1214 THE COLLIERY GUARDIAN June 29, 1917. SUBSIDENCE RESULTING FROM MINING* By L. A. Young and H. H. Stoek. (Continued from page 1079.J LABORATORY EXPERIMENTS AND DATA. In the laboratory various experiments and tests can be made to secure data which will be of assistance in the study of subsidence; more particularly general tests of the materials entering into the problem, and effect on superimposed material of the removal of part or all of the supports. Probably the most extensive experiments along this line are those made by H. Fayol, which included a variety of materials, as iron, fibre, canvas, rubber, sand, clay, and plaster. He placed iron bars 1*9 in. by 0’19 in. (50 mm. wide by 5 mm. thick) one above the other horizontally, the whole being supported by blocks of wood resting upon an iron table. A strong rule was placed upon the upper bar of iron, and the rule and bars were fastened together and to the table by means of stays and bolts. The wooden blocks were removed over a length of about 4 ft., and the sagging of the iron bars was noted. It was found that the deflection of the lower bar was 5 mm. (0-19 in.), of the 10th bar from the bottom 3'25 mm., of the 20th 1'75 mm., and that after the 30th bar there was no more bending. The same experiment was tried with flat aloe ropes and with straps of canvas and indiarubber in place of the iron bars. With straps of canvas and india- rubber the curve of the limits of deflection, that is to say, the limit of the zone of subsidence, had a height nearly equal to the distance between the points of sup- port. This height was about one-third of the same dis- tance for the ropes, and one-sixth for the iron bars. Wood and rocks also bend in a manner similar to the materials mentioned. In order to study the movement in beds of loose materials and in strata that might have been crushed by subsidence, Fayol used artificial beds of earth, sand, clay, plaster, or other materials, and constructed boxes of various dimensions having one side of glass. The box usually employed was 2 ft. 7 in. (80 cm.) long, 1 ft. (30 cm.) broad, and 1ft. 7 in. (50 cm.) deep. On the bottom of the box were placed, side by side, small pieces of wood of equal thickness, a few centimetres in width, and as long ds the breadth of the box. Experi- ments were made both with one row of these little pieces of wood and with several placed one above the other. Upon them were laid successive layers of arti- ficial strata, varying from 1 mm. to several centimetres in thickness. To note the movements, small pieces of paper about f in. (2 cm.) in length and f in. (1 cm.) in width, were put into the planes of stratification, and, on the glass, lines were marked in ink, covering exactly the lines formed by the paper. These lines enabled the least movement to be followed. By withdrawing the pieces of wood, excavations were formed and move- ment produced in the artificial strata. "i-ju SISiMBWMiililjBMSM iMMaawa— Fig. 19.—Subsidence of Artificial Beds. Fig. 19 represents the movements by taking away, in the order indicated by the numbers, the upper row of wooden pieces, where there were three rows each 0-3937 in. (1 cm.) in thickness. The first bed (dry sand), which rests directly on the pieces of wood, falls in as each pillar is withdrawn. The second bed commences to sink only when a certain number of pillars have been taken away. The sinking is shown at first by a slight curve, which has its greatest deflection toward the centre of the excavation. Then the third bed follows the second. The movement gradually extends in depth, and reaches the upper bed after the removal of the 12th pillar. After the removal of the 17th, the beds have become bent, as shown in the sketch, the limits of the deflection being the curves Z13 and Z17. (The index figure of the curves is the number of the last pillar taken away— namely, the curves Z8 Z4 indicate the extent of the movements after the removal of pillars 4 and 8.) It is apparent that the zone of sinking is a sort of expand- ing dome, which grows in proportion as the excavation extends. The bending of the first bed, hardly observable at first, is considerably increased. The second bed sinks rather less than the first, the third less than the second, and the sinking of each diminishes regularly in pro- portion as it is higher above the excavation. This sinking takes the form of a basin, the centre of which is on the vertical axis of the excavation. The lines A4, A7, A8, A9,. A11? A13, A17 are lines followed by the greatest deflections of the sunken beds after the removal of the pillars 4, 7, 8, 9, 11, 13, 17. These lines nearly coincide with the axes of the domes, which show the limits of the movement. * From University of Illinois Engineering Experiment Station Bulletin No. 91. Throughout the experiments it was evident after the removal of a certain number of the pillars that the pressure of the superincumbent mass was strong at the centre and weak at the circumference of the excava- tion. The second row of wooden pillars was taken away, and thus the depth of the excavation was doubled. The sinking of the lower beds increased; some of them fell in; and the broken ground occupied much more space. The disturbance was greater below, but not at the sur- face. The line of maximum deflection did not remain vertical, but some of the limiting domes were inclined. Removal of the third row increased the disturbance caused by the removal of the two former; the fractures of the beds and the spaces between the strata were multiplied; some opened more, others closed. As before, the movement started at the lower beds, and reached the upper as the excavation extended. The removal of each row of supports results in a new state of stability, which continues if no more pillars are taken away. Similar experiments were made with beds at various inclinations, and it was found that the line of greatest deflection was between the vertical and the normal, and that it departed further from the normal (that is, the perpendicular to the inclination of the beds) in proportion as the beds became more inclined. What- ever the inclination, the subsidence of each bed had always the form of a basin. When horizontal beds were covered over by beds dipping at various inclinations—that is, resting uncon- formably on them — the zone of settlement took the direction of the inclination of the beds, and its axis tended to become perpendicular to the beds affected. The lines drawn through the maximum bend of each bed were no longer continuous, but in passing from one set of beds to another were broken and shifted in the direction of the dip of the new set. In all cases, the sinking of each bed and of the surface was in the form of a basin. An experiment was made with horizontal beds, which showed that a block of coal left between two worked- out places may be of no use to protect the surface above it, because the zones of subsidence due to the excava- tion on either side, which, as already seen, take the form of domes, may overlap each other between the coal and 'the surface. As the area of subsidence increases in proportion as the excavation is extended, it may be asked whether there is any limit in depth to the propagation of the movement when the excavation extends indefinitely. To answer this, a mass of horizontal beds was isolated round about by a space being left between them and the vertical sides of the box, and then the wooden pillars (in this case 0-03937 in. thick) were taken away from under the whole area of the mass. Being entirely free at the sides, it might be considered to represent a mass of strata lying over the middle of a working of large extent. On taking away the pillars, the zone of sinking was seen to increase little by little, and to stop at a certain depth; the movement did not reach the surface. The expansion of the lower beds filled the space excavated, and the upper beds rested on the fallen rock. The pressure exerted by the upper strata was very much greater in the middle than at the circumference, and in this case, too, the sinking of the strata was in the form of a basin. The effect of faults was tested by inserting in a mass of horizontal beds a thin plate of metal, placed at an inclination, and extending the whole width of the beds. This broke the continuity of the beds, and represented a fault without throw. Its tendency was to stop the movement from extending above it, though the sinking occurred as usual on its low side, leaving an opening- in the plane of the cut, which extended to the surface. Fayol also made experiments upon the angle of frac- ture of rocks, the increase in volume of crushed rock, and the compressibility of crushed rock of various sizes. Effect of Lateral Compression upon Stratified Materials. Elaborate experiments were made by Willis in order to study the deformation of strata by compression. The substance used was beeswax, with plaster of paris to harden it and Venice turpentine to soften it, so that by using different proportions of these materials, beds of a wide range of consistency could be constructed. A load of shot was applied upon the beds when con- structed, in order to approximate the conditions under which strata at depth are deformed. The machine used for compressing the piles of strata endwise was a massive box of oak, provided with a piston which could be advanced by a screw. The pressure chamber was 3 ft. 3f in. long by 6 in. wide. The depth of the box was 1 ft. T. M. Meade made a number of experiments, and considered in detail the types of surface which may be developed. He used various kinds and combinations of bars, and applied pressure in various ways. An elaborate set of experiments was made to demonstrate circumferential compression. He used for this purpose discs of clay placed within a circumferential band which could be shortened. Effect of Vertical Compression upon Beds of Stratified Materials. Various tests upon bedded materials used for filling in mines have been made by the United States Bureau of Mines. Incidentally these tests have demonstrated the movement or flow of material in beds under pressure. Effect of Lateral Tension upon Stratified Material. Not very much work has been done to determine the tensile strength of rocks, and practically nothing has been done upon beds of stratified material. General Experiments. Experiments to illustrate geological phenomena and to discover the properties of rocks under conditions of pressure and temperature which may exist at great depths have been conducted by Daubree, Adams, and Coker, and various other scientists working at times privately, and at other times under the auspices of scientific bureaus of governments and of societies. Behaviour of Artificial Supports. Extensive tests have been made by the United States Bureau of Mines in various Government laboratories and by various mining companies in order to deter- mine the actual and the relative strength of different types of supports. Suggested Experiments and Tests. (1) In order to study surface subsidence resulting from the removal of supports, it is suggested that a model be constructed, say, on a 1/100th scale, both hori- zontal and vertical, approximating relatively the geological sequence of beds in a given district. The beds should have the same strength relatively in pro- portion to their weight, or the weight applied, as exists in the geological section which the model represents. The model should be of sufficient extent laterally to represent several panels of a pillar-and-room mine laid out on the panel system. Provision should be made for removing supports so that conditions such as would exist when pillars are drawn may be created. Observations should be made upon the height of sur- face from time to time, and, after surface movement has ceased, the model should be dissected so that the effects of subsidence below the surface may be noted. Similar models should be constructed to demonstrate working beds of various thicknesses, depths, and dips, and under other systems of mining. (2) Strength tests of roof materials should be made. The tensile strength and the angle of fracture in bend- ing tests should be determined. (3) The bending power of the various materials which constitute the mine floor should be measured. (4) In typical mines and under normal working con- ditions, the pressure or weight of roof should be measured and recorded over as long a period as possible at each point selected. (5) A study should be made of the composition and physical properties of the rock strata between the beds mined and the surface, and also immediately below the beds mined. PROTECTION OF OBJECTS ON THE SURFACE. The surface may be protected by the use of natural or artificial supports. Probably the most general method of preventing subsidence and of protecting objects on the surface is by leaving unmined a portion of the mineral deposit, with the idea that the pillar thus left will have sufficient strength to support the overlying rocks. In considering the service which a pillar may render, and in determining the size of the pillar or other sup- port for protecting specific mine openings or objects on the surface, it will be necessary to consider some of the following factors, and in some cases all of them: — (1) The unit strength of the material forming the pillar. (2) The height of the mine opening. (3) The dip of the mineral deposit. (4) The angle of break of the overlying rock. (5) The angle of draw or drag or pull over the pillars, as observed in the district or under similar conditions. (6) The strength of the overlying rocks. (7) The nature and amount of filling in the mined- out area adjacent. (8) The depth at which mining may be carried on without affecting the surface. (9) The bearing power of the bottom or floor. (10) The weight of overlying materials which must be supported. To determine the size of pillar necessary to protect mine openings of a given width, it is customary in some text-books to assume a span of roof and overlying rock to be supported, to estimate the total weight of such a block for the depth of workings, and then, with the known or assumed unit crushing strength of the material to be left in the pillar, the cross-section may be calculated. Such calculations are seldom used in practice, and they are open to the objection that they assume a pillar to be uniform throughout, while, as a matter of fact, all bedded deposits are composed of a large number of layers that may vary widely in hard- ness. For instance, some beds of very hard coal con- tain thin layers of mother coal which reduce the strength of the bed, thus vitiating any calculated results for strength of pillar based on tested specimens taken from the solid part of the bed. Shaft Pillars. Numerous rules have been formulated for the calcu- lation of shaft pillars in flat seams. Among the best known are the following : — Merivale.— D 50 in which S equals length of side of pillar in yards, and D equals depth of shaft in fathoms. Andre.—Up to 150 yds. deep, a pillar 35 yds. square. Up to 175 yds. deep, a pillar 40 yds. square. Up to 200 yds. deep, a pillar 45 yds. square, and so on, increasing 5 yds. for every 25 yds. of depth. Dr on.—Draw a line enclosing all the surface build- ings, such as engine houses, fans, etc. Make the shaft pillar of such a size that solid coal will be left in around this line for a distance equal to one-third the depth of the shaft. Wardle.—The shaft pillars should not be less than 120 ft. square, and the deeper the shaft the larger the pillars. Supposing the minimum to be 120 ft. for a depth of 360 ft., 30 ft. should be added for every 120 ft. in depth. Hughes.—Leave 1ft. in breadth for every foot in depth; that is, a shaft 600 ft. in depth should have a pillar 300 ft. in radius. Pa.mely.—For any depth to 300 ft., it may be suffi- cient to have a pillar 120 ft. square. Adopting this size as a minimum, we may fix any size of pillars for greater depths by increasing the pillar 1 ft. for every 4 ft. in depth.