Abstract
SummaryThe simulation of the viscous behavior of some clays is of high importance in many geotechnical problems. The literature offers a vast amount of constitutive models able to simulate the rate dependence observed on these materials. Although most of these models are calibrated to very similar experimental observations and share similar definitions of material parameters, some discrepancies of their response have been detected, which are related to their mathematical formulations. In this work, the causes of these discrepancies are carefully studied. To that end, four different model families are analyzed, namely, nonstationary flow surface (NSFS) models, viscoplasticity with overstress function (OVP), viscoplasticity with Norton's power law (NVP), and visco‐hypoplasticity (VHP). For the sake of a fair comparison, single constitutive models using the same set of material parameters, and following other requirements, are developed for each model family. Numerical implementations of the four resulting models are performed. Their response at different tests are carefully analyzed through simulation examples and direct examination of their constitutive equations. The set includes some basic tests at isotropic stress states and others as responses envelopes, undrained creep rupture, and an oedometer test with loading, unloading‐reloading, creep, and relaxation. The article is concluded with some remarks about the observed discrepancies of these model families.
Highlights
The consideration of these relations in the formulation of rate-dependent constitutive models has allowed users to simulate rate effects on some boundary value problems dealing with viscous clays
One can demonstrate that substitution of the latter relation in Equation (11), and assuming that creep volumetric strains are produced by viscoplastic strains ε.vvis, yields to the following relation: ε. vvis where Dr = Cα∕t0 is a reference creep rate, already introduced in other works, and Iv is, by some authors, referred to as the viscosity exponent or the Leinenkugel index,[76] which reads
With the aim of performing a fair comparison, the viscous models are required to account for a set of characteristics, which are established in the following lines: 1. Isotropic compression under the reference strain rate, ie, ε.v = Dr, yields asymptotically to the reference isotach
Summary
Most clays show a rate-dependent mechanical behavior as presented in many experimental works.[1,2,3,4,5,6,7,8,9,10,11] Experience has shown that rate dependency influences many geotechnical problems, for example, long-term settlement of structures,[12,13,14,15,16,17] slope creep,[18,19] pile shafts penetration,[20,21] and ground anchor relaxation.[22,23,24,25,26] Clearly, the study of the soil's viscous behavior is of relevant importance. One may believe that two different viscous models whose formulations are adjusted to identical critical state relations, and to the same empirical relations for viscous effects, should deliver the same response. This is, not true: The literature offers different theoretical approaches to develop the model formulation, presenting important discrepancies between them, of which users should be aware of. The set of simulations include isotach curves, creep tests, stress relaxation tests, response envelopes, and undrained creep tests This analysis is complemented with the simulation of an oedometric test of a Kaolin clay showing many rate effects.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
