The Rock Load (Terzaghi) Approach, the RQD-, RSR-, and "Geomechanics" methods and the Q-system were applied at Porter Square Station, a 21 m span subway station cavern in Cambridge Argillite. Rock classification and support predictions were made in four phases (boring records, exploration shaft, pilot tunnel, cavern survey) by two independent investigators. The predictions were compared to actual pilot tunnel and cavern support. Results of the study show that, rock classification and support predictions are affected to some extent by the increasing information as one goes from phase to phase; subjectivity only affects prediction of spatial variations. For most probable, (average), rock conditions the results of all empirical methods are comparable, while most methods are not suited to handle unfavorable conditions in such a large, cavern.
La classification des roches dans la station souterraine de Porter Square a ete effectuee en utilisant les methodes de Terzaghi, les methodes RQD, RSR et "Geomechanics" et le système Q. Il s'agit d'une galerie souterraine d'une largeur de 21 m dans de l'argilite'de Cambridge. La classification des roches et la prediction des dimensions des supports ont ete effectuees separement par deux personnes en quatre phases (sondages, puits de reconnaissance, galerie pilote et observations dans la galerie finale). Les predictions ont ete comparees aux observations dans la galerie pilote et dans la caverne. Les resultats indiquent que les predictions sont legèrement influencees par l'accumulation d'information d'une phase à l'autre; par contre la subjectivite joue un rôle seulement dans l'estimation de la variabilite spatiale. Pour les conditions geologiques moyennes les predictions obtenues par toutes les methodes sont comparablesi par contre pour une galerie d'une telle largeur, la plupart des methodes se sont revelees peu adaptees à des conditions geologiques defavorables.
Die Terzaghi-, RQD-, RSR- und "Geomechanics"-Methoden und das Q-System wurden fuer die Gebirgsklassifizierung der Porter Square Untergrundbahnstation angewendet. Es handelt sich urneine Kaverne von 21 m Spannweite in Cambridge Argillit. Gebirgsklassifizierung und Ausbaudimensionen wurden unabhangig von zwei Personen und in vier Phasen bestimmt, namlich aufgrund von Bohrungen eines Erkundungsschachtes, eines Pilotstollens und aufgrund der Aufnahme in der ausgebrochenen Kaverne. Es zeigt sich, daß die von Phase zu Phase zunehmende Information einen gewissen Einfluß auf Gebirgsklassifizierung und Ausbauvorhersagen hat. Subjektivitat hingegen beeinflußt nur die vorhergesagte Variation langs der Kavernenachse. Die Resultate alle Methoden sind vergleichbar, solange die geologischen Bedingungen durchschnittlich sind. Fuer unguenstige Extrembedingungen in einer so großen Kaverne sind die meisten Methoden dagegen nicht geeignet.
Formal empirical rock classification methods for tunneling, have played a role, for over 100 years and new methods have developed at regular intervals during the last 40 years particularly since the late 60's, the creation of new methods has accelerated: Terzaghi (1946): Stini (1950), Lauffer (1958), NATM (1965, Rabcewicz (1965), Muller (1978), RQD-Deere (1969). RSR (1974), Q (1974), Geomechanics-RMR (1973–79). Louis (1974), Franklin (1975,1976). Users and creators of the methods are naturally interested in the accuracy of prediction and correlation between methods. A significant number of comparative studies have thus been conducted either for specific cases or as general comparisons (e.g., Barton, 1977; Blackey, 1979; Bieniawski, 1980, 1979b; Steiner et al., 1980; McCusker, 1980; Rose et al., 1981) without reaching any definite conclusions as to superiority of a particular method or generally valid correlations. The investigation on which this paper is based does not pretend to provide this answer either. However, the study's context and several of the issues it addresses are different and should therefore provide additional insight. These differences lie in the size of the opening (a 21m (70ft), span subway station), in the fact that not only empirical, but also analytical and numerical methods are compared, and in the investigation of effects of subjectivity and of available information. The comparison of five empirical methods (Rock Load -Terzaghi, RQD, RSR, Geomechanics-RMR, Q) with regard to the influence of subjectivity and of available information is the topic of this paper. Information increased through 3 phases of exploration to the 4th phase, a survey in the final cavern. In most of these phases the methods were applied by two independent investigators.
Porter Square Station is an underground rock chamber of 21 m (70 ft) span,14 m (45 ft) height and 150 m (500 ft) length, with a transverse crossover tunnel (span 12 m, height 14 m (Figures 1,2). The rock is Cambridge Argillite. Figure 2 shows the staged excavation procedure that started from an access shaft, which together with the pilot tunnel had been previously excavated (see 2.2). Regular cavern supports consist of bolts (dowels), W8 × 40 steel sets at 1.5 m (5 ft) spacing and continuously blocked and embedded in shotcrete with welded wire fabric. (Figures 2a,b). Support installation followed immediately behind the excavation of 1 round (5 ft).
