The design of a 100,000 lb capacity 'stiff' machine recently completed by the Department of Mining of the Colorado School of Mines is described and a few typical results obtained using the machine are presented. This machine, which was built for about $2000, is intended primarily for use in undergraduate and graduate laboratories rather than as a research tool since strain rates cannot be easily controlled.
The basic loading frame is constructed using a 5-ft long steel tube (8" O.D., 1" wall thickness). End caps 3" thick are bolted into the ends of the tube and windows are machined near the upper end of the tube for inserting the samples. The rock sample is preloaded to a value of about 90% of its compressive strength using a 30-ton capacity hydraulic jack. Flat jacks positioned about the lower loading ram are then pressurized. This firmly fixes the lower platen to the wall of the tube and essentially removes the 'soft' hydraulics from the system. Additional load (actually displacement) is applied to the sample by heating the upper aluminum loading head with the use of a heating tape. The load on the sample is measured using strain gages mounted on the lower platen and the specimen deformation is monitored using two DCDTs connected in series.
The apparent stiffness of the machine is about 2.0 x 106 lbs/in. which has been sufficient to control failure in a number of rock types and specimen sizes. Typical results obtained when using the machine to test sandstone, limestone, shale, and coal samples are presented.
A laboratory experiment that is usually included in the beginning course in rock mechanics is the determination of the stress-strain curve and the compressive strength of a rock cylinder loaded in uniaxial compression. If the test is performed using an ordinary hydraulic compression testing machine upon reaching the peak strength (compressive strength), the sample fails in a more or less violent fashion, depending on the brittleness of the rock. Extrapolating these observations to the field, one might logically expect that a rock pillar upon being loaded to its peak strength would fail in a similar violent fashion and be reduced to a few large pieces. All of the load previously carried by this pillar would then be transferred to neighboring pillars, which if they were sufficiently close to their peak strengths at the time of failure of the first pillar, might also fail, transferring their loads to adjacent pillars, etc. If continued, this process would result in the collapse of the entire mine in a house-of-cards fashion. In general, mine pillars do not fail in this manner, but rather will slough, crack, and ravel in a quasi-stable fashion when displacement greater than that corresponding to the peak load is applied. They continue to carry load, although at a reduced level
Only rather recently it has been shown that the behavior of rock in the post-failure region (Fig. 1) depends upon the relative stiffnesses of the loading system and the rock.