The design and testing of pipe-sticks for

underground stope support

by P. J. J. VAN VUUREN*, B.Sc. (Eng.)
SYNOPSIS
Pipe-stick support in deep-level gold mines is first described and illustrated, a pipe-stick being a mine pole trimmed
to the required size and sleeved by a pipe. An account is then given of an investigation into this form of support
that involved load-compression tests in a IOOO-ton press, tests in a rapid-loading machine simulating rock-burst
conditions, a study of the load-compression behaviour of pipe-sticks installed underground, and a cost evaluation.
Pipe-stick support was found to be very effective on the face and to cost less than conventional matpacks.

SAMEVATTING
Pypstokbestutting in diepgoudmyne word eers beskryf en ge'illustreer. 'n Pypstok is 'n mynpaal wat tot 'n vereiste
grootte afgeskaaf word en omhul is deur 'n pyp. Daarna word daar verslag gedoen oor 'n ondersoek in verband met
hierdie vorm van bestutting wat las-druktoetse in 'n IOOO-tonpers, toetse in 'n snellaaimasjien wat rotsbarstoestande
naboots, 'n studie van die gedrag van pypstokke wat ondergronds aangebring is onder belasting en samedrukking,
asook 'n koste-evaluering behels het.
Daar is gevind dat pypstokbestutting baie doeltreffend is op die werksfront en minder as die konvensionele
matpakke koso

Introduction The prime function of stope support is therefore to

The design of stope support for tabular orebodies in generate sufficient force very soon after installation to
deep-level gold mines is very complex, and continuous control the rate of bed separation and deformation of
research is being carried out to find the optimum support the immediate hangingwall within acceptable limits.
system to satisfy the great variety of parameters. Much Since the elastic convergence is irresistible, the support
of the research is centred around the in situ behaviour of elements should have a yield characteristic that allows
stope support elements and the behaviour of the stoped- shortening whilst support is maintained.
out area itself. It is well-known that the various types of stope sup-
Towards the end of 1975, the Rock Mechanics Depart- port elements used in gold mines react differently when
ment of General Mining and Finance Corporation de- deformed3-6. Ideally, stope support should be very stiff
cided to initiate an experimental programme on pipe- initially so as to reach the design yield load with a
stick support. Tests were undertaken at the C SIR minimum amount of compression, and should then yield
laboratories, and installations were incorporated in a at a constant load.
stope at Stilfontein Gold Mine under protected condi- It is significant that a comparison of laboratory tests
tions, i.e. in between the matpacks forming the normal and underground in situ tests indicate a much lower
stope support, in order to establish the characteristics load-bearing capacity for the latter, probably owing to
of pipe-stick support. The conclusions drawn from these the time-dependent reaction and decaying of the packs
tests were very encouraging, and it was decided that the underground 7.
experiment should be continued on a larger scale.
This paper describes this support system and its pos- Features of Pipe-stick Support
sible application in the gold-mining industry. Pipe-stick support consists of a mine-pole and a pipe
forming a sleeve round the mine-pole (Fig. 1). When a
Stope Support Requirements mine-pole is installed with its long axis parallel to the
Stoping creates a slot in the rock, which disturbs the direction of the loading force, it very rapidly reaches its
equilibrium and results in a redistribution of the weight maximum load-bearing capacity, i.e. within 2 per cent
of the superincumbent strata to the solid areas, where compression. After this point has been reached, failure
additional stresses are induced. Owing to structural commences and the support very soon becomes ineffec-
weaknesses and stress fractures in the hangingwall rock, tive (Fig. 2).
bed separation occurs, creating hazardous conditions in When a stick is enclosed by a pipe, that pipe provides
the stope!' 2. lateral restraint to the failing timber and therefore
As stoping progresses, elastic and dilatation converg- enables the support system to carry considerable load
ence takes place, the rate of convergence depending on long after the stick has failed. The load-bearing capacity
the rate of face advance, depth of mining, strength of of a pipe-stick is higher than either that of the pipe or
hangingwall rock, and type of support installed. the stick. The pipe sleeve prevents destruction of the
Although the elastic convergence is virtually irresistible, stick, whilst the stick gives stiffness to the pipe.
it is possible to control the amount of convergence due To allow for variation in stoping width, a short length
to bed separation and maintain the integrity of the of stick is left protruding at either end of the pipe. When
immediate hangingwall strata. load is applied, sufficient creep takes place to allow the
and Corporation Limited, stick to be pushed right into the pipe before failure
* General Mining Finance
occurs. The protruding stick enables the person instal-
Johannesburg.

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY DECEMBER 1978 125
 Sloping Wldlh Pipe (15 or 20cm)

Fig. I-A pipe-stick for stope support

ling the pipe-stick to cut off a short piece to suit the

stoping width at the point of installation.
To highlight the differences in load-deformation be-
haviour, the load-percentage deformation characteristics
of a matpack (60 cm by 60 cm saligna), a mine pole,
and a pipe-stick are shown in Fig. 3.
The following conclusions can be drawn from Fig. 3.
(a) A mine pole reacts very rapidly, but, once its failure
load has been reached, it soon becomes ineffective.
It is therefore unsuitable as permanent support in
stopes experiencing more than a few centimetres of
elastic convergence.
(b) Pipe-sticks offer a higher maximum load than a
similar stick, maintaining a considerable load during
deformation. Once the maximum load has been
reached, the slope of the load-deformation curve
tends to be downwards. This behaviour may appear
unstable, but, since it does not happen to all the
pipe-sticks at the same time, the overall system is
considered stable. The stiffness of the hangingwall
also contributes considerably to an overall stable
system.
(c) Matpacks offer very little support during the first
10 per cent of deformation and become superior to
pipe-sticks only after about 20 per cent deformation.
The initial stiffness is increased by wedging and
prestressing1, 5 but is eliminated in practice by un-
even hangingwall or footwall and poor installation.
Support is not given in the face where it is needed.
Matpacks maintain the integrity of the hanging by
preventing blocks from falling out by virtue of the
area supported.

Support Pattern Fig. 2-A mine pole after maximum load had been reached
The usual pattern is a row of packs installed on either

126 DECEMBER 1978 JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY
 2000

1500

1000
~ ~
In'::

15cm Plp.-Stlck
Z ~-'-i ,../ -.. '.
:0:
.550 I ~.. '.'u'-""',-
.0
"...
' ""--/'
.."-'.
\, r
," ,-------

I I , , , I , , , , , , I , I , , , , , I I I , I ,
1'0 2'0 '3'0' '4'0' 50 '
Compr...lon In cm

Fig. 3--A comparison between mine poles, pipe-sticks, and saligna matpacks

side of dip and strike gullies, every sixth row of pipe-

sticks often being replaced by a row of packs. The row of
packs adjacent to gullies is considered essential because
(I) punching into the fractured footwall could render
the support much less effective,
0 (2) dislodgement of the pipe-sticks at any stage could
0 result in a collapse in the gully, with serious conse-
quences, and
0 (3) the mat packs maintain the integrity of the gully
0 'r'
.. .
hangingwall.
The replacement of pipe-sticks by packs in every sixth
0
."i .
row is a precaution in the case of sudden failure along a
~
0 parting plane over an extensive area, which would
J result in mass collapse and dislodgement of the pipe-
sticks.
i Fig. 4 illustrates typical stope-support layouts using
~
.
15
pipe-sticks.

01 . Manufacture of Pipe-sticks

~ 5
O! . ,
'1
I
The pipe-sticks being used at present have a diameter
of 15 cm. They are manufactured in different lengths,
which increase by 10 cm from 80 cm to 140 cm.
DJ- ..
E
The pipes are cut to a suitable length, and matching
sticks are cut to a length 20 cm larger than the pipe.
Dt ' " j Both are then painted the same colour, a different
ll, colour for each length. The stick is then placed in a lathe
O Q.
and trimmed to fit firmly into the pipe (Fig. 5). This lathe
11. consists of two electric-motor driven discs with a number
of TO cutters on the disc and a rotating spindle holding
the stick. Once cut, the stick is pushed into the pipe with
a long-travel piston ram so that it protrudes from the
pipe by 10 cm on either side.
Fig- 4- Typical stope-support layouts using pipe-sticks in The capital requirements for a plant that is capable
20 m panels at spaclngs of 2,5 m by 2,5 m and 3,0 m by 3,0 m of producing 300 pipe-sticks per day are R35 000.

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY DECEMBER 1978 127
 Advanta~es of Pipe-stick Support Economic Evaluation
(I) Owing to better load-compression characteristics up An accurate economic evaluation needs to include a
to 20 per cent compression, improved support close detailed analysis of the transport, labour, face advance,
to the face is achieved. This results in better centares per labourer, required spacing, actual spacing,
hangingwall control and safety on the working-face and material cost. The following conclusions were drawn
area. from such an evaluation.
(2) Friable hanging can be effectively supported by (i) This support system has a potential for reducing the
close spacing, and normal sticks can be used in stope timber by approximately 50 per cent overall.
conjunction with pipe-sticks. This method has been (ii) In view of the considerable reduction in the volume
used effectively, resulting in better stoping-width of pack material required, the use of pipe-sticks
control. would effect substantial savings in transport, labour,
(3) Cleaning and sweeping operations have been made and shaft time. This would result in a large direct
easier because of better utilization of the scrapers. saving, with improved delivery of other material
Less rock is left behind between the supports. required for production. The volume of material
(4) There is a considerable saving of transport and required is only 10 per cent of that for matpacks,
installation labour. and the installation time is 20 per cent.
(5) Because less timber is used in the stopes, the fire (iii) Pipe-stick panels have a potential increase in
hazard is reduced. productivity of 10 to 15 per cent over those with
conventional support. Other improvements such as
Disadvanta~es of Pipe-sticks sweepings, regular blasting, and better face support
(a) The support density of pipe-sticks appears to be result in consistently high face advance.
inadequate owing to their small cross-sectional The following is a cost comparison made in 1977 for
dimensions compared with packs. Miners therefore a stoping width of I m.
install them at a greater density than that laid down, 2000
or change to timber packs when bad hangingwall units
conditions are encountered. This practice is very Unit per
costly. cost month
(b) Pipe-sticks are unsuitable for use adjacent to gullies. Pipe-sticks R R
A different design is required there to ensure effec- Cost of 15 cm pipe (new) 9,00 18000
tiveness until total closure is reached. The dislodge- Cost of 1,5 m at 15 cm pole (50%
ment of pipe-sticks can result in a collapse over wastage) 1,00 2000
the gully. Machining costs ,50 1000
(c) The angle of installation is critical, being the main Shaft and haulage transport costs ,50 1000
cause of instability. This should be emphasized in Stope transport and installation
the training and supervision of miners. costs ,60 1200
(d) The maximum stoping width for the stability of a
particular diameter of pipe-stick limits its applica- Rll,60 R23 200
tion to narrow reefs.
(e) The negative slope of the load-deformation curve Solid matpacks (60 cm by 60 cm)
beyond the yield point, and the fact that the maxi- 10 mats per pack at 87c each 8,70 17400
mum load is much less than for mat packs, may Wedges ,50 1000
place pipe-sticks in disfavour. Shaft and haulage transport 3,00 6000
Stope transport and installation 2,50 5000

R14,70 R29400

Paired matpacks
(1l0 cm by 1l0 cm)
20 pairs per pack at 83c each 16,60 33 200
Wedges ,50 1000
Shaft and haulage transport 5,50 1l000
Stope transport and installation 4,60 9200

R27,20 R54 400

These costs apply only to the number of matpacks

that are actually replaced by pipe-sticks, which indicates
that, even when pipe-sticks are spaced at a greater
density (about 30 per cent greater), they remain
economically viable. The use of reject or secondhand
pipes would effect an additional saving of about R6000
Fig. S-Manufacturing pipe-sticks per month.

128 DECEMBER 1918 JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY
 Comparative costs per centare mined are represented
graphically in Fig. 6. These are based on the atove costs ()o
and are for a typical mine using one of the support pat- ,0 't

terns shown in Fig. 4. .,

i;

fr ,
Test Results and Deductions ;,
t
. j 1
A series of load-compression tests were conducted in ~ tl ~
T" Hf
a lOOO-ton press at the CS I R. The effect of variations 8
~ I "I' j. "i;
in components was tested on a large variety of pipe- ,",

sticks, and load-deformation graphs were plotted and

+ h "1. ,! 'I
if 1£.;: I 'r

the behaviour and mode of failure of these elements f!jt':'ji;;:lIif::! 'iiiP

I :
.i Ipl if
o.

observed. In addition, pipe-sticks were tested at the I:::

' :11H
Chamber of Mines Research Laboratory on a rapid-
loading machine simulating rock-burst conditions. 6 1;
""'. : ill I:ii I I
iIffi i'
1.! "111 in:
Finally, pipe-sticks equipped with load cells were in- ,
stalled underground to monitor load-compression be- . ; j
111
1;;llj ,I:
I!,!
:.
haviour over a period of time. , ., lfi, rrH: :~I~b;"
The following deductions were drawn from these tests. 1; ,
ni'"!'I.#'
,
I:,,;;.>.,
(a) Diameter I
111 I
. ",:f:HI l(i.U~.'1ii
The load-bearing capacity of a pipe-stick of 20 cm i " + " i:l, iLl Ill:
diameter is about 25 per cent greater than that of a ! iil' iJI.
Ill: :111,
pipe-stick of 15 cm diameter. A 20 cm pipe-stick
l! j
/!U jl~ ,.. ILi. ',:
.,'L' '11 ' :1 i"
I, I Ill. IJ: 'I:,
would also be more stable at higher stoping widths.
I -IIf ~ Ht
~,
iU J!.bL!~ q ;i;: illil
i 1:,
1,'1; HP 11,
-: , J ,,
~

On the other hand, the 15 cm pipe-stick is more u iJ: .., ~, I' :' 1 ;11': I
..' ..' "".. ' I ''
'11'
economical and easier to handle, and is therefore + : le! i, '

2
." lIt 1111iL': ,: I:;:
preferable for low stoping widths. The critical width- : :i I:::: :I I::' IiI1 :I:
tt ,~HIi: iIT!
to-height ratio limits the use of 15 cm pipe-sticks to '

a length of 140 cm. Beyond that length, the fre- ",,~1;jiljU:iii'ililifli!lU'i

quency of failure due to bending becomes too high. " h u: : ,I:
!!
~ : II
Fig. 7 is a typical load-deformation graph for a
t" I ti
pipe-stick of 20 cm diameter and a pipe-stick of t
+
15 cm diameter with a similar wall thickness. 0,8 1,0 1;4 1,6
(b) Wall thickness StOi>In8 WIdth In ,Mot...

The thickness of the pipe walls has a great effect on Fig. 6-Comparative costs of matpack and pipe-stick support

2000

1+-

1500 . ~~
'

z
'"
.E ,~+ r
+ :t;:'
;~ 't r
~
...
1000
j.
+,
+
,

500 '
. too
t'
-; , "
1
~1. ;
,

"
:1
,,0"+ 1

10 20
Compression In cm

Fig. 7- The deformation of 20 cm and 15 cm pipe-sticks

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY DECEMBER 1978 129
 ~

,r
.
~ . '"
t-;
1
30%
L
+,

r .
'"

20%
.. ~ rH .f
tt

10% .
"

+,
;- -
.~
~t "

Fig. 8- The distribution of the various modes of failure

their load-bearing capacity. The selection of wall deformation behaviour. If the stick protrudes too
thickness enables a designer to vary the maximum far, the end of the stick snaps at the initial peak on
load-bearing capacity between about 800 kN and the graph, the load drops off immediately, and in
1500 kN. The cost of piping increases rapidly with many cases the pipe-stick is dislodged. It has been
increased wall thickness. It is recommended that established that the maximum allowable length of
piping with a wall thickness of less than 4 mm (thin) protrusion is approximately 10 cm at each end of
should not be used for pipe-sticks. Pipes with a wall the pipe.
thickness of about 5 mm are considered most suit- (d) Distortion
able. As the pipe-stick is compressed, it yields by distort-
(c) Stick protrusion ing in various ways (Fig. 8). The ideal mode of
When the end of the stick protrudes from the pipe, distortion is the formation of a series of consecutive
the initial load rises to the maximum strength of the folds from one or both ends of the pipe (Fig. 9). The
stick and then decreases while the stick is com- wavy nature of the graphs indicates this behaviour,
pressed into the pipe. When the hanging and foot- each downward turn in the graph signifying the
wall come into contact with the pipe, it starts to start of a new fold.
carry load as well, and the total strength of the (e) Combinations of pipes and sticks
element is experienced. Fig. 7 shows typical load- By the use of various combinations of pipes and
sticks, the ultimate load-bearing capacity can be
greatly increased. However, the necessity for
greater load-bearing capacity and the increased cost
are questionable at this stage. Mention must be
made also of the possible danger of punching of the
support into the foot and hangingwall strata. When
a 15 cm pipe was inserted with a stick into a
20 cm pipe, a maximum load of 3840 kN was
achieved (Fig. 10).
(j) Rapid compression
At the Chamber of Mines Research Laboratories,
when pipe-sticks were tested under shock-load
conditions, the mode of yielding was found to be
similar to that experienced in the slow compression
tests. From the result of these tests, it can be de-
duced that pipe-sticks would stand up reasonably
well against seismic events.
Fig. 9-Deformatlon of a pipe-stick after yielding load was
(g) Underground test
reached Load~cells were installed on the top of two pipe-

130 DECEMBER 1978 JOI,JRNAL OF THE SOUTHAFRICAN.INSTJTUTE OF MINING AND METALLURGY

 ~
I
t

3500

-tT'
3000
+'
it

.,' ~1
.,'
2500
~.+

~-
2000
~: . _+1
.
+ j. ..
~ 1500 :, d

.E .:0
-H-.+
."
:; .-,
-.'
. H
...J
+ I +++4
1000 i ,'-+-
-+-
.+'
1: ..

500

10 20
Compression In cm

Fig. Ill-Loading attainable with double~pipes

2000
D

1500

z
"1000
.5

j
."

500

10 20 3'0
Compression In cm

Fig I I-Results of strain-gauge test on an in situ pipe-stick

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY DECEMBER 1978 131
 sticks in a stope at Stilfontein and monitored over only a few supports that were leaning over at the
a period of 6 weeks. Fig. 11 represents the results of time were dislodged. The hangingwall was undam-
one of these load cells installed 7 m from the face aged by the pipe-sticks.
and monitored until it was 42 m from the face, at (vi) A large number of panels have been stoped out
which stage 38 per cent closure had occurred. The successfully on pipe-sticks to date. Trouble experi-
initial flat portion of the graph is due to the load enced could be traced back to either poor installa-
cell and bottom of the stick having to make solid tion or to geological reasons.
contact with the uneven hangingwall and footwall.
The test was terminated when the load cell eventu- Problems Encountered
ally failed, shown by the downward trend at the
A number of problems have been encountered,
end of the graph. The maximum load of 1300 kN
some of which have been solved; others are in-
obtained agrees very well with that found in the
tests done at the C SIR. The flat section of the herent in the system and will require con-
graph from C to D on the underground installation, tinuous attention.
which is absent in the laboratory tests, suggests (a) By far the major problem, and often the cause of
time-dependent failure and creep, which occur early failure due to dislodgement, is the angle of
underground but not under 'rapid' loading condi- installation. The pipe-stick must be installed
tions such as in the laboratory. Deformation, tilting, accurately at right-angles to the hangingwall ;
and failure mode were recorded. At the end of the otherwise, convergence of hangingwall and footwall
test, the total tilt was 200. forces the pipe-stick over until it is dislodged.
Correct training and strict supervision are essential
Performance in Practice to ensure that pipe-sticks are installed at the
correct angle.
During the first installation period, 21 pipe-sticks (b) The end of the stick snaps when it protrudes too
were installed in 23-23B stope at Stilfontein Gold Mine far from the end of the pipe. This has been overcome
in January 1976. This stope was examined regularly, by limiting the protruding end of the stick to 10 cm
and the behaviour of all the pipe-sticks was recorded on each side, by securing the pipe in position with
until closure was greater than 50 per cent. Similarly, a a nail, or by deforming the end of the pipe slightly.
stope (28-49) at Buffelsfontein Gold Mine where pipe- (c) When broken ground near dykes and faults is
sticks were being used as support in four panels was encountered, part of the hanging is often lost.
visited regularly over a period of six months (March to Although pipe-sticks appear to be a better support
September 1977). Other stopes were examined whenever in this case because of their stiffness, it is found that
the opportunity arose, some salient features being ob- production personnel tend to favour matpacks.
served. (d) Lack of adaptability by stope workers is initially
During these visits, the following valuable information one of the major problems. Owing to the apparently
was obtained with respect to the behaviour and effective- less-dense installation, people feel unsafe in the
ness of pipe-sticks. stopes. The tendency is then to install pipe-sticks
(i) Pipe-sticks can yield in a number of ways such as
at a greater density or to revert to conventional
curling, folding, splitting, or punching into the support at the slightest change of conditions.
footwall. Under all these modes of yielding, the
pipe-sticks deform in a stable manner and maintain Conclusions
a reasonable supporting load.
(ii) Some pipe-sticks bend and then lose their load, but 1. The steep load-deformation curve prior to yield and
an insignificant number of these appear in the the constant load characteristics after yield mean
vicinity of the working face. This mode of failure that pipe-sticks are a very effective face support.
normally occurs about 15 m from the face, where it 2. When it is necessary to keep the stope open for a
is not of great importance provided it is restricted to distance greater than 40 m from the face, pipe-sticks
individual supports. are not suitable owing to their low ultimate loading
(iii) The failure of pipe-sticks is negligible up to 30 per characteristic.
cent convergence (i.e. 0,3 m) in low stoping widths. 3. Pipe-sticks reacted favourably to simulated rock-
When the convergence increases beyond 30 per burst conditions, and successfully stood up during
cent, a sharp increase in failure is observed. At a a large seismic event at Buffelsfontein Gold Mine.
convergence of more than 50 per cent, the majority 4. The overall cost and labour-saving features make the
of pipe-sticks become ineffective. use of pipe-stick support very attractive.
(iv) Pipe-sticks installed ahead of the blasting barricade 5. Good supervision and correct installation are essen-
are not damaged and, if securely installed, are not tial if ineffective support and early dislodgement are
easily blasted out. to be avoided.
(v) A seismic event with a Richter magnitude of 3,7
occurred at Buffelsfontein Gold Mine during July References
1977. The epicentre was in the vicinity of a stope 1. WIGGEL, R. R. The effects of different support methods on
that was supported on pipe-sticks at that time. The strata behaviour around stoping excavations. S. Afr. Inst.
pipe-sticks withstood the shock of this event very Min. Metall., Symposium on Rock Mechanics and Strata
Control in Mines, Apr. 1963-Jun. 1965. pp. 1-35.
well. They deformed in the normal manner, and 2. MoRE O'FERRALL, R. C., and KERSTEN, R. W. Stress and

132 DECEMBER 1978 JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY