RHEOLOGY OF THE CRUST BASED ON LONG-TERM CREEP TESTS OF ROCKS

For 20 years Kumagai and Itô have carried out creep tests on granite beams each of 215 × 12.3 × 6.8 cm. They have concluded that granite flows plastically with a vanishingly small yield stress, that is, viscous flow with a viscosity of 3-6 · 1020 P. Itô and Sasajima have carried out more accurate creep tests by bending three granite test-pieces (each 21 × 2.5 × 2.0 cm) and three gabbro test-pieces (16 × 2.0 × 1.5 cm). Deformations were measured using a light-interference technique. The results obtained over a three-year period show that the creep in granite is comparable with that found formerly, and that gabbro has a slower rate of creep than granite. Consequently, over a long period of time, rock may be represented rheologically as a Maxwell liquid. We may thus imagine the crust to be modelled by a Maxwell liquid, which floats on a more fluidal underlying layer supporting the crust isostatically. Based on this conception, subsidence of guyots and atolls has been investigated and the viscosity of the ocean crust has been found to be 1025-26 P. Analysing the Quaternary crustalmovements in the Himalayas and in southwest Japan, the viscosity of the orogenic crust has been estimated to be 1022 P. The viscosity of the cratonic crust seems to be greater than that of the orogenic crust. From the viscosities estimated above and the generally accepted value of rigidity of the crust, we can draw the stress-strain diagrams of the crust for different strain rates. For example, the diagram obtained for the orogenic crust suggests that the crust would behave elastically for times shorter than the relaxation time of 3000 years and would flow as Newtonian liquid for longer times, and if a shearing strength of the crust were greater than 10 bar and the shearing strain rate were smaller than 10-15/sec, the crust would flow without fracture. © 1979.