Results of Experiments of Secular Bending of Big Granite Beams Extending for 10 Years and Their Analyses

Abstract

This paper is a continuation of the first one entitled“Method to Find Secular Bending of Big Granite Beams and Results Obtained for the First Seven Years”, Journ. Soc. Materi. Sci. Janan, Vol. 14, No. 141, pp. 507∼519, 1965.
In order that the secondary creep (steady flow) can be experimentally confirmed, the creep-test should be continued for a period that equals the relaxation time of the material tested. From our experiments the relaxation time of granite is estimated to be 30∼300 years.
The deflection curve of the test-piece undergoing bending is assumed to be represented by y=T (t)·X(x), where T(t)is a function of the time t that has the reciprocal dimension of Young's modulus E and T(0)=1/E. Plotting the values of T(t) found from our experiments against t, the mean curves of T(t) for both the center-loaded and unloaded beams that are represented by broken lines in Fig. 4 strongly oscillate along some monotone and slightly increasing curves (not represented in the figure) during the first 4 to 5 years, and afterwards the oscillation for the center-loaded beam becomes markedly weak and that for the unloaded remains similar yet strong. It is noticed that the oscillations for both the two test-pieces run in the general trend parallel to the annual variation of humidity inside the laboratory. Elimination of the effect of humidity on the observed values of T(t) is one of our tasks that will be done next.
In the later part of this paper an interpretation of the above mentioned strong oscillations of T(t) is made to suggest that the observed values of T(t) would be apparent and the variation of the true values of T(t), corrected for humidity, would be monotone and slightly increasing with respect to time.
The conclusions obtained in this paper are: (1) Granite makes viscous flow or plastic one that has a very small yield stress beyond which the flow takes place. Since the unloaded beam has been flowing and its maximum stress is 12.8kg/cm², the above mentioned yield stress, if it existed, would be much smaller than this value. (2) Assuming T(t)=1/E+t/3η, the viscosity η of granite is calculated to be 10²⁰∼10²¹ poises, of which the greater value equals one tenth to one hundredth of the viscosity of the earth's upper layer under continents computed from the post-glacial uplift of Fenno-Scandia and equals that from the Pleistocene lake uplift in Utah, U.S.A.