The decomposed granite soils known locally as "Masado" are regarded as a special type of soils, because the soil grains are relatively unstable under mechanical or chemical actions. The properties of individual soil grains, which are important for the engineering purposes are examined by means of physical, chemical and mineralogical analyses.From the results of these analyses, it is concluded that the specific gravity of feldspar and the content of coloured minerals are the important factors to determine engineering properties of the soils.

View full abstract

Maximum heaving pressures caused by ice lens growth were determined experimentally for saturated specimens consisting of fragmental particles in several size ranges. Theoretical predictions show that the pores produced by the smaller particles in the system are responsible for the maximum heaving pressures measured. The suggestion is that the ice-water interface assumes an undulating configuration over the smaller pores when ice proliferation stops, and hence heaving pressures rise to a maximum.Particle size seems to be an adequate basis for predicting frost susceptibility in practice although the exact amounts and size limits permissible have not been evaluated.The nature of the failure caused by frost action is not the same for all engineering structures. Failures by "rate of heave" and "heaving pressure" should both be considered if the possibility of frost damage exists. The results substantiate earlier conclusions that there is no sharp dividing line between frost-heaving and nonfrost-heaving soils.Recent studies by the author have shown that the theory of Everett (2) and Everett and Haynes (3) provides a realistic basis for the growth of ice lenses from pure water in porous systems of simple and uniform geometry. Difficulties arise when the theory is applied to soils which have a more complex porous structure because a grain size, and hence a pore size, distribution exists. For a porous system of this complexity, the theory as it stands is not adequate for prediction, since little is known about the size of pores, in a given pore size range, that determine the heaving pressure characteristics.The two laboratory approaches for assessing frost susceptibility ("rate of heave" and "heaving pressure") are not entirely compatible, but it is clear and consistent that soils which are not frost-susceptible should not exhibit any measurable heaving pressure or rate of heave. Soils that have a low heaving pressure may show a considerable rate of heave under little or no restraint. On the other hand, some soils that exhibit a high heaving pressure from ice lens growth, such as dense clays, may due to permeability limitations have a sufficiently low heaving rate to be acceptable for some engineering structures subject to freezing conditions, (Fig. 1).[graph]The trend in soil engineering has been to predict the frost susceptibility of a soil from its particle size characteristics through relationships determined by "rate of heaving" experiments. Many organizations have carried out such tests but there are no generally-accepted limits to particle size or percentage of fines which determine frost susceptibility.In this paper attempt is made to elucidate further the mechamism of heaving, particularly the development of "heaving pressures" by ice lens formation in relation to the particle size distribution of the sample. It will be shown that from measurements of heaving pressure it is possible to obtain an estimate of the sizesof particles which are responsible.

View full abstract

Many investigators have studied piping (internal erosion) which was one of the causes of failure in many polder dikes, but most of the formulas proposed for the critical hydraulic gradient for piping were not sufficient in practical application. Namely, Bligh's and Lane's coefficients (they are referred to as C_b and C_w respectively) and C', which is expressed by C'=l/h=5∼10 are not adequate in the case of polder dikes.Mereover, the following coefficient C, which is derived from Terzaghi's theory, cannot fully represent the practical phenomena.C=C'(1-p)(γ_s-γ_w)/γ_w The author studied the piping phenomena with prototype tests and succeeded in deriving a new formula for the critical hydraulic gradient.In this paper, piping phenomena are classified into three types : (1) subsurface erosion, (2) heave and (3) internal erosion. Investigations on the first two types have been carried out and their definitions were established. The author newly defined internal erosion which had not been studied before, and derived an experimental equation from model tests and prototype tests in order to prevent the failure by piping in enclosure banks for land reclamation.The basic equation is given as I_f=i_s{(1-p)(γ_s-γ_w)/γ_w}ⁿ in which I_f is the critical hydraulic gradient, p is the porosity, γ_s and γ_w are the densities of sand and water, respectively, and i_s and n are the coefficients corresponding to soil characteristics and flow condition. As for sandy banks, the values of i_s and n were found to be 0.2 and 2, respectively, and the following equations were derived.I_f=0.2{(1-p)(γ_s-γ_w)/γ_w}²=(1-p)(γ_s-γ_p)/γ_w γ_p=γ_w+χ χ={(γ_s-γ_w)/γ_w}{(γ_w-0.2(1-p)(γ_s-γ_w} in which χ is a factor corresponding to ouopiping.From the above relationships, the safe seepage length L for a critical head H is given by the formula L=C_n·H·F=5{γ_w/(1-p)(γ_s-γ_w)}²·H·F in which F is the safety factor and C_n is a coefficient derived by the author.The theoretical analyses of these equations are also given.The propriety of the above-mentioned relation was suggested in the restoration work of the enclosure bank at Kasaoka Land Reclamation Project in 1956. This has given a possibility of rational design of the polder dikes of Nagasaki Land Reclamation Project and many other projects in Japan.

View full abstract