MOXHAM, K.A. A hard rock narrow reef mining machine—ARM 1100.

International Platinum Conference ‘Platinum Adding Value’, The South African

Institute of Mining and Metallurgy, 2004.

A hard rock narrow reef mining machine—ARM 1100

K.A. MOXHAM
Lonmin Platinum

The ARM 1100 was described in a paper presented at the 6th International Symposium on Mine
Mechanization and Automation in September 2001. It started operation underground at Rowland
Shaft in early 2002. This paper covers the results of those trials that resulted in more than 1 100
square metres of UG2 being mined.

Introduction removal of blasting fumes. This change has vast potential in

During the last thirty to forty years, major efforts have been asset utilization and return on investments. An alternate
made to mechanize the South African hard rock mining mining method such as rock cutting will also deliver many
industry. However, most if not all the narrow hard rock reef additional benefits such as:
mines where ‘trackless mechanized mining’ was introduced • Safety—stope conditions without blast induced damage
have reverted back to conventional handheld mining. This • Safety—removing people from danger
trend continues to this very day with two major issues not • People—reducing hard arduous forms of work
yet being resolved: • People—attracting young talent to the industry
• Dilution—defined stope width control
• Dilution and
• Change management.
Lonmin mechanization and automation
Dilution has been and will continue to be the demise of
any attempt to mechanize the mining method. In the higher
strategy
reefs were the envelope is in excess of >1.8 m mechanized Lonmin Platinum has embarked on an aggressive ten-year
mining has become the option of choice. While most of the vision to re-engineer its mining operations. A
gold and platinum mines in South Africa have very narrow mechanization and automation strategy has been formulated
reefs (<1 m) that are not conducive to ‘off the shelf’ copper to transform the mining process into the 21st century.
or coal mining equipment. This implies that new equipment Broadly, the strategy focuses on three main areas:
making use of tried and tested components needs to be • The worker
creatively and innovatively adapted for this restrictive • The workplace
environment. • The manager.
The second issue is the ability to successfully implement Within Lonmin Platinum mines, alternate mining
new and alternate mining methods. These new methods methods are being sought to re-engineer the traditional
make use of machinery and equipment that are handheld explosive based cyclic mining method to that of a
technologically advanced. Mine managers, and engineers non-explosive and continuous nature. One of the pillars of
need to accept the fact that new approaches and an alternate the mechanization and automation (M&A) strategy being
methodology is now required. A systematic approach, pursued is continuous mining and the vision of non-
which holistically views the environment, needs to be explosive based mining. The move from cyclic explosive
adopted, functional boundaries must be challenged, based mining to that of continuous non-explosive mining
alternatives sought and skills developed to sustain this new methods has been the dream of many mining engineers.
way. Much of the vast Bushveld Complex has favourable
Our current mining methods are still dependent on dipping orebodies, mostly in the range of 8 to 13°. This,
blasting techniques to break the rock. The cyclic nature of when compared to the steeply dipping orebodies of the
mining by blasting places severe constraints on the rate of South African gold mines, the bushveld invites the potential
face advance that can be achieved and consequently the to mechanize the stoping process.
utilization of the invested capital. For narrow reef hard rock However it is Lonmin’s view that the orebody will
mining to break out of these constraints and to really make always dictate the mining method, hence there will always
progress in the 21st century it is necessary to follow the be a place for the most appropriate methodology.
lead set by the soft rock mining industry. In the
underground coal mining industry, coal cutting has been
proven to be a most cost-effective solution; thus in narrow
Project history
reef hard rock mining the future must be based on the In early 2001 Lonmin Platinum, Voest Alpine and Sandvik
development of non-explosive methods of rock breaking Tamrock entered into a partnership to co-develop and trial
that in turn are integrated into continuous mining systems. the narrow reef miner. The machine was designed and
Non-explosive so that the mining operation can be manufactured and surface tests were completed in October
conducted on a continuous basis, with no delays for the 2001. The prototype ARM 1100 was then commissioned at

A HARD ROCK NARROW REEF MINING MACHINE—ARM 1100 207

Rowland Shaft and underground operations started in process. In undercutting the crack propagation is to a free
February 2002. face and rock fragments generated are substantially larger.
The trial had two major objectives:
• To demonstrate that the machine could cut hard rock Narrow reef miner’s specification
within a confined space and
Evaluation of the initial results from the pre-construction
• That the machine performance and cutter life could
phase led to the conclusion that the reef miner should have
ultimately deliver a viable alternate stoping method.
the following generalized specifications:
The depth of cut largely dictates the production
• All the cutting discs would be mounted in the same
performance of the machine and that in turn is controlled by
the rock failure mechanism. By varying depth of cut and cutting plane
penetration rate an optimum performance has been • The reef would be cut in an undercutting mode, thus
determined in the UG2, given the torque and force minimizing the power and cutting forces and
limitations are a product of machine size. It has been shown maximizing the chip sizes
that the machine can complete an advance of 850 mm in • The effective undercutting depth was estimated to be
sixty minutes. Production is estimated to be between between 50 and 60 mm
6–8 000 tons per machine per month working a three- shift • Machine stability and stiffness was essential, thus,
22-day cycle. - the machine must be staked while cutting,
The undercutting process generates loads on the cutter - the cutter boom must be as short and stiff as
bearings and cutter carbides that are markedly different possible,
from more conventional hard rock cutting. This initially • The machine will not be fitted with crawlers and will
caused carbide and bearing failures. After an extensive walk in the stope using its staking system
testing programme carbide button rings shrunk onto the hub • The machine must be capable of cutting its own entry
have successfully cut in access of 50 metres advance with into the stope
only 5.5 mm of wear. To achieve economic carbide life an • The machine should excavate the maximum amount of
advance of 100 metres is the target. reef from a single set-up
Rock handling in the face and from here to the interface • The machine in operation must be flexible enough to
into the mine infrastructure has received major attention. follow the major reef undulations
Mechanized drilling and installation of roof bolts have • All rock cuttings must be removed from the immediate
provided the instope support. cutting area and then transported out of the stope.

Theory of disc cutting

Mechanical rock failure is a complex process influenced by
nearly all rock physical and geological rock properties. The
dominant mode of failure is still the subject of much
research. One aspect shared by all the theories is the
existence of a zone of highly crushed rock material beneath
the cutter tip prior to chipping. As the cutter penetrates the
rock, a pressure bulb or crushed zone is formed due to the
extremely high stresses generated in the rock under the tip
of the cutter. The pressure in the crushed zone causes
tensile cracks to initiate and propagate into the rock mass. If
the stresses developed in the crushed zone are sufficiently
high, one or more cracks extend far enough to reach one of
the tensile cracks developed from an adjacent cut, or
alternatively a free face. In conventional disc cutting, rock Figure 1. Theoretical rock breaking process for conventional
failure is in the form of chipping. Figure 1 shows this disc cutting

Figure 2. Examples of existing technology used in the machine design

208 PLATINUM ADDING VALUE

 Additional features of the narrow reef miner obviously very abrasive. This therefore provided an
extremely difficult test for the cutting ability of the ARM
This generalized specification was fleshed out and the reef 1100.
miner conceived is shown in Figure 3 and has the following The ARM was installed on the east side of a raise-winze
main advantages: connection and is cutting from the solid face to the free face
• The machine design was simple in concept and made in the raise. The cutting plan is to cut up-dip for a distance
use of tried and tested components of approximately 70 metres and then turn the machine
• The machine was easy to automate and this had already around and cut down-dip for 80 metres. Once completed,
been done with other coal mining machines the ability of the machine to take a blind cut in a strike
• The machine consists of easily maintainable direction will be tested. Figure 6 shows some early pictures
components with most of the main functions being of the machine underground.
driven by hydraulic cylinders
• Most importantly, because the total system is simple
and flexible, the machine can be accommodated in a Data collection
wide variety of mining layouts. The first two weeks of testing was spent largely on
determining best slewing forces, depth of cut, cutter
penetration, cycle times, etc. Figure 7 is a typical example
Underground trial of the data collected for every cut of the machine.
Rowland |Shaft
Figure 3 shows the ARM 1100 test site at Rowland Shaft.
The cutting trial commenced in January 2002 on 25 Level
with the first cut made in February 2002. The first six
months of the trial was spent on testing different cutting
set-ups and cutter discs and a total of 25 metres were
achieved in the period up to 17 July 2002. At this point the
trial was stopped and new cutters were ordered and
modifications to the machine were undertaken. Cutting
recommenced in the first week of October 2002 and a
further 40 metres were cut in the period to the first week in
November 2002. By the end of January 2003 the machine
had cut a total advance of 100 metres.
The trial team consisted of the following full-time
personnel:
• Artisan/operator - Western Platinum Mine
• Technicians - VAB
• Technician - Sandvik Tamrock
The structure of the UG2 orebody at the trial site is
Lonmin’s deepest, a UCS of ±120 Mpa, extremely hard and Figure 3

Figure 4. Rowland Shaft ARM 1100 test venue

A HARD ROCK NARROW REEF MINING MACHINE—ARM 1100 209

 Figure 5. Typical cut face and detail of footwall contact

Figure 6. Assembly of the ARM 1100, launching the machine and cutting into the raise

Current Penetrat(mm)

Stall Counter []

Cutter Motor Press [bar]

System Pressure [bar]

180

150

Power Cutter 120

motor [kW]
90
FSlew [kN] 60

30
0

Figure 7. Typical machine and cutting data collected for each cycle of the machine

210 PLATINUM ADDING VALUE

Cutter configuration depth of cut. Figure 10 is an example of the detailed
After operating parameters were set different cutter disc analysis carried out and shows how penetration and depth
designs were tested. Cutters tested were manufactured with of cut are related for different carbide and cutter
varying angles of attack, different bearing designs and configurations. However, the fracture mechanics of the rock
different cutting edges, as shown in Figure 8 breaking process limit the depth of cut that can be achieved.
After extensive testing of various disc cutters, a best Figure 11 shows how undercutting with the current cutter
design was selected and it was decided to manufacture design becomes ineffectual as depth of cut is increased.
according to this design with varying tungsten carbide
grades on the inserts. Cutter performance
During the trial period 11 different cutter types and two
Cleaning system modifications different cutter heads have been tested and evaluated. The
During this time the cleaning mechanisms were evaluated cutters are the single most important and least understood
and modified to improve the effectiveness of rock removal component of this project. Furthermore the cutters and the
from the face. Early examples of the different rock removal associated cutter life are vitally important to the economic
system are shown in Figure 9. viability of this alternate method. Cutter costs make up
around 70% of the total costs per ton produced.
Narrow reef miner performance The cutters tested differed in three basic areas, these are:
Over 500 cuts have been recorded to date, by means of the • button orientation (15° and 20° degrees)
on-board data-recording system, and evaluated. Information • shape of buttons/cutting edges (conical buttons, chisel
has thus been collected of the following: type buttons, steel rings)
• bearing configurations
• Cutter arm slewing force
• angle of cutter axis.
• Cutter head power consumption
• Cutter arm slewing speed and mean penetration To date, 15° conical button cutters have performed the
• Advance per cut best in terms of wear, performance (Performance being
• Number of cutter head stalls. defined as cutting rate) and material produced (more chips
This information obtained to date has been used to and fewer fines). Steel-ring cutters performed better and
determine the most suitable cutting geometry and the produced a bulkier sized final product but wore out
determination of optimum cutting parameters. extremely quickly.
With the cutter arm force set at 230 kN the cutter head Figure 12 shows how rock fragmentation is affected by
does not stall however, cutter head penetration is limited by cutter type.

Figure 8. Different carbide profiles tested on the machine

Figure 9. Scoop arm used to load cut rock onto the belt conveyor

A HARD ROCK NARROW REEF MINING MACHINE—ARM 1100 211

 EFFECTIVE PENETRATION / DEPTH of CUT
The effective penetration reflects the siewing speed based on the 6 cutter cutterhead
The ‘set penetration’ is 7 to 9 mm
The effective penetration is limited by the max. Cutter arm siewing force of 230 kN

10

6 Cutter Conical 15° worn

eff. penetration rel.to 6 cutters (mm)

8
6 Cutter Conical 20° slightly worn
7 Cutter Conical 15° worn
7 7 Cutter Conical 20° new
6 Cutter Chiseltyp new
6

4 eff. penetration rel.to 6 cutters (mm)

3

0
20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
depth of cut (mm)

Figure 10. Effective penetration against depth of cut

70 mm depth of cut 100 mm depth of cut

no undercutting no more cutting

Figure 11. With increased depth of cut undercutting ceases to be effective

Chips from steel-ring cutters Chips from chisel-type cutters Chips from 20° conical buttons

Figure 12. Rock fragmentation produced from different cutter configurations

212 PLATINUM ADDING VALUE

 A significant difference in chip size and consequently • Installation of a new loading shovel and increasing the
specific energy was realized from different cutter types. speed of the shovel
• The biggest chips and lowest volume of fines was • An additional hydraulic pump was installed to enable
produced with the steel ring cutter the loading shovel to operate independently from the
• The conical buttons generated the smallest chips and cutting arm.
highest volumes of fines • Manufacture and installation of new muck chute
• The chisel shaped buttons produced fragmentation sizes • Loading plough at the end of cutter boom to remove
somewhere in between the other two cutters. ore more efficiently from the face
• Stopping rotation of cutter head during loading stroke.
Cutterhead
The original cutter head was designed and manufactured General learning and cutter observations
with six cutter discs. It was determined early on in the trial
that more cutters on the head will result in a smoother and The ARM cutting principle is based on the principle that
more effective cutting process. A new cutting head was the cutting action will propagate cracks and fractures in the
built and installed in the middle of July 2002 and it proved rock along which the chips will break off. The nature of the
to be a better option with higher slewing speed and greater UG2 reef however does not support the formation of
depths of cuts being achieved. Future generations of the fracture propagation, which results in relatively low cutting
machine will have to have an increase in installed power. efficiency. The UG2 in the trial area is also of the hardest in
This increase in available power will result in improved the Bushveld Complex and better performance can be
cutting rates and less grinding, resulting in lower fines’ expected in softer UG2 ores and the more brittle Merensky
generation. reef.

Cleaning process
Roof support
From the outset the cleaning of ore away from the cutter
head was identified as a major obstacle and indeed proved Currently roof support is being installed by hand. The
to be problematic. The main issues being: necessity to have an integrated permanent roof support
system was recognized early on in the project. With the
• Cleaning of ore from the cutting path. Too much ore availability now of a drifter small enough to operate in a
remained in the cutter path after completion of the
1.1 M stoping height it was decided to start immediately
cleaning stroke. This is demonstrated in Figure 13
with design of an integrated bolting system.
below
• The ore remaining is subject to a regrinding action,
resulting in high fines content. Ore handling system
• The cleaning arm which feeds the ore into the loading
chute to the conveyor is inefficient The cleaning system behind the ARM is another problem,
• The conveyor design is inadequate and results in which requires urgent attention. At the moment a ULP
excessive breakages. This is the result of both design loader is being successfully used, this provides a flexible,
(return pulley too short) and the accumulation of fines reliable cleaning method however, a very expensive one.
under the conveyor. Other alternatives are also being perused such as:-
A number of changes was made to improve the situation, • Vacuum system
Figure 14 shows the general arrangement, and more • Mobile conveyors
changes are in the process of being made. These are: • Shuttle cars

Face scraper
Loading
Loading chute arm

Figure 13. Material left on the face after the cutter head has
completed its cleaning stroke Figure 14. Modified conveyor belt loading

A HARD ROCK NARROW REEF MINING MACHINE—ARM 1100 213

 Narrow reef miner cutting performance To this end, the first production machine (MK II) is
(Proto-Type Machine) currently underground and cutting at Lonmin Platinum’s
Rowland Shaft.
The proto-type machine cut a total of 270 linear metres. It
is undeniable that the fact that the concept was able to cut Acknowledgement
these metres there is renewed hope that this will become a The author would like to thank the directors of Lonmin
viable alternative to current conventional drill and blast Platinum and the Sandvik Mining & Construction (SMC)
mining methods. for the permission to publish this paper.

Figure 15. The ULP loader in the cut stope

Figure 16. Prototype cutting performance

A HARD ROCK NARROW REEF MINING MACHINE—ARM 1100 214