Jean Biarez: His Life and Work - Multiscale Geomechanics: From Soil to Engineering Projects - page 6

At the same time, Biarez initiated research studies using new numerical techniques
that culminated in several PhD theses at the University of Grenoble.
These advances, however, produced as many problems as they did solutions. The
constitutive models introduced into the FEM were more complex and the
measurement of the parameters required more precise experimental protocols in the
laboratory. Most of all, it was difficult to obtain the values through in situ testing, on
the one hand, and to take into account the dispersion of the soil properties, on the
other hand.
Progress could not be made if these aspects could not all be treated
simultaneously. Since no single establishment could take on the huge task by itself,
Ecole Centrale joined EDF and the engineering consulting firm COYNE et
BELLIER in 1984 to develop a computer code called GEFDYN ( Géomécanique
Eléments Finis DYNamique ) to analyze problems involving soil. GEFDYN became
a useful tool in designing and justifying construction. Its aim was to take into
account the complexity of soil under the solid, liquid and gas phases during static or
dynamic loadings, including its stress history before construction (Marc Boulon in
1970 in Grenoble and then Denis Aubry in Paris).
The development of GEFDYN required parallel research work, notably on
constitutive models able to reproduce the complexity of the soil's behavior (Félix
Darve 1974, Jean-Claude Hujeux 1979) and the determination of their parameters,
all of which represented a new field of research. This approach was aimed at
maintaining a close link between the modeling of soil-structure behavior, its
application in civil engineering, and the introduction of significant parameters. This
last demand was the hardest to satisfy, since it involved measuring the properties of
in situ soils, as well as their sampling and laboratory testing conditions. It would not
be unfair to think that the progress achieved in this domain was not as significant as
in the computational field.
To ameliorate this domain, a small number of researchers developed a statistical
tool for analyzing soil data and observations of the behavior of soil masses in the
field (Jean-Louis Favre 1972). A complementary approach consisted of gathering
data from the construction site so that the real soil conditions from the start of
excavation throughout the construction process could be analyzed. This approach
can be likened to the rules defined by Terzaghi in 1927 as being one of the most
important tasks in soil mechanics.
The significant development of the laboratory in the 1970s can be seen in the
evolution of its name. The soil mechanics laboratory at Ecole Centrale was called
“Continuous Media Mechanics Group” in the early 1970s, then “Continuous Media
Mechanics Laboratory”, then “Soil-Structure Mechanics Laboratory” in 1979. The