An approach to the study of four practical problems of rock engineering using a micro-computer program is presented. The intent is to illustrate how a computer model, in these cases involving simple elastic behavior of the rock medium, sufficed to provide useful practical insight for the design. More sophisticated elastic/inelastic analysis, although available in the computer program, was felt to be not warranted in these cases, due principally to the absence of data on the mechanical properties of the rock mass. Specifics of the calculations and computations are minimized in this paper, but are explained in detail in several of the listed references.
The development of powerful, relatively inexpensive, micro-computers and computer programs able to model rock engineering problems has given engineers a potentially very valuable design tool. The microcomputer can be located at the job-site, even in remote locations, allowing the actual behavior of the rock to be compared with predictions; and design change options to be examined, with minimum delay.
However, the arrival of the microcomputer has not changed the rock. The large-scale heterogeneities and discontinuities, the variability and uncertainty of in-situ conditions remain, and must be considered in the design. Also, one has usually very limited detailed information on the mechanical properties of the rock mass prior to excavation, i.e., rock engineering problems are usually "data-limited." As has been noted,
A rock mass is a complex assemblage of different materials and it is very unlikely that its behavior will approach the behavior of the simple models which engineers and geologists have to construct in order to under stand some of the processes which take place when rock is subjected to load.
"Understand" is a key word in this statement.
Flexible, 'user-oriented' computer programs are now available which can help the engineer develop an understanding of how a structure in or on rock is performing, and indicate possibilities for improving the design. Empirical design rules based on successful designs for (apparently) similar projects, are valuable and should be used as a check on a proposed design. They can not provide the understanding-and the associated awareness of possibilities for improving the design-that can come from the proper use of models.
Starfield and Cundall (1988) have presented valuable guidelines on the use of models in rock mechanics, and note that Simplification is a crucial part of rock mechanics modelling. A model is an aid to thought, rather than a substitute for thinking..... They further advise.....plan the modelling exercise in the same way as you would plan a laboratory experiment. In particular, the modeler should ask, why am I building a model; how can I justify the particular model I am using; what did I learn from the modelling exercise?
We have attempted to follow this approach to modelling and offer four examples as illustrations. All of them are, to varying degrees, "data-limited" and were studied using the two-dimensional finite-difference computer code, FLAC, (See References) on a microcomputer.