6
23 times offers acceptable results. Thus, the intercon-
nected fracture zone is considered 53-m thick. Following
Khademian et al. (2021), the friction of the subvertical
DFN fractures and bedding planes are assumed 60% of
the internal friction of rock in each stratum, except for the
strata within the interconnected fracture zone with a con-
stant friction angle of 10 degrees.
The initial aperture of the fractures in the model needs
to be defined before solving the model under in-situ stresses.
It is concluded that an initial aperture of 0.4 to 0.5 mm for
bedding planes and sub-vertical fractures gives a relatively
good agreement between the pre-mining values of perme-
ability from site measurements and models.
Gob Model
In longwall mining, formation of gob material due to the
roof caving prevents overstressing the chain pillars. Gob
compaction degree and its residual voids affect the overall
deformation and stress distribution around the longwall. A
gob model in the roof of longwall panels is installed during
mining each panel in the model.
Simulation of gob materials in the continuum model-
ing environment such as FLAC3D can be done with an
equivalent element method (Esterhuizen, et al. 2010). In
this approach, the caving zone in the model is replaced with
zones whose cohesion is reduced during the model run.
However, this approach in 3DEC may lead to an exces-
sive overlap of blocks within the equivalent element zones
because the DFN-generated blocks can experience large
displacement, overlapping the equivalent element zones.
To avoid the excess overlap, Khademian et al. (2022)
simulated the effects of the gob materials instead of
explicitly modeling the caving process. In this approach,
counteracting loads are applied onto the mine roof in
the model and are gradually reduced until the mine roof
is fully relaxed. At the same time, incremental stresses are
applied as gob effects with a=5 MPa and a maximum strain
of 44%. Details of the approach can be found elsewhere
(Khademian et al., 2022).
Once the model reaches equilibrium, the surface sub-
sidence and pillar stresses are recorded. Figure 7 shows the
super-critical surface subsidence recorded in the model
with a maximum value of 1.5 m, about 70% of the mining
height, that is expected from a supercritical longwall panel.
Figure 8(a) and (b) show the modeled pillar vertical stress
after the first and second panels are completed, respectively.
In Figure 8, the in-situ vertical stress at the mine level is
about 8.9 MPa. The average vertical stress within the abut-
ment pillar (between the second and third panels) increases
to 13 MPa and to 24 MPa after the second and third
panel mining. Figure 8 also shows the gob stress after min-
ing each panel. With the simulated gob effects, the stress
within the roof gradually approaches the in-situ stress level
of 8.9 MPa, which reflects the expected gob effects after
mining the panels.
Aperture Variation
During mining of the two panels, apertures of all fractures
are recorded in the model. Figure 9 shows the aperture of
the subvertical fractures evolving with the progress of min-
ing in mm. The heat maps are based on a discretization of
the 3DEC model with a 17×17 m squares from the roof to
the surface of the mine. Values assigned to each square are
the average of the aperture of fractures falling within the
defined area of the square. The aperture values in Figure 9
correspond to the first, second, and third panel mining.
All fractures are initialized with an aperture of 0.3 mm.
In-situ stresses expand the range of apertures from 0.3 mm
to between 0.005 and 10 mm. This range further expands
as mining continues. For visualization purposes the range
of the aperture legends are limited between 0.25 mm and
2 mm. The subvertical fractures within the predefined set
of DFNs are activated due to the mining-induced deforma-
tion and stresses. The activation in shear or tensile modes
increases their aperture from the initial values.
Figure 8. Vertical stress at the mining horizon after mining
the second panel (a) and after mining the third panel (b)
Previous Page Next Page