THE 25mm GARFORD BULB ANCHOR FOR CABLE

BOLT REINFORCEMENT
PART 1: LABORATORY RESULTS

A.J. HYETT

W.F. BAWDEN

Department of Mining Engineering,

Queen's University
Kingston,
Ontario.
K7L 3N6

N. HEDRICK
Garford Pty. Ltd.
3/24 Attwell Street
Landsdale
Western Australia.

Submitted to the CIM Bulletin

MAY 1994
ABSTRACT

The Garford bulb anchor is a modified cable geometry developed and patented by
Garford Pty of Australia, and which has recently become widely available in Canada. It is
produced by simply deforming the 7 wires (including the central kingwire) of a standard
cable into a bulb1 (see Figure 1). Laboratory cable pull tests indicate that the bond
capacity of the 25mm Garford bulb anchor is comparable to the nutcase cable, in excess of
the birdcage cable, and significantly higher than standard 7-wire strand. Its comparative
advantage over the latter is most significant for tests at low radial confinement and high
water:cement ratio grout, which corresponds to operational conditions of low quality
rock (ie: weak, highly fractured rock masses, foliated hanging walls, failed ground etc.)
and poor installation quality control (perhaps the most common explanation for cable
bolt failures). It is under such conditions that cable bolting is often ineffective. The
most impressive result is that for a reduction of the effective rock mass modulus by a
factor of five, the Garford bulb cable capacity remains virtually unchanged. For tests
conducted at a 0.4 w:c ratio grout the dominant failure mechanism involved rupture of
the strand (tensile strength of the strand = 250kN) and as such the capacity of Garford
Bulb Anchor was independent of the effective rock mass confinement. Field test results
will be presented in a companion paper.

1 As a matter of terminology, this ungrouted configuration is called a bulb, and the grouted
configuration a bulb structure.
1. INTRODUCTION

Perhaps the most important development in recent cable bolt research has been to
highlight the limitations of conventional 7-wire strand. Both laboratory and
underground tests have demonstrated that cable bolt failure occurs primarily due to slip
at the cable-grout interface, and that the bond strength is associated with frictional-
dilational rather than adhesional resistance. Cable pull tests conducted at constant radial
pressure in a Modified Hoek cell (Hyett et al., 1994) indicated that the radial dilations
induced during bond failure, which are responsible for the frictional strength of 7-wire
strand are minuscule: 10-20µm for higher radial confining pressures. Herein lies both
the advantage and disadvantage of conventional 7-wire strand. In very good quality rock
masses, under favourable mining conditions, it represents an effective yielding support
system, mobilizing around 60-80% of the strength of the steel for a 10"(250mm)
embedment length. However, in poor quality ground, or where the rock mass is
destressed or has failed - namely in cases where cable support is often of critical
importance - excessive borehole deformation or mining induced borehole relaxation may
result in very low capacities. Such incidences may lead some engineers to falsely conclude
that cable bolting is always ineffective. In such circumstances, modified cable geometries,
which force much higher radial dilations, will guarantee higher bond strengths which are
less sensitive to installation quality control, rock mass characteristics and mining induced
effects (ie: stress redistribution around openings).

The Garford Bulb Anchor is a modified cable geometry developed and patented by
Garford Pty Ltd in Australia. It is produced by simply deforming the 7 wires (including
the central kingwire) of a standard cable into a bulb1 (see Figure 1). This can be done
without reducing the tensile strength of the strand. The versatility of the manufacturing
process allows both the bulb diameter and the bulb spacing to be readily varied.
Preliminary laboratory test results by CSIRO and Strata Control Technologies indicated
that the bond strength of the Garford bulb cable may be significantly higher than
standard cable and birdcage cable. The objective of the laboratory test programme
presented in this paper, was to determine the bond strength of the Garford bulb anchor
under different radial confinements, different water:cement ratio grouts, and, in
particular, to evaluate its prospective behaviour under adverse conditions of low radial
confinement (ie: in low quality, failed or destressed rock masses) and high water:cement
ratio grouts (poor operator quality control) that are all too common in many mines
worldwide. In a companion paper, Part 2, results from a field test programme will be
presented.

1 As a matter of terminology, this ungrouted configuration is called a bulb, and the grouted
configuration a bulb structure.

3
Figure 1. The Garford Bulb Anchor geometry

4
2. THE LABORATORY TEST PROGRAMME

The experimental setup is shown in Figure 2c, and is contrasted with the standard
laboratory pull test setup used by previous workers (Fuller and Cox, 1975; Goris, 1990)
(Figure 2b). All tests were conducted with a 300mm embedment length, the bulb being
located at the midpoint before testing. The pull rate was maintained constant at 0.3
mm/s with a servo-controlled MTS testing frame.

From an operational perspective, the test programme was intended to evaluate the bond
capacity of the Garford bulb anchor under various conditions of installation quality
control and rock mass quality. Therefore, as outlined in Table 1, two experimental
parameters were varied:
(i) the w:c ratio of the grout (see Table 2 for details and compare with Hyett et al.,
1992b)); and,
(ii) the radial stiffness of the confining pipe (see Table 3 for details).
The purpose of using confining pipes of different radial stiffness was to simulate different
in situ rock mass conditions. A correlation between laboratory and field test was
demonstrated by Hyett et al. (1992a).

Twelve tests were conducted for the 25mm Garford bulb (see Table 1), as well as twelve
additional tests for the larger 40mm bulb. These included tests at a 0.5 w:c ratio (poor
grout quality control), and using Sch. 40 Aluminum pipe (very low radial confinement
ie: low rock mass quality).

5
Figure 2. (a) Mining scenario and the laboratory pull test setups intended to simulate it:
(b) the conventional pull test and (c) the modified push test used in this study.

6
 Test # w:c ratio Confinement Embed. Length

1 (A-D) 0.4 Sch 80 - Alum 300mm

2 (A-D) 0.4 Sch 80 - Steel 300mm
3 (A-B) 0.4 Sch 40 - Alum 300mm
4 (A-B) 0.5 Sch 80 - Alum 300mm

Table 1. The 25mm bulb test programme. In addition these tests were duplicated for
the 40mm bulb.

w:c ratio Pd (28 day) UCS E V

(g/cm3) (MPa) (GPa)

0.40 1.96-1.99 58 - 69 11.5-12.7 0.18

0.50 1.85 - 1.81 42-51 8.9 - 9.7 0.18

Table 2. Properties of the Type 10 portland cement grout. P d is the bulk density after
28 day cure: UCS - Uniaxial compressive strength according to ASTM C39-84. E -
Young's modulus. v - Poissons ratio.

Pipe Spec. I.D. O.D Kr Equiv Rock Mass

(mm) (mm) . (MPa/mm) Modulus (GPa)
Sch. 80 - Steel 49.30 60.13 1604.2 49.7

Sch. 80 - Alum 49.02 60.45 599.5 17.7

Sch 40 - Alum 52.22 60.38 332.5 10.3

Table 3. Dimensions and radial stiffnesses of the pipes used to provide radial
confinement Kr is the radial stiffness of the pipe, which may be thought of as the pressure
in MPa required to induce 1mm of radial deformation. The number on the RHS
provides an indication of the equivalent rock mass Young's modulus. A value of 49.7
represents a relatively good quality rock mass (Q' = 10-25) at surface, or a fair/good
quality rock mass (Q' = 5-15) at depth, 10.3 represents a fair (Q'=1-5) quality rock mass
at surface, or a poor (Q'=0.5-1) rock mass at depth.

7
3. LABORATORY TEST RESULTS AND INTERPRETATION

The full load displacement plots are presented in Figure 3. Comparable results for a
standard cable are presented by Hyett et al. (1992a). Failure always occurred in the test
("top") section and never in the anchor ("bottom") section (refer to Figure 2). Using the
results of calibration tests, all deformations associated with the bottom section and the
test frame have been subtracted, so that the displacement along the abscissae of the plots
in Figure 3 are exactly those at the exit point of the cable (see Figure 2). A
comprehensive summary of the results is presented in Table 4.

Test # Peak Load Failure Broken

Mechanism Cable?

1A 254 #1 N
1B 258 #1 Y
1C 260 #1 Y
1D 247 #1 N

2A 253 #1 N
2B 259 #1 Y
2C 262 #1 N
2D - - -

3A 252 #1 Y
3B 237 #1 N

4A 191 #1 N
4B 231 #1 N

Table 5. Comprehensive summary of the results.

Failure Mechanisms

Below 25mm of axial displacement, failure involved axial displacement of the bulb
structure through the surrounding cement annulus by the mechanism illustrated in
Figure 4. Thereafter, for specific tests the bond strength at the cable-grout interface was
sufficient to exceed the tensile strength (250 kN) of the 7-wire strand. For tests in the
lowest confinement (Sch. 40 Aluminum), rupture of the cable did not occur, and a
noticeable hardening was observed, which may correspond to an increase in the wedging
action of the bulb as it was progressively displaced through the grout annulus. After
25mm of axial displacement failure was unstable, either due to (i) the rupture of
individual strands, or (ii) stick-slip displacement of the bulb through the cement annulus.
For tests in the lowest confinement (Sch. 40 Aluminum), rupture of the cable did not
occur, though a noticeable bulging of the pipe was observed, and in one case the pipe
actually split (Figure 5). This clearly indicates the potential dilatancy that can be induced
during bond failure of the Garford Bulb Anchor. Figure 6 shows successive sections
through failed samples. Although the grout within the bulb structure is highly fractured,
there was no evidence that the bulb was able to collapse.

8
9
Figure 4. The progressive sequence of failure

10
Figure 5. Specimen 3B: the aliminium Svh. 40 pipe redially split.

The effect that each of the two experimental parameters had on the bond capacity of the
Garford bulb anchor will be discussed below.

Effect of radial confinement

A comparison between Figures 3a, 3b and 3c indicates that radial confinement had
almost no effect on the bond capacity of the 25mm Garford bulb. A peak load of 240kN
was obtained even at very low radial confinement (Sch. 40 Aluminum). The distinct
bulge of the confining pipe for the Sch. 40 Aluminum, and to a lesser degree the Sch. 80
Aluminum, indicates the much higher potential for radial dilation induced during failure
of the Garford bulb than a standard 7-wire strand. Consequently, the confining
pressures generated in response to radial dilation and the corresponding bond capacity
attained are relatively independent of the radial stiffness of the confinement.

Effect of grout water:cement ratio

Even for a 0.5 w:c ratio grout (Figure 3d) high bond capacities (190kN and 230kN) were
attained. Again, this emphasizes that the potential for radial dilation during bond failure,
which induces high radial confining pressures, is able to compensate for the lower
strengths, lower Young's modulae, and higher shrinkage strains associated with higher
water:cement ratio grout.

Effect of Bulb Size

Intuitively one would expect an increase in bond strength corresponding with larger bulb
diameters, because of the greater mechanical interlock between the grout and the cable.
However, whereas the 25mm bulb resulted in capacities close to the tensile strength of
the strand and in many cases failure involved rupture of the steel1, for the 40mm bulb
bond failure occurred at lower loads (150 - 210 kN).

1 Rupture occurred wire by wire, either at the exit point or internally within the test section,
presumably within the bulb.
11
In view of this unexpected result, following testing to failure, several of the specimens
were sectioned (ie: cut perpendicular to the cable axis) for visual observation. This
categorically established that the grout had fully penetrated the bulb structure, but more
importantly revealed that a different failure mechanism had operated for the 40mm bulb.
In contrast to the behaviour for the 25mm Garford bulb described above, this involved
slip and twist of the individual wires through the bulb structure, so that, although the
cable had been pulled in excess of 50mm, the centre of the bulb structure remained in its
original position (Figure 7). The unscrewing characteristic of this failure mechanism can
be reconciled with the observation of cable "splaying" at the entry point (ie: free end) of
the test section. The corresponding bond capacities were lower and considerably more
scattered than for those for the 25mm. Therefore, for underground applications, it is
recommended that the bulb size should be restricted to 25mm.

Figure 7 Sections through a 40mm specimen following failure

12
DISCUSSION AND CONCLUSIONS

Comparison with other types of cable bolt

In Table 5, the results presented in Table 4 have been averaged and normalized with
respect to the bond capacity (or predicted bond capacity) for standard cable (Hyett et al.,
1992a). Note the marked comparative advantage of the Garford Bulb cable under
conditions of low radial confinement and high water:cement ratio (weak) grout. It is
interesting to note that, for the Garford bulb, a reduction of the effective rock mass
confinement by a factor of five, (Test Series GB1, GB2 and GB3), had minimal effect on
cable bolt capacity. Reduction of the grout from 0.4 w:c to 0.5 w:c (GB4) had a more
significant effect, albeit less than the equivalent test for a standard 7-wire strand. This
indicates that the operator should strive for optimal grout quality control irrespective of
the type of cable being used. Conventional strand will, however, be much more sensitive
to poor grout quality control than will the Garford bulb anchor. The use of a 0.35 - 0.4
w:c grout is recommended with the Garford bulb cable in order to ensure penetration
into the bulb structure; in other words slightly thinner than the 0.35 w:c ratio
recommended for the installation of standard 7-wire strand. Pumping a grout in this
range should be routinely possible using most grout pumping systems, and will remain in
upholes without the use of plugs.

Table 6 provides a comparison between laboratory determined cable capacity for the
25mm Garford bulb anchor cable and the nutcase cable (Hyett et al., 1993) for identical
confinement and grout quality test conditions. The geometric similarity between the two
types of cable is reflected in almost identical bond strengths and failure mechanisms. The
nutcase cable has already proven extremely effective during field tests and is currently
being used in several Canadian mines. At present no field tests have been conducted for
the bulb cable, although, based on the results in this report, its prospects are considered
to be excellent.

13
 Confinement

Water:cement
ratio Sch. 80 Steel Sch. 80 A1 Sch. 40 A1

0.4 2.5 3.33 4.0

0.5 3.5

Table 5. The comparative advantage of a 25mm Garford bulb over standard cable based
on 300mm embedment length tests of a single bulb. A factor of 4 means the Garford
Bulb has four times the peak load of single cable.

Confinement

w:c ratio Sch. 80 A1. Sch. 40 A1.

0.4 0.94 1.03

Table 6. A comparison of mobilized capacity for 25mm Garford bulb to a comparable

nutcase cable. A factor of 1.03 means the Garford bulb anchor has 1.03 times the peak
load capacity of the nutcase cable.

Implications for cable bolt design

From a design perspective, based on the results presented above, it is expected that failure
of cable bolted ground reinforced using the 25mm Garford bulb cable will most likely
involve rupture of the cables rather than bond failure. In such cases, the mining engineer
knows the capacity of his cable bolt system (250kN or 25 tonnes per cable) and is able to
modify his design, either by using twin Garford bulb cables or changing the cable bolt
spacing, based on this value. In contrast, for failures of cable bolted ground involving
bond failure of conventional cables, the only possible effective action is to improve grout
quality, and even then, the quality of the ground and the mining induced stress changes
may dictate that no engineering solution exists. Furthermore the engineer doesn't know
whether the bond capacity is 10 tonnes or 20 tonnes.

14
In summary, the most important advantages of the Garford bulb include the following:

(1) High bond capacity: the Garford bulb anchor and most modified cable geometries,
offers distinct advantage over conventional cable under the following conditions:

- highly fractured, weak rock masses

- yielded (ie: failed) ground
- conditions where significant mine induced stress change (reduction) occurs normal to
the cable axis.

The above are typical of many mining conditions (eg: destressing of stope hanging walls;
yielding of stope backs and pillars etc).

In addition, although increasing the w:c ratio of the grout decreases the capacity of both
the Garford bulb anchor and standard strand, the decrease in capacity is much less critical
for the Garford Bulb Anchor. At 0.5 w:c ratio grout the Garford Bulb Anchor attained
more than three times the capacity of standard strand.

(2) Replace surface fixtures: It is believed by the authors that modified geometry cables
can largely replace the need for surface cable fixtures (ie: plates and/or straps) as the bulb
closest to the hole collar should act as an effective surface fixture. This suggests that
recessed cables and/or cable bolts placed from remote access (ie: hanging wall cable bolt
drifts) should prove much more effective with modified cable than they have in the past
with conventional strand. It is also important to recognise that conditions of high
stresses and competent, high quality rock often result in stress slabbing of the excavation
periphery. This reduces the effective cable embedment length and the increased capacity
of the bulb anchor may be critical in maintaining such highly fractured rock in place.

(3) Design flexibility: The versatility of the manufacturing process, and in particular the
fact that the bulb spacing can be varied, will offer a definite advantage with respect to
specific design problems. Under specific conditions of very competent rock and very
high stress the stiffness of modified geometry cables may be a concern (note: depending
on the characteristics of the rock mass, Garford Bulb Anchor cables are roughly twice as
stiff as conventional strand under high confining conditions). With the Garford bulb
anchor such concerns can be addressed through appropriate design modifications. Bulb
anchors can be strategically located while other sections of cable are then debonded in
order to allow required rock mass movement. Alternatively, it may be sufficient to
simply vary the bulb anchor spacing.

15
5. REFERENCES

Fuller, P.G., & Cox, R.H.T., Mechanics of load transfer from steel tendons to cement
based grout. Proc. 5th Aust. Conf. on the Mechanics of Structures and Materials,
Melborne, pp. 189-203, 1975.

Goris, J.M., Laboratory Evaluation of Cable Bolt Supports. 92nd Annual General
Meeting of the CIM, Ottawa, May, 1990.

Hyett, A.J., Bawden, W.F. and Reichert, R.D., The effect of rock mass confinement
on the bond strength of fully grouted cable bolts. Int. J. Rock Mech. & Min. Sci.
& Geomech. Abstr. 29(5), pp. 503-524, 1992a.

Hyett, A.J., Bawden, W.F. and Coulson, A.L. 1992. Physical and mechanical
properties of normal Portland cement pertaining to fully grouted cable bolts.
In: Rock Support in Mining and Underground Construction, Kaiser, P.K., and
McCreath, D.R. (eds.), 1992b.

Hyett, A.J., Bawden, W.F., Powers, R. and Rocque, P. The nutcase cable bolt. In:
Innovative mine design for the 21st century eds. W.F. Bawden and J.F. Archibald,
1993.

Hyett, A.J., Bawden, W.F. and MacSporran, G.R. In prep. to be submitted to the Int.
J. Rock Mech. & Min. Sci. & Geomech. Abstr., 1994.

16