International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Contents lists available at ScienceDirect

International Journal of
Rock Mechanics and Mining Sciences
journal homepage: www.elsevier.com/locate/ijrmms

Gravity flow characterization of fine granular material for Block Caving T

a,b a,b,⁎ a,b
Vanessa Sánchez , Raúl L. Castro , Sergio Palma
a
Advanced Mining Technology Center, University of Chile, Santiago, Chile
b
Block Caving Laboratory, Department of Mining Engineering, University of Chile, Santiago, Chile

A R T I C LE I N FO A B S T R A C T

Keywords: Block and Panel Caving are based on the action of gravity for ore breakage and transport, so that material
Fine caved rock reaches the drawpoint through gravity flow. The study of caved rock flow is of vital importance for defining the
Moisture content design, planning, ore recovery and consequently the profitability of a mining project. This study experimentally
Unconfined gravity flow investigates the influence of moisture content and fine material on the draw-zone and hang-up frequency.
Flow-zone
Laboratory experiments were carried out on a physical model (scale 1:75) using a Load Haul Dump system
And hang-up
(LHD). The results show that the dimensions, the diameter, and the height of the draw-zone depend on three
variables: moisture content, particle size, and accumulated extraction mass. For the case of tests with moisture,
the increase in diameter reaches a maximum value and becomes almost constant. The ratio of the flow zone
width was found to be in the range of 2.7–3 times the drawpoint width, which is within the range of current
block cave design guidelines. Additionally, with moisture contents over 6%, “non-flow” conditions occur due to
formation of a cohesive arch. Results of our study were compared with data from the Esmeralda mine where the
effects of cohesive arch formation associated with fine material and moisture content were also observed.

1. Introduction fragmented rock is in the range from 1.4 to 3 times the drawpoint
width.
Block caving is a mining method that can achieve high extraction To improve the understanding of gravity flow behavior and verify
rates and low operational costs.1,2 The caving process transforms the data or theories, much research has been performed using physical
initially in situ rock into broken rock, which flows by the action of models, numerical models, and full-scale trials. Using physical mod-
gravity towards the drawpoints. Here, the gravity flow characteristics eling, many efforts have recently concentrated on understanding the
have a significant influence on engineering design including drawpoint flow of coarse caved rock using gravel as the model media.7,8 Current
spacing.3 The material flow could present different behaviors related to research has concluded that the height and width of the isolated ex-
the type of ore broken. The rock fragmentation within the ore column is traction zone (IEZ) and the isolated movement zone (IMZ) continuously
probably the most important parameter governing the performance of a increase with the extracted mass and the height of draw.7 Moreover, it
caving operation. Rock fragments are reduced as the column extraction was found that the movement zones of adjacent drawpoints do not
percentage increases due to flow, a process known as secondary frag- interact if the drawpoint spacing is greater than the width of the IMZ.9
mentation. Tests conducted to quantify the influence of the amount of fines, ver-
Empirical guidelines have been proposed to define drawpoint spa- tical load, and water content on the flow behavior of caved rock con-
cing.4,5 The empirical design guideline considered Rock Mass Rating clude that the amount of fines has a large impact on the frequency of
(RMR) or the fracture frequency, the width of the drawpoint, and ex- hang-ups.10
pected fragmentation. In the case of coarse fragmented rock, with a Research using physical modeling to understand the flow of fine
granulometry between 2 and 9 m, a flow diameter range of 8–13 m has material to date is rather limited. Most researchers have used sand as
been estimated, while for fine material (100% low to 0.3 m), it will the model media under dry conditions.11–14 The influence of water on
reach 6–6.5 m.4 Thus, the draw zone is 2.17–3 times the drawpoint's fines remains unknown. In terms of hang-ups, experiments have shown
width. Recent studies in Sublevel Caving mines that involve the flow of that particle size and opening size are key parameters to study inter-
blasted rock have indicated the width of the extraction zone is 1.4 times ference during the flow of coarse fragments. There are, however, no
the drawpoint width.6 Therefore, the width of the flow-zone for fine studies for fine caved rock.15,16

⁎
Corresponding author at: Advanced Mining Technology Center, University of Chile, Santiago, Chile.
E-mail address: [email protected] (R.L. Castro).

https://doi.org/10.1016/j.ijrmms.2018.12.011
Received 23 August 2017; Received in revised form 19 November 2018; Accepted 7 December 2018
Available online 25 December 2018
1365-1609/ © 2018 Elsevier Ltd. All rights reserved.
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

In this work, we studied the flow behavior of fine fragmented rock

through controlled experiments. We defined fine caved rock as material
that has experienced a secondary fragmentation such that it consists of
a heterogeneous mixture of blocks and fines rocks, where these latter
are present in a higher percentage. For this purpose, we first conducted
an analysis of similitude to define the boundary and initial conditions
for the experiments. We then built an experimental setup and carried
out trials under unconfined conditions to evaluate the influence of
moisture content on the extraction zone and hang-up frequency.
Finally, we compared the experimental results to available mine data.

2. Analysis of similitude for fine granular flow

The concept of similarity considers two systems: the prototype and

the model.17 The theory indicates that a scaling system and a real
system are similar if three similarities are satisfied: geometric, kine-
matic, and dynamic.17,18 In terms of dynamic similitude, the main
forces acting on the system must be well represented. This analysis
guarantees that the results obtained in the physical model are re-
presentative of the prototype. In the following section, the justifications
for the selection of the model media and physical model geometry are Fig. 2. Maximum shear strength for the media tested.
presented.

2.1. Model media

2.1.1. Size distribution

To characterize tested fine caved rock, samples were obtained from
drawpoints located at the Esmeralda Sector at El Teniente mine. The
material is considered by the mine personnel as fine material and cor-
responds to drawpoints with a high percentage of extraction. Fig. 1
shows the size distribution of the samples indicating that the average
particle size (d50) is 3.1 mm with a maximum size of 50 mm, and a small
fraction of coarse material. The media has been defined as well-graded
with a Cu of 43.3 and its content consists of 26.2% gravel, 66.7% sand,
and 7.10% silt and clay.

2.1.2. Shear strength characterization

The shear strength characteristics of the tested material were de- Fig. 3. Unconfined compression for three samples at Diablo Regimiento mine
termined by shear tests. The applied normal forces ranged from 49.0 to [20].
441.3 kPa. Fig. 2 shows the results of the test which indicated that
cohesion is 21.1 kPa, and the friction angle reaches 16.5°. These values
The shear strength parameters, therefore, depend on the confine-
of cohesion and friction angle are characteristic of cohesive soils.19
ment. For low confinement, the cohesion would be between 0.4 and
The above test does not consider low-stress conditions. Previous
0.8 kPa with a friction angle of 35°. For high confinement, the cohesion
unconfined compression tests were conducted for the same media for a
would be 21.1 kPa with a friction angle of 16°.
range of relative densities as shown in Fig. 3.20 The results indicate that
the strength is low for low relative densities.
2.1.3. Analysis of forces
The gravity flow in block caving mainly involves the forces of
gravity, friction, and cohesion.21 The similitude analysis for fine frag-
mented rock consists of defining the impact of water and fragment size
on capillary and electrostatic forces. It also defines the experimental
conditions of the model media. In Appendix A, the relationships for the
capillary forces are stated. Fig. 4 indicates the relationship between
gravity force (FG) and capillary forces (FC) as a function of fragment
size. The results indicate that the higher the moisture content, the
larger the capillary forces and also the smaller the fragment size, the
higher the influence of water. As noted, capillary forces could be larger
than gravity forces for a given set of fragment sizes and moisture con-
tent. For example, for a moisture content of 8%, capillary forces are
larger than gravity for particles below 2.7 mm. In the case of 2%
moisture content, capillary forces have more effect than gravity for
particles below 2.3 mm. The effect of these forces on our sample will be
discussed further in Section 5.2.
Electrostatic forces are another source of cohesion. These forces
Fig. 1. Size distribution of fine caved rock at Esmeralda mine. exist in all material, independent of its chemical nature.22 Electrostatic

25
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 4. Relation between gravity force (fg) and capillary forces (fc) as a function of fragment size for four moisture contents.

forces are larger than gravity for particle diameters below 100 µm.23 In (7.0 m and 14.60 m scaled). The dimension of the drawpoint drift was
the case of tested material, gravity forces are larger than electrostatic 6 × 6 cm (4.5 ×4.5 m scaled). The extraction system considered a
forces for 84% of the total mass. scaled load haul dump (LHD) bucket with a volume of 2.5 × 10−5 m3
Fig. 5 indicates the fragmentation for different geometrical scales (10.6 m3 scaled). The physical model has two-scaled LHDs, which allow
(e.g. 1:100, 1:75, 1:50) and the critical size under which capillary forces different extraction conditions: draw from one or two drawpoints.
would be dominant (under 2.7 mm for a moisture content of 8%). The As noted previously, the drawpoint (DW) has a width of 60 mm, and
results show that if the material is reduced in size to any scale, capillary the characteristic sizes such as d50, d80 and d100 are presented in Fig. 1.
forces would be larger than gravity for 100% of the total mass. The relation between drawpoint width and fragment size is indicated in
Therefore, in order to study the gravity flow of fine caved rock, the the following table.
same samples from Esmeralda mine described above were used. On the The literature indicates that if the ratio between the drawpoint
size distribution curve selected, 49% of particles would be affected by width and largest block size is greater than 5, the flow will be almost
capillary forces when the moisture content reached at 8%. certain. With a value in a range from 3 to 5, the flow will be uncertain,
and finally with a ratio lower than 3 almost certainly there will be no
2.2. Geometry of the physical model flow.25 In this way, the data of DW/dp show that most of the media will
not present hang-ups, but 20% of the size distribution of the material
After defining the model media, we designed and built the physical greater than 13 mm may be a cause of hang-ups during the flow
model. The laboratory scale selected was 1:75. The main advantage of a (Table 1).
smaller scale compared with full-scale is the low cost and reduced work
time. A smaller scale also offers a good view of the physical phenom- 2.3. Boundary conditions
enon. Moreover, recent studies indicate that satisfactory results can be
obtained using a scale factor 1:100 and 1:200.7,24 With gravity flow, it is important to define the column height as this
Fig. 6 shows the geometry of the physical model. The dimensions of defines the geometry of the extraction zone and the vertical loads acting
the flow chamber were 66.7 cm, 52.1 cm, and 33.3 cm, height, width at the production level. In cohesive granular materials, vertical stresses,
and depth, respectively (50.0 m, 39.1 m, 25.0 scaled m). σv are calculated by
The dimensions of the drawbell were 9.33 cm high and 19.5 cm long
γRh − cM ⎧ −zM tan ϕ ⎤ ⎫
σv = 1 − exp ⎡
M tan ϕ ⎨
⎩ ⎢
⎣ Rh ⎥⎬
⎦ ⎭ (1)

where γ is the specific weight of the material (20.8 kN/m ), Rh is the 3

hydraulic radius of the flow-zone (area/perimeter), z is the depth, and

M depends on the friction angles and is given by
M = 1/[1 + 2 tan2 ϕ]

This formula is an extension of the previous formula of Janssen and

takes into account any tendency for fines to adhere to ore pass walls, as
well as the friction of mud.26,27
In order to establish the vertical loads during flow at the mine, the
mechanical parameters shown in Section 2.1 were considered along
with different draw-zone diameters (see Fig. 7). As shown, the larger
the draw-zone diameters, the larger the vertical load will be. In this
context, for the flow-zone diameters within a range 5–20 m, the max-
imum loads are between 30 and 350 kPa.
In the case of the physical model, we applied Eq. (1), considering
the mechanical properties presented in Section 2.1.2, indicating that
the vertical loads are in a range from 3.8 to 4.8 kPa. Table 2 shows the
Fig. 5. Effect of interparticle forces as a function of material fragment size. Fc: flow factor for the physical model and prototype, demonstrating that
capillary forces; Fe: electrostatic forces. the flow factors are similar.

26
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 6. Experimental set-up – scale 1:75 - physical model, distribution of markers in ore column and drawbell dimensions.

Table 1 Table 2
Ratio between drawpoint width and characteristic size. Flow factor between prototype and physical model.
Characteristic size, DW dp, mm DW/dp Cohesion, (c) kPa Vertical load (σv) kPa σv/c

d50 3.1 19.4 Prototype 66.0 150.0 2.3

d80 13.0 4.6 Physical model 3.6 4.8 1.3
d100 50.0 1.2

the material d50, thus they did not alter the material flow. Extraction
zones were measured through the back analysis of recovered labeled
markers. This technique has been tested successfully in experimental
research on physical models that measure the flow-zone geometry.7
During the experiments, when approximately 20% of the total mass
was drawn, it was observed that the flow reached the surface. In this
case, the model was refilled with the extracted material to prevent
subsidence and to maintain the column height constant and conse-
quently, the vertical strength of the system. Additionally, it was ob-
served that during the extraction, no flow conditions and hang-ups
occurred. The events were noted that occurred in a series of consecutive
steps: (a) full drawbell or flow condition, (b) start of hang-up or void at
the drawbell, and (c) no flow condition as shown in Fig. 8.
In order to achieve the research objectives, ten experiments were
conducted (see Table 3). In the experiments, the amount of water was
changed from 0% to 8%. Also, the number of drawpoints was varied
from one to two. Finally, to quantify the experimental error, additional
experiments or replicas were carried out.

Fig. 7. Vertical load expected at the mine as a function of depth and draw-zone 4. Results and discussion
diameters.

This section shows the results of the effect of moisture content on

3. Experimental methodology the geometry of the draw-zone for extraction from one versus two draw
points. In addition, the frequency of hang-ups found in each experiment
To determine the extraction zone geometry, 768 painted numbered is quantified.
markers were positioned inside the model and recovered at the draw-
point. Each marker had a code and a unique location (coordinates X, Y,
and Z). Moreover, their sizes corresponded to the average diameter of

27
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 8. Evolution of a “no flow” condition - cross-sectional view of the drawbell.

4.1. Geometry of draw-zones diameter. Thus, for 0% and 2% moisture content, the diameter of the
flow-zone increases at a lower rate but does not attain a threshold.
The geometry of the flow-zone as a function of mass is obtained by
interpolating the location of the recovered markers. Fig. 9 shows an 4.2. Frequency of hang-ups during the flow
example of the flow-zone for an isolated draw, while Fig. 10 for two
drawpoints. In the case of an isolated draw, the extraction zone follows To quantify the interference during the flow, Table 4 contains data
a cylinder shape that increases in height with the extraction as theory about hang-up frequency. Frequency of hang-ups was quantified in
dictates. In the case of two drawpoints, initially the draw-zones acted in terms of the extracted mass between two consecutive hang-ups grams/
isolation but due to the close spacing, the flow-zones later merged into hang-up. The results indicate that as the moisture content increases
a single extraction zone. there is a decrease in the amount of mass drawn between consecutive
hang-ups.
4.1.1. Extraction from one point The results also show a statistically significant difference between
In the case of an isolated draw (experiments 1, 3, 5, 7 and 9), the drawing from one or two drawpoints: drawing from two drawpoints
geometry of the flow-zone was characterized in terms of its height and increases the amount of draw between hang-ups.
width (see Figs. 11a and 11b, respectively). Fig. 11a shows that the In terms of the hang-up mechanism in the case of dry material, as
height of extraction increases with mass for all tests where moisture expected there were events due to the presence of large fragments re-
content was below 6%. For 6%, the draw height reached a maximum of lative to the size of the drawpoint, fragments with a size equal to or
36.10 cm as a stable arch developed. In the case of experiment 9 that greater than d80 (13 mm). In the case of wet samples, hang-ups were
considers an 8% moisture content, 4.17 kg of mineral was extracted due to the cohesion that water induced in the fines.
before a cohesive arch with a height of 13.6 cm occurred.
Fig. 11b indicates that the diameter of extraction increases with 5. Discussion
mass for all tests. However, the rate of growth depends on the moisture
content. For samples with 2% of moisture content and above, the draw- 5.1. Geometry of flow-zone
zone rapidly increases and reaches a maximum. For dry samples, the
draw-zone increases at a lower rate but does not appear to reach a To compare the results with the available literature on flow of fine
maximum. caved rock, we compared our experimental data with the data available
Five replicas were carried out with a moisture content of 4% to in the literature. In the case of an isolated draw-zone for fine caved
calculate the experimental error. The relative error for the diameter rock, Laubscher proposed an expected diameter 6–6.5 m for drawpoint
reached 9.67%, while that of the height reached 6.69%. width of 2 and 3 m respectively, when within the flow, the fragment
size is lower than 0.3 m.4 In this context, Table 5 shows the di-
4.1.2. Extraction from two points mensionless number as a ratio between flow-zone width and drawpoint
To determine the difference in geometry and hang-up frequency width.
between drawing from one versus two extraction points, we carried out As noted, the flow-zone (scaled) from the experiments is larger than
experiments with the same moisture intervals as in the isolated case. In that expected using Laubscher methodology. However, the dimension-
the following analysis for experiments 2, 4, 6, and 8 when the moisture less flow width number (WFZ/Dw) indicates that the flow-zones are
content varies from 0% to 6%, the results show that the increase of within the range from 2.7 to 3 times the drawpoint width, which is
draw-zone height occurs for all tests, even for experiment 8 that con- within the range of expected results.
siders 6% moisture content (see Fig. 12a). On the other hand, Fig. 12b As a way of observing the ore recovery of fine fragmented rock in a
indicates that the increase of the flow-zone diameter depends on the modern LHD layout, Fig. 13 shows a plan view of the draw-zones for a
moisture, and each moisture content has a different rate of growth of 4 m drawpoint width. This indicates that current drawpoint spacing is

Table 3
Summary of experiments conducted.
Description Units Experiments

Exp 1 Exp 2 Exp 3 Exp 4 Exp 5 Exp 6 Exp 7 Exp 8 Exp 9 Exp 10

Drawpoints – 1 2 1 2 1 2 1 2 1 2
Moisture Content % 0 0 2 2 4 4 6 6 8 8
Accumulated extraction mass kg 21.4 50.8 22.0 45.5 21.4 44.3 22.3 44.2 4.8 29.0

28
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 9. Evolution of the ellipse to different accumulated extraction mass with 4% moisture content - isolated draw - longitudinal view to the drawbell.

Fig. 10. Evolution of the ellipse to different accumulated extraction mass with 4% moisture content - tests with draw from two drawpoints - longitudinal view to the
drawbell.

not adequate for recovering fine fragmented rock ore as only approxi- distribution when 80% of hang-ups are with a range of 1–2 hang-ups/
mately 58% of the total area is under draw. 1000 t. This information confirms that there are hang-ups associated
with fine material and moisture content, but with a lower frequency
compared with experimental data. The difference in hang-up frequency
5.2. Frequency of hang-ups related to cohesive arches
between the physical model and the prototype probably occurred be-
cause the sample was not scaled to the same level as the physical model
To evaluate the impact at mine scale, it is necessary to scale the data
in order to avoid overestimates of hang-ups due to capillary force ef-
to ton/hang-up as follows:
fects. However, observing the formation of cohesive arches generated at
g (scale3) laboratory scale by the combinations of moisture content and fine
tons = =g
1 × 106 (3) material variables did provide information about the mechanisms in-
volved which are summarized below in our conclusions.
The results are summarized in Table 6 and show that the number of
hang-ups per 1000 t extracted have similar results between isolated and
interactive flow, however, their difference is statistically significant. 6. Conclusions
Thus, the extraction from two points reduces the hang-ups due to a
greater movement of particles. In response to a paucity of literature on the flow of fine fragmented
Using a database of the Esmeralda Mine, we analyzed the hang-up rock and moisture content, experiments were carried out on a physical
values for extraction points associated with characteristics: fine mate- model to investigate the influence of these variables. The results con-
rial (fragment size < 25 cm) and moisture content (0–7%). The ana- firmed that moisture content affects the geometry of flow-zones and
lyses show an average of hang-ups of 1.2 ± 2.08 hang-ups/1000 t and hang-ups caused by cohesive arches. In this context, the experimental
present the following frequency distribution (see Fig. 14). data indicated that the diameter increased with a different rate of
Fig. 14 indicates that the distribution could have an exponential growth depending on moisture content. We found that with wet

29
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 12. a) EZ height and b) EZ width as a function of accumulated masses, for

different contents of moisture.
Fig. 11. a) IEZ height and b) IEZ width as a function of accumulated masses for
different moisture contents.
Table 4
Summary of hang-up frequency obtained during experiments.
samples the diameter of the drawzones tended to stabilize over time,
Experiments Drawpoints Moisture Number of Hang-up
while the drawzone diameter with dry samples (0–2% moisture con-
content hang-ups frequency grams/
tent) continued to increase with the mass drawn. [u] hang-up
Through this study we were also able to identify that the diameter of
the draw-zones is 2.7–3 times the drawpoint width for fine fragmented 1 1 0 53 428.0 ± 73.2
2 2 0 48 464.4 ± 97.9
rock in comparison to what is being used for design guidelines; for this
3 1 2 55 392.8 ± 75.7
reason, in modern LHD drawpoint spacing, ore recovery is expected to 4 2 2 54 400.2 ± 90.9
be low. The mass drawn between hangs-ups shows a dependency on 5 1 4 59 352.0 ± 68.4
moisture content and number of drawpoints. The higher the moisture 6 2 4 55 380.7 ± 43.7
content, the less mass can be drawn between events. Drawing from two 7 1 6 63 337.7 ± 66.4
8 2 6 63 341.8 ± 76.5
drawpoints allows for drawing more between events than when
drawing in isolated conditions.
The results obtained through small, scaled physical modeling have
Table 5
been useful to understand the mechanisms of granular material for Ratio between flow-zone width and drawpoint width, considering Laubscher
block caving purposes. However, there is still work to be done to clarify data and experimental data (scaled).
the role of height of the draw-zone and the effect of a large number of
Data Fragment Flow- Drawpoint Dimensionless
drawpoints, which would require building a larger physical model.
size, m zone - Width, DW, m flow width
There is also a need to conduct full-scale trials on fine fragmented rock Width, number WFZ/DW
using smart markers such as those carried out for coarse caved WFZ, m
rock.28,29 Another important and complementary tool would be to de-
Laubscher [4] < 0.3 6.5 3.0 2.2
velop numerical models to increase current knowledge and confirm
6.0 2.0 3.0
industry standards for mine design and operation. This kind of study Experiment 1 < 0.37 12.1 4.5 2.7
using a physical model can support the development and calibration of Experiment 3 12.5 4.5 2.8
numerical models to analyze gravity flow in cohesive scenarios, and to Experiment 5 13.3 4.5 3.0
consider strategies to mitigate the results of such scenarios. Experiment 7 13.7 4.5 3.0

30
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 13. Projection of the diameters obtained from two drawpoints with moisture contents of 0%, 2%, 4%, and 6% in a current layout.

Table 6 Laboratory, the Advanced Mining Technology Center of the University

Summary of hang-up frequency scales to prototype. of Chile and Codelco El Teniente in providing useful data for hang up
Experiments Moisture Hang-ups frequency, ton/ Hang-ups/
analysis. Also, we would like to thank the scholarship PIENSACOBRE
Content hang-up 1000 t with their sponsors: Mathematic Modeling Center (CMM), and National
Copper Corporation of Chile (CODELCO) for their financing. This paper
1 0 180.6 ± 30.9 5.5 ± 1.2 describes a component of the work carried out within the project
2 0 196.0 ± 41.3 5.1 ± 1.2
“Gravity flow technologies and fundamentals” run by the University of
3 2 165.7 ± 31.9 6.0 ± 1.7
4 2 168.9 ± 38.3 5.9 ± 1.2 Chile's Advanced Mining Technology Center which is funded by the
5 4 148.5 ± 28.8 6.7 ± 1.6 Chilean government through the CONICYT Project FB0809. V Sanchez
6 4 160.6 ± 18.4 6.2 ± 1.0 would like to thanks the scholarship PIENSACOBRE with their sponsors:
7 6 142.5 ± 29.7 7.0 ± 1.6
Mathematical Modeling Center (CMM) and CODELCO for their support
8 6 144.2 ± 32.3 6.9 ± 1.8
during her studies.

Fig. 14. Frequency distribution of hang-ups – Esmeralda mine.

Acknowledgments

The authors gratefully acknowledge the support of the Block Caving

Appendix. : The effect of interparticle forces on the model media

To determine the influence of capillary forces (FC), we calculated a simplified relationship between capillary forces with gravity force in a
granular media. The result determined the size of the particle in which the capillary forces could govern the flow. The following analysis considers
two identical solid particles, negligible liquid surface curvature in planes containing the sphere centers, and constant hydrostatic pressure in the
liquid volume.
In Fig. 15, R represents the radius of a spherical particle; R0 and R1 are the curvature radius of the air-liquid interface in the narrowest region of
the liquid volume and θ is the contact angle due to surface tension. The force FC acting in the liquid meniscus has two components, one due to the
surface tension FTS, appearing on the liquid perimeter, and another as a result of the liquid pressure FP30:

31
V. Sánchez et al. International Journal of Rock Mechanics and Mining Sciences 114 (2019) 24–32

Fig. 15. Capillary effect between two particles.

FTS = 2πγc R sin β (A1)

FP = 2π (R sin β )2 (A2)
where γc is the surface tension of the liquid, β is the angle given by the line defined by the interface solid-liquid-air, and the pressure, P, comes from
Laplace's equation. This relates the difference of pressures on both sides of a liquid membrane. P is given by
P = γc / R sin β (A3)
With these equations, the expression corresponding to the capillary forces is:
FC = FTS − FP (A4)

FC = πγc R sin β (A5)

References 15. Castro RL, Fuenzalida MA, Lund F. Experimental study of gravity flow under con-
fined conditions. Int J Rock Mech Min Sci. 2014;67:164–169.
16. Hustrulid W, Sun C. Some remarks on ore pass design guidelines. In: Massmin 2004
1. Brown ET. Block Caving Geomechanics. Brisbane: JKMRC, The University of Proceedings. Santiago; 22–25 August 2004: 301–308.
Queensland; 2007. 17. Barenblatt GI. Scaling, Self-Similarity, and Intermediate Asymptotics. Cambridge:
2. Heslop TG. Block caving–controllable risks and fatal flaws. In: MassMin Proceedings of Cambridge University Press; 1996.
. Brisbane; 29 October–2 November 2000: 280–291. 18. Langhaar HL. Dimensional Analysis and Theory of Models. New York: Wiley; 1951.
3. Chitombo GP. Cave mining: 16 years after Laubscher’s 1994 paper “Cave mining – 19. Lambe TW, Whitman RV. Soil Mechanics. New York: Wiley; 1969.
State of the Art”. In: Proceedings of the Second International Symposium on Block and 20. Castro RL, Basaure K, Palma S, Vallejos J. Geotechnical characterization of ore re-
Sublevel Caving, Caving 2010. Perth; 2010; 20–22 April 2010: 45–61. lated to mudrushes in block caving mining. South Afr Inst Min Metall.
4. Laubscher DH. Cave mining-the state of the art. South Afr Inst Min Metall. 2017;117(3):275–284.
1994;10:279–293. 21. Gómez R, Castro RL, Casali A, Palma S, Hekmat A. A comminution model for sec-
5. Laubscher DH. A Practical Manual on Block Caving. Report for the international caving ondary fragmentation assessment for block caving. Rock Mech Rock Eng.
study. Brisbane: JKMRC and Itasca Consulting Group; 2000. 2017;50(11):3073–3084.
6. Hustrulid W, Kvapil R. Sublevel caving–past and future. In: Proceedings of the 5th 22. Israelachvili J. Intermolecular and Surface Forces. 3rd ed. San Diego: Academic Press;
International Conference & Exhibition on Mass Mining, Massmin. Lulea; 9–11 June 2008: 2011.
107–132. 23. Seville JPK, Willet CD, Knight PC. Interparticle forces in fluidisation: a review.
7. Castro R, Trueman R, Halim A. Study of isolated draw zones in block caving mines by Powder Tech. 2000;113(3):261–268.
means of a large 3D physical model. Int J Rock Mech Min Sci. 2007;44:860–870. 24. Castro R, Pineda M. The role of gravity flow in the design and planning of large
8. Power G. Modelling Granular Flow in Caving Mines: Large Scale Physical Models and Full sublevel stopes. South Afr Inst Min Metall. 2015;115(2):113–118.
Scale Experiments [Ph.D. thesis]. University of Queensland; 2004. 25. Hadjigeorgiou J, Lessard JF. Numerical investigations of ore pass hang-up phe-
9. Trueman R, Castro R, Halim A. Study of multiple draw-zone interaction in block nomena. Int J Rock Mech Min Sci. 2007;44(6):820–834.
caving mines by means of a large 3D physical model. Int J Rock Mech Min Sci. 26. Beaus M, Pariseau W, Stewart B, Inverson S. Design of ore pass. In: Hustrulid W,
2008;45:1044–1051. Bullock R, eds. Underground Mining Methods. Littleton, Colo: SocMining Eng.;
10. Olivares D, Castro R, Hekmat A. Influence of fine material, humidity and vertical 2001:627–634.
loads on the flow ability of caved rock. In: Proceedings of the 49th US Rock Mechanics/ 27. Pariseau WG. The Gravity Induced Movement of Materials in Ore Passes Analyzed as A
Geomechanics Symposium. California; 28 June–1 July; 2015. Problem in Coulomb Plasticity [Ph.D. thesis]. University of Minnesota; 1966.
11. Kvapil R. Gravity flow of granular materials in hoppers and bins in mines - II. Coarse 28. Garcés D, Viera E, Castro R, Meléndez M. Gravity flow full-scale tests at Esmeralda
Material. Int J Rock Mech Min Sci. 1965;2(1) [277-04]. Mine’s Block-2, El Teniente. In: Proceedings of the Seventh International Conference &
12. Marano G. The interaction between adjacent draw points in free flowing materials Exhibition on Mass Mining. Sydney; 9–11 May 2016: 349–358.
and it application to mining. Chamb Mines J. 1980;22:25–32. 29. Whiteman DS. Smart Marker System – a new tool for measuring underground ore-
13. Kvapil R. Sublevel caving. SME Mining Engineering Handbook. Litteton, CO: Soc. Min. body flow in block and sublevel mines. Min Tech. 2010;119(3):182–192.
Metall. Explor.; 1992. 30. Kane JW, Sternheim MM. General Physics. 2nd ed. New York: Wiley; 1991.
14. Susaeta A. Theory of gravity flow (Part 1). In: Massmin 2004 Proceedings. Santiago;
22–25 August 2004: 167–172.

32