Environmental Engineering Reference
In-Depth Information
Table 3.
ethiopian tuff. Properties of saturated cores and compacted specimens (standard Proctor optimum).
Wetting
under stress
of 100 kPa
Type of
sample
Peak
strength
Post peak
strength
Residual
strength
saturated
permeability
Dispersivity.
Pinhole
structure
saturated
core (soft
rock)
expansion
ε
v
= -1.6%
(from
Rh = 50%)
non-dispersive
nD-1
c′ =
176 kPa
φ′ =
44°
c′
=32 kPa
φ′
= 37°
c′
= 0
φ′
= 29°
3.5 × 10
-10
m/s
compacted
to optimum
standard
proctor
collapse
ε
v
= 0.4%
(from
S
r
= 0.80)
c′
= 36 kPa
φ′
= 43°
c′
= 36 kPa
φ′
= 43°
c′
= 0
φ′
= 29°
3.5 × 10
-10
m/s
non-dispersive
nD-1
This is a low value, which makes the compacted tuff
a suitable material to build impervious barriers.
in the pinhole dispersion test (colloidal erod-
ibility) performed on the compacted nB-4 sam-
ple water run through a 1 mm hole drilled in the
compacted sample for a maximum of a10 minute
period at increments of water head of 85, 285, 590
and 1560 mm. no effluent turbidity was detected
and the hole did not erode, indicating an erosion
resistant material (non-dispersive nD1) according
to the pinhole test
(
Figure. 16
)
.
strongly from point to point although the effects
of sampling (coring) disturbance cannot be mini-
mized. however, the presence of montmorillonite in
the fine fraction of the tuff forces to be cautious.
The tuff appears to be non-dispersive in the pin-
hole tests performed. some volcanic tuffs around
the world are known to be dispersive and erodible,
but in general these phenomena are present when
the plasticity of the material is low.
specimens of crushed tuff and compacted at
standard Proctor energy reached dry densities
similar to the “in situ” intact soft rock. it is then
interesting to compare the properties of both
materials. This is done in Table 3.
There are two main differences when the table is
examined. The first one is the high peak strength
of the intact material and its brittle behaviour.
This behaviour is interpreted as an effect of the
natural structure and cementation. interestingly,
immediately after the peak strength is reached the
intact specimen adopts strength values similar to
the compacted specimen at the same dry density.
The second difference concerns the volumetric
behaviour when wetted under equivalent stresses.
The intact material tends to expand (but meas-
ured strains are moderate) whereas the compacted
specimen collapses. collapse increases with con-
fining stress in this case (up to a vertical stress of
400 kPa), a typical behaviour of compacted clayey
soils. The interesting observation is that the smec-
tite content is not able to counteract the collapse
behaviour, which is explained by the significant
macroporosity of the compacted specimen.
The remaining properties examined here are
similar for the two types of material.
it appears that the compacted tuff may be con-
sidered for dam construction as a preliminary con-
clusion. its properties as foundation material seem
to be also adequate. a word of caution is necessary
because of the heterogeneity of these formations
and the presence of smectite.
3
DiscUssion anD conclUsions
The “in situ” tuff is a hard soil of high porosity
(void ratio approaching 1). a main mineral compo-
nent identified is montmorillonite. The remoulded
soil is identified as a high plasticity clay (ch)
(liquid limits of 55-75%). This implies a low per-
meability, which is a positive feature. Possible foun-
dation preferential paths for the water include the
contacts between different volcanic units and the
presence of discontinuities in rhyolitic and basalt
levels. however the tuff matrix is an impervious
mass (saturated permeability of 3.5 × 10
-10
m/s). its
stiffness is high and it should not pose settlement
problems. The presence of montmorillonite could
imply, a priori, reduced friction angles, especially
under conditions leading to remoulding or weath-
ering of the rock. however, the tests performed
reveal that the tuff is a frictional material. espe-
cially surprising is the high residual friction (29°)
which is probably a consequence of the amorphous
mineral content and its non-platy nature.
The intact material displays a marked brittle
behaviour. after peak states it maintains a signifi-
cant friction, however. The natural tuff exhibits a
significant heterogeneity. other tests, not discussed
here, show that strength parameters depend mark-
edly on the tuff density. Tuff density seems to vary
