|
(11) | EP 0 138 446 A1 |
| (12) | EUROPEAN PATENT APPLICATION |
|
|
|
|
|||||||||||||||||||||||||||
| (54) | Ground support coating |
| (57) @ In the excavation of underground openings, an improved method of supporting the
excavation is provided. This comprises the use of a reinforced resin to support the
excavation. This material offers considerable advantages over the concrete supports
used hitherto. If necessary further support by bolting and screening may be used. |
[0001] The present invention relates to ground support coatings, and to a method of supporting an underground excavation or opening.
[0002] The excavation of underground openings in operations such as shaft sinking, tunnelling, raising and stoping almost always require the use of means for supporting the roof and walls to ensure that the opening remains in a stable condition to prevent collapse.
[0003] Various types of ground support have been used over the years, the earliest forms of support being timber sets. As technology progressed, steel frames, arches, roof bolts, grouted rods and cables, steel mesh and reinforced concrete augmented or replaced the basic timber set.
[0004] In particular, sprayed-on concrete commonly known as shotcrete or gunite (which is shotcrete minus aggregate) has achieved considerable success. In areas not experiencing geological stress it is often possible to dispense with the use of roof bolting and screening and to use shotcrete to support the excavation without compromising crew and equipment safety.
[0005] Shotcrete appears to assist in the development of a compressive Voussoir arch. The hardened concrete supports the opening and distributes the weight of the ground in compression. However shotcrete poses material handling problems being heavy, bulky and having a relatively short life in the wet mix mode. It creates undesirable dust problems and results in adverse skin reactions in certain individuals.
[0006] The present invention is based on the discovery of an alternative material for use in ground support systems where there are material handling problems.
[0007] According to the present invention, there is provided a method of supporting an underground excavation by forming a coating on the interior of the excavation to secure the excavation from collapse, characterised in that the coating is of reinforced resin of a predetermined thickness.
[0008] The blend of resin and reinforcement is preferably applied to the roof and walls of the excavation or opening by spraying apparatus, although other means of application may also be used. A preferred resin for use in the method of the invention is epoxy, reinforced with fibreglass or other fibrous or metallic reinforcement. Such combinations provide temporary and long term support to the opening or excavation and may be used in combination with other support systems. The reinforced resin compound bonds together the fractured or jointed rock pieces that normally occur as underground openings are excavated, thereby forming a reinforced network of individual rock pieces and generating a self supporting structure. Screens and roof bolts may be installed if conditions dictate.
[0009] Tests have shown that fibreglass-reinforced epoxy is the preferred material for use in processes of the invention, and that this is a satisfactory replacement for shotcrete. The components of the epoxy have a long shelf life; relatively small quantities are necessary; and it is a low flammable material. There is little danger of explosion during spraying operations and its combustion resistance in the cured form is equal to or better than that of wood. Fibreglass reinforced epoxy has the following advantages over existing methods and materials:
1) Low toxicity.
2) Economical.
3) Fast curing time.
4) Handling ease.
5) Low allergic sensitization potential.
6) Compatible with existing spray equipment.
7) Adhesion to dry and/or damp rock.
8) Low odour.
9) Light colour.
10) Does not corrode.
11) Has a low Young's modulous - flexible yet strong.
[0010] The reinforced resin uses its tensile strength to provide support. The overall support mechanism and interaction with the fractured material forming the walls and roof of the opening differs greatly from that provided by concrete sprayed lining systems which are dependent on the compression strength of the sprayed material.
[0011] Some examples will now be described by reference to the accompanying drawings in which:-
Figure 1 plots strength versus thickness.
Figure 2 plots strength versus dampness of rock.
Figure 3 plots strength versus percent fibreglass.
Figure 4 plots top surface sag versus lid drop.
Figure 5 plots top surface sag versus lateral pressure.
Figure 6 plots load versus deflection.
Figure 7 plots load versus deflection.
Figure 8 plots load versus deflection.
Example 1
[0012] Fibreglass reinforced Compox* roofing compound, an epoxy resin, was sprayed underground at Inco Limited's Copper Cliff South Mine using standard available spray equipment to coat the interior of the opening. The sprayer ensures the proper mix of resin base reactor and chopped fibreglass. In this example the following spray equipment was used:-
1. Bulldog* 2:1 mix ratio Hydra-Cat* with heaters.
2. 1:1 Stubby* pumps with 1.2 m feed hose.
3. 17.7 m of heated hose and solvent hose.
4. Monark* solvent pump.
5. Solvent flush gun.
6. Glass-craft* chopper gun (6 blade).
7. Band heater. * Trademark
[0013] These trials proved to be successful, some initial minor problems posed by the spraying equipment being easily surmounted. However, there was no existing way to relate the strength of the sprayed material to a reference standard such as concrete. Laboratory tests were then performed to establish strength and bonding evaluations of the fibreglass reinforced epoxy coating. These demonstrated that the optimum coating is one containing the maximum length and quantity of fibreglass that can be consistently sprayed to form a compact, thin layer. A coating of well sprayed fibreglass reinforced epoxy about 0.25 cm thick appears to provide equivalent support to that of a 1.27 cm thick layer of shotcrete. The tests showed that shotcrete asissts in the development of a compressive Voussoir arch while fibreglass reinforced epoxy coating creates a reinforced beam with the fibreglass working in tension. The tests are described in detail in the following examples.
Example 2
[0014] A punch test programme was devised to duplicate a flat configuration with no end constraint and to assess the effects of coating thickness, fibreglass content and length, and the condition of the rock surface on the behaviour of the coating. The results were output in terms of resistance per cm of crack length and compared to values quoted for shotcrete in the literature. The test was to determine if bond strength limits the coating strength as is the case of unreinforced shotcrete. Since adhesion plays a critical part in the interaction between a sprayed coating and the rock structure it is helping to support,mine rock had to be used in the test and sawn surfaces were required to eliminate surface conditions as an uncontrolled variable. The test apparatus was designed for regular sawn rock shapes; however the size of the rock speciments was limited by the size of the available rock sawing equipment and the number of tests by the time required to prepare the rock specimens.
[0015] The test strength range was also a limiting factor. It was decided that the objective strength of the fibreglass reinforced epoxy coating should be equivalent to 1.27 cm thick layer of shotcrete. The quoted strength for a punch test is 606.74 kN/n2 of crack length for seven day old shotcrete with a good bond. The size limit of the rock specimens and the low strength values required a test system with a fair degree of accuracy in the zero to 4500 N range.
[0016] Since fibreglass reinforced epoxy coatings are thin and fail easily when kinked because of the high stress concentration factors associated with sharp bends, the test apparatus also had to guarantee a straight thrust with no application of bending moments either before or during the test.
[0017] The punch tester consisted of a rock specimen holder in the shape of a rectangular frame and a four tiered test frame. Disposed below the tester was a hydraulic jack. A platen disposed in the test frame transferred the load of the jack to the rock specimen holder from below.
[0018] Three sawn rock pieces were clamped within the holder. The pieces were adjacent to one another. The two outer rock specimens were also touching the inner specimens. The cracks between specimens were only open enough to be seen when the assembly was held up against the light. The bevelled surfaces of the rectangular frame were covered with paraffin wax and the gaps between the frame were filled with paraffin wax to contain the epoxy within the test area. The epoxy formulation would not adhere to wax. The epoxy was then spread onto the surfaces of the rocks, and the pre-chopped fibreglass was sprinkled on top and patted into the epoxy with a plastic sheet.
[0019] The test frame consisted of four tiers in the form of four parallel horizontal flat plates and four vertical columns. The base plate and the four columns were firmly attached. The plate above the base held the platten which transferred the jack load to the base of the central rock pieces in the holder. It slid on the four columns to ensure that the jack thrust had no side or tilt component. The third plate from the bottom was fixed to the columns and had a rectangular hole for the passage of the jacking platten. There were four levelling screws to level and support the test piece until the jacking began. The top plate was movable and was capable of being raised, lowered or locked into position. The top plate was raised to move test pieces in and out of the frame and then lowered onto the test piece and locked in place. There was a hole drilled through the centre of the top plate to hold the linear potentiometer used to measure the upward displacement of the central rock piece during tests.
[0020] The rock specimen holder was inserted into the test frame above the platen. By energizing the hydraulic jack, a measurable increasing load was gradually introduced to the bottom of the central rock specimen. After a maximum pressure was achieved, the platten would punch the central rock specimen out. It was at this point that the coating would fail.
[0021] A total of fifteen tests were conducted, four to establish the relationship between coating thickness and strength, three to test for the effect of varying the fibreglass content, five for surface conditions, one for fibre length, one for cure time and one to see if the test would work for shotcrete.
[0022] A total of about 270 kg of rock sample were collected off the walls of the main footwall drift at the 1200 foot (.366 km) level at Limited's Copper Cliff South Mine in Sudbury, Ontario. Each sample was close to the size of a concrete foundation block. The rock type was a fine grained, quartzite, metasediment, some of which was brecciated.
[0023] The rocks were first ripped into slabs 2.22 cm thick and then cut up into either 7.62 x 12.7 cm or 10.16 x 12.7 cm blocks. One long edge of each 10.16 x 12.7 cm block was slightly bevelled on a polishing wheel. More than 50% of the samples failed along fractures or joints in the rock during sawing operations. This provided seventeen 7.62 x 12.7 cm blocks, thirty 10.16 x 12.7 cm blocks and several flawed spares.
[0024] The following procedure was used.
1. The plate assemblies for a test were cleaned and coated in Turtle Wax*. Hot paraffin was then painted on the bevelled edges of the top clamping pieces to guarantee no bonding between the clamping pieces and the epoxy.
2. Two 10.16 x 12.7 cm rock blocks and one 7.62 x 12.7 cm rock block were selected and washed with soap and water. When dry, they were loosely clamped in their respective assemblies and the assemblies bolted together with two tie bars. The bevelled sides of the outer larger blocks were placed adjacent to the smaller centre piece so that the apex of the slim "v" formed by the adjacent rock blocks was up. The * Trademark blocks were snugged together to leave a crack separating the blocks visible only when the assembly was held up against a light. The tie bars were removed and the clamping assemblies tightened. The tie bars were replaced and the gaps checked. The procedure was repeated as many times as necessary to achieve the correct gaps after all parts of the assembly were tightened.
3. The gaps between the upper clamping pieces were plugged with paraffin wax to keep the expoxy from flowing outside the test area and extending the lengths of the coated cracks.
4. The planned quantity of precut fibreglass rope fibres was weighed into a plastic cup.
5. The reactor and epoxy base were weighed into the plastic mixing cup and stirred for 60 seconds. The mix was poured and uniformly spread with a putty knife over the surfaces of the rocks in the test assembly. The quantity of expoxy was recorded.
6. The fibreglass fibres were sprinkled uniformly over the surface of the epoxy and patted and rubbed into the epoxy with a piece of plastic sheet. Air bubbles were worked out from under the plastic cover. The plastic cover was left on the sample.
7. The plastic cover was peeled off the coating when it no longer adhered to the epoxy, normally after a minimum of 2 hours and 15 minutes at the laboratory temperatures of 20°C.
8. After overnight curing, the specimen was prepared for testing by punching out the paraffin plugs separating the upper clamping pieces.
9. The jack was set in the frame and the movable platten jacked up and pushed down until the jacking load to raise the platten normalized at around 1330 N. The movable platten was then set at a specific height relative to the sample support plate.
10. A dummy piece consisting of one 7.62 x 12.7 cm rock block and the upper and lower clamping pieces for the block was set on top of the jacking platen. A graph recording was made as the dummy piece was jacked slowly up 1.27 cm.
11. The dummy piece was removed and the jacking platten lowered to the prescribed elevation. The sample was then placed on top of the levelling screws which were used to level and elevate the specimen to a point where just a few µm separated the base plate and the top of the jacking platten.
12. The loading frame top plate was lowered onto the tops of the upper clamping pieces on the two outer rock blocks and locked in place with the upper set of nuts.
13. A second graph tracing was made as the jack slowly pushed the central rock block up through the coating.
[0025] In these tests, some epoxy filled the cracks (the joints between the rocks) to a depth of about .25 cm below the top surface of the rock. The combination of the high adhesion between the epoxy and the rock and the presence of this material in the crack influenced the results. Calculations were made assuming that the area of the failed rock surface times the rock resistance plus the area of the failed fibreglass reinforced epoxy surface times its resistance was equal to the total force required to fail each specimen. The least squares solution for rock resistance was 979.1 kN/m2 and that of the fibreglass reinforced epoxy, 661.9 kN/m2.
[0026] It was decided that it was not necessary to prevent epoxy from getting into the cracks since this also occurred in underground trials. The results are given in Table 1.
[0028] Tests 1 to 4 were for the relationship of coating thickness to strength. The graph in Figure 1 shows an almost straight line relationship of strength vs. thickness up to the maximum thickness tested.
[0029] Tests 5, 6, 7, 12 and 13 checked the effect of the surface condition on the strength of the coating. Test 5 was dry. In test 6, a wet paper towel was placed on the rock surface for 20 minutes before the epoxy coating was applied. The towel was removed and the excess surface water sponged off before the coating was spread. At least 50% of the rock surface appeared dark and therefore damp when the coating was applied. For test 7, a wet paper towel was placed on the sample for 20 minutes before the coating was applied and was not removed until the moment before the coating was spread. Surface water was not sponged away.
[0030] In test 12, mine dust,scraped off one of the samples collected from 1200 level at Copper Cliff South Mine,was lightly dusted over the dry surface of the specimen before the epoxy was applied. In test 13, the specimen surface was covered in oil and the excess oil wiped off. Mine dust was lightly dusted over the sample surface before the epoxy coating was applied.
[0031] Figure 2 is a graph of strength versus dampness and shows reduced strength with dampness. In the extreme case in test 7, where adhesion failure exclusively occurred, it still took 1334.4 N of force to peel the coating. The dry dust and oil dust test results were near identical insofar as strength was concerned but the oily surface behaved better in toughness. Tests 6, 12 and 13 all showed some adhesive failure where the coating parted from the rock surface and took along only a few small rock grains.
[0032] Tests 8, 9, 10 and 16 were to determine the relationship between fibreglass content and strength. Test 16 was a repeat of test 8 because of a mistake made during the testing procedure. The tie bars were left on at the start of jacking and were removed only after a high load level had been reached. Figure 3 is a graph of strength versus fibreglass content and shows increased strength with increased fibreglass content.
[0033] Test 11 used 0.71 cm long glass fibres. The strength was very high for the percentage of fibreglass but the behaviour was brittle. The apparent toughness may be misleading. It was believed the results were a function of the softness of the jacking system and the acceleration limits of the recording devices employed.
[0034] Test 14 was a 4 hour cure time test of the epoxy. The epoxy behaved as a viscous fluid and the jacking load was a function of the jacking rate. As soon as jacking stopped, the load dropped to zero.
[0035] Test 15 on shotcrete used the same mix as was applied on the simulated mine roof. The surface of the shotcrete layer was kept damp as it cured. It was tested 4 hours after mixing and exhibited no bond strength to the saw cut rock surface.
Example 3
[0036] A second test regimen was planned to simulate a blocky mine roof where the lateral constraint presssure and the roof load could be controlled.
[0037] A blocky mine roof is one where the span of the opening ranges between 10 and 100 times the average joint spacing. The size of the frame to hold a simulated mine roof must therefore be greater than 10 times the length and width of the blocks used to simulate the jointed rock.
[0038] The smallest available block was a refractory brick with the dimensions of 11.43 x 7.62 x 7.62 cm, each brick weighing 1.5 kg. For flexibility in the arrangements of the blocks in the model, the frame measurements were designed to allow placing the bricks with their long axis along either the length or the width of the frame.
[0039] The apparatus for simulating a mine roof consisted of a simulated mine roof frame, a support stand and two lids. The frame had axles welded to each end which fitted into steel bushings on the stand. The frame was rotatable a full 360 degrees within the stand and was held in either a flat or upright position by pins which fitted through holes drilled through the bushings and axles. One lid (the bottom lid) was used whenever the frame was turned over. The bottom lid had a hole drilled close to its centre point to measure roof sag relative to the lid position during tests.
[0040] The frame held a total of 480 bricks in four layers of 120 bricks each. The inside of the frame was lined on two adjacent sides with 1.91 cm plywood. The other sides were lined with thin pieces of cardboard behind which were placed two truck tire tubes, acting as low pressure bladders to provide lateral pressure uniformly against the sides of the bricks. Holes were drilled through the sides of the frame for the tube stems after the frame was constructed. With pressure limited to less than 48.3 kN/m2 in the tubes, there was no problem with the tube protruding out between the top of the bricks and the frame.
[0041] Water was used to pressurize the tubes but no effort was made to bleed out trapped air. Pressure control was provided by an overflow device which could be raised to a maximum elevation of 4.6 m above the mid height of the test frame.
[0042] A jacking beam was disposed above the frame and a hydraulic jack was situated between the beam and the top layer of bricks. As the shaft within the jack was increasingly brought to bear on the beam, the bottom of the jack, lying directly on the bricks, caused the bricks to ultimately sag and then collapse. Initially the simulated mine roof tests were used to establish a model which failed consistently under its own weight at a given lateral pressure. The model was then used to evaluate the relative strength of a shotcrete and a fibreglass reinforced epoxy coating.
[0043] Experiments 1, 2 and 3 established the baseline performance of the model without any coating. Experiments 4 and 5 tested shotcrete coatings. Experiment 6 tested a poured fibreglass reinforced epoxy coating and experiment 7 tested a sprayed fibreglass reinforced epoxy coating.
[0044] The following test procedure was used:-
1. The lower lid was installed flush with the lower ledge in the roof test frame.
2. 120 bricks making up the first layer in the test frame were placed with the long axis of the bricks parallel with the long axis of the frame. The bricks butted up against the two plywood lined sides.
3. Thin cardboard liners were placed along the two sides of the frame without plywood liners.
4. The two truck tyre tubes were carefully installed behind the cardboard liners in the frame so that the bodies of the tubes were uniformly spread along the lengths and heights of the two frame sides. (A 10.00 x 22 tube goes along the long side and a 9.00 x 20 tube along the short end of the frame.)
5. The frame was filled with three more layers of bricks, maintaining the bricks in the same orientation as used in the first layer.
6. The overflow pipe was set to its maximum elevation, the drain valve closed and the overflow valve opened. The water was turned on to the frame lateral pressure system. The water flow was adjusted that there was always a slight overflow of water from the system to maintain the lateral pressure head.
7. After about one hour, the top lid was installed snug to the top layer of bricks.
8. The frame was turned over 180 degrees in a clockwise direction, looking from south to north, so that the weight of the bricks in the frame upright position was against the long side plywood liner.
9. The lower lid was removed.
10. The cracks between the bricks were filled with fine sand to prevent the epoxy from leaking down inside the frame. A narrow bank of sand was left along the edge of the frame to keep the epoxy from running under the frame edge. Excess sand was brushed away.
11. The depth of the bricks below the upper edge of a 7.62 cm steel channel along a preselected location were measured and the brick rows and the frame marked so that the measurements could be duplicated after the coating has been applied and cured.
12. The frame area was divided into 11 sections such that the boundaries between sections did not coincide with the brick joints. Each section was numbered and a sequence of pours used such that adjacent sections were not poured sequentially.
13. The quantities of reactor, base and fibreglass to cover each section to a depth of 0.21 cm with 10% fibreglass by weight was calculated. The nominal 1.27 cm long fibreglass strand was used.
14. The ingredients for the first section were weighed and the reactor and base mixed for a full 60 seconds and spread over the planned area of the frame. The fibreglass was sprinkled uniformly over the epoxy and the fibreglass patted in gently with a piece of plastic sheet working the larger air bubbles to the edge of the section to bleed them off.
15. This was repeated for the remaining sections which do not butt up against previously poured sections.
16. After the adjacent sections had cured to the point where the plastic sheets could be peeled cleanly away from the epoxy,14 was repeated for the intermediate sections.
17. 12 hours or more after the last section of epoxy had been poured, the depth measurements in step 11 was repeated.
18. The lower lid was installed snug on top of the fibreglass epoxy coating and the frame rotated 180 degrees in a counterclockwise direction.
19. The upper lid was removed, the jacking beam mounted and the upper surface sag measured at a centrally located brick.
20. The lower lid was lowered uniformly in small increments and the degree of sag and the depth of the bottom of the lower lid below the epoxy surface of the frame recorded.
21. When the lower lid was completely clear of the fibreglass epoxy coating, it was lowered in one step to the limit of the support rods.
22. The lateral pressure on the frame was reduced in stages to a net of 1.23 m recording the upper and lower surface sag at each stage.
23. The jack was set under the jacking beam to bear on some centrally located bricks.
24. The frame was loaded with the jack in approximate 444.8 N increments to a maximum of about 8000 N and the upper and lower surface sag at each increment of load recorded.
25. Where the coating had not failed, the jack load was reduced to zero and the upper and lower surface sag readings recorded.
26. The overflow drain valve was opened and the water to the lateral pressure system shut off.
27. When drainage stopped and the lateral pressure was zero, steps 25 and 26 were repeated.
28. The frame was loaded to about 8000 N and lower surface sag readings recorded.
29. The frame was loaded in about 1350 N increments to the jack limit of 13,334 N, recording upper and lower surface sag readings at each stage of loading.
30. Where the coating had not failed, the load was reduced to zero.
31. The zero load sag of the upper and lower surfaces was recorded.
32. The frame was loaded in about 3500 N increments until the coating failed. The upper and lower surface sag readings at each increment of load being recorded, at specific intervals of sag after the coating had failed.
[0045] In experiment 1, the test frame was filled as in steps 1 to 6 with the exception that 1.905 cm plywood separated the tubes from the bricks.
[0046] Although some blocks fell out when the lateral pressure head was reduced to zero, it was felt that there was too much residual lateral pressure due to the tight fit between the tubes and the plywood liners.
[0047] Experiments 2 and 3 were identical in procedure to experiment 1 except that the plywood liners separating the tubes and bricks were replaced by thin pieces of cardboard.
[0048] Figure 4 is a plot of top surface sag versus lid drop for tests 2 to 7 and Figure 5 is a plot of the top surface sag as the lateral pressure was reduced. They show that for experiments 2 and 3, the magnitude of roof sag differed but the inflection points were identical.
[0049] For experiment 4, the bricks were dampened and coated with a simulated shotcrete mix using a combination of one part Cement Fondu* to 3 parts portland cement to achieve the desired four to six hour strength range of between 1379 and 2068.4 kN/m2. There were problems mixing the coating and a flash set occurred before the coating could be spread. This was characterised by the mix drying.
[0050] In experiment 5 the ratio of Cement Fondu to portland cement was changed to one part Fondu to five parts portland. This combination had some unusual properties. The initial set was rapid but began about 15 minutes after mixing. It reached a 12.7 cm cube strength in excess of 1379 kN/m2 by four hours and then maintained this strength for at least an additional two hours before beginning to increase in strength again.
[0051] In experiment 4 the coating failed at a jack load of 2224 N while the confining lateral pressure was a 12.53 kN/m2. The failure started with the coating peeling away from the bricks along the long edge of the frame. About two thirds of the coating fell at a jack load of 2446.4 N and the remainder fell at the 3558.4 N mark.
[0052] In experiment 5 the failure load was just 7561 N and the failure was quite sudden. A shear plane developed along the long side of the frame with a break angle of 15 degrees. It corresponds to the type of failure hitherto in the literature for a moment- thrust failure of end-supported layers in an arched configuration. There was also adhesion failure extending from the shear crack through to the midline of the frame, but none was noted until after the shear fracture developed.
[0053] In experiment 6, the poured fibreglass reinforced epoxy coating was applied as outlined in the general test procedure. Failure could not be induced with the 13,344 N test jack. A 88,960 N jack was used for the final failure load cycle.
[0054] The fibreglass reinforced poured epoxy coating failed at a load of 27,577.6N. Failure was gradual and started as a split along the centre-line long axis of the frame. The load carrying capacity of the coating reduced as the crack lengthened and new cracks developed.
[0055] In experiment 7, the reinforced fibreglass epoxy coating was sprayed on using the spray equipment discussed heretofore. Otherwise, the test proceeded in a similar manner to experiment 6. The coating failed at a load of 14,011.2 N during the fourth loading cycle.
[0056] Failure started with a brick rotating out of the roof and tearing the epoxy away from the bricks. Failure was again gradual but there was more debonding than cracking whereas in the first epoxy test there was no debonding.
[0057] The mix was off ratio when the test frame was sprayed in that it was too blue, indicating an excess of reactor. It remained sticky and after the failure it was noted there were a few internal layers of pure white base. The sand used to plug the cracks between the bricks and to form a dam around the perimeter of the frame had blown loose and formed layers on the surface which caused bonding problems. It appeared as if 75% of the adhesion difficulties could be attributed to the off-ratio mix and the other 25% to the sand.
[0058] Figure 6 shows the load deflection curves for a lateral pressure of 12.53 kN/m2. Experiment 4 failed in this loading cycle.
[0059] Figure 7 shows the load deflection curves for the second loading cycle when the lateral pressure was reduced to zero. The good shotcrete coating in experiment 5 failed even though deflections were far less than those measured for experiment 7.
[0060] Figure 8 shows the load deflection curves for the fourth cycle of loading and shows the toughness of the fibreglass reinforced epoxy after initial failure.
[0061] The tests given in Examples 2 and 3 demonstrate the effectiveness of the invention. This means that after an excavation is opened in a traditional manner such as by drilling or blasting the practice of bolting and screening can be dispensed with, or at least postponed. In a preferred method of the invention the interior of the opening or excavation is spray coated with a fibreglass reinforced resin. For mine openings it is preferred that a fibreglass length of about 0.7 cm disposed in a coating about 0.25 cm thick is preferred. The resin should include at least 10% fibreglass in the preferred method, and most advantageously more than 20%.