Modelling of Strain Accumulation Due to Low Level Vibrations in Granular Soils Christian Karg

248 Pages·2007·2.1 MB·English

Modelling of Strain Accumulation
Due to Low Level Vibrations in Granular Soils
Christian Karg
Promotor: prof. dr. ir.W. Haegeman
Proefschrift ingediend tot het behalen van de graad van
Doctor in de Ingenieurswetenschappen: Bouwkunde
Vakgroep Civiele Techniek
Voorzitter: prof. dr. ir. J. De Rouck
Faculteit Ingenieurswetenschappen
Academiejaar 2007 - 2008
ISBN 978-90-8578-179-0
NUR 955, 956
Wettelijk depot: D/2007/10.500/53
Dit proefschrift is goedgekeurd door de promotor:
Prof. dr. ir. W. Haegeman Universiteit Gent
Samenstelling promotiecomissie:
Prof. dr. ir. R. Verhoeven Universiteit Gent
Prof. dr. ir. P. Troch Universiteit Gent
Prof. dr. ir. R. van Impe Universiteit Gent
Prof. dr. ir. W. van Impe Universiteit Gent
Prof. dr. ir. G. Degrande K.U. Leuven
Dr.-Ing. F. Wuttke Bauhaus University Weimar
Dr. ir. P. Mengé Dredging International N.V.
Published and distributed by:
Faculty of Engineering
Ghent University
Jozef Plateaustraat 22
B-9000 Ghent
Keywords: strain accumulation, cyclic triaxial test, bender element test, non-
cohesive granular soil
Copyright ©2007 by C. Karg
All rights reserved. No part of the material protected by this copyright
notice may be reproduced or utilised in any form or by any means,
electronic or mechanical, including photocopying, recording or by any other
information storage and retrieval system, without written permission from
the publisher: Faculty of Engineering, Ghent University.
Printed in Belgium

Preface
The work presented in this thesis could not have been conducted without
support of many people I would like to express my acknowledgements. First
of all, I wish to gratefully acknowledge Prof. Wim Haegeman, the supervisor
of this thesis, for his support, encouragement and helpful comments during
the course of the study.
This work is carried out within the frame of the SBO project IWT
03175 ‘Structural damage due to dynamic excitations: a multi-disciplinary
approach’. The financial support of the Flemish Community is also
gratefully acknowledged. In this regard I would like to express my gratitude
to the co-workers of the research project, especially to Stijn François for
stimulating and invaluable discussions and Prof. Geert Degrande from the
K.U. Leuven for his numerous useful suggestions and comments.
In addition, I am indebted to the staff of the Laboratory of
Geotechnics at the Ghent University and the Department of Industrial
Engineering Sciences at the KHBO in Ostende for their direct and indirect
support. Special thanks deserve Luc Boehme for his patient support during
the installation of the testing device and Kenny Deswelgh, who performed
most of the extensive laboratory works. Furthermore, I thank Dr. Lutz Karl
for many helpful discussions and creative ideas and Dr. Frank Wuttke for
his encouragement and inspiring discussions.
Finally, I appreciate the great support and encouragement of my
family for all intents and purposes. I deeply thank my wife, for her support,
patience and appreciation. The birth of our children has provided the
happiness that makes life much more complete.
Christian Karg
Ghent, September 2007
V

Abstract
Vibrations in the ground are a matter of growing concern, especially in
densely populated areas. Beside natural vibration sources as seismics and
wave loading, manmade vibrations gained in importance in recent years as
recent developments lead to a significant increase of the traffic volume. This
comes along with higher construction activities in urban areas as the built
environment needs to be adapted to new demands. Machine foundations
and wind power plants are further examples for manmade vibrations
sources. As a result significant higher vibration levels as in the past are
observed. This increased dynamic impact may lead to a reduction of
serviceability and diminish the lifetime of buildings. For this reason it is
important to investigate consequences of vibrations in the ground with
respect to their damage risks for buildings. Another important aspect is the
consideration of permanent or long term vibrations at low strains in the
design process. More detailed understanding of the mechanisms in the soil
during dynamic long term impacts would allow for optimisation in
dimension foundations, leading to economically advantageous design.
The understanding of mechanisms in the soil caused by low level
vibrations, their consequences on the built environment, their modelling and
prediction are in the focus of this work. In particular the elasto-plastic long
term behaviour of non-cohesive granular soil under cyclic loading at the low
strain level is studied in experimental investigations as well as in theory. The
conducted literature study revealed that single dynamic events at the low
strain level can be considered to be purely elastic as no plastic deformations
are detectable. Accordingly, the dynamic soil behaviour can be well
modelled using simple elastic constitutive laws. As a consequence, residual
deformations cannot be predicted using common dynamic soil models. Both
field measurements and laboratory investigations, however, have proven
VII
VIII ABSTRACT
non-negligible permanent deformations after application of large numbers
of dynamic load applications.
This process of strain accumulation is explained as a kind of aging
process of the soil, comparable to creeping due to dynamic loading in other
materials. In general, the accumulation process is characterised by three
main influencing aspects: the current state of the soil, the current state of
stress and the characteristics of the dynamic impact. The combination of
these parameters defines the intensity as well as the direction of
accumulation. Theoretical considerations in combination with experimental
studies on these accumulation phenomena are subject of main interest in this
thesis. Constitutive models for accumulation prediction in the literature are
studied and evaluated regarding their power and applicability.
Based on these cognitions a new accumulation model is proposed
for prediction of accumulation problems. The essential advantages of this
approach are the small number of model parameters, the simple formulation
and the proper implementation in a commonly used finite element code. All
model parameters can be determined from a single kind of laboratory test,
namely the cyclic triaxial test. The simple formulation allows for easy
comprehension, extension and general applicability for a wide range of
practical problems. The implementation in the finite element code supports
the use of the model in practical applications.
The validation of any accumulation model is difficult. Up to now
experimental investigations considering all, current state of the soil, current
state of stress and the characteristics of the cyclic loading, are quite rare both
in the field and in laboratory tests. For this reason advanced cyclic triaxial
and bender element test equipment is set up at the Laboratory of
Geotechnics, Department of Industrial Engineering Sciences, KHBO in
Ostende, Belgium in the frame of the present work. A test procedure
including sample preparation, saturation, consolidation, bender element and
cyclic triaxial testing of cylindrical soil samples is defined. The apparatus
allows for the investigation of elastic material properties and the
accumulation behaviour of soil. A wide range of test options allows for
consideration of many influence parameters in laboratory tests.
In testing campaigns the influence of different parameters on the
elastic material behaviour as well as on the accumulation properties of the
sand of Mol is investigated. Thereafter a hyperbolic non-linear constitutive
model well describes the degradation of soil stiffness with increasing cyclic
ABSTRACT IX
strain amplitudes. An empirical formulation, developed on the basis of
resonant column tests to describe the maximum shear modulus as a function
of the void ratio and the mean effective stress, is also valid for bender
element test results. An influence of the cyclic loading frequency on the
elastic material properties is not detected in the range of 0 to 10Hz.
Test results confirmed the applicability of elastic constitutive laws
for description of the dynamic soil behaviour. The elastic material
parameters are found to be independent from the number of load cycles
already applied. During accumulation of large numbers of cycles, however,
non-negligible residual deformations are observed. Increasing void ratios
and cyclic strain amplitudes lead to increasing strain accumulation rates. An
influence of the cyclic loading frequency on the accumulation behaviour is
not detected in the range of 0 to 10Hz. Therewith the frequency content of a
dynamic impact is of minor importance for accumulation predictions. Test
results of long term cyclic triaxial tests are used for validation of the
proposed accumulation model. The procedure of parameter identification is
described. Good agreement of test data and the accumulation model under
consideration is achieved.
Christian Karg
Ghent, September 2007