To evaluate how the scale effect regression curves must be seen with reference to works, it is essential to clarify whether they are normal or reverse. This matter is controversial in the field of joint roughnes and strength, but in this paper is examined in detail. We concluded that the present opinion which says that those scale effects are usually normal ones (Bandis, 1980; Barton, 1990) must be revised and checked, since, on the contrary, in fresh and unweathered joints they are usually reverse scale effects. Some additional experimental conclusions in this field are added. The inadequacy of Patton's model to solve these problems is cited.
Pour evaluer la signification des essais à petite echelle pour les oeuvres il est fondamental determiner si l'effet d'echelle est normal ou inverse. Cette matiere controverse dans les domaines de la rugosite et de la resistance des discontinuites est examinee en detail. Nous inferons qui a besoin de revoir l'actuelle opinion qui affirme que leurs effets d'echelle sont normals, selon les idees de Bandis (1980) et Barton (1990), Pour le contraire, en des discontinuites recentes, pas alterees, en general l'effet d'echelle est inverse. Nous additionnons quelques conclusions fondees dans des essais experimentaux. Nous referons l'insuffisance du modele de Patton pour resoudre ces problems.
Um abzuschatzen, wie die Maßstabseffektregressionskurven bezueglich der Bauten gesehen werden muessen, ist es wesentlich, zu klaren, ob sie normal oder urngekelut sind. Dieses Thema ist im Feld der Kluftrauhigkeit und -festigkeit strittig, aber, in dieser Mitteilung, wird es ausfuehrlich untersucht. Wir kamen zu der Schlußfolgerung, daß die gegenwartige Meinung, die sagt, daß jene Maßstabseffekte gewöhnlich normal sind (Bandis, 1980; Barton, 1990), revidiert und nachgeprueft werden muß, da sie, im Gegenteil, in frischen und unverwitterten Klueften, im allgemeinen, umgekehrte Maßstabseffekte sind. Einige zusatzliche experimentelle Schlußfolgerungen in diesem Feld werden vorgestellt. Daß des Modell von Patton unpassend ist, um diese Probleme zu lösen, wird erwahnt.
Based on a model with weak links organized in series alternating with strong links, where the whole set collapses when one weak link fails, Weibull (1951) showed that the failure probability of a body increases with the growth of the test volumes. This philosophy, which strictly speaking, is only valid for the tensile strength of the rocks (Brown, 1971), was unsuitably extended to almost all scale effect fields. So, there is the generalized idea resulting from the Weibull philosophy that the normal scale effects (represented by an average negative exponential function) are the habitual ones for all the rock mass properties and the reverse (or inverse) (represented by an average positive exponential function) are the exceptions. Leal Gomes (1998) showed experimentally that the inverse scale effects are frequent depending on the strength and deformability properties of the rock being explained by weak links organized in parallel with strong links as 1with the strength and deformability of some intact porphyritic granites.
Actually, it happens that the appearance of new "characteristic dimensions" or higher heterogeneity hierarchies, usually, diminishes the values of the properties, with the growth of the size tests and the scale effects result normal.
