Experimental investigations of soils under conditions of a general state of stress [1,2,3] showed that their deformation properties are described by" quite complex nonlinear relations of the form eav= ea~av,=i, =),where Oar and eav are the average stress and strain; oi and e i are the intensities of stresses and strains; a and m e are Nadai-Lode parameters of the type of state of stress and strain. Below in solving physical nonlinear problems we will use as the deformation characteristics of the soil the generalized shear modulus G and the volume strain K determined with the use of the relations of Eq. (1) for condition c~ e and c~.At present there are several methods of solving physical nonlinear problems: iterative, step (incremental), mixed step-iterative[4], etc. In the iterative method the secant moduli Gs and K s are used; the calculation is made for the complete load and is reduced to an iteration process continuing unffl the difference in the values of certain controlled quantities (e.g., displacements) for successive iterations becomes permissibly small. In the step method the load isdivided into partsand the calculaffon is carried out stepwise, by load increments, on the assumption that the deformation characteristics of the soil within a given n-th step do not change and correspond to the state of stress and strain after n-1 steps. The step method of loading permits using both the secant Gs, Ks and the tangential (differential) Gt, Kt moduli, in the mixed method incremental loading is combined with iteration processes at individual steps. The last two methods permit better control of the solution, since with their use the development ofthestress-strain state with increase of load is followed.Nonlinear problems are solved on the basis of the finite-element (FEM) and finite-difference (FDM) approximations (methods), methods of local variations, and others. The FDM with the iterative method was used, e.g., in calculating earth foundations and dams [5,6,7]. Problems for earth dams were solved by the iterative method with a finite element approximation [8]. The FEM is used below for solving physically nonlinear problems. A great virtue of the FEM is its indifference with respect to the character of the boundaries and stabilization of the region, laws of variation of the rigidity parameters and external load.To solve two-dimensional physically linear and nonlinear problems by the FEM, a program ('FLINZ-GP') was written in ALGOL-60 language for the BESM-6 computer (authors A. K. Bugrov and K. K. Grebnev) at the Departrnent of Underground Structures, Bases, and Foundationsof the Leningrad Pol}~echnic Institute. The program provided for the use of triangular (TFE) and rectangular (RFE) finite elements, and also their combined use. For certain outlines of the boundaries of the region its division into RFE is done in the computer, and in other cases preliminary division is used. Calculation of the mesh-point (node) forces as a function of the soil's weight is provided for in the computer. Analytic expressions or ...