A reevaluation of the performance of the Lower San Fernando Dam during the February 9, 1971 San Fernando earthquake was conducted to evaluate methods of determining the residual shear strengths of liquefied soils. The use of Standard Penetration Test (SPT) N-value correlations, as proposed by H. B. Seed and his colleagues at the University of California, and laboratory tests to obtain steady-state strength as proposed by Gonzalo Castro and his colleagues at GEI, Inc., are evaluated. A recommended approach for seismic safety analysis is presented. The approach, which is intended for use by practicing engineers and not for problems of a research nature, generally consists of the techniques developed by Seed and his colleagues and incorporates routine use of a field performance-based correlation between soil strength and SPT blow counts.

Abstract-The complex viscosity of a material is a two-component quantity, comprising real and imaginary parts. The real part of the complex viscosity is often very useful because in many materials it approaches the ordinary steady-flow viscosity at low frequencies, Because many materials with high viscosities are very slow in reaching a steady-flow condition, the determination of the steadyflow viscosity may be very difficult; however, an approximation can often be obtained from lowfrequency values of the real part of the complex viscosity. In this study, the complex viscosity of a Georgia kaolin has been determined by measurements made on specimens subjected to oscillatory simple shear, over three decades of frequency. Other independent variables in the study are the water content of the clay and the shear strain amplitude. Data were obtained from experimental measurements in the form of values of the magnitude, or absolute value, of the complex viscosity, and the phase angle between the imposed oscillatory strain and the stress response. Empirical functional relationships are developed to relate these quantities to the independent variables, and these are in turn used to obtain the real and imaginary parts of the complex viscosity as functions of the independent variables. The results of this study indicate that the complex viscosity is not linear, but decreases approximately as a power function of the strain amplitude; the relation between the complex viscosity and the water content is approximately an inverse logarithmic one, and changes very rapidly at water contents near the liquid limit; and the phase angle increases with increasing strain amplitude approximately as a power function. INTRODUCTION THERE are two important types of anomalies which may appear in the mechanical properties of a material. Stress anomalies occur when strain or ~train rate depends on stress in a manner other than a direct proportionality, and time anomalies occur when stress depends on both strain and strain rate together, as well as higher time derivatives of strain. In general, both kinds of anomalies will coexist to some degree in all materials. Evaluation of these anomalies is important in determining the macroscopic mechanical properties of a material for direct applications; in addition, some information on the nature of the microscopic material behavior can often be inferred from the macroscopic behavioral characteristics.In the work reported herein, the complex viscosity of a Georgia kaolin clay is determined from measurements made on specimens subjected to oscillatory simple shear. The independent variables in this study are shear strain, frequency of oscillation, and water content, whereas the dependent variables are shear stress and phase angle. Despite observed stress anomalies, the response is expressed in the terminology of linear viscoelastic theory; in terms of response functions, the real and imaginary components of the complex 101

No abstract

No abstract

Abstract-A Georgia kaolinite, at water contents from 55 to 95 per cent, was tested by means of a Weissenberg Rheogoniometer under conditions of pure shear with sinusoidally varying deformation over a frequency range of 3 decades. The results, including time-dependent effects, are expressed in terms of the magnitude of the complex modulus and the phase angle, as developed in the theory of linear viscoelasticity, and stress-strain hysteresis curves. The complex modulus is a two-component quantity, which has a real part associated with the elastic or energy storage characteristics of the material and an imaginary part associated with its viscous or energy dissipation characteristics. Although the complex modulus interpretation is very good for linearly viscoelastic materials, its applicability and usefulness diminishes as the material departs from linear viscoelastic behavior. On the other hand. the determination of energy dissipation from stress-strain hysteresis curves does not depend on any assumption concerning material behavior, because the area enclosed by the curve gives a direct measure of the energy dissipated in a single cycle of deformation. The dissipation characteristics obtained by the two methods are compared and used to illustrate the degree of validity and some limitations of linear viscoelasticity theory.