Characteristics of Different types of Support Systems for Static and Dynamic Conditions

HTML> Characteristics of Different types of Support Systems for Static and Dynamic Conditions

Characteristics of different types of support systems in rockburst conditions

X. Yi

Flairbase Inc., 6600 Trans-Canada Highway, Suit 519, Pointe-Claire, Quebec, H9R 4S2

ABSTRACT: The load-displacement curves from testing of support elements are used to assess the mechanical characteristics in static and dynamic loading conditions. Element Load Capacity (LC) and Energy Capacity (EC) are defined and obtained from test curves. These two parameters reflect the mechanical performance of an element in static and dynamic conditions, respectively. A simple method is devised to estimate the mechanical characteristics of support systems based on element properties. A CAD computer program as well as a database is designed to automate this procedure. The load-displacement curves from the literature and the typical support levels found in Canadian hard rock mines are used to demonstrate that the EC-LC point graphs can be used for comparisons of different support elements and systems. Therefore, a EC-LC point graph can be used to assist in the support selection and design in rockburst conditions.

1 INTRODUCTION

Most underground support design procedures are derived from case studies. Examples are the Q-based NGI support design procedure for permanent underground openings by Barton, Lien and Lunde [1], and the Q-based cable bolt design chart for temporary underground stopes by Potvin and Milne [2]). Although the factor of safety method is often not explicitly used due to the variability of the rockmass material, the method is the basis for any empirical procedure. Barton et al. [1] used the method to rationalize the design tables based on the collected data, i.e., the ground pressure exerted by the rockmass was compared to the capacity of rockbolts and concrete lining.

The mechanical properties of support elements and systems must be known if the factor of safety analysis is to be performed. In cases where the loading conditions can not be definitely determined in a rockmass, the comparison of different support systems can assist in the selection and design of the most appropriate support. Other than the mechanical properties, consideration must also be given to factors such as corrosion protection, cost of material, and time and ease of installation.

This paper focuses on the load-displacement characteristics of support elements. The load and energy capacity (LC and EC) values of the elements are obtained from test curves. Since testing of support systems is difficult, if not impossible, to perform, a simple method is devised to estimate the

mechanical characteristics of support systems based on element properties. This procedure allows one to:

1. use a EC-LC point graph for elements to pick the appropriate elements for a support system,

2. use a EC-LC point graph for systems to compare different systems, and

3. answer the question "what are the load and energy capacities of your support system?".

An AutoCad-based computer program is designed to implement the procedure. The advantage of a CAD program includes visualization and flexible user-computer interaction. The user can pick load-displacement curves to obtain the element EC and LC values, and points in the EC-LC element point graphs to obtain the system EC and LC values. The load-displacement curves and the points are linked to a database where other information, such as detailed testing conditions, element cost, time of installation and manufacturer, is stored. This database can be used to pin-point the specific load-displacement curves.

2 SUPPORT DESIGN RATIONALE

The rock ejection model may be used for analyzing rock failure and support design in rockburst conditions (Wagner 1984, Yi and Kaiser 1993). This model recognizes that a rock block can be ejected by seismic waves, and the energy of the ejected block must be dissipated with the help of the support. The rock block in the model may be a single rock block or a volume of fractured rockmass ejected as an entity.

Laboratory impact tests demonstrated that the energy balance approach can be used for the rock ejection model (Yi and Kaiser 1994). In this approach, the sum of imparted energy terms equals the sum of consumed energy terms. The imparted energy consists of the kinetic energy of an ejected rock block, and the gravitational potential energy released by the rock block as it moves downwards (relevant mainly for excavation roof). The consumed energy terms include the energy dissipated in rock fracturing and friction between the ejected rock block and surrounding rockmass, the energy dissipated in friction between support element and the grout or borehole wall, and the energy absorbed in the elastic and inelastic deformation of the support elements. The objective of support design is, therefore, to exploit the energy absorbing capabilities of a combination of support elements and to help the rockmass absorb the imparted dynamic energy. In practical terms, the opening integrity must be maintained to control the release of gravitational potential energy, while a sufficiently large displacement is accommodated to allow for energy dissipation in rock fracturing. In short, a support system must be designed to:

(i) maintain opening integrity while permitting a sufficiently large displacement, so as to minimize rock ejection, and

(ii) contain the ejected rocks.

These requirements were recognized by Ortlepp (1983), Wagner (1984), Roberts and Brummer (1988), and others.

In order to resist dynamic loading, a support system must have a high energy absorbing capability, a high displacement accommodating capability, and a high degree of areal coverage. However, in rockburst conditions, the loading is not necessarily dynamic alone. Dynamic rock failure can result in a fractured rockmass which exerts static pressure on the support. The support must also have a sufficiently high load bearing capability. In other words, the design principles for static loading should be followed or combined with those for dynamic loading. The following two examples illustrate this type of considerations. In the next section, support requirements will be formulated in terms of the relative magnitudes of rockbursts.

Example 1: In the side walls of an excavation, the gravitational potential energy of a rock block can be ignored if the wall integrity is maintained. The frictional resistance of the surrounding rock mass increases with increasing block depth (in a direction normal to the opening surface). Consequently, a very deep rock block (with depth larger than about one third of opening size) may impart little dynamic energy to the support. Shallow rock blocks can be tied together to form a deep block using short but densely placed rockbolts. Ortlepp (1983) described a tunnel subject to a large rockburst, where 16 mm diameter and 2.5 m length short grouted bolts spaced at 1m x 1m were combined with 6-strand 7.5 m long pre-tensioned cable anchors spaced at about 2.8 m. The long cable anchors were designed to absorb the dynamic energy.

Example 2: A deep block in the may behave as a "dead weight" if its fall is triggered by a seismic event. For a support element with low load and energy capacity, such as the standard Split Set, the energy absorbed may not be sufficient to compensate for the released gravitational potential energy as the block moves downwards. This type of support may not be effective in the excavation roof if used alone. Elements with higher load bearing capabilities should be used.

3 PARAMETER DEFINITIONS AND

FORMULATION OF SUPPORT DESIGN

RATIONALE

3.1 Support Elements

The load-displacement curve for a support element contains all the element properties needed in either an analytical or an empirical design. Parameters should be defined to assist in analysis, to compare different support systems, and to formulate the support design rationale. In the present procedure, it is assumed that a testing is performed with point load and the displacement measured at the load point.

It is generally accepted that the load capacity (LC) of a support element is the controlling parameter for static loading. LC is simply the peak load on the load-displacement curve.

For dynamic loading, the previous section showed that the energy capacity (EC) and the displacement capacity (DC) are the controlling parameters. The displacement capacity DC is defined as the

Fig.1 Definitions of energy, load, and displacement capacities (EC, LC, and DC) for support elements.

displacement sustained before either element rupture or loss of opening integrity (Fig. 1). For practical purposes, it equals either the displacement at element rupture, or, an empirical limit value for those non-rupturing elements which relies on frictional resistance. Wagner (1984) suggested a 300 mm limit for stopes in South African gold mines. In this paper, this 300 mm limit is adopted. The energy capacity EC is defined as the area under the load-displacement curve bounded by the above defined displacement capacity (Fig. 1). Since the element displacement capacity DC is not independent of both the load capacity LC and the energy capacity EC, DC is not explicitly included in a comparison procedure.

For rockburst conditions and certain static conditions (poor rockmass quality), the degree of areal coverage is critical. To deal with this parameter, support elements are grouped into two categories: in-rock elements which are installed inside the rockmass, and surface elements which are installed on the excavation surfaces. In-rock elements, such as a grouted rebar, perform the reinforcing and holding functions, while surface elements, such as the shotcrete, perform the areal coverage and retaining functions.

In terms of the defined parameters, the support design rationale can be translated into the following:

(i) The higher the LC, the more resistant the element is to static loading, and the higher the EC, the more resistant it is to dynamic loading.

(ii) The higher the degree of areal coverage, the more resistant the system is to rockbursts.

In terms of the relative magnitudes of rockbursts and the defined parameters, the support design rationale outlined in the previous sections are formulated as shown in Table 1. Some explanations follow.

In low in situ stress static conditions, it is desirable to restrict rockmass deformation to the lowest degree. Fuller (1978) suggested that rock reinforcement design should be based on control of joint displacement, so as to increase joint strength and help the rock support itself. On the other hand, in very high static stress conditions or situations where large displacement is induced due to large excavation spans and/or soft rockmass, it is necessary to accommodate the large displacement such that rock stress can be transferred to the surrounding competent rockmass with increasing displacement. Bywater and Fuller (1983) designed cables with 2 m long debonded segments at the free surface to allow for increased displacement for the hanging wall of a large open stope at Mount Isa Mines Limited, Australia. In an experimental tunnel

at 2600 m depth at Buffelsfontein Gold Mine in South Africa, side wall convergence as large as 421 mm was measured as mining front passed by the

Table 1 Support requirements for support elements and systems

(a) In rockburst conditions

Relative burst magnitude	Small strain bursts	Medium to large bursts
EC	low	high
LC	high	low - high
Areal Coverage	high	high
System EC	low	high
System LC	high	high

(b) In static conditions

Relative rock deformation	Small rock deformation	Large rock deformation
EC	low	high
LC	high	high
Areal Coverage	low	low - high
System EC	low	high
System LC	high	low - high

tunnel. This demonstrates that the support requirement for large rockbursts is similar to that for static loading with large deformations.

In conditions where rockburst occurrence is frequent but the magnitude of each is small (e.g., strain bursting on the surfaces of openings), it may be desirable to prevent bursting by providing a sufficiently large confinement. Support elements of high load capacity serve this purpose. Hence, support design principles for static loading should be applied to this type of low intensity rockburst conditions.

3.2 Support Systems

The response of a support system depends on the behavior of individual elements, properties of connections between in-rock and surface elements, and interactions between the rockmass and the support. In reality, the elements in a system act neither in parallel nor in series, but, in a complex manner. The system energy capacity is defined as the sum of EC values for all elements. The system load capacity is defined as the weighted sum of in-rock elements (with a weight of 1) and surface elements (with a weight of 2) for the following reasons: (i) in-rock and surface elements complement one another in a system, and (ii) LC for surface elements under uniform loading (more realistic) is approximately twice that under point loading (testing condition).

Support requirements for support systems in terms of the defined parameters are the same as for elements (Table 1). Elements of high EC can be combined with those of high LC to form a system of both high EC and LC. Again, the degree of areal coverage for a support system can not be quantified. This should not affect the comparisons of realistic support systems based on the EC-LC point graphs, since the degree of areal coverage is approximately represented by the EC and LC value of surface elements, and a surface element is nearly always present in support systems in rockburst conditions.

4 RESULTS FROM PUBLISHED DATA

Limited published load-displacement curves from laboratory and field static tests of different types of supports are available, and those from dynamic testing are rare. In the following sub-sections, a few published results are used to illustrate the procedure. The procedure is automated by using an AutoCad-based computer program. The results are presented to help understand the procedure. For engineering support selection and design, it is recommended that more relevant curves, e.g., from field testing at the specific mine or mines of similar rockmass and stress conditions, are used.

The EC and LC parameters defined in the previous section depend on specific testing techniques and conditions. Tests of the same support element but using different test techniques and conditions are often not comparable. However, with the same testing technique and conditions, different support elements can be compared. For the purpose of comparing different support elements, it is assumed that the performance comparisons of different types of support, derived from static tests, are valid for rockburst conditions.

4.1 In-Rock Elements

Ortlepp (1983) conducted static tests on grouted support elements (Table 2). A thick-walled steel

Table 2 Brief description of the cement grouted in-rock elements tested by Ortlepp (1983) using steel cylinders

No.	Description of support element	Displ. capacity DC (mm)
A	f16mm end anchored and grouted bolt	300*
B	6 f7mm smooth rods	300*
C	f16mm 3x7 strands	23
D	f22mm old hoist cable	300*
E	f13mm old hoist cable	19.7

Note: * The 300 mm displacement limit is chosen for

non- rupturing elements.

cylinder cut in halves was used to simulate two separating rock blocks. The EC-LC points are plotted in Fig. 2a (marked A, B, C, D and E). Compared to the other elements in this testing program, Element B (cement grouted multi-rods, partial slip) has both high LC and EC values. This makes the element suitable for both static and dynamic conditions. Elements C (cement grouted cable) and E (cement grouted old hoist cable) may be suitable only for static loading due to the low EC values. These two elements ruptured in the testing.

Stillborg (1990 and 1993) performed static tests on twelve different support elements using two concrete blocks simulating separating rock blocks (Table 3). The length for all elements was 3 m. Data measured from load-displacement curves are

Table 3 Brief description of the in-rock elements tested by Stillborg (1990) using concrete blocks

No.	Description of support element	Displ. capacity DC (mm)
I1	f17mm mechanical	49
I2	f24.5mm tubular, post grouted	107
I3	f20mm cement rebar	36
I4	f20mm resin rebar	26
I5	f22mm resin fibreglass	30
I6	f30mm cement 2x7 strand cable	51
I7	Cement 2x7 strand Birdcage cable	13
I8	f39mm Split Set	> 150*
I9	Standard Swellex	12
I10	Yielding standard Swellex	> 150*
I11	Super Swellex	30
I12	Yielding super Swellex	>150*

Note: * the symbol ">" indicates that the loading

was discontinued. DC is set to 300 mm and the corresponding EC is calculated assuming constant LC.

plotted in Fig. 2a (marked I1 to I12). Compared to other elements in this testing program, element I6 (grouted cable, bond slip) and I12 (yielding Super Swellex) have both high LC and EC values. These two elements are suitable for both static and dynamic conditions. Element I7 (2x7 strand grouted Birdcage cable, cable rupture) and I5 (resin fiberglass) have very low EC values but high LC values. They are only suitable for static loading. Element I9 (standard Swellex) has a very low EC value, which is not suitable for dynamic loading. I8

(Split Sets) has a high EC but a low LC value. It may be effective in the walls for rockburst support, but not in the roof due to static gravitational loading.

For an in-rock element with both a free length and a bond length in the rockmass, such as the rock anchor used in civil engineering, EC value depends on the free length where stretching occurs, if the bond length is sufficient to prevent bond slip. EC

can be calculated from energy density (energy per unit volume) and ultimate strain of steel. For different grades of steel bars, the energy density and ultimate strain parameters have different values. A cable from Scan Rope, Sweden (Stillborg 1986) with a nominal ultimate strength of 812 MPa and an ultimate strain of 3 %, for example, EC = 7.7 kJ and LC = 255 kN per unit meter free length. The installation of cables in the rock mass is not restricted by opening size. If the free length is increased to more than 3 m, it can be effective for both static and rockburst conditions.

4.2 Surface Elements

Pakalnis and Ames (1983) performed in-situ tests on 1.2 m x 1.2 m panel mesh (Table 4). Parameters obtained from load-displacement curves plotted in Fig. 2b (marked M1 to 4). Compared to the in-rock

Table 4 Brief description of mesh tested by Pakalnis and Ames (1983) using point-load

No.	Description of mesh *	Displ. capacity DC (mm)
M1	Weld, 4 gauge wire 25.8 mm2, aperture 102 mm	310
M2	Weld, 9 gauge wire 11.1 mm2, aperture 102 mm	362
M3	Chain, 9 gauge wire 11.1 mm2, aperture 51 mm	425
M4	Chain, 11 gauge wire 7.3 mm2, aperture 51 mm	435

Note: * "weld" for weld mesh and "chain" for chain-

link mesh.

elements in Fig. 2a, all the mesh elements have low LC and EC values. Surface and in-rock elements can complement each other in a support system in which the former provide superior areal coverage and the latter provide high EC and/or LC.

Kirsten and Labrum (1990) tested 1.6 m x 1.6 m (bolts spaced at 1 m) panels of mesh and fibre reinforced shotcrete (Table 5). EC-LC points are plotted in Fig. 2b (marked S1 to S6). Mesh and

Table 5 Brief description of reinforced shotcrete panels tested by Kirsten and Labrum (1990) using point-load

No.	Type of reinforcement *	Thickness (mm)	Displ. capacity DC (mm)
S1	Mesh	73	108
S2	Mesh	148	113
S3	Mesh	180	125
S4	Fibre	52	142
S5	Fibre	104	129
S6	Fibre	141	134

Notes: * Approximately 1.6 m x 1.6 m panels with 4 bolts in 1m x 1m pattern. Mesh was 75 mm aperture and 3.1 mm wire diameter. Steel fibres were Dramix ZP, 30mm long and 0.5mm diameter.

fibre reinforced shotcrete plates thicker than 140 mm (S2, S3 and S6) have higher EC and much higher LC values than for the mesh in Table 4. The reinforced shotcrete should perform better than mesh alone, particularly in static loading conditions. However, the reinforced shotcrete panels, thinner than 73 mm, have lower EC values than for mesh alone. It must be noted that shotcrete has a higher degree of areal coverage than mesh alone. Shotcrete may strengthen fractured rock on opening surface via cement penetration into the rock fractures, and protect mesh from corrosion. It is more tolerant to damage caused by production machinery than mesh alone.

Archibald et al. (1993) performed in situ pull tests on Minegard coatings of 4.6 mm thickness. The EC and LC values were approximately 0.7 kJ and 11.2 kN respectively. Minegard may strengthen fractured rock via material infiltration, and it may provide additional support through the caisson effect during rock deformation, i.e., the air pressure behind the coating becomes slightly less than in the opening (Archibald et al. 1993).

In cable lacing, the energy capacity of the entire length of a cable (more than 10 m) can be mobilized for local support, if the cable is not fixed to the holding bolts, but, passed through the eyebolts. For a 16 mm diameter, 10 m long cable with yield point of 380 MPa and ultimate strain of 8 %, the displacement, load, and energy capacities are calculated to be 566 mm, 86 kN, and 61 kJ (assuming local rock ejection within 4 bolts, and the 10 m length is mobilized). For practical purposes, we conservatively define DC = 300 mm to obtain LC = 62 kN and EC = 12 kJ. For local rock ejection, therefore, cable lacing would have higher EC than nearly all the commonly used surface elements (Fig. 2b, marked L). In addition, properly installed cables are connected using clamps which should allow substantial slip before complete failure of the cable. Cable lacing and mesh can complement each other in a system so that that the former provides high EC and the latter offers higher degree of areal coverage.

4.3 Support Systems

In a support system, the in-rock elements provide support to rock blocks, and surface elements support those rocks whose load can not be transferred to the in-rock elements through rock arch effects.

Fig. 3 shows the system EC and LC point plots for two separate classifications of support systems. The one by McCreath and Kaiser (1992) (Table 6a) is marked 0 to V, and the one by Brummer et al. (1993) (Table 6b)is marked L1 to L5. Data for in-rock elements in Table 3, and for surface elements in Tables 4 and 5 are used. For the former classification, from systems 0 to III, both EC and LC increase, but, from III to V, essentially either LC or

Table 6a Six levels of typical supports in Canadian

hard rock mines, after McCreath and Kaiser (1992)

No.	Roof support description
0	Mechanical bolts in 4'x5' diamond pattern
I	No. 0 plus mesh
II	No. I plus grouted rebars at diamond centers
III	No. I plus grouted double 7-strand cables
IV	No. III plus 100 mm shotcrete
V	No. III plus lacing

Table 6b Five levels of support systems, after Brummer et al. (1993)

No.	Support description
L1	Spot bolts only - density 1 bolt per 3 m sq.
L2	Pattern bolts - 1.2m by 1.2 m
L3	L2 plus weld mesh screen
L4	L3 plus shotcrete
L5	L2 plus shepherd’s crooks, chain-link screen & lacing

EC increases. This trend is also observed for the latter classification, where from L1 to L3, both EC and LC increase, but, from L3 to L5, essentially either LC or EC increases.

Therefore, the comparisons of support systems based on the system EC-LC point graph correctly reflect the experience gained in Canadian hard rock mines, that is, a higher level of support is used for a more severe rockburst condition.

The direct costs (labor, drilling and material) for installing the different support systems are also calculated by Brummer et al. (1993). This data are used to construct the graph in Fig. 4. It is interesting to see that the trend for system EC values reflects the trend for the costs, i.e, gradual increase from L1 to L3, but a jump at L4.

5 DISCUSSIONS

Elastic or yielding cushions may be placed between the rock surfaces and the head-plates of in-rock elements. An elastic cushion is one with a lower stiffness, and an yielding cushion is one with a lower yield load than the in-rock element. Soft wood blocks and rubber pads were tested by Yi and Kaiser (1994). It was found that these cushions reduce elastic stresses in mechanical bolts and enhance the bolt-mesh connection. However, if bolts are designed to undergo plastic deformation, then the cushions absorb insignificant amount of energy.

Bond slip on bolt-grout interface for fully grouted bolts, frictional slip on bolt-rock interface for frictional bolts, and slip of yielding nuts on the threads of bolts are examples of slip mechanisms. Bolts with bond slip possess high EC values.

From the analyses in this paper, the following observations are made:

1. Fully grouted bolts or cables should be installed to allow bond slip to occur before bolt rupture.

2. Based on test results from Kirsten and Labrum (1990), Mesh or fibre reinforced shotcrete thicker than about 100 mm have both higher EC and LC than mesh alone. Hence, reinforced shotcrete thicker than about 100 mm performs better than mesh as retaining elements.

3. For local rock ejection, cable lacing provides the highest EC among all surface elements because the entire length of a cable (more than 10 m) can be mobilized. It is therefore the best retaining element for dynamic loading.

4. Cushions underneath the head plates are beneficial if meshes are used.

5. Accumulation of plastic deformations under repeated small impacts may cause support to fail in time. Additional elements should be added to ensure safe access.

6. In a support system, elements of different energy and load capacities, and degrees of areal coverage can be combined to obtain the desired EC and LC.

6. CONCLUSIONS

The energy capacity (EC) versus load capacity (LC) point graphs (Fig. 2 and 3) can be conveniently used to compare the mechanical performance of different support elements and systems in rockburst conditions. The results from such comparisons reflect the experience gained in the Canadian mining industry.

ACKNOWLEDGMENTS

A portion of the material used in this paper was collected during the author’s post-doctoral fellowship at the Geomechanics Research Center of Laurentian University. Drs. P. K. Kaiser, D. R. McCreath, R. K. Brummer and D. D. Tannant provided guidance and assistance.

REFERENCES

Archibald J.F., Mercer R.A. and Lausch P. 1993. Innovative ground support using rapid deployment of spray-on liners, presented at the 95th Annual General Meeting of CIM, (13p). Alberta, Calgory.

Barton N., Lien R. and Lunde J. 1974. Enginering classification of rockmasses for the design of tunnel support. Rock Mechanics, 6 (4), 189-236.

Brummer R.K., Nose J-P., Hay L., and Kaiser P.K. 1993. Risk-cost-benefit analysis applied to the design of support in burst-prone mines. In Bawden W.F. and Archibald J.F. (Eds.), Innovative Mine Design for the 21st Century - Proceedings of the International Congress on Mine Design, (969-980). Kingston, Ontario, Canada.

Bywater S. and Fuller P.G. 1983. Cable support of lead open stope hangingwalls at Mount Isa Mines Limited. In O. Stephansson (Ed.), Proceedings of the International Symposium on Rock Bolting, (539-555). Abisko, Sweden.

Fuller P.G. and Cox R.H.T. 1978. Rock reinforcement design based on control of joint displacement -a new concept. Proceedings of the 3rd Australasian Tunnelling Conference, (28-35). Sydney, Australia.

Hepworth N. and Gay N.C. 1986. Deformation and Support Performance in a Tunnel in a high Stress, Hard Rock, Mining Environment. Rock Mechanics Laboratory, Research Report No. 11/86, Chamber of Mines of South Africa.

Kaiser P.K. 1993. Support of tunnels in burst-prone ground - toward a rational design methodology. In P. Young (Ed.), Proceedings of the 3rd International Symposium on Rockbursts and Seismicity in Mines, (13-27). Kingston, Ontario, Canada, A.A. Balkema/Rotterdam/Brookfield.

Kirsten H.A.D. and Labrum P.R. 1990. The equivalence of fibre and mesh reinforcement in the shotcrete used in tunnel-support systems. Journal of the South African Institute of Mining Metallurgy, 90(7), 153-171.

McCreath D. R. and Kaiser P. K. 1992. Evaluation of current support practices in burst-prone ground and preliminary guidelines for Canadian hard rock mines. In Kaiser P. K. and McCreath D. R. (Eds), Proceedings of the International Symposium on Rock Support, (611-619). Sudbury, Canada, June 16-19. Rotterdam / Brookfield: A.A.Balkema.

McCreath D. R. and Kaiser P. K. 1992. Current Support Practices in Burst-Prone Ground, for Rockburst Project Management Committee of Mining Research Directorate, Geomechanics Research Center, Laurentian University, Sudbury, Canada.

Ortlepp W. D. 1983. Considerations in the design of support for deep hard rock tunnels. Proceedings of the 5th Congress of the International Society for Rock Mechanics (D179-187). Melbourne, Australia. Rotterdam: A. A. Balkema.

Pakalnis V and Ames D. 1983. Load tests on mine screening. In Udd J.E. (Ed.), Underground Support Systems, special volume 35, The Canadian Institute of Mining and Metallurgy.

Potvin Y. and Milne D. 1992. Empirical cable bolt support design . In McCreath D.R. and Kaiser P.K. (Eds.) Rock Support in Mining and Underground Construction, Proceedings of International Symposium on Rock Support, (269-275). Sudbury, Rotterdam: Balkema.

Roberts M. K. C. and Brummer R. K. 1988. Support requirements in rockburst conditions. Journal of South African Institute of Mining and Metallurgy, 88(3), 97-104.

Stillborg B. 1986. Professional Users Handbook for Rock Bolting, (145p). Trans Tech Publications.

Stillborg B. 1990. Rockbolt and Cablebolt Tensile Testing across a Joint, (19p). Unpublished report, Mining and Geotechnical Consultants, Lulea, Sweden.

Stillborg B. 1993. Rockbolt Tensile Loading across a Joint, (19p). Report to Atlas Copco, Lulea University of Technology, Lulea, Sweden.

Wagner H. 1984. Support requirements for rockburst conditions. In N. C. Gay and E. H. Wainwright (Eds), Proceedings of the 1st International Symposium on Rockbursts and Seismicity in Mines, (209-218), Johannesburg, 1982, SAIMM, 1984.

Yi X and Kaiser P.K. 1993. Mechanisms of rockmass failure and prevention strategies in rockburst conditions. In P. Young (Ed.), Proceedings of the 3rd International Symposium on Rockbursts and Seismicity in Mines, (141-145). Kingston, Ontario, Canada, A. A. Balkema / Rotterdam / Brookfield.

Yi X. 1993. Dynamic Response and Design of Support Elements in Rockburst Conditions, (273p). PhD Thesis, Dept. of Mining, Queen's University, Kingston, Ontario, Canada.

Yi X. and Kaiser P.K. 1994. Impact testing for rockbolt design in rockburst conditions. International Journal of Rock Mechanics and Mining Science & Geomech. Abstr. 31(6), 671-685.