Environmental Engineering Reference
In-Depth Information
The previously referred 'sandwich' of massive
flows h and i, in which the diversion tunnels were
driven, are likely to have attracted higher levels of
horizontal stress than their neighbours, and this
can be indirectly assessed by the relative magni-
tudes of deformation moduli that can be estimated
from the following equation, again for near-surface
(nominal 25 m depth) and low porosity.
We were clearly mostly within the 'stress-slabbing'
sRF class ( Table 1 ) of 5-50, Barton and Grimstad,
(1994), which implies a σ c 1 ratio of 5 to 3, or an
'elastic behaviour' tangential stress ratio assump-
tion ( σ θ c ) of 0.5 to 0.66, i.e. a tangential stress
high enough to cause failure with rock strength
scale effects considered.
in the case considered here, the major princi-
pal stress is of course σ h and the above ratios are
suggesting that its value might be in the approxi-
mate range 47 to 56 MPa, when using the 140
and 280 MPa uniaxial strengths in the logical way
in relation to the above strength/stress ratios of
approximately 5 to 3.
Measurements performed at the site with an
older lnec sTT (stress tube tensor) method were
inconsistent, but maximum stresses of 29, 43 and
54 MPa were recorded, and, significantly, the core
removed from the 9 m deep holes above the over-
coring sites, showed 'disking', which is a sure sign
of strong stress anisotropy and large magnitude.
There was also some limited core-disking in deeper
parts of two investigation boreholes.
an alternative way of back-calculating the
possible horizontal stress level is to use the set
of empirical 'depth-of-failure' data assembled
in Figure 14. With depths of failure as seen in
Figure 15 in the range 2 to 3 m for an average tun-
nel 'radius' of about 8 m, we see in Figure 14 that
ratios of σ max/σ c of about 0.6 to 0.7 are implied
when D f /a is in the range of (8 + 2 or 3 m)/8 = 1.25
(
)
13
/
e
=
10
Q
GPa
(3)
mass
c
The estimated contrasts in rock mass deforma-
tion moduli were perhaps in the range 17 to 37 GPa for
flows G and J, and 31 to 74 GPa for flows h and i, in
fact roughly a doubling of moduli due to the more
massive nature of the central, and eventually very
troublesome basalt flows. With greater horizontal
stress in the h and i flows, an anisotropic distribu-
tion of moduli would probably have been in opera-
tion, but this possibility has been ignored in the
simple treatment that follows.
4
Back-calcUlaTion oF PossiBle
sTRess leVels
We can first address the magnitude of the defor-
mations actually recorded at up to twenty meas-
urement locations along each of the five diversion
tunnels. The convergences were plotted by sadly
departed colleague nelson infanti, in the approxi-
mate log 10 Q/(span or height) versus log 10 (con-
vergence) format of Barton et al., (1994). even
at the top heading stage, the deformations, which
ranged from 0.5 to 13 mm, were mostly higher
than expected from the central empirical trend of
numerous data:
Table 1. extract from Q-system sRF, concerning stress-
failure of massive rock. Barton and Grimstad (1994).
sPan or heiGhT
Q
∆(
mm
)
(4)
in the case of TD-5, ten of the twenty instru-
ment locations that were monitored again after
benching down to the full 17 m height, showed
magnitudes of convergence at the triangular moni-
toring stations that ranged from 13 to 50 mm, with
a median value of 22 mm, and a mean of 25 mm.
Back-calculation according to equation 4 sug-
gested much lower 'stressed' Q-values, 20 mm
deformation implying Q ≈ 0.8, and 50 mm imply-
ing Q ≈ 0.3. so characterization prior to tunnel
excavation, was suggesting Q-values for the mas-
sive h and i flows of the order of 15 to 200, while
classification for tunnel design was, through back-
calculation from deformations, suggesting Q-values
in the approximate range of 0.3 to 1.5.
Figure 14. empirical data for stress-induced depths of
failure in relation to stress/strength ratios. Martin et al.,
(2002).
 
Search WWH ::




Custom Search