Sie sind auf Seite 1von 7

Numerical Modeling as a Tool to Refine

Bench Designs in Open Pits


RODRIGO SILVA GUZMN a,1 and SERGIO YEZ SAAVEDRA a,1
a
Itasca S.A., Santiago, Chile
Abstract. The correct design of open pit mines depends, among other things, on the slope design at
different scales, which must deal with the geotechnical stability and the safety of personnel and equipment. In
this regard, the bench scale design procedures that rely on prediction of the spill length of failed wedges
currently lack of a robust definition of the conversion from a solid volume of rock forming a wedge into a
pile of broken material. At present, the assumption of a bulked density is the main consideration for
estimating the maximum spill length of the broken material. This approach, however, disregards the inherent
complexity of rock masses and the likely influence of the internal discontinuities and/or defects into the
actual potential for disintegration of a failed volume.
The work presented in this article addresses a first approach to this issue by proposing the use of
discontinuum numerical modeling codes (such as Itascas 3DEC) to estimate and define rules for the physical
behavior of the spilled material once the kinematic failure mechanisms (planar in this study) are activated.
The main objective of this study is an assessment of the final configuration of the spilled material, its volume,
the final geometry of the slope, the spill length and establishing possible correlations of these variables with
the rock mass characteristics. The study is applied to a limited number of rock masses, ranging from a very
isotropic material with many joint sets within the failed wedge, to a very competent rock mass with only a
few persistent discontinuities.

Keywords. Numerical Modeling, Bench Design, Open Pit, 3DEC, Spill Length, Angle of Repose.

1. Introduction
The code KATS (Kinematic Analysis Tool for Slopes) is a tool developed by Itasca S.A.
(the Santiago office of Itasca International Inc.), aimed at assessing instabilities caused
by day-lighting wedges and planar failures formed when different structural sets
interact with the orientation of a given slope [1,2].
The aim of this work is to check the geometrical assumptions adopted in KATS to
calculate the spill length, which is "the extent of the spill that is originated by the failed
(and broken) volume" [2] and it is frequently used as a reference to determine the berm
width necessary for containment and therefore safety of the operation. A threedimensional distinct element model (using the code 3DEC) [3] was developed to
estimate the extent and pile shape of the failed material taking into account the rock
mass strength and shape of the particles of material inside the planar failure.

Corresponding Author. E-mail: rodrigo.silva@itasca.cl and sergio.yanez@itasca.cl

2. Planar Cases Defined in KATS


KATS defines three alternative configurations for the failed volume based on relations
between several geometrical variables. Figure 1 shows those cases, which are described
below:

Case 1: There is no contact between the pile (blue) and the wedge sector (red)
after the failure, see Figure 1a.

Case 2: There is contact between the pile (blue) and the wedge sector (red)
after the failure and the spill angle (r) is less than the dip structure (f), see
Figure 1b.

Case 3: Equal to Case 2 but the spill angle (r) is greater or equal than the dip
structure (f), see Figure 1c.

a)

b)

c)
Figure 1. Planar Cases. a) Case 1, b) Case 2 and c) Case 3.

These cases assume that the material is deposited according to a predefined spill
angle (r) which is assumed equal to the repose angle of the material, and the spill
length calculation is developed for each case along with the variables defining the spill
length (LD). The spill length is calculated by following geometrical considerations as
illustrated in Figure 1. Details about this are provided in [1,2], which describes the
methodology behind KATS. As an example, the spill length for Case 1 is defined below.

(1)

(2)

3. Construction of the Numerical Model and Analysis Considerations


A three-dimensional distinct element model (3DEC) was developed to estimate the
spill lengths for several geometrical combinations, in order to compare with the planar
spill length formulation used by KATS in terms of the 3 analysis cases illustrated in
Figure 1. The model parameters are shown in Table 1; and they are described in Figure
1. In all cases the planar structure is assumed to be cohesionless; and since the friction
angle () is lower than the structure orientation angle (f), the factor of safety (FoS) is
always less than 1.0.
Table 1. Geometrical and strength parameters used to build the 3DEC models
Case
1
2
3

Hb (m)
15
15
15

b ()
70
70
70

B (m)
8
8
15

f ()
50
50
32

()
25
25
25

Hi (m)
6
12
9

In 3DEC, the structures are represented by discontinuities; while the rock mass is
described by the surrounding block behavior. In this study, the blocks are assumed to
be rigid and so the rock mass deformation is ignored. On the other hand, the
discontinuities (contacts) were modeled with a Mohr-Coulomb constitutive model. The
contacts not only represent the planar structure, but they are also used to represent the
fragmented rock of the failed rock mass material (failed wedge in Figure 1). This
fragmented material is described by small blocks interacting by the contacts. These
contacts have zero cohesion and an initial 35 friction. Sensitivity studies related to the
block shape and size, along with contact friction angle describing the fragmented rock,
are presented in the subsequent sections.

4. Results and Discussion


4.1. Evaluation of the pile shape
The 3DEC models simulate the behavior of the failed wedge, by capturing the
fragmentation of the failed rock mass located on top of the planar structure. The
resulting spill as simulated by 3DEC is shown in Figure 2 for all cases. It is worth
noting the close resemblance between the pile shapes representing the spills, and the
ones used by KATS (see Figure 1) to derive the spill lengths.

a)

c)

b)

Figure 2. 3DEC Model. a) Case 1, b) Case 2 and c) Case 3.

4.1. Spill Length/ Spill Angle Comparisons- Base Case Scenario


The spill angle in KATS represents a predefined angle given by the repose angle of the
material, which was equal to 38 in this study. On the other hand, in 3DEC the spill
angle represents a modelling result, depending on the fragmented rock description. The
Base Case (BC) scenario in 3DEC considered cubes of 0.5 m size, and 35 friction for
the contacts within the failed wedge. Figure 3 compares the results between KATS and
3DEC, showing a reasonable match for the spill length but larger discrepancies for the
spill angles, which varies between 33 and 38 in the 3DEC models. Note that Case 1
gives identical results in KATS and 3DEC, corresponding to a situation where all the
fragmented material is moved away from the plane of the structure, while in the other
cases, the spill was partially lying on that plane.
The 3DEC results show that, apparently, the spill angle is an output of the
modelling rather than an independent parameter. However, based on the lower values
of spill angle recorded in the sensitivity analysis, this is probably a numerical artifact.
Further analyses about this parameter are discussed in the next section.
Spill length KATS

Spill angle 3DEC

11.0

Spill Length (m)

10

9.0
8.0

39
38

10.0

37
36
35

6
4.5

34

4.5

33

Spill angle ()

spill length 3DEC


12

32
2
31
-

30
Case 1

Case 2

Case 3

Figure 3. Spill length and Spill angle comparison (Base Case Scenario).

4.2. Spill angle Comparison- Sensitivity Analysis


Figure 4a shows a relationship between repose angle, shape and size of the
cohesionless particles that is used in hydraulic engineering [4]. The typical repose
angle for particle size greater than 4 inches (0.1 m) is around 38 to 40 regardless of
its shape and it is in accordance with the KATS assumption of 38. As an example,
Figure 4b shows results of spill angle based on Case 1 3DEC analysis using spheres
instead of cubes for blocks and varying contact friction angle. Although the trend is

correct (more rounded particles lead to a flatter angle of repose), there is a large effect
in spill angle when spheres are used, especially for lower friction angles.
45

a)

b)
Spill angle ()

40

35

Case 1 cube-0.5m
30

Case 1 sphere-0.5m

25

20
=5

=15

=25

=35

=45

friction angle between blocks ()

Figure 4. a) Relationship between repose angle and size and shape of the particles [3] and b) 3DEC analysis
using spheres and cubes for blocks with different friction angles contacts.

4.3. Spill Length Comparison- Sensitivity Analysis


If we consider a cohesionless material within the planar wedge, the change in friction
angle between the blocks (surface quality) and the resizing (block interlocking)
indirectly we are talking about rock mass quality. Several combinations of size and
friction angle were included in the 3DEC analysis in order to seek for some
correlations between Spill Length and rock mass quality. The results obtained based on
60 different 3DEC models (20 per case) do not show any correlation between rock
mass quality and spill length. However, Figure 5 shows a good match between average
spill length (3DEC) and values obtained from KATS.
12
11

Case 3
3DEC-Average= 11 m
3DEC-Std.Dev= 1.8 m
KATS= 11 m

Spill Length - KATS (m)

10
9
8

KATS=3DEC

Case 2
3DEC-Average= 9.0 m
3DEC-Std.Dev= 1.1 m
KATS= 8.0 m

case 1

case 2
case 3

Case 1
3DEC-Average= 4.5 m
3DEC-Std.Dev= 0.5 m
KATS= 4.5 m

5
4
2

10

12

14

16

18

Spill Length - 3DEC (m)

Figure 5. Statistics of spill length results (20 3DEC models per case).

5. Conclusions
Several 3DEC models were implemented in order to validate geometrical assumptions
used in the Itasca code KATS in a deterministic approach. The results obtained are
encouraging because the average spill lengths calculated in the numerical models are
significantly close to the KATS results, especially when all the fragmented material is
moved away from the plane of the structure (Case 1) and also the sensitivity analyses
show larger standard deviation for cases where material is partially lying on planar
structure.
As expected, analyses using spheres instead of cubes for blocks show an effect in
spill angle which is very large especially for lower friction angles between the blocks
(contacts). This subject should be further investigated in future studies. Also, more
studies are recommended in order to validate a kinematic probabilistic analyses based
on the usage of Discrete Fracture Networks (DFN).

Acknowledgements
The authors would like to acknowledge Mr. Ignacio Jara, for his valuable work
during the preparation of the article

References
[1] S. Yaez, R. Silva, Herramienta de Anlisis Cinemticos para Taludes en Minas a Cielo Abierto
(KATS). VIII Congreso Chileno de Ingeniera Geotcnica Santiago-Chile, (2014).
[2] P. Gmez, R. Silva, A Kinematic Analysis Tool for Slopes (KATS), Rock Mechanics for Natural
Resources and Infrastructure SBMR 2014 ISRM Specialized Conference 09-13 Goiania, Brazil,
(2014).
[3] Itasca Consulting Group, "3 Dimensional Distinct Element Code (3DEC)", Inc., version 5.0., 2014
[4] E. W. Lane, The importance of fluvial morphology in hydraulic engineering. Proceedings, American
Society of Civil Engineers, No. 745, (1955).

Das könnte Ihnen auch gefallen