The paper first analyses existing rock mass classifications. The authors propose a classification which is developed in stages, similar to the evolution of the design. Classification should start right from the choice of site and be developed all the way through to the final completion of the work. All the stages of classification together must constitute an integral system. In each successive stage the classification is filled in with new, more reliable and specific data, with more and more quantitative parameters, as produced by current investigations and required by the current stage of design. Thus its character changes from the most general in the initial stage to become completely specific at the end.


Cette communication commence par analyser les classification existantes. Ensuite les auteurs proposent une classification en plusieurs phases. Elle devrait être commencee dès le choix de la location des installations etfinir à la fin de la construction. Toutes les phases de la classification doivent presenter un système unique. A chaque nouvelle phase on doit ajouter de nouvelles donnees plus sûres et plus pecises avec des paramètres de plus en plus quantitatifs: autant que les recherches en cours le permettent et que les besoins progressifs du projet le demandent. De cette manière le caractère de la classification se changerait en devenant de plus en plus precis au fur et à mesure de l'avancement du projet.


In dieser Arbeit werden die bestehenden Methode der Gebirgsklassifikation analysiert. Die Autoren schlagen vor, daß die Klassifikation in Phasen gemacht wird, sodaß sie sich dem Entwurf ahnlich entwickelt. Mit der Klassifikation sollte schon bei der Wahl des Bauortes begonnen werden, und am Ende des Bauens sollte sie fertig sein. Alle Phasen der Klassifikation muessen ein System darstellen. In jeder folgenden Phase wird die Klassifikation mit immer mehr quantitativen Parametern erganzt,sodaß der Charakter der Klassifikation dem konkreten-Objekt immer naher kommt.


Underground construction engineering has for sometime now been a rapidly expanding field. Apart from the ever increasing need for road and water tunnels ensuing from the ramification of highway networks and the rapid multiplication of hydroelectric plants, many other works are now built underground: underground railways, subways warehouses, shelters, power: houses, recreation facilities, etc. Serious consideration is already being given to the construction of entire underground communities (F. Moreland, 1977). The increasing scope and-diversity of underground engineering projects has elicited advances both in investigation and design methods and in excavation and building technology. A specific feature of underground works compared with most other projects is, their very great dependence on geological conditions, so that the investigation and forecasting of these conditions is especially important. State-of-the-art design practices for underground works are based shiefly on classifications of rock masses, the results of site investigations, including in situ tests, and mathematical stress/strain models. These three groups of methods in fact characterize the three main design approaches. However none of them by itself can provide sufficiently reliable data at the right time for design purposes. In practice they are therefore usually combined and integrated. This may also be seen as the most likely trend in the future development of investigation and design methodology. This-paper critically reviews some of the most recent rock mass classifications, and presents a new approach which in our judgement would greatly augment the applicability of site investigation data.


An analysis of various classifications which have found wide application in underground engineering projects reveals that despite their indisputable value they all have certain shortcomings which limit their usefulness. One of the oldest classifications, widely used especially in the USA, is that of Terzaghi (1946). It is used mostly in the case of steel arch supporting, but is not applicable with more modern excavation methods. More recent classifications are adapted to the so-called New Austrian Method of tunneling, using, shotcrete, iron mesh and rock bolting. The classifications of Rabcewicz (1957), Lauffer (1958), and Pacher, Rabcewicz and Gosler (1974) introduce the excavation stand-up time into the classification scheme. In 1970 Deer proposed a classification for application in underground engineering projects in which rock mass quality is estimated from the length of core segments yielded by drilling (RQD). In 1972 Wickham, proposed a classification based on a number of parameters, to each of which he assigned numerical values. He recommends the kind of support to be used in different classes of rock mass. Apart from the RQD quality index, Louis also introduces a strength index from which the mechanical behaviour of the rock mass can be Important innovations in the classification system were introduced by Bieniawski (geomechanical classification) and by Barton, Lien and Lunde (Q-system). These two are also the currently most comprehensive classifications, derived for the most- part from -analysis of ground conditions during the construction of a large number of tunnels.

This content is only available via PDF.
You can access this article if you purchase or spend a download.