Aluminum bars were selected to form a physical model for the simulation of the specific features of highly jointed rock stratum. Various factors such as the length and number of the rock bolts were investigated for tunnel stability using trapdoor apparatus to understand the mechanical behavior of rock bolt in an underground opening. Comparisons between strain distribution and configuration of rock bolt were also made to clarify the mechanism of the supporting system in the underground opening. The results show that if the major discontinuity of the rock mass was horizontal, the extent of the plastic zone around the underground opening can be restricted in a narrow range due to the installation of rock bolts. The elastoplastic zone near the opening converged to a constant value as the rock bolt length was twenty percent higher than the equivalent radius of the opening (R). At this stage, the bolt is at the most economical embedded position. Further investigations indicated that the optimum embedded length of the rock bolts is 2R.
When engineers try to excavate an opening through rocks, the original stress field would be disturbed. It might deform the rocks especially around the opening, and then lead to the failure of tunnels. In order to control the deformation of tunnels, engineers always design the supporting system based on their accumulated experiences. In order to verify the stability of the openings, observations by means of displacement measurements such as extensometers used in the field as well as physical model tests are desirable because of the simplicity of instrumentation and reliability of the data obtained. Hence, the concept of pilot tests through simple physical modeling approach was utilized for the understanding of in-situ behaviors of rock mass. The physical modeling approach has become more powerful and effective when examining and assessing the failure mode of underground structures. Goodman (1969) first stacked the sugar cubes to study the feasibility of plane base friction models. Later on, scaled physical models were used by Barton (1979) to investigate the two-dimensional deformation of very large underground opening. The models consisted of at least 20,000 discrete blocks. Joint orientation and stress levels were varied, and some models were dynamically loaded to simulate strong earthquakes. From 1974 to 1979 Chappell studied the deformational responses and internal stresses acting among the units in blocky model experiments by using photoelastic techniques and numerical simulation. In 1972 Goodman and Heuze studied the laminated rock using rock bolts as reinforcing members. Mark (1982) had also conducted similar tests, and suggested that the most effective roof-bolting pattern in uniformly-bedded roof is that of the bolts clustered at a distance from the rib equal to 1/5 of the total span. Sakurai (1985) conducted the study on rock slope protection of toppling failure by using two-dimensional aluminum block models. In his presentation, varying degrees of discontinuity and different joint patterns in relation to bolting technique were investigated. Application of the critical strain to the estimation of failure slopes were also discussed.