Cable-Types & Sizes

Cable bolt application in mining

For several decades now, cable bolting systems have been used for ground stabilisation and reinforcement in mines. There are around 19 types of cable bolts used in Australian mines, classified into main five categories. These are; (a) smooth or plain surface cable bolt; b) Bulbed; c) Nut caged; d) Spiral and indented cable bolts; e) a mix of plain and spiral cable bolts. Figure 1 above shows a selection of cable bolt types used in Australian underground coal mines.

Most cable bolts used in Australian mines are of seven or 19 wire construction. The seven wire cables has six outer wires wrapped around the central core wire, which is known as centre or king wire. However, the 19 wire cable  has two layers of wires consisting of nine 5 mm diameter outer wires; nine 3 mm inner layer wires, all wrapped or laid around 7 mm inner or king wire. The lay of the cable can be determined by the direction in which the outer strands of the cable are wrapped or laid around the centre king wire. The king-wire can be slightly larger in diameter to ensure the outer wires contact each other. For further information on the cable components and construction methodology, the reader is referred to Hutchinson and Diederichs  ( 1996)[CLICK], [http://icgcm.conferenceacademy.com/papers/detail.aspx?subdomain=icgcm&iid=965].

For a cable bolt support system to be effective, the loads have to be successfully transferred from the rock to the cable through the grouting materials.  These axial forces can be applied via the bearing plate, or as a result of horizontal movement of the rock mass at shear planes and bed separations. Thus the anchorage applied at the top of the hole, can be enhanced by buttons, opening the strand- called birdcages or bulbs.

Cable bolts are normally evaluated for strength and load transfer properties. The strength of cable can be carried out by tensile failure tests, while the load transfer strength is evaluated by pull and shear strength tests as well as cable stiffness.

Pull test

The pull test and / or ultimate tensile strength test  is carried out using a double embedment pull testing method in accordance with the British Standard BS7861: Part 2: 2009.  [Specification for Flexible systems for roof reinforcement]

Failure of a grouted anchorage by load transfer mechanism will occur either by; failure at the cable-to-grout interface; failure at the grout-to-rock interface; failure through the grout column; or by failure through the rock around the borehole wall.  Thomas (2012) [Click] [http://icgcm.conferenceacademy.com/papers/detail.aspx?subdomain=icgcm&iid=1011][Click].  documented a total of 19 cable bolts, which were subjected for load transfer studies by pull testing.

 

The tested cable included, plain-strand, bulbed, nut-caged and spiral or indented cables (see opposite). Figure 2 (a,b) opposite shows load-displacement results of various cables. Each tested cable ends are encapsulated in steel tubes and subjected to pulling.  Alternatively the cable end can be encapsulated in concrete cylinders in the laboratory or in overhead concrete blocks as show if Figure 2c. Alternatively tests can also be carried out underground in what is now known short encapsulated pull test (SEPT).

Ultimate tensile strength/ Failure load

For ultimate tensile strength failure, Figure 3 shows load-displacement profiles of three cable sections pull tested to failure. Tests were carried out in accordance with the British Standard (BS7861-Parts 2: 2009).  Pull test is carried out in embedment tubes with an internal diameter equal to the size of the hole as recommended by the manufacturer for tendon installation and with the steel tube wall thickness of at least 10 mm.  The internal surface of the tubes is threaded to a 2 mm pitch and 1 mm deep to maintain bonding of the resin / grout between and steel wall  http://ro.uow.edu.au/coal/560/

Cable encapsulation bonding Length

Two methods are available to determine the optimum length of the cable bolt initial top section encapsulation (without de-bonding).  One method is by direct pull testing and the other by shear testing.

The direct pull test is a common method of determining the optimum bonding length of the cable bolt encapsulation prior to its snapping. This test can be carried out both in the field and in the laboratory.  Naturally the length of encapsulation may vary according to the rock type (formation), borehole diameter, cable bolt diameter, capacity and purpose of installtion. Normally such tests are carried out effectively without the cable bolt being rotated during pull-out test.

Shear test

Shear strength of  cable bolt can be carried by;  a) a simple guillotine shear test; b) a single shear test of the cable bolt encapsulated in concrete structure; c) double shear test in concrete or rock.

The optimum de-bonding length for cable bolt installation cannot be determined by shear testing. It is a fact that the shear strength of steel is 60-70% of the tensile strength of the cable tested in tension. Not all the cable strands (wires) fail in tension or in shear, when tested in rock or concrete; it is a mixture of both. Hence the optimum length that can be determined by shearing will be less that is produced by normal pull out test. This means that if the cable bolt de-bonding length is to be studied, it should be determined from the cable pull-out test and not by shear testing. To rephrase, the de-bonding length of encapsulation in shear test will be less than that achieved with the cable failure by pure pull testing.

Two methods of testing cables in shear. The single shear and double shear test  The single shear test can be carried out either by double embedment method of encapsulating a cable length in steel tube as shown in Figure 4 or testing of cable installed in concrete cylinders as shown in Figure 4 b

Figure 5 shows the load-displacement graphs of three cable bolt sections tested for shear using the British System of double embedment shown in Figure 4. The sheared load values are likely to be lower that the tests results obtained from testing of cable installed in concrete cylinders. It is clear that cable element will be affected by the steel embedment walls as shown in Figure 5. http://ro.uow.edu.au/coal/560/

Figure 5 shows the axial load profile of a cable tested in shear using a double shear system with initial pretension load of 5 kN (almost 0 kN load) as illustrated in Figure 6 opposite. The maximum pretension load at the peak shear load shear load of 1291 kN force was 380 kN.   The pretension load of 380 kN was monitored by the load cell mounted at the outer side of the of the outer concrete block in the double shear test.

The actual recorded force is due to (a) the sharing force of the cable on both side of the central block plus the shear forces to overcome frictional force of the concrete joint surfaces.  The net cable shear failure is likely to be 75% of the total sharing force. Work is now in progress to carry out cable double shear testing without the concrete joint faces coming in touch with each , hence zero face shear force.

Further reporting on double shear testing and other tests visit: http://ro.uow.edu.au/coal/559/ ; http://ro.uow.edu.au/coal/560/;http://ro.uow.edu.au/coal/509/

Facebooktwittergoogle_plusredditlinkedinmailFacebooktwittergoogle_plusredditlinkedinmailby feather