Actes du colloque

Proceedings of the 18

International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013, volume 6, 2016

French Innovations in Geotechnics: The National Research Projects

English translation of the Special Lecture in French, “Innovations Françaises en Géotechnique: les Projets

Nationaux de Recherche”,

Proc 18

Int Conf Soil Mechanics and Geotechnical Eng, Paris 2013, Volume 1, 163-182.

F. Schlosser; C. Plumelle; R. Frank; A. Puech; H. Gonin; F. Rocher-Lacoste; B. Simon

French Society for Soil Mechanics and Geotechnical Engineering (CFMS)

C. Bernardini

Institute for Applied Research and Experimentation in Civil Engineering (IREX)

ABSTRACT: Full-scale experiments have been considered extremely useful to French civil engineering since the 1960's as a means

of studying structural behavior and new process mechanisms. At the end of the 1970's, the innovative concept behind France's

National experimental research Projects (or NPs) was devised by a French civil engineer named M. Martin. The originality of this

concept lies in the fact that 80% to 85% of funding is generated by project members in the form of subscriptions and especially in-

kind contributions (allocating research time and experimental sites, conducting tests, providing equipment, etc.), with the assigned

Ministry then financing just 15% or 20% of the total budget. The first NP, labeled "

Clouterre

" (1986-90), focused on soil nailing for

retaining walls and was followed by 30 more civil engineering projects, 7 of which involved geotechnical engineering. The IREX

Institute (Institute for Applied Research and Experimentation in Civil Engineering), created in 1989, supervised and managed such

projects. This paper presents the initial steps and the procedure for the NPs, and describes 5 of them in the field of geotechnical

engineering..

KEYWORDS: research, project, innovation, instrumentation, physical and numerical modeling, full-scale experiments.

INTRODUCTION

Soil behavior is a complex phenomenon, and no theory is

available to accurately calculate the stresses and strains of a soil

subjected to any kind of loading. As such, the skeleton of a soil

is neither elastic nor even elastoplastic. Moreover, the water-

skeleton coupling is typically difficult to assess. Despite the

tremendous developments in computing power, it is still

impossible to obtain a set of relations between stresses and

strains both capable of correctly representing the behavior of a

soil composition and usable in practice. All theories are merely

approximations.

The experimental approach to soil behavior thus remains a

key element, especially for verifying a theory's validity.

Mandel's similarity laws (1961) had already revealed the

limitations of reduced-scale sand models, under static loading,

subsequent to the scale effect, which has gradually led to

developing centrifuges in the field of geotechnical engineering.

Furthermore, the widespread development over the past several

decades of measurement instrumentation has not only enabled

studying certain aspects of the behavior of geotechnical facilities

in operation, but has also produced full-scale experimental

structures, which have greatly advanced the state of knowledge.

In France, J. Kérisel (1962) produced the first type of full-

scale experimental structure dedicated to pile behavior. After

conducting, on the Maracaibo Bridge in Venezuela, the first pile

loading test by means of separately measuring both the tip load

and total load at the pile cap, J. Kérisel built a large-sized testing

station on the sandy St Rémy-lès-Chevreuse site, where piles

were being driven into a vast and deep concrete tank filled with

compacted sand. He proceeded by separately measuring the tip

load during pile-driving and found that it varied linearly at first

until reaching a depth of roughly three times the pile diameter,

then remained constant beyond that point. This result, now

widely renowned, has contributed extensively to changing the

pile cap strength calculation with respect to previously applied

theories.

In France, another full-scale experimental structure was built

at the same St Rémy-lès-Chevreuse site, this time by Tcheng

(1975), on the CEBTP Institute's testing station to study large

sand masses that had gradually undergone a state of thrust or

bearing. The station's primary element was a very stiff metal

screen, 5 m wide by 3 m high, containing in its central part 6

embedded measurement cells outputting both the vertical and

horizontal stress components. This screen had been suspended

by 8 hydraulic jacks; then, assisted by a servo control system, it

could be rotated around an axis lying close to the base and

translated horizontally. Two sands were tested: Fontainebleau

sand, characterized by a homogeneous particle size distribution;

and Loire sand, with a much broader distribution. These results

were instructive, notably as regards deviations between theory

and reality, yet they also exposed the difficulties tied to such an

experimental campaign, given an initial state (K

) that depends

on the level of compaction and varies considerably from top to

bottom of the screen.

As of the mid-1960's, the LCPC (French Central Laboratory

for Bridges and Highways) has undertaken, in collaboration with

the LRPC (its regional laboratories), research on soft soil

embankments (1973), ground slope stability (1976), deep

foundations and new retaining structures. In each case, one or

more full-scale experimental structures had been built

specifically for this research effort. Regarding slope stability, a

naturally unstable hill slope was dedicated to this endeavor and

heavily instrumented; monitoring could then take place over

several years.

Research on France's new Reinforced Earth retaining

technique, invented by Henri Vidal in 1963, helped give rise to a

set of Recommendations and State-of-the-Art Practices (1979),

produced jointly by LCPC and the SETRA Road and Highway

Research Center. An experimental Reinforced Earth wall was

built in 1968 by the Eure Departmental Public Works Office and

then instrumented by the Western Paris LRPC Laboratory. For

the first time, this wall made it possible to demonstrate that the

tensile force in reinforcement strips was not being maximized at

the level of the facing, but instead at a given distance inside the

wall (see Fig. 1).

Volume 6 - Page 75