sand is a research topic for many decades and several
methods and models are developed. A crucial parameter is the
local density (porosity or void ratio). The sensitivity of granular
material becoming liquefied is expressed by the so-called
liquefaction potential, which can be determined by special
laboratory tests on samples from site at various manufactured
densities. This counts for dynamic liquefaction (Barends &
Ruygrok 1997) and for static liquefaction (Stoutjesdijk, de
Groot & Lindenberg 1998). Laboratory results should then be
calibrated with the in-situ characteristics to determine the local
likelihood of liquefaction. Unfortunately, local characteristics
are difficult to measure.
Under a critical loading, e.g. a trigger caused by dynamic
shaking (Ishihara 1993) or a static slope slide (Stoutjesdijk e.a.
1998; de Groot e.a. 2006), excess pore pressures will arise and
when they reach the actual effective stress level, the granular
structure changes into a mud, the state of liquefaction. The
liquefied zone depends on the intensity of the trigger. Next, the
sand restructures following a sedimentation, characterized by
non-linear dispersion and consolidation (Pane & Schiffman
1985). Particularly, fine loose sands are sensitive to this process.
When sands are densely packed, negative pore pressures may
arise. When the soil liquefies, it behaves like a heavy fluid that
induces excess pore pressures into the surrounding soil, which
thus may cause shear strength reduction in a much larger area.
Some 1D tests and 3D tests have been performed and elaborated
to investigate the effect of the intensity of the liquefaction
trigger and the extent of induced excess pore pressures in the
surroundings.
2 EXCESS PORE PRESSURES AT LIQUEFACTION.
2.1
The process and effect of mud sedimentation
In the laboratory, 1D sedimentation tests have been
performed on fine saturated sand with some silt (
d
10
~ 60
),
taken from lake IJssel in the Netherlands. A tube (51 cm high, 8
cm diameter), shown in Figure 1a, is completely filled, covered
and rotated 180 degrees, and put at rest in a vertical position
while quickly uncovered, in order to simulate dumping this type
of sand under water. During the subsequent sedimentation the
actual pore water pressure is measured at two positions, at the
bottom and half way, by sensors recordings sampled at 200 Hz.
As observed, during the sedimentation, segregation of water and
fines occurs. In this case, in about 7 minutes a loose sand
column of 45 cm high is formed with on top a layer of expelled
water (height 3 cm) and a layer of silt (height 2.5 cm).
