While impact loading was used to represent the effects of seismicity as realistically as possible, quasi-static load tests were performed to obtain accurate load-deformation characteristics of the system and its components. This model allows a direct assessment of the effect of relevant parameters and is therefore very useful in practice. The model can also be used to better understand the mechanisms involved in the support function of the mesh and ligaments.
One of the most revealing findings of this project was the general incompatibility between wire mesh and lacing and the compliant tendons in terms of stiffness and strength. The effective strength of the mesh was found to be severely reduced by stress concentrations associated with the tendons and, to a lesser extent, the lacing. The properties of the yielding tendons largely determine the performance of the tested support systems and it is therefore important that these properties match the specifications under all loading conditions.
One possible explanation for this finding is the sheer size of the current tests, compared to the small-scale tests performed previously.
G ENERAL
To further complicate the design problem, the proportional influence played by each support component is affected by the "integrity" of the rock mass. With increasing deformations, unraveling and loss of integrity can have a pronounced effect on the load distribution between the rock mass and the various support components. As soon as relatively large rock mass deformations can occur between the support units, energy is absorbed locally by the fabric support and inside the unraveling rock mass.
The stiff fabric is therefore more prone to failure, but is able to prevent the rock mass from unraveling as long as it does not fail. The softer substances have a much smaller capacity to prevent the rock mass from unraveling, but they have a relatively large (elastic) energy absorption capacity. A situation as shown at the top of Figure 1-2 is typical of most of the tests performed: relatively soft fabric combined with stiff tendons leading to bulging of the rock mass and ultimate failure of the relatively weak fabric.
Most of the impact energy is absorbed by the expanding fabric and rock mass; tendons hardly contribute to this scenario.
A IMS AND O BJECTIVES
T UNNEL S UPPORT D ESIGN C ONSIDERATIONS
The concept of creating a structurally competent reinforced rock mass arch is the basis of most of the empirical rock bolt design methods (Stillborg, 1986) and therefore generally applies in the South African gold mines as well. In its simplest form, the tributary area theory is assumed to be directly applicable to the tendons, which are supposedly anchored in stable rock outside the boundaries of the fractured and fragmented sheet (Wojno and Jager, 1987, Stilborg, 1986). Although the presence of mesh and laces provides a local support function in the areas between tendons, it is clear that this support function is relatively limited compared to that of the tendons themselves.
The support components within a typical tunnel support system are not compatible, often resulting in premature failure of the mesh, which is the weakest link in the system. According to this manual, the function of the retaining elements (i.e. the tendons) includes absorbing energy under severe conditions. Interpretation shows that initial stiffness and strength are useful to accommodate less severe conditions, while high energy absorption capacity, associated with high rock mass accumulation, is recommended for the more severe cases.
Nonlinear stiffness characteristics (in the direction of loading) and their effects on strength and energy absorption capacity can be easily evaluated and investigated with the help of the model.
C ONCEPTUAL M ODEL
The in-plane stiffness and strength of the lace or mesh will determine the response to vertical loading. If the in plane stiffness of the lace or mesh, including "boundary stiffness" is expressed by. C=100 / and .C=1000ton/m Note that both the vertical load and the in-plane force are linearly related to the in-plane stiffness.
A direct comparison between Figure 3-8 and Figure 3-6 is not easy, because the actual stiffness is linearly related to the in-plane stiffness in the conventional case, as shown in Figure 3-6. In-plane stiffness has a similar effect on load-carrying capacity as it does on energy absorption capacity. The in-plane strength of the support system has a greater effect on energy absorption capacity than on load-bearing capacity.
An increase in the strength of the plane to 20 tons simply results in a corresponding increase in optimal stiffness of the plane.
R ESULTS
Calibration of Load Cells and High Pressure Pump
To get the maximum capacity from the hydraulic propeller that loads the system, a high-pressure pump has been developed, with a maximum output pressure of 80 MPa. Using a manometer, the output pressure of the pump could be manually set to any desired value. By recording the propeller force for certain pump pressures, a relationship between pump pressure and propeller force could be obtained.
This relationship is linear and calibration has shown that each pump pressure of 10 MPa induces a force of approximately 55 kN in the hydraulic propeller (5.5 tons).
Summary and Discussion
A shorter span should in theory lead to an increase in stiffness, and this result is therefore indicative of the influence of stress concentrators. To avoid stress concentrations and to ensure maximum load transfer through the web and/or lacing, the remaining tests were performed using brick walls. The latter finding is consistent with the results of the present tests, as the strength of the welded mesh was always lower than that of the diamond mesh.
It is also possible that at relatively higher mesh density (wires/span) the increase in the stiffness of the rhombic mesh is disproportionate. From the results of tests where the support force was directly applied to the web and/or lace, a deformation of 93 mm would be associated with a support force of about 66 kN. The net was clamped on shims and therefore the net had to reduce the horizontal movement of the tendons to a certain extent.
The results of these static tests are directly applicable to the impact tests, as they allow the estimation of the support forces generated by the net and lacing in relation to monitored deformations.
Conclusions
However, by making effective use of existing materials, a similar increase in load-bearing capacity can also be achieved without resorting to larger amounts of support material. The system response during the crash tests is comparable to the system response during a static test, with regard to the load-deformation properties. If the system stiffness of the rock mass and tissue support is too soft, or the system strength of the rock mass and tissue support is too low, compliant tendons will not serve this purpose.
The increased stiffness results in an increase in (dynamic) forces due to subsequent impacts and can lead to ultimate failure if the strength of the system is less than the yield strength of the tendons. A relatively high initial stiffness of the system, which can be achieved for example by using thick shotcrete, from a competent rock mass, by pre-deforming the mesh and ties or in any other way, would prevent the rock mass from buckling and would immediately attract relatively large impact forces. However, if the strength of the system again exceeds the yield strength of the tendons, these tendons will absorb the deformation and energy according to the original design philosophy.
These considerations will help interpret the results of the dynamic tests performed by SRK for this project.
R ESULTS
Summary and Discussion
Failure of the mesh may be sufficient to cause rock mass failure in the case of sufficiently fragmented and unraveling rock. As a principle, the strength of the mesh should not be exceeded before yielding begins in the tendons. The strength of the fabric support must match the yield strength of the tendons.
In the absence of any load-bearing capacity of the rock mass, this implies that the strength of the fabric must be greater than the yield strength of the tendons. The remainder of the load is most likely transferred directly to the ground (see Figure A -9). At a support force of 110 kN, excessive bending of the mesh frame was observed and the test was terminated (the frame subsequently had to be straightened).
However, in the case of the brick columns, most of the load will be distributed directly to the brick columns. In the next test on the same system, failure of the system occurred when the strut force exceeded 165 kN. It is not obvious where the rest of the energy is dissipated, but (elastic) energy absorption in the mesh and stringing is a major contributor.
This explains the reduced contribution of the center bolts and the increased effect in the outer bolts. Slippage of the lacing was also observed in the special compliant devices in this test. During the first impact, a relatively large amount of slippage occurred in the lacing yielding devices.
A mass of 3 tons was dropped from a height of 1.25 m onto the load spreader in the center of the four bolts. No yielding occurred in any of the tendons, but slip was observed in the yielding apparatus connected to the lace. A mass of 3 tons was struck on the load spreader located in the middle of the four tension cables.
A mass of 3 tonnes was dropped from a height of 1.25m onto the load spreader in the center of the suspended assembly of bricks and support. No postloading could be observed on any of the tendons, while localized damage was induced in the mesh around some of the tendons. However, the increased stiffness of the net and lacing should also lead to greater forces being generated during the impact.
