Geoscience Reference
In-Depth Information
318
Multiscale Geomechanics
9.6. Conclusions
The general principle for establishing a model that represents the behavior of a
natural soil at a given spatial scale is to identify the essential mechanisms that govern
its behavior. We have seen in this chapter that a relatively simple basic macroscopic
model using the finite element method, when it contains the basic mechanisms of
deformation, can be used to study diverse and complex problems. We have also shown
how this model can be generalized; that is adapted to different geologies by taking into
account the physical phenomena involved.
We have also discussed the importance of the model's key parameters and how
they need to be identified. The latter presents two advantages: to ensure, on one hand,
the consistency of model parameters, and to propose, on the other hand, an acceptable
set of parameters in the absence of real data. Most of the work is carried out on
disturbed, that is, remolded soils. The validity of the approach must, of course, be
verified for natural soils. As obtaining undisturbed samples can be a very delicate
task, it is also very important to extend the identification strategies to in situ tests
such as the use of the pressuremeter, the penetrometer, etc. This can only be achieved
after establishing a well-informed database. These various models respond to numerous
loadingconditions,butsomeunexploredwaystoprovidecompletesolutionstopractical
problems may still be found. We could take into account, for example, aging or the
chemical interaction between soils in any given environment. Any development of this
nature will require a rigorous theoretical framework, based on the thermodynamics
of the irreversible processes taking place, complemented as always by careful
experimental observation. The modeling tool can be used for several purposes. One
objective is to obtain predictive models, i.e. models that have been sufficiently mastered,
validated and tested so that they can be used for the design of structures. Another use
for numerical models is as a feasibility study of a real problem that can be made by
studying model parameters.
In this chapter we have presented some static and dynamic geomechanical
problemswithcoupledhydro-mechanicorsingle-phaseprocesses.Althoughthemodels
presented are not yet commonly used in the design of structures, they are sometimes
applied in the exploration phase and, more specifically, in case of the failure, that is,
the malfunctioning of a structure. More frequent applications of these models to study
geo-structures and calculation feedbacks, based on real data, will allow better model
validation and will help establish, in a particular model, validity domains.
Let us not forget that these models do need real data; hence the need to develop
measurement techniques and structure monitoring campaigns. A database containing
the characteristics of these materials will help establish a correlation between the
properties measured in situ and in the laboratory, on the one hand, and the physical
and model parameters, on the other hand. Hence, we can limit the number of trials
Search WWH ::




Custom Search