In order to develop a new method which predicts a occuring time of the slope failure due to a rainfall, the laboratory and field experiments were carried out.
The changes of the pore pressure with time by the laboratory model tests are shown in Fig. 2. The pattern and occuring time of the slope failure are shown in Fig. 4. The results show that the prediction of a occuring time of the slope failure is possible by the measurement of the pore pressure.
The changes of the pore pressure with time by the field experiment are shown in Fig. 14. The changes of the displacement gauge and strain gauge with time are shown in Fig. 15. The pore pressures increased sharply from fifteen minutes before the failure, while almost no displacement of the slope surface and no strain in the slope was measured. Then the prediction based on the measurement of pore pressure in a slope during a seepage of a rainfall seemes to be probable.

View full abstract

In order to study the process of slope failure caused by heavy rainfall, some large scale slope models of loamy and sandy materials were constructed. Artificial rainfall was applied to these models until a large scale failure occurred. Surface displacement were monitored by extensometers set on the models from the beginning of experiment to the final failure. Through analysis of the data, a rule on slope movement in the process of failure was found.
The increment of the logarithm of velocity of surface displacement is proportional to the logarithm of acceleration of the surface displacement in the final stage of the slope movement just before the failure. That is d²χ/dt²=a (dχ/dt) ^αwhere χ is downward surface displacement along the slope, t means time, d²χ and dχ/dt show acceleration and velocity of surface displacement, a and αa are constants. Usually, α took 1.5 to 2.2 in range in the experiments.
From the integration of the equation for the range of α>1, an equation, 1/υ = {a (α-1)} ^{1/ (α-1)} (t_γ-t) ^{1/ (α-1)} is obtained, were 1/υ is 1/ (dχ/dt), that is inverse number of velocity and t_γ is a constant of integration. The equation shows that the curve written by each point of t and 1/υ is linear if α=2, convex if α>2 and concave if 1<α<2 and that the value uniformly decreases. In the case of α=2, the failure time of a slope can be predicted exactly from the time (t_γ) when 1/υ equals 0 (infinite velocity).
Furthermore, the equation can be modified as (1/υ) / (d (1/υ) /dt) = (α-1) (t_r-t). The equation shows that the curve written by each point of t and (1/υ) / (d (1/υ) /dt) is linear for the range of α>1. Based on the relations mentioned above, the failure time of a slope can be predicted exactly in the case of α≠2. Also a graphical method to show the failure time, based on the equation, is proposed.

View full abstract

According to the results of the exploration of slide surface by the test pits, it was proved that the slide surface could be classifed into two types-the shearing type and the frictional type. Of these two types the shearing type used to be called “the slide surface”, and on constants of the shearing strength a report has already been made.
The slide surface of the frictional type is the surface on which thin water layer works as a lubricating agent. So, in layers, such slide surfaces exist discontinuously with their angles of friction being very small and the surfaces unsteady.
On the other hand, the slide surface of the frictional type isn't affected by the pore water pressure, and so when the slide surface of this type is distributed in the layers, the landslide occurs only when the pore water pressure works on the part of the slide surface of the shearing type.

View full abstract

Movement type of landslides seems to be affected by clay fraction (CF) of landslide materials. Landslide susceptibilty will be also predictable by the clay fraction
In mudstone areas, the highest potential of landsliding will be induced by reduction of shear strength in landslide materials to residual strength and elevation of water level up to the slope surface. Safety factor and critical slope angle in this case are determined.
The safety factor is relatively low in the areas of high susceptibility which is inferred from clay fraction (CF). Those areas usually have large difference between the critical slope angle and a real dip of a slope.

View full abstract

The Hayama-Mineoka ophiolite belt, in which landslides have frequently occured, is exposed in the southern portion of Boso and Miura Peninsulas about 50 km south of Tokyo. Ophiolite complexes occur in a narrow uplifted zone in the direction of ESE-WNW. Pillow and dyke basalts are exposed throughout this zone. ³⁹Ar-⁴⁰Ar age of Mineoka basalts ranges from 40 to 50 Ma. These ages show the dates of eruption at the ocean floor.
The paleomagnetic investigation indicates that the underlying basalts layer is tilted by about 28° nearly southward and the direction of the paleostress was NNW at that time of the formation of these ophiolite complexes. For the structual model derived from paleomagnetic investigation, the gravity and magnetic anomalies are theoretically calculated and found to be consistent with the observed anomalies.
The relation between the topography characterizing landslide occurence and the geography of Hayama-Mineoka ophiolite belt has been investigated. It is recognized that in Boso and Miura Peninsulas landslides have been observed appreciably at the area composed of igneous rocks in Mineoka and Hayama belt. It is noted that the frequency in landslide occurence is high in the southern slope and low in the northern slope of Mineoka Hill.
The obvious correlation has not been found between the distribution of landslides and water vein streams and also an angle of inclination of topography. It is speculated that ophiolite complexes was remarkably fractured by the NNW stress when it was uplifted on the land, and that the change of paleostress direction caused the release of stress in the cracks of underground basalts layer. The occurence of landslide in this district may be related intimately to the shape of underground rock body.

View full abstract