A UNIFIED CONSTITUTIVE MODEL FOR
CLAYS AND SANDS
by
Juan Manuel Pestana-Nascimento
BSc (Eng.) Civil Engineering (1985)
Universidad Catdlica Andrés Bello.
Master of Science Civil Engineering (1988)
Massachusetts Institute of Technology.
SUBMITTED TO THE DEPARTMENT OF
CIVIL AND ENVIRONMENTAL ENGINEERING
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF SCIENCE IN CIVIL ENGINEERING
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
May, 1994
@ Massachusetts Institute of Technology
Signature of Author
Department of Civil and Envirenmental Engineering >
May, 1994
Certified by
Protestor'Andrews- Whittle
Thesis Supervisor
Accepted by
Professor Joseph M. Sussman
Chairman, Departmental Committee on Graduate Students
nieCHIVES,
m7 908A UNIFIED CONSTITUTIVE MODEL FOR CLAYS AND SANDS
by
Juan M. Pestana-Nascimento
Submitted to the Department of Civil and Environmental Engineering
on May 4, 1994 in partial fulfillment of the requirements for the degree of
Doctor of Science in Civil Engineering.
ABSTRACT
This dissertation describes the formulation and evaluation of a new constitutive model,
MIT-S1, which is capable of predicting the rate independent, effective stress-strain-strength
behavior of uncemented soils over a wide range of confining pressures and densities,
The MIT-S1 model introduces a simple four parameter elasto-plastic equation to
describe the non-linear volumetric behavior of freshly deposited cohesionless soils in
hydrostatic and one-dimensional compression. Sand specimens compressed from different
initial formation densities approach a unique condition at high stress levels, referred to as
the Limiting Compression Curve (LCC), which is linear in a couble logarithmic void ratio,
¢, effective stress space, 0’. Inecoverable plastic strains develop throughout first loading
and represent mechanisms ranging from particle sliding and rolling at low stresses; to
crushing, which is the principal component of deformation for LCC states. The model also
characterizes hysteretic and bounding surface plasticity observed in unload-reload cycles
based on previous model development at MIT.
Shear behavior is modeled by a single yield surface, which is a function of the current
void ratio and stress state, with hardening rules to simulate the evolution of anisotropic
properties. A separate failure criterion controls the frictional strength mobilized at large
shear strains. These features enable MIT-S1 to describe characteristic transitions from
dilative to contractive shear response as the confining pressure increases, Critical state
conditions are derived analytically as a function of model input parameters and form a non-
linear locus in loge-loga' space which is consistent with measured data reported recently in
the literature,
‘The shear behavior of sands in the LCC regime is normalizable and qualitatively
similar to that of clays resedimented from a slurry along a virgin consolication line (VCL}
Hence, the proposed model unifies the constitutive description of clays and sands. The
overall formulation requires the selection of fourteen input parameters (material constants)
that can be determined from standard types of laboratory tests or estimated from
correlations with other physical properties. Detailed comparisons with data for Toyoura
Sand show that MIT-S1 can predict reliably: a) the behavior of loose and dense sand with a
transition from fully collapsible to dilative behavior as a function of stress level and density:
b) variation of peak friction angle in drained shear tests and c) critical state conditions
measured in undrained shear tests. For clays, only thirteen input parameters are required,
and predictions for Boston Blue Clay demonstrate that MIT-S1 provides a significant
advance in modeling the anisotropic stress-strain-strength behavior especially at high
overconsolidation ratios.
Thesis Supervisor: Professor Andrew J. Whittle
Associate Professor of Civil and Environmental EngineeringACKNOWLEDGME
This work would have not been possible without the collaboration of many individuals
and institutions. I am first indebted to my thesis committee member
Professor Andrew J. Whittle supported my work and provided superior intellectual
challenges. In many senses, he has been a mentor and a friend; I have learned enormously
from him. 1 would also like to thank him for his meticulous review cf my thesis and for
trying to teach me (in vain) the use of the Greek alphabet.
Professor Charles C. Ladd provided input, careful review and guidance for this thesis.
In addition, he gave me the opportunity to teach several of his classes.
Dr. Jack Germaine offered his technical advice and provided numerous sources of
experimental data,
1 gratefully acknowledge the support of the Fundacion Gran Mariscal de Ayacucho
during my first two years at MIT, and the Instituto Tecnoldgico Venezolano del Petréleo,
INTEVEP, for their partial support of this research.
I would also like to thank the following individuals:
Dr. Franco Lupini for first motivating my inierest in soil behavior,
Professor Mohsen Baligh who motivated me to pursue research work in constitutive
modeling;
Dr. Manuel Vicente Mendez for initiating me into civil engineering practice and for
invaluable mentoring advice;
Dr. Juan jose Bolinaga, Dr. Ignacio Rodrigues Iturbe and Isaac Foinquinos for, in
their own particular ways, determining my stay at MIT;
Prof. Robert V. Whitman for guiding me into environmental geotechnics;
Professors Antonio Gens, Scott Sloan, M. Oda, and Cesar Sagaseta, for their
‘comments and feedback during the various stages of this research;
Dr. Lambe, Francisco Silva and Allan Marr (through T.W.Lambe, Inc.) and Isa
Romero and Franco Lupini (through GEODEC and GEOHIDRA, respectively) for
allowing me to participate in very interesting engineering projects while at MIT.
1 am specially grateful to my “soil-free” friends: Prof. Tim Sheahan, Dr. Spike
Hashash, Eduardo and Elena Mier-y-Teran, Andrés and Rianet De Leon, and my MIT
friends: Dante Legaspi, Samir Chauhan, Marika Santaga:a, Diana Zreik and Gebran
Karam. I would like to thank specially Doug Cauble for helping in reviewing the final
document and Mike Geer who made my time at the computer terminal more bearable.
To my parents I owe the most. This work would have not been possible without their
unconditional support and love. This thesis is dedicated to them, who always wanted a
Doctor, although I am absolutely sure this is not the kind they had in mind.
I would like to thank my wife, Sandra Mattar, for her innovative ways of providing
support and for her invaluable understanding during this very demanding period of my life
This thesis is completed literately thanks to her. Finally, 1 would like to thank my daughter,
Maria Teresa, for providing me with very useful insights in solving puzzles and giving
generously of her vitality, and joy in living. Her laughter has soothed my impacient walk
and has illuminated the sometimes not-so-bright path of my life
Juan Manuel Pestana-Nascimento.